Repositório da Universidade de Lisboa: Página principalrepositorio.ul.pt/bitstream/10451/36053/1/ulsd... · I will always be grateful to my dearest friends and LaSIGE colleagues

��

��

��

� � ��

�� !��

"��#��!��!��

$��!��%

&��'!��(��!�&��)��'��!�

"0��%

&��!�%

1��'!��'��'��!��2��3��4&��!�� !��4��'��5��3��(��

��%

1��'!��6��4��!��!'��4��!��!��2�!��7!�8'��7!��3��!�!��4

��9

1��'!��'��2��'��(��4&��!�� !��4��'��5��$��3��3�

��(��9

1��'!��2��'��"�+��2��!��4&��4��'��5��3��

��(��9

1��'!��:'��;��$�3��2��4&��';��4��'��5��3��(��9

1��'!��(��!�&��)��'��!�4&��';��4��'��5��3��(��4

��!��9

��'��!��!��(��(!��*+��'��'!��

��!�!��(��'��!��!�!'�*<��%��$=�'��*+��5��$��>?��%��:@��@/...A@,-..

��B�4��'��5��3��(��?��%��@��@--C-/@,-.D

,-./

Acknowledgments

I want to express my profound gratitude to my advisor, Professor Carlos Duarte for his

immense transmission of knowledge and the way he allied high demand with friendship

abundance. Thank you Carlos for being an inspiration on a daily basis, for the times you’ve

always been there for me, for being both compassionate and strict, for motivating me to

be a better researcher, for inspiring me into becoming a professor. Thank you for being a

great advisor, and a dear friend.

I would like to thank Pedro Goncalves, a constant presence during great part of my

PhD and a person who tough me very much about commitment and about being a great

professional, and for always having a smile on his face every single day I’ve worked side

by side with him at LaSIGE.

I thank the support and the contributions of Goncalo Gracas and Fabio Rito which

were very important to my thesis, with whom I loved to work with and only regret not

having had the opportunity to work again.

I want to thank all the institutions that helped me carrying out this thesis work and

without which this PhD would never be possible: Universidade Internacional para a

Terceira Idade (UITI), Universidade Senior de Massama e Monte Abraao (USMMA),

Universidade Senior de Odivelas, and an uncountable number of Senior Home facilities

across the country which kindly took part in several of the studies.

I would very much like to thank professors Luıs Carrico and Tiago Guerreiro for the

valuable contributions and advises given that helped developing this thesis work or helped

me becoming a better researcher.

I thank to various people with whom I’ve crossed my path during this PhD and whom

I’ve always treasured the presence, friendship and advices. Thank you Andre Nogueira,

Christoph Jung, Luıs Almeida, Miguel Garcia, Vinicius Cogo, Andre Rodrigues, Rui

Lopes, Ricardo Pinto, David Costa, Daniel Costa and Hossein RZ. I’d especially would

like to thank Jan Egil Nordvik whose unexpected friendship and advices helped me greatly

- Jan, I will now happily drink that Portvinsfat.

I will always be grateful to my dearest friends and LaSIGE colleagues Diogo Lima,

Luıs Duarte, Joao Craveiro, Nadia Fernandes and Tiago Goncalves who helped me get

here with invaluable good times and never-ending support. It’s the time I spent with you

the one I treasure the most.

i

To the best Benfiquista and cafe owner I know Nuno Rodrigues I thank for the com-

panionship and delightful breakfasts I had the pleasure to enjoy at C5 during great part of

my PhD.

I’m grateful to my friends Luıs Sousa, Diogo Albuquerque, Diogo Pinto, Joao Alves,

Filipe Rodrigues, Ricardo Batista, Catarina Marques, Catarina Venceslau, Carolina Simoes,

Renata Milhano, Rui Pedro Caldeira, Jaime Carvalho, David Silva and Ruben Reis, for

giving me all the friendship and support I ever needed during the last five years and for

always motivating me to keep going.

Thank you Fabio Silva and Clara Patrıcio for giving me the unexpected and effortless

friendship I always looked for but never though it existed. And thank you for the motivation

and strength you gave me during the final two years of my thesis.

To my girlfriend and best-friend Beatriz Szwarc I thank you for giving me your World

and for all the support in the darkest hours of this PhD. I don’t think I would ever finished

this without you in my life.

I’d like to thank my brothers Paulo and Pedro for always being my closest friends and

for teaching me so much about life and commitment.

A final word goes to my parents, to whom I owe everything I am, and who I will always

look up to. The education you gave me is something I would only dream to pass on to my

children. It’s the sacrifices, the advices and the love you gave me that brought me here.

ii

To my parents.

Resumo

A populacao na maioria dos paıses do mundo esta a envelhecer, e esse envelhecimento

traz consigo um aumento da dependencia na terceira idade juntamente com outros proble-

mas. Desses problemas um dos mais preocupantes e o isolamento social, com os idosos

a ficarem mais tempo em suas casas provocando uma diminuicao drastica da quantidade

e qualidade de interacoes sociais com os outros. Essa diminuicao pode por sua vez levar

a situacoes drasticas de degradacao da saude e ao aumento do risco de mortalidade. Os

servicos de redes sociais (SNS), como o Facebook, constituem-se como o veıculo ideal para

aumentar essas mesmas interacoes sociais muito devido ao facto da maioria dos adultos ou

da maior parte dos familiares dos idosos serem tipicamente utilizadores destes servicos,

tornando-os um meio atrativo para este segmento da populacao. Ainda assim, e embora

o uso dos SNS entre a populacao idosa tenha aumentado significativamente nos ultimos

dez anos, actualmente apenas cerca de um terco dessa populacao faz uso destes. Razoes

para esta falta de adesao estao nao so relacionadas com a complexidade de desenho e de

interface destes servicos, com problemas de privacidade e com incapacidades sensoriais

resultantes do envelhecimento mas tambem com o facto do PC, tecnologia tipicamente

usada para lhes aceder, nao se mostrar a forma ideal de intermediacao entre esta populacao

e estes servicos. Desta forma, existe nao so a necessidade de perceber e resolver problemas

de interface relacionados com o SNS mais usado (Facebook) por forma a torna-lo mais

inclusivo ou acessıvel, como tambem existe a necessidade de possibilitar que o acesso a

este servico seja efectuado atraves de tecnologia que incentive os idosos a procurarem de

forma activa a comunicacao com os seus amigos e familiares.

Esta tese defende que o desenvolvimento deste tipo de solucoes deve basear-se em

TV, nao so porque esta e, nos dias de hoje, a tecnologia mais utilizada pela segmento

mais velho da populacao, mas tambem porque a mesma e entendida por estes como

uma tecnologia menos intimidante que todas as outras, sendo facilmente associada a

sentimentos de relaxamento bem como usada para apoiar ou desencadear atividades sociais

em ambientes domesticos. Alem disso, argumentamos tambem que, para lidar com as

capacidades e diferencas que caracterizam os idosos bem como a sua interacao com a

tecnologia, o desenvolvimento de solucoes inclusivas baseadas em TV tambem se deve

apoiar em diversas alternativas de interacao e por isso em interfaces multimodais.

v

Partindo desta base, esta tese focou-se primeiramente na amplo estudo e compreensao

do trabalho mais significativo desenvolvido nos ultimos anos, sobre a adocao de tecnologias

por parte da populacao idos. Focou-se tambem na forma como diversos trabalhos baseados

em aplicacoes para TV ilustram esta tecnologica como um potencial ponto de entrada

vantajoso para o uso de redes sociais, bem como, na forma como trabalhos passados

experimentaram diferentes formas de interacao, mantendo o controlo remoto como a

modalidade principal.

De seguida, o trabalho desenvolveu-se em diversas etapas relacionadas com a identificacao

de recomendacoes de desenho de aplicacoes TV baseadas nas caracterısticas e preferencias

da populacao idosa. Foram usadas varias tecnicas de recolha de requisitos bem como de-

senvolvidos e testados com idosos variados prototipos funcionais. Os resultados mostraram

acima de tudo que aquando do desenvolvimento de aplicacoes deste tipo para idosos

devem ser suportadas varias formas de interacao a fim de adequar a mesma as amplas

caracterısticas desta populacao e por forma a compensar os declınios sensoriais e mo-

tores relacionados com o envelhecimento. Adicionalmente como resultado desta analise

identificamos mais de trinta recomendacoes desenho deste tipo de aplicacoes.

No que diz respeito as tecnologias sociais, esta tese comecou por aprofundar atraves

de uma extensa pesquisa bibliografica o entendimento das diversas razoes que contribuem

para a falta de adocao de SNS por parte da populacao idosa bem como se traduziu na

identificacao dos diversos domınios referentes a interacao desta populacao com este tipo

de servicos. A isto seguiram-se varias etapas de recolha de requisitos atraves de ques-

tionarios online, discussoes em grupos de foco bem como de entrevistas semi-estruturadas.

Nestas foi discutido o papel das redes sociais, a forma como os idosos percecionam e

interagem com estas, e a receptividade desta populacao para o uso da TV como forma

de mediar essas mesmas redes. Como resultado de todas estas fases de trabalho, identifi-

camos recomendacoes de desenho de redes sociais direccionadas para as caracterısticas

da populacao idosa e abrangendo treze domınios diferentes: dos mais importantes rela-

cionados com o papel da famılia, o controlo de mecanismos de privacidade e princıpios

de desenho de interfaces, a importancia da multimodalidade, da capacidade de agrupar

contactos e funcionalidades, fotos, reciprocidade de interacoes e tangibilidade das funcoes.

Seguindo as recomendacoes relativas ao desenho de aplicacoes TV e de SNS direc-

cionadas para a populacao idosa e tendo como principal objectivo aumentar o uso de SNS

por parte da populacao idosa e por consequencia reduzir o isolamento social dos mesmos,

desenhamos e implementamos um prototipo multimodal accessıvel baseado no Facebook

e em TV, o “You, Me & Tv”. Este prototipo foi desenvolvido seguindo uma abordagem

de desenho centrado no utilizador bem como de desenho participativo. Neste processo,

baseamo-nos na simplificacao da interface do utilizador mantendo apenas as funcional-

idades mais importantes para a populacao idosa, alterando parcialmente a localizacao

e/ou o agrupamento destas funcionalidades, e recorrendo a mecanismos multimodais e

vi

adaptativos. Como resultado das varias fases de requisitos, desenvolvemos tambem duas

funcionalidades nao existentes no Facebook. Uma delas suporta a hipotese de converter

fotos fısicas em formatos digitais utilizando uma camera colocada por cima da TV e

oferecendo a hipotese de partilhar a foto resultante atraves do prototipo. A outra suporta a

possibilidade de partilhar o que os utilzadores estao a ver na TV. O prototipo foi usado

e avaliado durante um perıodo de 3 a 6 semanas por tres idosos com os resultados a

indicarem um claro potencial deste sistema para aumentar as interaccoes sociais online e

offline entre idosos e (principalmente) os seus parentes, bem como o importante papel da

TV na mediacao deste acesso sobretudo para idosos que ainda nao eram utilizadores de

Facebook, a importancia do controlo remoto e a forma como este deve ser mantido como o

principal meio de interacao, e como tanto a fala como os gestos podem ajudar em tarefas

de producao de conteudo (escrita ditada) e de visualizacao (navegar entre fotos). Atraves

da avaliacao manifestamente positiva do prototipo em termos de usabilidade, verifica-se

como relevantes muitas das recomendacoes identificadas durante os passos anteriores desta

tese.

Em tracos gerais, o trabalho desenvolvido nesta tese responde a tres diferentes questoes

de pesquisa: 1) quais os factores que limitam a adopcao de redes sociais por parte da

populacao idosa; 2) como estes factores devem ser considerados no desenho de servicos

de redes sociais que consideram as caracteristicas dos idosos; e 3) quais as caracterısticas

dos sistemas multimodais baseados em TV que podem contribuir para aumentar a adopcao

de redes sociais por idosos. Adicionalmente, esta tese tambem responde a duas diferentes

hipoteses, validando que atraves do uso de servicos de redes sociais focados em TV e em

capacidades multimodais, a populacao idosa pode aumentar o numero de interacoes sociais

com membros da famılia ou amigos proximos, mas nao validando que o mesmo sistema

seja capaz de ajudar no desenvolvimento de novas relacoes de amizade.

Palavras-chave: idosos, multimodalidade, redes sociais, televisao, Facebook,

desenho centrado no utilizador

vii

Abstract

World population is ageing with social isolation increasing as a result. While Social

Network Services (SNS) like Facebook constitute the ideal vehicle for increasing social

interactions, only about one-third of this population makes use of them. Reasons for this

lack of uptake are related to design complexity, privacy issues and age-related impairments,

with traditional technology like the PC also proving to be insufficient for mediating older

adults access to these services. This thesis argues for the development of SNS solutions

based on Television (TV) and multimodal mechanisms, as TV is not only older adults’

most frequently used technology, but also perceived as less intimidating and associated

with social interchanges, and as resorting to several modalities can help accommodating

differences and to overcome disabilities typically present in this population. With this

in mind, we drew recommendations for the design of TV-based applications targeting

older adults by performing inquiring techniques and testing distinct TV-based multimodal

prototypes. We followed this with a full literature survey on older adults’ use of social

applications as well as several other inquiries regarding SNS and receptiveness of TV

for social tasks. As a result, we formulated recommendations targeting the appropriate

design of an inclusive SNS which span thirteen distinct domains. Recommendations focus

mainly but not only on family, privacy, multimodality and UI design. Building on these

and focusing on user-centered design (UCD) and participatory design we implemented a

multimodal SNS and TV based prototype (You, Me & TV) and evaluated it during a period

of 3 to 6 weeks with 3 participants. Results clearly indicate a potential for the system to

increase online and offline interactions for older adults, the relevancy of TV in mediating

this access for non-adopters of Facebook, the importance of remote controller to be kept as

the primary way of interaction, how speech and gestures can help in content production

and visualization purposes and how photos also play a relevant role.

Keywords: older adults, multimodality, social network services, television,

Facebook, user-centered design

ix

Contents

List of Figures xviii

List of Tables xxi

1 Introduction 1

1.1 Problem and Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Research Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.3 Relation to GUIDE Research Project . . . . . . . . . . . . . . . . . . . . 5

1.4 Research Methodology Employed . . . . . . . . . . . . . . . . . . . . . 6

1.5 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.5.1 List of Publications . . . . . . . . . . . . . . . . . . . . . . . . . 9

1.6 Dissertation Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2 Related Work 15

2.1 Older Adults Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.1.1 Sensory Characteristics . . . . . . . . . . . . . . . . . . . . . . . 16

2.1.2 Cognitive Characteristics . . . . . . . . . . . . . . . . . . . . . . 18

2.1.3 Motor Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 21

2.2 TV and Older Adults . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.2.1 TV-based Interaction . . . . . . . . . . . . . . . . . . . . . . . . 25

2.2.2 TV as a Social Solution . . . . . . . . . . . . . . . . . . . . . . 32

2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

3 Design of TV Applications targeting Older Adults 37

3.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

3.2 Stakeholders Research Stage . . . . . . . . . . . . . . . . . . . . . . . . 39

3.2.1 Inquiring Older Adults . . . . . . . . . . . . . . . . . . . . . . . 39

3.2.2 Inquiring Developers . . . . . . . . . . . . . . . . . . . . . . . . 40

3.3 Multimodal Interaction User Study . . . . . . . . . . . . . . . . . . . . . 42

3.3.1 User Test Application . . . . . . . . . . . . . . . . . . . . . . . 42

3.3.2 User Study Procedure . . . . . . . . . . . . . . . . . . . . . . . 43

3.3.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

xi

3.4 Adaptive Interaction User Study . . . . . . . . . . . . . . . . . . . . . . 49

3.4.1 Prototyping User Initialization Application . . . . . . . . . . . . 49

3.4.2 Prototyping Electronic Program Guide Application . . . . . . . . 51

3.4.3 User Study Procedure . . . . . . . . . . . . . . . . . . . . . . . 53

3.4.4 User Study Results . . . . . . . . . . . . . . . . . . . . . . . . . 55

3.5 Recommendations for the Design of TV Applications targeting Older Adults 58

3.5.1 Interaction Modalities . . . . . . . . . . . . . . . . . . . . . . . 59

3.5.2 Multimodality . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

3.5.3 Adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

3.5.4 TV-based UI . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

4 Drawing Recommendations for the Design of a Social Network Service tar-

geting Older Adults 67

4.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

4.2 Gathering Literature Requirements . . . . . . . . . . . . . . . . . . . . . 68

4.2.1 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

4.2.2 Social Applications and Older Adults . . . . . . . . . . . . . . . 71

4.2.3 SNS use by Older Adults . . . . . . . . . . . . . . . . . . . . . . 76

4.2.4 Social Network Services Developed for Older Adults . . . . . . . 81

4.2.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

4.3 Gathering Older Adults Requirements . . . . . . . . . . . . . . . . . . . 93

4.3.1 Inquiring Older Adults About SNS . . . . . . . . . . . . . . . . 93

4.3.2 Generating Discussion on Older Adults’ Facebook Use . . . . . . 103

4.3.3 Generating Discussion on Older Adults TV-Related Use . . . . . 108

4.4 Recommendations for the Design of a SNS targeting Older Adults . . . . 114

4.4.1 Family . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114

4.4.2 User Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

4.4.3 Privacy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

4.4.4 Photos and Media Relevance . . . . . . . . . . . . . . . . . . . . 119

4.4.5 Multimodality . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

4.4.6 Direct Communication . . . . . . . . . . . . . . . . . . . . . . . 122

4.4.7 Knowing New People . . . . . . . . . . . . . . . . . . . . . . . 123

4.4.8 Personalization and Adaptation . . . . . . . . . . . . . . . . . . 123

4.4.9 Grouping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

4.4.10 Tangible Value . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

4.4.11 Offline Role . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

4.4.12 Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

4.4.13 Reciprocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

xii

5 Designing and Evaluating a TV-based SNS 129

5.1 Participatory Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

5.1.1 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

5.1.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

5.2 You, Me & TV prototype . . . . . . . . . . . . . . . . . . . . . . . . . . 132

5.2.1 Design Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . 132

5.2.2 Implementation Decisions . . . . . . . . . . . . . . . . . . . . . 135

5.2.3 Main Functionalities . . . . . . . . . . . . . . . . . . . . . . . . 136

5.2.4 Multimodal Architecture . . . . . . . . . . . . . . . . . . . . . . 139

5.2.5 Adaptation and Personalization Interface . . . . . . . . . . . . . 140

5.3 Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

5.3.1 Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

5.3.2 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

5.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

5.4.1 Older adults acceptance . . . . . . . . . . . . . . . . . . . . . . 142

5.4.2 Multimodality importance . . . . . . . . . . . . . . . . . . . . . 145

5.4.3 Benefits of personalisation and adaptation . . . . . . . . . . . . . 146

5.4.4 Usability and User Experience . . . . . . . . . . . . . . . . . . . 147

5.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

5.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152

6 Conclusions and Future Work 155

6.1 Dissertation Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

6.2 RQ1: What are the factors limiting SNS adoption by older adults? . . . . 156

6.3 RQ1.1: How these factors should be considered when designing SNS

targeting older adults? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

6.3.1 Age-related impairments . . . . . . . . . . . . . . . . . . . . . . 158

6.3.2 UI complexity . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

6.3.3 UI functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162

6.3.4 Older Adults Self-Belief . . . . . . . . . . . . . . . . . . . . . . 164

6.3.5 Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165

6.4 RQ2: What characteristics multimodal TV-based platforms have that could

contribute to older adults adoption of SNS? . . . . . . . . . . . . . . . . 165

6.5 H1: By using TV-based multimodal SNS, older adults increase interactions

with family members and close friends. . . . . . . . . . . . . . . . . . . 166

6.6 H2: By using SNS, older adults establish new friendships. . . . . . . . . 167

6.7 Limitations and Future Work . . . . . . . . . . . . . . . . . . . . . . . . 168

A Survey Script: Stakeholders Research - Inquiring Older Adults 171

xiii

B Survey Script: Stakeholders Research - Inquiring Developers 181

C User Trial Script: GUIDE Adaptive Interaction User Study 193

D Survey: Older Adults, Technology and Facebook 209

E You, Me &TV: Focus Groups Pre-Questionnaire 217

F Evaluation of You, Me & TV: SUS and UEQ Questionnaires (Portuguese ver-

sion) 219

Bibliography 242

xiv

xvi

List of Figures

1.1 Representation of the methodology of work followed during this thesis. . 7

2.1 Applications built with visual impairments in mind. BigLauncher (left),

Simple Senior Phone (middle) and GrandPhone Launcher (right). . . . . . 17

2.2 Applications built with hearing impairments in mind. BioAid (left), Hear-

ing Aid with Replay (middle) and uSound (right). . . . . . . . . . . . . . 18

2.3 Screens of mElderly, an application that prioritizes most used older adults

functions on the smart-phone. . . . . . . . . . . . . . . . . . . . . . . . . 20

2.4 Screens of the Stepmania application where users: are alerted to start

a session; follow video tutorials for interaction; play games; positive

reinforcement and self-monitoring is used with a flower metaphor . . . . 21

2.5 Screen-shots of Clevermind user interface. . . . . . . . . . . . . . . . . . 22

2.6 Visualization of the growing plant metaphors within the app to motivate

older people through conditioning, goal setting, and self-monitoring. . . . 23

2.7 Flower user interface consisting of a general overview (left), and a day

overview (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.8 An example of a SimplePromenade virtual walk (left), GUI displaying a

virtual instructor remembering a particular movement protocol and user

scores (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.9 Iatrino and Modeo (2006) writing task approaches:: Multi-press with

timeout keyboard and visual feedback or TNT (left), multi-press with

timeout keyboard (middle), virtual keyboard (right). . . . . . . . . . . . . 26

2.10 Remote controls used in the study (left), mean and standard deviations

values of writing speed per method (right). . . . . . . . . . . . . . . . . . 27

2.11 WeSlide: selection of ’G’ . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2.12 Some smart-phone touch-based RCs examples: the reRemote (left); the

redEye remote (middle); and the Verizon FIOS app . . . . . . . . . . . . 28

2.13 RemoteTouch controller (left) and the mapping from the RemoteTouch

controller to the screen (right) . . . . . . . . . . . . . . . . . . . . . . . 29

2.14 Touchless Finger-Count in the context of interactive television . . . . . . 29

2.15 Pointing devices and virtual keyboards used in Barrero et al. (2016) . . . 30

2.16 IntegraTV general interface (left); and speech-based interface (right). . . . 31

xvii

2.17 Promotional image of the Philips Motiva system . . . . . . . . . . . . . . 33

2.18 T-Seniority software screenshot . . . . . . . . . . . . . . . . . . . . . . . 33

2.19 Eldy TV Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

2.20 Screens of iNeighbour TV features . . . . . . . . . . . . . . . . . . . . . 34

3.1 The User Centered Design Approach . . . . . . . . . . . . . . . . . . . . 38

3.2 Screen-shots of the UTA prototype . . . . . . . . . . . . . . . . . . . . . 43

3.3 User-trials photos. Colour selection screen (on the left), and participant

using a Wii remote (on the right). . . . . . . . . . . . . . . . . . . . . . . 45

3.4 Screenshots of the UIA Flash-based (low fidelity) prototype . . . . . . . . 50

3.5 Screen-shots of the UIA high fidelity mockup (left) and UIA high fidelity

prototype (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3.6 Screen-shots of the UIA final prototype . . . . . . . . . . . . . . . . . . 52

3.7 Screen-shots of the EPG mock-ups . . . . . . . . . . . . . . . . . . . . . 53

3.8 Screen-shots of the EPG high fidelity mock-ups . . . . . . . . . . . . . . 54

3.9 Similarities between all adapted versions of the EPG. . . . . . . . . . . . 55

3.10 Modality preferences by task . . . . . . . . . . . . . . . . . . . . . . . . 58

4.1 Graphic showing distribution of answers regarding participants technologi-

cal self-characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

4.2 Reasons for limited use of Facebook for actual users (left) and for non-

users (right). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

4.3 Reasons behind the use of TV . . . . . . . . . . . . . . . . . . . . . . . 102

4.4 Differences between users of Facebook and non-users of this SNS. . . . . 110

4.5 Older adults preferred modalities for interact with TV. . . . . . . . . . . . 113

5.1 The low-fidelity prototypes used in the focus groups. . . . . . . . . . . . 130

5.2 Home page with the news-feed . . . . . . . . . . . . . . . . . . . . . . . 133

5.3 Group and sub-group manager . . . . . . . . . . . . . . . . . . . . . . . 133

5.4 Volume and font size configuration (personalization) . . . . . . . . . . . 133

5.6 The news-feed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137

5.7 Pictures that user took with Kinect camera. . . . . . . . . . . . . . . . . 138

5.8 A group with 3 members of user family. . . . . . . . . . . . . . . . . . . 139

5.9 Frequency of use of U1, U2 and U3 . . . . . . . . . . . . . . . . . . . . 143

5.10 Feature usage by U1, U2 and U3 . . . . . . . . . . . . . . . . . . . . . . 145

5.11 Social interactions of U1, U2 and U3 . . . . . . . . . . . . . . . . . . . . 146

5.12 Evolution of SUS scores for U1, U2 and U3 . . . . . . . . . . . . . . . . 148

5.13 Evolution of UEQ scores for U1, U2 and U3 . . . . . . . . . . . . . . . . 149

xviii

xx

List of Tables

3.1 K-Means Cluster Centers for visual, hearing, cognitive and motor variables 40

3.2 Subjective ratings of the UIA. . . . . . . . . . . . . . . . . . . . . . . . 56

4.1 Journals and Conferences included in the survey . . . . . . . . . . . . . . 70

4.2 Additional relevant Journals and Conferences included in the survey . . . 71

4.3 Related Work and Domains . . . . . . . . . . . . . . . . . . . . . . . . . 85

4.4 Participants’ age range distribution. . . . . . . . . . . . . . . . . . . . . 94

4.5 Willingness to use Facebook. . . . . . . . . . . . . . . . . . . . . . . . . 97

4.6 Facebook use vs Technical Expertise) . . . . . . . . . . . . . . . . . . . 99

4.7 Facebook use vs Number of people the participant lives with . . . . . . . 100

4.8 Participants’ age range distribution. . . . . . . . . . . . . . . . . . . . . 109

5.1 Participants description . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

5.2 Usage of different modalities relative to specific tasks. . . . . . . . . . . 147

xxi

Chapter 1

Introduction

This introduction constitutes a broad overview of all the research reported in this thesis. It

begins by discussing the motivation and main problems for the research. It follows with

the definition of a set of resultant research questions and hypothesis, continues with a

description of the methodology employed and the contributions made and concludes with

a review of the dissertation’s structure.

1.1 Problem and Motivation

Population all over the world is ageing (Eurostat European Commission, 2012). At the

beginning of this millennium, the world population aged 60 years or over numbered 600

million, triple the number reported in 1950. In 2009, these numbers had surpassed 700

million. By 2050, 2 billion older persons are projected to be alive, implying that they

will once again triple over a span of 40 years (United Nations, 2010). Additionally, the

population of older persons is itself ageing. Today, persons aged 80 years or over account

for close to 1 in every seven older persons. And by 2050, this ratio is expected to increase

to nearly one person among every five older person (United Nations, 2010).

As people get old, old-age dependency ratios increase (Commission, ECFIN) along

with several other problems, such as cognitive impairments and motor disabilities as well as

social limitations (Lutz et al., 2008; McLaughlin et al., 2009; Age, 2010). One of the most

concerning issues is loneliness, which typically “kicks in” at retirement age. Older adults

spend a lot of time alone in their homes resulting in low quantity and quality of contact

with others which contributes to increasing social isolation (Cornwell and Waite, 2009).

This resulting decrease of social contact can lead to drastic health degrading situations

(Joinson, 2008; Sundar et al., 2011) and the increase in their mortality risk (Alaoui and

Lewkowicz, 2013).

For older adults, loneliness and social isolation can mean, above everything else, the

absence of contact with their stronger ties, usually family members but sometimes also

close friends. For that reason, these are the primary motivations for the use of social and

1

Chapter 1. Introduction 2

communication technologies since these appeared (Santana et al., 2005; Lindley, 2012;

Raffle et al., 2010; Judge et al., 2011). Thus, (off-line) social networks are the only way

to fight isolation as these serve a significant number of functions, providing social and

emotional support, information resources and ties to other people (Sundar et al., 2011).

Social Network Services (SNS) are the online equivalent to (offline) Social Networks.

In the words of Burke, “SNS are designed to connect people with friends, family and

other strong ties, as well as to efficiently keep in touch with a larger set of acquaintances

and new ties, without having to share the same space” (Burke et al., 2011). A growing

body of evidence suggests there are many social and cognitive benefits for older adults

that make use of these technologies to create content and actively participate in reciprocal

information sharing with family and friends, or even for other purposes like information

seeking and learning (Karahasanovic et al., 2009; Bothorel et al., 2011; Judge et al., 2011;

Cornejo et al., 2013; Lindley, 2012; Baecker et al., 2014; Harley et al., 2014).

Facebook, as the most popular SNS (Statista, 2018), constitutes the ideal vehicle for

this use. With their family and friends already using this service, it becomes an attractive

medium for enticing older adults to engage in an online presence. This is, in fact, the

main reason why SNS developed specifically for older adults have not achieved popularity,

with their younger counterparts not being interested in using these tailored new solutions

(Chen, 2009; Waycott et al., 2013). Recent numbers also show the increased tendency for

Facebook popularity among older adults, with about half of online elderly making use of

it (Pew Research Center, 2015b). Still, and although the use of SNS among the oldest

segment of the population has increased a lot in the last ten years (e.g. from 11% to 35%

in the U.S.), only about one third makes use of them. This is especially relevant if we

consider that about 65% of all adults use these services and that this percentage increases

to 90% in younger adults (Pew Research Center, 2015a).

Reasons for this lack of uptake are complex, and much of the literature overlaps.

Design complexity is one of the central problems with the majority of SNS UI overlooking

age-related impairments (Gibson et al., 2010; Cornejo et al., 2013; Baecker et al., 2014;

Lindley, 2012; Hope et al., 2014; Norval et al., 2014). Additionally, privacy issues (not

understanding who can see what, being afraid of revealing too much, etc.) (Gibson et al.,

2010; Norval et al., 2014; Hope et al., 2014; Waycott et al., 2013; Baecker et al., 2014),

preconceptions about Facebook and SNS use (security problems, misuse, excessive use, or

using SNS as a tool for making new friends) (Harley et al., 2014; Hope et al., 2014; Gibson

et al., 2010; Lehtinen et al., 2009), and loss of more profound ways of communication

(Burke et al., 2011; Hope et al., 2014) are other very relevant issues. Consequently, and

because few recommendations or guidelines exist, there is the need of understanding and

studying how these issues should be tackled to achieve a better inclusive design of this

kind of services.

Additionally, while traditional technology like the PC could compensate some of the


age-related changes in older adult’s physical, social and cognitive resources, they in many

cases are unable to learn or make use of them (McLaughlin et al., 2009; Sayago et al.,

2007). This is also the case for SNS access (Xie et al., 2012; Michailidou et al., 2015).

Hence, the challenges are not only resolving interface problems related with well-known

(and used by everyone) SNS like Facebook but also providing technology that encourages

older people to actively engage in technology-mediated communication and support of an

active life (Lindley et al., 2008).

In this context, we argue that the development of this kind of solutions should be

based on familiar technology like Television (TV) especially if this can be allied with

accessibility concerns. Not only TV is the most used technology by older adults (Bobeth

et al., 2012; Heart and Kalderon, 2013; Mitzner et al., 2010), but is also perceived as a less

intimidating technology (Coleman et al., 2010; Obrist et al., 2008). Additionally, watching

TV is also associated with relaxation, safety and comfort, all factors that contribute to

familiarity and ease of use, particularly in older adults (Tsekleves et al., 2009; Gawlinski,

2003). And research has shown that when performed in the right contexts and “contrary

to the popular belief, watching TV is far from an isolating, antisocial, experience” (Lull,

2014): contributing to “structuring daily life, to satisfying old peoples’ needs for reflection

and contemplation and to remain socially integrated” (Ostlund, 2010); promoting commu-

nication and well-being particularly in geriatric care and with family and friends (Rice

et al., 2007; Geerts et al., 2008); and supporting social activities in the domestic setting

(Bernhaupt et al., 2008; Chorianopoulos, 2008; Geerts et al., 2008). Thus, because of all

these reasons, this platform should be explored to introduce social applications and reduce

the gap between old and new technology (Karahasanovic et al., 2009; Plaza et al., 2011).

Moreover, to deal with older adults disabilities and differences when interacting with

technology, we advocate that the development of inclusive TV-based solutions should also

resort to multimodal interfaces or the use of several alternative modalities of interaction.

This is supported both by primordial findings regarding these interfaces advantages in

terms of user satisfaction (Hauptmann and McAvinney, 1993; Oviatt, 1997; Oviatt et al.,

1997), robustness (Keates and Robinson, 1998; Oviatt, 1999), efficiency (Oviatt, 1997;

Brewster, 1998; MacKenzie and Soukoreff, 2002; Oviatt, 1997), and especially flexibility

(Keates and Robinson, 1998; Brewster and Crease, 1999; Brewster, 1998; Oakley et al.,

2001) and by several examples of how multimodality has made interaction easier for older

adults in the most recent years, and especially in the communication and SNS context

(Lindley et al., 2009; Vetere et al., 2009; Raffle et al., 2010; Judge et al., 2011; Baecker

et al., 2014; Neves et al., 2015; Karahasanovic et al., 2009; Cornejo et al., 2013; Waycott

et al., 2013).

Finally, there is the need to understand how adaptation and personalization mechanisms

can be paired with multimodal interaction to help older adults interact. This is drawn from

several past projects which resorted to adapting dynamic characteristics (Stephanidis et al.,


1998) or changing layout and font size (Gajos et al., 2007) to fit older adults’ characteristics.

Or by other more recent research, which has verified that personalization and adaptation

are highly appreciated by older adults when considering SNS profile presentation (Kara-

hasanovic et al., 2009), and SNS features (Burke et al., 2011; Lindley, 2012). Or even other

studies focused on the differences concerning awareness and communication needs (Judge

et al., 2011), and regarding distinct chunks of information (Grosinger et al., 2012) among

older adults. In common, all these studies evidenced the need for these differences to be

considered when designing a service usable by older adults, and again, the importance of

adaptation mechanisms to tackle that.

1.2 Research Goals

From the description just provided, several research goals can be identified for this thesis,

which are formulated into a set of research questions and hypothesis:

RQ1 What are the factors limiting adoption of SNS by older adults?

As the use and adoption of social network services, and particularly Facebook, by older

adults is still low, there is the need for, first of all, understanding what the reasons behind

this limited use are. Also contributing to this understanding should be the factors behind

applications developed in recent years targeting older adults communication with family

members and friends, as these also constitute the basics of fighting social isolation.

RQ1.1 How these factors should be considered when designing SNS targeting older

adults?

After identifying the relevant factors limiting adoption, there is the need of understanding

and studying how these issues should be tackled to achieve a better inclusive design of this

kind of services. In this, the focus should be on learning which features should be kept and

how they should be designed and which UI elements are more understandable and easy to

interact with. The focus should also be on which type of contacts should be favoured and

which are less important, which interaction modalities can help overcoming limitations,

or which functions are missing that would contribute to greater use. Condensing all these

findings into the design of a more accessible and inclusive service can be the step forward

needed for increasing older adults interaction with the service and their social inclusion as

a consequence.

RQ2 What characteristics multimodal TV-based platforms have that could contribute to

older adults adoption of SNS?


Considering all the reasons appointed for older adults TV adoption, and the potential role

of multimodality, there is the need to understand to what extent these can help in the

adoption of SNS. For this to happen it is necessary to inquire about the receptiveness of

older adults regarding modern TV capabilities and multimodal mechanisms or the use

of modalities that can be alternative to remote controls. Additionally the study of what

modalities are preferred for what TV-related activities can generate further understanding

on both modern TV, older adults interaction, accessibility manners, and this population

use of SNS. In this context, input modalities like speech and gestures (pointing) should be

compared with the traditional remote control interaction (and between each other). Finally,

visual and audio output and the way they are adapted to each user characteristics should

also be studied concerning representing another potential advantage.

H1 By using TV-based multimodal SNS, older adults increase interactions with family

members and close friends.

In this context, it is expected that by using the prototype, older adults increase interaction

with family members. This increase should be considered for both the online and the

offline contexts and could be in the form of likes, comments, and shares for the first, and

in the way of visits, phone calls or any other kind of human interactions, for the latter.

Additionally, while several examples of social applications exist, most of them use PC

and Tablet devices as a vehicle for older adults adoption, and very few solutions existed

regarding the use of TV for older adults social interaction with the very few being focused

only on Remote Control (RC) interaction. Is also of particular interest understanding if

by adding multimodal features to a TV-based Facebook-based, older adults can increase

social interactions, be it online or offline ones.

H2 By using SNS, older adults establish new friendships.

Similar to the previous hypothesis, a Facebook-based prototype built taking into considera-

tion both accessibility concerns and recommendations identified in RQ1.1, is expected to

be able to forge new relations between older adults and other online users, establishing

new friendships. We will consider as a new friendship any people who are added as a

friend on the service which they did not know before or are “friends of friends” of one

of their contacts. Similarly, any interactions “in the flesh” with people they did not know

before will be considered a new friendship, if this is a result of an online interaction.

1.3 Relation to GUIDE Research Project

Part of this thesis’ work was done in the context of a research project, GUIDE (Gentle

User Interfaces for Elderly People). This European FP7 project focused on the provision

of alternative ways for older adults to interact with TV-based applications in a more natural


and accessible way. It made use of distinct UI configurations and multimodal feedback

to adapt the experience to each user. It also provided a toolbox to assist developers in the

implementation of accessible TV applications, by making use of simulation mechanisms

and code annotation to respectively let developers understand the effects of any impairment

on the perception of the rendered UI and support an adaptive mechanism capable of

adjusting or replacing UI components to fit specific user characteristics.

This thesis’s work relates to the GUIDE project in several phases. Work performed

during the development of several test applications as well as the project’s final applications

concluded with various findings which are part of this document. More specifically, the

identification of guidelines for the development of TV-based applications targeting older

adults contributed to informing the design of TV-based applications like the SNS prototype

developed for this thesis work. The analysis of several user trials which culminate in strong

evidence that TV and multimodal features should be used as the main entry point when

considering older adults adoption of technology, also contribute to answering to RQ2. Thus,

GUIDE acted as a catalyst for the beginning of this thesis, providing valuable guidelines

for the development of following applications and supporting the type of technology that

should be used.

1.4 Research Methodology Employed

This thesis work followed a traditional user-centred-design (UCD) methodology (depicted

in fig. 1.1), strongly based on iterative prototyping (Goldman and Narayanaswamy, 1992)

and contextual design (Beyer and Holtzblatt, 1997). The main justification for this is

the fact that UCD allows for a detailed understanding of users’ difficulties as well as

for the identification of cultural and social issues, very relevant for matters relating to

SNS and technology adoption. At the same time, by resorting to iterative prototyping,

solutions developed to test both TV and SNS approaches have its origin in the requirement

gathering process, and evolve both as a result of participatory design phases, and practical

user studies. Finally, contextual design tackles all design steps taking into consideration

requirement gathering stages centred on the target population perspectives. Thus, the

methodology followed offered some agility, necessary for supporting the revisit of the last

three phases as more results were obtained.

This thesis work began with the necessary study of related research performed on the

subjects of TV and SNS, all in the context of older adults. Two of the main tasks were the

requirement analysis of all relevant findings performed in the context of GUIDE, and the

realisation of a full literature survey concerning the use of social applications and SNS by

the target population. While the latter was necessary for a better understanding of all the

dimensions influencing older adults use of these services, the first culminated with a set of

recommendations which informed the design of TV-based applications.


Figure 1.1: Representation of the methodology of work followed during this thesis.

Following this, we adopted several requirement gathering techniques, inquiring older

adults mainly about SNS and traditional and modern TV capabilities and habits by per-

forming online and offline questionnaires, and contextual interviews on senior universities

and health-care institutions. These were conducted for gathering a better understanding of

technology perception, both in the context of the GUIDE project, as well as in early phases

of this thesis and the context of several user studies and prototype design.

The participatory design was also one of the primary tools used for the contextual

design of prototypes. Focus groups were run for discussing the design of both TV and

Facebook-based prototypes. Paper-based prototypes were used in the context of these

focus groups.

Informed by all previous phases, prototyping happened in several stages of this thesis.

In GUIDE with the development of several TV-based multimodal functional prototypes.

And in the final steps of the PhD with the development of SNS-based applications for

TV (as a way of evaluating how older adults interact in an SNS context, and if social

interactions increase).

Starting in the requirement gathering, and extending to the prototyping phase, we

tried to validate with older adults every decision and recommendation implemented in the

prototypes. We also evaluated several use-cases of longitudinal use of the TV-based SNS

prototype developed.

1.5 Contributions

The following details the main contributions made by this thesis:

• Design of Accessible Multimodal TV-based Applications Under the context of

the GUIDE project, a contribution was made regarding the design of accessible

multimodal TV-based applications which are part of this thesis.


– A list of guidelines concerning the design of multimodal TV-based applications

for older adults. Throughout the realisation of several user-trials with TV-based

prototypes and test-applications, it was possible to collect and conclude on

several interaction habits, user interface preferences and features, and main

problems that characterise older adults interaction with TV-based applications

and multimodal features. By testing several options, it was also possible to

turn several solutions into guidelines for the development of more inclusive

and accessible TV-based applications for this segment of the population.

• Design of Social Network Services targeting Older Adults As part of this thesis

work several contributions were made regarding the understanding of older adults

interaction and limitations using SNS, and consequently regarding the design of

TV-based multimodal prototypes targeting the increase of social interaction and

tackle loneliness.

– A full literature review of social applications and SNS and their use by older

adults. This was performed taking special consideration to the main journals

and conferences in the relevant areas of HCI and Social aspects of HCI and

relevant cited publications from these sources from the last ten years. A list of

main domains approached by these publications was identified from which a

list of recommendations targeting the inclusive design of SNS was reasoned

and discussed.

– A list of guidelines concerning the design of Social Network Services tar-

geting older adults characteristics. Taking the previous steps as the basis for

performing online questionnaires and semi-structured interviews, we draw rec-

ommendations and formulated guidelines to be followed when designing new

solutions and new prototypes of SNS or Facebook-based solutions, especially

relevant for the contexts of TV-based applications.

– Design of a TV-based Facebook-based prototype with multimodal features

capable of inciting older adults to increase social interactions. A fully func-

tional prototype based on Facebook was constructed and evaluated with three

older adults over a period of three to six weeks. The prototype was built with

multimodal features like the possibility of interacting with speech and gestures.

Functionalities related to TV sharing were also integrated. The domains previ-

ously identified, as well as the guidelines for developing TV-based applications,

were considered. Results show the possibility of increasing older adults social

interactions with family and close friends, both online and offline, by making

use of these technologies. Results also confirm the importance of multimodality

and focus on family, as well as the relevance of domains and recommendations

identified in the previous steps of this thesis.


1.5.1 List of Publications

All results presented in this thesis have been validated through the publication of conference

and journal articles and a book chapter. A complete list of publications is shown next.

1.5.1.1 Design of Accessible Multimodal TV-based Applications

These are the publications related with the development of TV-based applications with

multimodal features developed mainly within the context of GUIDE project.

• Coelho et al. (2011).Developing accessible TV applications. Jose Coelho, Carlos

Duarte, Pradipta Biswas, Patrick Langdon. In Proceedings of the ASSETS11,

13th international ACM SIGACCESS Conference on Computers and Accessibility,

Dundee, Scotland, pp. 131-138, ACM Press, 2011.

In this paper the GUIDE accessibility approach on how to develop accessible TV applications

is explained. Integrating multimodal interaction, adaptation techniques and the use of

simulators in the design process, it is shown how to adapt User Interfaces to the individual

needs and limitations of older adults. For this, we relied on the identification of the most

relevant impairment configurations among users in practical user-trials and drawn a relation

with user specific characteristics. As the main contribution for this thesis, the paper also

provides guidelines for more accessible and centred TV application development.

• Coelho et al. (2012b).Involving All Stakeholders in the Development of TV Ap-

plications for Elderly. Jose Coelho, Carlos Duarte, Tiago Guerreiro, Pedro Feiteira,

Daniel Costa, David Costa, Bruno Neves, Fernando Alves, Pradipta Biswas, Patrick

Langdon. In International Journal on Advances in Intelligent Systems Volume 5,

Number 3 & 4, pp. 427-440, IARIA journals, 2012.

This paper continues the description of the GUIDE accessibility approach focusing on the

multimodal interaction techniques and on saving efforts from the developers’ point of view.

TV users, developers and manufacturers were confronted and evaluated new interaction

and design paradigms with particular interest for a novel way of acquiring and providing

knowledge from and to the users with an application called User Initialization Application.

• Coelho et al. (2013a).Designing TV Interaction for the Elderly - A Case Study

of the Design for All Approach. Jose Coelho, Tiago Guerreiro, Carlos Duarte. In

A Multimodal End-2-End Approach to Accessible Computing, Human-Computer

Interaction Series 2013, pp. 49-69, Springer, London, UK, 2013.

This chapter builds on the knowledge gained from designing, prototyping and evaluating

different TV applications, supported by multimodal interaction, and aimed at the older

segment of the population, to offer an extensive list of guidelines specific for designing TV

interaction.


1.5.1.2 Design of Social Network Services targeting Older Adults

These are the publications related to the study of the main domains of SNS and older adults

and the design of SNS solutions based on taking the characteristics of the older segment of

the population into consideration.

• Coelho and Duarte (2016).A literature survey on older adults’ use of social net-

work services and social applications. Jose Coelho, Carlos Duarte. In Computers

in Human Behavior, Volume 58, pp. 187-205, Elsevier, 2016

Considering the vast amount of research surrounding the emergent field of SNS, along with

other meaningful social applications, and its use by the older segment of the population, in

this paper we reviewed all the work performed in this area with particular relevance for the

last ten years. A total of thirteen domains were identified related to how these services can

be improved to consider older adults characteristics: from the most important associated

with family role and privacy control, to issues related to the design of user interface, the

importance of multimodal interaction and adaptive solutions to compensate age-related

declines, to several other focusing on the importance of groups, photos, cultural and health

information. Main contributions are given in a set of recommendations which result from

discussions on each domain and which aim at the design of a more inclusive SNS solution.

• Gomes et al. (2014).Designing a Facebook Interface for Senior Users. Goncalo

Gomes, Carlos Duarte, Jose Coelho, Eduardo Matos. In The Scientific World Journal,

Volume 2014, 8 pages, Hindawi Publishing Corporation, 2014

In this paper is reported the work performed through a series of direct observations, interviews

and focus groups for identifying recommendations for the design of SNS targeting older

adults. It is also reported how these recommendations served as the design basis for a

prototype for older adults which supported sharing and viewing of Facebook content. Lastly,

the paper also indicates how this prototype was compared with Facebook’s native mobile

application. Main results showed that while the latter does not meet older adults concerns

like privacy and family focus, the built prototype supported relevant use cases in a usable

and accessible manner.

• Coelho and Duarte (2015).Socially Networked or Isolated? Differentiating Older

Adults and the Role of Tablets and Television. Jose Coelho, Carlos Duarte.

In Proceedings of INTERACT 2015, 15th International Conference on Human-

Computer Interaction, Lecture Notes in Computer Science, Volume 9296, pp. 129-

146, Springer, Bamberg, Germany, 2015

This paper describes a study to investigate the main difficulties and motivations towards

Facebook use and adoption among older adults and inquire about the possibilities of using

Tablet and Television as alternatives to the traditional PC for accessing this kind of services.


Main contributions are findings that show correlations between self-belief in technical

skills, motor limitations and tablet use and the use of Facebook and the way UI complexity

constitutes a barrier. Additionally in the paper are also identified distinct groups and distinct

feelings about the use of TV as a vehicle for social interaction.

• Coelho et al. (2015).Prototyping TV and Tablet Facebook Interfaces for Older

Adults. Jose Coelho, Fabio Rito, Nuno Luz, Carlos Duarte. In Proceedings of

INTERACT 2015, 15th International Conference on Human-Computer Interaction,

Lecture Notes in Computer Science, Volume 9296, pp. 110-128, Springer, Bamberg,

Germany, 2015

In this paper we focus on the potential of TV and Tablet devices to improve the ability of

older adults to interact with SNS. The paper reports findings from a study composed of

semi-structured interviews and focus groups which show good receptiveness from older

adults to perform social tasks on both TV and Tablet-based applications, for interacting using

alternative modalities like speech or back-of-device tapping, and for the use of adaptation

mechanisms. Informed by the study results the main contributions are the two prototypes,

and a collection of recommendations regarding the design of TV and tablet-based interfaces

for this population.

• Coelho et al. (2017)”You, me & TV” - Fighting Social Isolation of Older Adults

with Facebook, TV and Multimodality. Jose Coelho, Fabio Rito, Carlos Duarte.

In International Journal of Human-Computer Studies, Vol 98, pp. 38-50, Elsevier,

2017

Finally, in this paper are presented three in-depth case studies of the use of TV-based

multimodal Facebook prototype developed and reported in the previous publications. The

prototype is presented in more detail as well as its TV shows sharing functionality, its feature

for turning printed photos into digital versions, and multiple interaction modalities - remote

control, voice and gestures-. Main contributions are the results which show that participants

accepted the prototype, rating it highly regarding usability, and showing that (especially for

previously non-Facebook-users) it improves both their online and offline interaction with

relatives.

1.5.1.3 Other Related Publications

These papers express secondary work performed during this thesis which was also accepted

for publication after being peer-reviewed:

• Coelho et al. (2013b). TV Applications for the Elderly: Assessing the Accep-

tance of Adaptation and Multimodality. Jose Coelho, Pradipta Biswas, Tiago

Guerreiro, Gokcen Aslan, Carlos Duarte, Patrick Langdon. In Proceedings of the


ACHI2013, 6th International Conference on Advances in Computer-Human Interac-

tions, pp. 232-242, IARIA, Nice, France, 2013.

• Coelho et al. (2012a). Building Bridges Between Elderly and TV Application

Developers. Jose Coelho, Carlos Duarte, Pedro Feiteira, David Costa, Daniel Costa.

In Proceedings of the ACHI2012, 5th International Conference on Advances in

Computer-Human Interactions, pp. 53-59, IARIA, Valencia, Spain, 2012.

• Coelho and Duarte (2011b). The contribution of multimodal adaptation tech-

niques to the GUIDE interface. Jose Coelho, Carlos Duarte. In Proceedings of the

International Conference on Universal Access in Human-Computer Interaction, pp.

337-346, Springer, 2011.

• Coelho and Duarte (2011a). Building Supportive Multimodal User Interfaces.

Jose Coelho, Carlos Duarte. In SUI2011, Workshop of Supportive User Interfaces,

ACM SIGCHI Symposium on Engineering Interactive Computing Systems (EICS

2011), Pisa, Italy, 2011.

The author has also co-authored other peer-reviewed publications, describing minor or

parallel aspects of the hereby described work:

• Duarte et al. (2011a). Eliciting interaction requirements for adaptive multi-

modal TV based applications. Carlos Duarte, Jose Coelho. In Proceedings of the

International Conference on Universal Access in Human-Computer Interaction, pp.

337-346, Springer, 2011.

• Biswas et al. (2011). Multimodal adaptation through simulation for digital

TV interface. Pradipta Biswas, Patrick Langdon, Carlos Duarte, Jose Coelho.

In Proceedings of Euro ITV 2011, 9th International Interactive Conference on

Interactive Television, pp 231-234, ACM Press, Valencia, Spain, 2011.

• Duarte et al. (2011b). GUIDE: Creating Accessible TV Applications. Carlos

Duarte, Patrick Langdon, Christoph Jung, Jose Coelho, Pradipta Biswas, Pascal

Hamisu. In Proceedings of AAATE 2011, 11th European Conference for the Ad-

vancement of Assistive Technology, Everyday Technology for Independence and

Care, Volume 29, pp 905-912, IOS Press, Maastricht, The Netherlands, 2011.

1.6 Dissertation Structure

The introduction has provided an overview, motivation, challenges and goals for the

research reported in this thesis. The contributions were also briefly described. The

remainder of this thesis goes into further detail and is structured as follows:


Chapter 2 analyses previous related work in the fields of older adults characterisation

and relation with TV-based technology. Special relevance is also given to the integration of

multimodal mechanisms and how these have proven to be valuable in the past concerning

older adults interaction.

Chapter 3 describes all the work performed regarding the identification of recommen-

dations to follow in the design of TV-based solutions targeting older adults. Taking related

work findings as the basis, it describes several requirement gathering, participatory design,

prototype development phases and user studies performed both in and out of GUIDE

context.

Chapter 4 describes all the work performed regarding the identification of recom-

mendations to follow in the design of an SNS-based solution targeting older adults. An

extensive survey reviewing SNS and social applications use by older adults, a description of

a set of inquiries performed with this population, and the design and use of Facebook-based

prototypes to pre-validate some of the dimensions identified, are described.

Chapter 5 focuses on both the design and evaluation of a TV-based Facebook-based

prototype. It describes how all design recommendations regarding both TV and SNS

were considered, how longitudinal use-cases were carried out with older adults using the

prototype and draws results regarding both the number, type and nature of interactions

resulting from its use.

Finally, Chapter 6 concludes on the relevant findings of all this thesis work, answering

each research question and hypothesis formulated, and provides directions for future work.


Chapter 2

Related Work

This chapter constitutes a broad overview of the most meaningful work developed in the

past years concerning the most relevant areas of this thesis work. It begins by focusing on

older adults age-related limitations, giving a general understanding of what can influence

older adults adoption of technology, and how past research has tackled these issues. It

continues with a thorough explanation of how the use of TV and TV-based applications

has evolved into being an advantageous entrance point for more complex technology and

services by this segment of the population, especially when tied together with multimodal

features. The chapter ends with an overview of how TV-based technologies have been

used as social solutions.

2.1 Older Adults Limitations

It is common to define “Aging” as the progressive loss of function with advancing age and

increasing rates of health problems (Sundar et al., 2011). Additionally, Parra (2013) also

describes it as the process we all undergo from the moment of our birth, being characterised

by a continuous growth of knowledge but a decline of our capabilities after our 50s or 60s.

Therefore, and independently of health, there are three main areas where ageing

manifestation affects older adults. Sensory limitations, or the ones which hinder the

capacity to capture and interpret information, like visual and hearing. Cognitive limitations,

or the ones which block the ability to process, reason on, and produce information. And

motor limitations, or the ones which hinder the capacity to perform physical activities. As

a consequence of all these limitations, people also tend to aggravate their social context,

which typically leads to an increase of dependence from others and a decrease in quality

of life (ECFIN), 2008). In this section, we detail each type of older adults limitation

and explain how these influence their capacity for daily living and interacting with any

technological devices.

15

Chapter 2. Related Work 16

2.1.1 Sensory Characteristics

Sensory changes with old age may affect the processing of information, as it impacts the

perceptual abilities, especially sight and hearing, which are essential to function in certain

aspects of everyday life (Fisk et al., 2009).

2.1.1.1 Visual Impairments

Visual impairment commonly afflicts the older population and inconveniences their daily

lifestyle. Most people over age 50, will eventually experience some visual impairment,

which can vary in degree from person to person (Pak and McLaughlin, 2010), and will

need to wear glasses to correct their visual defects.

The two main impairments related to vision, are visual acuity and colour perception. Vi-

sual acuity declines rapidly after age 60, translating into more difficulties on the perception

of details in visual images (Pak and McLaughlin, 2010). Decreases in colour perception

make distinguishing between colours with similar hues and low contrast harder (Farage

et al., 2012). Additionally, with age, there is also a decline in the eye’s accommodation

capability which is called presbyopia, which makes focusing harder causing difficulties

in near objects viewing (Fisk et al., 2009). All these impairments (and others like loss of

peripheral vision and disability glare) are not only explained by age-related modifications

but also by diseases which are prevalent in old age like Cataracts (Saxon et al., 2009),

Glaucoma (Shari, 2008) and Diabetic Retinopathy (Jackson and Owsley, 2003).

These visual impairments have a substantial impact on the generality of daily activities

performed by older adults (reading, driving, etc.) where technological related activities

are included. As most of the information is still conveyed visually (Pak and McLaughlin,

2010), and although the majority of design guidelines cannot be extrapolated from their

interaction context, there are some which should be taken into account: both Shari (2008)

and Farage et al. (2012) evidenced that small texts, and tiny icons should be avoided,

and highly saturated, or contrasting colours are recommended; Becker (2004) indicated

that elaborate font styles or pattern backgrounds should also be avoided; Morrell (2001)

and Phiriyapokanon (2011) advised for limiting the amount of information on screen

to improve visual search efficiency and attention (although these are also related with

cognition). Additionally, multimodal interfaces could also play a role in compensating

visual problems by employing feedback in other modalities (audio, touch, haptic).

Past research also developed applications which took into consideration these rec-

ommendations. The best examples are applications for smart-phones and tablets which

focused on making fonts, icons and colours more suitable for older adults by increasing

their size and maximising contrasts or compensating visual impairments with audio feed-

back and input. BigLauncher, Simple Senior Phone, GrandPhone Launcher ((fig. 2.1)

shows screen-shots of these), Phonotto, Large Launcher, and Necta Launcher (Zapata et al.,

2015; de Souza and Silva, 2016) are examples of these applications.


Figure 2.1: Applications built with visual impairments in mind. BigLauncher (left), Simple

Senior Phone (middle) and GrandPhone Launcher (right).

Most of IT applications with accessibility concerns are also suitable to compensate

visual impairments that come with age, mainly by employing multimodal mechanisms or

dictation. Examples of these are given with the following: No-Look-Notes incorporated

multi-touch input (Bonner et al., 2010), WebAnywhere facilitated navigation using voice

and audio output for outputting website content (Bigham et al., 2008), and Penpal was an

electronic pen which made possible to scan and read out loud any text document (Visu

et al., 2011).

Finally, this type of impairments in old age can have not only physical consequences

like injuries from falling or traffic accidents but also related emotional effects that under-

mine the psychological well-being. Therefore, age-related visual impairments can by itself

lead to social isolation.

2.1.1.2 Hearing

Hearing or auditory age-related changes are the second main limitation affecting older

adults and begin gradually from 40 years old accelerating after age 55 (Davis, 1991). At

the beginning of the century, about 78% of people over 65 had a disability of this kind,

with 7% having problems hearing any sound, 19% loud sounds and 68% speech (Dalton

et al., 2003).

Hearing impairments in old age is mainly due to changes which cause disturbances

in the inner ear metabolism (Fozard and Gordon-Salant, 2001). These changes can be

grouped into three different categories (Saxon et al., 2009): a conductive hearing loss

which typically results in a reduction of sound level perception; a sensorineural hearing loss

which usually occurs from disorders in the inner ear or damage to the nerve pathways to

the brain; and mixed hearing loss which involves the two previous losses. As a result of all

these, older adults have not only problems in understanding speech-like sounds (especially


Figure 2.2: Applications built with hearing impairments in mind. BioAid (left), Hearing

Aid with Replay (middle) and uSound (right).

in noisy or echoing environments) but are also usually unaware of those problems. And

even though in its generality this type of loss can be offset with the use of hearing aids,

these require an adaptation process that for some older people presents difficulties.

Ideally, to compensate for the hearing loss in older adults, auditory signals should be

combined with other sensory channels, such as flashing lights or vibrations, to convey the

same information to the hearing-compromised older adult (Farage et al., 2012). Taking

this as a prominent example, multimodal interfaces could play an important role when

compensating hearing declines by providing additional feedback (visual, touch, haptic).

Several applications developed to ease this population interaction with smart-phones

and tablets also focused on the hearing-loss dimension, and while some were already

mentioned, others like WisePhone concentrate on keeping the volume to maximum or

adapted to each user necessities (de Souza and Silva, 2016), and BioAid iPhone app, the

Hearing Aid with Replay, and uSound (Tiwari et al., 2016) processed sound from the

microphone and delivered it over headphones in real time (fig. 2.2 shows screen-shots for

the last three), thereby compensating hearing impairments of each user.

Finally, considering we mainly communicate with others using verbal communication,

hearing loss naturally limits the ability to understand and participate in social activities

that require communication. This results, among other psychological problems, in loss of

independence and social isolation (Cavanaugh and Blanchard-Fields, 1990; Saxon et al.,

2009), reducing the quality of life (Dalton et al., 2003).

2.1.2 Cognitive Characteristics

As they start ageing, a significant number of older people starts having mild symptoms of

cognitive decline that occur as part of the normal ageing process (Craik and Jennings, 1992;

Douglas et al., 2008) and non-communicable diseases (like Dementia and Alzheimer’s


disease) (Cheng et al., 2014; Cholerton et al., 2016). According to Lane and Snowdon

(1989), 35% of subjects with 65 years and over have an age-related cognitive impairment,

and this number increases to 85% on individuals aged over 80 years (Larrabee and Crook,

1994).

As a way of defining all cognitive characteristics influenced by ageing, we can say

that a product adequate for younger adults might be unusable by their older counterparts

because of increased difficulties related with memory, attention and spatial cognition.

In terms of memory, there are several distinct types to be considered (Pak and McLaugh-

lin, 2010): working (or the capability of retaining information during a task), semantic (or

the capacity for storing factual information), prospective (or the capability to remember to

perform functions in the future) and procedural (or the ability to perform activities using

previously acquired skills). From these, only long-term memory (composed by semantic

and prospective) remains relatively unchanged with ageing (Fisk et al., 2009; Cavanaugh

and Blanchard-Fields, 2014), while working memory is the most affected with older adults

holding many items for shorter periods of time (Pak and McLaughlin, 2010; Farage et al.,

2012). Thus, declining memory capacity negatively impacts daily activities such as speech

and language comprehension (Cavanaugh and Blanchard-Fields, 2014), reasoning and

problem-solving (Fisk et al., 2009), as well as the ability to learn new procedural tasks or

forge new memories.

Attention can be described as older adults capacity to process information, and can

be divided into three aspects (Fisk et al., 2009): attentional capacity (or the capability of

splitting attention resources across various tasks), selectivity (or the capacity of focusing on

a single item or task while ignoring interferences) and vigilance (or the ability to maintain

concentration on a particular task over a more extended period of time). All three decline

significantly with age (Pak and McLaughlin, 2010). However, selectivity is notably more

difficult to retain as older adults seem to have more trouble concentrating on one factor

while ignoring other distracting stimuli. Additionally, vigilance can also be a big issue

as older adults are more susceptible to being distracted by irrelevant surrounding stimuli

and being slower in performing tasks that take several seconds or minutes to complete

(Cavanaugh and Blanchard-Fields, 2014; Fisk et al., 2009).

Spatial cognition is the ability to retain and reconstruct location-based representations

of the world, and are usually related to the ability to navigate information hierarchies (Fisk

et al., 2009). Research has shown that older adults not only have more difficulties than their

younger counterparts on tasks involving the constructing of mental models like navigating

mobile phone menus (Gilbert and Karahalios, 2009) but also prefer shallow navigation

systems for online navigation (Zaphiris et al., 2002).

Considering all these three dimensions, cognition is of extreme relevance when consid-

ering interface design (Brewster, 2002; Fisk et al., 2009). In the first place, to compensate

for declines in working memory, displaying all necessary information for each specific


Figure 2.3: Screens of mElderly, an application that prioritizes most used older adults

functions on the smart-phone.

user to perform a task (instead of expecting the user to remember it) would be ideal (Pak

and McLaughlin, 2010). In the second place, and considering procedural memory, it is

advisable to make use of older adults crystallised knowledge by building upon previously

learned mental-models and procedures, and avoiding technical expressions (Fisk et al.,

2009). In third place, and when focusing on attention issues, both Fisk et al. (2009) and

Pak and McLaughlin (2010) suggested that not only avoiding clutter and unnecessary

items that may distract the user but also draw him/her to essential elements through other

sensory channels could help. This last recommendation highlights the potentially decisive

role of multimodality also for related cognitive issues. Additionally, drawing attention

to important or most frequently performed actions can be a form of avoiding all three

dimensions issues - attentional capacity, selectivity and vigilance - at the same time (Pak

and McLaughlin, 2010). Finally, clear structure of tasks and consistency of information

are additional vital requirements to reduce complexity and cognitive load (Phiriyapokanon,

2011): namely by employing essential function unity (i.e. one key one function), page

function unity (i.e. one page one function), using wizards for complex tasks, coloring and

labeling information, and providing consistent navigation bars or menus.

Research has focused on these recommendations. Examples of that are GrandPhone

Launcher, WisePhone, Help Launcher Deluxe and mElderly (shown in fig. 2.3), appli-

cations which focused on prioritising most essential contacts and actions, making more

accessible for older adults to use a smart-phone (de Souza and Silva, 2016).

Moreover, research on serious games is also relevant as these serve the purpose of

cognitive stimulation. Good examples are games like Cogniplus (GmbH, 2007) (providing

a wide range of animated cognitive games for improving health prevention), Posit Science

(Ball et al., 2010) (offering a wide range of brain-train programs), Byun and Park (2011)

monitoring cognitive games (keeping track of older adults cognitive metrics and informing


Figure 2.4: Screens of the Stepmania application where users: are alerted to start a

session; follow video tutorials for interaction; play games; positive reinforcement and

self-monitoring is used with a flower metaphor

doctor about them), and Stepmania (Far et al., 2012) (designed specifically for improving

older adults attention by making them concentrate on multiple moving objects at the

same time) (fig. 2.4). Other games developed for tabletop devices also offered a range

of cognitive training games, like Hermes Maze (Buiza et al., 2009) a training memory

game which was used for preventing older adults from missing medical appointments, and

an adaptation of Tangram (Zapirain et al., 2010) used to support psycho-motor activity

therapies.

Another way in which technology can compensate cognitive declines is by making

use of alerting applications (McCall et al., 2013). The Ubimeds (Silva et al., 2009) and

the AIS set (Garcıa-Vazquez et al., 2011) are good examples of reminder applications

and technological artefacts which made use of physical objects (dispensers), auditive and

visual alerts, and smart-phone capabilities to facilitate the retrieval of pills.

Finally, decreased cognitive functions have proven to sustain engagement in social

and productive activities (de Bruin, 2012; Scheibe and Carstensen, 2010). As a result,

some researchers focused on developing social applications which are directly allied

with the training of cognitive functions in old age. Examples include the application

developed by Meza-Kubo et al. (2009) which concentrate on involving the family network

for assisting cognitive training, as well as a similar application called Clevermind (http:

//www.myclevermind.com) which made use of an UI specifically designed for older

adults with decrease cognitive ability and used contact with loved ones through Facebook

to positively affect older adults (fig. 2.5), and Derboven et al. (2012) shopping game where

older adults had to remember the shopping list (improving memory skills) and could be

helped remotely by family members when buying the products in a virtual reality shop.

2.1.3 Motor Characteristics

Motor control refers to the response time and accuracy of movement (Pak and McLaughlin,

2010). The ageing process involves modifications to the muscular-skeletal system (Kurni-


Figure 2.5: Screen-shots of Clevermind user interface.

awan, 2008) that contribute to a reduction of capabilities such as strength and endurance

(Cavanaugh and Blanchard-Fields, 2014), also reducing touch sensitivity or tactile acuity

(Wickremaratchi and Llewelyn, 2006). As a result of these modifications not only older

adults have issues performing activities of daily living (walking, climbing stairs, stooping,

kneeling) but the probability of falls increases and several repercussions occur that affect

the task performance of manipulating technological devices (such as double-clicking with

a mouse, striking a key, or even selecting a touch button) (Hawthorn, 2000; Laursen

et al., 2001). The amount in which these modifications occur is such that reports showed:

approximately 17% of the men and 23% of the woman aged over 65 have some limiting

physical limitations (Eurostat European Commission, 2012); around 20% of those aged

above 76 have a severe mobility limitation (Steel et al., 2004); older adults take 30% to

70% longer than their younger counterparts to perform specific motor tasks (Ketcham

et al., 2002) being also less accurate in performing those movements (Pak and McLaughlin,

2010).

Tremor is the most common form of movement disorder or motor impairments in the

older segment of the population. It can be observed over several age-related pathologies

like essential tremor, Parkinson’s disease, dystonic disorders and cerebellar disease of head

trauma (Strickland and Bertoni, 2004). And even though it affects only 0.4% of the world

population, it increases to 6.3% for people aged between 60 and 65 and to 22% for people

aged above 94 (Louis and Ferreira, 2010). Additionally, tremor is also directly related

to decreased hand function (Ranganathan et al., 2001; Carmeli et al., 2003), and for that

reason strongly influences the use of technological devices.

Thus, with the global increase of older adults population, understanding, modelling

and dealing with tremor is a significant concern in designing assistive technologies. In

that sense and considering indirect input devices (such as the mouse or pointer-based

interaction), Pak and McLaughlin (2010) suggested that reducing cursor speed, and gain

(the compensation between users’ actual movement and the distance the cursor moves on


Figure 2.6: Visualization of the growing plant metaphors within the app to motivate older

people through conditioning, goal setting, and self-monitoring.

screen) when nearing targets, allows older adults to acquire a target more efficiently and to

perform more extensive (and less precise) motor movements. These concerns regarding

tremor are even more relevant when considering touch-screen interface design, where

precise fine movements are required. However, Stobel and Blessing (2010) showed that

older adults are capable of accurately performing finger gestures on touch-screens if they

can perform those actions at a slower rate. To these findings, others added that provided

the appropriate button size and space they can also perform effectively with this kind

of technology overcoming tremor issues (Jin et al., 2007), reducing usage barriers and

avoiding frustration (Kobayashi et al., 2011).

Similarly to cognitive impairments, both training applications and exergames have been

used for compensating motor age-provoked declines. Concerning the first, applications

have been used in order to improve motor control and rehabilitation (Annett et al., 2009;

Silveira et al., 2013), increase walking through socialization (Mubin et al., 2008), or

by using a metaphor of a growing flower (Albaina et al., 2009; Silveira et al., 2013)

(fig. 2.6 and fig. 2.7). Regarding the latter, most exergames are designed to prevent,

manage or improve motor capabilities by helping and motivating older adults to engage

in physical activities. In this context, several Nintendo Wii games have been tested with

positive results (Lawrence et al., 2010). Others such as SilverPromenade (Gerling et al.,

2011) used console components for motivating older adults to take virtual walks (fig. 2.8).

Additionally, physical training often demands custom-built devices and sensors (de Morais

and Wickstrom, 2011), like the walk-boards (Kim et al., 2012), or the work performed

with sensing carpets (Smith et al., 2009) to guide and motivate older adults into exercising

more (fig. 2.8).

Once again, and like all the other types of age-related limitations, mobility impairments,

due either to motor impairments, or to current constraints, cognitive declines or disease,

often compromise independence, and the ability to accomplish basic daily routines leading

to a decrease in social interactions and quality of life (Mazzeo et al., 1998).


Figure 2.7: Flower user interface consisting of a general overview (left), and a day overview

(right).

Figure 2.8: An example of a SimplePromenade virtual walk (left), GUI displaying a virtual

instructor remembering a particular movement protocol and user scores (right).

2.2 TV and Older Adults

During the past years, digital TV as a media consumption platform has increasingly

turned from a simple receiver and presenter of broadcast signals to an interactive and

personalised media terminal, with access to traditional broadcast as well as Internet-based

services. TV panels currently on the market offer embedded digital processing platforms

(Connected TV), which turn the TV into an application platform and service terminal.

TVs are now used for web browsing, social networking, and to share data, in an attempt

to provide viewers with the ability to create, edit, share, and control contents (Cesar and

Chorianopoulos, 2009).

This evolution, however, was accompanied by the lack of assistive technologies which

tends to leave out a significant portion of potential users. User groups like disabled

or older adults face problems when using the services offered by the typical smart-TV.

Concentrating on the latter group, they often face problems which span from issues

connecting their TV to the Internet, to problematic barriers that digital menus and electronic-

program guides (EPG) raise (Cooper, 2008; Carmichael et al., 2006), to problems related

with the available means of interaction. All these are strongly linked to the age-related

impairments that were described in the previous section. For older adults with visual

impairments, TV menu text can be difficult to make out and read. Too much information

on the screen can represent a cognitive overload for older adults who get confused with the


options available. And it can be laborious for them to learn and remember the sequence of

steps required to achieve a goal, putting additional strain on their memory requirements.

The development of more accessible TV and its relation with common ICT applications

(especially social media but also education and health-related services) could make a big

difference for older adults living quality (Trinh et al., 2012). They have the potential to

enable or simplify participation and inclusion in their surrounding private and professional

communities. Taking this into consideration, several researchers focused on delivering

new interaction opportunities and services tailored for older adults and focusing on the TV.

In this section, we detail how TV-based technologies and interaction have evolved and

how these represent constraints or advantages to the older segment of the population. We

also provide detail on how TV-based technologies have been used in the past for fostering

social engagement within this population.

2.2.1 TV-based Interaction

Past work has focused on several ways of interacting with TV-based applications. In the

following sections, we summarise the main constraints and evolutions regarding RC, text

input, pointing, touch, speech and multimodality.

2.2.1.1 Remote Control

One of the most discussed research topics surrounding TV-based applications and older

adults is the one related to the possibilities of interaction, or in other words, the alternatives

to the traditional RC.

As TV evolved, and with the appearance of TV-based applications, the RC also evolved

and started to be looked upon as a more elaborate second interface (Springett and Griffiths,

2007). With the need of controlling additional functionalities, more and smaller buttons

were included. This raised issues that although appropriate to all users, were even higher

for older adults, mainly because of age-related declines. For those with visual impairments,

smaller buttons are harder to find. For those with motor impairments, the smaller the button

the most problematic it can be to press the required key. And for those with cognitive

limitations, the most functions a button has associated or the more buttons there are with

distinct functions, the harder it is to remember each one. Adding even more complexity

is the need for a TV user to constantly switch focus between the screen and RC while

trying to complete a task. For older adults, this is not only associated with aggravation

of cognitive requirements but also means the need for clear corrective lenses for each

interface (Carmichael et al., 2006). Also, not only RC usability is poor (Lessiter et al.,

2004), but also its design varies widely (Eronen et al., 2004).

Going into more detail, as soon as TVs evolved to incorporate tasks related with more

than switching between channels, several difficulties emerged regarding RC interaction.


The main ones were related with not having an any easy way of inputting text (Roibas

et al., 2005).

2.2.1.2 Text Input

Ingmarsson et al. (2004) was among the first to identify this issue and to present an

alternative. The Number Typer method (TNT) showed writing speeds similar to Personal

Digital Assistants (PDAs) handwriting methods, and as a result its prototype was later

adopted by Iatrino and Modeo (2006), who compared it against other two alternatives

(fig. 2.9): a multi-press feature associated with a time-out interface but with no visual

feedback (like in TNT); and a virtual keyboard.

Figure 2.9: Iatrino and Modeo (2006) writing task approaches:: Multi-press with timeout

keyboard and visual feedback or TNT (left), multi-press with timeout keyboard (middle),

virtual keyboard (right).

These experiments were the basis for the emergence of two text-entry techniques which

became the main alternatives still used to date: the virtual keyboards and the multi-tap

technique. In the first, and although the simplicity of interactions, character selection can

be slow because of the long distance between them and the number of keys needed to be

pressed (Iatrino and Modeo, 2006). In the second, even if this problem is reduced (as a

button matches multiple characters), the need of dividing attention between remote and

television screen is the main issue impacting the interaction (Nichols et al., 2006; Cooper,

2008). Additionally, and with both techniques, the issues are aggravated when considering

older adults because of the reduced motor and cognitive capabilities. Thus, although

pioneers, findings on both these interactions only showed the need for new devices and

interaction techniques to appear to accomplish satisfactory writing tasks (even more among

older adults).

In the same way, and when focusing on the traditional RC interaction, other studies

like the one by Barrero et al. (2014) showed that RC design does not have a significant

impact on older adults performance, being slower for both simple and complex tasks and

across all solutions. Once again, this showed the need for alternative ways of interacting

with TV (fig. 2.10).

Additionally, as a result of TV evolution, and by not letting users directly manipulate

items on UI when using an RC (Shneiderman, 2005), other problems were raised, like the


Figure 2.10: Remote controls used in the study (left), mean and standard deviations values

of writing speed per method (right).

ones related with interacting with complicated digital menus like EPGs (Cooper, 2008;

Carmichael et al., 2006). Thus, several research has focused on developing different kinds

of interactive methods like pointing devices (Kunert, 2009), touch-controlled mechanisms

(Roibas et al., 2005) and speech recognition (Roibas et al., 2005).

2.2.1.3 Pointing

When considering pointing devices, a popular approach was the development of gyroscopic

RC, an idea that first emerged as the gyro-mouse (MacKenzie and Soukoreff, 2002). By

detecting hand movements and translating those into variations in the position of a cursor,

this type of control allowed users to move the cursor (or the mouse) as if they where

pointing remotely with a finger.

A typical case of a gyroscopic controller that become popular with many researchers for

performing typical TV-based tasks was the WiiMote (Lawrence et al., 2010; Neufeldt, 2009;

Aoki et al., 2010) although other works revealed problems with this type of interaction

related to fatigue, response delay and precision (especially for older adults as it is affected

by hand tremor) (Myers et al., 2002; Nacke, 2010; Konig et al., 2009), it continues to be

used in more recent research. The best example of that was the WeSlide (Godard et al.,

2013), a gestural text-entry technique that by using only the WiiMote as an input device

for approximate pointing (fig. 2.11), managed to be faster, less error-prone and strongly

preferred by older adults over traditional multi-tap.

2.2.1.4 Touch

Shifting away from pointing, and first suggested by Enns and MacKenzie (1998), the

use of touch-screens has also been the main research point when considering TV and

older adults. The possibility of providing direct control over interface elements (one of

RCs main disadvantages) allied with the rapid spread of smart-phones, merged this idea

with the possibility of using these devices as alternative controllers (Silva, 2011). Google

TV remote applications, Android remote, reRemote, redEye, and the Verizon FIOS were


Figure 2.11: WeSlide: selection of ’G’

among several examples of this (fig. 2.12). However, these solutions were also criticised

as they divide visual attention between two screens.

Figure 2.12: Some smart-phone touch-based RCs examples: the reRemote (left); the

redEye remote (middle); and the Verizon FIOS app

Trying to tackle this issue, Choi et al. (2011) developed an optical touch-pad for

tracking the thumb hovering over its surface and designed a TV application to demonstrate

a possible new interaction style entitled RemoteTouch. In this, a shadow representing

the users’ thumb touches the screen, presses a button, flicks a cover-flow list, or draws a

simple stroke, while the thumb stays and moves on and above the touch-pad. This solution

showed the potential of touch-based RCs in TV context. However, its main limitations

were related to the use of non-standard technology and the lack of testing with older adults

(fig. 2.13).

Taking half a step away from touch-screens, Bailly et al. (2012) introduced Finger-

Count gestures. This consisted of a set of simple and fast to perform multi-finger and

two-handed gestures which worked for both touch-screen and gesture interaction from a

distance (fig. 2.14). Results showed that by being easy to use and learn, this could alleviate

problems related to occlusions, accuracy and reachability. However, not only they also did


Figure 2.13: RemoteTouch controller (left) and the mapping from the RemoteTouch

controller to the screen (right)

not test it with older adults, but their conclusions also showed that typing on a physical RC

remained faster than using gestures for that purpose.

Figure 2.14: Touchless Finger-Count in the context of interactive television

Finally, quite recently, another study from Barrero et al. (2016), by comparing a

touchpad solution with previous TV-entry methods (keyboard, gyroscopic, etc.) showed

that discomfort and problems could occur among older adults using any of these alternatives.

Thus, solutions that foster older adults use should not be focused on using only one of

these interaction possibilities (fig. 2.15).

2.2.1.5 Speech

Using speech as a means of interacting with technology originated from very early works

in natural language processing (Furnas et al., 1987) and online systems (Elder, 1970). Only

at the beginning of the century, Shneiderman (2000) pointed out that speech interaction

can provide an improvement in the quality of life for specific user groups like older adults.

And only about five years later it was considered in a TV context.


Figure 2.15: Pointing devices and virtual keyboards used in Barrero et al. (2016)

Komine et al. (2004) was the first to do this by conducting a Wizard-of-Oz experiment

to design an “operation interface easy enough for anybody to use”. Their results were

promising, with participants considering easy to make programme selections and the

ability to access necessary information without the need to go through hierarchically

structured menus, is much appreciated. Still, these results were influenced both by the use

of natural language (where everything can be said), and the Wizard-of-Oz approach (where

everything is recognised).

One year later, the IntegraTV-4all fully implemented a speech module targeting both

TV context and older adults (Ceccaroni et al., 2005). Aiming at facilitating personal

autonomy and social integration through services of leisure and information, this platform

among several of its features made it possible for users to interact with voice and navigate

without the need of visual references, providing all possible options by voice (fig. 2.16).

Although RC could also be used, and speech was used more in a context of severe visual

impairments, this was a good solid example of how this type of interaction could play an

important role in TV-based interaction for older adults.

From that point on, few systems or projects focused on targeting older adults interaction

with TV only through speech. Instead, the importance of using this type of interaction was

further considered in the context of multimodal TV systems (along with pointing and RC

alternatives). Turunen et al. (2009) presented a first multimodal media centre interface,

which combined voice, gesture and physical touch. The speech was relevant in the way

users could fully control the media centre not using any other modality. However, the

results showed that both touch and gestures were preferred to speech justifying it both

with the limited number of words which could be recognised (about 100) or with the

microphone angle.


Figure 2.16: IntegraTV general interface (left); and speech-based interface (right).

Finally, both Bernhaupt and Pirker (2014) and Aylett et al. (2014) reviewed several

obstacles in combining Human-Computer Interaction (HCI) and speech technology and

identified several technical challenges. Among these, relevance should be given to the need

for each user to actively activate speech recognition to avoid any misleading interpretation

of the ambient noise or sound in the living room, and for the fact that distinct languages

have different recognition accuracies. They also indicated that speech interaction implies

more cognitive load and social intrusion than any other kind of interaction.

2.2.1.6 Multimodal TV

From all systems and interaction modalities referred, conclusions can be drawn which

support the advantage of adopting a multimodal approach: While for some tasks RCs are

too slow or insufficient (like writing text or navigating complex menus), alternatives such

as touch imply dividing attention between two screens (even more critical for older adults

with decreased cognitive functions), while pointing is not suitable for individuals with

motor impairments like tremor, and speech can be too complex to recognize and learn or

too limited to provide a broad interaction.

Additionally, preliminary research on multimodal interfaces revealed that people favour

it over unimodal interaction (Hauptmann and McAvinney, 1993; Oviatt et al., 1997). Some

facts play a role in this preference. Increased accessibility is one, as “using a number of

modes can increase the vocabulary of symbols available to the user”(Keates and Robinson,

1998) and yield a “highly synergistic blend in which the strengths of each mode are

capitalised upon and used to overcome weakness in the other” (Oviatt, 1999). Another

factor is efficiency and reliability, with multimodal interfaces speeding up tasks completion

by 10% (Brewster, 1998; Mankoff et al., 2005). Moreover, they also improve error handling


and reliability, with users performing about one third fewer errors (Oviatt, 1997), and

provide increased flexibility, because of the ability to switch among different modes of

input to whichever is more convenient to a user (Keates and Robinson, 1998). All these

are of extreme importance when considering older adults, which impairments can span

across several dimensions by resorting to several modalities they have the opportunity of

compensating incompatibilities of using others.

Coincidence or not, as modern TV emerged, manufacturers and service providers

introduced multimodal platforms which incorporated several types of RCs, touch features

(including smart-phone compatibility), and gesture-based interaction, on top of which they

added speech capabilities. SAMSUNG was among the first to add some of these to their

premium-class TV sets, and even though they lacked robustness, this showed that new

and multiple ways to control the TV should be taken seriously. Panasonic also presented

a Text-to-Speech feature with market quality recognition capabilities, integrating it with

a whole set of other features. And more recently, Google TV, Apple TV and Samsung

Internet@TV also presented systems supporting multiple interaction options, including

speech commands, and both gyroscopic and traditional RC.

Even though all these examples are suited for the entire population and did not cater to

specific older adults needs, by supporting multimodal interaction they were a step closer

to being more inclusive. And even though few research exists focusing on multimodal

TV for older adults, some of the already mentioned speech-based examples made use of

multimodal features, like the IntegraTV-4all project (Ceccaroni et al., 2005) and the system

built by Turunen et al. (2009). Finally, the GUIDE project also implemented a TV-based

older-adult-based framework which made use of several interaction modalities. As work

developed on GUIDE is part of this PhD, more explanations and report on the use of these

multimodal mechanisms will be presented in the next chapter.

2.2.2 TV as a Social Solution

Although the majority of social network solutions developed having older adults in mind

were not focused on TV, some few examples worth mentioning exist.

The first meaningful attempt was a medicine platform for older adults developed by

Philips (2005). The Philips Motiva (fig. 2.17), monitored older adults progress outside

the hospital promoting self-management and facilitated two-way communication between

these and their care providers and family. As a result of this project, several conceptual

prototypes (Social TV, Windows Media Center, Collabora TV, and ConnecTV) were later

developed by Alaoui and Lewkowicz (2013). Although these were not validated in real

scenarios with the target population, they constituted the first examples of how making use

of several ways of communicating with caretakers (voice communication, texting, chatting,

video conferencing or even providing TV recommendations) could be a way of achieving

better social support through the TV.


Figure 2.17: Promotional image of the Philips Motiva system

In 2007, the T-Seniority project 1 also addressed social inclusion implementing its

communication functionalities around the use of a traditional RC. Being used by hundreds

of users across Europe, this was a clear demonstration that service-oriented technological

platforms for active ageing could be widely and successfully deployed through the TV.

It also showed that the simplicity factors around RC use when merged with adaptation

features could also be fundamental characteristics of a service to be adopted by this

population (fig. 2.18).

Figure 2.18: T-Seniority software screenshot

Since 2010, the Ocean Blue Software (OCB) 2 has been a clear demonstration of how

integrating Google solutions (Nexus TV applications) with assistive technology can help

deliver home support to vulnerable and impaired older adults. The project focused on

providing access to various services including a TV-based SNS. As main points to consider

were the development of a “Talking TV” enabling EPG and menus be voiced to compensate

vision declines and the possibility of changing colour schemes and increase or decrease

1http://tseniority.idieikon.com2www.oceanbluesoftware.com


font sizes manually. Additionally, they also developed a single-button version of an RC to

make it easy to turn on and off some of these features. Still, the SNS based services proved

to be inefficient because of its restricted use which was typically associated with familiars

being Facebook users and not keen on being introduced to a whole new service (more on

this subject can be found on the next chapter).

Figure 2.19: Eldy TV Software

Founded in 2012, ELDY 3 (fig. 2.19) Is focused on offering technological tools and

human support to reduce older adults isolation and loneliness. They developed Eldy TV,

software which provides services like E-mail, Web Browsing and Messaging TV, designed

in a way suitable for this segment of the population. Still active, since its deployment,

it has been used by around 300 thousand users. However, as the main limitation, ELDY

suffers from the same problem as OCB, being dependent on younger counterparts adopting

it.

Finally, three years ago, the iNeighbour TV (Abreu et al., 2013) also tried to enhance

opportunities in older adults’ social life. Using TV as “a more familiar UI for them” it

allowed the access to common Internet-based services making it possible for older adults

to both explore who can help with a task and offer themselves help to others (fig. 2.20).

As the main limitation, the service also suffered from the same problem described in the

3www.eldy.eu

Figure 2.20: Screens of iNeighbour TV features


previous examples: having been implemented to foster the increase of social interactions

with other older adults, left out (younger) family members, consequently leaving out the

main source of older adults’ social interactions.

2.3 Summary

In this chapter, we started by presenting relevant research concerning both older adults

characteristics and limitations. Regarding sensory constraints, we’ve seen that particular

attention should be given to visual impairments by avoiding small texts and tiny icons,

providing highly saturated and contrasting colours, and using big enough legible-enough

fonts. Additionally, hearing loss should also be tackled by mainly combining auditory

signals with other types of feedback to convey the same information (or resorting to

multimodal feedback). When considering cognition declines, we’ve seen the necessity

of always displaying the necessary information for each specific user to perform a task,

employing previously learned mental models and procedures, and drawing attention to

the most frequently performed actions. Additionally, in this context, we’ve also seen the

importance of resorting to multimodal features to avoid clutter and drawing attention to

important elements through other sensory channels. Finally, and as the main physical

limitation, we’ve seen that tremor should be compensated by reducing cursor speeds when

making selections, also by decreasing the rate of finger gestures and touch selections and

providing the appropriate button size and space.

Following this, we described TV-based solutions which made use of several interaction

modalities as a way of easing older adults adoption and interaction. In this context, we

have seen that RC although remaining the default way of interaction, has proven to be

insufficient for older adults especially when trying to manipulate complicated menus or

performing text input tasks. Additionally resorting to new types of RCs, like gyroscopic or

touch-based also seems insufficient mainly because of older adults impairments. Moreover,

and although some research has been showing that resorting to speech might help in a

TV-based context, others have shown that it might be too challenging for this segment

of the population. Therefore, providing TV with multimodal mechanisms which support

several of these modalities might be the way of compensating age-related declines.

Finally, we’ve also provided an overview of existing social solutions targeting both TV

and this segment of the population, showing that although some were implemented, none

is widely used mainly because they leave younger counterparts out.


Chapter 3

Design of TV Applications targeting

Older Adults

This chapter presents the steps followed in this thesis for drawing recommendations

targeting the appropriate design of TV-based applications which consider older adults

characteristics and preferences. It focuses mainly on the work developed in the context of

the GUIDE European Project, more specifically in the development of several multimodal

TV-based prototypes that serve as requirement gathering phases relating older adults

characteristics with the TV context of use. This chapter starts with a description of

the methodology that integrates all these steps. It follows with the description of every

prototype development phase of the GUIDE project which are relevant to this thesis. In

each section, the procedure and a summary of the results obtained are described. The

chapter finalises with the formulation of all relevant recommendations originated from all

the steps.

3.1 Methodology

In the context of the work reported in this chapter, end-user requirements have been col-

lected using a mixed methods approach. Data were drawn from multiple approaches like

literature, quantitative and qualitative data analysis. A triangulated approach then saw

to establish the predominant weight of evidence by agreement or disagreement between

sources, knowing the strengths and limitations of the methods and increasing the validity

and reliability of qualitative findings (Langdon et al., 2003; Flick, 2009; Miles and Huber-

man, 1994). The framing of all phases as design problems constrained the triangulation

and assisted it by directing the focus of comparison to the design of the final interactions

between the user, system, technology and usage context. Therefore, the particular methods

were based on finding relevant results in literature studies and UI accessibility standards,

performing interviews and surveys with both developers and older adults and performing

user trials with the latter. From these, technical and interaction requirements have been

drawn, laying the basis for the design of multimodal TV-based interfaces for older adults.

37

Chapter 3. Design of TV Applications targeting Older Adults 38

Figure 3.1: The User Centered Design Approach

Departing from this basis, we divide the user-centred design process into four main stages

(figure 3.1):

1. The first stage was the Early Requirement Stage composed by both the Literature

Requirements and the Technical Review Stage, where data from literature about

models of older users, technical specifications and standards of accessibility was

collected from several sources and compiled into requirements and use cases.

2. The second stage was the Stakeholders Research Stage, where two user inquiring

phases (surveys) were conducted to infer primary data about the target population and

technical specifications relating to the use and development of TV-based multimodal

interfaces. This phase was divided between methods targeting research on older

adults, and others targeting TV-based application developers.

3. The third stage was the Prototype Development Stage, where several application

prototypes were developed using the requirements collected in the previous stages.

These prototypes were tested with older adults in the following stage. In the context

of GUIDE user studies three distinct applications were developed at this stage, the

User Test Application (UTA), the User Initialization Application (UIA), and the

Electronic Program Guide (EPG) application. These prototypes were also refined

with the results from the first user studies.

4. The fourth stage was the User Study Stage, where multimodal interaction aspects

were understood and refined, as well as adaptation aspects. Additionally, at the

second iteration of this stage, an overall evaluation was performed with older adults

to cover all multimodal and interaction aspects and guarantee a suitable design,

being a confirmatory step where all parts of the GUIDE system and applications

were validated.

In the following sections, these stages are described in detail both regarding method-

ology as well as regarding results and how these can be translated into design guidelines.

Relevant here is the fact that stage 1 is not described in this chapter as relevant work was

included in the previous, while stage 2 is described in 3.2.Stakeholders Research Stage,


and stages 3 and four are described together in the context of two iterations, with the first

iteration being described in 3.3.Multimodal Interaction User Study and the second in the

following section (3.4.Adaptive Interaction User Study).

3.2 Stakeholders Research Stage

After the early requirements were concluded, the Stakeholders Research stage began tar-

geting two distinct groups of users: older adults, and developers of TV-based applications.

For each, a distinct procedure was followed constituted by questionnaires and focus groups.

In the following subsections, we describe both.

3.2.1 Inquiring Older Adults

An exhaustive inquiring process (survey) was developed targeting older adults and using

the operational variables and measurements identified in the previous requirement stage.

The aim was to help develop a user model based on extensive user data, which divided

users into several clusters, each one with different characteristics. Additionally, this served

as the first step for making an adaptation of UI elements to older adults possible.

3.2.1.1 Procedure

For this phase, 46 (forty-six) older adults with different age-related disabilities were re-

cruited (30 female and 16 male). The average age was 70.5, and all participated voluntarily.

All participants were Spanish. All data was audio recorded and analysed after the survey.

All activities performed involving older adults were safeguarded from the ethical point of

view.

The survey was conducted as a face-to-face 15 to 45-minute interview, performed by

psychologists with a background in neuropsychology, and administered individually in

order to establish the profile of every participant. Questions involved more than 100 items,

combining objective and subjective assessment of the individual capabilities. Information

was collected concerning the following areas: socio-demographics, attitudes towards

technology, coping styles with new technologies, attributional styles toward technological

devices, daily living activities, vision, hearing, mobility, cognition, and personality traits.

The detailed questionnaire can be found in Appendix A.

3.2.1.2 Results

Considering the large number of variables contained in the dataset, these were submitted

to a two-stage process of analysis where correlations were made and a k-mean cluster

(with a value of k=3) analysis (Kanungo et al., 2002) was performed, reducing the results

to only significant data. This K-mean clustering algorithm works like an Expectation


Maximization (EM) algorithm, which partitions n observations into k clusters in which

each observation belongs to the cluster with the nearest mean. Table 3.1 uses the result of

the k-means clustering to characterise what a typical member of the high, medium and low

impairment group would be capable of, based on the calculated cluster centres.

Table 3.1: K-Means Cluster Centers for visual, hearing, cognitive and motor variables

Vision LOW MEDIUM HIGH

Close Vision: level able to read perfectly 20/20 20/60 20/80

Distant Vision: level able to read perfectly (meters) 5 5 20

General Eyesight good excelent normal

Seeing at Distance good poor poor

Seeing at Night normal poor poor

Colour perception good bad bad

Hearing LOW MEDIUM HIGH

Able to hear a sound of 500Hz? Yes Yes No

Able to hear a sound of 2KHz? Yes Yes Yes

Conversation from a noisy background Excellent Normal Normal

Cognition LOW MEDIUM HIGH

TMT (seconds) 30 49 136

Cognitive Executive function (no impairment) (low impairment) (high impairment)

Motor LOW MEDIUM HIGH

Mobility diagnosis none hernia / slipped disc none

Muscular Weakness never A few occasions Frequently

Write No difficulty No difficulty Mild difficulty

Tingling of limb difficulty No Mild Mild

Rigidity difficulty No Mild Moderate

The profiles were formed by combining and grouping all modalities simultaneously

such that a specific grouping may represent capability on users perceptual, cognitive and

motor ranges. A particular user may present, for example, low vision, medium cognition

and high motor capability levels. This resulted in the definition of three types of user profile

clusters for every variable - low, medium and high impairments profiles. Following this,

adaptation may utilise more accurate quantitative data to locate the user’s profile further.

This may entail a short visual acuity, cognition, physical or hearing test. Thus, these

profiles would be initially used to index the GUIDE user profiles following a successful

initialisation application interaction.

3.2.2 Inquiring Developers

Concerning developers, the general goal following the preliminary requirements was to

explore and understand the common practice among developers working on Set Top Boxes

(STBs) and TV-based applications. Thus the focus was on inquiring developers to obtain

data about current tools and APIs used in Set-top box/connected TV platforms and to

investigate how accessibility is perceived and applied in the industry. As a second goal, we

explored developer knowledge to identify which tools would developers need, to efficiently

integrate accessibility features into their applications. Finally, we aimed to stimulate new

ideas through discussions and to identify new relationships between objects embodying


GUIDE concepts and objects embodying common practice. Both subjects were covered

through both focus-groups and online surveys.

3.2.2.1 Procedure

Two focus group sessions were carried out with TV-based applications experts in a natural

and interactive focus group setting. Two moderators conducted the sessions (for ensuring

progress and topic coverage), and each session had between six and eight participants

lasting around 120 minutes. Sessions were initiated with presentations of scripts containing

development and use cases that cover different aspects of the project and its concepts,

including requirements and proposed guidelines from the previous stage. Interactive

brainstorming and discussions followed.

Additionally, a questionnaire was designed to investigate how accessibility is currently

perceived and applied in the industry. This survey also allowed respondents to vote on

the most essential features of the envisaged discussion. In total, 81 participants from 16

countries, and 30 companies all over the world participated. The questionnaire can be

found in Appendix B.

Findings from both inquiring procedures assisted in the specification of the interfaces,

architecture and core processes to be implemented in the project’s multimodal constitution.

3.2.2.2 Results

Developers agreed that if users are involved in every development phase of the applications

(or in the maximum phases possible), the resulting UI will be more usable. It was concluded

that for older adults, UIs should be maintained clear and simple, however, without giving

the impression, they have been designed for someone with impairments (or of being a

“system for seniors”). Additionally, they also agreed costs are the current major reason

for reduced application of user-centred design in their industry (followed by time and

lack of accessibility awareness). As the current most important device on interaction with

TV-based applications, they agreed on the RC, stating it should have a central role in

the interaction, and only be relegated to a secondary role if that is a result of each user

interaction preferences. Gesture control and speech input were recognised as secondary

technologies.

Concerning adaptation, in general, participants agreed UI automatic adaptation has the

potential to help older adults accessing the generality of ICT services including modern

TV. Still, they also agreed adaptation mechanisms should avoid changing interface aspects

unless needed for specific user interaction. Also, radical changes in the UI must be avoided,

so the user feels in control and not lost in the interface. If a radical change is indispensable,

the UI must inform the user of the proposed changes. Identified as the main obstacle to UI

adaptation is the fact that older adults present too many differences between each other.

Therefore, for adaptation to fit each user, the system (or the application) should find a


way to know user impairments, preferences or characteristics. This “discovery” should

occur the first time the user interacts with the system, while also being short, not too much

intrusive, and entertaining. The most important conclusion debated in this subject was the

one saying the system should support UI markup as an interface between application and

adaptation. In their opinion, this would allow developers to keep tools and development

environments and without too much additional effort, taking the first step into an accessible

design.

For most participants, TV platforms and TV-based applications will be the most

relevant platforms in the future, while web-based application environments will become

more important for TV. Finally, they also pointed out that a system like GUIDE must

consider situations where multiple users are using the TV at the same time.

3.3 Multimodal Interaction User Study

The first iteration of prototyping and user trials was a primary source of user requirements

engineering, with the aim of observing older users in a lab situation interacting with

alternative modalities of interaction, gathering both qualitative and quantitative data.

Thus, a multimodal application was developed during this stage, the User Test Applica-

tion (UTA). Concrete and practical user trials were conducted with the UTA to identify the

most relevant UI configurations, viable usage methods (e.g. gestures, command languages,

etc.) and user impairments, as well as to understand older adult’s typical patterns when

performing the unimodal and multimodal interaction. Findings obtained from its evalua-

tion also fed the development of later prototypes in the second prototyping iteration and

respective user studies.

In the following sections, we start by describing the UTA application regarding general

goals and design considerations, following with a description of the user study procedure

followed. Lastly, we provide an extensive description of the results obtained in the trials.

3.3.1 User Test Application

After conducting the survey and deriving the user models, we developed a test prototype,

the User Test Application (UTA). The UTA goal was to collect user preferences regarding

specific UI elements and interaction contexts, and also for understanding user dimensions

and requirements during practical trials, and before experimenting with real applications

with older users. The UTA was a simple application that asked for user input using

different modalities to perform several tests regarding interaction preferences and UI

elements preferences. It also made possible to simulate several interaction contexts to

understand what are the risks of making older adults interact with TV’s using non-standard

modalities like speech and pointing.


Figure 3.2: Screen-shots of the UTA prototype

The UTA was the most straightforward application in GUIDE consisting in a low

fidelity prototype with a set of white background pages with text labels, images and buttons

(usually four to eight per screen) for performing selections (figure 3.2). The UTA was

developed without any previous mock-up design. It was based on a set of tests focused on

sensory, cognitive and motor capabilities of older adults. UI elements were defined in XML

documents. This allowed for the construction of different interaction scenarios by merely

modifying the elements on a text editor, defining button size, location and labels and also

images to be placed and what type of output feedback would be provided (visual, audio,

both). A Virtual Character (VC) was also integrated to serve as a friendly intermediate

between the application and the user. Both the VC and the pointing recognition mechanism

(which made use of the Microsoft Kinect) worked as independent applications which

received and sent messages from and to the main application by socket connections. For

speech recognition, the wizard of Oz approach was used.

3.3.2 User Study Procedure

Making use of the UTA, users had the opportunity to experiment with the different

modalities and devices of interaction. Seventeen people (4 male and 13 female) between

the ages of 55 and 84 (averaging 65.8 years old) participated in the user-trials. All of

them were recruited in Spain and presented different levels of expertise in technology,

as well as different physical and cognitive abilities. In each user-trial, every participant

was presented with all possibilities of interaction (pointing with the hand, using the Wii

remote, interacting by speech, or RC) and asked to perform a series of tasks related with

TV interaction. The tasks were divided into several scripts concerning different types of

interaction or different UI elements:


• Visual: The TV screen presented buttons on different locations. Participants were

encouraged to say which buttons’ location, size and inter-spacing would suit their

preferences better. Afterwards, they were asked to select their colour preferences

according to the button and background. Subsequently, they were requested to read

some texts with different size fonts and comment which one they would prefer. For

each task, they were required to point at a particular button on the screen using their

preferred modality.

• Audio: Participants heard different messages at various volumes. They were asked

to repeat what they heard and to select the most appropriate volume for them. This

was also asked while watching TV (with the sound on). Additionally, female and

male voices were compared, and participants reported their preference.

• Motor: To examine which gestures participants prefer, they were requested to

perform a variety of gestures in the air (with one or both hands). They reported

which ones they were comfortable performing. Participants also had to select

different buttons on the screen using either a conventional RC or their fingers. They

reported which one suited them better taking into account the button placement on

the screen: bottom or top, right or left. Participants were also asked to interact with

Tablet-based applications, entering numbers in a visual keyboard and controlling

virtual maps.

• Cognitive: Cognitive tests in the form of “games” were performed to evaluate

participants’ visual memory and attention capacity. Images were shown on the screen,

and participants were requested to memorise them. The images then disappeared,

and participants were asked to select the location where a specific image was. Time

of selection was measured.

• VC: The same message was presented to participants, either by a VC or by an audio

voice. Participants then reported their preference. This was also done for VC and

text. Participants were asked to choose one of three types of avatars (face avatar,

half-body avatar and full-body avatar).

• Multiple modalities: Participants were asked to perform several selections using

only one modality or several modalities at the same time. This was also asked while

simulating different contexts of interaction, and while interacting from different

locations on the room. They were also asked which way of presenting information

they preferred, combining presentation in one modality and several modalities at the

same time.

The measurements taken in this study were the number of errors, the time rate, and

observation of participant’s actions and behaviours. Every user-trial lasted around 60


Figure 3.3: User-trials photos. Colour selection screen (on the left), and participant using a

Wii remote (on the right).

minutes, and during the trials, participants had to sit down in front of the screen ensuring

the distance was the same for every user. Every assessment was recorded with two cameras,

one of them focused on the participant and the other on the TV screen. Analysis of the

user-trials was made by analysing those videos. A specific image taken from the two

cameras can be observed on fig. 3.3.

3.3.3 Results

The following results are from behavioural patterns observed in the user-trial videos as

well as from participant comments. The analysis was grouped into eight distinct domains

concerning UI design aspects and interaction patterns and observations.

Button configuration: Users preferred big buttons (83%), well-spaced between each

other (83%) and centred on the screen (65%). Also, there was no clear preference regarding

the colour of buttons, except for all users preferring high contrasts between the button

background colour and its text colour. Regarding size, big buttons were preferred by

users because they were easier to see (and older users typically have some kind of visual

impairments), and also easier to select as a more significant area was available for users to

interact when pointing. More space between buttons was preferred because it added more

precision to pointing selection actions, and centred buttons were preferred because it was

easier to point to the centre of the screen, than to the peripheries. However, the preference

for UI button location, more than based on user abilities or preferences, should be based on

tasks being performed or the type of application being executed. For example, older adults

behaviour indicated that buttons centred on the screen can be less suitable while watching

TV, but more suitable when using an application for exploring photos. Additionally, if

the input recognition mechanisms are accurate enough, less space between buttons would

bring the advantage of having to travel less distance with the finger to make a selection.

Text Configuration: From six sizes of fonts presented to users, almost 50% of them


preferred the third bigger, and the rest of them divided themselves between the second

(18%) and fourth (24%) bigger fonts. No one chose the two smaller text sizes, and only

one user chose the biggest one. Also, more than 50% of users with minor vision difficulties

(users who can’t see all sizes of text presented) preferred the second and third bigger

sizes, and 90% of users who didn’t have any vision difficulties preferred the third and

fourth bigger sizes. Additionally, for dark background colours, more than 75% of users

preferred white text, and vice-versa. As for coloured backgrounds, there was a division in

the preference of text colour, from white (32%) to black (31%) to dark blue (25 %). When

asked for the reasons for these choices, older adults stated they did not prefer the bigger

text fonts because otherwise too much area would be taken on the screen, leaving no space

for other UI components. Also, another reason for choosing smaller fonts was because

they provide a more comfortable reading. However, smaller fonts were more challenging

to read for people with visual impairments like the majority of the participants, and for

that reason, a medium sized font was indicated as the best option for them.

Audio Configuration: From five different volumes presented, almost 65% of the

users preferred the second loudest, and the rest of them divided themselves by the first

(12%) and third (18%) loudest. No one chose the fourth and fifth volumes. Also, 100%

of users with minor hearing difficulties (users who couldn’t hear all volumes presented)

preferred the first and second loudest. All users who didn’t have any hearing difficulties

preferred the second and third loudest. From these numbers is evident users with hearing

difficulties preferred higher volumes than users without those difficulties. Regarding the

audio feedback voice genre choice, there was no clear preference between male and female

voices (6 preferred male, four female, and 7 had no preference). Also, it was also evident

in the analysis that no relation existed between the gender of the user interacting and the

gender of the voice chosen, with both female and male users choosing both female and

male voices for audio feedback.

Gesture Configuration: Users had to choose from swipe, pinch and circle gestures,

with one and two hands. They all said to be comfortable doing any of them. When making

gestures on air, the preference has gone to the one-hand swipe (47%) and two-hand swipe

(or clapping) (47%) gestures, while any user didn’t choose two-hand-circle or any variation

of pinch. 12% of users didn’t like to interact using gestures on air. This means that while

participants preferred gestures easier to make, and had no problem whatsoever interacting

by gestures, this type of interaction should not be imposed on them, but instead be available

as an alternative (and intuitive) option for interacting with the TV. The existence of specific

gestures for specific tasks, like raising or lowering volume, is needed and should be

explained to the user before he or she starts using the system.

General Interaction: In these trials more than half of the users (53%) preferred to

interact using speech while 35% chose to point. These numbers evidentiate that alternative

ways of interacting with the TV (in this case speech and finger pointing) were preferred to


the traditional way (RC interaction). One reason for this could be the novelty effect of both

modalities, while other could be the lack of realism of the UTA. Still, this should not be

ignored. Additionally, it is also a fact that training can even improve interaction with these

alternatives as the user learns what is required to move the cursor on the screen to each

button or menu or what speech command must be issued to make that selection. However,

it was also evident from discussions with participants that regarding pointing interaction,

support would be necessary for interacting from different locations (front or side to TV)

and positions (sitting or standing), as they target distinct situations and interaction contexts

at home. On another note, and if not suggested, very few users (11%) alternated hands

or arms to make pointing interaction easier, and even when suggested only 2 in 6 tried it.

From these two who tried it, none continued the interaction with the opposite arm even

after noticing improvements. It was also observed that not only participants did not change

the hand they were interacting with when they could achieve a task with that hand, but also

they did not alternate hands when they could not select a button on the opposite side of the

hand they were interacting with. This means that if older adults do not intuitively alternate

between hands to perform better pointing interaction, not only it cannot be imposed on

them, but also pointing might not be the ideal interaction modality or at least the way

pointing was presented might not have been intuitive enough.

Multimodal Interaction: More than half of the users (53)% wanted both multimodal

input and output when they were using the system (even if only in specific contexts

of interaction). Moreover, 59% of users said they wanted to interact using more than

one modality at the same time, leaving only 29% wanting to interact using one single

modality. 82% of users said they preferred multimodal feedback from the system, leaving

less than 20% wanting information presented in only one way (just visual, or just audio).

Additionally, every participant was able to interact multimodally with the system and

combine speech and pointing, even when they preferred only one modality. However,

participants exhibited different multimodal interaction patterns during the trials. For

instance, in a first interaction, a user could speak first and point afterwards, and in the next

interaction make the opposite, or point and speak at the same time. This was observed

across several participants, indicating a lack of specific multimodal interaction pattern for

each or at least a lack of guarantees that even if most of the times a user interacts in the

same manner, he/she will do that every single time. Participants also changed the way they

interact depending on the type of feedback given while interacting, further supporting this

argument. Regarding user preferences in input and output modalities, there were apparent

differences between what participants stated their preference was, and what they asked

for when interacting. In fact, 100% of the users wanted multimodal output every time

information was presented to them because every user who said to prefer only one type of

feedback admitted differently when in specific interaction contexts. The same happened

concerning input modalities, with almost half of the users admitting they were wrong when


they stated their original preferences. For all these reasons, the main takeaway finding

from this is the notion that several modalities of interaction should be made available at any

time. In that way, users may choose which modality they prefer interacting at any given

time, as well as change between modalities when for some reason they cannot proceed to

interact with the original one.

Selection Preferences: From 11 users who performed pointing selection using their

fingers, all were right-handed, and 9 (82%) preferred to make selections on the right side

of the screen. Also when performing the same type of interaction, nine from 11 preferred

to select buttons on the top of the screen rather than on the lower part. This shows the

existence of a strong relationship between the arm used for interacting when pointing with

a finger to the screen, and the preferred area of selection (right-arm interacting almost

always means preference in selecting items or buttons in the right area of the screen).

Although this would need to be further validated with a sample of left-handed users, it

indicates a preference that suggests menus or interactive elements should be placed on the

top half of the user’s dominant side. When focusing on speech, when buttons on a UI were

labeled like “Button1”, about half of users when interacting with speech, and about one

third when interacting multimodally (speech and pointing at the same time), said words

like “one” or “the first” to select it, instead of reading what was written on it. While with

buttons labelled with content like “kitchen” or “football” this did not happen once, and

users always said what was written on the button. This also suggests a tendency that should

be taken into consideration when labelling buttons or other UI elements.

VC and Interaction Preferences: When asked about VC preferences participants

divided themselves between the three possible avatar options (head-only, full-body and

half-body). More relevant was one participant observation regarding how the chosen option

would depend on the application and interaction contexts. Regarding the relevance of

VC presence, participants largely preferred information presented by it than this being

presented only in text (53% vs 18%) or with voice output (47% vs 30%). Still, and again,

about 12% referred these preferences would also depend on the interaction context.

Summarizing the findings, it was evident that alternative ways of interacting with the

TV - if simple enough for not requiring too much learning - were well received (and can

be preferred to the traditional RC interaction). However, any type of interaction should not

be imposed on the users, but be available as an intuitive option for interacting with the TV.

Additionally, when not used intuitively, modalities of interaction should be explained to the

user before he or she starts using the system. But above all, these trials showed older adults

express a clear preference for multimodal interaction in both input and output contexts.

They also enforce the need for a multimodal system to be complemented with adaptation

mechanisms (to compensate declines and differences between older adults individuals).


3.4 Adaptive Interaction User Study

The second iteration of prototyping and user trials was performed with the main goal to

understand older adults interaction with realistic TV-based applications further and gather

further knowledge related with how adaptation and multimodal mechanisms can facilitate

the access to these type of applications.

In detail, our objectives included validating the User Initialization Application (UIA) -

which had its development begun much sooner at the requirement gathering stage - and

its acceptance by the end-users. We wanted to assess if the users understood the role of

the UIA and if they would use it to benefit from the adaptations it may provide. Also,

about adaptation, we wanted to assess if the users were able to perceive it and, most of

all, acknowledge the suitability of the adaptations to their needs. Therefore, in this stage

was performed a validation of all the requirements identified in the previous stages of this

UCD approach, related with adaptation and new ways of providing multimodal interaction

to older adults through modern TV contexts. Knowledge gained in the initial study and

the stakeholder’s research phase was used to develop an EPG application focusing on

the engagement between users and a realistic TV interaction scenario. Ultimately, we

wanted to verify the objective benefits of the usage of an adapted EPG over its non-adapted

counterpart.

All verified requirements are then presented as guidelines in the following section of

this chapter.

3.4.1 Prototyping User Initialization Application

The UIA was developed with the goal of being an introductory application in GUIDE,

a tool which runs only the first time a user initialises the system. When a new user is

recognised, the UIA is initialised by presenting a step-by-step introduction to the system

and how to interact using the different modalities. Another purpose of the application is

to expedite the user profiling procedure. As it would not be feasible to ask each user of

the final product to complete an extensive survey before using a system like this, the UIA

presents a much-reduced set of questions and tasks to the user to allow the user modelling

component to assign the user to one of the previously created profiles. Even though this

does not allow for a profile perfectly fitted to the user, it makes a good starting point for

adaptation processes as the profiles emerging from the stakeholders’ research stage already

considered a large pool of representative users. Additional information collected during

system usage later refines these individual profiles. Additionally, not only it works as a

tutorial, but it also collects information about the user characteristics, impairments and

preferences. It uses that information to assign a profile and personalise both the UI and the

interaction.

UIA development begun during the Stakeholders research stage (as it is easily verifiable


Figure 3.4: Screenshots of the UIA Flash-based (low fidelity) prototype

by the adaptation requirements drawn at this stage), with a first version being implemented

as a demonstration application for purposes of discussion with both developers and older

adults (figure 3.4). For that purpose, it was based on a Flash mock-up serving as a preview

for what type of tests could be included, and to simulate the general behaviour of the

application. As a proof of concept, this Flash mock-up was seen as a low fidelity prototype

from the beginning because it only made possible to interact with a mouse and was

constituted only by ten different tests. It made use of an independent module of a VC to

present the UIA goals to all interested parts.

Following this first low fidelity prototype, several other versions were implemented

with the goal of being as realistic and as close to a final product as possible. The second

version of UIA was developed with the goal of being heavily tested in the first iteration

of user trials with older adults. For this to be possible, all UIA-related development

decisions and goals, resulted from findings of the previous requirement gathering stages -

the need for having the capability of acquiring information from each older adult in order

to initiate an adaptation process, and the notion that older adults have the need of learning

new interaction modalities in a TV-based context. In the same way, UIA-related features

resulted from the discussion with developers in the previous requirement stage, while the

basic profiles resulted from the clustering process also described in the previous stage. The

development was first based on a new mock-up design (left side of figure 3.5) which guided

the development of a Web application (HTML, JQuery, CSS and JavaScript) composed

of simple pages with buttons, slide menus and the same virtual character present in UTA

(right side of figure 3.5). Thus, after designing the mock-up, the metrics to be present in the

application were defined with the help of people with expertise in the areas of psychology

and gerontology. These could be summarised to this set of tests/screens:

• Modality Introduction: Self-explanatory videos of how to interact with each modal-

ity, followed by “do-it-yourself” tasks;


Figure 3.5: Screen-shots of the UIA high fidelity mockup (left) and UIA high fidelity

prototype (right).

• Button and Menu Personalization: Button size and font, and background color

configuration;

• Cursor Personalization: Cursor size, shape and color configuration;

• Audio Perception: Hearing capabilities and preferences.

Additionally, some standardised measurements were adopted to tackle the more relevant

features: for colour blindness tests, plates 16 and 17 of Ishihara test (Clark, 1924); and

for dexterity, both grip strength and active range of wrist motion were estimated from age,

genre and height of users following ergonomics research (Angst et al., 2010).

In general, a very simple UI was considered for UIA, with a different screen for every

task and metrics identified. Few buttons presented per screen (to keep simplicity and

prevent user confusion). Every screen preserved the same navigation model - an area with

“next”, “previous” and “repeat” buttons, and another visually distinctive area for presenting

information and requests. For every metric to be measured, tests were presented as simple

games or questions about preferences. As it was already briefly described, during several

screens, there was a VC present, responsible for reading to the user every question and

instructions on how to interact and select the different options available. Still, as we were

not responsible for the development of this component, not much further details or results

will be presented concerning this. Finally, other results obtained while testing both the

UTA and the UIA also informed the refining of UIA’s design (e.g. high contrast colours,

big, centred and well-spaced buttons, etc.). Throughout the lifespan of the project, the UIA

was iteratively updated until a final version was validated in the second user study iteration.

Figure 3.6 shows four distinct screens of the UIA final version).

3.4.2 Prototyping Electronic Program Guide Application

Similarly to the UIA, the knowledge gained in the first user study was used to develop an

EPG application focused on the engagement between users and a realistic TV interaction


Figure 3.6: Screen-shots of the UIA final prototype

scenario. By developing this application, the goal was to validate previously developed

notions about older users interaction. For instance, do users favour alternative and mul-

timodal ways of interaction with the TV or when they are in the presence of a realistic

application they prefer the traditional interaction devices, like an RC.

The development of this application started with mock-up designs (fig. 3.7), which were

then reviewed and approved by experts in the development of TV and STB applications.

Later, all elements were implemented using HTML, JQuery, CSS and JS languages

(fig. 3.8). Concerning the different modalities of interaction, pointing recognition used

the same components as the UTA, a speech recognition module was implemented using

Loquendo’s Automatic Speech Recognition, which was capable of recognising speech

commands and sending them via a socket connection. Additionally, a simplified RC (with

fewer buttons than the traditional ones) was designed and built. This also included a

gyroscopic sensing alternative which enabled the control of a pointer on the screen.

Regarding content the EPG had real information about channels and respective sched-

ules and shows, to confer more realism to the application. Adapted versions were developed

to fit distinct user profiles. Output modalities, including the VC, were made available and

selected according to the user profile. Finally, in terms of available actions, the EPG was

integrated with a live TV stream and made possible to switch between channels, navigate

menus using proper buttons on the UI, with alternate commands (speech), keys (RC up,

down and ok arrows), or by performing scrolling gestures with the hands. It also provided a

dialogue with full details on each show, as well as additional “fake” options like a “record”


Figure 3.7: Screen-shots of the EPG mock-ups

option.

3.4.3 User Study Procedure

Three distinct adapted versions of the EPG were developed. The assigning of each version

was dependent on the result of running the UIA. These adapted versions were distinct

regarding UI elements specification of button spacing, buttons font size and UI colours.

However, and as we can see from figure 3.9, only small augmentations were necessary (as

the non-adapted version of the EPG was developed already addressing accessibility issues).

Therefore, for each participant, a user profile and the correspondent EPG adapted version

was assigned based on the info collected by the UIA.

Regarding the UIA, there was the need to evaluate how understandable the application

was regarding its goals and the instructions it provided. Additionally, we also wanted to

understand how easy it was for older adults to interact with it, and if they would still use it

if it were part of their daily lives. For this, not only quantitative data was retrieved from

UIA (user profile and UI preferences), but a satisfaction questionnaire was also performed

to assess the participants understanding and acceptance of the application.

Regarding the use of the EPG prototypes, the main goal was to assess the use of

multimodal interaction. In particular, the aim was to find which modalities were the most

used by older adults in a realistic TV interaction scenario. Additionally, there was also


Figure 3.8: Screen-shots of the EPG high fidelity mock-ups

a focus on adaptation, mainly in understanding if users perceived it and if they were

more satisfied using UIs adapted to their characteristics than non-adapted versions of

the same application. Every user interaction was recorded in a video with log files with

every interaction also being saved. Extensive analysis was performed based on both data

and observation. Given the semi-supervised methodology, and the fact that when first

interacting with novel technology several doubts and conversations can occur between

the participants and the supervisors, objective comparisons were not performed (e.g. task

times, task errors).

The evaluation session, for each participant, comprised three distinct phases: (1)

Running the UIA and going through all its tasks; (2) performing tasks (changing a channel,

activating subtitles or just going through the schedule of an entire day for a specific channel)

with the adapted EPG version; and, (3) performing the same tasks with the Non-Adapted

EPG version. Therefore, a within-subject design was used, where every participant ran all

three prototypes. The order in which they used both versions of the EPG was randomised to

counteract learning effects. Participants were briefed about the project and the applications

regarding its goals and the tasks to perform. Given the specific character of older adults,

traditional usability unaided tasks were sometimes unfeasible. Therefore, all help requests


Figure 3.9: Similarities between all adapted versions of the EPG.

and interventions were logged and taken into consideration in the analysis. A total of

40 older adults (24 female and seven male) with age-related disabilities. Recruitment

was performed in two countries, with 21 participants (14 female and seven male) being

recruited in Spain and 19 participants (10 female and nine male) in the UK. The average

age was 70.9 years old, and the different user profiles were assigned to the participants in

the following manner: 16 participants with profile A, 20 with profile B, and 4 with profile

C. Participation were volunteers, and all activities involved in this study were safeguarded

from the ethical point of view. Efforts were focused on creating a similar environment and

technical conditions in both labs. Trials were conducted by usability experts. Participants

were given freedom to interact (the trial conductor would only intervene when needed, or

participant asked for help).

3.4.4 User Study Results

Discovering Older Adults Profiles with UIA. All participants in the study performed

both a pre-survey and ran the UIA. 29 out of 40 profile assessments were performed

similarly by the two methods (representing 74%). The interrater reliability between the

profiles assigned with the pre-survey and the UIA was found to be Kappa = 0.58 (p<0.001)

revealing a moderate agreement. It is relevant to notice that the UIA enables users to input

preference values, something that goes beyond ability profiling. This is likely to explain


Table 3.2: Subjective ratings of the UIA.

Questions about the UIA Median IQR

1. Have you understood the purposed of UIA? [1- Yes; 2- No] 1 0

2. If you had the system at home, would you go through UIA or

would you skip it? [1- Would do it; 2- Would skip it]1 1

3. Do you think the UIA process takes too long? [1- Yes; 2 -

Neutral; 3- No]3 0

4. Were the instructions easy enough to understand? [1- Yes; 2-

No]1 0

5. Did you noticed any changes in the application while you were

using it? [1- Yes; 2- No]1 1

part of the mismatch (e.g. a user with no visual impairments is likely to prefer a higher

contrast button when is confronted with such a hypothesis). Another source of uncertainty

may be the under-statements made by some of the participants in the pre-survey. Still,

the 74% reflect the UIA’s capacity to understand user characteristics and assign them to

a specific user profile that reflects those specificities. From this, we acknowledge that

adapted TV applications based on simple initialisation profiling are feasible and likely to

improve over traditional methodologies.

UIA evaluation by the Older Adults. Participants were not used to using something

like UIA, so it was essential to assess how they see this approach and if they are willing to

use such a tool to improve their performance. They took between 12 and 37 minutes to

complete the UIA (Avg=22.8, SD=5.9). Although they were discouraged from engaging in

long conversations, participants were free to express their opinions and doubts during UIA

usage, which certainly increased the time to finalise the process.

Table 3.2 presents the subjective ratings given by all participants to the questions posed

in the evaluation questionnaire. Regarding understanding the UIA purpose (question 1), 9

out of 40 (22%) did not understand it. This indicates that such a process could be better

motivated or else the users will likely ignore it. In line with this, 11 out of 40 (28%) stated

they would skip the process if they had it as a system at home (question 2). 5 participants

stated to find the process too long while four others were neutral about it (question 3).

All the remaining thought it was neither too long nor tiring. Most users (35) felt the UIA

was easy to follow and understand (question 4). Regarding the adaptations felt during the

process (question 5), 26 participants stated to have noticed them. This is easily explained

as 16 were classified as profile A, which means they had few adaptations done during the

UIA.

In sum, the participants seem positive towards the UIA, and even though some could

find it discouraging, the right motivation allied with the fact the application should only be

run once easily supports its use. This is also supported by the overall understanding of UIA

purpose and by the fact that the vast majority of participants considered it easy to follow.


Evaluating multimodal interaction Using the adapted versions of the EPG, partici-

pants could perform tasks with any of the available modalities of interaction. After each

task, they were asked about their preferences. Figure 3.10 presents their preference count

after each task. Overall, it shows that both Speech and Tablet interaction was seen as

positive improvements for interacting with the EPG. Conversely, the standard RC was seen

by part of the participants as a safer option which they are not willing to trade easily. The

gyroscopic remote was mildly used while pointing at the screen was the least preferred

option. The participants were also able to use Pointing and Speech together, but this

option was not revealed as interesting enough. This might be related to the complexity

of using more than one modality and the reduced timespan of training. Further, is likely

the simplicity of the tasks was more appropriate for simpler selection approaches while a

multimodal option could have been directed at more complex tasks. Nonetheless, given

the population and the overall goal of simplifying processes and the trends mentioned

above, it seems reasonable to suggest that a variety of unimodal interfaces seems to be

the adequate fit to an older adult needs and preferences. Interestingly enough, when the

available modalities were explained as being used with the adapted versions of the EPG,

most participants state they would stick with the standard RC. However, when looking

closer at the order of the trials, those that run the adapted version first seemed to be eager

to try out the different modalities. Particularly the gyroscopic RC, the tablet and speech.

While those that performed the tasks previously with the non-adapted version tended to use

a more conservative approach in the adapted setting, giving preference to Speech followed

by the standard RC and the tablet, and ignoring the remaining.

In both scenarios, about half of the participants shifted their preference during the

adapted trials while the other half stuck with their initial preference (first task). Preferences

also diverged between users from different profiles. Participants from Profile A (no

adaptation) showed a preference for both the standard RC and speech and mild preference

for using the tablet. All the other modalities were ignored. As to Profile B (medium

adaptation level), participants showed a less consistent selection. The tablet was the most

preferred modality, but all the other modalities were seen as useful. As to Profile C (higher

adaptation level), consistency was once again revealed with preferences going for both RC

modes, pointing and speech.

In summary, results suggest that older adults do not reject the use of alternative

modalities of interaction if these are adapted to their profiles and the tasks at hand. Further,

age did not show a significant correlation with acceptance nor any observable tendency on

the set of preferred modalities. This suggests that this overall acceptance is pervasive to

the age groups in the sample.

Evaluating Adaptation. Subjective acceptance of both versions of the EPG was

evaluated through a statement to classify using a five-point Likert scale (1- the system has

not supported me at all; 2- the system has supported me only in some parts of the task;


Figure 3.10: Modality preferences by task

3- undecided; 4- the system has supported me in almost all the tasks; 5- the system has

supported me in every moment). A Wilcoxon Signed Rank Test revealed a minor difference

between the EPG adapted (Mode=4; IQR = 1) and non-adapted versions (Mode=3; IQR=2),

z=-1,665, p<0.1, with a medium effect size (r=0,37). Although not statistically significant,

subjective acceptance of adaptation showed to be slightly higher by participants in profile

C, suggesting the (more noticeable) adaptations improved the relationship between the

participants and the interface. Comparisons of acceptance regarding icon and text properties

(evaluated as a dichotomous response to the suitability of the property) in both adapted and

non-adapted settings revealed no significant differences. Apart from the larger number of

users with no adaptation, the baseline EPG was already designed as an accessible version.

This might have been the reason for the participants to rate the non-adapted EPG as adapted

to their needs as well.

3.5 Recommendations for the Design of TV Applications

targeting Older Adults

Throughout all the described stages of the work developed under the context of GUIDE,

results demonstrated clear preferences and findings which can be derived into specific

recommendations for the design of TV-based systems targeting the older segment of the

population. Even though several were identified in more than one stage of the described

UCD methodology, they are only presented once. Recommendations are divided into four

categories: interaction modalities, multimodality, adaptation, and design of the TV-based

application. These recommendations are now highlighted being also referred their support

of previous guidelines from Nielsen (1995) and Norman (2013).


3.5.1 Interaction Modalities

Several interaction modalities were used in the context of GUIDE, and results have con-

sistently provided findings in the same directions. First and foremost, and concerning the

traditional interaction modality that is the RC, findings have shown RC device should be

simplified so that older adults can understand it better or in alternative, interaction

should be based on a simpler use of the RC. Examples of the latter are given like sup-

porting interactions where only the directional buttons and the “ok” or “select” keys are

used (section 3.3). Additionally both recommendations are in line with Norman’s guide-

lines concerning “constraints” and more particularly concerning the need for simplification

of interaction possibilities (Norman, 2013).

When older adults are not capable of interacting with this modality, speech-based

interaction should be supported as the main alternative. That is if users do not have

specific impairments that make it harder to use speech of course (section 3.4). Additionally,

several other recommendations were drawn regarding speech, like the importance of

automatic speech recognition to consider the user and interaction context in order to

improve recognition, the need for UIs to avoid presenting unknown terms to the user as

it can result in speech recognition errors or misunderstanding due to misspelling, confusion

or lack of knowledge of technological terms, the need for systems to support the user

both in preventing and recovering from speech recognition errors by appropriately

presenting recognized commands or providing multimodal feedback, and the need for

systems to consider speech characteristics of older adults like pauses for recalling a

word or a sentence, voice tremors, etc. All these recommendations were referred on the

first described stage of this chapter (section 3.3). Additionally, they are also in line with

previous guidelines like the ones regarding input and output characteristics, adaptivity and

consistency of multimodal user interface design by Reeves et al. (2004) and Nielsen (1995)

guidelines on “match between system and the real world”, “user control and freedom” and

“help users recognize, diagnose and recover from errors”.

Additionally, to RC and speech, pointing interaction was also used in some TV-based

tasks. As a result, some recommendations can also be drawn. Findings from the first

described user-tests (section 3.3) showed that applications should always provide appro-

priate means of feedback to the user when interacting by pointing, clear examples of

this is to where the user is pointing when no object is selected, which UI elements are

being selected or which button is being pressed. This is further supported by findings

on section 3.3, which further suggested that changing button colour or augmenting the

action with audio feedback would be sufficient. Also related to this is the fact that ap-

plications should also minimize the risk of inadvertent activation of functionalities or

button selection, by unintended gestures, be it by confirming the proposed action or by

employing algorithms capable of detecting real intention. These recommendations are

also in line with previous guidelines regarding multimodal interaction, more specifically


findings by Stanney et al. (2004) concerning pointing interaction, findings by Reeves et al.

(2004) concerning adaptivity and consistency, and previous guidelines from Nielsen (1995)

concerning visibility and error prevention and Norman (2013) regarding feedback.

Moreover, and in the context of the user studies described in section 3.3 several other

recommendations can be drawn concerning pointing. Probably the most important is

the one related with the fact that to prevent the user from making unwanted selections,

each pointing selection must only be validated after the intention was maintained

for a given amount of seconds. However, unnecessary physical demands in pointing

should also be reduced to the minimum possible or the user will get tired quickly. In

this sense, making the amount of time needed for selection something dependent of (or

configurable by) each user, could be a solution. We’ve also observed that when users are

required to select something they tend to use their dominant hand, independently of where

the icons are placed. This should be taken into account by, when possible, placing icons

on the dominant-hand side of the screen, possibly increasing efficiency and decreasing

tiredness. All these recommendations are in line with Nielsen (1995) guidelines on “error

prevention”, more particularly guidelines regarding “preventing errors from happening”

and providing “confirmation options”.

Finally, when focusing on how a TV-based system should consider output, several

recommendations can also be extracted from the described user-studies. From previous

research (chapter 2) and when concerning voice feedback, we can easily identify that voice

output has to be clear and easy to understand. In a more general way, and confirming

previous guidelines from McGee et al. (1998) regarding multimodal feedback, findings

from section 3.3 showed the need to always ask for a confirmation whenever a message is

issued from the application to the user. At the same stage, it was also clear that interaction

mechanisms should also be capable of recognising if the user has received that message,

and when it was not, repeat the process (or change the modality involved) like defended by

Oviatt (2003).

3.5.2 Multimodality

Multimodal mechanisms have also been employed in great part of the user studies described

previously in this chapter. In the context of chapter 2 comes the general finding regard-

ing the necessity for providing modality help regardless of the interaction used. That

means that help and explanation for using any device or alternative option to the traditional

RC must be available at any time in the application/system. This is in line with Nielsen

(1995) description of “help and documentation”. From the same user-studies comes the

findings that were validated all through the consequent studies regarding the notion that

every task must be possible to achieve using any modality or device of interaction

available which supports Norman (2013) notion of “consistency”. Also to achieve these

findings, subsequent studies (section 3.3) also showed the need for the use of activation


procedures for both gesture and speech interaction. Therefore, to prevent errors occur-

ring when switching between distinct interaction modalities, specific modalities should be

activated by specific procedures (e.g. speech recognition should be activated by specific

keywords, gesture recognition should be activated by a specific gesture). Additionally,

preliminary findings also showed that multimodal systems should provide training steps

for the user, for each interaction modality, and for each interaction context. Finally,

and when focusing on using several modalities in a combined manner, findings obtained

through section 3.3 show that a multimodal system should support the recognition of

universal speech terms to simplify multimodal interaction. This means that when in-

teracting multimodally with an application, speech terms like “Select”, “This”, “Yes”, “No”

or “Confirm”, should be included in the recognition dictionary. These commands can be

used as confirmatory steps to select an option while the user is pointing at it. Additionally

terms related with numeric labels (“One”, “Two”, “Three”, or “First”, “Second”, “Third”),

should also represent keywords. These can be used to support redundancy when a user is

pointing and speaking, or when only making use of speech.

3.5.3 Adaptation

While not initially focusing on adaptation, work on multimodal interfaces lead us to some

findings which are related to how features should be personalised or adapted to each user

characteristics.

In that sense, when focusing on how a system should support user input with multimodal

features, it is especially relevant that each interaction should be personalised for or by

each user. This came from section 3.2 results, and the notion that when the user is

using the system for the first time, the application should ideally be adapted to the user

parameters (interaction preferences, colors, font size, button size, audio volume, etc.)

based on an initial set of configurations performed by the user. Also, as older adults

may experience changes in their characteristics and abilities because of the age-related

phenomenon, these parameters should also be re-configurable by the user at any time.

In this same sense, that initial set of configurations could be achieved by making use of

a User Initialization Application like the one described in section 3.2. This application

should also be capable of providing tutorials on new ways of interaction the first time a

user interacts with the system. If used, that application should also be as entertaining and

short as possible, and the user should also be offered the possibility to cancel (or skip) the

initialisation process at any time. These notions are in line with previous recommendations

by Reeves et al. (2004) regarding how user preferences should be considered across

modalities and with Nielsen (1995) guidelines on “flexibility and efficiency of use”.

Additionally, to further complement the first recommendation, the system should ideally

support automatic recognition when an older adult is in front of the TV, to adapt the

system to his/her specific needs (section 3.2). This could be done through recognition


of facial characteristics, voice, thumb, etc. Moreover, and still considering findings on

input adaptation, findings coming from section 3.2, also support the need for adaptation

to distinct user contexts. I.e. in different situations distinct methods should control the

application according to the older adult’s needs. For example, if a user usually controls

the application by voice but has a cold and has lost his/her voice, the system should

adapt either by suggesting other modalities of interaction or by ways of adapting the

interaction. An example of the latter was also suggested as a recommendation: when a

motor-impaired user selects a UI element through pointing, the area of selection should be

increased, ideally by making the surrounding area of the button also selectable. Finally,

when an older adult accidentally gives contradictory commands to the system while using

distinct modalities at the same time (e.g. saying “Channel 1” and mistakenly pointing at

“Channel 2”), only one of them should be recognised. Therefore, it must be possible for

the user to pre-establish a preferred modality that will prevail over any other in case

of contradiction (section 3.2).

When focusing on how the output should be adapted, several recommendations can

also be extracted from the described user-studies. From the first stage (section 3.3) we

could identify some older adults’ interest in customization regarding male and female

voice representation, as well as volume levels, to satisfy different user preferences or

impairments. In a more general way, findings from section 3.3 suggested that output

should always depend on the interaction context, for example when the user is watching

TV the feedback messages given back should be different from the ones rendered when the

user is, for example, looking at pictures on the TV. A solution could be to add mechanisms

that would support each user in selecting in which context he/she wants messages rendered

in what modality. Alternatively, the system should learn these patterns through user

interaction history. Findings on section 3.4 followed a similar line, with the notion

that a system like this should consider different ICT skills and adapt presentation

accordingly (e.g. people with better ICT skills might prefer complex menus which offer

more information while people with fewer ICT skills will probably prefer simple menus

with less information but easier to control). This type of pre-configuration could be

achieved with the help of a configuration application like the one presented previously as

UIA. In line with this, a TV-based system should also adapt presentation considering

distinct user impairments (section 3.2) (e.g. people with visual problems prefer bigger

buttons on the screen, while people with hearing problems prefer to have information

provided in both visual and audio forms). Still, while adaptation to screen layout (UI

elements) should be offered in modern TV systems targeting older adults, that adaptation

should only be applied when related with user impairments, or as a result of direct

user requests (or user explicit acceptance of proposed changes) (section 3.2).

When considering adaptation in a more general context, findings on section 3.2 also

showed that adaptation to older adults behaviors and habits should be provided (e.g.


adaptation to user’s daily schedule, or interaction habits) in order to favor their use of the

system, and that the same user profile (necessary for adaptation mechanisms) should be

maintained across multiple applications and situations whenever possible. Finally, and as a

helpful finding towards the development of TV applications targeting older adults’ we’ve

also identified (section 3.2) the main distinctive factors between older adults as being the

following: capability to read perfectly from close and distant vision; capability of seeing

at night, and color perception; capability to hear sounds of different frequencies and to

distinguish conversations in a noisy background; cognitive impairments; and mobility

diagnosis like muscular weakness and tremors. Therefore, when designing for these type

of users, distinct profiles should be created focusing on these differences.

3.5.4 TV-based UI

When focusing specifically on the development of TV-based UI targeting older adults,

several other recommendations can be extracted from all the user studies described in this

chapter.

Starting off, any graphical UI design targeting older adults should be based on a a

simple layout and a limited amount of functionalities. Additionally, instructions should

be “easy to understand”, and icons should be perceived as “intuitive” (chapter 2). In line

with this, following user-studies supported both the notion that UI should be designed

to be clear, simple and with uncluttered screens focusing on big and well-spaced

buttons (section 3.2 and section 3.3) and the notion of simplification of novel features

(section 3.2). Both these notions are in line with Norman (2013) definition of “visibility”.

Every time a user wants to perform a task he/she is not familiar with, the number of steps

required should be as few as possible, and additional feedback or help from the output

components is highly recommended. Building on this need for simplification, findings

on section 3.2 have also shown that ease of use is more important than branding or

UI appearance and that user interaction must be favoured over anything else. Also in

line with this recommendation, we also identified the need for the most important or

most used functionalities to be accessible via simple interaction schemes (or shortcuts)

like “one-click” / “one-touch” / “one-command” / “one-key” interactions. So, TV-based

applications should look to provide one-touch features for most used functionalities

to simplify older adults interaction. Once again, this recommendation confirms previous

recommendations by Reeves et al. (2004) concerning multimodal consistency, and more

specifically the importance of enabling shortcuts as well as Nielsen (1995) notion of

allowing users to tailor frequent actions to provide flexibility and efficiency of use. Still

and while talking about simplicity, it is also important to keep the UI simplified, but not

senior (section 3.2). This means that TV-based UIs must support older adults’ needs with

a simplified look-and-feel but should not appear to be designed only for their needs or give

the impression they have been designed for someone with impairments or they the risk not


being used by them.

Advancing to feedback importance, most of the times and throughout all described

user studies (with particular incidence on section 3.2), participants asked for additional

feedback to any input task they’ve performed (e.g. a beep, a graphical highlight, a

vibration), so we should assume that any TV-based application or interaction should

support this. Additionally, we also identified the need for feedback messages that catch

visual attention (section 3.3) so that older adults can clearly differentiate system messages

from the ones that usually appear on the TV (broadcast). These recommendations support

Nielsen (1995) guidelines on “visibility of system status” and more particularly on how

the system should always keep users informed through appropriate feedback.

Moreover, and further contributing to the notion of the importance of RC interaction,

findings resulting from section 3.2 also advanced the importance of keeping well-known

TV concepts. As older adults have problems to adapt to new paradigms in interaction, con-

cepts related to traditional RC interaction and TV-based application paradigms should be

kept whenever possible. The same findings also state the consistency of user experiences

across platforms is also important so that when switching from one platform/device to

another and while consuming the same service, older adults should be able to keep the

same interaction paradigms (or the ones they are already familiar with). This recommenda-

tion is further supported by findings in final user studies (section 3.4) which showed that

interaction paradigms should be maintained across all tasks, if not, typical “select and

confirm” tasks usually can result in blocking situations. Additionally, this recommendation

is also in line with Norman (2013) definition of “consistency”.

Finally, and especially relevant to the following chapters, in the preliminary findings

(chapter 2) we have also seen that older adults applications should be designed to avoid

loneliness and foster social inclusion. In fact, older adults identified the need to commu-

nicate and share experiences with friends and relatives, as the main reason for using a

TV-based application (rather than watching television).

3.6 Summary

In this chapter, we presented all requirement gathering steps and user studies performed

with older adults concerning the development of TV-based applications targeted at them.

As a result, a list of recommendations is presented falling within the scopes of interaction

modalities, multimodality, adaptation and UI inclusive design. These recommendations

provided the first basis for the development of SNS prototypes developed in the context of

this thesis, and which will be detailed in the next chapters.

In a general manner, the results obtained are an indication that when designing for

this segment of the population, new types of interaction modalities should be supported

in order to fit their full range of characteristics. However results also show that for most


older adults, and whenever possible, RC interaction should be maintained as the main

modality, while speech and pointing are seen as good alternatives. The results also echo

for the need to know the users before making use of any adaptation or personalisation

mechanisms (which are welcomed), and in that sense, they present a type of application

(the UIA) which could perform that profiling mechanism while introducing every modality

to users. This chapter also reinforces the benefits of resorting to UCD approach when

designing systems centred on older adults.

On top of these recommendations, it is also necessary to understand in which ways

TV-based applications could be designed not only to be inclusive but also to target social

features. The next chapter focus on this subject.


Chapter 4

Drawing Recommendations for the

Design of a Social Network Service

targeting Older Adults

This chapter presents the steps followed in this thesis for drawing recommendations

targeting the appropriate design of an SNS which considers the characteristics of the older

segment of the population. It focuses on both literature and user inquiring techniques.

Thus, this chapter starts with a description of the methodology that integrates these steps. It

follows with an extensive review of the related literature surrounding the emergent field of

SNS use by older adults, encompassing conference and journal papers as well as research

projects, to lay the ground for a critical analysis of this technology. It continues with a

description of a set of online questionnaires, focus groups and semi-structured interviews

performed at two distinct older adult institutions. The chapter finalises with the formulation

of all the relevant recommendations generated by all these steps.

4.1 Methodology

One of the main goals of this thesis is to understand how to draw SNS (or Facebook) closer

to older adults so they can take advantage of its features as a means of increasing social

interactions. To be able to do that, we must understand the relationship between older

adults and SNS, and the technology that might support SNS dissemination.

With this in mind, and taking the related work described in the second chapter as

the basis, we saw the necessity of looking deeper into how SNS have been considered

or designed in the past and how this influenced older adults interaction. Therefore, as a

preliminary step, we performed a literature review (described in detail in section 4.2), which

made possible to identify the most relevant domains surrounding the subjects mentioned

above and to focus sequential steps into gathering older adults opinions and requirements

on the most influential subjects. After gathering literature requirements, there was also the

need to directly inquire older adults about all possible domains involved in the adoption

67

Chapter 4. Drawing Recommendations for the Design of a Social Network Service

targeting Older Adults 68

(or lack of) and use (or no use) of SNS. With this goal in mind, we conducted a series

of user studies (described in detail in section 4.3), following a mixed methods approach,

combining questionnaires, focus groups and semi-structured interviews. Our preference

for these approaches was again (like in work described throughout the previous chapter)

based on previous studies which have demonstrated the advantages of data triangulation

techniques (Taylor and Bogdan, 1984; Langdon et al., 2003).

Thus, this user requirement gathering phase divided itself into two phases. First, to

pool options regarding both SNS, Facebook and technology contexts, we conducted ques-

tionnaires (both online and off-line). This decision was supported by the goal of collecting

data from the broader older adult population possible (Ackroyd, 1992). Questionnaires

were performed with individuals living at home, older adults living in senior health-care

facilities, and participants which can be considered more active and were at that time senior

university students. To note here the fact that in this step we did not design questionnaires

to focus specifically on the domains identified in the previous phase. This decision was re-

lated to the fact that when first confronting older adults we did not want to restrict any data

that could be excluded by doing so. At the same time, this first user requirement procedure

could also work as a confirmation of the relevance regarding the domains identified in the

literature phase. These questionnaires are described in detail in section 4.3.1. Secondly, and

to make possible to collect richer data and foster discussion between individuals (Adams

and Cox, 2008), we conducted focus groups and paired interviews with older adults on two

distinct subjects, SNS (described in section 4.3.2) and TV (described in section 4.3.3. For

both, we focused on both senior health-care facilities and universities. In these, discussions

were mainly focused around the major domains identified in the literature requirement

gathering phase.

4.2 Gathering Literature Requirements

The need to analyse past work on the use of social technology by older adults is a necessary

step to understand the main characteristics, and the ”whys” and ”hows” it can be used

and designed to suit this population better. This section starts with a description of the

whole literature analysis process followed to achieve this goal. It follows with three

subsections which divide research in three contexts: (1) older adults’ most distinctive

social experiments (based on the use of applications with social purposes); (2) studies

related with the use of standard SNS like Facebook; and (3) research that investigated the

main reasons why SNS targeted specifically at this population do not fulfill their goals.

This clear division is justified by the way researchers in each section had distinct goals.

While researchers in the first context were more concerned with small and specific social

problems related with older adults communication, researchers in the second context were

concerned with the understanding of what is missing from typical SNS to be used by this



population, and researchers in the last context focused only on the older population without

even considering the younger generations as part of SNS (or as the imperative force behind

its use by the older population). For each of these, articles are presented in a chronological

order to give a full overview of how the research field has evolved. Finally, the section

finalises with a discussion where the most relevant dimensions present in all these works

are identified.

4.2.1 Procedure

Many mainstream journals and conferences in the area of human factors, ergonomics, and

HCI, as well as many specialist journals and conferences on ageing, disability and technol-

ogy, were selected for this survey. These quality conferences and journals were selected

based on their impact factor Thomson Reuters (2013) and rankings by the Australian Re-

search Council’s ranking of journals and conferences Australian Research Council (2010).

Conferences which, although not featuring on these rankings, are especially relevant to the

field - like ASSETS - were also considered. Additionally, papers originating from other

sources were selected based on the fact that papers in the main sources strongly cited them.

Table 4.1 presents the sources from where papers were selected for analysis.

A total of 7384 papers from 11 distinct sources were analysed in the literature search

and review. Of these papers, 31 were selected for the survey which represents less than

0.5% of the entire pool of selection. Additionally, ten more papers were selected from

additional sources as they were heavily cited in the selected pool of papers (table 4.2).

To give a full understanding of how these 41 relevant papers were selected we now

explain in detail each of the five steps that composed the review process:

1. Venue selection: Taking into consideration their impact factor and rankings, a list of

relevant conferences and journals were selected. Determinant factors were related to

how the conferences and publications focus specifically on older adults, on aspects

of Human-Computer Interaction, and on the social aspects of technology. Sources

related to psychological and social aspects of old age which were not related to

technology tended not to be considered. Additionally, this literature review focused

especially on the last seven years of research, so the majority of sources analysed

belonged to this time interval. The list of conferences and journals are presented

in table 4.1.

2. Keyword-based pre-selection of papers: For each conference and journal selected, a

pool of papers for analysis was built from the ones featuring specific keywords in

the title from four specific categories: terms related with (1) the older segment of

the population like “elderly”, “older adults”, and “seniors”; terms related with (2)

older adults characteristics like “vision impaired”, “cognitive decline”, “dementia”,

etc.; terms related with (3) SNS and social applications like “social network”,



Table 4.1: Journals and Conferences included in the survey

Mainstream journal or conference Years Analyzed Selected

ACM Conference on Human Factors in

Computing Systems (CHI)

2005 - 2015 3178 13

ACM SIGACCESS International Confer-

ence on Computers and Accessibility (AS-

SETS)

2011 - 2014 108 1

ACM Transactions on Computer Human

Interaction (TOCHI)

2012- 2015 112 1

ACM Transactions on Accessible Comput-

ing (TACCESS)

2012 - 2015 27 0

British Computer Society Interaction Spe-

cialist Group Conference (BCS HCI)

2009 - 2015 348 3

Computers in Human Behavior 2009 - 2015 2119 4

Conference on Human-Computer Interac-

tion (INTERACT)

2011 - 2015 620 3

Human Computer Interaction 2012 - 2015 57 0

International Journal of Human-Computer

Studies

2009 - 2015 408 3

International Conference on Intelligent

User Interfaces (IUI)

2011 - 2015 271 0

Universal Access in The Information So-

ciety

2012 - 2015 136 3

Total 7384 31

“social network services”, “social services”, “social applications”, “SNS”, etc.; and

terms related with (4) issues like isolation and social skills like “social isolation”,

“loneliness”, “social decline”, “communication”, “keeping in touch”, etc.

3. Abstract-based pre-selection of papers: From the preselected papers, each abstract

was analysed to gain a general overview of each work. Taking into consideration the

way the abstract focused on the general categories of the study (identified on point

2), papers were selected for the next step. The ones which the abstract gave a clear

idea of not being in the scope of the study were filtered out.

4. A priori coding: All selected papers were read and coded following a a priori ap-

proach (Weber, 1990) and taking into consideration the categories and the keywords

identified in point 2. This process allowed to filter out papers which approached the

main subjects in a too superficial manner and at the same time enabled the placement

of each relevant paper into one of the three categories previously identified.

5. Emergent coding: Papers were annotated and classified using emergent coding (Miles

and Huberman, 1994) from which, as a result of inductive reasoning, several do-



Table 4.2: Additional relevant Journals and Conferences included in the survey

Additionally relevant journal or confer-

ence

Years Analyzed Selected

Australian Computer-Human Interaction

Conference (OzCHI)

2012 1 1

Behavior & Information Technology 1993 1 1

Educational Gerontology 2012 1 1

IEEE Internet Computing 2005 1 1

Journal of Community Informatics 2012 1 1

New Media & Society 2003 1 1

Nordic Conference on Human-Computer

Interaction (NordiCHI)

2010 1 1

Personal Ubiquitous Computing 2007 1 1

Transactions on Information Technology

and Biomedicine

2010 1 1

XRDS Crossroads ACM 2007 1 1

Total 10 10

mains resulted (each of these domains is individually explored more ahead in this

document). As an example of this coding process, we have the domain related

with ‘family” (and how both SNS and social applications revolve around family

members and family relationships) which was not considered as a keyword in the

previous steps and was identified as the most relevant domain at this step. The most

relevant papers, i.e. the ones which spanned over several domains, or the ones who

introduced relevant domains, were selected as papers to include in the survey. A total

of thirty-one papers were selected using this process, and the way they span over the

selected sources can also be consulted on table 4.1 by giving special attention to the

“selected” column.

6. Additional papers: As a result of point 5, some papers cited by the selected papers

were also identified as relevant and went through the same reading and coding

steps. In total, an additional ten papers were selected, and their sources are listed on

table 4.2.

4.2.2 Social Applications and Older Adults

Although not being full SNS, social applications developed in the most recent years,

have focused on characteristics typically present in this type of services. Their review is

crucial to capacitate the development of SNS entirely understandable and accessible to

the older segment of the population. Additionally, their use in many cases constituted the

first experiment with social technology, and that itself can give insights into how these

technologies should be designed having older adults in mind.



The first relevant social application targeting older adults was Brunette et al. (2005)’s

Meeteetse application, which focused on place attachment to connect older adults in the

same community. They wanted to strengthen shared identity into a community centre

(through a location-aware digital camera and a large public display) and create a tangible

way of communication at older adults’ home (through a touch-screen scheduling device

and a digital picture frame). Photos were automatically uploaded and associated with

events and seniors could share slideshows on the public display at the community centre,

and vote “thumbs up” or “thumbs down” on using an RC to determine how long they are

kept in rotation. Problems with taking or deleting photos and navigation issues created the

necessity for animated transitions, visual cues and colour coding to give users a visceral

sense of change between pages. Additionally, the mechanisms for sharing experiences

were too basic. Allowing community members to share stories in a more expansive

fashion (Izadi et al., 2003), and the possibility of making use of a messaging portal would

make Meeteetse more interactive, inclusive and close to an SNS.

Evolving towards a family approach, Romero et al. (2007)’s ASTRA application

focused on providing context for conversation within a household by using photos. It

served as a foundation for a breath of subsequent research into situated displays, as the

work of delivering specific objects and the appropriation of messaging devices can be seen

as ways of sustaining family members relationships.

After this work, several other projects followed the same goal of establishing better

ways of communication between older adults and family members: Lindley et al. (2009)

and Vetere et al. (2009) work focused on sharing of photos between older adults and

relatives, while Raffle et al. (2010) and Vutborg et al. (2010) prototypes focused on

storytelling between grandparents and grandsons living apart.

Lindley et al. (2009) developed two social applications: ShoddyPop, an asynchronous

email server which allowed for the recipient to take his time before replying or recip-

rocating the attention; and PersonCards an application which applied Brunettes’ early

work (Brunette et al., 2005) and allowed for lightweight information to be sent to a situated

display in an older adult’s home. In these, they made use of diverse channels for sending

messages. Recipients could also reply with handwritten messages for a more personal

and easy to produce a form of contact. With both applications they promoted use cases

for a number of design implications: (1) contact should allow for a level of intimacy that

is personalized – e.g., the use of “personal touches”, like the tone of someone’s voice or

obtained through the recognition of a loved one’s handwriting – thus supporting a level of

expression that is lost with text messages and even in email; (2) technology should allow

for more focused, intense communication, as older adults are prepared to devote time to

activities – e.g., talking on the phone or writing letters – that demand thoughtfulness and

skill, therefore, while time is dedicated to contact, this contact should be non-intrusive; and

(3) design should support reciprocity, because while older adults appreciate the potential



for new technologies to offer insights into the lives of their loved ones, they want to be

able to give something back in return.

Vetere et al. (2009) developed the Collage, a technology probe to allow the sharing of

everyday experiences, story-telling, family history and the creation of new experiences.

They designed the system to provide asynchronous exchanges of photographs and text-

messages between smart-phones and medium-sized touch-screens located in prominent

areas. Users could stop, drag and resize items and actions would reflect on both screens

(WYSIWYG alike). Although only exploratory, its use showed electronic photographs and

text became both the subject and object of intergeneration playful encounters. It had an

ambiguous character with its use being both sanctioned and celebrated. It involved the

capability of maintaining a meaningful relationship and at the same time the feeling of

familial obligation. Lastly, although authors evidenced that older adults’ possible physical

impairments should be one of the main points of focus, these were not the in fact considered

in their work.

Raffle et al. (2010) took inspiration from Romero’s (Romero et al., 2007) and Lind-

ley’s (Lindley et al., 2009) previous work and created a custom-made device to be used by

grandparents and grandchildren for a shared reading of physical storybooks. It included

an audio channel and page sensing technology to determine if the child was on the same

physical page. The evaluation showed that it not only made children more engaged in

long-distance communication than when they used other technology, like Skype, but also

that the quality of the intergenerational interactions improved as well. Their work also

reinforced the value of intergenerational play first evidenced by Vetere et al. (2009).

Similar to Raffle et al. (Raffle et al., 2010), Vutborg et al. (2010) implemented and

evaluated a technology to provide adequate communication between grandparents and

grandchildren. It allowed the sharing of virtual objects, through a shared display. It

consisted of two LCD monitors and two computers, one set for each household. It followed

the WYSIWIS concept and was equipped with an audio channel through which users could

talk to each other, as on a regular phone. Their work highlighted four essential themes

for the design of a system with social goals and focusing on older adults: conversational

context (to have something to talk about), facilitation (to be given the opportunity to talk),

supporting grandparent caring for grandchild, and diversified interaction forms (to maintain

attention of different users with distinct characteristics). The latter theme was regarded as

the most important and employed by following systems focused on this segment of the

population.

Still related with solutions regarding interaction with family, both Giorgi et al. (2011)

Life Frame framework and Gaver et al. (2011) device employed the use of digital mementos

and photos as a tool for increasing social interaction, while Judge et al. (2011) prototype

focused on the inclusion of a “personal touch” on the communication between older adults

and relatives.



Giorgi et al. (2011) Life Frame was composed of multiple interactive digital frames

distributed in different places around the house. Each one worked like a tangible object

with augmented services added to physical mementos for memory preservation and en-

hancement. They concluded these technologies could support specific needs connected

with ageing: the need for safety and the need for company. However, the process of appro-

priation was slow and relied closely on mediation by family members or close friends, and

on the capability of suggesting functions, they are familiar with. Still, the added value of

generating opportunities for real encounters and relationships was seen as fundamental for

satisfying the older adults’ need for company.

The Photostroller (Gaver et al., 2011) was a device designed for use by residents

of a care home for older people which showed a continuous slide-show of photographs

retrieved from the Flickr image website using a set of six predefined categories modified

by a tunable degree of ‘semantic drift’. It was composed by a large screen, a removable

control unit and wheeled legs to allow it to be used both in large groups (dinner room)

or more intimate sessions (bedrooms). It was well received because it allowed elderlies’

memories and conversations to drift along with the imagery serving as a tool for sociability

and reminiscence (and mentally engaging). Older adults saw it as a simplified version of a

TV with a clear and recognisable purpose and enjoyed the interaction it offered. It was

novel because it only made use of photos to make older adults engage and socialise and

confirmed the role of alternative modalities of interaction (like TV) for social engagement.

Judge et al. (2011) showed how a multifamily media space called Family Portals

increased feelings of connectedness and awareness. They based their system on an always-

on approach that provided a continuous video connection between the households and

supported asynchronous mobile (around the home) interaction by touch-writing on a Tablet

PC. The dedicated area for writing resulted in increased information sharing between family

members, and the combination with video improved domestic awareness. Additionally, the

device also increased communication outside the family circle and activity participation

in each other’s homes. Their results also showed differences regarding awareness and

communication needs between distinct family members and the necessity for adaptation

features to tackle these.

Following these works, Lindley (2012) also focused on asynchronous communication

between older adults and relatives, while, taking a distinct approach, Garattini et al. (2012)

work presented the development of a device which made use of a telephone and several

input modalities to support social interaction among older adults.

Confirming their previous indications (Lindley et al., 2009) that older adults express a

preference for heavyweight interaction, Lindley (2012) presented findings from a field trial

in which three generations of a family were linked through a situated messaging device

called Wayve, which, while designed to support lightweight messaging, also afforded

rich and expressive contact. The main features to be acknowledged as necessary by the



participants included the possibility of one-click message sending to a set of favourite

contacts, the asynchronous nature of the system which made it possible for them to work-

around the schedules of their relatives, or the ability to scribble messages. Wayve was

perceived as very easy to use. However, older adults tended to use mobile phones only for

talking instead of text messaging or sending picture messages. Older adults also valued the

fact that Wayve was a closed network, only available to their family as it gave them a sense

of privacy. Lastly, they tended to highlight the possibility of using alternative modalities,

which could better meet their communication habits, like speech interaction or other sorts

of multimodal approaches.

Garattini et al. (2012), through the Building Bridges project, developed a device

consisting of a 12-inch touch screen computer and a phone handset with cradle and

speakers. This device allowed older adults to listen to regular broadcasts and supported

both audio and written exchanges between each other and their contacts. This provided

further understanding of how this segment of the population engages through technology:

(1) contact with strangers requires a certain amount of effort and, although it is possible

with time, very few are interested; (2) informative and educational content acts as an

effective channel for encouraging social interaction; (3) usage can be influenced by gender

and levels of loneliness but not by prior technological experience; (4) there are concerns

about privacy and personal identity related with being seen on a web-cam and the necessity

of having control over who can see what; and (5) the necessity for a system like this to

work in conjunction with meeting face-to-face. Thus, Garattini et al. confirmed indications

related with the importance of the offline role and introduced new factors such as the need

for privacy control, the importance of cultural and healthy subjects and the sensitivity of

having to deal with strangers as key ones when using social applications.

Later, while Grosinger et al. (2012) developed an application targeting the promotion

of social motivation by using issues related with physical activity, Kim et al. (2013) turned

their attention to the empowering of social relationships between older adults and their

relatives by including gifts as a motivation for social touch.

Grosinger et al. (2012) developed a tablet-based prototype called Agile Life that

provided “Physical Activity Information Chunks” (PAICs) and promoted awareness of

friends’ activities and opportunities to join in. Their findings supported the preference for

the nurturing of close relationships over the forging of new ones and suggested that the

content of information chunks needs to be tailored to each older adult, or group of older

users.

TimelyPresent (Kim et al., 2013) main goal was to connect three-generation families

across different time zones. It supported the ability to create video-based gifts to stimulate

local-time-based conversations and collective reflection. Families found the prototype

valuable for their communication with the presence of very few buttons being the main

reason for its high usability. They further supported the importance of having family-



related content, and family-related features could help boost the adoption of socially-based

services by older adults. They revisited the discussion related with avoiding the creating of

new social obligations and provided indications that video-clips may have advantages over

photos as they allow for non-verbal social cues such as laughter, smiles or funny gestures

to be communicated more effectively.

More recently, Baecker et al. (2014) confirmed a lot of the previous findings regarding

factors such as the importance of asynchronous communication, photos and the presence

of some personal form of communication for keeping contact with existing family ties.

Additionally, they identified other relevant factors with identical relevance. Following,

Neves et al. (2015) used the same device as Baecker et al. and further showed the

importance of favouring additional forms of interaction, distinct uses and paying attention

to the diversity of older adults.

Baecker et al. (2014) developed the “InTouch” prototype focusing on several recom-

mendations originated from a user-centered design approach: they avoided traditional

computer aesthetics and conventions; supported expressions of person-hood by providing

alternative interaction (touch and gesture) and non-typing forms of communication like

videos, photos and audio; used pictures of family and light forms of social exchanges

like waves and smiles; focused on features for maintaining existing contacts and family

communication; supported both asynchronous and synchronous communication; and were

non-language specific by using icons. In their tests, they found the necessity to enhance

privacy features with the ability for receiving only messages from a set of contacts and

identified three types of profiles for social, active older users. Finally, results showed gains

in enabling new varieties of display media such as TV, family albums and smart-watches

and exploring multimodal input.

Neves et al. (2015) deployed the same communication appliance as Baecker et al.

(2014) and, making use of the same properties, they suggested factors that facilitate

the adoption of social technologies such as social support by family members and the

importance of factoring security and considering cultural issues (such as the meaning of

gestures) into the interaction design. They also found that different types of media, types

of use (active vs passive), and adaptive mechanisms to compensate for motor limitations,

should be used as necessary. Finally, they provided further recommendations for the design

of more inclusive social service, like avoiding the use of ambiguous symbols and using

explicit stimulus, such as icons to indicate how to perform specific gestures to achieve

specific tasks.

4.2.3 SNS use by Older Adults

The most relevant studies performed in the last years regarding the use of standard SNS

(like Facebook) by the older segment of the population, contribute to characterise the main

causes for its lack of use, and the main motivations for its adoption.



In 2005, Santana et al. (2005) performed the first study in which SNS was appointed as

being fundamental for social relationships in the old age. They revealed that phone calls

were the main method to connect with relatives and friends, but these are used more to

update information or on important dates, rather than as a proper SNS. They also showed a

preference of older adults for family contacts over friends.

After them, Sayago et al. (2007) study focused on the role of images when older adults

share information for social purposes, and Joinson (2008), Karahasanovic et al. (2009) and

Lehtinen et al. (2009) followed with investigations on the views and motives of SNS use,

and on the involvement of older adults on generating content for these platforms.

Sayago et al. (2007) looked at the subject of online sharing by the elderly. They

generated novel recommendations for that time: (1) elderly learned more from past

experiences than from tests or other learning activities, and more efficiently together than

from studying alone; (2) text-based instructions proved to be less useful than video-based

ones; and (3) avoiding computer jargon enabled them to understand new technologies

better.

Joinson (2008) was the first to investigate the use of a real SNS (Facebook), and

identified several unique uses and gratifications: social connection, shared identities,

content, social investigation, social network surfing, and status updating. Additionally,

results suggested the main goal of “keeping in touch” comprised two main functions:

surveillance, as the function where Facebook is used to see what family and friends are

“up to”; and the use of Facebook to research offline contacts. As older adults are strongly

associated with the first function, this study was the first to support the direct value SNS

can have for this population.

Karahasanovic et al. (2009) showed that although older adults are resistant to participate

in online communities, given the right circumstances, they would be very interested in

social services: (1) a TV with additional functionalities could offer a solution because

of its simplicity to turn on and operate; (2) co-creation of content would make it easier

for them; (3) personalization would be highly appreciated not only in adapting their page

but also out of curiosity about what others do with their pages; and (4) they should be

able to choose which people can see which of their content and at all times feel in control

regarding privacy settings.

Lehtinen et al. (2009) were aligned with Karahasanovic et al. (2009) findings con-

cerning privacy control as they showed older adults perceived the Internet as a socially

unsafe place. They stated that the functions for controlling the recipients of the content

published should be clear and simple and further suggested that privacy settings should be

by default set for keeping everything private and then give options for its public disclosure.

Their other findings indicated the main issues address were users’ computer skills and

transferring or constructing a common experience.

All these studies served as the basis for other, more relevant and recent, studies like the



one performed by Gibson et al. (2010), which published findings regarding the main causes

and reactions about the use of this technology by the older segment of the population. On

their studies, older adults compared SNS to a “round robin” letter they might put in a

Christmas card. Additionally, and building from Karahasanovic et al. (2009); Lehtinen et al.

(2009) findings, older adults felt vulnerable with the customary binary or tertiary privacy

settings and counter-intuitive nature of default settings and found critical to distinguish

between what they want to share with family, close friends and acquaintances. Furthermore,

the importance of varying degrees of reciprocity was identified as motivational factors,

and there was dissatisfaction with the way the SNS made use of the word “friends” as a

catch-all term for anyone. Finally, older adults felt more secure with a group identity rather

than highlighting themselves as an individual.

Later, Burke’s (Burke et al., 2011) understanding of social capital and how different

users can relate with different uses and Xie et al. (2012) perceptions on how older adults

can learn to use SNS also contributed to understanding the use of SNS by older adults.

Burke et al. (2011) distinguished between two types of social behaviour in SNS: direct

communication consisting of personal, one-on-one exchanges (messages, wall posts and

synchronous chat) which are associated with increases in bonding social capital Gilbert

and Karahalios (2009); and passive consumption of social news and its association with

bridging social capital Joinson (2008). In their conclusions, they rejected previous findings

and stated that Facebook might work best for increasing the value of weaker relationships

among the older population, and that status updates and photos could increase the feeling of

connectedness among less socially skilled elderly. They suggested several guidelines: the

UI should focus on creating incentives for users to produce stories that create opportunities

for more inbound communication; adapting the display of content so users can be prompted

to interact with contacts in need of communication; suggesting reconnecting with old

friends or reordering lists of people to highlight acquaintances over regular contacts;

and resurfacing prior content as a means of stimulating communications for inactive

relationships.

Xie et al. (2012) identified privacy (e.g. knowing if there were ways of making

information available only to some contacts) as being the critical perceptual barrier to

adoption. However, and contrary to prior studies which considered privacy issues, instead

of focusing on influencing the design, they developed educational strategies to overcome

privacy concerns like introducing the concepts before introducing the functions, responding

to privacy concerns and making social media personally relevant. The most relevant aspect

of the study was the fact that participants changed from the initial unanimous, strongly

negative to a final eventual willingness to participate in SNS. This means that, as older

adults become more familiar with SNS, they will likely integrate it into their daily lives.

Still, efforts are needed on educating about the existing settings.

Considering all the previous studies, Hope et al. (2014), Norval et al. (2014) and Harley



et al. (2014) performed broader studies on the main causes and reactions of using this

technology.

Differentiating themselves from previous studies, Hope et al. (2014) focused on aware

technological users who expressed interest in improving communication with their social

network. They supported earlier findings on “heavyweight” communications (Lehtinen

et al., 2009; Burke et al., 2011), but extended this by analysing the importance of tangible

communications, like paper-based and hand-written communications. They supported

“letter writing” as a social practice that should be explored in an SNS context when targeting

this population, as it supports varying degrees of formality depending on the content and

tone used and enables both immediate and later reflection on interactions. Other results

showed that older adults are concerned about the time required for legitimate participation

and the loss of deeper communication, content irrelevance, and privacy. A major curiosity

here is the suggestion that “by not calling anyone” (in the family) they can, in fact, be

communicating their well-being and independence. Additionally, seniors believe they can

fill gaps in the communication with people of their age who do not use SNS or have an

active role in the mass communication of illness or identical situations to the whole family.

Thus, they provided additional recommendations regarding the design of inclusive SNS:

(1) foster strong tie relationships; (2) afford expression of thoughtfulness and concern

by supporting social communications with tangible value; (3) support older adults’ value

regarding the ability to foster communication and help younger relatives; and (4) offer

content from venerated sources only.

Norval et al. (2014) identified a series of recommendations from a focus group study

with older adults: (1) the need to clarify and simplify privacy settings; (2) avoid technical

terminology and use a simple layout; and (3) avoid friend suggestions. As a result, they

developed a system with two UI versions, one for control which was strongly influenced

by Facebook default UI, and a modified UI which adapted the control UI to follow these

recommendations (and others from previous works). They changed feature location, colour

and size aspects and the default privacy settings and buttons related to it. Although no

direct communication or photo upload features were present, participants preferred and

thought it would be easier to use the Modified UI regularly rather than the default UI.

Further recommendations were suggested such as modifying the navigation panel for

privacy purposes and removing details related to an assumption of prior knowledge. Their

implementations of prior research in a usable prototype showed Facebook could be much

more usable by the older population.

Harley et al. (2014) looked at the significance of family and local community and how

these influence online engagement among older adults. Their findings spanned over the

most common domains identified until date. They reported problems related to privacy

and disclosure and the necessity for a clear transition from public to private interactions.

They also stated a clear preference for meeting face-to-face and for SNS to work as a tool



for achieving this aim. They suggested the inclusion of several features: “sharing rooms”

focused on sharing ideas and content not only dedicated to family members but also for

anonymous sharing; focus on initiatives to connect with local communities off-line; and

the existence of formal roles within an SNS to make other people engage more in the

purpose of each older user.

More recently, while Michailidou et al. (2015) focused on the differences between

the use of Facebook by older and younger adults, other authors continued to focus on the

importance of privacy aspects, the key role of photos, direct communication and adaptation

for keeping connected to family through SNS (Lang and Barton, 2015; Hayes et al., 2015).

Michailidou et al. (2015) investigated the use of online social-based services by older

adults in culturally diverse countries (Greece and UK). They found cultural differences

regarding preference for online or off-line support, perceptions of technology use and pref-

erence for using SNS only with family members or close friends. They found similarities

in the way older adults rely on off-line means of communication and physical presence

in cases of deep support, and consider themselves less capable of using and learning how

to use technology than their younger counterparts. They mainly defended the need to

incorporate the “language of the user” which differs with the type of culture by adopting

concepts such as culturally adaptable UIs (Reinecke and Bernstein, 2011). They also

recommended that SNS should provide features to understand content in any language,

make use of pictures to increase member visibility and support the ability to have a video

conversation in an instance.

Lang and Barton (2015) explored the methods individuals use to manage perceived

invasions of privacy in the form of undesirable Facebook photos uploaded and tagged

by other users. They showed older users are less likely to untag or request deletion than

younger counterparts as they are less technically skilled and less likely to experience

undesirable Facebook photos. Still, when they face this type of problem, older adults tend

to comment on the photo or contact the uploader via chat or private message instead of

untagging.

Hayes et al. (2015) hypothesised that older adults use Facebook differently and less

often than younger users and that they are more resilient to the adverse effects of Facebook

use. Results showed younger adults use the tool more frequently and are more emotionally

impacted by it than older adults. In more detail: (1) there was a negative relationship

between age and the amount of time spent viewing own photos, with younger adults

doing it more frequently; (2) there was an inverse relationship between time spent and

frequency of checking Facebook per day and age; (3) older adults use features like chat

and posting photos less than younger users mainly because they are more sceptical and

less skilled about its usage; and (4) older adults update their personal information and

status significantly less than younger adults. However, this study suffers from limitations,

including the fact that, not only they did not adequately consider differences between



younger and older adults (only 19% of participants were older adults), but also performed

no comparison between younger older adults and older older adults (no participant was

aged more than 70 years of age).

Lastly, Vroman et al. (2015) confirmed already consensual findings with the majority

of older adults in a study using ICT for connecting with family and other close social

relationships or to access health and routine-related information and rejecting activities

related with knowing new people. However, they distinguished themselves by both

rejecting the notion that older adults would like to use direct communication in this

context, and by revealing the influence of other unacknowledged factors. They showed

both education and age can have a substantial impact and also revealed that being alone

makes older adults less likely to a self-initiated exploration of these services. Moreover,

they showed the use of ICT for social purposes is also associated with the perceived

importance of the activities and with feeling healthy. Finally, they envisioned a three-tiered

person-centred model to explain and promote ICT social use among older adults with the

attitudes, needs and ICT capacity of an older adult at the centre surrounded by personal

relationships, utility factors, and the virtual community.

4.2.4 Social Network Services Developed for Older Adults

In past years, and as a response to the lack of use of traditional SNS like Facebook, the

focus has also been given to the development of SNS solutions specially tailored for older

adults.

Czaja et al. (1993) developed the first meaningful attempt of prototyping an SNS to

target older adults specifically. It consisted of a simple messaging system to support social

connectedness for users with little computer experience. Participants could operate the

system with a minimal amount of difficulty and indicated it facilitated social interaction

and provided a chance to meet new people. Still, a decline in use was observed over time,

as people were busy and the system had limited functionalities. Also not favouring its use

was the fact that it focused exclusively on the female gender.

It lasted about one decade until another SNS for older adults was developed. Morris

(2005) solar concept focused around giving health feedback to older adults and provided

meaningful ideas on how to present content in SNS, while four years later Sillanpaa

et al. (2009) concept focusing on people with cognitive and language impairments was

relevant as it supported alternative ways of interaction when considering user characteristics

typically present in the older population.

In Morris (2005) symbolic representation, the older person was the sun at the centre of

the solar system and was surrounded by stars representing their relatives. It made use of

sensors and self-reported data and conveyed minimal information to avoid stigmatising

the elder. It kept a history of recent interactions by tracing a line between each person’s

initial and current level of contact and modifying the distance and thickness of the line.



Elders could switch to a line graph (with aggregate contacts with everyone in the social

network) or a bar graph (with the levels of contact with each contact on a given day). This

display was considered to act as a catalyst for people to contact the ones they were losing

touch with. It also incited elders to be less passive, by preparing food for visitors, initiating

phone calls, or driving somewhere to see someone. In general, this SNS constituted an

excellent example for monitoring social ties and alerted for the importance of family and

photos when designing for older adults and to the relevance of off-line interaction.

Sillanpaa et al. (2009) introduced solutions used in Kaveripiiri.fi web service, an SNS

targeting young people and older adults with intellectual and developmental disabilities,

or any kind of communication, attention control or guidance problems. Kaveripiiri’s

layout was focused on grouping items by proximity, and the appearance of UI elements

and functionalities was unified throughout the service so that active areas and essential

functions could be located more easily. They showed that registering and logging in was

challenging in many ways and that users with reading and writing difficulties could not

fully participate, as communication in SNS is mostly done in written form. Having this in

mind, they added text-to-speech functionality for all messages, a text input tool with ready-

made phrases to choose from and log in by selecting a sequence of symbols. Concerning

privacy they reorganised the content based on privacy level (“me and my friends”, “all

users” and “private messages”) to make it more obvious what was the audience for the

content posted. Their experiments showed that adaptations had a major impact on the

service’s accessibility. Additionally, their findings rejected the notion that older adults

do not want to know new people online. Lastly, they provided a discussion about the

difficulties of situations where people are communicating but cannot see or hear each other

and on how that places an extra burden on older adults interactions in social media.

Following these studies, we find one of the most relevant studies performed until date

in Chen (2009), which focused on several SNS solutions targeting older adults and their

main pros and cons. One of the first recommendations given was that elderly benefit most

from a simplified and obvious UI, i.e., less is more. This fact is in line with Sillanpaa et al.

(2009) findings regarding simplicity. However, Chen’s expanded on this by contributing

to the notion that offering fewer possibilities for interaction would be better for older

adults. Additionally, he defended the importance of supporting domestic languages in SNS

targeting older adults, as well as further supporting the notion of family members being

the central focus point. Moreover, and resulting from his analysis of several SNS, several

good practices can be resumed: (1) resulting from Bettie 1, non-technological savvy older

adults, could place a relative pass (a card) at three distinct areas which supported distinct

functionalities (sending a message, showing updates, showing details of that contact). The

existence of this type of short-cuts for specific functionalities as well as the existence

of cards for specific persons evidences the importance of the notions of grouping and

1http://www.bett.ie/



favourites; (2) from Oldkids 2 and Eons 3, emerges the discussion that SNS based on both

social approval between the old-age community, E-commerce and gaming approaches,

rather than on the benefits of social communication with family and close friends are not

appreciated by older adults; (3) finally, resulting from AboutMyAge 4 and Sagazone 5,

comes the reflection that not adapting content and UI to the older segment of the population

also makes them opt out.

After Chen study, other three more recent SNS were proposed: Burmeister (2012)

study on an online community of seniors called GreyPath; Cornejo et al. (2013) proposition

for a novel ambient SNS which was based on a strong multimodal component to drive

elderly closer to this kind of systems; and finally, Waycott et al. (2013) SNS prototype

build for older adults sharing of photographs and messages.

Burmeister (2012) GreyPath portal supported from technologically novice users, to

technologically literate people. Greypathians could enrol in courses, contribute to or

receive information on a wide variety of topics, and there were rooms and forums, each

moderated by a senior volunteer who had a name and a face on the site. Older adults saw

it as an inclusive “ neighbourhood” community defined by a social exchange that goes

beyond the act of information exchange and encourages personal contact. Additionally,

GreyPath showed that an SNS targeting older adults has to aim for off-line interaction (just

like Morris (2005) first indicated) and to somehow construct a community around each

user. Be it a community of family members and close friends, or a community of people

who live close to each other.

Cornejo et al. (2013) developed Tlatoque, a multi-touch screen PC resembling a photo

frame, which communicated to Facebook to expose photographs in the user’s home and

provided means of reciprocating information back. It was implemented as an Ambient

Social Network System (aSNS)6, in which the SNS component could automatically share

on the user’s behalf, having privacy and interests in consideration. It further enabled the use

of ambient feedback services to share meaningful routine and hobbies related information

(like in Tentori et al. (2010)). Along with touch-based gestures, it supported a set of

natural gestural interaction to achieve tasks like turning on or off the display and browsing

photographs. Tlatoque made use of four distinct feedback services: the first enabled older

adults to share interests; the second enabled them to rate the content being displayed; the

third enabled the sharing of their whereabouts, and the last was an exergame that enabled

users to socialise through exercise accomplishments. Results of using Tlatoque, showed

2http://www.oldkids.cn3http://www.eons.com4http://www.aboutmyage.com5http://www.sagazone.co.uk6An aSNS is a ubiquitous system with a set of SNS services moved outside the desktop into the

environment where we live and interact embedded into common objects (e.g. portraits, lamps, and notebooks)

or tangible artefacts available in the user environment providing intuitive means for browsing, managing, or

generating social media content.



that displaying information through sentient objects and the inclusion of touch, body and

gesture-based interaction eased older adults adoption of technology and manipulation of

social information. Additionally, older adults used the content uploaded by relatives to

enrich both in-person and online conversations and increased interaction with relatives

also due to the content of family events.

Waycott et al. (2013) investigated the nature and role of digital content that has been

created by older adults, to forge new relationships. They used a prototype application called

Enmesh to create and share photographs and messages. It made use of the iPad in-built-

camera and the virtual keyboard for exchanging photographs and messages which would

float down the screen in a semi-random fashion displaying recent items more frequently.

Each object could be manipulated on the screen in real-time by one user at a time, building

a sense of social presence with the goal of creating a community. Findings showed that

participants did not shy away from sharing some of the bad and good aspects of their daily

life, often using humour to describe their experiences. Thus, this enabled them to build

rapport and find common interests. Additionally, while sharing information, participants

were not overly concerned with privacy as all of them knew each other through face-to-face

meetings before. Moreover, the combination of images and text-enabled participants to

describe and share personal objects and spaces that communicated details of their histories

and everyday lives. Therefore, Waycott et al. (2013) findings evidenced the importance of

constructing a community around the older adult, just like Burmeister (2012) had proposed

before. However, they further showed how older adults who are nor previously acquainted,

can interact with each other in a social-technology manner, without too much privacy

concerns, if the UI is transparent regarding each user’s actions.

4.2.5 Discussion

From all the research presented it was possible to identify and distinguish a total of thirteen

different domains. Table 4.3 enumerates and shows how those span across all the related

work. In this section, we discuss each domain, by summarising how related work embraced

the domain through time.

4.2.5.1 Family

Across all the work performed in recent years concerning SNS and older adults, there is a

clear indication of what makes them adopt these tools: the possibility of keeping in contact

with family members which by some reason, or a combination of factors, physically or

emotionally, got away. Correspondingly, more than half of the literature reviewed puts

family as the foundation for the use of social services and tools among elderly.

Since the first meaningful social studies and applications started to target older adults

social issues, that family has been appointed and used as the main motivator (Santana et al.,



Table 4.3: Related Work and Domains

Study Fam

ily

Inte

rfac

e

Pri

vac

y

Ph

oto

san

dM

edia

Mu

ltim

od

al

Dir

ect

Co

mm

un

icat

ion

Kn

ow

ing

New

Peo

ple

Per

son

aliz

atio

nan

dA

dap

tati

on

Gro

up

ing

Tan

gib

leV

alu

e

Offl

ine

Gen

der

Rec

ipro

city

Brunette2005 ✓ ✓

Romero2007 ✓ ✓ ✓

Lindley2009 ✓ ✓ ✓ ✓

Vetere2009 ✓ ✓ ✓ ✓ ✓

Raffle2010 ✓ ✓

Vutborg2010 ✓ ✓

Giorgi2011 ✓ ✓ ✓ ✓

Gaver2011 ✓ ✓

So

cial

Ap

pli

cati

on

s

Judge2011 ✓ ✓ ✓ ✓ ✓

Lindley2012 ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Garattini2012 ✓ ✓ ✓ ✓

Grosinger2012 ✓ ✓ ✓

Kim2013 ✓ ✓ ✓

Baecker2014 ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Neves2015 ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Santana2005 ✓ ✓

Sayago2007 ✓

Joinson2008 ✓ ✓ ✓

Karahasanovic2009 ✓ ✓ ✓ ✓ ✓

Lehtinen2009 ✓ ✓

Gibson2010 ✓ ✓ ✓ ✓ ✓ ✓ ✓

Burke2011 ✓ ✓ ✓ ✓ ✓

Xie2012 ✓ ✓ ✓

Hope2014 ✓ ✓ ✓ ✓ ✓ ✓

SN

SU

se

Norval2014 ✓ ✓ ✓

Harley2014 ✓ ✓ ✓ ✓

Michailidou2015 ✓ ✓ ✓ ✓ ✓

Vroman2015 ✓ ✓

Lang2015 ✓ ✓ ✓ ✓

Hayes2015 ✓ ✓ ✓

Czaja1993 ✓ ✓

Morris2005 ✓ ✓ ✓ ✓

Chen2009 ✓ ✓ ✓ ✓

Sillanpaa2010 ✓ ✓ ✓ ✓ ✓ ✓

Burmeister2012 ✓

Cornejo20013 ✓ ✓ ✓ ✓ ✓

Eld

erly

SN

S

Waycott2013 ✓ ✓ ✓ ✓ ✓



2005). This can be seen in the metaphor of the interactive solar system which focused

on keeping touch by using representations of family members’ contacts (Morris, 2005),

and other systems which also used them as the “bread and butter” for its goals (Vetere

et al., 2009; Raffle et al., 2010; Lindley, 2012; Judge et al., 2011; Kim et al., 2013). In

fact, exchanging messages and pictures between younger and older generations (Romero

et al., 2007; Vutborg et al., 2010), was also appointed by the elderly as vital in the process

of learning new technology (Karahasanovic et al., 2009; Giorgi et al., 2011; Vroman

et al., 2015). Finally, the family was also appointed as the main focus of SNS targeting

specifically old age (Chen, 2009) and situational display applications (Lindley et al., 2009),

and appointed as the main cause for using existing SNS like Facebook (Gibson et al., 2010;

Xie et al., 2012; Grosinger et al., 2012; Cornejo et al., 2013; Hope et al., 2014; Baecker

et al., 2014; Neves et al., 2015; Michailidou et al., 2015; Vroman et al., 2015).

4.2.5.2 User Interface

Real life social networks are a product of the interaction between individuals and have no

real interface, except the one provided by each person. SNS has a typical interface which,

regardless of minimal changes, is provided to every user in the same way. Although this

principle could favour interaction with SNS when compared with real-life social networks,

users of all ages have complaints about SNS interfaces being too complex, having too

many options, and not being appropriate for “ non-typical” users who suffer from physical

or mental impairments, or to users without a technological background. By taking another

look at table 4.3, it is evident that the SNS UI issues are the second most debated domain

in the last ten years of research, or since SNS was launched globally. Current SNS UIs

are not only one of the main causes for many to not adopt SNS, but also one of the main

complaints from the ones who adopted.

A lot of research on SNS has focused on suggesting alternative UIs (Morris, 2005;

Chen, 2009; Giorgi et al., 2011), or simplifying UI elements (Sayago et al., 2007; Sillanpaa

et al., 2009; Gibson et al., 2010; Hope et al., 2014; Norval et al., 2014; Baecker et al., 2014;

Neves et al., 2015; Hayes et al., 2015), especially when considering older adults as a target.

Additionally, several other social applications focused on different ways of getting closer

to older adults by providing dedicated UIs focused on message exchange between older

adults in the role of grandparents and the rest of family (Vetere et al., 2009; Lindley, 2012;

Cornejo et al., 2013; Kim et al., 2013), or on a community context (Gaver et al., 2011;

Waycott et al., 2013).

In their conclusions, researchers are unanimous in terms of what a SNS like Facebook

offers: its features are placed in a non-intuitive manner in the UI; specific functionalities

are hard to find; terms used in buttons and links are not familiar to those who have not

been using this kind of services before; and it is not usable by impaired users, older adults,

or users which shift away from the typical 20 to 40 year segment.



4.2.5.3 Privacy

All the papers focusing on SNS use by this segment of the population, refer privacy as the

main barrier to its adoption. While this issue is relevant for a lot of users, as older adults

have a more restrictive goal when using SNS, privacy assumes extreme relevance. In the

same way, they are most interested in keeping contact with relatives and close friends; they

also need to feel secure when favouring SNS over more traditional ways of communication.

Currently, SNS, as they are, are not safe for older adults to adopt, as privacy definitions

and configurations are to complex for them to use or get used too.

The fact that users need to have some knowledge to understand where the information

they share ends up makes older adults opt out of these services (Sillanpaa et al., 2009;

Gibson et al., 2010). Therefore, there is a common view that older adults should be

able to feel in control of privacy settings and choose which people can see which con-

tent (Karahasanovic et al., 2009; Xie et al., 2012; Cornejo et al., 2013; Hope et al., 2014;

Neves et al., 2015) and that privacy settings should be kept clear and simple for them to

use (Lehtinen et al., 2009; Norval et al., 2014). This is particularly relevant concerning

family and close friends. Keeping a “closed network”, available only to their relatives, is

known to be a reason for adopting social applications (Lindley, 2012; Baecker et al., 2014),

as well as gradually granting access to more information as they get familiar with each

contact (Garattini et al., 2012; Waycott et al., 2013).

4.2.5.4 Photos and Other Media Relevance

The relevance of certain types of media is a subject discussed in great part of the literature

concerning SNS. More specifically, the role of photos in drawing older adults closer to

social applications and services has been studied by several researchers as a relevant one.

Starting ten years ago, there has been strong evidence that older adults prefer all image

modalities to text (Morris, 2005) and that bringing pictures to older adults’ homes through

the use of digital frames can build attachment to a place like a community centre (Brunette

et al., 2005) as well as provide context for conversations with family members living

apart (Romero et al., 2007; Vetere et al., 2009). Additionally, the use of photos as the

focus of SNS profiles and sharing activities (Chen, 2009; Burke et al., 2011; Waycott et al.,

2013; Baecker et al., 2014) has been shown to have a crucial effect on the ones with lower

social communication skills, increasing their feelings of connectedness. Moreover, and

although some studies reported that older adults tend to have and upload fewer photos than

their younger counterparts (Hayes et al., 2015; Lang and Barton, 2015) the implementation

of interaction around the direct manipulation and visualization of photos (Vetere et al.,

2009; Gaver et al., 2011; Waycott et al., 2013; Baecker et al., 2014; Neves et al., 2015;

Michailidou et al., 2015) has also shown to be important in building attachment to the

system and their contacts.



4.2.5.5 Multimodality

One of the major limitations of current SNS like Facebook, and also one of the main causes

for its lack of use by users with age-impairments, is the absence of alternative ways of

interaction, different from the traditional keyboard and mouse, RC in TV contexts, and

touch in smart-phone contexts. In almost half (14 out of 31) of the literature considered in

this survey, researchers refer alternative modalities; adaptation mechanisms focused on

complementary ways of interaction, and the support for several modalities at the same

time, as indispensable for compensating the old-age associated impairments.

In the vast majority of the cases, suggestions and findings regarding multimodal

interaction in the context of SNS, have been a result of older adults’ suggestions when

faced with accessibility problems (Lindley, 2012), or when asked about the main reasons

for not adopting this type of services (Karahasanovic et al., 2009; Lindley et al., 2009;

Judge et al., 2011). As a response to these suggestions, solutions have been proposed

by the community with the goal of facilitating the use of social services, with different

modalities being proposed by almost every researcher: location-aware cameras, large public

displays and RCs (Brunette et al., 2005); TV with additional functionalities (Karahasanovic

et al., 2009); digital frames, handwritten input and smart-phones (Lindley et al., 2009);

smart-phones and touch-screens (Vetere et al., 2009); ready-made phrases for text input

and symbolic input for logins (Sillanpaa et al., 2009); laptops, audio and page sensing

technology (Raffle et al., 2010; Vutborg et al., 2010); speech input augmentation (Lindley,

2012); touch and gestures (Cornejo et al., 2013); combination of text and images or text

and videos (Judge et al., 2011; Waycott et al., 2013); and waving, tapping and gesture

based interaction (Neves et al., 2015; Baecker et al., 2014).

4.2.5.6 Direct Communication

Although SNS can be used for many different goals, for older adults, the main reason for

using social technologies is to keep in touch with relatives. The main way of keeping in

touch is by establishing some sort of direct communication. On the studies performed with

this segment of the population, it was common for older adults to talk about features related

to receiving direct feedback from their contacts. Consequently, many social technologies

focused on the implementation of communication channels between them and their family,

which could work as an alternative to standard communication devices like the telephone

or traditional, and now, less used practices, like writing letters.

A study performed around 2005 first showed a strong preference towards alternative

ways of communicating with relatives rather than making phone calls (Santana et al.,

2005). Following, other works showed the importance of chat and direct messaging

related functionalities, or a preference for interactions that better approximate a “real

conversation”, especially when focusing on family contact (Romero et al., 2007; Vetere

et al., 2009). Studies also showed that older adults are usually the ones who initiate



great part of the “conversations” (Lindley, 2012). This can be directly related to findings

showing that these type of functionalities where people communicate but cannot see or

hear each other, are known to place an extra burden on interaction, as they usually involve

some kind of response obligation (Sillanpaa et al., 2009; Vetere et al., 2009; Hope et al.,

2014). Because of these obligations, the use of asynchronous messaging as a way for older

adults to work around the busy schedule or availability of their relatives (or just keeping

contact with new friends) (Vetere et al., 2009; Judge et al., 2011), the use of hybrid services

which provide both asynchronous and synchronous communication (Baecker et al., 2014;

Neves et al., 2015), as well as the creation of incentives for users to produce stories that

create opportunities for more inbound communication (Burke et al., 2011), were adopted

in several applications and studies. The importance of direct communication for dealing

with privacy issues, like unwanted tags on a photo, was also revealed (Lang and Barton,

2015).

4.2.5.7 Knowing New People

Along with re-connecting with old contacts, the original goal of SNS was the possibility

of making new friends and meet new people. This goal was first accepted by users who

first adopted this kind of social services on the Web, but later dropped by the majority of

people (Boyd, 2007). Nowadays, there is a lot of research showing a relation between the

way SNS are used to establish new friendships and the age of the user, with younger users

being the ones associated with being more capable of meeting new people on-line (Burke

et al., 2011). Research also showed that for older adults, the use of SNS for this purpose

is not only unattractive but also scary and appointed by the majority as a cause for not

making use of tools like Facebook.

Early research showed that SNS needed to be perceived as useful for them to be adopted

in the long term, and meeting new people, most of the times, was not enough (Czaja

et al., 1993). In the same way, research which followed showed that: it is necessary

to make a distinction between close relatives and people the user does not necessarily

know (Gibson et al., 2010); for older adults, to call and talk with someone they do not

know, requires courage (Garattini et al., 2012); connecting with family and close friends

instead of establishing new friendships, builds up social capital in the way older adults

need (Burke et al., 2011; Grosinger et al., 2012; Hope et al., 2014); SNS should not disrupt

social ties with existing family or friends by making new ones (Baecker et al., 2014);

avoiding friend suggestions and unsolicited email is a necessary step for older users to

adopt Facebook (Norval et al., 2014); and that there is, in fact, a tendency to reject social

activities on-line which are related to knowing new people (Vroman et al., 2015). On the

other hand, research also suggests that with time people get to know each other and have

less frequent but longer and more meaningful conversations, like the ones they have with

family and close friends, and that in fact, these new relationships can be more reciprocal



than the others (Sillanpaa et al., 2009; Garattini et al., 2012; Judge et al., 2011). Still, even

these studies conclude that providing opportunistic, light-hearted social interaction among

strangers is hard, closely related with taking risks (Harley et al., 2014) and again usually

not enough to encourage the use of an SNS.

4.2.5.8 Personalization and Adaptation

Older adults suffer from age-related impairments which, by affecting each one’s sensory,

motor and cognitive functions in a distinct way, create a great heterogeneity of charac-

teristics in this segment of the population. In past works concerning the use of social

features, these differences between older individuals were also evidenced, and several

considerations concerning the use of personalisation and mechanisms to adapt to each

situation have been presented.

Since at least 2009, researchers have been verifying in user trials that older adults

highly appreciate personalisation and adaptation, especially when considering SNS profile

presentation (Karahasanovic et al., 2009) and features (Sillanpaa et al., 2009; Burke et al.,

2011; Lindley, 2012; Xie et al., 2012). Although regarding profiles and pages, it is more a

matter of taste, when talking about functionalities and the way each user can, not only have

access to it, but also make use of it, it is a matter of necessity. Additionally, other studies

focused on the differences between distinct family members with differing awareness

and communication needs (Judge et al., 2011), and in the differing older adults’ needs

regarding chunks of information (Grosinger et al., 2012), and how these differences also

have to be considered when designing a service usable by the elderly. More recently,

other researchers have focused on distinct dimensions which evidence the necessity for

personalization and adaptation features: the identification of different older adults’ styles

of social users (Baecker et al., 2014; Hayes et al., 2015); the necessity to address physical

factors and how they should be compensated regarding the weight of the device and

dexterity levels (Neves et al., 2015); and the influence of culture regarding the choice for

on-line and off-line support by family members (Michailidou et al., 2015).

4.2.5.9 Grouping

Even for the younger generations, SNS like Facebook are generally too complex regarding

the vast number of features they offer, and in the way, they can be accessed or reconfigured

to fit each one’s tastes and needs. Older adults and the way they cannot make use of the

majority of the features are the best examples to understand how SNS features are not

designed in an inclusive or accessible way. In the past five years, solutions contemplating

the grouping of functionalities and contacts were appointed by several older users in several

distinct studies and projects, as one of the simpler ways of making SNS more usable.

In the last seven years, a set of technology and studies suggested grouping func-

tionalities into family (Chen, 2009; Gibson et al., 2010), favourites (family and close



friends) (Gibson et al., 2010; Lindley, 2012; Harley et al., 2014), and community (Burmeis-

ter, 2012) oriented features. Grouping was suggested considering how content can be

displayed and shared with others for content production to be a rewarding and engaging

activity (Karahasanovic et al., 2009; Waycott et al., 2013; Harley et al., 2014). Many

different solutions were tested. Older-adult specific applications, for example, focused

on co-creation of content (Karahasanovic et al., 2009), restricting the contact list to the

essential (Chen, 2009; Lindley, 2012) and adapting the UI to highlight only meaningful

contacts (Waycott et al., 2013).

4.2.5.10 Tangible Value

Most research performed in the last decade concerning social applications in the scope of

older adults focuses on the family before anything else. However, in most cases, the family

is not around, and the virtual interaction is more meaningful when supported by photos, or

some kind of media and personal artefacts related to family members. This is similar to

what older adults have around their houses: framed pictures and objects that remind them,

every day, of the ones they want to keep contact with or, at least, the ones who make them

feel accompanied. The tangible value or the presence of this kind of artefacts in digital

format can also have a similar role, at the same time as they help older adults in keeping

contact with the ones they want the most. In the last seven years, there were some works

which evidenced this tangible value and made suggestions about ways of merging this with

SNS.

In the last years, there was a lot of research on digital frames capable of presenting or

exchanging family photos between older adults and their relatives (Giorgi et al., 2011).

This is the most direct and basic way of providing tangible communication, and also

one of the main ways of making older adults adopt this type of social technologies.

Additionally, there were other suggestions on supporting scribbling, handwritten messages

and notes on pictures, as other types of exchanged media which constituted a more

personal form of contact going deeper into the tangible value(Lindley et al., 2009; Lindley,

2012). Furthermore, a recent study showed the importance of paper-based communication,

even in a virtual environment (Hope et al., 2014), mainly because it affords varying

degrees of formality and gives users a chance of immediate reflection. Lastly, a couple

of studies (Baecker et al., 2014; Neves et al., 2015) reinforced this need for supporting

expressions of person-hood through tangible UIs based on real-world objects and practices.

4.2.5.11 Offline Role

Although SNS are all about social relationships on a virtual environment, and the use of

these social contexts can have an important role in decreasing isolation and loneliness,

there is still a dimension related with how the opportunities provided by these virtual



environments can act as catalysts for off-line interactions. This relation between online

and offline relationships is referred by a substantial part of the chosen literature.

Morris (2005) solar system, which situated the older adult user as the sun in the centre

of a system constituted by their relatives, showed a direct relation with an increase of

off-line activities initiated by the older users. In fact, this first indication was later con-

firmed as a relevant incentive for the use of SNS by older adults: SNS should work best

in conjunction with meeting face to face (Garattini et al., 2012); generate opportunities

for real encounters (Giorgi et al., 2011); provide a clear transition from public to private

interactions and provide opportunities to connect with local communities (Harley et al.,

2014); and because older adults rely on off-line communication for more urgent situa-

tions (Michailidou et al., 2015). Additionally, topics of conversation and affinity themes

found in online interactions are also used by older adults as a way to enrich in-person

conversations (Cornejo et al., 2013).

4.2.5.12 Gender

Although there are clear human factors which influence the use of SNS, little is known

regarding other factors directly related to each one’s characteristics. Especially regarding

older users, and resulting from the extensive research performed in the last years, some

intrinsic factors have also been appointed as having an influence. The gender of the elderly

user is one of those factors.

Since one of the first studies regarding SNS was performed only with women between

the ages of 55 and 95 (Czaja et al., 1993), there was the notion that gender would have a

direct influence on the use of these tools. Still, in the literature reviewed for this survey,

only other three researchers compared the use of SNS by users of different genders (Gibson

et al., 2010; Garattini et al., 2012; Lang and Barton, 2015). Although their results show

that women are more frequent users of social tools than men, none of them has advanced

any justifications for why this happens. Additionally, and from a technology point of view,

as both old women and old men suffer from the same kind of impairments, there seems

not to be any explanation related with their physical or perceptual characteristics.

4.2.5.13 Reciprocity

The main goal of social tools like Facebook, especially when focusing on older adults,

should be on establishing and keeping contact with loved ones. However, the way in

which communication is supported by the tool can be determinant for this goal. Older and

younger adults are very different from each other (and older adults between each other)

not only in the level of technology awareness but also regarding interests. Being capable

of attracting each to contents which enable reciprocity of communication will determine if

both ends of communication are able and interested in maintaining the type of constant



contact which results in social advantages. Still, until recently, not many researchers have

evidenced this domain of reciprocity.

The first relevant study referring the notion of reciprocity and its importance in the

scope of SNS concluded that a service like Facebook was already capable of affording

expressions of reciprocity, mostly by focusing on similar interests (Lehtinen et al., 2009).

In their opinion, the job was more on constructing a common experience for both older

and younger users than anything else. These notions were also confirmed by a study

focusing on older adults’ opinions, which showed reciprocity is a factor in the adoption of

SNS (Gibson et al., 2010), and later picked up and “enhanced” by subsequent studies where

content could be sent to digital frames at an older adult’s house by adapting channels of

communication which reflect multi-use by younger generations (this way reciprocity could

be supported by fitting interaction modalities to both older and younger adults capabilities

and habits) (Lindley et al., 2009). In another example, resurfacing prior content, like status

updates or photos with comments, was used as a means for stimulating communications

for relationships that have been inactive for some time (and once again favour interactions

in which reciprocity was missing) (Burke et al., 2011).

4.3 Gathering Older Adults Requirements

This section is divided into three parts. First, we describe the process and results of a

questionnaire with older adults about the concepts behind SNS and particularly about

Facebook use or adoption possibilities. Secondly, we describe the interviews and focus

groups we performed with representatives of this population to foster broader discussion

on these topics. Thirdly, we described the results of semi-structured interviews conducted

with older adults and focused on using the TV as the main platform for accessing SNS.

4.3.1 Inquiring Older Adults About SNS

To further understand how to draw Facebook closer to older adults a questionnaire was

designed and applied. More particularly, the goal was, first of all, to understand what

distinguishes Facebook users and non-users (across the older population). We wanted to

check if older adults not using Facebook have the desire to use it, and try to understand

the main reasons behind its non-adoption or limited use. We also wanted to understand

the differences between these older adults regarding their education, gender, age-range,

age-related limitations and the way they believe in themselves concerning technology use.

Additionally, we wanted to investigate if these factors can influence the adoption and use

of an SNS. Moreover, and building from previous findings supporting the use of TV in an

SNS context, it was also a goal of this questionnaire to learn about older adults opinion

on the use of this platform as an alternative for accessing Facebook or related features,

identifying distinct groups and feelings about its use.



We considered both collecting richer data from interviews or focus groups, and pooling

a larger number of sources with user surveys. Because richer data would be helpful for

more in-depth characterisation of some of the topics to be addressed, we opted for first

conducting a larger user survey and leave other instruments to the subsequent stages.

4.3.1.1 Procedure

The questionnaire (attached as Appendix D) was designed in three distinct languages

(English, Portuguese and Spanish) to reach an international audience. An online version

was made available for four months and announced on several mailing lists with the focus

on health-care and research. Additionally, and to make sure that not only older adults with

access to the Internet could participate, an off-line version was also administered in person

at three institutions (two senior universities and one retirement home).

In total, 141 participants (64 male, 77 female) answered the questionnaire. From

these, 129 were performed on-line, while 12 were administered in person. All participants

were volunteers and more than 55 years old. We did not restrict participants to more

than 60 or 65 years of age, as we also wanted to reflect the importance of individuals

which will be considered older adults in the next five to ten years. Age was not asked

directly, and participants selected the appropriate age range from 6 different categories

which spanned from “less than 60” to “more than 80”. Table 4.4 shows how participants

spanned over those categories. Anonymity was kept at all times, with each user being

identified exclusively by an id number. All data was saved in a secure repository.

Age categories Frequency Percentage

Less than 60 46 32.6

Between 60 and 64 29 20.6




More than 80 12 8.5

Total 141 100.0

Table 4.4: Participants’ age range distribution.

Regarding the questionnaire composition, profiling questions on age, gender, edu-

cation, household composition, the way participants see themselves concerning the use

of technology, and regarding difficulties and impairments, were asked first. Secondly,

participants were asked about Facebook awareness, how frequently they make use of it,

or how many times they have tried it. These questions were followed by Likert scale

questions where participants had to classify their agreement (on a scale from 1 (strongly

disagree) to 5 (strongly agree)) towards possible reasons why they use or why they would



like to use Facebook. Sentences were related to getting information about what’s going

on in the world, family and friends, making new friendships, privacy issues, usability and

satisfaction. The option for structuring the questionnaire in this way was supported by the

fact that we did not want in this first user requirement phase to restrict older adults answers

to the set of domains identified in the previous requirement gathering phase (section 4.2

Gathering Literature Requirements).

Moreover, questions related to the use of TV were asked next. No questions related

with awareness or frequency of use were asked, as it was assumed that every user had a

TV (in fact one participant questioned this option, and noticed she neither had or wanted

to have one). Instead, a set of Likert scale questions (with the same scale as above) were

asked regarding reasons why participants use TV and how easy it is for them to interact

with traditional components of this platform. These were followed by matters related with

their familiarity with smart-TVs, access to the Internet through TV, and their willingness

to use the TV as a way of accessing Facebook or related social activities and content.

On a last note, although specific questionnaires for accessing technological expertise

and other skills’ characterising instruments already exist we decided against using such

standardised tools due to 1) the increase it would represent in what constitutes already a

long questionnaire; and 2) most of them being based on self-reported data nonetheless.

4.3.1.2 Results

From all answers provided, it was possible to structure this user study results into five

specific areas. While the first two concern technological expertise, age-related limitations

and Facebook knowledge/use, the following two are more focused on the differences

between older adults who use Facebook and the ones who don’t, or on how we can

distinguish older adults regarding these technologies. Results regarding the use of TV

and its relation with SNS are presented at last. Moreover, whenever relevant, the relation

between these results and the previously identified SNS domains is also referred.

4.3.1.2.1 Technological Expertise and Age-Related Limitations We inquired partic-

ipants about their self-characterisation regarding technical expertise and the technologies

widely available in the market (figure 4.1). More than half of the participants (58%)

considered themselves as being capable of using typically available technology and only

about one fourth (24%) considered to have some difficulties interacting with technology.

Concerning the matter of age-related limitations, each participant answered from their

perspective selecting the type and severity of impairments they feel to have. About one

third considered not having any impairment, while 45% to have only one type, with about

10% having two types, 9% three different types of impairments and a total of 5 partici-

pants (3.5%) having the whole range of hearing, vision, motor and cognitive impairments.

Additionally, regarding the severity of limitations severity, 53% of the participants consid-



Figure 4.1: Graphic showing distribution of answers regarding participants technological

self-characterization

ered having only minor impairments, leaving only about 14% considering to have more

severe limitations. Lastly, and concerning the type of impairments, the most frequent

ones were related to vision, with more than half (51%) considering to have this kind of

difficulties. Additionally, 20% considered to have hearing issues, about 15% to have

motor limitations and other 15% to have cognitive difficulties as a result of the ageing

process. These results showcase the differences between older adults individuals and also

show how age-related limitations span overall physical domains. Both of these points are

characterisation specificities supported by the previous literature.

4.3.1.2.2 Facebook Knowledge and Use Regarding Facebook, almost everyone (97%)

was aware of what it is, and about 67% reported using it, with about 40% doing it every day

(representing 62% of the ones which use Facebook). These results were quite unexpected

for us, with only one-fifth of the participants reporting to never having used the SNS before.

Additionally, the number of Facebook users was lower within in-person respondents (50%)

than for online participants (69%). While this was expected, it did not cause any significant

differences in the findings.

When questioned about the reasons behind the use of this SNS (table 4.5), participants

gave the most relevance to keeping in contact with family and close friends. This is in-line

with the findings obtained in the literature requirements (section 4.2) which identified the

main importance of family in the use of SNS. Additionally, they have an almost equivalent

relevance to keeping in contact with friends. This was not identified previously as a relevant

domain, and might be dependent on the discussion regarding the meaning of the word



“friend” and the degree of proximity that a “friend” can, in fact, have for an older person.

Moreover, they’ve reported being neutral or reject (mode 1) the use of Facebook as a means

for accessing information about what is going on in the world, and tend to reject the use

of Facebook as a platform for making new friends. This also supports findings from the

previous phase regarding how new relationships are not what older adult pursue (and in

fact want to avoid) when using this kind of services.

I use

Facebook

to...

know what’s

going on in

the world.

know about

my relatives

and close

friends.

know what’s

going with all

my friends.

make new

friendships.

Mean 2.68 3.56 3.26 1.63

Median 3.00 4.00 3.00 1.00

Mode 1 4 4 1

Table 4.5: Willingness to use Facebook.

When posing the same questions to participants who do not use Facebook, all possible

reasons for using the platform in the future were rejected (averages between 1.96 and 1.82

and modes of 1). This shows that the majority of older adults who do not make use of

Facebook do not see any advantage in doing so, which in turn can indicate that even if

they had heard about Facebook, they lack real knowledge about what it can offer. This

is also supported by the fact that from the 30% of participants which are non-users of

Facebook, only one third had experimented with the SNS. Therefore, until older adults

become acquainted with the possibilities of Facebook, they will most probably reject its

utility and consequently its use.

When confronting Facebook users with the problems which contribute the most for its

limited use (left of figure 4.2), participants appointed privacy issues (Mean = 3.18, Median

= 3.00 and Mode = 5) as the major factor. At the same time, they also rejected not liking

Facebook (Mean = 2.23, Median=2.00 and Mode = 1) or not knowing how to use it (Mean

= 1.69, Median = 1.00 and Mode = 1) as reasons for not taking full advantage of it.

In the same way, when asking non-users of Facebook about these reasons (right of

figure 4.2), privacy was even more strongly appointed as the main barrier for making use

of it (Mean = 3.73, Median = 4.00 and Mode = 5). However, in these participants case, the

preconceived idea of not liking Facebook is also appointed as having a direct influence in

not adopting it as a social tool (Mean = 3.39, Median = 3.00 and Mode = 3).

These results are also in-line with previous related work on these matters, as privacy

issues related both with the way privacy settings are designed and the controversial stories

about invasion of privacy are known to have an influence on the Facebook use and adoption

especially among older adults.



Figure 4.2: Reasons for limited use of Facebook for actual users (left) and for non-users

(right).

4.3.1.2.3 Differences between Facebook users and non-users From the findings re-

garding Facebook use, we got some insights into the way older adults see this tool and the

general problems that lead to its limited use or adoption. However, more insights were

needed regarding what makes an older adult, a Facebook user.

One major insight we wanted to take into consideration was the possible influence

of older adults’ self-belief in their capabilities for using Facebook (table 4.6). We tested

for possible correlations between the way participants classify themselves regarding

technical expertise and Facebook use. Performing a Pearson’s correlation test we found a

negative correlation (r=-0.243, p=0.004) between both variables. This means that there is a

relation regarding the way participants believe in their capabilities and the use of Facebook.

Consequently, this shows the importance of older adults’ self-belief, or how by considering

themselves capable of interacting with technology or being confident in their technical

skills, older adults are more likely to use Facebook.

Therefore, knowing that less skilled users see Facebook as a more complex service

(or UI) than users who believe in their technical capabilities, the effort should be on

simplifying the UI or the way older adults have access to it. This is even further supported

by the fact that, not only non-adopters but also Facebook users considered privacy settings

as the major factor for its limited use, which shows that the way UI are designed is not

even tailored for them. Additionally, it is also necessary to promote this simplicity so that

older adults can give the SNS a second chance.

Taking into consideration that self-belief could be influenced by the limitations each

person has, we also checked for possible correlations between limitations and the use of

Facebook or technology use. In this sense, we found no correlations regarding the number

of limitations reported and the use of technology, neither between the severity of each

ones’ limitations and the use of technology, or the type of limitations and Facebook use.



Technical Expertise Classification Facebook

Non-user

Facebook

User

I use any typical technology available

nowadays4 30

I’m comfortable working with some

technology19 30

I have minor issues dealing with

technology6 13

I have some difficulties with technology 14 20

I have a lot of difficulties with technology 4 1

Total 47 94

Table 4.6: Facebook use vs Technical Expertise)

The only correlation found regards the way motor limitations evolve (in severity) and

Facebook use. The Pearson correlation test indicated a statistically significant negative

value (r=-0.17, p=0.043) which means that as motor limitations evolve in severity, users

tend to use Facebook less. Therefore, older adults with tremors, trouble moving arms,

difficulties grasping a mouse, positioning and controlling a cursor, or hitting a button, will

tend to use Facebook (or the technology which permits accessing Facebook) less than older

adults with just a small portion of these problems, and much less than users with none.

The main reason for this would be the one related with the traditional way of accessing

Facebook being through a PC, where limitations on the use of a mouse typically result

in an inability to use the whole service. Concerning these limitations, solutions should

focus on 1) providing additional or alternative ways of accessing Facebook (like the TV),

2) making alternative modalities of interaction available, particularly voice, another type

of gestures or a combination of both, or 3) providing adaptation mechanisms capable of

attenuating the differences. All these three suggestions were previously identified in the

literature survey phase as important domains to consider when targeting the design of a

more inclusive SNS.

Considering the possible influence of education, or the possibility that older adults with

more or fewer studies have distinct probabilities of using Facebook, we also performed a

Pearson’s correlation test on these variables. However, not only no statistical significance

was found, but the biggest percentage of Facebook users (78%) were older adults with the

12th grade (middle of the education scale), while about 65% of participants with the 9th

grade, a graduate degree or a Masters/PhD were Facebook users. Therefore, education

seems not to have a relevant role in influencing the use of Facebook.

We also checked for possible differences between female and male participants in this

study, i.e. a possible influence of gender on Facebook use. Although a slight difference

favoured man (69% uses Facebook against 65% in woman), this difference was non-

significant. This partially contradicts previous findings that older woman might be more



Facebook Non-user Facebook User

Alone 14 18

One other 22 35

2 to 3 other 10 25

4 to 5 other 1 13

more than 5 0 3

Total 47 94

Table 4.7: Facebook use vs Number of people the participant lives with

predisposed to use this kind of services than older man, supporting the argument that

gender has no effect on Facebook use and that both should be targeted similarly.

Lastly, we also inquired about the possible influence of living alone or accompanied,

or how the number of people with who the older adult lives with, influences the use of

Facebook (table 4.7). Based on the results obtained by cross-checking both variables,

we performed a Pearson’s correlation test and found a significant correlation (r=0.233,

p=0.006). This showed the smaller the number of people there is in a household the least

probable is for an older adult to use Facebook, i.e., the more isolated (at least physically)

older adults are, the less they tend to use Facebook. This suggests the way an SNS like

Facebook is currently designed approximates more to a tool for expressing existing social

connections - as valid as a phone call, exchanging messages or even meeting face-to-face

-, than a tool for regaining contact with family members and friends. Still, even though

designed measures should be considered to support the latter better, serving as a tool for

fostering already existent social relationships is also one way of fighting social isolation.

4.3.1.2.4 Distinct Groups During the analysis, we looked for some distinct groups of

users which could provide additional findings regarding the differences between Facebook

users and non-users.

First, following previous findings concerning older adults who live alone, we checked

for differences between the ones who live alone and do not use Facebook (group 1)

and the ones who, in the same conditions use the SNS (group 2): firstly, about 28% of

participants in group 1 consider themselves capable of using any typical technology without

difficulties, while this percentage raises to 60% in group 2; secondly, and concerning age,

all participants in group 1 were older than 71 years of age, with about 43% being older

than 80. This value decreases significantly in the second group, with only about one fourth

(26%) being older than 71 and less than 3% (2.6%) having more than 80 years of age. These

numbers suggest the way participants consider themselves capable of using technology and

age, as factors which can influence the use of Facebook, especially when living alone. The

first suggestion regarding technological self-believe is, in fact, a confirmation of previous

findings on this user study. The second provides evidence that, if we consider isolation as

a sum of living alone and also not connecting to others using Facebook, the most isolated



users are also the most aged ones.

Additionally, when comparing these values (age and self-belief) regarding older adults

who live accompanied and make use of Facebook and the ones that in the same conditions

do not, we found that both factors are more critical, the less company, the older adult has.

Therefore, the oldest older adults are also the ones who believe less in their capabilities

concerning technology, and that leads them not to make use of Facebook.

Finally, we checked for differences concerning the reasons appointed for not using

Facebook and the way they vary with older adults self-belief in their technical capabilities.

Relevant results were found for two of these reasons: the first regarding “I do not use

Facebook because of privacy issues” where the value remained constant regardless of the

technical expertise of the participants, which suggests that privacy issues are not related

with technical skills and are a transversal problem to the whole older adult population;

and secondly the concordance with “I do not make use of Facebook because I don’t know

how to use it” increased as older adults consider themselves less technologically skilled.

These are not only further confirmations regarding the importance of self-believing in their

technical skills and the limitations posed by privacy issues, but also suggests that Facebook

does not look simple enough to be used by anyone, and that the way it is designed leaves

both less-skilled and less-brave older adults out, or simply frightens these older adults.

4.3.1.2.5 Older Adults use of Television Concerning the use of TV, and when asked

about the main reasons that characterize its use by the older population (figure 4.3), almost

everyone agreed on it being easy to use (mean of 4.56), on the importance of watching

their favorite programs and shows (4.14), or getting information about what is going on in

the world (4.12). This shows the importance of using TV both as a means of accessing

new information and as a vehicle of entertainment. These uses are not well aligned with

Facebook’s main purposes, which focus on communication with family and friends. In fact,

when asked about the possibility of using TV for distinct purposes of the ones they are used

to, only about 26% of the participants showed interest in using it for accessing the Internet

or to get information about family and close friends and an even smaller percentage (about

16%) showed interest in accessing Facebook. Additionally, and also concerning the latter,

about 57% of participants were in fact against it.

Although these preliminary findings could be interpreted as an indication that TV

may not be an appropriate alternative for reducing the gap between Facebook and older

adults, additional findings can somehow rebut this. We have to take into consideration

that participants were not familiarized either with the concept of smart-TVs, or accessing

the Internet through TV and, as we have seen earlier, it is common that by not having

experienced new ways of using the platform they refuse using it for anything else than what

they are already used to. Additionally, there are also differences regarding the acceptance

of these new ways of using TV when comparing Facebook users and non-adopters, with



Figure 4.3: Reasons behind the use of TV

the mode increasing by one value. This also supports the previous argument, showing that

the most knowledge older adults have about technology the most receptive they will be

about adopting it, or at least, experimenting with it.

Summing up, the fact that older adults use the TV for different purposes of Facebook,

that is, mainly for entertainment, might discourage its role in targeting social isolation.

This might even be aggravated by older adults’ tendency for, not only rejecting the use

of technology they do not know for purposes they know well (like the use of Facebook

for keeping in contact with family and friends), but also rejecting the use of technology

they know well for purposes they are not familiar with (like the use of TV as a bridge for

that Facebook access). However, at least for older adults aged more than 80 years old who

live alone and are generally not familiar with Facebook in its traditional form, integrating

Facebook features in a simplified form on the TV might help. Furthermore, while older

adults do not experience solutions, we cannot know for sure, and with the rise of smart

TVs, it might be only a matter of time until Facebook reaches the big screen. When this

happens, older adults should be considered the main use case for this transition.

4.3.1.3 Summary

We inquired older adults about habits and concerns related with the most used SNS and

understood differences between users and non-users of the service. We also explored

older adults’ receptiveness to the possibility of using alternative technology like TV, to

facilitate Facebook adoption. First of all, results supported previous findings (presented

in section 4.2) regarding privacy, family and age-related differences. They also provided

further indications regarding a strong necessity for simplifying the design of the SNS when

targeting the older population. Additionally, new relevant findings showed that believing



in their technical skills, and having different degrees of motor limitations can make a

difference regarding Facebook adoption for this segment of the population. Finally, there

were differences between older older adults and younger older adults regarding social

context, with the first being characterised as more isolated and as more receptive for the

use of TV as a vehicle to use social tools.

4.3.2 Generating Discussion on Older Adults’ Facebook Use

Following the questionnaires, we wanted to understand further how Facebook can be

improved for older users, and emphasise some discussion about different habits, typical

uses, limitations and likings concerning UI. To achieve those goals, we conducted paired

interviews (and focus groups) in two different institutions, a senior health-care institution

and a senior university. In both, we wanted for participants to enter in discussions around

these thematics. Additionally, by targeting distinct institutions, we wanted to understand if

these older adults have different habits concerning technology and SNS.

4.3.2.1 Procedure

In the senior health-care institution we conducted paired interviews with 20 participants

over a period of three days. In the senior university, it was not possible to gather participants

in pairs to conduct interviews, and a single focus group session was performed instead

following the same line of discussion designed for the interviews. 9 participants participated

in the latter with a session running for a 1 hour period.

On both processes participants were previously asked if they have ever used Facebook

(or seen it being used by others), and about their relationship with technology, i.e. which

devices they use in their everyday life. Following that, the discussion was based around

the domains previously identified in the previous section (4.2. Gathering Literature

Requirements) - family, privacy, grouping, types of media, etc. - with a special focus on

the most relevant and irrelevant functionalities, and what could be improved considering

older adults use. For non-Facebook users, the discussion also focused on the reasons why

they do not use it (or have knowledge about it), receptiveness and contexts for using it in

the future.

A total of 29 people (22 female, seven male) participated in the study. Their ages

ranged from 55 to 93 years old (AVG = 69.5, SD = 13.0). Like in the questionnaires

(4.3.1.Inquiring Older Adults About SNS), we also did not want to exclude individuals

who will soon be older adults. Considering the different institutions, the 20 participants

from the senior health-care institution averaged 71.8 years old (SD=15.2), while the 9

participants from the senior university averaged 64.6 years old (SD=1.8).



4.3.2.2 Results

While the discussion on both focus-groups and interviews was encouraged towards the

previously identified domains, for a better understanding of the conversation flow we

introduce the results grouped into more general contexts. Thus, after providing detail on

Facebook use and familiarity, we group results in reasons for using and not using this

service and relevant functionalities. In each of this, we acknowledge any relation with the

domains identified in section 4.2. Gathering Literature Requirements and also establish

relations with the results obtained in the questionnaires (4.3.1. Inquiring Older Adults

About SNS). After these, results are also presented concerning UI aspects and the way

these span across the identified domains, and finally regarding photo and media relevance.

4.3.2.2.1 Facebook Use and Familiarity From all 29 participants, 18 (62%) knew

what Facebook is while 11 were not familiar with it. From the ones who knew about it,

15 (83%) stated they used it regularly at that time. This means that just over 50% of the

participants in the study used the SNS. Additionally, the ones who used Facebook were on

average 62 years old, while the ones who did not, were on average 16 years older. This

tendency confirms what we’ve seen earlier in the questionnaire regarding older older adults

not using SNS as much as younger older adults. Concerning the frequency of use, about

47% of the participants that were on Facebook, used it more than once per day, while 53%

accessed it once, twice or three times per week. This means that from all older adults

interviewed or that participated in the focus groups only about one quarter were daily users

of Facebook.

4.3.2.2.2 Reasons for using Facebook Regarding main reasons for using Facebook,

half of the participants (10) of the Senior Health-care Institution referred to talking with

family and friends using the chat functionality. This indicates that direct communication,

identified in previous requirement gathering phases, as a very necessary domain. Addition-

ally, the second most appointed reason was the sharing or posting functionality, because

of the ability to share what they feel through Facebook network (4 participants), or see

what their friends are saying (2). Seeing photos (photo relevance), give advice to family

and friends (direct communication and reciprocity), finding events (off-line role), and

consulting information related with groups (grouping) were other reasons indicated by one

participant each which also refers to previously identified domains. Finally, the possibility

of knowing new people, although referred as irrelevant in the literature and questionnaires,

was appointed as the main reason for using this tool by one person, as also were reading

news, and playing games. Still, regarding the latter, more than half of the participants

complained about game-related activities and notifications issued by Facebook, turning

gaming into an unpopular functionality. Additionally, more than 50% (8 in 13) of senior

health-care participants who have not adopted the SNS expressed a desire to accompany



their family members life on virtual tools (even considering almost all of them never had

any contact with personal computers before). Again this reinforces the importance of a

family-based dimension when targeting older adults use of SNS.

Regarding participants from the Senior University, almost all said to use Facebook

more for consuming information than for updating their status or sharing information with

others. However, and contrary to this argument and in line with the senior health-care

participants’ results, when asked for the main function used when using Facebook, the

majority (5 participants) agreed on functionalities related with sharing information with

friends and family. Still, regarding these participants and Facebook adoption, other fact

popped out as curiously relevant. Two participants started using the tool for reporting

activities (like registering a complaint on a specific page only available on Facebook)

but ended up using it almost every day for every other purpose. In fact, one of these

participants stated “I only use Facebook for complaining, and to say bad things about the

government” and also “I don’t like Facebook at all” but admitted using it every day for any

reason (she also did not use it for family purposes as she stated not having any relatives

alive).

4.3.2.2.3 Reasons limiting Facebook use When asked about the reasons for not adopt-

ing Facebook, most of the participants in health-care facility appointed security reasons, as

they do not know how to change the privacy policies and restrict to who, and from who,

they post and read content. Additionally, one participant indicated not having any family

members left who she would be interested in keeping contact with as the main reason for

not using the application.

When discussing reasons for not adopting or using Facebook more frequently, all

Senior university participants except one, agreed on the problem with privacy issues as the

main difficulty when using the service. This shows that regardless of older adult’s activity,

privacy is the number one problem when accessing or using Facebook, as it was identified

in both previous requirement stages. Other issue appointed, was getting frequently lost in

the application because it has too many functionalities on the same UI (3 participants). This

shows, once more, the lack of simplicity that the Facebook UI presents is critical for this

population. Two users also reported having difficulties when uploading and posting photos:

“We have to do a lot of things, and I get lost in the process”. One of them admitted she

ended giving up from posting photos on Facebook. Also, they are afraid of getting addicted

to Facebook, as they see others do: “I see people who used to be close to us, without any

time available because they are on Facebook almost all the time”. All also agreed when

one participant stated “It makes me sick, those people who publish in Facebook everything

they do: ’Now I’m eating, now I’m going to the bathroom, now I’m seeing a movie’. It’s

ridiculous!”. One of the participants agreed she used Facebook all the time to share her

poetry with her friends. However, she disagreed on that being necessarily negative: “I’m



there all the time. But I think it’s a good thing since I finish writing more than I would

without it.”

4.3.2.2.4 Relevant Functionalities When asked about the most important functionali-

ties on Facebook, Senior University participants indicated the necessity for a more practical

(simple and easy to use) UI, and the capabilities of direct (virtual) contact with others,

and organising events with family members, as the main ones. All these reflect domains

identified in the literature requirement phase (4.2), respectively the need for simplifying

the UI, fostering direct communication, and focus on functions that have an off-line role.

Concerning participants from the Senior Health-care institution, the majority (5 partici-

pants) agreed on “talking with friends” as the main function. Hence, particularly relevant

here is the fact that both chat and posting on each other “wall” were appointed as main

functionalities, but both also regarding friends and not only the family. Other main func-

tionalities appointed, were the ability to find and keep in contact with old friends and

family (2 participants), and receiving notifications and messages on each others birthdays:

“I felt better when I received all those messages, somehow I felt less lonely”. Moreover, all

participants agreed on the less important functionality as being the ones offered by games,

explaining they did not understand how to play the majority of them because “they are

too complicated”. There were also major complaints about invitations received all the

time from games they do not know or want to play and about advertisement all over their

pages. These opinions are in line with the ones referred by the previous group, however

a higher focus on family is present on the first group: only about 30% admitted to using

Facebook focused only on family, and agreed that “talking with friends” meant both family

and people with who they share no family ties. Finally, these older adults also wanted to

restrict interaction only to people they know and had little interest in making new friends

online. This indication also reflects findings obtained in the literature concerning the notion

of “knowing new people” and how it can be hard for older adults to engage in new social

relationships through SNS.

4.3.2.2.5 User Interface At the end of the interview, each health-care facility partici-

pant was inquired about the current Facebook UI, i.e. the main difficulties accessing it and

what they would change to make it more accessible to the older segment of the population.

Main difficulties pointed out were the frequent layout changes operated by Facebook,

which obligate them to a new period of learning, along with the privacy definitions, the

existence of non-native language text, and the difficulties when publishing or uploading

photos. Again this reinforces the need for several improvements to be made on the Face-

book UI, along with the relevance of other domains which are indirectly related to this,

respectively the need to simplify privacy mechanisms and the relevance of photos as the

main type of content.



Additionally, when asked about missing functionalities, the same participants ap-

pointed the need for more speed (“should be faster”), the inclusion of the type of con-

tent/group/thematics filters, and the possibility of giving greater relevance to favourite

contacts, or people with which participants interact more. Two older adults also stated the

need for a help mechanism - “because sometimes we get into situations we do not know

how to get out off”. Some of these functions already exist on Facebook, but the fact that

senior participants did not know about it, shows they should be accessible in a different

way so they could use it. These are also related to domains identified in the literature

requirement phase and in the questionnaires regarding the way Facebook UI should be

simplified and allow for better “grouping” functionalities.

In the same way, when asked about the Facebook UI and if they would like to have

a different one centered in older adults specificities, some of the Senior University par-

ticipants worried they would be treated differently and opposed to the removal of some

functionalities they do not use: “Right now we do not use it, but probably in the future we

will want to use those features”. Other half recognized they would enjoy a more simple

and personalized UI and once again appointed privacy issues and functionalities as the

main factor of improvement: “I do not want to see information about people I don’t know”,

“If I could not be apprehensive while using it, I would use it all the time, because there is a

lot of information there that is not possible to find outside Facebook”. One participant also

said “Old people reject what they don’t know well. If we could have something more easy

and practical to use, focused on our interests, we would use it a lot more”.

Moreover, when specifically asked about the features they would expect to find: they

indicated the necessity of having some kind of cultural feed related with their age - they

wanted this to be automatically created for them -, and the possibility of receiving more

information related with health homes and services or social events. Also, the majority

expressed the necessity for some kind of functionality they could use to filter the type of

content they receive/read as well the possibility of posting different information to distinct

groups of friends or family. They recognised Facebook already makes this available to

them, but they don’t know how to use these functionalities.

Once again, if we compare these necessities with the functionalities suggested by older

adults in the health-care institution, we notice a big shift from family oriented to cultural

and health-oriented content. One possible reason for this is that more active older adults

are less dependent on their family and seek information related to a more active lifestyle,

feeling less isolated and dependent on family members. Still, in this context, one Senior

University participant expressed the desire for the capability of remembering the childhood

of every family member and other two were interested in the possibility of having some

family oriented functionalities, such as a family news-feed. Both these indications also

reflect domains identified in the literature requirement phase, more specifically family and

reciprocity.



4.3.2.2.6 Photo and Media Relevance Finally, and as we previously stated on the

previous domain, when discussing the preferred type of media on Facebook, almost all

participants agreed with each other on images as the type of content which attracts more

attention. Conversely, only 2 participants said to pay attention to videos, with all agreeing

that as videos take too long to watch, they tend not to pay attention or giving up on watching.

This tendency also supports findings regarding prioritising photo content obtained in the

previous two requirement gathering steps.

4.3.2.3 Summary

We conducted interviews and focus groups on characterising older adults Facebook use

or the reasons for not using it. One of the main conclusions was that Facebook usage

among older adults could be increased since older participants were willing to use it

and do perceive several benefits associated with it. However, for this to happen, special

considerations must be given to specific issues related, mainly, with privacy, which is

not only the main reason for this population to not adopt Facebook but also the main

problem for the ones who already use it. Additionally, studies also report Facebook UI, to

be tailored for older adults, could be improved in what regards its family focus, grouping

and filtering functionalities, and simplification of some functionalities and UI complexity

aspects. We also found further support for several of the domains identified through the

two previous requirement gathering phases.

4.3.3 Generating Discussion on Older Adults TV-Related Use

As a final step in gathering older adults requirements regarding the use of SNS technology,

we considered paramount to characterise them concerning their technological expertise

and age-related limitations. Additionally, and taking into consideration previous findings

(of both previous requirement steps described in this chapter, and the previous chapter),

we also wanted to access further their opinions regarding the use of TV as a technology to

access SNS. To that end, we conducted additional semi-structured interviews with older

adults. Again the main reason for using this method is because it allows us to collect

more data than simple questionnaires as when necessary, the interview conductor can ask

reasons for a particular response.

4.3.3.1 Procedure

We started by inferring what distinguishes older adults who use Facebook from the ones

who do not and what are the main reasons for non-adoption or limited use. For that

purpose, we asked profiling questions concerning gender, age, education and household

composition, as well as self-characterisation regarding general and emotional health, social

feelings and age-related impairments. Next, we asked questions related to the TV as a



Age categories Frequency Percentage





More than 80 1 3.1

Total 31 100.0

Table 4.8: Participants’ age range distribution.

vehicle for older adults to adopt SNS like Facebook, i.e. we tried to understand if older

adults are receptive to use social applications on TV. Continuing the discussing around

the TV, we also inquired about which modalities older adults prefer for interacting with

this platform. Finally, the last points were related with the concepts of adaptation and

personalisation, i.e. which of these concepts is preferred by older adults, and what issues

should be taken into account in the implementation of each concept.

A total of 31 participants (11 male, 20 female) participated in the semi-structured

interviews. These had a duration of approximately one hour each. All participants were

volunteers and more than 60 years of age. Age was not asked directly. Participants

indicated which of the age ranges they were. There were six different categories and table

4.8 shows how participants spanned over those categories.

Anonymity was kept at all times, with each user being identified exclusively by an id

number. All data was saved in a secure repository. On a last note, although specific ques-

tionnaires for accessing technological expertise and other skills’ characterising instruments

already exist we decided against using such instruments due to 1) the increase it would

represent in what is already a long interview; and 2) most of them already being based on

self-reported data nonetheless.

4.3.3.2 Results

Results from the semi-structured interviews are presented below. Special consideration

is given towards answering each of the five main fields already described: profiling ques-

tions, distinguishing users and non-users of Facebook, TV-SNS receptiveness, preferred

modalities for TV-SNS interaction, and adaptation and personalisation concepts.

4.3.3.2.1 Age-Related Limitation About 93% of the participants indicated having

visual impairments, 36% hearing disabilities and about 33% complained they forget events

that happened some time ago. In a Likert scale of 1 (nothing satisfied) to 5 (very satisfied)

participants averaged 3.55 (SD=1.31, MED=4.00, mode=5) regarding social satisfaction.

Concerning the use of technology, participants reported difficulties, with only 23% stating



Figure 4.4: Differences between users of Facebook and non-users of this SNS.

to be comfortable using some technology and about 52% having some to a lot of difficulties

interacting with technology.

4.3.3.2.2 What distinguishes older adults who use Facebook from those who do

not? From the 30 older individuals interviewed, about half (48.8%) used Facebook

in their daily lives. This serves as an appropriate basis for a comparison between the

two populations (Facebook users and non-users). We first observed there were no major

differences between these two groups regarding age-related limitations. Following, we

found differences related to the age range of both. An association showed the younger older

adults population was the one who used Facebook more (graphic A on figure 4.4). This

association was particularly visible from 60 to 70 years of age, as a percentage of Facebook

users drops. Further, there was also an association between the use of Facebook and the

way participants appreciate using a PC (graphic B on figure 4.4), meaning that non-users

liked PC less than Facebook users. In the same way, non-users also liked TV more than

Facebook users (graphic C on Figure 4.4), and they would enjoy less accessing it on a TV

context than Facebook users (graphic D on Figure 4.4). This can be partially explained if



we consider that during the interviews it was observed a tendency for non-users to reject

all alternatives for accessing Facebook. The main reason for this population group to not

use this SNS were related with privacy issues (“I think that Facebook is not safe. How

do I know if talking to a person on Facebook, that person is the only one that sees what I

write”). Another reason appointed was related to negative feelings about what people do in

the SNS (“I do not like Facebook, because people expose their lives there”).

These results showed that younger older adults use Facebook more. Users of this

SNS prefer using the computer more than non-users, and vice-versa concerning the use

of TV. We also found that the reason for non-users of Facebook to reject all alternatives

for accessing this SNS are related to privacy issues and negative feelings about the way

that people expose their lives. These findings support both previous findings regarding

the use of Facebook by this population (like Gibson et al. (2010); Cornejo et al. (2013);

Norval et al. (2014) in section 4.2) and findings from the previous studies (section 4.3.1

and section 4.3.2).

4.3.3.2.3 Are older adults receptive to use social applications on TV? The just re-

ported results on the usage of Facebook and its relation with PC and feelings towards TV

(chart B and C on Figure 4.4 respectively) also allow us to discuss that TV could be a way

for non-Facebook users to adopt the SNS. By not using a PC, they currently can not make

use of Facebook, but their general appreciation towards TV could make this context ideal

for the use of SNS. However, when considering their perceived interest in doing so, they

rejected it (chart D on Figure 4.4). Still, as we have seen in the previous paragraphs, this

rejection could be strongly associated with bad feelings about the word “Facebook”. In

this sense, Smart-TV technology could be one way for older adults to access SNS like

Facebook. However, this technology has not yet reached the older population: 94% had

never used Smart-TVs, and 71% did not know about the concept. Still, when we explained

what it was about, they were very enthusiastic. Some of them even said they would buy a

Smart-TV if they had the chance. From several tasks they would like to perform using one,

the ones related with SNS, such as viewing photos and videos of family and close friends,

creating events with these groups and sending messages were highlighted. Furthermore,

they were also interested in talking to family and friends through video conference, as this

would allow them to see people on a larger screen. One participant even said “So I could

talk and see my granddaughter on TV when she went to Covilha (a distant city)”. These

results are an important outcome of this study as they constitute an indication that older

adults might want to perform social tasks through the TV.

Moreover, older adults enjoyed the idea of recording and sharing TV content with

family and friends if those contents are related to news, documentaries and debates. In fact,

some of them already do this without the technology: “I’m watching TV and call friends

to tell them to switch to what I am seeing”. This leads to the notion that including a TV



content sharing functionality on a TV-based SNS can be a good solution. This also provides

excellent feedback to the notion that older adults are receptive to use social applications

on TV: through a TV-based SNS, older adults have the opportunity to share TV contents

with their family and close friends directly. Finally, some older adults proposed a new

functionality. As they have plenty of printed photos laying around in their homes, they

would like to share them with their family and friends through the TV. Thus, the possibility

to digitise their digital photos and share them through their online network is one that

should also be considered.

These results clearly show older adults’ interest towards using a TV-based SNS ap-

plication, as they want to make social tasks on TV, such as talk with their family, and

sharing TV content and printed photos with their family and close friends. These findings

contradict preliminary findings regarding the use of TV for mediating access to SNS or

how older adults reject this technology (section 4.3.1).

4.3.3.2.4 What modalities older adults prefer for interacting with the TV ? When

asked about the need for help when interacting with TV-based SNS, older adults showed

interest in a broad range of alternatives. The preferred manner of assistance was voice

feedback (81%), followed by help through a VC (avatar) (71%) and textual help (55%).

The preference for voice support was justified as being the one closer to the traditional

ways of communication they use daily. Some participants also underlined that having an

application talking to them could make them feel less alone. Several structured questions

showed the preferences concerning seniors’ interaction with a TV-based SNS. Figure 4.5

shows older adults prefer both the traditional way (RC) and speech for this interaction.

Additionally, they are also open to interacting through gestures. In respect to new ways

of interaction with the TV, it is important to highlight older adults found interesting to

have a tablet as a device to help in two distinct situations: when they can not see specific

TV-content they would like to have a tablet displaying it at a larger size (M=3.81, MED=4,

mode=5); and for showing content related with what they are watching, meaning providing

additional information about TV-content. (M=4.00, MED=4, mode=4).

These results show that older adults are strongly interested in having alternative ways

of interacting with a TV-based SNS, such as voice and gestures, at the same time as they

preserve the traditional interaction (RC) and are helped by contextual information in both

audio and visual forms. Therefore they confirm previous indications regarding the use of

multimodal features when targeting older adults TV use (chapter 3).

4.3.3.2.5 Concerning the concepts of adaptation and personalisation, which is pre-

ferred by seniors ? The concepts of personalisation and adaptation were explained to

older adults before asking them about any preference. After acknowledging the concepts,

they were asked about the possibility of changing content size, and increase/decrease



Figure 4.5: Older adults preferred modalities for interact with TV.

audio volume within the application, both by doing it manually (personalisation) or being

done automatically (adaptation). Results showed a strong interest in both features in both

concepts averaging 4.39 (SD=0.72, MED=5.00, mode=5) for personalisation and 4.32

(SD=1.013, MED=5.00, mode=5) for adaptation. We also verified that there was not a

clear preference for either one.

We also observed older adults opinions about other conditions related to both concepts.

In general, they found interesting to have a technology tailored to their needs (M=3.52,

SD=1.48, MED=4.00, mode=5) and would not mind if both adaptation (87%) and per-

sonalisation (83%) would make their UIs different from others. However, almost every

participant (90%) argued for the necessity of having control over what UI features could

be adapted. Finally, when asked if they would mind if the technology collected data

about the way they interact with it for adaptation purposes, participants showed tolerance

(M=3.90, SD=1.30, MED=4.00, mode=4). Still, lack of tolerance was generally justified

with concerns of privacy (“I do not know what “they” are going to do with my data”).

These results indicate that seniors have a strong interest in adaptation and personal-

isation concepts, showing a preference for neither. Additionally, they also indicate that

older adults do not mind if adaptation and personalisation make UIs different from others,

although it is imperative that they retain control over what can be changed. Finally, only

a few portions minded about the privacy issues related to the necessity of collecting data

about the way users interact with the system. They also further support previous indica-

tions regarding the value of adaptation or personalization and how they can compensate



age-related declines (chapter 3, section 4.2 and section 4.3.1).

4.3.3.3 Summary

We conducted a study consisting of 31 semi-structured interviews where we sought to

differentiate older adults who use Facebook from the ones who do not, and to perceive the

opportunities for using a multimodal adaptive TV as a way for non-users of SNS to gain

access to them. Results showed older adults are receptive to use SNS on TV. However,

these services need some modifications, with the main ones regarding the provision of

alternative modalities to the RC like speech and gesture. Older adults also highlighted

the importance of always getting contextual help through several outputs and appreciate

the concepts of adaptation and personalisation. Finally, the study also suggested new

features such as sharing of printed photos through TV-SNS and also sharing TV-content or

TV-related activities.

4.4 Recommendations for the Design of a SNS targeting

Older Adults

In this section we build from the domains identified in 4.2. Gathering Literature Require-

ments to, along with the results obtained on both 4.3.1. Inquiring Older Adults About

SNS, 4.3.2. Generating Discussion on Older Adults’ Facebook Use and 4.3.3. Generating

Discussion on Older Adults TV-Related Use discuss and derive recommendations for the

design of a SNS targeting the older segment of the population.

4.4.1 Family

From all the related work which focused on social solutions and older adults, as well

from all the user gathering requirement steps we’ve gone through, it becomes evident the

necessity for SNS to give a clear and central role to family or contacts with whom the user

maintains or wants to maintain a strong relationship.

Thus, taking into consideration several indications from older adults that participated

in all requirement gathering steps, SNS should be designed to better support the main-

tenance of family relationships. This means that SNS should not only support a better

specification of family roles but also consider the implementation of mechanisms capa-

ble of notifying users regarding situations related to their family contacts. Some of the

suggestions provided by literature (section 4.2) include alerting to periods of inactive

interaction or favour interaction contexts with these users as a way of mitigating situations

related to loss of contact. Similar suggestions were referred by older adults in our studies,

like providing a family news-feed (section 4.3.2), the ability to create events with family

members (referred on both section 4.3.2 and section 4.3.3), establish video-conference



calls with family members (section 4.3.3) or even the capacity of reviving old connections

(section 4.3.1).

Moreover, several researchers (section 4.2) implemented alternative ways for grandpar-

ents to communicate with grandsons or other elements of the family living apart, suggesting

that SNS should support traditional forms of expressing family engagement. An exam-

ple of this is the support for grandparents caring (or another family role) or the necessity of

providing solutions for that type of role to be reflected in SNS functions (Hope et al., 2014).

Other support for this recommendation could be related with: media implementations

capable of being translated to “animated gifts” or the integration of “gestures” which

reflect family communication (Neves et al., 2015; Michailidou et al., 2015); functionalities

which can produce an effect close to storytelling (Lindley et al., 2009; Lindley, 2012)

like “reviewing the childhood” of each family member (suggested in section 4.3.2); direct

messaging or chat features (Vroman et al., 2015) (also suggested in section 4.3.3); and the

capability of sharing printed photos of themselves or family members through the online

network (section 4.3.3) . Other solution several times suggested (by every step of the

requirement gathering process) as a way of giving a more central role to the family was to

provide a family group with its own settings to give users the sense of a close network,

available only to family, where safety and company can be found. In this group, older

adults could also perform their active role in the mass communication of family situations

(e.g. illness) to the whole family, contributing to their self-identity and self-worth. As

SNS, like Facebook, already creates a family group by default, this recommendation would

only require the setting of different privacy mechanisms inside the group and, possibly, the

implementation of small functions related to family (like the ones already referred). Which

can develop a sense of connectedness between family members, not only online but also

off-line (Hope et al., 2014; Chen, 2009). On alternative technological contexts like is the

use of TV as a vehicle for older adults to access SNS, the use of features related with that

context and the way older adults are used to use it with their relatives can also be important.

One example is the ability to share what they see on the TV, both by recommending it

through the service or by using service information to suggest an off-line interaction (like

it was suggested in section 4.3.3).

Still, the implementation of this kind of family-based mechanisms is not easy, mainly

because of two constraints. In the first place, technology suffers from an ambiguous char-

acter, being both sanctioned and celebrated when involving familiar obligations. Which

means that by exposing older users to these features, systems can be both favouring and

forcing interactions, with previous findings showing that, sometimes, by not communicat-

ing with family, an older adult could be in fact showing that “everything is alright” (Hope

et al., 2014). Secondly, we’ve also seen that some older adults (e.g. the ones which

maintain a more socially active life, the ones who have no family left, etc.) are afraid

that by giving a central role to family (and close friends), they lose the ability to also



focusing on other friends (like senior university colleagues) (supported by results on

section 4.3.1,section 4.3.2). Both issues could be dealt if SNS functionalities could be

automatically adapted to each older user profile. However as this would require additional

profiling steps (and constitute additional privacy concerns), the most feasible solutions

would be the possibility of opting between the use of family features (section 4.3.2),

or the ability of including “friends” or “ favourite” contacts as part of those features.

Regarding the latter, the creation of these groups should be based on automatic mech-

anisms (like the number of interactions or frequency of interactions), as the Facebook

existing ability for creating groups like these is not used or understood by older adults

(section 4.3.2).

4.4.2 User Interface

In all requirement gathering steps, Facebook UI complexity is appointed as a reason for

Facebook limited use by the older segment of the population.

Reasons appointed are the lack of intuitiveness and the way privacy settings are de-

signed), having too many features which older adults do not grasp the concept (section 4.2

and section 4.3.1) or which cause them to get lost (section 4.2 and section 4.3.2), the way

it is offered to the older segment of the population without considering their different

needs and limitations (section 4.2 and section 4.3.1), or the facts that not all functionalities

are on the native user language, or that older adults tend to have difficulties uploading

photos through the UI (section 4.3.2). Therefore, findings on all requirement gathering

steps suggest that the UI should be the simplest possible. Older adults would be the first

but not the only ones to benefit from a more simplified and obvious UI where it is safe for

anyone to interact, easier for anyone to learn and less-intimating for anyone to try.

In terms of solutions, several ways of achieving this recommendation have been

proposed by relevant work in the area (section 4.3.2): grouping items and functionalities

by proximity would help users who have no experience in finding every function and at the

same time give users only the essential information needed for each context (also referred

as the need for grouping, filters or favorites); giving clear evidence to active areas so that

users can locate them more easily; move secondary functionalities – or the ones less used

by users – to secondary interaction areas so that users can more easily locate primary

functions; change the color and increase the size of header and content, or give the option

to increase and decrease size, so that users with visual impairments can also interact with it.

All these solutions would be feasible and do not implicate radical changes in the standard

UI of an SNS like Facebook. However, identifying the relevance of each function and what

to collapse or highlight, can be tricky, as different users use distinct functions. A solution

would be to start off with secondary features, and automatically “promote” them if they

are frequently used.

Additionally, other solutions have also been suggested by the several user requirement



gathering steps. From our questionnaires (4.3.1) we understood the need for a UI to

foster older adults self-believe on being capable of making use of it and at the same

time the ability to learn through experimentation. For this to be possible, the UI must

be capable of avoiding errors, introducing functionalities step by step, or focusing only

on the most-needed functions. This is also supported by discussions with older adults

(4.3.2), along with the need for the UI to provide “safe points”. These could be a “home”

page or a clearly identified profile page, as these would be obvious points for a user to

access when feeling lost. In the same discussion process, older adults have also identified

continuous UI changes as disruptive for the use of this kind of services and claimed they

should not be forced to change their behaviour or to adapt abruptly to a new UI every time

a new update is made. Therefore, the UI should be constant, and layout changes should

be avoided or made optional.

From both the related work requirement gathering process and inquiries with older

adults (section 4.3.1) also comes the notion that the UI should act as a catalyst by

motivating interaction and keeping users from losing touch with their contacts. To achieve

this, there are several related recommendations which could be considered when designing

a more inclusive SNS: the ability to support the various family roles (also referred in

the previous section); the possibility of more recent items to be shown more frequently,

which could help users at both ends of the interaction to keep motivated and interact

more; providing support for more self-expression and creativity when composing content,

by implementing approaches related with letter writing for older adults (this is further

discussed when focusing on tangible value), or providing users with more interaction

possibilities with already existing contacts like making possible to send gifts and visualizing

those media artifacts. This last design idea would enable older participants to describe and

share details of their histories and everyday lives in a closer way to what they are used to.

As a product of both related research and discussions with older adults (section 4.3.2)

comes a set of findings related to the importance of language in inclusive SNS design.

From related work come several indications that the use of easy-to-understand language

instead of computer jargon is crucial to, not only enable older adults to better understand

and interact with this type of technology, but also to attract new users. This translates

into two distinct recommendations: (1) removing details related to assumptions of prior

knowledge (meaning of icons, features which appear on mouse hover and enter-to-submit

boxes). Although this seems easy to implement, there is still a major problem with this

approach. Changing language terms to which frequent users are used to, introduces a

radical change which may not be well received by these users; and (2) make use of icons

which can suggest the distinct SNS functionalities. However, these icons should avoid both

ambiguous and explicit affordances. As an alternative, augmenting the terms already used

with descriptions or other types of feedback could also help new users without changing

the interaction paradigm too much. Additionally, on both interviews and focus groups



with older adults, we’ve seen the necessity to avoid non-native language UI terms and

content. This means that terms used in the UI should always be in older adults’ native

language. In the same way, content shared by the user’s friends which is not in the

appropriate language should be hidden, otherwise the user will have the feeling that his or

her personal space is being ‘invaded’ with terms he or she is not capable of understanding

(which can eventually result in user’s frustration).

Lastly, there must also be concerns about providing easy-to-understand privacy

options, or on providing interaction to accommodate users’ social, physical and in-

tellectual differences. Still, as these by themselves constitute separate SNS domains,

elaborate discussions and specific design recommendations can be found in their respective

subsections.

4.4.3 Privacy

Common to every requirement gathering step is the problematic of privacy, or how SNS (or

Facebook) privacy settings are the main limitation behind the use of this type of services

both by current older users and by possible adopters among the older segment of the

population. Additionally, both related literature and requirement gathering steps also agree

regarding the way these should be solved.

Thus, from related work findings (section 4.2), privacy settings have to be designed

to be clear and simple, and the SNS UI should focus on facilitating the discovery and

access to privacy settings. One particular example was given on how to achieve this: in

Facebook, modifying the panel in the top right to include text in addition to the white arrow

of the drop down box would better identify the menu. Additionally, several researchers

evidenced the need for changing the way privacy settings can be accessed when sharing

any status - either by making it more evident or changing its size -.

Moreover, both past research (section 4.2) and older adult inquiring (section 4.3.1)

suggested that instead of changing the way users can access privacy settings, designers

should rethink the disclosure and promotion of content and information keeping privacy

values in mind. It was suggested changing default settings to be more private and give

options to reach more contacts (or groups) gradually. By changing default settings to

post only to family or favourites, or even to be only visible to contacts with which the user

interacts most, would be a way to make sure the only way information becomes available

to others would be by explicitly declaring that to the service. Additionally, by doing so,

older adults self-belief in the service would probably increase, which according to results

from questionnaires, would also tend to increase its use. Finally, and also having this into

consideration, the possibility of re-organising content based on the level of privacy or the

confidence in its origin, could also help to make the user feel safer, as the majority of posts

and information he or she visualise would be from their most worthy SNS contacts.

Additionally, results from the interviews (section 4.3.2) have also focused on this



subject, identifying the main issues with Facebook privacy options in the way they are

available or can be configured. Thus, results from this phase revealed Privacy options

“Post to” and “View from” must be simplified with older adults having a set of explicit

options available for specifying who can see what they share every time they want to make

a post, and from whom they want to consume information (instead of privacy options

accessible through an arrow-activated menu). By integrating distinct groups into these sort

of explicit privacy options, SNS would also be following older adults recommendations

regarding focus on groups.

Although findings and research also suggest that changing Facebook UI could introduce

significant changes in the way it is used, and be disruptive for users who are already used

to it, this should not prevent changes. Not only because they represent a crucial step for

older adults to adopt SNS, but also because the majority of “typical” users are also not

fully aware of privacy definitions, and currently do not make conscious use of SNS. For

the ones who are already fully aware of all privacy definitions, and have in fact changed

these definitions since they began using the service, there should be a careful (or optional)

transition to the new implementations, making sure their privacy settings are not changed

in the transition.

Finally, and considering the importance that alternative ways of interaction can assume

in the future regarding interaction with SNS (which is a common finding to all requirement

gathering steps) there will also be the need to tailor privacy settings to fit each mode

of communication or interaction. This is not only relevant when considering ways

of making use of privacy settings with distinct modalities of interaction, but also when

considering ways to automatically share older adults context, like the ones used by (Cornejo

et al., 2013). Distinct ways of interacting should not be intrusive to the privacy settings

defined by a user. However, for distinct interactions must be considered distinct privacy

settings (for example, when considering mechanisms which share information regarding

the presence of an older adult at home, this information should only be available to the

family, even if the user has changed default privacy settings to the public).

4.4.4 Photos and Media Relevance

When answering questions related to the content they can find on Facebook (section 4.3.2),

older adults clearly identified photos as the main content that gets their attention when

browsing the news-feed. And in their understanding, more relevance must be given to

Photos or Images than to other types of content. Two possible solutions to achieve this

is to make photos occupy a bigger range of the screen, or to support the existence of a

photo news-feed. Additionally, as older adults also reported difficulties when uploading

photos or images to Facebook (briefly mentioned in the User Interface section), special

concern should be given to providing ways to simplify and increase the intuitiveness of

uploading photos into the news-feed. This relevance or preference for photos is also a



subject widely discussed over past literature (section 4.2). In fact, several researchers have

drawn recommendations which focus on simplifying SNS interaction with photos: one of

these is to automatically upload photos associated with events and family or making

it easier for older adults to share this type of information in the form of photos. This is

seen as a way of boosting interaction between older adults and their relatives; another

recommendation is to provide interaction around photos like the use of natural touch

and gestures to perform like, share or comment actions, which could also increase older

adults use of this type of systems, making them feel more involved with the system and its

members.

Other solutions were suggested by distinct stages of user requirement gathering which

are related both with this domain and the tangible value domain (described further ahead in

detail). Therefore, when discussing photo relevance (section 4.3.2) and TV social use (sec-

tion 4.3.3), older adults suggested replicating real (physical) photos through the SNS or

in other words, the possibility of sharing printed photos, as a way of remembering family

members childhood. Additionally, they also suggested (section 4.3.2) the implementation

of passive social consumption modes based on photos (like the automatic creation of a

family photo slide-show using family members photos shared on Facebook). Both ways

can be seen as potential ways of spurring social interactions through the use of photos.

However, the use of photos is not the only one to have been highlighted in SNS related

studies. The use of videos has been appointed both by related research and user studies.

However, findings on this are somewhat contradictory or at least bi-faced.

In recent studies, there were indications that video clips can be better than photos for

expressing non-verbal social cues such as laughter, smiles or personal gestures (Kim et al.,

2013). Not contradicting the use of photos as the main type of media, these researchers

point to the possibility of videos also having a role in making SNS more inclusive. However

the generation of this type of media and the way it can be shared over the Internet has to

be made more accessible for older adults, and in that sense they suggest the possibility

to provide short-clip sharing through the direct use of the device’s camera as a sharing

functionality of SNS. This matter seems not to be consensual, since when being interviewed

(section 4.3.2) older adults showed some indifference regarding video content, describing

it as “boring” or as “a waste of time’, indicating they would like to give less relevance to

videos than to other types of content. Again opposing this, when asked about the use of

social features through TV, older adults referred to both media types as the ones they would

like to watch (section 4.3.3). Thus, taking into consideration these differing opinions,

focusing on functionalities like providing options for reducing or increasing the size

occupied by videos on the Facebook news-feed, or providing a video-only news-feed

could satisfy older adults distinct preferences. The possibility of supporting combined

media items which make use of photos and audio was also a matter debated by several

researchers (4.2), and which could also favour older adults interaction by compensating



visual impairments and enriching the interaction.

Finally, another way of considering this duality between photo and video relevance

was also suggested during the interviews (4.3.2), with the recommendation that news-feed

should allow filtering by type of media. This recommendation is also related with the

“grouping” domain identified as relevant when considering SNS inclusive design (and

which is later described in more detail).

4.4.5 Multimodality

The use of self-expression and creativity to describe and share personal details of each

one’s history and everyday lives is fundamental to raise the adoption of SNS by the older

population (section 4.2). Additionally, both phases of requirement gathering have shown

that SNS solutions focused on including older adults should focus on making alternative

modalities of interaction available, particularly voice, but also gestures and modalities

related with the typical technological context like RC for TV. Thus, multimodality can not

only compensate for age-related declines but also provide this bridge.

Related work findings (section 4.2) also indicate that for alternative modalities to be

used, special concern should exist regarding their intuitive use. If users have to learn

how to use a modality, they will not adopt the service. Therefore, SNS should make

alternative modalities of interaction non-instructional so that users (older or younger

ones) may simply “pick up and use” any way of interaction. At the same time, interactions

should be automatically associated with functionalities to avoid the feeling of using

a replacement modality. Both of these recommendations heavily depend on different

contexts of interactions, and on the characteristics of each modality, and therefore, their

implementation may depend greatly on user testing or in the definition of some profiling

mechanism. However, the first step of user requirement gathering (section 4.3.1) also

indicated that older adults typically make use of alternative modalities for tasks they

cannot accomplish using the traditional ones (or in other words as a way of compensating

declines or age-related differences like tremors). For that reason, traditional ways of

interaction should be kept as the main interaction mode. Additionally, by augmenting

those results with findings related with older adults preferred modalities in a TV-based

SNS (section 4.3.3), we can verify that while voice feedback and speech are the main

alternatives for both text input tasks or tasks not achievable with RC, gestures should

be provided as an alternative for spatial and navigation tasks. In fact, the necessity to

adapt availability of modalities to each user is also indirectly suggested by some of the

selected literature (section 4.2). Still, and contradicting this recommendation, many may

say that, if modalities are designed and integrated intuitively, users will use them if they

feel the need or not even notice them otherwise.



4.4.6 Direct Communication

As a result of both the related research regarding the use of SNS and solutions developed

for older adults to keep in contact with relatives (section 4.2), and all inquiries and

discussions surrounding the subject of communication on SNS and social use of TV

(section 4.3.1, section 4.3.2 and section 4.3.3), it became evident the need for SNS to

favor direct forms of communication. The most immediate solution suggested by older

adults was to favour “chat” functionality. Thus, chat should have a central role in SNS.

This could be done either by giving chat functionality higher importance in the UI or by

placing other SNS functionalities around it. In the same way, other possibility suggested

by related work (section 4.2), was to create incentives for more direct communication

in SNS. This could be accomplished by issuing different notifications related with chat,

establishing simple connections between chat and other functionalities (like commenting

a post), introducing simple features related with family or close friends contact, or by

making changes so that share and comment features could follow a more direct way of

communication. The focus on these type of functionalities was also supported by findings

of the questionnaires concerning technology use by older adults (section 4.3.1) or in

discussions concerning TV social uses (section 4.3.3). Findings of the first suggested

providing users with more interaction possibilities with already existing contacts or as

a way to make possible the revival of old connections. Findings of the latter suggested

the possibility of giving more importance to video-conference calls with family members.

Still, these solutions would imply radical changes in the way Facebook works, which could

be dramatic to everyone who is already used to its UI. Therefore, solutions should focus

on small augmentations of the UI like keeping the chat expanded, or incorporating it more

visible in each contact profile page. Other less radical or “intrusive” way of achieving

the same goal could be the resurfacing or prioritisation of information as a means of

stimulating communication. For example, making status updates, photos or comments

related to an older adult more visible in their relatives and close friends news-feed would

create more opportunities for interaction without changing the way they use the service.

Additionally, and as further conclusions of the works discussed: “direct forms of com-

munication should be capable of capturing both the permanence of asynchronous

exchanges and the interactivity of synchronous play”(Vetere et al., 2009); while func-

tionalities related with direct communication should be kept interesting and conserve the

properties of real off-line exchanges. Moreover, Michailidou et al. (2015) also supported

findings regarding video-conference (section 4.3.3) evidencing the necessity of supporting

or increasing the visibility of video-chat features as the older segment of the population

uses these for obtaining a more direct and rich online contact with their family or close

friends.

Finally, and as a result of older adults general concerns when interacting with social

applications (section 4.2), it is also important to avoid the creation of new social obliga-



tions, especially when communicating with close relatives. This concern extends to both

sides: older adults do not want to be a burden for their contacts, and at the same time,

younger users do not want to look bad or unconcerned, especially when focusing on older

family members. There are a time and place for communication, and users purposefully

structure their lives and home environments to support this goal, so technology should also

do this. Again, a solution could be to take an asynchronous approach to communication or

restrict direct messaging features to the availability of each contact (and therefore provide

mechanisms for configuring, or making others aware, of these periods of availability).

4.4.7 Knowing New People

All relevant research (section 4.2) agrees that features related to meeting new people should

be avoided or made optional for the majority of older SNS users. This is also identified by

older adults as a main concern when discussing the limited use of Facebook (section 4.3.2).

Thus, taking into consideration that removing these features would require a radical change

in the paradigm of SNS like Facebook, other solutions could achieve the same result. One

way would be to highlight acquaintances over “new friends”. This could be achieved

by reordering the lists of contacts and prioritising the first group of people over the latter

(this was also suggested as a privacy-related recommendation). Another way of achieving

the same result could be by giving less importance to functionalities related with “friends

of friends” or “people you may know”. In this sense, to provide appropriate terms

when referring to individuals the user does not know instead of using familiar terms,

which somehow try to impose a friendship degree, is fundamental. Additionally, it is also

important to provide options for avoiding friend suggestions and unsolicited emails as

these are not only hated by older adults but also by almost every other type of user. Finally,

avoiding, making optional or clearly identifying news-feed content not originated

from Facebook contacts (like photos where these are tagged, or content posted on their

profile by their friends) might be another way of avoiding older adults feelings which are

related with loss of control within their online social experience.

4.4.8 Personalization and Adaptation

From past works concerning SNS (section 4.2), there is the common understanding that

supporting an SNS UI with adaptation mechanisms like adapting interaction with text-

to-speech, text-input, speech commands and other augmentative alternatives, could have a

positive role on helping users with age-related impairments to make use of the SNS func-

tionalities. This understanding also surfaced when inquiring older adults about technology

and SNS (section 4.3.1) and when discussing SNS use with older adults (section 4.3.2).

Both results show that not only (the wide variety of) age-related impairments leads to

differences in the use of Facebook, but also preferences and habits of the older population



- e.g. the self-belief in their technology skills - play a differentiation role. As a result,

SNS and technology through which older adults access these services should offer mech-

anisms which tailor both presentation and interaction to each user characteristics

or in other words, make space for adapting UI to distinct older adults’ user profiles. In

this way, SNS should not only be capable of adapting to both skilled and less-skilled,

younger and older, confident and less-confident, and traditional or less-traditional users.

But also contemplate factors related to older adults limitations, such as physical, sensory

and cognitive impairments, as well as cultural and social characteristics. Examples of

these adaptations are features like reducing the sensibility of touch operations for scrolling

up and down and zooming in/out, increasing the size of selectable items, increasing the

space between UI items, reducing the number of elements shown at once, or decreasing

the number of elements per menu. Additionally, all requirement gathering phases support

that these adaptation mechanisms can also be related with the provision of multimodal

features (already discussed).

Finally, discussions within the context of using these mechanisms on TV-related

social functionalities (section 4.3.3), suggested a strong interest in both adaptation and

personalisation (which is seen as a manual adaptation), showing no clear preference

between these concepts. Results also showed older adults are open to UI changes and to

processes targeting the collection of user information as a result of these mechanisms if

they can control to which extent these modifications are performed, and this information is

used.

4.4.9 Grouping

Focusing on the solutions investigated by several researchers (section 4.2), various recom-

mendations can be extracted regarding the simplification of SNS features by resorting to

grouping or filtering: making group-specific functionalities or options is the necessary

step to not only focusing on meaningful contacts but also to simplify the majority of

functionalities or solve issues related to its use (like the already discussed privacy-related

issues). One example of a feature like this would be the possibility of creating “rooms”

within a specific group (or with specific contacts) so users could share ideas and contents in

a more group-oriented way. Other suggested way of achieving the same result, would be to

simplify a feature’s access focusing on frequency of use. This could be accomplished

by either, simplifying functionalities and changing default sharing settings, or by giving

priority and evidence to each user’s most used functionalities and most frequent contacts.

In fact, converging with the latter, from discussions with older adults (section 4.3.2) several

individuals felt the necessity of SNS automatically creating a “ favourites” group with

the most “interacted with” contacts. This is justified by the fact that although features

like these already exist on Facebook (groups based on location, schools, places worked and

manual creation of groups), older adults do not know how to use (or find) them. Adding to



this is the fact that users do not necessarily know which contacts they like more and share

most content, or talk the most using the chat functionality. Moreover, to give the option of

naming each group can also be a simple way for older adults to get around some issues

related with phrasing and naming identified in previous studies (section 4.2) while making

them feel to be in control of the service, and not the opposite. In all these solutions, the

main challenge is to not radically change the UI so that users do not feel they are using a

completely different service than their younger counterparts.

Finally, but still related to grouping functionalities, when discussing interesting func-

tions with more active older adults (section 4.3.2), they requested the implementation

of cultural and health-based functionalities and groups. Functionalities like a cultural

news-feed, an health-related news-feed, the existence of cultural groups, or the possibility

of creating events related to cultural and health topics were suggested. Again implementing

such features without raising the complexity of the general UI is the main challenge.

4.4.10 Tangible Value

Based on others work regarding older adults social applications (section 4.2), several

solutions emerge regarding the integration of tangible communications in inclusive SNS.

The support for the transmission of handwritten messages as a type of media is the

most obvious one, however for this to be possible requires having some kind of hardware

associated with the service. Alternatively, providing a “letter writing” mode or look-

and-feel for sharing functionalities would be a better solution, as it could be optional

and would make possible for older adults to identify with the process of sharing as a

way of exchanging digital letters (like email). Additionally, supporting features that

enable the virtual representation of real artefacts such as pictures and objects was also

suggested by relevant work as a good incentive for the creation of new interactions between

SNS users, turning the oldest segment of the population into the main vehicle of these

interactions. In fact, this latter suggestion was also supported by discussions around the use

of TV as a vehicle for SNS (section 4.3.3), with older adults suggesting the need for sharing

printed photos around the network, as something that would spur interaction with family.

Moreover, when focusing on interviews and focus groups with older adults (section 4.3.2),

there was also the demand for making games more simple and traditional. This is

because older adults do not play the majority of games on Facebook, and often hate the

fact of constantly receiving updates on those games they do not play. Making them focus

on playing more traditional games (or the ones they were used to play in childhood or with

family) through Facebook, could not only foster online interactions with others but also

bring back memories. So the tangible value here can also contribute for a healthy Facebook

use. Finally, resorting to functionalities with tangible value older adults can relate and

identify with, can also contribute to increasing their self-belief in using the service, which

according to our inquiries (section 4.3.1) is directly related with the increased use of the



SNS.

4.4.11 Offline Role

A lot of literature evidence (section 4.2) showed the line between online and offline

interactions should be very thin, and that SNS should also focus on features to boost offline

interactions, at least, when considering older adults as target users. One possible way of

achieving this would be to give more attention to functionalities related to organising

events, for example by locating them in a more central role regarding the general SNS

UI. In fact, support for this type of functionality is confirmed across all user requirement

steps, with a lot of discussions focusing around the offline value (ideally, the majority of

older adults would prefer meeting face-to-face). Another alternative for inciting offline

interaction would be to incorporate phone numbers into SNS contacts as a way to make

it possible to follow up an online interaction with a phone call. Still, this is a delicate

suggestion, as it should be dependent on each contact’s availability and privacy preferences.

Moreover, the possibility of generating or promoting local events close to the user

could also boost offline interactions with members of the same community or people who

live close by. Finally, mixing both family-focused features (like the ones already described)

with trying to increase offline interactions would be ideal for the context of using SNS to

increase older adults social interactions and consequently decreasing loneliness.

4.4.12 Gender

Although at first some research (section 4.2) reported a difference between male and

female behaviour concerning SNS use, more recent works showed the opposite. In the

same way when inquiring older adults about the use of Facebook 4.3.1 results also did not

find any statistically significant differences between man and woman. This means that

no particular recommendation can be given rather than different genres should not be

targeted differently when designing SNS.

4.4.13 Reciprocity

Issues related to social isolation should be the main target of an SNS like Facebook when

considering older adults, especially by providing opportunities and functionalities to keep

in contact with family and close friends. This need for reciprocity was not only the

subject of previous related work (section 4.2) but also appointed in discussions with older

adults (section 4.3.2) as the main reason for using Facebook. Additionally, and contrary

to what has been defended in the past, these functionalities should focus not only on

family and close friends but also on other friends, especially to provide users with more

interaction possibilities with already existing contacts and for making possible the revival

of old connections. A good example of this was also given with senior university students



involved in the discussions stating they would like to keep contact with their colleagues

through the SNS. In this sense, and although not many recommendations were given,

works like the ones performed by Lindley (2012) and Burke et al. (2011) are the first step

towards this notion. Therefore, designing features which enable reciprocity and multi-

use can be determinant for the adoption of SNS by the older segment of the population,

be it by evidencing and resurfacing content capable of stimulating communications

or by adopting different ways of interaction for each type of user so that everyone can

use social tools without having to adapt to supported modalities of interaction, and that

way supporting the easiness of reciprocating any interaction. Finally, good examples of

this need for reciprocity were also given on the last step of user requirement gathering

(section 4.3.3) with participants stating the need for functionalities like video-conference

or sharing printed photos. These are the type of functionalities that can increase reciprocity

between SNS contacts, be it by making use of face-to-face virtual interactions or by

spurring discussion around photos.

4.5 Summary

In this chapter, we presented all requirement gathering steps taken to identify the main

domains involved in designing SNS targeting the older segment of the population.

We started by describing how we performed a literature survey on both SNS studies

and social applications developed in the last ten years in the context of older adults. From

this, we identified thirteen relevant domains: from the most important related with the

family role and privacy control to issues related with the design of the UI, the importance

of multimodal interaction, to several others focusing on the importance of groups, photos,

gender, reciprocity and tangible value.

Following this, we focused on the description of how further requirements were col-

lected directly with older adults by performing questionnaires, semi-structured interviews

and focus groups on both senior health-care institutions and senior universities. From

these, we also draw relevant discussions on both the use and lack of use of Facebook and

SNS related features by older adults, also establishing relations with technology use and

older adults characteristics.

Considering both requirement gathering phases, we provided a discussion centred on

each of the SNS domains identified, and draw a set of recommendations which aim at the

design of a more inclusive SNS solution.

Finally, it’s very relevant to mention that since these recommendations were published

two years before this thesis delivery, Facebook has evolved a great deal, with some of the

modifications performed on the service going hand-in-hand with these recommendations.

It is the case with the revision on some privacy mechanisms (even though at this stage

they are still too complex which partially contributed to the whole GDPR phenomenon),



some redefinition on the possibility of creating groups and the implementation and focus

on features related with tangibility and direct communication.

Chapter 5

Designing and Evaluating a TV-based

SNS

Following the recommendations drawn regarding the design of a SNS targeting older adults

(chapter 4), previous findings regarding TV (chapter 3), the fact that TV is the most used

technology in this segment of the population (Bobeth et al. (2012)), and having as the main

goal to increase the likelihood of older adults using technology and contributing to lower

their social isolation, we implemented an accessible UI for Facebook based on TV - the

You, Me & TV prototype. Additionally, and also following previous findings on how to

deal with age-related disabilities and differences when interacting with technology, we

also resorted to the use of multimodal and adaptive mechanisms in this implementation.

This chapter presents all steps taken to develop and validate the You, Me & TV

prototype. We start with a description of the UCD process followed for the design

of the application. We follow this with a description of the prototype both regarding

its implementation, main functionalities, multimodal architecture and personalisation

features. After this, we describe the user study performed for the purposed of evaluating

the application with older adults, including the procedure and a detailed analysis of the

findings. We end the chapter with the main conclusions.

5.1 Participatory Design

Based on the design requirements we have collected regarding both the design of TV-based

applications (chapter 3) and SNS (chapter 4) targeting older adults, we developed the

TV-based Facebook prototype following a participatory design approach.

5.1.1 Procedure

In this process were carried out two focus-groups (12 participants), composed of two

groups of 6 participants. Each focus group had a duration of about one hour, and all

participants were volunteers and more than 60 years of age (no specific question regarding

129

Chapter 5. Designing and Evaluating a TV-based SNS 130

Figure 5.1: The low-fidelity prototypes used in the focus groups.

age was asked though). In each focus group, we used low fidelity prototypes with a high

degree of modularity, which allowed older adults to build together the system’s information

architecture as well the distinct screens which later would be implemented in the You, Me

& TV high-fidelity prototype. In this process, the UI was built by moving, placing and

ordering the distinct components on the cardboard (which simulated the screen) (fig. 5.1).

Every time they felt necessary, older adults also had the opportunity of creating additional

components (screens).

These focus-groups were performed at a Senior University and a local community

Gym for the old age. Participants were asked to have active participation in the discussions,

always intervening when they felt necessary or had a different opinion. We started by

explaining the goals and general behaviour of the service to be built. We also asked

participants to answer a brief pre-questionnaire about technology and Facebook aware-

ness (appendix E). These questionnaires made it possible to characterise the participants

regarding technological competence.

The discussion began around functionalities older adults saw as more relevant, and

how they would like to see the menu and contact information presented. It continued

around how to group information and contacts or how to filter media content. Finally,

some discussion was generated around the concepts of personalisation and adaptation.


5.1.2 Results

All participants used TV frequently, and only one stated having difficulties interacting with

it in his daily life. Additionally, only 4 (33%) used a PC, with only two considering it

easy to use. In what respects Facebook use, from the three (25 %) who used the service

frequently, only two considered it easy to use. In summary, few participants made use

of technologies other than the TV. This made them an ideal group to generate discussion

around this platform.

Regarding the results obtained in the focus groups, we divide them into five distinct

domains: (1) preferred functionalities; (2) preferred type of content; (3) how to manage

and organise SNS contacts; (4) general opinion regarding adaptation and personalisation;

and (5) UI composition.

Regarding preferred functionalities, when asking participants what functionalities

they were more interested in, they were unanimous. They see the possibility of getting

information from family members and close friends (seeing the information posted by

these) as the most important one. They also indicated that, for them, all other functionalities

were also indirectly related to this. That is, functionalities like posting, photo viewing,

etc. only made sense if they could be used with the goal of establishing bridges between

them and their family members or close friends. Following this, and after briefed about the

News-feed functionality, participants were asked about the name and if they would change

it, and to what. Opinions diverged with names like “diary”, “events”, and “news” being

suggested. They eventually converged to “diary” (“This is about my diary. Or my diary of

activities”).

Regarding the preferred type of content, participants considered almost every type of

content as important, in the way all of them make it possible to obtain and to transmit

information about and to their relatives. Still, photos and TV content were seen as the

most important. They also added that the importance of the latter is obvious in the way

they feel most conversations between themselves and with relatives evolve or are initiated

around TV-content. Regarding photos, a common statement was “I really enjoy seeing

other persons’ photos!”.

When discussing how to manage contacts, participants suggested an unexpected solu-

tion regarding group management in Facebook. In their opinion, having two main groups

of contacts (family and friends) is not ideal, and most of them felt as necessary to create

sub-groups inside these two. One of the participants stated “I have friends which I see only

when I go to the gym classes and have some other friends which I see when I go to the

coffee house’, so I want them to be presented as distinct groups”. When faced with the

question regarding what happens when a friend belongs to the two groups, he answered

that person should be presented on both sub-groups. Following this, participants were

inquired about how should the UI present members of each sub-group - just by listing the

contact name, name and a photo, or name, a photo and the last shared status. Participants


agreed on the second approach as they felt it was the one that managed to achieve the

best balance between the information needed to recognise a contact and the number of

contacts presented on the screen. As they wanted to see the biggest amount of contacts on

the screen, the third approach was quickly excluded. Additionally, the first approach did

not make possible for them to univocally identify each context, as it is common for several

contacts to have the same name. Additionally, and when presenting information about a

family member, participants also stated they would like the family degree to be presented.

These findings related with sub-groups are in line with previous findings (identified in the

previous chapter) related both with privacy and the necessity of grouping elements of the

Facebook UI to increase its usability and adoption.

Moreover, when discussing concepts related to adaptation and personalisation, par-

ticipants were not capable of providing examples of functionalities they would like to

see automatized or personalised. However, after asking participants to exemplify on the

low fidelity prototype how they would configure the audio volume or the brightness of

the screen, or how they would order the menu options, they agreed they would like to

personalise these, and would also like if these could be automatically set based on their

characteristics. They stated this would be something that would save them some effort and

for that reason would be welcome.

Regarding UI composition, and after discussion evolved around the UI item location,

order, number and naming, we ended up with a clear vision of the prototype to be built.

Figures 5.2, 5.3 and 5.4, show screen representations of a medium fidelity prototype based

on those discussions. The colour scheme adopted for these was coherent with the one used

in Facebook, to ease adoption and to not transmit the feeling older adults would be using a

distinct service.

5.2 You, Me & TV prototype

As a result of the previous participatory design focus groups and supported and informed

by the findings obtained in the previous chapters, the You, Me & TV prototype was built. In

this section, we provide details on each design decision, implementation step and resulting

main functionalities, architecture, and UI.

5.2.1 Design Decisions

From the participatory design focus groups just described resulted the identification of

the most desired features in a TV-based SNS for this age group. That is, a list of the most

relevant recommendations to be implemented in the You, Me & TV final prototype:

Use TV as a vehicle for older adults to access existing SNS services: Older adults

demonstrated interest in performing social activities like talking with family members and

close friends through chat features, video-conference, and visualizing photos and videos


Figure 5.2: Home page with the news-feed Figure 5.3: Group and sub-group manager

Figure 5.4: Volume and font size configura-

tion (personalization)

of both these groups on a Smart-TV (or on a TV with the necessary integrated technology).

This means that TV should be used as an alternative capable of making older adults adopt

SNS already used by their family, like Facebook.

Photos and TV content should be the preferred types of content to be used in an

SNS targeting Older Adults: Photos trigger nostalgic feelings which lead to the creation

of new social opportunities (conversations) between older adults and their relatives. TV

content is important since it is also a big part of offline conversations - it was referred as

a common behaviour to call someone on the phone to discuss or alert to something they

watched or are still watching on the TV.

Make it possible the digitally share printed photos: Old (printed) photos can be

effective in creating social opportunities between older adults and their family (or the ones

portrayed in the photos) because of their nostalgic potential. The process responsible for

digitalising this photos and sharing them through the SNS must be simple and easy to use

and only capable of reflecting the photos they have framed at their homes.

Favor older adults preferred interaction modalities: Older adults see the RC as the

main interaction modality since they are already used to it seeing it as a natural interaction.

In this sense, RC interaction should be simple and focus on using the directional arrows

and the OK button. Remaining keys on the RC can be used to perform other (social) tasks,

such as making a post with a press of a button. Voice and gestures should be considered

the main alternative modalities. Regarding speech, the implementation should focus on


simple commands (as natural speech interaction may create problems), while for gestures,

the mouse cursor metaphor approach should be considered. Finally, all three modalities

should be available for use at all times, and the older adult should never be forced into

using a specific modality.

Focus should be on three main functionalities: Content publishing, content visual-

isation and group management should be the three main functionalities made available

through an SNS prototype. Other than these, focus on family first and on direct and

reciprocal communication is essential.

Need for sub-group management functions: Older adults exalted the importance of

managing sub-groups inside each group of contacts and the fact that should be possible for

a contact to belong to several sub-groups at the same time, to reflect real-life situations.

Easy identification and navigation between contacts: Not only because of concerns

related with privacy or the need for using SNS only with people they know but also as

a design need, older adults must be able to identify clearly to which person a profile

corresponds. They also want to visualise the maximum amount of contacts per screen.

Therefore, the best solution should be one based on the contact name and a photo, and a

family degree if any.

Use of personalisation and adaptation functions: Older adults are interested and

receptive to both contexts. However, adaptation is seen as more advantageous and requiring

less effort, while they feel personalisation necessarily implies learning. For both older

adults highlight the importance of always getting contextual help through several outputs.

Additionally, they also feel the need and importance for controlling any change resulting

from adaptive mechanisms.

Thus, following these recommendations, we developed You, Me & TV, a prototype for

a TV-based Facebook application aiming to meet the needs of older adults regarding both

content and interaction.

Interaction-wise the prototype tried to retain some of the looks and feel of Facebook

while abiding by previously described interaction design recommendations targeting senior

users and TV-based (chapter 3) social applications (chapter 4): considering asynchronous

messages and communication that is quick to create and send, and highly visible when

received (Lindley, 2012); support older adults understandings of friendship (Lehtinen

et al., 2009); creating Internet access methods that are strongly preferred to mainstream

ones (like the TV) (Gibson et al., 2010); providing new topics of conversation and affinity

themes (Cornejo et al., 2013); fostering strong tie relationships, material-based social

communications which afford expression of thoughtfulness and concern, and the older

adults role in filling communication gaps in the family (Hope et al., 2014); removing

unnecessary or unused features of the Facebook interface (Norval et al., 2014); provide

TV with an important role when designing applications for sharing and co-creation of user-

generated content for older adults (Karahasanovic et al., 2009); make sure that approaches


to support sociability are explored not only regarding physical care but also emotional and

social support, with photos having a central role in achieving this (Gaver et al., 2011);

when focusing on TV-based UI always present a short number of interactive elements for

each screen focusing on big buttons, on the support of different modalities of interaction,

and on making possible personalisation of UI items (Coelho et al., 2011, 2013a).

Content-wise, given that the relevant content is mostly the one produced by family

members and close friends (Lindley, 2012; Judge et al., 2011; Gibson et al., 2010; Xie

et al., 2012; Cornejo et al., 2013; Hope et al., 2014), the prototype integrated directly with

the Facebook SNS through the Facebook’s Graph API. This also supported the prototype’s

publishing features, endowing the user with the ability to produce and consume content.

5.2.2 Implementation Decisions

Taking into consideration the TV context we decided for the implementation of a Web

application as the generality of TV-based services are based in this type of solution (Costa

et al., 2014). Therefore, this prototype makes use of web-based traditional development

technologies. In this case, as several sub-systems were developed and integrated into the

prototype, the following technologies were used:

• HTML5: Language used for developing the UI.

• CSS3: Language used for modifying the UI look-and-feel to make it aesthetically

more attractive.

• JavaScript, JQuery and AngularJS: Languages used to provide the prototype with

dynamic functionalities. This means that by using this, the prototype can change its

presentation and elements without being refreshed or reloaded.

• NodeJS: Language responsible for making the communication between the several

sub-systems more efficient by being based on asynchronous communication (i.e.

two pieces of code can be executed at the same time).

• PHP: Language used for the server to communicate with the database and return

information to the user or to update the database with data sent by the user through

the UI.

• Neo4J: A graph-based database in NOSQL with no predefined structure making

possible for every piece of information being treated as a new node and therefore

related with all the previous nodes. This was especially useful for constructing the

social network, with nodes representing each profile being associated with other

profiles, establishing the social relationships.


• Microsoft Kinnect’s module: Made use of Microsoft languages for the implemen-

tation of several modalities: (1) Speech: were defined a set of voice commands and

also an algorithm for speech-to-text; (2) Pointing: was implemented as an algorithm

for controlling the mouse cursor by mapping finger coordinates using the Kinect

camera sensors.

• RC module: Made use of an Arduino set with an infra-red receptor to receive

commands from the RC keys. Each key was pre-associated with a specific mes-

sage/command.

• Data module: Module responsible for creating, modifying and removing specific

nodes and database (Neo4J) relationships.

• Facebook API: Communication API responsible for inserting and getting informa-

tion from Facebook.

Additionally, and for structuring all the code, we followed an MVC (Model View

Controller) approach. Therefore we divided code into three distinct categories: (1) Models:

which grouped functions responsible for interacting with the database; (2) Views: respon-

sible for creating the UI and presenting on the screen; (3) Controllers: responsible for

receiving events related with user interaction and activating the respective models or for

getting from these all the information needed to be presented through views. A general

overview of all these modules and components can be found in fig. 5.5.

5.2.3 Main Functionalities

The “You, me & TV” prototype focused on three main features: 1. News-feed; 2. Managing

groups of friends; and 3. Photo viewing and sharing.

The news-feed (fig. 5.6) allowed older adults to see what their family and friends were

publishing in Facebook. It was also in the news feed that users were able to publish content

for their family and friends to see. The self-generated content was also displayed in the

news-feed together with likes and comments that family members and friends input on the

content. This provided a single place for a user to overview her or his social interactions.

The prototype also made it possible for users to like a content (the like button when toggled

would say “I don’t like” or the ability to cancel the like) or to see all comments or make a

new comment on the content. Relatively to the native Facebook applications, “You, me

& TV” allowed its users to publish two new types of content: the TV show the user was

currently watching and digital copies of printed photos he or she owns.

During the semi-structured interviews, it was perceivable older adults have a particular

interest in the possibility of viewing and sharing photos with family members and friends.

Accordingly, we included in the prototype a feature that allowed users to share with their

contacts the old printed photos they treasure at their homes (fig. 5.7). We hoped for this to


Figure 5.5: You, Me & TV infrastructure

Figure 5.6: The news-feed


Figure 5.7: Pictures that user took with Kinect camera.

increase their social possibilities while using the prototype. While designing this feature,

we decided on a solution that favoured interaction simplicity over the quality of the final

digital artefact to share, thus deciding against complex scanning mechanisms. The final

implementation makes use of a Microsoft Kinect connected to the prototype. The user

prompts the system (through one of the available modalities) for taking a picture, the

system presents the Kinect’s live picture, the user frames the printed photo in the image,

and a snapshot is taken after a small period has elapsed. This snapshot, after confirmation

by the user, is posted to the user’s Facebook stream, promoting social interaction with

family and friends. The feature supporting the user to post the TV-show being watched

to the Facebook stream was not completely integrated with the TV set in the prototype.

The prototype runs on a PC connected to a TV-set that worked only as a display. This

means that the prototype was not aware of what channel the user is watching. This required

extra effort from the user that would have to stop watching TV switch to the prototype

application and select the channel that was being watched (the prototype collected from a

Web service the shows being broadcasted in the different channels, freeing the user from

having to specify the show). An upgraded version of this application would allow the user

to post this info directly from the TV set.

Another important feature resulted from the discussions with older adults (chapter 4)

regarding the distinction they feel needs to be made between family and friends. Conse-

quently, we decided to separate them into two groups. This is implemented in the prototype

by having two entries in the main menu, one for each group: family and friends. Addi-

tionally, in the focus-groups preceding the prototype development older adults requested

the ability to specify smaller sub-groups. The prototype supports this recommendation by


Figure 5.8: A group with 3 members of user family.

allowing the user to create new sub-groups of the family and friends groups, and name

them as desired (e.g. “grandsons” or “reading club”). It was possible to add and remove

friends and family to and from these sub-groups (fig. 5.8).

5.2.4 Multimodal Architecture

Taking into account the different needs and characteristics that characterise the members of

this user group, we decided on the early stages of design to support multimodal interaction.

Given the TV interaction context, users could interact with the prototype using an RC in

addition to speech commands and gestures.

To interact with the application using the RC users mainly employ the arrow keys (up,

down, left and right) to control what element is selected on the screen. Additionally, there

were four main buttons on the RC: Confirmation, Back, Help and Personalization. The

confirmation button allowed users to confirm the execution of a task or make a selection.

For example, if users wanted to select one of the options within a news-feed post (e.g. like)

they would use the keys to select the desired post and then click the confirmation button

to access the different options. The back button allowed users to return to the previous

functionality. The help button allowed users to access a modal box with contextual help.

The personalisation button will be explained later in this section.

Regarding gesture interaction, we used the mouse cursor metaphor recommended by

Bobeth et al. (2012). With this metaphor, hand movement maps to the movement of a

cursor on the application. Selections were done by the hand closing action. Hand tracking

and gesture recognition were implemented through a dedicated application that uses the


Kinect SDK v1.8 and communicates events to the Facebook-based prototype.

Finally, the voice modality was divided into two modes: voice commands and dictation

mode. Voice commands were implemented using the Annyang framework (Annyang, 2015)

allowing command definition in Web applications and recognition of voice commands.

To prevent unwanted command recognition, in its initial state the application would only

recognise the keyword “Speech”. Thus, any voice instruction must be preceded by this

statement. Dictation mode was implemented using the Web Speech HTML5 API and

allowed users to write by dictating. This mode would only work every time the element in

focus was a text box.

5.2.5 Adaptation and Personalization Interface

The deployed prototype also explored the concepts of adaptation and personalisation. The

aim was to understand what characteristics of each concept would be more beneficial in the

context of an application targeting older adults. In what concerns these two concepts, and

its implementation on this prototype, adaptation was defined as the act of automatically

calibrating features to each user characteristics or interaction patterns. Personalization

was defined as the act of changing UI features by manually choosing or modifying them.

Adaptation and personalisation mechanisms were applied to two features: the order of the

elements in the main menu, and the order of the elements of the family and friends groups

and sub-groups. Regarding automatic adaptation, the element order was adjusted based

on the number of accesses made by the user. For example, elements included in the main

menu appeared in the order they were most commonly used. Thus, the most used features

appeared at the top of the menu. Concerning the order of family members and friends,

the same principle was applied – the most popular elements appeared in first places. The

personalisation concept was applied to the same features. For a user to personalise the

order of the elements in the different groups or the order of features in the main menu, the

element to be reordered would have to receive the focus using the arrow keys in the RC,

followed by a press in the personalisation button. After that a message would be presented

to the user, asking for the new position of the element. To ensure that users had control

over these concepts, personalisation options always took priority over adaptation options.

5.3 Validation

The overall goal of the prototype was to improve the quantity and quality of social

interactions of their senior users. To be able to assess this goal we conducted an in-depth

case study, where we deployed the prototype in the homes of three participants. Originally,

the idea was to perform a longitudinal study as we wanted to find out the impact the

prototype would have over a long period of use. However, because of Facebook API’s

limitations (forced the migration to Graph API v2.0 leading to the friends’ data API


Table 5.1: Participants description

User U1 U2 U3

Age 65 67 73

Gender Female Male Female

Family (on Facebook) 6 12 8

Friends (on Facebook) 4 7 3

Duration 6 weeks 3 weeks 3 weeks

Facebook user No Yes No

Lives with Alone Wife and Stepson Husband

Retired Yes No Yes

shutting down, and a few other changes) and limited resources (each system needs a PC, a

Kinect, an RC, and a TV) we ended up performing three case studies. Nonetheless, these

could give us insights into the prototype acceptance, frequency of use, more used features

and the number of technology-mediated social interactions performed by older adults.

5.3.1 Participants

We planned to have the prototype deployed in the homes of four older adults for three

weeks each. All participants were selected via convenience sampling. However, one

participant requested to keep the prototype running after the three-week period ended.

We agreed to the request, and since we had no access to further hardware for prototype

deployment, we ended up with three participants. Table 5.1 shows the profiles of the older

adults who participated in the study. The table also shows the number of family members

and close friends who also contributed by “friending” the senior participant and placing

content in their native Facebook news-feeds.

As seen in Table 5.1, two of three participants were not Facebook users previously

in this study. However, User 2 (U2) used Facebook on a daily basis. We assisted the

participants who were not Facebook users in creating their accounts and inviting their

relatives.

All three participants lived in the same city (Lisbon) although in distinct neighbour-

hoods, and all had family members living in the same city. Only User 1 (U1) had family

living close by, in this case, her daughter and grandsons. None of the participants had

significant visual or mobility impairments, rather than occasional tiring associated with

age.

5.3.2 Procedure

The study, which was performed from mid-March to mid-April 2015, started with the

installation of the system in the participants’ homes. Firstly the system was introduced,

and an explanation of all features and modalities that could be used was given. Each task


was performed several times by each user, and when necessary by the trial mediator, to

exemplify or clarify doubts, until the user agreed he/she understood and could perform

them without help. Contact channels with the trial mediator were also established so that

any doubt during the trial period could be clarified by phone or with a technical visit. All

features were available during all weeks, except for adaptation and personalisation features.

These were used separately because they could cause incoherent results when used together.

U1 and U3 used personalisation in the first half of the study and adaptation in the second

half. U2 used these features in reverse order. All accesses to the different features were

logged, and the findings gathered through the analysis of these logs are presented in the next

section. We also collected qualitative data through weekly checkpoints (interviews) where

we visited the participants’ home and discussed (with no previously defined structure)

their opinions of the system and difficulties they faced. Additionally, we administered

two surveys to measure their experience with the system: the System Usability Scale

(SUS) (Brooke, 1996) and the User Experience Questionnaire (UEQ) (Cota et al., 2014).

SUS calculates a score of the application’s usability. The average SUS score is 68. Thus,

a SUS score above 68 would be considered above average. UEQ measures the user

experience through six different scales that evaluate the quality of design (stimulation and

novelty) and quality of use (dependability, perspicuity, efficiency and attractiveness). U2

and U3 answered these questionnaires every week while U1 answered it every two weeks,

for a total of three samples for each participant. Both questionnaires were administered as

one, and the Portuguese version of the full questionnaire can be found in appendix F.

5.4 Results

Overall, older adults received our prototype well. Since the first days, all participants

and their relatives showed the willingness to use the application. Although this would

be expected to happen, as all three participants agreed to participate in the study, their

experience improved as they kept using the prototype. In the following sub-sections, we

discuss different aspects of the overall experience.

5.4.1 Older adults acceptance

In this section, we address acceptance and use of the prototype. We base our discussion on

the frequency of use of the system and individual features, complemented with opinions

collected during the study.

5.4.1.1 System Usage

U1 was the most active user of our system. She said that she wanted to use the application

every day. Figure 5.9 illustrates this statement. As it is possible to see, U1 used the


Figure 5.9: Frequency of use of U1, U2 and U3

application almost every day. In the 44 days, she had the prototype available to her,

she used the application 159 times and only in 7 days she did not use the “You, me &

TV” prototype. She averaged 3.6 uses per day. No pattern correlating non-usage with

the weekday was found (the prototype was not used on one Monday, one Tuesday, one

Wednesday, two Thursdays and two Fridays). In general, she really appreciated our system

because it allowed her to interact more with her relatives. She said: “many times after I

use the application, I call my relatives and talk with them about the photos that I saw in

“You, me & TV””.

On the other hand, U2 made limited use of the system (Figure 5.9). He used the

application only 34 times in 23 days, averaging 1.34 uses per day. In the checkpoint

meetings, he said that he was already accustomed to using Facebook on his personal

computer and he did not see advantages in using it on a TV-based application. However,

he was the user that made the most use of the different modalities, in particular, voice

and gestures. He stated that he enjoyed these modalities and the fact “they were always

available for all features”.

U3 made growing use of the application, especially because the number of shared

photos between her and her family also increased (Figure 5.9). She accessed the prototype

74 times in 23 days with an average of 3.2 uses per day.

In general, since the first week, participants felt the system was quite beneficial to their

lives as they could see what family members did daily (if they published it on Facebook).

On every checkpoint, participants indicated that features such as sharing printed photos

and viewing family photos were the most important ones in this application.

As expected, the family was the major reason for adopting “You, me & TV”. In total,

we recorded 42 accesses to family member profiles compared to only 12 to friend profiles.


This reflects that the participants’ social networks were composed mainly of their family

members.

Another important feature of this system was the ability to share what the user was

watching on TV. In the weekly checkpoint discussions, we learned that the usage of this

feature could have been considerably higher if the integration with the TV set was full,

instead of requiring the user to change between TV and prototype application. Still, they

showed enough interest in this functionality to use it, even with the required effort. They

also suggested the possibility of having a button on the RC that could automatically share

with family members what they were watching on TV.

Participants appreciated the features that allow fast content sharing with their relatives.

U1 stated: “Sometimes I can not call my family, and they often are unable to call me, but

with this application, I can always know what they are doing and can easily show them

what I’m doing”. In the context of being able to communicate somehow proximity with

family members, the ability to “like posts” in the news-feed was appreciated. In fact, she

indicated that it was easier to select a button to like one post than to write a comment in

that post. We observed that content sharing motivated older adults to establish contact with

their relatives. U1 indicated that she reestablished contact with an aunt who she had not

seen for a long time. This fact is representative of the potential of our prototype to increase

relations between older adults and their relatives. The importance of photos, consuming

family content, but also of creating content can be seen as main functionalities preferred in

Figure 5.10, where the number of times each feature was used by each user is presented.

Considering U2, this participant made limited use of You, me & TV’. However, he

indicated appreciating the sharing printed photos feature because “sharing pictures on

Facebook is really difficult and through this application, it is easier”.

U3 focused particularly in photos, both the ability to take a photo and share it with the

family and the ability to see the photos of her relatives. She used the application as a photo

sharing system: “I used the application to share old printed photos with my family. They

also send me old photographs, and this allowed me to revive old times.” In three weeks,

U3 shared forty-three photos with her family.

5.4.1.2 Interactions with others

We looked into how SNS could increase social interactions between older adults and their

relatives by measuring the number of sharing actions – liking posts, posting, commenting,

sharing TV content, sharing printed photos – that older adults performed in You, me &

TV’ and observing their evolution over time for each participant.

U1 interacted 83 times (averaging 1.89 interactions per day) with his/her family mem-

bers. The evolution of these interactions can be seen in Figure 5.11. We can see that the

number of interactions increases with time, after an initial period where the participant

was not yet comfortable enough to share content.


Figure 5.10: Feature usage by U1, U2 and U3

U2 interacted only 15 times (averaging 0.65 per day), and growth was contrary to the

one observed in U1 (Figure 5.11). However, the data for this participant must be labelled

as inconclusive, since he was already an assiduous user of Facebook and he continued to

use the Facebook native application during the study period, establishing interactions that

we could not account for.

U3 was the participant with the greatest number of interactions with others (Figure 5.11).

She engaged in 75 interactions in total (an average of 3.26 interactions per day). Like

participant U1, U3 showed a tendency to increase interactions over time.

5.4.2 Multimodality importance

Regarding the modality most used in the study, we could verify that the RC was the

main way participants used to interact with the prototype. Additionally, this was the only

modality for which they did not require assistance. They felt comfortable interacting using

the same device they always have used to interact with the TV set. However, we also found

that there was interest among the participants to try the other (new) modalities, especially

for features that were not so easy to use with RC, like writing a post. In fact, we can

state that the participants did not appreciate using the RC to write posts, given that it was

a very time-consuming process. As a consequence, we witnessed an increase in posts

written by voice over time. U2 found that “by speaking. It is much easier to write on TV.”

Gestures were also used, especially to select large buttons, such as buttons representing

family members and friends. However, participants considered it a little difficult to use,


Figure 5.11: Social interactions of U1, U2 and U3

given that they already exhibit slight hand tremors which caused the on-screen cursor to

jump considerably.

Table 5.2 presents the overall distribution of modalities used for each feature, across the

three participants. As aforementioned, it is clear that the RC was used for most interactions.

However, it is informative to look at what the other modalities were most used for. Speech

commands emerge as a valid alternative for producing textual content. In fact, its overall

usage surpassed the use of the RC to write posts and comments. Albeit much lower, it also

had a non-negligible use for liking posts. Gestures, on the other hand, had more use for

content viewing tasks, including accessing profiles of friends and family members and

accessing photos posted by the members of the participants’ network. Overall, results

indicate that voice is more useful in content producing tasks, while gestures are more

appropriate for content browsing tasks.

5.4.3 Benefits of personalisation and adaptation

The concepts of adaptation and personalisation elicited different reactions from the partici-

pants.

Personalization concepts were perceived as very complex, which caused some dis-

comfort. U1 stated “it is a feature that will not be missed.” U2 indicated that “he did not

felt that its use would cause the application to become simpler to use.” U3 considered

personalization, in general, a feature for advanced users. She stated: “you might think that

this is a good feature if you know how to use it. I don’t see benefits in using it. But that’s

me!”

In what concerns adaptation, the opinions were more positive, in particular regarding


Table 5.2: Usage of different modalities relative to specific tasks.

Task / Modality RC Voice Gestures

Total Percentage 77.31% 7.98% 14.71%

Content production 77% 17% 6%

Posting 39% 61% 0%

Comment a post 45% 55% 0%

Like a post 64% 26% 10%

Take a photo 88% 4% 8%

Share TV content 100% 0% 0%

Content visualization 76% 0% 24%

See family digital frame 100% 0% 0%

See family profile 68% 0% 32%

See friends profile 58% 0% 42%

See member’s photos 75% 0% 25%

See member’s videos 100% 0% 0%

See member’s interests 100% 0% 0%

Group management 80% 0% 20%

Create a subgroup 100% 0% 0%

Add member to subgroup 100% 0% 0%

Remove member from subgroup 71% 0% 29%

the adaptation of the order by which friends and relatives are listed. U1 said: “I have

more interest in seeing pictures of my grandchildren and my children and I like that they

automatically appear first.” U3 had the same opinion, adding the following regarding the

adaptation of the order of entries in the main menu: “I use the application to view photos

of my family, and for taking and sharing pictures with them, so I like that the first two

menu options are my family and my photos. Best of all is that I did not have to do anything

to change that.” However, U2 alerted us to the fact that sometimes the menu changed

without him being aware. This leads him to select the wrong menu option occasionally.

However, at the end of the trial, he recognised that it was a feature that had plenty of

benefits especially because it made it faster to perform certain actions.

5.4.4 Usability and User Experience

As mentioned, each checkpoint participants filled out both SUS and UEQ questionnaires.

We now present and discuss these results.

U1 was the participant that better classified our prototype regarding usability and

user experience. The usability level, as measured by SUS, was consistently rated high

and increased at each checkpoint (Figure 5.12). UEQ scores constantly remained high

throughout the checkpoints with a single exception: dependability (Figure 5.13). This

was expected given that the deployed system was a prototype, and thus not completely


Figure 5.12: Evolution of SUS scores for U1, U2 and U3

robust and stable. Another factor that contributed to the low dependability score was the

participant’s perception regarding the security of Internet applications, as can be seen in

U1 statement: “sharing personal photos on the Internet is a little insecure”.

U2 was the participant who worst classified our system regarding usability and user

experience. In the first checkpoint, he gave a high usability level as measured by SUS

(Figure 5.12) but the following checkpoints’ results were substantially worse. This could

mean that his valuation of You, me & TV’s usability decreased after the initial novelty

factor wared off. This is particularly evident in this participant since he continued using

Facebook’s native application, which has a higher level of usability than our prototype.

However, it is important to observe that only checkpoint number 2 presented a SUS score

marginally below 68 points, which is the average SUS score. UEQ scores reflect the same

tendency (Figure 5.13). Only the novelty score maintained a higher value throughout

the study, reflecting the participant’s opinion about the novelty that speech and gestures

brought to the interaction with Facebook. Finally, the attractiveness score, albeit decreasing

in each checkpoint, was always in the positive range of the UEQ scale, meaning that the

overall experience of this participant was still positive.


Figure 5.13: Evolution of UEQ scores for U1, U2 and U3

U3 also classified our prototype positively. The SUS score lowered from the first to the

second checkpoint but increased in the final checkpoint to a final score above 77 points

(Figure 5.12). UEQ scores remained positive and stable across the period of the study, with

the same exception that had already been observed for U1, regarding the dependability

score (Figure 5.13).

In general, results indicate that our prototype achieved high usability marks and

provided participants with great user experience, especially taking into account that it was

still a prototype.

5.5 Discussion

Findings from this study strongly support previously related work notions (section 4.2)

that SNS wanting to meet older adults’ needs should focus primarily on family interaction.

In our study, accessing content and interacting with family members, even if done offline,

proved to be the main reason for adopting such technology. The content older people

prefer to consume is related to their family, allowing them to know more about their

relatives. Although this can be a way of increasing social interactions by fostering contact

with family members, it can also raise concerns regarding the possible lack of interaction

with non-family contacts, or even lack of SNS use if family members are absent from

these. In fact, among older adults, other uses are associated with the use of SNS (finding

information about what’s going on in the world, contacting friends, communicating inside

specific groups) and TV (watching information programs, watching entertainment shows)

(section 4.3.1), therefore, technology must not be seen only as a bridge for family contact,

but also focus on other functionalities for keeping in contact like the ability to increase

neighborhood interactions and communicating with groups of friends or colleagues, or on


other functionalities related with keeping a more active life spanning from accessing news

to developing new activities and interests.

Regarding the type of content that the participants in this study preferred, photographs

played the central role. There were two types of preferred photographs: (1) photos from the

senior’s past that provide nostalgic moments and encourages them to re-establish contact

with relatives; and (2) current photographs of family that allow older adults to know more

about their relatives’ status and activities, especially the ones that live far away.

Regarding novelty, and when compared with the traditional Facebook UI and func-

tionalities, the feature of our system which influenced participants the most was the one

related to sharing printed photos. The ability to use the Kinect camera to capture the

printed photographs and send it as a digital memoir to their relatives on the network was

seen, not only as an easier way to upload a photo but also allowed both older adults and

relatives to revive old memories and favoured reciprocity. This is mainly because by

visualising this content on their feed, relatives also uploaded more photos, favouring a

chain of interaction around this type of digital artefacts. These findings not only support

previous ones related with the importance of photo content to attract older adults to use

social technology (“Photos and Media Relevance” domain specified in section 4.4) but

extends these notions further showing viable ways of including new photo artefacts on

SNS systems based on TV.

TV is a technology that older people face with optimism and do not feel uncomfortable

using since they are already accustomed to it. In that regard, using the TV as a platform

for deploying the access to Facebook proved to be a contributing factor to their adoption.

The fact that they can share TV content with their contacts was also identified as a positive

factor. Participants indicated that the awareness of what TV shows they are watching

created new streams of conversations, which consequently led to an increase in the number

of interactions with their relatives. Another factor that made our participants comfortable

using the applications was the ability to interact with using the RC. This was the modality

most used by them. However, when participants felt that this was not the most appropriate

way to perform a specific task, then they looked for new ways of interaction. This was

visible when participants found the RC too cumbersome to write posts and comments

and explored alternatives, eventually switching to an easier modality for that task: usually

using speech to dictate the contents. This supports both others (section 4.2) and ours

(section 4.3.1, chapter 3) perspectives that a multimodal system is paramount given its

ability to provide alternative modalities for all tasks that can be carried out within the

application.

Two other concepts that can be important for improving the accessibility of an in-

terface for older adults are personalisation and adaptation. The personalisation features

implemented were not seen as positive by our participants, since it increased the usage

complexity of the application. They had to learn how to customise it before they could


reap the benefits of this feature, but they were not willing to spend that effort. Adaptation

features were, however, much better received by the participants, who enjoyed not having

to do anything to benefit from a better user experience. By adapting the order of the menus,

participants performed faster and had the feeling the system acknowledged what they

would like to do, making them feel confident. The only raised issue was related with the

lack of awareness of when adaptation was performed.

Perhaps the most impacting factor relative to our prototype adoption was the previous

participants’ experience with social networks. This is clearly seen in the usage patterns

exhibited by U2 on the one hand, and U1 and U3 on the other. U2 was already a Facebook

user and saw limited benefits in using the SNS via a TV set. He did point out the benefits

of having additional modalities to interact with, but overall he made limited usage of the

prototype during the three weeks it was available to him. U1 and U3 were not previous

users of Facebook, had none to limited experience with PCs, and were enthusiastic about

the prospect of learning to use one. Providing them with access to Facebook through a

platform they were already familiar with proved to be decisive for its use (and probable

adoption). The impact on their well-being was such that one of them asked for retaining

the prototype after her trial period ended (first three weeks). The results strongly suggest

that, for technology inexperienced users, TV can make Facebook accessible. Results

also suggest that having access to Facebook increased the quantity and quality of social

interactions experienced by older adults, both online and offline. Overall, we found

evidence that a solution with similar characteristics to what we have deployed has the

potential to help fight social isolation of older adults, especially considering that focusing

on family is the first step in that way.

We envision that these findings could inspire different initiatives among the oldest

segment of the population. Community and health centres should have ways of providing

and supporting this kind of access for older adults to share and receive information related

to their family and close friends. For the less active, however, the focus should be on

supporting and motivating the development of more basic and less complex (and more

familiar) technology like TV (or Tablets) as a way of accessing social and communication-

based services which are typically available from more complex technology (like the PC

or Smartphones) that these cannot or have trouble using.

Finally, we would like to discuss some of the limitations that constrained this study.

The main one being the small number of participants in it. We had to choose between

having more participants for shorter periods or less for longer periods. Given the nature of

the study we were conducting, we opted for the latter. Two factors constraining the number

of participants were the lack of further hardware to deploy the prototype and the fact that

the Facebook API was changed at the end of April, ending our study for all purposes.

Having such a small number of participants limits the generalisation of our findings. It

also prevented us to investigate further factors that can impact the adoption of the proposed


solution. However, we still believe that these findings are relevant, and supported by

having observed similar behaviours in participants U1 and U3, which had in common

not being Facebook users. Additionally, the lack of data and observations concerning

the number and type of social interactions pre and post-study hinder us from analysing

to what extent the increase in this type of interactions was associated with the use of the

prototype and the kind of technology used. Other limitations strive from the incomplete

integration of the prototype with the TV set and an improvable implementation of the

hand tracking and gesture recognition system. The former limited a more fluid exchange

of attention between the Facebook application and the TV broadcast, which might have

prevented further usage of our prototype. The latter limited the usability of gestures, which

might have prevented the more extensive use of this modality, given the combination of its

heightened sensitivity with participants exhibiting hand tremor. Additionally, it’s arguable

how much a prototype built on top of TV limits another kind of interactions like watching

TV, especially if the person using it does not live alone. Still, not only this system was

originally designed taking special consideration for people who live alone, but we can

also consider two other factors as indicatives of this not being a real problem: (1) 2 of the

3 users were accompanied during the period they used the prototype, and none of them

mentioned any trouble regarding this; and (2) the majority of activities performed focused

on family, and this could be something of interest to, not only the person using it, but also

to the other elements of the same household.

5.6 Summary

In this chapter we described all participatory design steps taken into building a TV-based

SNS prototype - “You, Me &TV” - focused on both recommendations identified in previous

studies and chapters and also on focus groups performed with older adults. The chapter

follows with the report of an in-depth case study with three senior participants focusing on

how the use of “You, Me &TV” prototype can help older adults to be more socially active.

The results indicated a potential for this system, or another with its characteristics, to

increase both online and offline social interactions between older adults and (mainly) their

family relatives. Providing TV-based access to Facebook was a contributing factor to the

adoption of the system, and for the resulting opportunities for social interaction. This was

a decisive factor for the participants that were not Facebook users at the beginning of the

study. However, its importance was smaller for the participant that already used Facebook.

The prototype supported two innovative features which were very well received by older

adults: the possibility of sharing printed photos and the support for sharing TV content.

Both these features showed potential for triggering offline conversations and, in that way,

help fighting social isolation of its users. Additionally, participants classified the prototype

very high regarding usability. Regarding interaction preferences, although the RC was seen


as the traditional way of interaction, speech was used mainly for content production tasks

while gestures were seen as a viable option for visualisation purposes. Finally, while the

personalisation features were seen as adding complexity to the system, participants really

enjoyed the adaptation mechanisms because they speed up the interaction and provide

them with a sense that the system understands them. This case-study showed that TV-based

Facebook provided with appropriate modalities of interaction (and adaptation features)

could help older adults tackle social isolation, not only by making them use Facebook-

related functionalities they do not feel comfortable to use on the PC but mainly because

these online interactions result in more offline social exchanges and bigger reciprocity with

their relatives. Finally, the novel features implemented in this prototype, by making use

of standard and available technology (a Kinect sensor or camera and an RC) are easy to

replicate or integrate into real TV-based systems.


Chapter 6

Conclusions and Future Work

In this final chapter, we provide a general overview of how the work performed in this

PhD answers the identified research questions and hypothesis. After an overview of all

the work developed in this thesis, we answer each research question and hypothesis in its

own section. For each we identify the relevant domains and subjects addressed throughout

the thesis and how they contribute with findings and explanations. Following these, we

provide a brief discussion regarding the main limitations of this PhD, as well as a brief

section identifying relevant future work.

6.1 Dissertation Overview

We started this thesis with an overview of contributions and papers published regarding the

work developed as well as a brief description of the methodology employed. We motivated

the thesis context regarding the need for SNS based on technologies closer to older adults

like the TV, and we enumerated several reasons for this need.

In the second chapter, we started by summarising relevant past research on older adults

limitations. We showed how attention should be given to visual impairments avoiding

small configurations of both text and UI elements and providing high contrasting colours

when possible. Additionally, hearing impairments should be compensated by combining

other types of feedback. Furthermore, How because of cognition declines design should

be based on avoiding too much information and functions at the same time, and drawing

attention to the most performed actions. And through all these, we have understood the

importance of multimodal features to avoid clutter and drawing attention to complementary

sensory channels. We have also seen that some TV-based solutions that make use of

multimodal features exist with the goal to increase its inclusiveness or as a way to provide

additional ways of accessing this technology. Common to these solutions was keeping the

RC as the main way of interaction. And finally, we showed how although some solutions

exist regarding integrating social mechanisms with TV, they were too much tailored for

this population typically leaving their younger counterparts out.

155

Chapter 6. Conclusions and Future Work 156

In the third chapter, we described how some of the work developed under the context

of the GUIDE project contributed with valuable recommendation guidelines regarding

the design of inclusive TV applications, from which we later build to implement our

SNS-based solutions. This work involved several requirement gathering and user-testing

phases with both developers and older adults, and varying from a series of interviews,

focus groups and questionnaires to the development of a user-test application and an

EPG. Recommendations span from how to implement different interaction modalities,

adaptation principles, inclusive design, and to a whole range of recommendations related

with TV-based interaction and user interfaces.

In the fourth chapter, we described the results of a literature survey and several re-

quirement gathering steps to extract relevant recommendation guidelines for the design

of an SNS-based TV-based application targeting older adults. We identified 13 relevant

domains with the most important being the ones related to family, privacy, multimodal

interaction, UI principles and photos. The requirement gathering steps confirmed these

domains, and we provided a detailed description in which we extracted about 60 design

recommendations which formed the basis for the development of the “You, Me & TV”

prototype.

Finally, in the fifth chapter of this thesis, we start by describing the steps we followed

for the design and implementation of a TV-based multimodal SNS prototype based on

Facebook - the “You, Me & TV” prototype. In these, we followed the recommendations

identified in the two previous chapters and performed low-fidelity participatory design

sessions with older adults. We also describe in detail the in-depth case study on the use of

the prototype, where three older adults made use of it for a period of 3 weeks to 6 weeks,

and in which the results clearly indicated a great potential for a system like this, resulting

in an increase of both online and offline social interactions between older adults and their

family relatives and friends. We also understood that the use might be different when

comparing Facebook adopters with non-adopters and that new functionalities like the ones

related with the ability to share printed photos as digital photos through the TV or sharing

what they see on the TV are among the ones with the bigger impact. Additionally, the high

usability of the prototype was achieved as a result of following the previously identified

recommendations.

6.2 RQ1: What are the factors limiting SNS adoption by

older adults?

As a result of this thesis findings, we identified five different factors that limit SNS adoption

by older adults: age-related impairments, (SNS) UI complexity, (SNS) UI functions, older

adults self-belief and technology.

Regarding age-related impairments, related work concerning older adults limitations


(chapter 2) showed visual, hearing, cognitive and motor impairments resulting from aging

are the four main factors that limit technology adoption among this population and also

contribute to social isolation (by compromising natural ways of communication or having

related emotional effects that undermine the psychological well being). Therefore, SNS

adoption/use is twice limited by these impairments. This was further supported by findings

when questioning older adults regarding factors that limit their use of Facebook (sec-

tion 4.3.1). We saw that those older adults with tremors, trouble moving arms, difficulties

grasping a mouse, positioning and controlling a cursor, or hitting a button, tend to use

Facebook (or the technology which permits accessing Facebook) less than older adults

with just a small portion of these problems, and much less than users with none.

Concerning UI complexity has been one general concern when designing applications

usable by older adults (chapter 2), more specifically when targeting these in TV-related

context (section 2.2.1 and chapter 3), as well as a problem preventing older adults access

to SNS (section 4.2). This problem was first tackled with the development of SNS with

particular UI characteristics targeted to older adult users, but as these applications are not

adopted by their family members and relatives (section 4.2.4), particular concern has to be

put into the following issues (chapter 4):

1. features placed in a non-intuitive manner in the UI;

2. specific functionalities being hard to find;

3. labels used in buttons and links unfamiliar to those who have not used these kind of

services before;

4. not being usable by impaired users, older adults, or users which shift away from the

typical 20 to 40 years old segment

In this thesis, and as a result of designing a particular SNS UI with older adults and

studying its prolonged use with them, we have been able to understand further the problem

or ways of mitigating it (presented in the next sub-section).

Strongly related with UI complexity, SNS functions are also a determinant factor in

SNS adoption by older adults as SNS are not designed to around the most interesting

functions for this population. In this thesis, this has been understood first when analysing

past studies on older adults use of SNS (section 4.2), as it was also understood when

discussing with older adults regarding their main goals behind using SNS and even TV-

based related features (chapter 4). From all these, particularly relevant are the findings

regarding the importance of family, offline value, and photo and direct communication

relevance, that were conversely drawn from every user study, as well as the rejection of

functions related with knowing new people or making new friends (section 4.3.2), which

was at first the main goal of an SNS like Facebook.


This thesis also highlights that self-belief in their ability to use technology is related

to the capability to use SNS. When inquiring older adults about SNS use (section 4.3.1),

we saw a direct relationship between the way older adults believe in their capabilities and

their use of Facebook, hinting that by considering themselves capable of interacting with

technology or being confident in their technical skills, older adults are more likely to use

these services.

Last but not least, technology is also a factor limiting the adoption and use of SNS by

the oldest segment of the population for many reasons:

• older adults tend not to have sufficient computer literacy, or have difficulties interact-

ing or learning to use a PC, which is still the main technology used for accessing

SNS like Facebook (section 1.1);

• smartphones, which are the other main access point to SNS, are also difficult to

use for great part of this population segment, be it because the majority of SNS

applications do not target older adults issues concerning UI complexity or because

they do not target their age-related specificities (section 2.1);

• while tablets represent a compromise for accessing SNS for older adults (because of

the screen dimensions allied with direct manipulation), SNS applications tend to be

adapted from the mobile context bringing the same issues identified with these.

• TV constitutes in many ways the ideal access point for older adults use of SNS

(mainly because they are used to it), but SNS designed specifically for this context

have been too distinct of the ones deployed on traditional technologies (section 2.2),

or are based on new ways of interaction (leaving the RC out), and as a consequence

have shied older adults and their relatives away from its use (section 2.2.1 and

section 2.2.2).

6.3 RQ1.1: How these factors should be considered when

designing SNS targeting older adults?

Concerning each one of the previously identified factors, several recommendations can be

extracted from all the work described in this thesis regarding the design of SNS targeting

the older segment of the population.

6.3.1 Age-related impairments

Regarding age-related impairments, several recommendations can first be drawn from the

related work identified in this document (chapter 2) concerning each of the impairment

types. From these we highlight the following:


1. making fonts, icons and colors more suitable by increasing their size, maximizing

contrasts, or compensating visual impairments with audio feedback and input, or

with the help of other multimodal mechanisms;

2. compensate auditory loss with other feedback like vibrations and multimodal mech-

anisms;

3. compensate cognitive impairments by:

• displaying all necessary information for each user to perform a task;

• building upon previously learned procedures/actions;

• avoiding technical expressions clutter and unnecessary items;

• drawing them to important elements through complementary sensor channels

(multimodal feedback), or by employing key function unity (i.e. one key one

function).

4. compensate motor impairments by considering input alternatives (like speech),

reducing cursor speed or provide appropriate button size and space.

Additionally, the whole chapter 3 focuses around the evaluation of alternative interac-

tion modalities on a TV-based context and how these can compensate age-related declines.

This chapter provides a detailed list of recommendations (section 3.5) regarding specific

modalities like the RC, speech and voice output and pointing, as well as regarding multi-

modal contexts. The same chapter also introduces adaptation or personalisation procedures

and showcases how these can further help compensate age-related impairments. From

several recommendations presented we highlight:

• the need to personalise each interaction for or by each user;

• the need to adapt to distinct user contexts

• the need to pre-establish a preferred modality of interaction;

• the need to consider different ICT skills and user impairments when adapting pre-

sentation.

Both chapter 4 and chapter 5 contribute with further findings that support the potential

of multimodality and adaptation to compensate age-related declines, and in that sense

suggest additional relevant recommendations. Regarding multimodal interaction and

adaptation, these should be highlighted:

• make alternative modalities of interaction as intuitive and non-instructional (with no

required fixed commands or steps);


• traditional ways of interaction should be kept as the main interaction mode (like RC

for TV);

• alternative modalities should be offered to compensate age-related declines or pref-

erences;

• speech should be the main alternative for both text input tasks or tasks not achievable

with RC;

• gestures should be provided as an alternative for spatial and navigational (or brows-

ing) tasks;

• the availability of each modality should be adapted to each user characteristics;

• support an UI with augmentation alternatives;

• offer mechanisms which tailor both presentation and interaction to each user charac-

teristics.

6.3.2 UI complexity

In what relates to UI complexity and how it can limit older adults adoption of SNS, several

recommendations have been drawn regarding four UI-related domains: privacy, grouping,

tangible features, and UI simplification.

Privacy issues were first identified as a problem because older adults had trouble

understanding who can see what and are afraid of revealing too much about their lives

(chapter 2). These issues were confirmed throughout the whole chapter 4 resulting in

several recommendations followed when designing the prototype described in chapter 5

and further validated by the results of its use:

• focus on facilitating the discovery and access to privacy settings (chapter 4);

• change default settings to be more private and make it optional to reach more contacts

or groups (chapter 4);

• simplify privacy options related with “post to” and “view from” (chapter 4);

• make more explicit who a contact is and his/her relationship degree with the user

(chapter 5).

Similarly, solutions contemplating the grouping of functionalities and contacts were

appointed as one of the simpler ways of making SNS more usable to older adults (chapter 4).

These indications were corroborated by discussions performed with older adults and refined

in the process of designing the “You, Me & TV” prototype (chapter 5), resulting in several

recommendations:


• group functionalities into family, favorites and community (cultural, health) oriented

features (chapter 4);

• adapt the UI to highlight only meaningful contacts (chapter 4);

• simplify the access to a feature depending on frequency of use (chapter 4);

• support the creation and use of sub-groups reflecting the several context of older

adults lives (chapter 5).

Moreover, other findings resulting from discussing SNS complexity with older adults

(chapter 4) showed that virtual interaction could be more meaningful when supported by

some personal artefacts or by the notion of providing a more tangible UI. This tangible

value helps by making the UI more familiar or similar to activities older adults are used

to do (like exchanging photos or writing letters). In this sense, several recommendations

were advanced like the ability to take advantage of physical artefacts older adults have at

their homes (like framed photos), or the possibility of supporting scribbling or handwritten

messages. One of these recommendations was not only implemented in the developed

prototype which made use of the Kinect to support the ability to take photos of framed

photos, but in fact proved to be one of the main triggers that increased interaction supported

by the SNS (and offline too) (chapter 5).

Finally, when focusing specifically on UI simplification, several other recommenda-

tions were drawn by this thesis which can contribute to a better SNS UI. When addressing

the development of TV-based UI targeting older adults (chapter 3) some recommendations

should be highlighted:

• UI should be based on an as simple as possible layout and limited amount of

functions;

• UI should be designed to be clear, simple and with uncluttered screens focusing on

big and well-spaced buttons;

• one-touch features should activate most used functions;

• UI should be simplified but not look senior-only;

• feedback should be constant and catch the user attention;

• well-known TV-related concepts should be maintained.

These recommendations were confirmed and considered when discussing SNS constraints

(chapter 4) from which resulted several others:

• the need for the UI to provide “safe points”;


• the need for the UI to be constant and with few layout changes in order to enhance

older adults learning;

• the need for favoring easy-to-understand language over computer jargon or non-

native language.

All these recommendations were considered when designing an accessible SNS prototype

(chapter 5) and while they were pre-validated at design-time by older adults own configu-

ration of the UI, the results on its extended use also show older adults satisfaction with the

resulting UI usability.

6.3.3 UI functions

In what respects UI functions, while several current SNS functions are usable and of

interest to older adults, others are seen as out of scope regarding the ways this segment

of the population wants to use these services. Also in that matter, throughout this thesis,

some functions were also identified as missing from a typical SNS like Facebook. From all

identified domains, recommendations were drawn concerning the avoidance of functions

related with knowing new people, and the support of functions related with family, offline

interactions, reciprocity, the provision of photo/media as a catalyst for interactions and the

importance of direct communication. Following, we summarise recommendations on each.

Regarding knowing new people, findings throughout the whole chapter 4 show that:

• acquaintances should be highlighted over “new friends”;

• distinguish content created by individuals the user know or does not know;

• options should be provided for avoiding friend suggestions;

• options should be provided for avoiding news-feed content not originated from

well-known contacts.

In direct relation to these, family comes in first place concerning functions that should be

offered by the SNS. In that sense more recommendations resulted from chapter 4:

• better support the maintenance of family relationships;

• provide the possibility of better specifying family roles within the SNS;

• support the ability to create events or initiate video-conference calls with family

members;

• provide the possibility of reviving old connections or feelings by resurfacing old

family content;


• provide the ability to include “friends” or “favorite” as part of family features;

• support traditional forms of expressing family engagement.

Following the tendency of these recommendations, discussions with older adults when

designing the SNS prototype (chapter 5) also showed their main interest for functions that

work as a bridge between them and their family members (or close friends). Therefore

several other recommendations were drawn:

• the presentation of family degree when navigating through contacts;

• the possibility to share printed photos with family;

• the ability to focus on family, or on direct communication with family members, as

the central function of the SNS.

In fact, direct communication was one of the main domains identified in the extensive

literature overview presented in chapter 4, and as a result, this matter was discussed with

older adults and generated recommendations related with creating incentives for this kind

of interaction:

• providing chat functions with a more central role in SNS;

• supporting the resurfacing or prioritization of information;

• increasing the visibility of video-chat features;

• providing mechanisms for configuring, or making others aware, of availability

periods.

While some of these were later implemented in the SNS prototype (chapter 5), chat

functionalities were not made available (to favour offline interactions). Therefore, the

focus was on functions that could facilitate communication with family members and

relatives (like the resurfacing of family-related content, default sharing privacy options

directed to the family group, and the sharing framed photos and TV-content). As a result,

when talking about direct communication, we are also talking about the value of functions

focusing on reciprocal and offline ways of interaction.

Several studies appointed reciprocity as one of the main goals of older adults for

using technologies related to social interaction (chapter 2). That notion was also discussed

and confirmed with older adults, and understood as the act of providing opportunities

and functions to keep in contact with family and close friends (chapter 4). As a result,

recommendations were suggested:

• designing features which enable reciprocity and joint use;


• evidencing and resurfacing content capable of stimulating communications.

When inquiring older adults, offline interactions were identified as a goal of using

SNS that would interest them more than any online interaction (section 4.3.1 and 4.3.2). A

lot of literature defended this same notion and further showed that the line between online

and offline interactions should be very thin (section 4.2). Recommendations for achieving

this include:

• give more attention to functionalities related to organising events;

• incorporate and make visible phone numbers of SNS contacts;

• generate or promote social events close to the user.

Also converging with the notion of offline interactions comes the inclusion of TV content

as one of the preferred types of content to share on an SNS, mainly because they are a big

part of offline conversations (chapter 5).

Finally, and common to all these functions, photos were understood as the main way of

fostering interactions and also as the main way of attracting older adults to the use of SNS.

This was a notion first collected from findings of the literature survey which was confirmed

by indications given in questionnaires and focus-groups (chapter 4) and validated with

the use of the “You, Me & TV” prototype (chapter 5). In the prototype, photos were not

only the main request for profile identification and to promote its use, but also the main

object around which social interactions occurred. As a result, recommendations such as

the following should be considered:

• give more relevance to photos/images than to other types of content;

• simplify and increase the intuitiveness of uploading photos;

• provide or improve ways of interaction around photos;

• provide ways of replicating real (physical) photos through the SNS;

• focus on photos as the way to provide easier identification of contacts;

• support ways of passively consuming photos posted on the SNS.

6.3.4 Older Adults Self-Belief

We have seen that older adults self-belief in their technological skills is directly related

with the use of SNS (section 4.3.1). Therefore, our recommendation is that in order to

increase their self-belief we should provide them with a platform to access SNS with which

they are comfortable. Taking this into account, we have resorted to TV as the basis of the

“You, Me &TV” prototype (chapter 5).


6.3.5 Technology

Concerning technology and how providing the ideal setup can support older adults in

accessing SNS, in this thesis we argue that TV is the ideal entry point. However for that

to be a reality specific recommendations must be met. First of all TV by itself does not

have sufficient features to make it possible to use SNS like Facebook. Therefore providing

multimodal features (like some of the systems referred in chapter 2) and simplifying and

keeping the RC as the main interaction method are two of the main recommendations

to consider (several recommendations regarding this can be found both on previous sub-

sections and originating from chapters 3 to 5). Secondly, TV-based interfaces should not

only be designed with accessibility concerns in mind but should also be concerned with

adapting themselves to older adults capabilities. While multimodal features can help in

that by supporting several modalities and broadening the user options for interaction, some

adaptation mechanisms should also be considered tying UI elements and specific interac-

tion modalities (input and output) to specific age-related impairments and technological

expertise (chapters 3 and 5). Thirdly, by using Facebook instead of developing specific

solutions tailored for older adults, we are also making it possible for older adults to use the

same system their (younger) family relatives and friends use, and therefore we are one step

closer to making its adoption possible. Last but not least, when designing Facebook-like

TV-based applications several recommendations should also be considered regarding UI

elements and how these should behave and look like. More on this can also be found ahead

when answering the following research question.

6.4 RQ2: What characteristics multimodal TV-based plat-

forms have that could contribute to older adults adop-

tion of SNS?

Related work findings (section 2.2.1.6) showcased several characteristics that multimodal

TV-based platforms have that can contribute to older adults adoption of SNS. With this

type of system they can (and should) continue to interact with the traditional RC they are

already used to, and if (or when) RC is too slow or inadequate (like in tasks involving

writing text or navigating complex menus), they can resort to alternatives such as speech,

pointing, gestures or touch. This is described as increased accessibility because using

more modes of interaction can result in the increase of vocabulary of symbols available

to the user and at the same time leads to a highly synergistic blend in which the strengths

of each mode are capitalised upon and used to overcome weakness in the other. This was

identified in this thesis introduction (and by earlier work on multimodal interfaces) as

increased efficiency, reliability, and flexibility (chapter 1). Because of these characteristics

past work on multimodal-TV also showed that even when an application is designed with


no accessibility concerns, the fact that it supports multimodal interaction represents a step

closer to being more inclusive.

Moreover, the work developed under the GUIDE project further supported multimodal

characteristics. In one of the user studies (section 3.3), while older adults typically preferred

multimodal input and output, they also exhibited different multimodal or unimodal patterns

depending on the context and feedback provided by the system. In the second user study

(section 3.4) not only RC was seen as a safer option for interaction but older adults were

able to interact with any other available modality, with shifts between modalities being

frequent, and simultaneous use of two modalities being seen as rather counter-intuitive

(and as something that would require training). Both user-studies further support the

previous findings that by either resorting to multimodal mechanisms or several possibilities

of unimodal interactions, multimodal TV-based platforms are an adequate fit for both

technology-trained and technology-untrained older adults.

Finally, both the notions of efficiency, reliability and flexibility and the notions related

with the preferential use of RC and alternate use of other available modalities were also

supported by the studies analysed in section 4.2, by discussions presented in section 4.3.3

and by observations regarding the use of the Facebook-based prototype (chapter 5).

6.5 H1: By using TV-based multimodal SNS, older adults

increase interactions with family members and close

friends.

When questioning older adults about the importance of TV (section 4.3.1), answers showed

not only that older adults relate TV with uses nonaligned with SNS (or Facebook) main

purposes but also that only one-quarter of them showed interest in those type of SNS

functions, while half were against it. While this preliminary indications disagree with H1,

we should consider these notions might have been aggravated by older adults’ tendency

for not only rejecting the use of technology they do not know for purposes they know

well (like the use of Facebook for keeping in contact with family and friends), but also

rejecting the use of technology they know well for purposes they are not familiar with (like

the use of TV as a bridge for that Facebook access). In fact, when discussing older adults

TV-related use (section 4.3.2), results showed their interest towards using a TV-based SNS

application is justified by their interest regarding social tasks on TV, such as talking and

sharing photos with family and friends or sharing TV-originated content. While these

findings were in accordance with H1, several indications were also given that a number

of modifications should occur (like the provision of alternative modalities to the RC, the

need for contextual help or making use of adaptation mechanisms). These indications

were followed in chapter 5, along with the implementation of other suggested functions

like the sharing of printed photos through the TV-SNS and the possibility of sharing TV-


content. Additionally, a prototype was built with the help of the target (older adult) users,

which helped in understanding and designing the UI according to their preferences and

expertise. As a result we have clearly seen an increase on both (logged) online interactions

and (reported) offline interactions of the two individuals which were non-adopters of

Facebook, while we have also acknowledged a consistent use of the new features and

the new modalities by the one which was already a user of the SNS. The observed and

reported findings support H1 at least for non-previous adopters of Facebook, while for

previous adopters the results are also encouraging. Still, we also acknowledge that only the

replication of these use-cases with more users could permit to extrapolate these findings to

a set of clearly identifiable older adults profiles.

6.6 H2: By using SNS, older adults establish new friend-

ships.

While SNS systems like Facebook were initially built on top of the notions related to

establishing new friendships or knowing new people, this thesis shows this is not the

case for older adults. We have seen examples of systems focusing on social and TV-

based applications that supported both family and friends and features to establish new

friendships (section 2.2.2). If this could indicate at first that both functions should be as

SNS goal, when dwelling into the extensive study of older adults and SNS applications

(section 4.2) we see that only 40% focused on studying functions related with establishing

new friendships and only about a quarter of these favoured the use of SNS for these

purposes. Reasons appointed by the majority were related with the fact that for older adults

to establish new friendships or connections with people they do not know through SNS

would require courage, would not necessarily build up social capital, or would disrupt

social ties with existing family or friends. Additionally, findings from these works also

stated older users would like to avoid functions like friend suggestions and unsolicited

emails and reject any activity related to establishing new friendships online.

These indications which do not support H2, were further reinforced by findings from

questionnaires (section 4.3.1) and discussions with older adults (section 4.3.1). When

inquired about the reasons behind Facebook use, older adults not only gave the most

relevance to keeping in contact with family and close friends (i.e. people they already

know) but they also rejected its use as a platform for making new friends. And when

discussing relevant functions, only one person referred this as something to have on an

SNS (with the vast majority again focusing on functions related with people they know).

Finally, as these kind of features were also not referred by older adults when designing the

“You, Me &TV” prototype, no feature related with these functions was implemented in the

final prototype (chapter 5).

All these findings and considerations clearly reject the hypothesis, not only denying


that by using SNS older adults establish new friendships but also that they, in fact, don’t

want to use SNS for any purposes related with this.

6.7 Limitations and Future Work

This work main limitations are the ones related with the final user study (or use-cases): the

small number of participants in it, the incomplete integration of the prototype with the TV,

and above all the Facebook API changes that ended the study for all purposes. As a result,

the lack of participants in the study prevent us for having results as generalizable as we

would have desired. Additionally, other limitations were also involved when designing

and performing questionnaires and discussions with older adults like the ones described in

chapter 4. For these, the main difficulty was finding participants and institutions willing to

participate in the studies, and the lack of distinct institutions where to perform and repeat

the same designed experiments (ideally we would like to have had a senior university,

an health-care institution and a senior nursing home over the whole course of the thesis).

This limited the understanding of how the nature of the institution could influence the

results obtained or the establishment of more relations regarding the level of isolation

and correspondent level of technological expertise and Facebook knowledge (to name

a few). Furthermore, and regarding the work developed in the context of the GUIDE

project (chapter 3), the fact that all were performed in the context of a European Project

with distinct goals from this PhD, made more difficult a clearer or better-suited design

of its tasks. This hindered, for example, the development of inquiries, discussions or

developments of applications/prototypes with a social context or which could approximate

more to an SNS context.

Regarding future work, after working for six years on areas related with older adults,

inclusive TV design and SNS we gathered some knowledge on which domains or future

research questions should be a priority.

First of all, and recurrent from the fact that several inquiries were made to older

adults at different times during this thesis, and most of them were performed to assess

how older adults embrace technology they do or do not know, we learned the lack of

appropriate ways to measure technology acceptance or expertise among older adults. By

this, we mean that although many models exist which explain ICT use and predict how

users embrace technology (like TAM (Davis, 1989), TAM2 (Venkatesh and Davis, 2000),

TAM3 (Venkatesh and Bala, 2008) or UTAUT (Venkatesh et al., 2003)), there are also

contradictory feelings regarding its validity regarding older adults (Schepers and Wetzels,

2007; Kim and Garrison, 2009; Bagozzi, 2007; Chen and Chan, 2011). Additionally, the

number of models which have been developed which focus specifically on older adults

(like SCT (LAM and LEE, 2006), AARP survey (Barrett, 2008), CE (Arning and Ziefle,

2008), CLS for older adults (Sengpiel and Dittberner, 2008) or STAM (Chen and Chan,


2014)) make it hard to understand which one is the most appropriate for which situation

without a detailed analysis of all of them. This analysis would also be very relevant for

other studies in this area, because at the best of our knowledge still, no other research has

focused on this.

Secondly, and as a result of the possible increase of social interactions derived from

Facebook use, or from a more inclusive SNS adaptation to older adults, it is also important

to focus on the understanding of how older adults family members or relatives perceive

older adults use of those tools. This would be important not only to understand in which

way they would be typically receptive or keen to help or to interact with them but also to

understand which ways they feel it would be excessive or damaging to their interactions

or daily-living. Relatives could also be inquired about functions they feel that could

be added to this kind of services that could help further in the social inclusion of their

dependent/independent older relatives.

Thirdly, and as a result of the importance that older adults give to both TV and photos,

it would be interesting to focus future research on ways of integrating the two for both

senior healthcare contexts and to seniors living at home. The Photostroller (Gaver et al.,

2011) is certainly a very good example of this, but more could be developed which could

be based both on older adults activities of daily living as well as on interests and on family

interaction or routines. We believe this kind of solutions by being easily integrated into

older adults daily routines and being capable of easily attaining older adults interest would

have a great capability of tackling social isolation without the need of using a full SNS.

Fourthly, and conversely to the previous point, research should also focus on developing

new inclusive technologies which incorporate the feelings of tangibility and nostalgia. On

almost every discussion we had with older adults under the course of this thesis, every

time a feeling like these emerged, more substantial and enthusiastic conversations occurred

as a result. While some technologies related to dementia and Alzheimer’s already done

this (Routledge et al., 2013; Chao et al., 2017; Wallace et al., 2012), this is not commonly

transported to general technology which could see its use by older adults increase as a

result.

Finally, the inclusion of persuasive mechanisms capable of reminding older adults

of certain social opportunities or tasks could also help further in the increase of social

interaction or on fighting social isolation among this segment of the population. Research

on this subject has already taken place as part of several European Union programs.

However, projects are mostly focused on fostering physical exercise and rehabilitation.

Developing tools capable of understanding older adults context (family, cultural, health and

interests) and reminding them to engage in more social opportunities (be it by sending them

messages/notifications or recommending certain events) might be a way for technology to

foster social inclusion without requiring heavy use of devices.


Appendix A

Survey Script: Stakeholders Research -

Inquiring Older Adults

171

9 Survey administered in the pre-trial

!

"#$%&! #'()!*+,-.!/(01)+! 23!4!565!

!

! "#$%&'()*+,-,./&$&*(,-(/0&(1$&2/$,)3(

71-(8!

%9+(8!

&:9;,9+1)8!

!!

!"#$%&'%()*+%,-./0'&#'12%-+.3'%1#

!!

5<!=>(8!????!

6<!"(@-()8!

!A9;(!

!B(C9;(!

!!

D<!A9)E+9;!*+9+,'8!

9<!'E@>;(!

F<!C9))E(-!4!;E:E@>!GE+H!9!I1,0;(!

I<!GE-1G!

-<!'(09)9+(-!

(<!-E:1)I(-!

!!

J<!&-,I9+E1@!K.(9)'L8?????!

!!

3<!M9:(!.1,!)(+E)(-N!

!!! !.('O!)(+E)(-!9'????????????????????????????????!

!!! !@1O!G1)PE@>!9'???????????????????????????????!

!!

Q<!%14%E-!.1,!H9:(!9@.!I1@+9I+!GE+H!+(IH@1;1>.!-(:EI('!9+!G1)PN!

!!!!!!!!! !R('O!K0;(9'(!'0(IES.L!GE+H!??????????????????????????????!

!!!!!!!!! !T1<!

!!

2<!M1G!-1!.1,!;E:(N!

!!!!!!!!! "#!"$%&'(!

!!!!!!!!! )#!*+,-!./%0.'1/"2,&'2(!

!!!!!!!!! 3#!*+,-!3-+$42'&(!

!!!!!!!!! 4#!*+,-!%,-'2!5"6+$7!6'6)'2.!%2!52+'&4.(!

!!

8#!%1!.1,!,',9;;.!'E+!-1G@!9@-!G9+IH!UV!+1>(+H()!9+!H1C(NN!

!R('!

!T1!

!!

W<!%1CE@9@I(8!

!!!!!!!!! !2+9-,:-"&4'4!

!!!!!!!!! !$'5,:-"&4'4!$'5,


!

"#$%&! #'()!*+,-.!/(01)+! 2Q!4!565!

!!!!!!!!! !"6)+4';,2%0.#!

!

4"#5-*6'%78#3*&01%9%,:#*;/*-'*1&*"#

5-%)7&3!

!!

<%# :%7#

%=1# 30*#

/-%)7&3>!

!! ?2# :*8@# &.1# :%7#

%/*-.3*# 30*#

/-%)7&3>!

5X!<%=!>?!@&$7!*+,-!

-'$/! 52%6! %,-'2.=!

A?!B+,-!"!$%,!'55%2,=!

C?!B+,-!"&7!-'$/!

!! !! !! A%=# =%79)# :%:#

)*8&-'B*# :%7-#

%/'1'%1# .B%73# 30*#

/-%)7&3>!

5X! D'27! &'9",+E'=!

>?! <'9",+E'=! A?!

<'0,2"$=! C?!

F%.+,+E'=! G?! D'27!

F%.+,+E'!

!! !! !! !! A%=# %23*1# )%# :%7# 78*# 30*#

2%99%='1,#/-%)7&38>!

YX!<'E'2(! H?! @&3'! %2! ,*+3'=!

>?! IE'27! 5'*! 6%&,-.=! A?!

IE'27!5'*!*''J.=!C?!@&3'!%2!

,*+3'! "! *''J=! G?! K'E'2"$!

,+6'.!"!*''J=!L?!M"+$7!

!! !! !! !! !! !!

!! C%! D*8!5! 6! D! J! 5! 6! D! J! 3! Y! 5! 6! D! J! 3! Q!

E%78*! Z! Z! Z! Z!Z!Z!Z! Z!Z!Z!Z!Z! Z!Z!Z!Z!Z!Z!

F*:B%.-)! Z! Z! Z! Z!Z!Z!Z! Z!Z!Z!Z!Z! Z!Z!Z!Z!Z!Z!

G%7&0#

8&-**1!

Z! Z! Z! Z!Z!Z!Z! Z!Z!Z!Z!Z! Z!Z!Z!Z!Z!Z!

GH! Z! Z! Z! Z!Z!Z!Z! Z!Z!Z!Z!Z! Z!Z!Z!Z!Z!Z!

E%B'9*#

/0%1*!


I%+/73*-! Z! Z! Z! Z!Z!Z!Z! Z!Z!Z!Z!Z! Z!Z!Z!Z!Z!Z!

?13*-1*3#

&%11*&3'%1!


H')*%(

&%12*-*1&*!


G*9*(

9*.-1'1,!


G*9*3*;3! Z! Z! Z! Z!Z!Z!Z! Z!Z!Z!Z!Z! Z!Z!Z!Z!Z!Z!

!!

$'!+H()(!9@.!1+H()!9''E'+E:(!+(IH@1;1>.!+H9+!.1,!,'(N?????????????????????!

M9:(!.1,!(:()!,'(-!+H(!UV!S1)!9@.+HE@>!F(.1@-!G9+IHE@>!0)1>)9C'N??????????????????!

M9:(!.1,!(:()!,'(-!+H(!C1FE;!0H1@(!S1)!9@.+HE@>!F(.1@-!C9PE@>!I9;;'N????????????????!

##

##

J"#K33'37)*8#3%=.-)8#3*&01%9%,:#)*6'&*8#


!

"#$%&! #'()!*+,-.!/(01)+! 22!4!565!

N-+.!O0'.,+%&&"+2'! +.!4'.+9&'4!5%2!2'.'"23-!/02/%.'#!N-'! +&5%26",+%&!3%$$'3,'4!*+$$!&%,!)'!0.'4!5%2!"&7!

%,-'2!0.'.#!N-'2'!"2'!&%,! 2+9-,!%2!*2%&9!"&.*'2.#!B'!*+$$! "//2'3+",'! 7%02! 3%%/'2",+%&!"&4!-'$/<!@02!

2'.'"23-!92%0/!+.!+&,'2'.,'4!+&!7%02!%/+&+%&!"&4!.J+$$.!+&!,'3-&%$%97#!

%E)(I+E1@'8![;(9'(!9@'G()!(9IH!\,('+E1@!F.!+EIPE@>!+H(!@,CF()! +H9+!I9@!F('+! E@-EI9+(!GH9+!.1,!)(9;;.!

+HE@P!9F1,+!+H(!\,('+E1@<!UH(!@,CF()'!'+9@-!S1)!+H(!S1;;1GE@>!)('01@'('<!

!!

5X=>)((!

6XT(,+)9;!

DX%E'9>)((!

!!

5<!$!+HE@P!+H9+!@(G!+(IH@1;1>.!-(:EI('!9)(!-(:(;10(-!C9E@;.!+1!F(!,'(-!F.!.1,@>!,'()'<!

5< !!! !!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

6<!$!+HE@P!+H9+!$!@((-!+1!,'(!@(G!+(IH@1;1>.<!

5< !!!! !!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

D<!$!I1@'E-()!E+!EC01)+9@+!+1!+).!+1!F(!10(@]CE@-(-!+1G9)-'!@(G!+(IH@1;1>.<!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! !D<! !

!!

J<!$!I1@'E-()!C.'(;S!H9:E@>!+H(!@(I(''9).!'PE;;'!+1!C9@9>(!+1!,'(!@(G!+(IH@1;1>.!+11;'<!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

3<!$!H9:(!0)1F;(C'!+1!,'(!+H('(!+(IH@1;1>E('!0)10();.!(:(@!GE+H!0)9I+EI(<!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

Q<!UH(!0)1F;(C'!1S!+(IH@1;1>.!-(:EI('!9)(!EC01''EF;(!+1!,@-()'+9@-O!'1!E+!E'!H9)-!+1!SE@-!9!'1;,+E1@<!

5<! !!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

2<!^H(@!+H()(!E'!9!0)1F;(C!GE+H!9!@(G!+(IH@1;1>.!+11;O!E+!E'!F(I9,'(!+H()(!E'!'1C(+HE@>!G)1@>!GE+H!+H9+!

-(:EI(<!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

8. I�m afraid to touch a new technology tool in case I�ll break it.!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

9. I don�t get advantage of using new technology tools.!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D<! !

!!

5Y<!$!0)(S()!+1!,'(!9@!1;-!S9'HE1@!+11;!GE+H!S(G()!S,@I+E1@'!+H9@!9!@(G!1@(<!

5< !!!!!!!!!!!!!!!! 6<! !!!!!!!!!!!!!!! D< !

L"#IM5?CN#$GDOP$#

N-'!+&5%26",+%&!3%$$'3,'4!*+$$!&%,!)'!0.'4!5%2!"&7!%,-'2!0.'.#!N-'2'!+.!&%!2+9-,!%2!*2%&9!"&.*'2.#!B'!*+$$!

"//2'3+",'!7%02!3%%/'2",+%&!"&4!-'$/<!F$'".'!)'!2'"$$7!.+&3'2'!*+,-!7%02!"&.*'2.#!


!

"#$%&! #'()!*+,-.!/(01)+! 2_!4!565!

!!

5<!^H(@!$!H9:(!9!0)1F;(C!,'E@>!9!@(G!+(IH@1;1>.!+11;O!$!,',9;;.`!

9<!U).!+1!'1;:(!E+!GE+H!C.!1G@!'PE;;'<!

!!

F<!$!I1@+9I+!GE+H!C.!S9CE;.!1)!S)E(@-'!'1!+H(.!I9@!'1;:(!C.!0)1F;(C<!

!!

I<!$!H10(!+H(!0)1F;(C!GE;;!F(!'1;:(-!;9+()!GE+H1,+!+1,IHE@>!E+!1)!-1E@>!9@.+HE@>!GE+H!+H(!@(G!

+(IH@1;1>.!+11;O!'1!$!a,'+!G9E+<!

!!

-<!$!a,'+!;(9:(!+H(!-,+.!$!G9'!-1E@>!9@-!$!-1!@1+!I1C(!F9IP!+1!'((!GH9+!H9:(!a,'+!H900(@(-!

GE+H!+H(!+(IH@1;1>.!+11;<!

!!

6<!^H(@!$!'+9)+!H9:E@>!1@(!0)1F;(C!9S+()!9@1+H()O!$!,',9;;.!10+!+18!

9<!U).!+1!'1;:(!E+!GE+H!C.!1G@!'PE;;'<!

!!

F<!$!I1@+9I+!GE+H!C.!S9CE;.!1)!S)E(@-'!'1!+H(.!I9@!'1;:(!C.!0)1F;(C<!

!!

I<!$!H10(!+H(!0)1F;(C!GE;;!F(!'1;:(-!;9+()!GE+H1,+!+1,IHE@>!E+!1)!-1E@>!9@.+HE@>!GE+H!+H(!@(G!

+(IH@1;1>.!+11;O!'1!$!a,'+!G9E+<!

!!

-<!$!a,'+!;(9:(!+H(!-,+.!$!G9'!-1E@>!9@-!$!-1!@1+!I1C(!F9IP!+1!'((!GH9+!H9:(!a,'+!H900(@(-!

GE+H!+H(!+(IH@1;1>.!+11;<!

##

Q"#K33-'B73'%1.9#83:9*8#3%=.-)#3*&01%9%,:#)*6'&*8#

!!

5<!$S!$!9C!@1+!9F;(!+1!'1;:(!+H(!0)1F;(C!$!H9:(!GE+H!+H(!@(G!+(IH@1;1>.!-(:EI(O!$!+HE@P!+H9+`!

!!

!!! 9<!$!9C!@1+!I909F;(!1S!'1;:E@>!+H('(!0)1F;(C'<!

!!

!!! F<!T(G!U(IH@1;1>E('!-(:EI('!H9:(!9!>)(9+!-('E>@!F,+!'1C(+EC('!+H(.!I9@!F(!F)1P(@!+11!GE+H1,+!

-1E@>!9@.+HE@>!G)1@><!

!!

!!! I<!T1F1-.!H9'!+9,>H+!C(!I1))(I+;.!'1!$!I9@!@1+!-1!E+<!

!!

!!! -<!$+!E'!C.!S9,;+O!F,+!E+!E'!I1CC1@!E@!9;;!+.0('!1S!0(10;(<!

!!

6<!UH(!,'9>(!1S!@(G!U(IH@1;1>E('!-(:EI('8!

!!

!!! 9<!79,'('!C(!01'E+E:(!(C1+E1@'!KC1+E:9+E1@O!I,)E1'E+.O!H900E@(''<<<L!

!!

!!! F<!79,'('!C(!@(>9+E:(!(C1+E1@'!K9S)9E-O!'+)(''O!'9-@(''<<<<<L!

!!

I<!$+!-1('!@1+!I9,'(!C(!9@.!(C1+E1@'<!

##

##

R"#<.'9:#9'6'1,#.&3'6'3'*8#


!

"#$%&! #'()!*+,-.!/(01)+! 2W!4!565!

!!

!!! 9<!M1G!G1,;-!.1,!-('I)EF(!.1,)!>(@()9;!H(9;+H!'+9+,'N!

!D'27!9%%4!

!P%%4!

!Q"+2!

!<%,!.%!9%%4!

!R"4!

!!

!!!!!!!!! !F<! $@! +H(! +9F;(! '1C(! 9I+E:E+E('! 1S! C1FE;E+.! 9)(! C(@+E1@(-<! *1C(! 0(10;(! H9:(! 0)1F;(C'!!!!!

! 0()S1)CE@>! +H('(! 9I+E:E+E('<! 79@! .1,! E@-EI9+(! +1! GHEIH! -(>)((! .1,! 9)(! 9F;(! +1! 0()S1)C! +H(!

! 9I+E:E+E('N!

!!

!! T1+! 9+!

9;;!

Z@;.! GE+H! H(;0! S)1C!

1+H()'!

Ê+H! 9! ;1+! 1S!

(SS1)+!

Ê+H! '1C(!

(SS1)+!

Ê+H1,+!

(SS1)+!

"(++E@>!E@!9@-!1,+!1S!F(-! Z! Z! Z! Z! Z!

^9;PE@>!,0!9@-!-1G@!+H(!'+9E)'! Z! Z! Z! Z! Z!

*E++E@>!-1G@!1)!>(++E@>!,0!1,+!1S!9!IH9E)! Z! Z! Z! Z! Z!

A1:E@>!9)1,@-!E@!.1,)!H1C(! Z! Z! Z! Z! Z!

A1:E@>!1,+'E-(!.1,)!H1C(!F,+!E@!+H(!:EIE@E+.!

1S!.1,)!H1C(!

Z! Z! Z! Z! Z!

A1:E@>!S9)!9G9.!S)1C!.1,)!H1C(! Z! Z! Z! Z! Z!

!!

%1!.1,!,'(!9@.!9E-!S1)!C1FE;E+.N!

!R('!

!T1!

$S!.('O!GH9+!PE@-!1S!9E-'!-1!.1,!,'(N!

?????????????????????????????????????????????????????????????????????????????????????

???????????????????????????????????????????????????????????????????????!

!!

!!

S"#5PTIP5GUKO#$F?OO$#

!!

!!!!!!!!! K"#H?$?MCV!

!!

$7BW*&3'6*#.88*88+*13#

5<!%1!.1,!H9:(!9@.!-E9>@1'E'!)(>9)-E@>!.1,)!(.('E>H+N!

9<!";9,I1C9!

F<!79+9)9I+!

I<!/(+E@10H9+.!

-<!Z+H()'!K'0(IESE(-L8!??????????????F(++()!'E>H+!1@!1@(!(.(!

!!


!

"#$%&! #'()!*+,-.!/(01)+! _Y!4!565!

6<!%1!.1,!G(9)!>;9''('!1)!I1))(I+E:(!;(@'('N!

9<!R('!

F<!T1!

!!

A#! K'$5! 2",'4! 9'&'2"$! E+.+%&! ]! K^H(@! G(9)E@>! .1,)! >;9''('! 1)! I1))(I+E:(! ;(@'LO! H1G! >11-! E'! .1,)!

(.('E>H+N!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

$S! @1+! >11-O! -1! .1,! H9:(! 9@.! 0)1F;(C'! GE+H! :E'E1@O! GHEIH! >;9''('! 1)! I1))(I+E:(! ;(@'! don�t

I1))(I+N!

??????????????????????????????????????????????????????????????????????????????

??????????????????????????????????????????????????????????????????????????????

??????????????????????????????????????????????????????????????????!

!!

J<!K'$5!2",'4!4+.,"&,!E+.+%&!]!K^H(@!G(9)E@>!>;9''('LO!H1G!>11-!E'!.1,)!(.('E>H+!S1)!'((E@>!+HE@>'!9+!9!

-E'+9@I(O!;EP(!)(I1>@E'E@>!9!01'+()!9I)1''!+H(!'+)((+N!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

3<!K'$5!2",'4!3$%.'!E+.+%&! ]! K^H(@!G(9)E@>!>;9''('LO!H1G!>11-!E'!.1,)!(.('E>H+!S1)!'((E@>!+HE@>'!,0!

I;1'(O!;EP(!)(9-E@>!9!@(G'090()N!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

Q<!K''+&9!",!<+9-,!]!K^H(@!G(9)E@>!>;9''('LO!H1G!>11-!E'!.1,)!(.('E>H+!S1)!'((E@>!+HE@>'!9+!@E>H+O!;EP(!

)(9-E@>!9!@(G'090()O!,'E@>!'+)((+!;E>H+E@>N!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

2<!%1!.1,!H9:(!0)1F;(C'!-E'+E@>,E'HE@>!+H(!I1;1,)'!C.10E9N!

!!

######### #MBW*&3'6*#.88*88+*13#

*H1G!+H(!09)+EIE09@+!+G1!'H((+'!K'ECE;9)!+H9+!+H(!1@('!,'(-!F.!+H(!10H+H9;C1;1>E'+!,'('!GH(@!+('+!+H(!

:E'E1@L8!1@(!S1)!+H(!-E'+9@+!:E'E1@!9@-!+H(!1+H()!S1)!I;1'(!:E'E1@L<!

!!


!

"#$%&! #'()!*+,-.!/(01)+! _5!4!565!

######### X"#APKT?CNV#

!!

$7BW*&3'6*#.88*88+*13#

##

5<%1!.1,!,'(!9!H(9)E@>!9E-N!

R('!

!T1!

!!

6<!K'$5!2",'4!9'&'2"$!-'"2+&9!]!K^H(@!,'E@>!9!H(9)E@>!9E-LO!H1G!>11-!E'!.1,)!H(9)E@>N!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

A#!K'$5!2",'4!-'"2+&9!KFIIST!*+,-!&%+.'!]!K^H(@!,'E@>!9!H(9)E@>!9E-LO!H1G!>11-!E'!.1,)!9FE;E+.!+1!

S1;;1G!9!I1@:()'9+E1@!ES!+H()(!E'!9!@1E'.!F9IP>)1,@-O!',IH!9'!UVO!)9-E1!1)!IHE;-)(@!0;9.E@>N!

!!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

J<!K'$5! 2",'4!4+.,"&,!KFIIST!-'"2+&9! ]! K^H(@!,'E@>!9!H(9)E@>!9E-LO!H1G!>11-! E'!.1,)!9FE;E+.! +1!

S1;;1G!+H(!-E9;1>,(!E@!9!C1:E(!1)!9+!+H(!+H(9+)(!K,'E@>!9!H(9)E@>!9E-!9'!,',9;LN!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

3<!K'$5!2",'4!-'"2+&9!UV<PV<P!*+,-!&%+.'!]!K^H(@!,'E@>!9!H(9)E@>!9E-LO!H1G!>11-!E'!.1,)!9FE;E+.!+1!

H(9)!9!0H1@(!)E@>!ES!+H()(!E'!9!@1E'.!F9IP>)1,@-O!',IH!9'!UVO!)9-E1!1)!IHE;-)(@!0;9.E@>N!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!

!!

Q<!K'$5! 2",'4!4+.,"&,!KFIIST!-'"2+&9! ]! K^H(@!,'E@>!9!H(9)E@>!9E-LO!H1G!>11-! E'!.1,)!9FE;E+.! +1!

H(9)!9!0H1@(!)E@>!E@!9!C1:E(!1)!9+!+H(!+H(9+)(!K,'E@>!9!H(9)E@>!9E-!9'!,',9;LN!

9<!&bI(;;(@+!

F<!>11-!

I<!S9E)!

-<!011)!

(<!:().!011)!(<!:().!011)


!

"#$%&! #'()!*+,-.!/(01)+! _6!4!565!

!!

MBW*&3'6*#.88*88+*13#

Administer the �Home Audiometre Hearing Test�.!

##

# ######I"##EMX?O?GD#

5<%1!.1,!H9:(!9@.!-E9>@1'E'!)(>9)-E@>!.1,)!C1FE;E+.N!

!R('O???????????!

!T1!

!!

#

!!!!!!!!!!

K-+#.1)#$0%79)*-#+%6*+*138#

!!

!!!!!!!!! 3<! [;(9'(! H1;-! .1,)! 9)C! 1,+! E@! S)1@+! 9'! S9)! 9'! .1,! I9@! I1CS1)+9F;.! 9@-! H1;-! E+! +H()(! S1)! 9!!!!!!!!!!!!!!!!!!!!!!!!

! I1,0;(!1S!'(I1@-'<!WV&,'2E+'*'2!,%!4'6%&.,2",'X!

!!=)C!S,;;.!(b+(@-(-!K@1+!F(@+L!

![9)+E9;;.!(b+(@-(-!KH9;S!G9.L!

!79@@1+!(b+(@-!9)C!9+!9;;!

!!

!!!!!!!!! Q<! [;(9'(! H1;-! .1,)! 9)C! 1,+! 9'! S9)! 1@(! 'E-(! 9'! .1,! I9@! I1CS1)+9F;.! 9@-! H1;-! E+! +H()(! S1)! 9!!!!!!!!!!!!

! I1,0;(!1S!'(I1@-'<!WV&,'2E+'*'2!,%!4'6%&.,2",'X!

!!=)C!S,;;.!(b+(@-(-!K@1+!F(@+L!

![9)+E9;;.!(b+(@-(-!KH9;S!G9.L!

!79@@1+!(b+(@-!9)C!9+!9;;!

!!

!!!!!!!!! 2<!79@!.1,!I)1''!.1,)!H9@-!9I)1''!.1,)!F1-.!9'! S9)!9'!.1,!I9@!I1CS1)+9F;.!9@-!H1;-! E+! +H()(!!!!!!!!!!!!!!!!!!!!

! S1)!9!I1,0;(!1S!'(I1@-'<!WV&,'2E+'*'2!,%!4'6%&.,2",'X!

!!=)C!S,;;.!(b+(@-(-!K@1+!F(@+L!

![9)+E9;;.!(b+(@-(-!KH9;S!G9.L!

!79@@1+!(b+(@-!9)C!9+!9;;!

!!

$'3#.1)#83.1)#

!!

!!!!!!!!! _<!Y.J!,-'!+&4+E+40"$!,%!.+,!+&!"!3-"+2!*+,-1*+,-%0,!"26.!"&4!.,"&4!52%6!+,!*+,-%0,!-'$/#!

!!&9'E;.!

!!Ê+H!-ESSEI,;+.!

!!79@@1+!-1!

!!

Y.9Z'1,#.1)#X.9.1&*#

!!!!!!!!! W<!79@!.1,!G9;P!9I)1''!+H(!)11C!S1;;1GE@>!+H(!;E@(!1@!+H(!S;11)!9'!S9)!9'!.1,!I9@!I1CS1)+9F;.N!

!!&9'E;.!

!!Ê+H!-ESSEI,;+.!

!!79@@1+!-1!

##

$7BW*&3'6*#.88*88+*13#

!!!!!!!!! 5Y<!U)(CF;E@>!(b0()E(@I(!


!

"#$%&! #'()!*+,-.!/(01)+! _D!4!565!

!!!!!! Y!�!T(:()!

!!!!!! 5!�!B(G!

!!!!!! 6!�!c,E+(!9!;1+!

!!!!!! D!�!A1'+!1S!+H(!+EC(!

!!

!!!!!!!!! 55<!A,'I,;9)!G(9P@(''!

!!!!!! Y!�!T(:()!

!!!!!! 5!�!B(G!

!!!!!! 6!�!c,E+(!9!;1+!

!!!!!! D!�!A1'+!1S!+H(!+EC(!

!!

!!!!!!!!! 56<!A,'I;(!*09'C'!

!!!!!! Y!�!T(:()!

!!!!!! 5!�!B(G!

!!!!!! 6!�!c,E+(!9!;1+!

!!!!!! D!�!A1'+!1S!+H(!+EC(!

!!

!!

MBW*&3'6*#.88*88+*13#

/9@>(!1S!C1:(C(@+!1S!G)E'+!K>1@E1C(+()L!

[)('',)(!K-.@9C(+()L!

!!

<"#IMNC?G?MC#

9<!U)9E;!A9PE@>!U('+!K=)C.!$@-E:E-,9;!U('+!d9++().O!5WJJL8!(b(I,+E:(!S,@I+E1@!

F<!=,-E1]V()F9;!e(9)@E@>!U('+!K/(.O!5WQJL8!A(C1).!

I<!%E>E+]'.CF1;!U('+!K',F+('+!S)1C!+H(!^=$*L8!0)1I(''E@>!'0((-!

Appendix B

Survey Script: Stakeholders Research -

Inquiring Developers

181

!!!!!!!

!

!"#"$%&"'()*+"(,-%../-'")

"#!$%&'!(%)*$+,!$%+!-&..+,+#$!*),$'!/.!$%+!'0,1+2!),+!-+'(,&3+-4!

010! 2.,'%3+4,-%.))

!

5%&'!&'!$%+!$+6$!$%)$!&'!*,+'+#$+-!$/!$%+!'0,1+2!*),$&(&*)#$'!)$!$%+!3+7&##&#78!

9)7+!:8!!

;3/0$!$%+!<=">?!*,/@+($8!

5%+! ?0,/*+)#! *,/@+($! <=">?! A<+#$B+! 0'+,! &#$+,.)(+'! ./,! +B-+,B2! *+/*B+C! DEEE470&-+F*,/@+($4+0G! &'!

-+1+B/*&#7!)!'/.$E),+!.,)H+E/,I!E%&(%!)BB/E'!-+1+B/*+,'!$/!+..&(&+#$B2!&#$+7,)$+!)((+''&3&B&$2!.+)$0,+'!

&#$/!$%+&,!)**B&()$&/#'4!;((+''&3&B&$2! &#!<=">?!)--,+''+'!)0$/H)$&(! &#$+7,)$&/#!/.!1),&/0'!0'+,! &#$+,.)(+!

$2*+'!D7+'$0,+J!1/&(+J!$)3B+$!9K'J!,+H/$+!(/#$,/B'J!+$(4G!)#-!)0$/H)$&(!)-)*$)$&/#!$/!$%+!*,+.+,+#(+'!)#-!

&H*)&,H+#$'! /.! )#! +B-+,B2! 0'+,4! "#! $%+! *,/@+($J! )! '/.$E),+! .,)H+E/,I! ./,! ,0#$&H+! )-)*$)$&/#! D<=">?!

.,)H+E/,IG!)#-!)!$//B3/6!./,!-+'&7#F$&H+!)-)*$)$&/#!)#-!'&H0B)$&/#!D<=">?!$//B3/6G!E&BB!3+!-+1+B/*+-4!

<=">?!*0$'!)!-+-&()$+-!./(0'!/#!$%+!+H+,7&#7!L23,&-!5M:!*B)$./,H'!)#-!'+,1&(+'!DK/##+($+-!5M'J!N+$F

5/*! O/6+'J! +$(4GJ! '0**/,$&#7! )**B&()$&/#! +#1&,/#H+#$'! )'! LOO5M! )'! E+BB! )'! *,/*,&+$),2! H&--B+E),+!

'/B0$&/#'! /.! 5M! H)#0.)($0,+,'4! 5%+'+! 5M! *B)$./,H'! %)1+! $%+! */$+#$&)B! $/! 3+(/H+! $%+! H)&#! H+-&)!

$+,H&#)B'!&#!$%+!0'+,'P!%/H+'J!-0+!$/!$%+&,!(/#1+#&+#(+!)#-!E&-+!)((+*$)#(+4!!

Q/,+!&#./,H)$&/#!/#!<=">?8!EEE470&-+F*,/@+($4+0!R)(+3//I!5E&$$+,!S!

!

9)7+!T8!!

;3/0$!$%&'!'0,1+28!

"#!$%&'!'0,1+2!$%+!<=">?!(/#'/,$&0H!E)#$'!$/!7)&#!-)$)!)3/0$!(0,,+#$!*,)($&(+!&#!N+$!$/*!3/6U(/##+($+-!

5M!-+1+B/*H+#$4! "$! &'! &#1+'$&7)$+-!%/E!)((+''&3&B&$2! &'!(0,,+#$B2!*+,(+&1+-!)#-!)**B&+-! &#! $%+! &#-0'$,24!

R&#)BB2J! $%+! '0,1+2! B+$'! $%+! *),$&(&*)#$'! 1/$+! ./,! H/'$! &H*/,$)#$! .+)$0,+'! /.! $%+! +#1&')7+-! <=">?!

.,)H+E/,I!)#-!$//B3/64!!

5%+!'0,1+2!)--,+''+'!*,/.+''&/#)B'!.,/H!$%+!./BB/E&#7!),+)'8!

V! ;**B&()$&/#!-+1+B/*+,'!

V! N/.$E),+!.,)H+E/,I!U!H&--B+E),+!U!$//B!-+1+B/*+,'!

V! ;((+''&3&B&$2!&#!"K5!

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

!

!!!!!!!

!

V! N+,1&(+!*,/1&-+,'!D+474!"95MJ!MW>J!L335M!*/,$)B'J!SG!

V! ='+,!&#$+,.)(+!$+(%#/B/72!-+1+B/*+,'!D&#*0$!)#-!/0$*0$J!+474!'*++(%!,+(/7#&$&/#!)#-!'2#$%+'&'J!

7+'$0,+J!%)*$&(!-+1&(+'J!&#*0$!-+1&(+'J!%+),&#7!)&-'J!7,)*%&()B!0'+,!&#$+,.)(+'J!SG!

V! L),-E),+! *B)$./,H! H)#0.)($0,+,'! D+474! N+$! $/*! 3/6+'J! 9K'J! H/3&B+! -+1&(+'J! K/#'0H+,!

?B+($,/#&('J!SG!

V! X+'+),(%!(/HH0#&$2!

V! N$)#-),-&Y)$&/#!3/-&+'!)#-!,+B)$+-!/,7)#&Y)$&/#'!

"#$%&'()*+*&',)-%./%)#$%-0(1$2%3*44%($+$*1$%'%-055'(2%./%)#$%-0(1$2%$1'40')*.,%($-04)-6%%

!

015! 6'%7"((-%./$))

70$-)*.,% 8,-3$(%

9B+)'+!&#-&()$+!2/0,!*,/.+''&/#)B!3)(I7,/0#-4!

DH0B$&*B+!'+B+($&/#'!*/''&3B+G!

!" ;**B&()$&/#!-+1+B/*H+#$!

!" Z+3!-+1+B/*H+#$!

!" L),-E),+!H)#0.)($0,&#7!U!-+1+B/*H+#$!

!" ='+,!&#$+,.)(+!-+'&7#!

!" X+'+),(%!

!" N$)#-),-'!

!" W$%+,8!

!

![[[[[[[[[[[[[[[[[[[[[[[[!

L)1+! 2/0! +1+,! E/,I+-! E&$%! +B-+,B2! /,!

-&')3B+-!*+,'/#'\!

DH0B$&*B+!'+B+($&/#'!*/''&3B+G!

!

!" ?B-+,B2! *+/*B+! D*/''&3B2! E&$%! )7+&#7F,+B)$+-!

&H*)&,H+#$'G!

!" >&')3B+-!*+/*B+!

>/! 2/0! $%&#I! $%)$! -+'&7#&#7! 'H),$! 0'+,!

&#$+,.)(+! ./,! +B-+,B2! /,! -&')3B+-! *+/*B+! ()#!

*,/1&-+! $%+H! 3+$$+,! )((+''! $/! '+,1&(+'! 0'&#7!

"#./,H)$&/#!5+(%#/B/72\!

!" ]+'!

!" Q)2!3+!

!" >+.&#&$+B2!#/!

!!!!!!!

!

)

018! 9+''".,)/&&$-4/,-%.:(;(,"<)3"(-=.)&'%4"(()7%')/&&$-4/,-%.)3"#"$%&<".,)

70$-)*.,% 8,-3$(%

Z%)$!%),-E),+!*B)$./,H!-/!2/0!(/#'&-+,!H/'$!

&H*/,$)#$! '! ./,! L23,&-! 5M! U! K/##+($+-! 5M!

'+,1&(+'!)#-!)**B&()$&/#'8!

D9B+)'+!*0$!#0H3+,'!.,/H!:F^!$/!*,&/,&$&Y+G!

!" N+$!$/*!3/6!

!" K/##+($+-!5M!*B)$./,H!D5M!-+1&(+!E&$%!

&#$+,#+$!)((+''G!

!" Q/3&B+!5M!*B)$./,H!

!" 9K!

!" <)H&#7!(/#'/B+!

!" "!-/#P$!I#/E!

Z%)$! /*+,)$&#7! '2'$+H! -/! 2/0! (/#'&-+,!

&H*/,$)#$! ./,! -+*B/2H+#$! /.! L23,&-! 5M!

)**B&()$&/#'!)#-!'+,1&(+'\!

D9B+)'+!*0$!#0H3+,'!.,/H!:F^!$/!*,&/,&$&Y+G!

!" Z&#-/E'!

!" ;#-,/&-!

!" Q++</!

!" <//7B+5M!

!" ;**B+!WN!.)H&B2!

!" _`^!

!" 9,/*,&+$),2!U!/$%+,8!

[[[[[[[[[[[[[[[[[[[[[[[!

!" "!-/#P$!I#/E!

Z%)$! )**B&()$&/#! +#1&,/#H+#$! U! ,0#$&H+! -/!

2/0! (/#'&-+,! H/'$! &H*/,$)#$! ./,! L23,&-! 5M!

)**B&()$&/#'!)#-!'+,1&(+'\!

!" Z+3!3,/E'+,!U!,+#-+,&#7!+#7&#+8!

[[[[[[[[[[[[[[[[[[[[[[[!

!" a)1)!

!" 4`?5!

!" ;-/3+!RB)'%!U!;&,!

!" KUKbbUW3@K!

!" W$%+,8!

[[[[[[[[[[[[[[[[[[[[[[[!

!" "!-/#P$!I#/E!

!!!!!!!

!

Z%)$!H&--B+E),+!U!'/.$E),+!.,)H+E/,I!-/!2/0!

(/#'&-+,! H/'$! &H*/,$)#$! ./,! L23,&-! 5M!

)**B&()$&/#'!)#-!'+,1&(+'\!

!" cS!&#*0$!5KSd!

!" W$%+,D'G8!

[[[[[[[[[[[[[[[[[[[[[[[!

[[[[[[[[[[[[[[[[[[[[[[[!

[[[[[[[[[[[[[[[[[[[[[[[!

[[[[[[[[[[[[[[[[[[[[[[[!

[[[[[[[[[[[[[[[[[[[[[[[!

!" "!-/#P$!I#/E!

KB)''&.2! $%+!),+)!/.!)**B&()$&/#'!2/0,! (/H*)#2!

)--,+''+'8!

!" O,/)-()'$&#7!D"95MJ!()3B+J!+$(4G!

!" N/(&)B!H+-&)!

!" K/HH0#&()$&/#!

!" ;H3&+#$!;''&'$+-!e&1&#7!

!" W$%+,8!

![[[[[[[[[[[[[[[[[[[[[[[[!

"#! H2! (/H*)#2! U! &#'$&$0$&/#! )((+''&3&B&$2! )#-!

&#(B0'&1+!-+'&7#!),+!(/#'&-+,+-!S!!

!" `/$!&H*/,$)#$4!

!" W**/,$0#&$&+'!./,!#&(%+!H),I+$'4!

!" K,&$&()B4!

Z%)$!I&#-!/.!$+'$'!-/+'!2/0,!(/H*)#2!)**B2!./,!

0'+,F(+#$+,+-!-+'&7#U+1)B0)$&/#\!

DH0B$&*B+!'+B+($&/#'!*/''&3B+G!

!" "#$+,#)B! $+'$'!E&$%!+H*B/2++'!$%)$!%)1+!3++#!

&#1/B1+-!&#!-+1+B/*H+#$!

!" N2'$+H)$&(!$+'$'!E&$%!+6$+,#)B!,+)B!%0H)#!+#-!

0'+,'!

!" "#1/B1+H+#$!/.!D0'+,F(+#$+,+-G!-+'&7#!+6*+,$'!

!" ;**B&()$&/#! /.! -+'&7#! 70&-+B&#+'! D)'! +474!

*,/1&-+-!32!?5N"J!ZfKJ!"5=JSG!

!" ;((+''&3&B&$2! $+'$! '/.$E),+! D+474! $/! 1+,&.2!

'$)#-),-!(/#./,H)#(+G!

!" ='+,!'&H0B)$&/#!DM&,$0)B!0'+,'G!

!" W$%+,8!

[[[[[[[[[[[[[[[[[[[[[[[!

!" `/#+!

!" "!-/#P$!I#/E!

!!!!!!!

!

".!2/0!(0,,+#$B2!'++!*,/3B+H'! &#!)**B2&#7!0'+,F

(+#$+,+-! -+'&7#! )#-! )((+''&3&B&$2! *,&#(&*B+'! &#!

2/0,! -+1+B/*H+#$! *,/(+''J! E%&(%! &'! $%+! H)&#!

&''0+\!

!" `/!+6*+,$!*+,'/##+B!)1)&B)3B+!

!" e)(I!/.!)1)&B)3B+!&#./,H)$&/#!

!" e)(I!/.!$&H+!

!" K/'$!

!" ;E),+#+''!./,!0'+,F(+#$+,+-!-+'&7#!

!" e)(I!/.!)((+''&3&B&$2!'/.$E),+!'0**/,$!

!" W$%+,8!

[[[[[[[[[[[[[[[[[[[[[[[!

!" "!-/#P$!I#/E!

Q2! (/H*)#2! U! &#'$&$0$&/#! *+,./,H'! 0'+'! $%+!

./BB/E&#7!)((+''&3&B&$2!'/.$E),+!'0**/,$8!

!" Z;"F;X";!

!" Q&(,/'/.$!Z&#-/E'!;((+''&3&B&$2!;9"!

!" _$!;((+''&3&B&$2!;9"!

!" Z+3F3,/E'+,!)((+''&3&B&$2!'0**/,$!

!" W$%+,8!

!

[[[[[[[[[[[[[[[[[[[[[[[!

!" `/#+4!

!" "!-/#P$!I#/E!

Z%)$! /*+,)$&#7! '2'$+H! -/! 2/0! 0'+! ./,!

-+1+B/*H+#$!U!-+'&7#!$//B'\!

DH0B$&*B+!'+B+($&/#'!*/''&3B+G!

!" Z&#-/E'!

!" e&#06!

!" Q)(WN!

!" W$%+,8!

[[[[[[[[[[[[[[[[[[[[[[!

!" `/#+!

!" "!-/#P$!I#/E!

Z%)$! -+1+B/*H+#$! $//B'! U! +#1&,/#H+#$'! -/!

2/0!(0,,+#$B2!H)&#B2!0'+!$/!-+1+B/*!./,!L23,&-!

5M!)**B&()$&/#'\!

DH0B$&*B+!'+B+($&/#'!*/''&3B+G!

!" Z+3FUL5QeF! ?-&$/,8!!

[[[[[[[[[[[[[[[[[[[[[[[!

!" QN!M&'0)B!N$0-&/!

!" ?(B&*'+!

!!!!!!!

!

!" ;-/3+!RB)'%!

!" W$%+,8!

!

[[[[[[[[[[[[[[[[[[[[[[[!

!" "!-/#P$!I#/E!

!

01>! ?"@+-'"<".,() 7%') A44"((-B-$-,;) C'/<"D%'E() 7%') F;B'-3) GH) A&&$-4/,-%.)

6$/,7%'<()

Z%)$! .+)$0,+'! '%/0B-! )#! )((+''&3&B&$2!

.,)H+E/,I!*,/1&-+\!

D9B+)'+!*0$!#0H3+,'!.,/H!:F^!$/!*,&/,&$&Y+G!

!" ='+,!H)#)7+H+#$!

!" X0#F$&H+!*,/.&B+!)-)*$)$&/#!

!" ='+,! &#$+,.)(+! )-)*$)$&/#! $/! $%+! 0'+,'!

*,+.+,+#(+'!)#-!&H*)&,H+#$'!

!" "#$+7,)$&/#!/.!1),&/0'!&#*0$U/0$*0$!-+1&(+'!

!" W$%+,D'G8!

!

[[[[[[[[[[[[[[[[[[[[[[[[!

Z%)$!),+!2/0,!,+g0&,+H+#$'!./,!)#!)((+''&3&B&$2!

.,)H+E/,I!/#!)!N5O!U!K/##+($+-5M!*B)$./,H8!

D9B+)'+!*0$!#0H3+,'!.,/H!:F^!$/!*,&/,&$&Y+G!

5%+!.,)H+E/,I!'%/0B-!3+!S!

!" S! (0'$/H&Y)3B+! D+474! )--&#7U,+H/1&#7!

(/H*/#+#$'G!

!" S!.,++!)#-!*03B&(B2!)1)&B)3B+!

!" S!/*+#!'/0,(+!!

!" S!&#$+7,)$+-!&#!+6&'$&#7!N5O!H&--B+E),+!

!" S! (/H*,&'+! 70&-+B&#+'! )#-! 3+'$! *,)($&(+'! ./,!

)**B&()$&/#!-+'&7#!

!" S!'$)#-),-&Y+-!

;,+! 2/0! &#$+,+'$+-! &#! -+1+B/*&#7! )**B&()$&/#'!

E&$%! )#! )((+''&3&B&$2! .,)H+E/,I! DB&I+! +474! $%+!

/#+!+#1&')7+-!32!<=">?G\!

!" ]+'4!

!" Q)23+4!

!" 9,/3)3B2!#/$4!

!" >+.&#&$+B2!#/$4!

Z%)$!I&#-'!/.!)0$/H)$&(!)-)*$)$&/#!E/0B-!2/0! !" 5+6$!'&Y+!

!!!!!!!

!

*,+.+,!H/'$!./,!2/0,!)**B&()$&/#'\! !" N(,++#!B)2/0$!

!" K/B/0,!

!" K/#$,)'$!

!" ;0-&/!.&B$+,&#7!

!" M/B0H+!

!" ;-)*$)$&/#!/.!&#*0$!-)$)!D+474!(0,'/,!'H//$%&#7!

./,!0'+,'!E&$%!$,+H/,G!

!" W$%+,D'G8!

[[[[[[[[[[[[[[[[[[[[[[!

[[[[[[[[[[[[[[[[[[[[[[!

[[[[[[[[[[[[[[[[[[[[[[!

5%+!<=">?!)**,/)(%!,+B&+'!/#!)!0'+,!+#,/BH+#$!

'$+*J!E%&(%!$)I+'!*B)(+!/#(+J!E%+#!$%+!0'+,! &'!

0'&#7! $%+! '2'$+H! ./,! $%+! .&,'$! $&H+4! "#! $%+!

+#,/BH+#$! $%+! '2'$+H! H+)'0,+'! 0'+,!

()*)3&B&$&+'U&H*)&,H+#$'!$/!)**B2!)-)*$)$&/#4!!

Z%)$! E/0B-! 3+! 2/0,! ,+g0&,+H+#$'! ./,! $%&'!

*,/(+-0,+\!

!

!

!" N%/,$!&#!$&H+!

!" 5%+!?#$+,$)&#&#7!

!" K0'$/H&Y)3B+!

!" O,)#-&#7!

!" 5%&'!*,/(+-0,+!E/0B-!#/$!3+!)((+*$)3B+4!

!

!

01I! ?"@+-'"<".,()7%')A&&$-4/,-%.)!"#"$%&<".,)6'%4"(()

<=">?! -+1+B/*'! )! '&H0B)$&/#! $//BJ! $%)$! ()#!

H&H&(! *+,(+*$&/#! )#-! (/7#&$&1+! ()*)3&B&$&+'! /.!

)#! +B-+,B2! 0'+,! $/! +1)B0)$+! )! 7&1+#! 0'+,!

&#$+,.)(+!-+'&7#4!

!

L/E! E/0B-! 2/0! B&I+! $/! 0'+! )! 0'+,! '&H0B)$&/#!

$//B!&#!$%+!)**B&()$&/#!-+'&7#!*,/(+''\!

!" 5/! !"#$"%%&! +6*+,&+#(+! %/E! )#! &H*)&,+-!

*+,'/#! E&BB! *+,(+&1+! $%+! 0'+,! &#$+,.)(+! D32!

*+,(+&1&#7!)!'&H0B)$+-!1&+E!/#!$%+!)**B&()$&/#G4!

!" 5/!"$'(!"')*"%%&! +1)B0)$+! $%+! .+)'&3&B&$2! /.!H2!

-+'&7#! ./,! '+1+,)B! 7,/0*'! /.! &H*)&,H+#$'! D32!

,0##&#7! )0$/H)$&(! 0'+,! &#$+,.)(+! $+'$'! &#! $%+!

3)(I7,/0#-G4!

L/E! E/0B-! 2/0! )**B2! $%+! '&H0B)$/,! &#! $%+!

-+1+B/*H+#$!*,/(+''\!

!" 5/!*+,./,H!&$+,)$&1+!,+.&#+H+#$4!

!" 5/! *+,./,H! )! .&#)B! g0&(I! (%+(I! &.! $%+! 0'+,!

&#$+,.)(+!*+,(+*$&/#! ./,! )#! &H*)&,+-!0'+,! &'! )'!

&#$+#-+-!32!$%+!-+1+B/*+,4!

!!!!!!!

!

!" "PH!#/$!&#$+,+'$+-4!

Z%&(%! )'*+($'! /.! $%+! 0'+,! &#$+,.)(+! -/! 2/0!

(/#'&-+,!H/'$!&H*/,$)#$!./,!'&H0B)$&/#\!

!" <,)*%&()B!0'+,!&#$+,.)(+!

!" ;0-&/!

!" "#*0$!-+1&(+'!D+474!,+H/$+!(/#$,/BG!

!" W$%+,8!

[[[[[[[[[[[[[[[[[[[[[[[!

".! 2/0! E/0B-! %)1+! $/! H)I+! 2/0,! )**B&()$&/#'!

)((+''&3B+! ./,! *+/*B+! E&$%! 1),&/0'! &H*)&,H+#$!

(/#.&70,)$&/#'J!E/0B-!2/0!*,+.+,!$/!7+$!S!

!" S!='+,!"#$+,.)(+!H),IF0*J!$%)$!"!()#!)--!$/!H2!

+6&'$&#7!="!-+'&7#!$/!H)I+!&$!)-)*$)3B+!

!" S! ='+,! &#$+,.)(+! $+H*B)$+'! $%)$! "! ()#! 0'+! $/!

)''+H3B+!)#!)-)*$)3B+!="4!

!" S!<+$! ,+(/HH+#-)$&/#'! .,/H! $%+! )((+''&3&B&$2!

'+,1&(+! /#! /*$&H&Y+-! ="! *),)H+$+,'! DB/0-#+''J!

./#$!'&Y+J!+$(4G!./,!(0,,+#$!0'+,4!

!" W$%+,8!

!

[[[[[[[[[[[[[[[[[[[[[[[!

!

!

01J! K+(-."(()

70$-)*.,% 8,-3$(%

Z&BB!$%+!)1)&B)3&B&$2!/.!)!<=">?!R,)H+E/,I!%+B*!

2/0!-+1+B/*!2/0,!)($&1&$2\!

!" >+.&#&$+B2!]+'!

!" _0&$+!]+'!

!" Q)2!3+!

!" 9,/3)3B2!#/$4!

!" >+.&#&$+B2!#/$4!

N+B+($! (,&$+,&)! $%)$! 2/0! (/#'&-+,! )'! H/'$!

&H*/,$)#$! &#! -+1+B/*&#7! 2/0,! )($&1&$2! 0'&#7! )!

<=">?!.,)H+E/,I4!

DR,/H!:DH/'$G!$/!6DB+''G!&H*/,$)#$G!

!" R)'$!-+1+B/*H+#$!

!" K/'$!+..+($&1+!

!" X+B&)#(+!/#!'$)#-),-'!

!" ;''&'$)#(+!

!" S!

!!!!!!!

!

!

01L! 6"'(%./$)3/,/)

70$-)*.,% 8,-3$(%

<+#-+,! !

R&,'$!#)H+! !

e)'$!#)H+! !

K/H*)#2! !

Q)&B!;--,+''!

Db!(/#.&,H!H)&B!)--,+''G!

!

9,/.+''&/#! !" K?W!!

!" K5W!

!" N+#&/,!+6+(0$&1+!

!" 9,/-0($!H)#)7+,!

!" 9,/@+($!H)#)7+,!

!" K/#'0B$)#$!

!" Z+3H)'$+,!

!" >+'&7#+,!

!" >+1+B/*+,!

!" X+'+),(%+,!

!" W$%+,8!

[[[[[[[[[[[[[[[[[[[[[[!

5/! E%&(%! ()$+7/,2! -/+'! 2/0,! (/H*)#2! U!

&#'$&$0$&/#!3+B/#7\!

DH0B$&*B+!'+B+($&/#'!*/''&3B+G!

!" ;**B&()$&/#!-+1+B/*+,'!

!" N/.$E),+! .,)H+E/,IJ! H&--B+E),+! /,! $//B!

*,/1&-+,'!

!" N+,1&(+!*,/1&-+,'!

!" N+$!$/*!3/6!!U!(/##+($+-!5M!H)#0.)($0,+,'!

!" W$%+,! %),-E),+! U! K/#'0H+,! +B+($,/#&('!

H)#0.)($0,+,'!

!!!!!!!

!

!" N+$!$/*!3/6!H&--B+E),+!*,/1&-+,'!

!" ='+,!&#$+,.)(+!$+(%#/B/72!-+1+B/*+,'!

!" ='+,!&#$+,.)(+!U!)**B&()$&/#!.,/#$+#-!-+'&7#+,'!

!" X+'+),(%!(/HH0#&$2! !

!" N$)#-),-&')$&/#! 3/-&+'! )#-! ,+B)$+-!

/,7)#&')$&/#'! !

!" W$%+,8!

![[[[[[[[[[[[[[[[[[[[[[[[!

L/E! H)#2! +H*B/2++'! ),+! E/,I&#7! &#! 2/0,!

(/H*)#2\!

!" :!h!:i!

!" ::!h!ji!

!" j:!h!jii!

!" ji:!F!:iii!

!" Q/,+!$%)#!:iii!

"!E/0B-!B&I+!$/!,+(+&1+!)!'0HH),2!/.!$%+!,+'0B$'!

/.! $%&'! '0,1+2! D'+#$! $/! H2! H)&B! )--,+''J! &.!

*,/1&-+-G8!

!" ]+'4!

!" `/4!

!

01M! N"."'/$)4%<<".,()

9B+)'+! E,&$+! )#2! 7+#+,)B!

(/HH+#$'! U! '077+'$&/#'! U! &-+)'!

2/0!(/#'&-+,!,+B+1)#$4!!

!

!

!

!

)

Appendix B. Survey Script: Stakeholders Research - Inquiring Developers 192

Appendix C

User Trial Script: GUIDE Adaptive

Interaction User Study

193

11 Script of the GUIDE Roadshow trials

!

"#$%&! #'()!*+,-.!/(01)+! 23!4!565!

!! "#$%&'()*('+,(-./01(2)345+)6('$%375((

!"#$%&'#()*+*,-*.,+*/)#,00-*1,+*/)##

$7+)1-,8(!+9(!0:)+;8;0:7+'!+9:+!<9:+!+9(.!:)(!=1;7=!+1!(>0();?(7+!;'!:!@A!+9:+!;'!:!B;++B(!C;+!-;DD()(7+!D)1?!

<9:+!+9(.!9:E(!:+!+9(;)!91?('F!@(BB!+9(?!+9;'!;'!:!0)1+1+.0(!1D!+9:+G!:7-!<9:+!<(!<:7+!;'!+1!(>0();?(7+!

<9:+!<;BB!C(!;D!+9(.!9:E(!:!'.'+(?!B;H(!+9;'!:+!+9(;)!91?('F!I(!:)(!;7+()('+(-!;7!+9(;)!10;7;17'G!+9()(!:)(!

71!);=9+!1)!<)17=!:7'<()'F!I(!<;BB!+).!+9(!'.'+(?!:7-!-;DD()(7+!<:.'!+1!,'(!;+F!!

!

*+(0!5J!

$7D1)?:+;17!0)('(7+(-!+1!+9(!,'()J!!

Explanation: �This is the UIA, it is what you would see when you start your TV for the first time� �Please listen to the system and try to perform the tasks they ask you to perform�!!

/(?:)H'!KD1)!+9(!',0()E;'1)LJ!!

§ @9(!,'()!'91,B-!M,'+!D1BB1<!+9(!;7'+),8+;17'F!

§ $D! ;+! ;'! 8B(:)! +9:+! +9(!,'()!7((-!9(B0G!=;E(! +9(! ;7'+),8+;17'!7(8('':).! D1)! +9(!0()'17! +1!C(!

:CB(!+1!0()D1)?!+9(!+:'H'!<;+9!+9(!-;DD()(7+!?1-:B;+;('F!!

§ N1?0B(+(!+('+!<;+9!?1,'(!;D!,'()4'.'+(?!D:;B'!KI;O:)-!1D!PQ!:00)1:89LF!!

§ /('+:)+!+('+!C.!)(B1:-;7=!0:=(!K(F=F!R3!H(.LF!

§ #'(!17B.!?1-:B;+.!:+!+9(!':?(!+;?(G!;D!71+!)(S,;)(-!-;DD()(7+!;7!+9(!'8);0+!!!!

§ Only if there�s something they can�t do, explain to them how they are supposed to interact K:7-!:'H! +9(?!<9:+! ;'!?;'';7=! D)1?!+9(!#$TG!1)!91<!81,B-!+9(!#$T! ;7'+),8+! +9(?!C(++()U!

A;-(1U!$?:=('U!TE:+:)!'91<;7=U!

§ @:H(!71+(!1D!(E().!10():+1)!;7+()E(7+;17!K(F=F!81?0B(+;7=!:!+('+!?:7,:BB.G!)('+:)+;7=!:!+('+G!

(+8FLF!!

!

P0():+1)!:8+;17J!*+:)+!#$T!:7-!'91<!?:;7!'8)((7F!

§ TD+()!+).;7=!+9(!)(?1+(!817+)1BG!)(B1:-!+9(!0:=(!KR3L!:7-!0()D1)?!+9(!+:'H!:=:;7!<;+9!+9(!

=.)1'810;8!)(?1+(!:7-!<;+9!+9(!;V:-F!

!

%:+:!+1!C(!81BB(8+(-J!

§ @:H(!71+(!1D!+9(!+;?(!<9(7!(:89!,'()!C(=;7'!,';7=!+9(!#$T!:7-!<9(7!+9(!,'()!(7-'!,';7=!;+F!

§ @:H(!71+(!1D!+9(!())1)'!:7-!+;?(!+9(!,'()!81??;+'!-,);7=!+9(!@W@!+('+F!

§ X1+(!+9(!:7'<()'!<9()(!)(S,;)(-F!

!

@:CB(+!S,('+;17J!

§ @:CB(+!S,('+;17!5J!T'H(-!:D+()!+9(!@W@Ygame: �What area on the tablet screen is preferred and possible to use?� Full screen, subareas in center, N, NE, E etc....__________________!!

2345&'6!#'()!$7;+;:B;O:+;17!:00B;8:+;17!K#$TL!Y!5!

TD+()!'91<;7=!+9(!:00B;8:+;17!

Z,('+;17J!


!

"#$%&! #'()!*+,-.!/(01)+! 2[!4!565!

· %1!.1,!+9;7H!+9:+!+9;'!0)18(''!;'!+11!B17=U!!

5! 6! \!

]('! X(,+):B! X1!

!

· %1!.1,!D;7-!;+!+;);7=U!!

5! 6! \!

]('! X(,+):B! X1!

!

· ^:E(!.1,!,7-()'+11-!<9.!<(!-1!+9(!#$TU!

· $D!.1,!9:E(!9:-!+9(!'.'+(?!:+!91?(G!<1,B-!.1,!=1!+9)1,=9!;+!1)!'H;0!;+U!

· $'!+9()(!:7.!S,('+;17!<9()(!+9(!01'';CB(!10+;17'!-1!71+!D;+!<;+9!.1,)!0)(D()(78('!K(F=F!+9(!C;==()!C,++17!'91<7!;'!71+!

C;=!(71,=9L!I9;89!17(U!

· I()(!+9(!;7'+),8+;17'!(:'.!(71,=9!+1!,7-()'+:7-U!

· %;-!.1,!71+;8(!:7.!89:7=('!;7!+9(!:00B;8:+;17!<9;B(!.1,!<()(!,';7=!;+U!K;7D()!:C1,+!+9(!:-:0+:+;17!;7!+9(!#$TLFU!

· %1!.1,!D((B!+9:+!+9(!E;)+,:B!89:):8+()!;'!9(B0;7=!1,+U!

· %1!.1,!9:E(!:7.!1+9()!81??(7+'!+1!:--!;7!+9;'!D;)'+!09:'(U!

!

!

*+(0!6J!A;)+,:B!89:):8+()!V/&R&/&XN&*!#

$7D1)?:+;17!+1!+9(!,'()!

�Now we will have a look 1D!-;DD()(7+! :'0(8+'!1D! +9(! E;)+,:B! 89:):8+()! +9:+! '91<(-!,0! ;7! +9(!C1++1?!

corner�#

P0():+1)!:8+;17J!!

*91<!+9(!0:)+;8;0:7+'! +9(!0;8+,)('!1D! +9(!-;DD()(7+!E;)+,:B! 89:):8+()'G!(;+9()!17! +9(!?:;7!VN!1)!17!:7!

(>+()7!B:0+10F!!

!

/(?:)H'!KD1)!+9(!',0()E;'1)LJ!!

P7B.!'91<!+9(!0;8+,)('!+9:+!81))('017-!+1!+9(!0:)+;8;0:7+'!(:)B;()!891;8(!;7!+9(!#$TF!R1)!(>:?0B(J!;D!+9(!

0:)+;8;0:7+!8911'(!?:B(!;7!+9(!#$TG!17B.!'91<!+9(!?:B(!0;8+,)('F!!

!

%:+:!+1!C(!81BB(8+(-J!

§ &B-()B.!:7-!:!.1,7=!E;)+,:B!89:):8+()G!<9;89!17(!'49(!0)(D()'U__________!

§ %;DD()(7+!-)('';7=!'+.B('!D1)?:B4;7D1)?:BG!<9;89!17(!'49(!0)(D()'U_____________!

!

!

P0():+1)!N1??(7+'J!!


!

"#$%&! #'()!*+,-.!/(01)+! 2`!4!565!

_______________________________________________________________________________!

________________________________________________________________________________!

!

7"#89:#!00-*1,+*/)#:&)&',-#()+'/;31+*/)#

$7D1)?:+;17!0)('(7+(-!+1!+9(!,'()J!

#@T!?:;7!'8)((7!

#

$7+)1-,8(!+9(!,'()!+1!+9(!817+(>+J!

§ $7+)1-,8(!+9(!,'()!C.!':.;7=!+9:+!<(!<;BB!71<!'((!<9:+!<1,B-!C(!+9(!?(7,!1D!+9(!+(B(E;';17!

:7-!-;DD()(7+!<:.'! +1! 817+)1B! ;+F!I(!<;BB! '((! +9(! -;DD()(7+! 89:77(B'! :7-!91<! +1! 7:E;=:+(!

C(+<((7!+9(?F!!

§ $7+)1-,8(!+9(!,'()!C.!':.;7=!+9:+!+9(!D;)'+!'8)((7!9(4'9(!'(('!<1,B-!C(!+9(!D;)'+!'8)((7!:D+()!

+,)7;7=!17!+9(!@AF!

§ &>0B:;7!+9:+!+9()(!:)(!71!E;-(1'!C(8:,'(!+9;'!;'!17B.!:!';?,B:+;17a!:'H!+9(!,'()!+1!;?:=;7(!

+9(!C:8H=)1,7-!;?:=(!;'!+9(!@A!0B:.;7=F!

/(?:)H'J!!

§ @:H(!71+(!1D! +9(!+;?(!<9(7!(:89!,'()!C(=;7'!,';7=!(:89!#@T!E()';17!:7-!<9(7!+9(!,'()!

(7-'!,';7=!;+b!!

§ @9(!D1B-()'!817+:;7!+9(!'(E():B!E()';17'!1D!+9(!#@T!(:89!,'()!?,'+!(>0();?(7+F!

§ @:H(!71+(!1D!(E().!10():+1)!;7+()E(7+;17!K(F=F!81?0B(+;7=!:!+('+!?:7,:BB.G!)('+:)+;7=!:!+('+G!

(+8FLF!!

§ N1?0B(+(!+('+!<;+9!?1,'(!;D!,'()4'.'+(?!D:;B'!KI;O:)-!1D!PQ!:00)1:89LF!!

§ $D! ;+! ;'! 8B(:)! +9:+! +9(!,'()!7((-!9(B0G!=;E(! +9(! ;7'+),8+;17'!7(8('':).! D1)! +9(!0()'17! +1!C(!

:CB(!+1!0()D1)?!+9(!+:'H'!<;+9!+9(!-;DD()(7+!?1-:B;+;('F!!

§ *+:)+!<;+9!+9(!717Y:-:0+(-!E()';17!<;+9!9:BD!1D!+9(!':?0B(!:7-!<;+9!+9(!:-:0+(-!E()';17!;7!

+9(!1+9()!9:BDF!

P0():+1)!:8+;17J!

§ *(+Y,0! A;',:B^,?:7*(7';7=! K8,)'1)! :)(:LF! V:.! :++(7+;17! 71+! +1! C(! E;';CB(! +1! +9(! H;77(8+G!

17B.!,'()!'91,B-!C(!E;';CB(F!!

§ #'()!9:'!+1!)(?:;7!;7!817'+:7+!01';+;17F!

§ c1:-!+9(!:00B;8:+;17!+1!+9(!'8)((7F!

!

<"#89:#!;,0+&;#=&'%*/)##

!

P0():+1)!:8+;17J!!

§ *+:)+!&V"!:-:0+(-!E()';17!K<;+9!1)!<;+91,+!:,-;1G!-(0(7-;7=!17!,'()!0)1D;B(LF!

§ c11H! :+! +9(! D;B(! )(=:)-;7=! +9(! 0)1D;B('! :7-! '((!<9;89! 0)1D;B(! +1! ),7! D1)! (:89!,'()F! $7'()+!

�?profile=X� at the end of the URL bar. X is the id of the profile!!

$7D1)?:+;17!0)('(7+(-!+1!+9(!,'()J!

§ T00B;8:+;17!?:;7!'8)((7!K@A!C:8H=)1,7-!;?:=(L!

§ @(BB!+9(!0:)+;8;0:7+!+9:+!9(4'9(!;'!:BB1<(-!+1!'<;+89!?1-:B;+.!<9(7(E()!+1!+).!<9;89!17(!;'!

+9(!0)(D()(-F!!

!

$7+)1-,8(!+9(?!+1!+9(!817+(>+J!!


!

"#$%&! #'()!*+,-.!/(01)+! 2d!4!565!

§ @9(!,'()!'<;+89('!17!9;'!@AG!;'!)(81=7;O(-!C.!+9(!"#$%&!'.'+(?G!'1!9;'49()!0)1D;B(!;'!:B)(:-.!

H71<7F!

§ T'! +9;'! ;'!71+! +9(! D;)'+!:88(''! +1!9;'49()!@AG! +9(! '8:77;7=!1D! +9(! B;E(!89:77(B'!9:'!:B)(:-.!

C((7!0()D1)?(-!'1!B;E(!89:77(B'!:)(!H71<7!:7-!8:7!C(!-;'0B:.(-F!@9:+!;'!+1!':.G!+9:+!+9(!

,'()!-;)(8+B.!:88(''('!+9(!D;)'+!B;E(!89:77(B!:E:;B:CB(F!!

!

$7+)1-,8(!+9(?!+1!+9(!?1-:B;+;('J!!

§ V1;7+;7=F!!

§ Speech. Say the keyword �GUIA/GUIDE/GEIDE� before giving the command.!§ Pointing and speech (pointing and saying �select�)!§ #';7=!+9(!=.)1'810;8!)(?1+(!817+)1B!K<9;+(!)(?1+(!817+)1BL!

§ #';7=!+9(!'+:7-:)-!)(?1+(!817+)1B!KCB:8H!)(?1+(!817+)1BL!

§ #';7=!+9(!+:CB(+!

#

/(?:)H'J!

§ They should know how to do this from the User Initialization Application. If they don�t, take a note, and try to understand why they didn�t get it. (Would they need an explanation for (E().!:00B;8:+;17UL!

§ I9(7!:'H;7=!:C1,+!+9(!?1-:B;+.!-1!;+!:B<:.'!:D+()!S,('+;17!5G!6G!3!:7-!2G!;D!+9(!,'()!<:7+'!

+1!'<;+89!?1-:B;+.!;7!:71+9()!S,('+;17!71+(!;+!-1<7G!1+9()<;'(!D;BB!;7!+9(!?1-:B;+.!+9(!,'()!

891'(!+1!0()D1)?!+9(!+:'H!<;+9F!K$+!?;=9+!=(+!+11!)(0(+;+;E(!:7-!:771.;7=!1+9()<;'(L!

#

>,+,#+/#5&#1/--&1+&;6!

X1+(!-1<7!:7'<()'!:D+()!(:89!+:'H!<9()(!)(S,;)(-!K4,'?&;#@*+A#5/-;L!

2345&'6!&V"!�!T-:0+(-!Y5!

Z,('+;17'J!

Please open the menu (open the menu by pressing �ok� on the remote control, or by saying �GUIA_MENÚ/GUIDE_MENU� )!

!"#$%&'($))&*+&*"(&($,"-./01&#2$&-'',/3-#/"0&*"(&#(4/01&5"#2&6".-,/#/$)!

· BC>!D(EF#9G8H8G82<8I#�#!J+&'#%/-K*)L#+A&#+,%?#@A*1A#/)&#*%#M/3'#0'&J&''&;#4/;,-*+MN##

· !DE8G2!E(=8# 78O!=(C$G# �# >/# M/3# -*?&# +A&%&# @,M%# /J# 4,?*)L# +A&# 4&)3# ,00&,'N# (J# )/+P# ;/# M/3# A,K&# ,)M#

%3LL&%+*/)#/)#A/@#+/#;/#+A*%N#

!

!

P0():+1)!N1??(7+'J!

!________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Y6!

Z,('+;17'!'+(0!5J!

#'(! +9(! D;)'+! ?(7,! +1! 7:E;=:+(! +9)1,=9! 89:77(B'! K,0! :7-! -1<7! :))1<'! 17! +9(! ';?0B(! O:00;7=! ?(7,L!!

Let�s try all the modalities: !


!

"#$%&! #'()!*+,-.!/(01)+! 22!4!565!

;F V1;7+;7=F!K01;7+;7=!+1!+9(!,0!:7-!-1<7!:))1<'L!

;;F Speech. (saying �GUIDE_CHANNEL_UP�, �GUIDE_CHANNEL_DOWN�)!;;;F Multimodal interaction (pointing to the up and down arrows and saying �select�!;EF #';7=!+9(!=.)1'810;8!)(?1+(!817+)1BF!K01;7+;7=!+1!+9(!,0!:7-!-1<7!:))1<'L!

EF #';7=!+9(!+):-;+;17:B!)(?1+(!817+)1BF!K0)('';7=!,0!:7-!-1<7!:))1<'L!

E;F #';7=!+9(!+:CB(+F!K,';7=!+9(!01;7+()!+1!01;7+!+1!+9(!,0!:7-!-1<7!:))1<'a!,';7=!+9(!&V"!17!+9(!+:CB(+F!

· BC>!D(EF#9G8H8G82<8I#�#!J+&'#%/-K*)L#+A&#+,%?#@A*1A#/)&#*%#M/3'#0'&J&''&;#4/;,-*+MN##

· !DE8G2!E(=8#78O!=(C$G#�#Q/3-;#M/3#-*?&#+/#+A*%#*)#,)/+A&'#@,MN#R'&L,';*)L#,)M#/J#+A&#4/;,-*+*&%"#

Z,('+;17'!*+(0!6J!

$7+()?(-;:+(!@:CB(+!S,('+;17'J!

!

()J/'4,+*/)#0'&%&)+&;#+/#+A&#0,'+*1*0,)+6#

!

&>0B:;7!+9:+!+9()(!:)(!D,)+9()!01'';C;B;+;('!<;+9!+9(!+:CB(+G!',89!:!:'!<(CYC)1<';7=G!)(:-;7=!7(<'0:0()G!<:+89!(Y8B:''('!(+8FL!

!

C0&',+/'#,1+*/)6#

!

P0(7!+9(!&V"!?(7,!17!+9(!+:CB(+!:7-!'91<!+9(!,'()!+9:+!+9(!?(7,!8:7!C(!'91<7!-;)(8+B.!17!+9(!;V:-F!P0(7!+9(!<(CYC)1<'()!

:7-!'91<!+9(!0:)+;8;0:7+!'1?(+9;7=!KD1)!(>:?0B(!7(<'0:0()LG!'91<!+9(!(Y8B:''('F!!

· E,5-&+# S3&%+*/)# T6# RC5%&'K,+*/)"# O/@# ;/# 3%&'%# A/-;# +A&# +,5-&+N# ()# +A&# -,0P# -*?&# ,# '&4/+&# 1/)+'/-#

&+1NUUUUUUUUUUUUUU#

· E,5-&+#S3&%+*/)#V6#>/#M/3#J&&-#1/4J/'+,5-&#3%*)L#+A&#+,5-&+#R+A&#13'%/'"N#WXY#D*?&'+#%1,-&#R(#-*?&#*+#,#-/+XI!don�t!like!it!

,+#,--"UUUUUUUUUUUUU#

· E,5-&+#S3&%+*/)#Z6#(%#*+#&,%M#+/#[%@*+1A[#5&+@&&)#K*&@*)L#+A&#+,5-&+#%1'&&)#,);#%&-&1+*)L#,#1A,))&-#,);#4/K*)L#5,1?#

/)#+A&#E=#%1'&&)#@*+A#+A&*'#&M&%N#WXY#D*?&'+#%1,-&#RK&'M#&,%MXK&'M#;*JJ*13-+"UUUUUUUUUUUU#

!

!

!

P0():+1)!N1??(7+'J!!

_______________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Y\!

Z,('+;17'J!

VB(:'(!'(B(8+!:71+9()!89:77(BF!R1)!(>:?0B(G!8:7!.1,!=1!to channel �<ChannelName>�? Let�s try with your preferred modality. !

· BC>!D(EF#9G8H8G82<8I#�#>/#M/3#1/)J*'4#)/@#R,J+&'#%/-K*)L#+A&#+,%?%"#+A,+#+A*%#*%#M/3'#0'&J&''&;#4/;,-*+M#J/'#+A*%#

+,%?N#

· !DE8G2!E(=8#78O!=(C$G#�#Q/3-;#M/3#-*?&#+/#+A*%#*)#,)/+A&'#@,MN#R'&L,';*)L#,)M#/J#+A&#4/;,-*+*&%"#


!

"#$%&! #'()!*+,-.!/(01)+! 5ee!4!565!

!

!

!

P0():+1)!N1??(7+'J!!

________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Y!f!

Z,('+;17J!

VB(:'(!:88(''!+9(!D,BB!-('8);0+;17!17!+9(!8,))(7+!0)1=):?!1D!+9;'!89:77(BF!N:7!.1,!+(BB!,'!<9(7!+9(!0)1=):?!'+:)+(-!:7-!<9(7!

;+! D;7;'9('U! @).! +1! D;7-! +9(! ;7D1)?:+;17! ,';7=! .1,)! 0)(D())(-! ?1-:B;+.F!!

X1+(J! @9;'! +:'H!<;BB! C(! 0()D1)?(-!<;+9! :7-!<;+91,+! @A! C:8H=)1,7-!71;'(! 17! +9(! &V"_T%TV@&%! K71;'(! 9:'! +1! C(! 8)(:+(-!

'1?(91<FFFL!

· Q/3-;#M/3#0'&J&'#A,K*)L#+A&#E=#%/3);#/)#/'#/JJ#@A*-&#0&'J/'4*)L#+A*%#+,%?NUUUUUUUUUUUUUUUUUUU!

· BC>!D(EF# 9G8H8G82<8I# �# >/# M/3# 1/)J*'4# )/@# R,J+&'# %/-K*)L# +A&# +,%?%"# +A,+# +A*%# *%# M/3'# 0'&J&''&;#

4/;,-*+MNUUUUUUUUUUUUUUUUUUUU!

· !=!E!G#�#>/#M/3#-*?&#+A&#!K,+,'N#>/#M/3#+A*)?#*+#%A/3-;#5&#0'&%&)+#,+#,--#+*4&%NUUUUU____!

!

!

!

P0():+1)!N1??(7+'J!

!________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Y3!

Z,('+;17J!

VB(:'(! +).! +1! =(+! :88(''! +1! +9(! ',C+;+B('F! N:7! .1,! :8+;E:+(! +9(! ',C+;+B('U! #'(! .1,)! 0)(D())(-! ?1-:B;+.F!!

X1+(J!@,)7!+9(!=):E;+.!<(BB'!1741DD!K$70,+!T-:0+:+;17L!-(0(7-;7=!17!+9(!,'()!0)1D;B(F!!


4/;,-*+MNUUUUUUUUUUUUUUUUU#

!

P0():+1)!N1??(7+'J!!

________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!


!

"#$%&! #'()!*+,-.!/(01)+! 5e5!4!565!

2345&'6!&V"!�!T-:0+(-!Y![!

Z,('+;17J!

VB(:'(!71<!=1!C:8H!+1!+9(!?:;7!'8)((7F!V()D1)?!+9;'!'+(0!,';7=!.1,)!0)(D())(-!?1-:B;+.F!!


4/;,-*+MNUUUUUUUUUUUUUUUUUUUUUUUUUU#

!

!

!

P0():+1)!N1??(7+'J!

!________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Y`!

Z,('+;17J!

Please choose one modality and switch to �<ChannelName>�"!


4/;,-*+MNUUUUUUUUUUUUUUUUUUU!

!

!

!

P0():+1)!N1??(7+'J!

!________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Yd!

Z,('+;17J!

VB(:'(!=1!+1!+9(!<91B(!B;'+!1D!89:77(B'G!C.!,';7=!:7.!?1-:B;+.F!&>0B1)(!+9(!B;'+!1D!89:77(B'!:7-!7:E;=:+(!+1!+9(!B:'+!89:77(B!17!

+9(!B;'+F!I9;89!17(!;'!+9(!B:'+!89:77(BU!


4/;,-*+MNUUUUUUUUUUUUUUUUUUUUUUUU#

!

!

!

P0():+1)!N1??(7+'J!!

________________________________________________________________________________!

________________________________________________________________________________!


!

"#$%&! #'()!*+,-.!/(01)+! 5e6!4!565!

________________________________________________________________________________!

!

2345&'6!&V"!�!T-:0+(-!Y2!

Z,('+;17J!

N911'(!:!89:77(BG!=1!+1!+9(!?1)(!10+;17'!?(7,!:7-!7:E;=:+(! +9)1,=9!+9(!0)1=):??('F!@9(7!=1!C:8H!+1!+9(!89:77(B!?(7,F!

#'(!.1,)!0)(D())(-!?1-:B;+.F!!


4/;,-*+MNUUUUUUUUUUUUUUUUUUU!

· :828G!D#$2>8GIE!>(2:#�!'&#M/3#)/@#J,4*-*,'#@*+A#+A&#,00-*1,+*/)N#>/#M/3#-*?&#*+NUUUUUUUUUUUUUUUUU!

!

!

!

P0():+1)!N1??(7+'J!!

________________________________________________________________________________!

________________________________________________________________________________!

________________________________________________________________________________!

!

D) GENERAL QUESTIONS

5F I$7\8<E(=8#!<<89E!2<8#Y!^1<!<(BB!-;-!+9(!'.'+(?!',001)+!.1,!;7!'1BE;7=!+9(!+:'H'U!

5! 6! \! f! 3!

@9(!'.'+(?!9:'!71+!

',001)+(-!?(!:+!:BB!

@9(!'.'+(?!9:'!',001)+(-!

?(!17B.!;7!'1?(!0:)+'!1D!

+9(!+:'H!

#7-(8;-(-! @9(!'.'+(?!9:'!

',001)+(-!?(!;7!:B?1'+!

:BB!+9(!+:'H'!

@9(!'.'+(?!9:'!

',001)+(-!?(!:+!(E().!

?1?(7+!

!

6F $D!.1,!+9;7H!+9(!'.'+(?!'91,B-!0)1E;-(!:!C(++()!',001)+!K?:;7B.!)('017'('!5!:7-!6LG!0B(:'(!

+9;7H!1D! +9(!+(8971B1=;('!.1,!:)(!8,))(7+B.!,';7=! ;7!.1,)!-:;B.! B;D(!:7-!<);+(!-1<7!4! +(BB!,'!

91<!;+!81,B-!C(!;?0)1E(-J!!

!

!

!

!

T]#!>!9E!E(C2##!

\F %1! .1,! 0()8(;E(! +9:+! +9(! D1BB1<;7=! (B(?(7+'! 9:E(! C((7! 81))(8+B.! :-:0+(-! +1! .1,! E;',:BG!

9(:);7=G!?1+1)!:7-!81=7;+;E(!:C;B;+;('U!!


!

"#$%&! #'()!*+,-.!/(01)+! 5e\!4!565!

X1+(J!K+9(!:7'<()!'91,B-!C(!.('!1)!71F!$D!;+!;'!71G!:7!(>0B;8:+;17!;'!)(S,;)(-F!

· $817'!

5F *;O(!

6F N1B1,)!!

\F c18:+;17!1D!+9(!C,++17'!!

· @(>+!

5F *;O(!

6F N1B1,)!!

\F R17+!

· TE:+:)!

· A1;8('!

5F A1B,?(!!

6F *0((-!1D!'0((89!

\F NB(:)!

· I:.!1D!;7+():8+;17!

5F @1!D1BB1<!17(!'+(0!+1!:71+9()!

6F @1!H71<!<9:+!;'!(>0(8+(-!C.!?(!

\F &:'.!+1!B(:)7!91<!+1!,'(!

!

fF I1,B-!.1,!B;H(!+1!,'(!:!81?C;7:+;17!1D!?1-:B;+;('U!I9;89!?1-:B;+;('!<1,B-!.1,!81?C;7(U!

R1)!<9:+!)(:'17U!!

!

3F #'(!1D!T,-;1!P,+0,+J!!

X1+(J!T'H!17B.!+9(!17(!+9:+!81))('017-'!+1!+9(!+);:BJ!

! R1)!+9(!,'()'!<91!+('+(-!17B.!&V"_T%TV@I*P#X%!

!

Q/3-;#M/3#0'&J&'#+/#A,K&#,3;*/#/3+03+#@A*-&#),K*L,+*)LN#RJ/'#&^,40-&6#,#K/*1&#%,M*)L#

+A&# ),4&# /J# +A&# 1A,))&-%P# '&,;*)L# +A&# 0'/L',44&%# ;&%1'*0+*/)%P#

&+1]"UUUUUUUUUUUUUUUU!

! R1)!+9(!,'()'!<91!+('+(-!17B.!&V"_T%TV@!

!

Q/3-;#M/3#0'&J&'#)/+# +/#A,K&#,3;*/#/3+03+#@A*-&#),K*L,+*)LN# RJ/'# &^,40-&6# ,# K/*1&#

%,M*)L# +A&# ),4&# /J# +A&# 1A,))&-%P# '&,;*)L# +A&# 0'/L',44&%# ;&%1'*0+*/)%P#

&+1]"UUUUUUUUUUUUU!


!

"#$%&! #'()!*+,-.!/(01)+! 5ef!4!565!

! R1)!+9(!,'()'!<91!+('+(-!C1+9!+9(!&V"_T%TV@!:7-!&V"_T%TV@I*P#X%J!!

!

QA*1A#K&'%*/)#A,K&#M/3#0'&J&''&;NUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUUU!

!

P0():+1)!N1??(7+'J!!

______________________________________________________________________________!

______________________________________________________________________________!

____________________________________________________________________________!

!

!

8"#89:#�#2/)#,;,0+&;#K&'%*/)#

P0():+1)!:8+;17'J!

§ c1:-!717Y:-:0+(-!&V"!:00B;8:+;17F!!

§ NB1'(!$70,+!T-:0+:+;17F!

/(?:)H'J!!

§ @9(! 1)-()! 1D! +9(! +('+'! <;+9! +9(! :-:0+(-! :7-! +9(! 717Y:-:0+(-! E()';17! '91,B-! C(!

):7-1?;O(-G!1)!E:).!'1!+9:+!3eg!1D!+9(!+;?('!+9(!:-:0+(-!E()';17! ;'!,'(-!D;)'+G!:7-!1+9()!

3eg!+9(!717Y:-:0+(-!E()';17!;'!+9(!D;)'+!+1!C(!,'(-F!

§ @:H(!71+(!1D! +9(!+;?(!<9(7!(:89!,'()!C(=;7'!,';7=!(:89!#@T!E()';17!:7-!<9(7!+9(!,'()!

(7-'!,';7=!;+b!

#

$7D1)?:+;17!0)('(7+(-!+1!+9(!,'()J!

&>0B:7:+;17J!!

�You will now see another ver';17! 1D! +9(! '.'+(?G! :! E()';17! +9:+! ;'! 71+! :-:0+(-! +1! .1,)! 0()'17:B!

0)(D()(78('F!!]1,!<;BB!0()D1)?!:!'();('!1D!+:'H'G!:7-!;7!+9;'!'(8+;17!<(!<;BB!17B.!,'(!+9(!'+:7-:)-!)(?1+(!

control (the black one).�!

!

2345&'6!&V"!�!X17!T-:0+(-!Y5!

Z,('+;17J!

VB(:'(!10(7!+9(!menu (open the menu by pressing �ok� on the remote control)!


!

"#$%&! #'()!*+,-.!/(01)+! 5e3!4!565!

BC>!D(EF#9G8H8G82<8I#�#Q/3-;#M/3#0'&J&'#+/#;/#*+#3%*)L#/+A&'#4/;,-*+M#/J#*)+&',1+*/)N!

!

!

!

P0():+1)!N1??(7+'J!!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

2345&'6!&V"!�!X17!T-:0+(-!Y6!

Z,('+;17'J!

#'(!+9(!D;)'+!?(7,!+1!7:E;=:+(!+9)1,=9!89:77(B'!K,0!:7-!-1<7!:))1<'!17!+9(!)(?1+(!817+)1BL!

!DE8G2!E(=8#78O!=(C$G#�#Q/3-;#M/3#-*?&#+/#+A*%#*)#,)/+A&'#@,M#R0'&%%*)L#/+A&'#?&M%N"!


!

!

P0():+1)!N1??(7+'J!!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

!

2345&'6!&V"!�!X17!T-:0+(-!Y\!

Z,('+;17J!

*<;+89!+1!hN9:77(B!7:?(ib!


!

!

P0():+1)!N1??(7+'J!!


!

"#$%&! #'()!*+,-.!/(01)+! 5e[!4!565!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

2345&'6!&V"!�!X17!T-:0+(-!Yf!

Z,('+;17J!

VB(:'(!+).!+1!=(+!:88(''!+1!+9(!',C+;+B('F!N:7!.1,!:8+;E:+(!+9(!',C+;+B('U!!


!

!

P0():+1)!N1??(7+'J!!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

2345&'6!&V"!�!X17!T-:0+(-!Y3!

Z,('+;17J!

VB(:'(!:88(''!+9(!D,BB!-('8);0+;17!17!+9(!8,))(7+!0)1=):?!1D!+9;'!89:77(BF!N:7!.1,!+(BB!,'!<9(7!+9(!0)1=):?!'+:)+(-!

:7-!<9(7!;+!D;7;'9('U!!


!

!

P0():+1)!N1??(7+'J!!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

!

2345&'6!&V"!�!X17!T-:0+(-!Y[!

Z,('+;17J!

VB(:'(!71<!=1!C:8H!+1!+9(!?:;7!'8)((7F!!


P0():+1)!N1??(7+'J!!


!

"#$%&! #'()!*+,-.!/(01)+! 5e`!4!565!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

2345&'6!&V"!�!X17!T-:0+(-!Y`!

Z,('+;17J!

Please select another channel. For example, can you go to �la sexta�? !

BC>!D(EF#9G8H8G82<8I#�#Q/3-;#M/3#0'&J&'#+/#;/#*+#3%*)L#/+A&'#4/;,-*+M#/J#*)+&',1+*/)N#

P0():+1)!N1??(7+'J!!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

!

!

2345&'6!&V"!�!X17!T-:0+(-!Yd!

Z,('+;17J!

VB(:'(!=1!+1!+9(!<91B(!B;'+!1D!89:77(B'F!&>0B1)(!+9(!B;'+!1D!89:77(B'!:7-!7:E;=:+(!+1!+9(!B:'+!89:77(B!17!+9(!B;'+F!I9;89!

17(!;'!+9(!B:'+!89:77(BU!



!P0():+1)!N1??(7+'J!!

______________________________________________________________________________

______________________________________________________________________________

______________________________________________________________________________!

2345&'6!&V"!�!X17!T-:0+(-!Y2!

Z,('+;17J!

N911'(!:!89:77(BG!=1!+1!+9(!?1)(!10+;17'!?(7,!:7-!7:E;=:+(!+9)1,=9!+9(!0)1=):??('F!@9(7!=(+!C:8H!+1!+9(!89:77(B!

?(7,F!


:828G!D#$2>8GIE!>(2:#�!'&#M/3#)/@#J,4*-*,'#@*+A#+A&#,00-*1,+*/)N#>/#M/3#-*?&#*+N!

!

!


!

"#$%&! #'()!*+,-.!/(01)+! 5ed!4!565!

H"#:&)&',-#_3&%+*/)%6##

!

!!

W]#I$7\8<E(=8#!<<89E!2<8#X#O/@#@&--#;*;#+A&#%M%+&4#%300/'+#M/3#*)#%/-K*)L#+A&#+,%?N#

!5! 6! \! f! 3!

@9(!'.'+(?!9:'!71+!

',001)+(-!?(!:+!:BB!

@9(!'.'+(?!9:'!',001)+(-!

?(!17B.!;7!'1?(!0:)+'!1D!

+9(!+:'H!

#7-(8;-(-! @9(!'.'+(?!9:'!

',001)+(-!?(!;7!:B?1'+!

:BB!+9(!+:'H'!

@9(!'.'+(?!9:'!

',001)+(-!?(!:+!(E().!

?1?(7+!

!

$D!.1,!+9;7H!+9(!'.'+(?!'91,B-!0)1E;-(!:!C(++()!',001)+!K?:;7B.!)('017'('!5!:7-!6LG!0B(:'(!+9;7H!1D!+9(!

+(8971B1=;('!.1,!:)(!8,))(7+B.!,';7=!;7!.1,)!-:;B.!B;D(!:7-!<);+(!-1<7!4!+(BB!,'!91<!;+!81,B-!C(!;?0)1E(-J!!

!

T]#>*;#+A&#89:#,00-*1,+*/)#-//?#-*?&#+A&#/)&#M/3#3%&#/)#M/3'#E=À/4&N#>*;#*+#J&&-#'&,-*%+*1NUUUUUUUUU#

#

V]#>/#M/3#0&'1&*K&# +A,+# +A&# J/--/@*)L#&-&4&)+%#A,K&#5&&)#1/''&1+-M#,;,0+&;# +/#M/3#K*%3,-P#A&,'*)LP#

4/+/'#,);#1/L)*+*K&#,5*-*+*&%N##

!"#$%&7#2$&-0)8$(&)2"9,.&5$&4$)&"(&0":&;*&/#&/)&0"<&-0&$=',/3-#/"0&/)&($>9/($.:?!

· $817'!

! :L!*;O(!!

! CL!N1B1,)!!

! 8L!c18:+;17!1D!+9(!C,++17'!!

· @(>+!

! :L!*;O(!

! CL!N1B1,)!!

! 8L!R17+!

Appendix C. User Trial Script: GUIDE Adaptive Interaction User Study 208

Appendix D

Survey: Older Adults, Technology and

Facebook

209

3/6/2017 Survey: Older Adults, Technology and Facebook

https://docs.google.com/forms/d/19JoTzeSSefBS2qVV2VYNmcy2JraDS_dC8uXCyeSvzEY/edit 1/8

Survey: Older Adults, Technology and FacebookMy name is José Coelho, I'm PhD student at Informatics Department of Faculty of Sciences at University of Lisbon, and my PhD is focused on older adults and the use of Social Network Service and TV by this segment of the population.

This survey goal is to collected valuable information regarding older adults habits concerning technology which can later originate the development of new technology focused on this population needs and difficulties.

The survey will take from 5 to 10 minutes to answer. Participants are asked to answer to questions related the use of Facebook, Television and also to classify themselves in terms of technological skills and necessities.

Any data collected will be anonymous. It will not be used to identify participants in any way and will be stored in a secure place to maintain confidentiality. The participant can, at any time, abandon the questionnaire, and no data will be saved.

We thank you for your collaboration.

*Obrigat!rio

Characterization

1. Please select the age range you are in: *

Marcar apenas uma oval.

Less than 60

Between 60 and 64

Between 65 and 70.

Between 71 and 75.

Between 76 and 80.

More than 80.

2. Please select your gender: *


Male

Female



3. Please select your level of education:


4th grade

9th grade

12th grade

Graduate Degree

Masters/PhD

None

4. With how many persons do you live? *


Alone

1 Other

2 to 3

3 to 5

More than five

5. How do you see yourself concerning technology skills and needs? *


I have different needs from younger users concerning the use of technology.

I have different skills from younger users concerning the use of technology.

I have different skills and needs from younger users concerning the use of technology.

I have no different skills or needs from younger users concerning the use of technology

6. How do you rate yourself in terms of technical expertise? *


I have a lot of difficulties with technology

I have some difficulties with technology

I have minor issues dealing with technology

I�m comfortable working with some technology

I use any typical technology available nowadays.



7. Which impairments or difficulties you have? *

Choose as many as you need.Marque todas que se aplicam.

Minor hearing impairments - trouble hearing some sounds

Some hearing impairments - trouble hearing what people say

Minor vision impairments - vision ok with glasses

Some vision impairments - trouble reading even with glasses, troubles seeing at distance.

Minor motor impairments - trouble moving, or making some movements.

Some motor impairments - tremors, trouble moving arms or making a lot of movements,

difficulties grasping a mouse, positioning and controlling a cursor, etc..

Minor cognitive impairments - trouble remembering short or long term events.

Some cognitive impairments - trouble remembering short and long term events and making

simple associations.

No impairments.

Outro:

Facebook

8. Are you aware of what Facebook is? *


Yes

No

9. Do you use Facebook? *


Yes

No Ir para a pergunta 18.

10. How regularly you use Facebook?


Rarely

Once a Week

2 to 3 Times a Week

Everyday

More than once a day

Classify the next sentences in terms of accordance regarding the use of Facebook.



11. I use Facebook to get information about what's going on in the world.


1 2 3 4 5

Strongly Disagree Strongly Agree

12. I use Facebook to know what's going with my relatives and close friends.


1 2 3 4 5


13. I use Facebook to know what's going with all my friends.


1 2 3 4 5


14. I use Facebook to make new friendships.


1 2 3 4 5


15. I make a limited use of Facebook because of privacy issues.


1 2 3 4 5


16. I make a limited use of Facebook because I don't like it.


1 2 3 4 5


17. I make a limited use of Facebook because I don't know how to use it.


1 2 3 4 5


Ir para a pergunta 26.



18. Have you ever tried to use Facebook?


Never

I've tried once

2 or 3 times

More than 3 times

Classify the next sentences in terms of accordance regarding the possibility of using Facebook.

19. I would like to use Facebook to get information about what's going on in the world.


1 2 3 4 5


20. I would like to use Facebook to know what's going with my relatives and close friends.


1 2 3 4 5


21. I would like to use Facebook to know what's going with all my friends.


1 2 3 4 5


22. I would like to use Facebook to make new friendships.


1 2 3 4 5


23. I do not use Facebook because of privacy issues.


1 2 3 4 5




24. I do not use Facebook because I don't like it.


1 2 3 4 5


25. I do not use Facebook because I don't know how to use it.


1 2 3 4 5


Television

26. I use Television to watch my favourite channels and shows.


1 2 3 4 5


27. In my opinion the Television very easy to use.


1 2 3 4 5


28. In my opinion the is very easy to interact with the Remote Control.


1 2 3 4 5


29. I watch Television with my family members.


1 2 3 4 5




30. I use the Television because my family members also use them.


1 2 3 4 5


31. I use Television to get information about what's going on in the world.


1 2 3 4 5


32. I'm familiar with the concept of smart TVs.


1 2 3 4 5


33. I use Television to acess the Internet.


1 2 3 4 5


34. I would like to use Television to acess to the Internet.


1 2 3 4 5


35. I would like to use Television to acess to Facebook.


1 2 3 4 5


36. I would like to use Television to see photos and information about my family members andclose friends.


1 2 3 4 5


Appendix E

You, Me &TV: Focus Groups

Pre-Questionnaire

217

!"#$$#!%$&'($)*$+,-./0(12-/3(42567/5089-::;925

4;<5($(-=($>8830?##@-.0A<--<B5A.-C#=-2C0#@#$D1E>%08FG"H$;I))2JKLM>LNN$OP14QKRRSTJ8=8U8V#329:8=-2C

Powered by

Focus Group Pre-Questionnaire*Obrigat�rio

1. Age:

Marque todas que se aplicam.

Less than 60 years of age

Between 60 and 64 years of age




More than 80 years of age

2. Easiness regarding the use of the following technology: *

Marque todas que se aplicam.

VeryDifficult

DifficultNeither difficult

or easyEasy

VeryEasy

Don't Know /Do not use

Television

Computer

Tablet

Facebook

Appendix F

Evaluation of You, Me & TV: SUS and

UEQ Questionnaires (Portuguese

version)

219

!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,(!(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

Question�rio de Usabilidade e Satisfa��o de

utiliza��oDiga se concorda ou n�o com as seguintes afirma��es:

1. 1. Gostaria de usar o sistema frequentemente.


1 2 3 4 5

Discordo totalmente Concordo totalmente

2. 2. O sistema � demasiado complexo.


1 2 3 4 5


3. 3. O sistema � f�cil de usar.


1 2 3 4 5


4. 4. Necessito de ajuda de um t�cnico para usar o sistema.


1 2 3 4 5


5. 5. As v�rias fun��es do sistema estavam bem estruturadas.


1 2 3 4 5


!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,($(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

6. 6. Muitas funcionalidades do sistema n�o se percebiam bem para que serviam.


1 2 3 4 5


7. 7. As pessoas ir�o aprender a usar o sistema rapidamente.


1 2 3 4 5


8. 8. O sistema � muito complicado de usar.


1 2 3 4 5


9. 9. Senti-me confiante e tranquilo a usar o sistema.


1 2 3 4 5


10. 10. Tive de aprender demasiadas coisas antes de come�ar a usar o sistema.


1 2 3 4 5


O que sentiu ao usar o sistema

O sistema �...

11. 11. *


1 2 3 4 5 6 7

Desagrad�vel Agrad�vel

!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,(Y(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

12. 12. *


1 2 3 4 5 6 7

Dif�cil deperceber

F�cil deperceber

13. 13. *


1 2 3 4 5 6 7

Criativo Sem criatividade

14. 14. *


1 2 3 4 5 6 7

F�cil deaprender

Dif�cil deaprender

15. 15. *


1 2 3 4 5 6 7

Valioso Sem valor

16. 16. *


1 2 3 4 5 6 7

Aborrecido Excitante

17. 17. *


1 2 3 4 5 6 7

Desinteressante Interessante

!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,(Q(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

18. 18. *


1 2 3 4 5 6 7

Imprevis�vel Previs�vel

19. 19.*


1 2 3 4 5 6 7

R�pido Lento

20. 20. *


1 2 3 4 5 6 7

Original Tradicional

21. 21. *


1 2 3 4 5 6 7

Restringe as decis�esque pode tomar

Permitefazer oquepretendesemrestri��es

22. 22. *


1 2 3 4 5 6 7

Bom Mau

23. 23. *


1 2 3 4 5 6 7

Complicado F�cil

!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,(I(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

24. 24. *


1 2 3 4 5 6 7

Aborrecido Atrativo

25. 25. *


1 2 3 4 5 6 7

Tradicional Inovador

26. 26. *


1 2 3 4 5 6 7

Inc�modo C�modo

27. 27. *


1 2 3 4 5 6 7

Seguro Inseguro

28. 28. *


1 2 3 4 5 6 7

Motivante Desmotivante

29. 29. *


1 2 3 4 5 6 7

Atende �sexpectativas

N�o atende �sexpectativas

!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,(Z(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

30. 30. *


1 2 3 4 5 6 7

Ineficiente Eficiente

31. 31. *


1 2 3 4 5 6 7

Claro Confuso

32. 32. *


1 2 3 4 5 6 7

Impratic�vel Pr�tico

33. 33. *


1 2 3 4 5 6 7

Organizado Desorganizado

34. 34. *


1 2 3 4 5 6 7

Atraente Feio

35. 35. *


1 2 3 4 5 6 7

Simp�tico Antip�tico

!"#!$#$%!&'(!!)!$*+,-./0123/0(4,(5-67/8/464,(,(96./-:6;<0(4,(+./8/=6;<0

>6?,(&(0:(@A..B-C##40D-E?00?8,ED0F#:03F-#4#!GH4IJKH4L&MM7NOPQRS-N9,TU",K=U>V"K0BW6-*AX?#B3/1.:03F

36. 36. *


1 2 3 4 5 6 7

Conservador Inovador

Opini�es

37. 37. Que aspectos melhorava?

38. 38. O que mais gostou?

39. 39. O que menos gostou?

40. 40. Sentiu que o sistema o/a aproximou da sua fam�lia e/ou amigos?

Bibliography

Abreu, J. F., Almeida, P., and Silva, T. (2013). ineighbour tv: A social tv application to promote.

Information Systems and Technologies for Enhancing Health and Social Care, 221:49–58.

Ackroyd, S. (1992). Data collection in context. Longman Group United Kingdom.

Adams, A. and Cox, A. L. (2008). Questionnaires, in-depth interviews and focus groups. Cambridge

University Press.

Age, U. (2010). Technology and older people evidence review.

Alaoui, M. and Lewkowicz, M. (2013). A livinglab approach to involve elderly in the design

of smart tv applications offering communication services. volume 8029 of Lecture Notes in

Computer Science, pages 325–334. Springer Berlin Heidelberg.

Albaina, I. M., Visser, T., van der Mast, C. A., and Vastenburg, M. H. (2009). Flowie: A persuasive

virtual coach to motivate elderly individuals to walk. In 2009 3rd International Conference on

Pervasive Computing Technologies for Healthcare, pages 1–7. IEEE.

Angst, F., Drerup, S., Werle, S., Herren, D. B., Simmen, B. R., and Goldhahn, J. (2010). Prediction

of grip and key pinch strength in 978 healthy subjects. BMC musculoskeletal disorders, 11(1):94.

Annett, M., Anderson, F., Goertzen, D., Halton, J., Ranson, Q., Bischof, W. F., and Boulanger, P.

(2009). Using a multi-touch tabletop for upper extremity motor rehabilitation. In Proceedings

of the 21st Annual Conference of the Australian Computer-Human Interaction Special Interest

Group: Design: Open 24/7, pages 261–264. ACM.

Annyang (2015). annyang! speechrecognition that just works.

Aoki, R., Maeda, A., Watanabe, T., Kobayashi, M., and Abe, M. (2010). Twist&tap: text entry for

tv remotes using easy-to-learn wrist motion and key operation. IEEE Transactions on Consumer

Electronics, 56(1):161–168.

Arning, K. and Ziefle, M. (2008). Development and validation of a computer expertise questionnaire

for older adults. Behaviour & information technology, 27(4):325–329.

Australian Research Council, T. (2010). Excelence in research for australia@ONLINE.

Aylett, M. P., Kristensson, P. O., Whittaker, S., and Vazquez-Alvarez, Y. (2014). None of a chind:

relationship counselling for hci and speech technology. In CHI’14 Extended Abstracts on Human

Factors in Computing Systems, pages 749–760. ACM.

227

Bibliography 228

Baecker, R., Sellen, K., Crosskey, S., Boscart, V., and Barbosa Neves, B. (2014). Technology to

reduce social isolation and loneliness. In Proceedings of the 16th International ACM SIGACCESS

Conference on Computers & Accessibility, ASSETS ’14, pages 27–34, New York, NY, USA.

ACM.

Bagozzi, R. P. (2007). The legacy of the technology acceptance model and a proposal for a paradigm

shift. Journal of the association for information systems, 8(4):3.

Bailly, G., MuLler, J., and Lecolinet, E. (2012). Design and evaluation of finger-count interaction:

Combining multitouch gestures and menus. Int. J. Hum.-Comput. Stud., 70(10):673–689.

Ball, K., Edwards, J. D., Ross, L. A., and McGwin Jr, G. (2010). Cognitive training decreases

motor vehicle collision involvement of older drivers. Journal of the American Geriatrics Society,

58(11):2107–2113.

Barrero, A., Melendi, D., Paneda, X. G., Garcıa, R., and Cabrero, S. (2014). An empirical

investigation into text input methods for interactive digital television applications. International

Journal of Human-Computer Interaction, 30(4):321–341.

Barrero, A., Melendi, D., Paneda, X. G., Garcıa, R., and Pozueco, L. (2016). Evaluation of text entry

methods for interactive digital television applications with devices alternative to conventional

remote controls. International Journal of Human-Computer Interaction, (just-accepted).

Barrett, L. L. (2008). Healthy@ home.

Becker, S. A. (2004). A study of web usability for older adults seeking online health resources.

ACM Transactions on Computer-Human Interaction (TOCHI), 11(4):387–406.

Bernhaupt, R., Obrist, M., Weiss, A., Beck, E., and Tscheligi, M. (2008). Trends in the living room

and beyond: results from ethnographic studies using creative and playful probing. Computers in

Entertainment (CIE), 6(1):5.

Bernhaupt, R. and Pirker, M. M. (2014). User interface guidelines for the control of interactive

television systems via smart phone applications. Behaviour & information technology, 33(8):784–

799.

Beyer, H. and Holtzblatt, K. (1997). Contextual design: A customer-centered approach to systems

designs (morgan kaufmann series in interactive technologies).

Bigham, J. P., Prince, C. M., and Ladner, R. E. (2008). Webanywhere: a screen reader on-the-go. In

Proceedings of the 2008 international cross-disciplinary conference on Web accessibility (W4A),

pages 73–82. ACM.

Biswas, P., Langdon, P., Duarte, C., and Coelho, J. (2011). Multimodal adaptation through

simulation for digital tv interface. In Proceddings of the 9th international interactive conference

on Interactive television, pages 231–234. ACM.

Bobeth, J., Schmehl, S., Kruijff, E., Deutsch, S., and Tscheligi, M. (2012). Evaluating performance

and acceptance of older adults using freehand gestures for tv menu control. EuroiTV ’12, pages

35–44, New York, NY, USA. ACM.

Bonner, M. N., Brudvik, J. T., Abowd, G. D., and Edwards, W. K. (2010). No-look notes: accessible

eyes-free multi-touch text entry. In International Conference on Pervasive Computing, pages

409–426. Springer.

Bibliography 229

Bothorel, C., Lohr, C., Thepaut, A., Bonnaud, F., and Cabasse, G. (2011). From individual commu-

nication to social networks: evolution of a technical platform for the elderly. In International

Conference on Smart Homes and Health Telematics, pages 145–152. Springer.

Boyd, E. (2007). Social Network Sites: Definition, History, and Scholarship. Journal of Computer-

Mediated Communication, 13(1-2).

Brewster, S. A. (1998). Sonically-enhanced drag and drop.

Brewster, S. A. (2002). Non-speech auditory output. Human Computer Interaction Handbook,

Lawrence Erlbaum Associates, Mahwah, NJ, pages 220–239.

Brewster, S. A. and Crease, M. G. (1999). Correcting menu usability problems with sound.

Behaviour & Information Technology, 18(3):165–177.

Brooke, J. (1996). Sus-a quick and dirty usability scale. Usability evaluation in industry, 189(194):4–

7.

Brunette, K., Eisenstadt, M., Pukinskis, E., and Ryan, W. (2005). Meeteetse: social well-being

through place attachment. In CHI’05 Extended Abstracts on Human Factors in Computing

Systems, pages 2065–2069. ACM.

Buiza, C., Soldatos, J., Petsatodis, T., Geven, A., Etxaniz, A., and Tscheligi, M. (2009). Hermes:

Pervasive computing and cognitive training for ageing well. In International Work-Conference

on Artificial Neural Networks, pages 756–763. Springer.

Burke, M., Kraut, R., and Marlow, C. (2011). Social capital on facebook: differentiating uses and

users. In Proc. CHI 2011, pages 571–580. ACM.

Burmeister, O. K. (2012). What seniors value about online community. The Journal of Community

Informatics, 8(1).

Byun, S. and Park, C. (2011). Serious game for cognitive testing of elderly. In International

Conference on Human-Computer Interaction, pages 354–357. Springer.

Carmeli, E., Patish, H., and Coleman, R. (2003). The aging hand. The Journals of Gerontology

Series A: Biological Sciences and Medical Sciences, 58(2):M146–M152.

Carmichael, A., Rice, M., Sloan, D., and Gregor, P. (2006). Digital switchover or digital divide: a

prognosis for usable and accessible interactive digital television in the uk. Universal Access in

the Information Society, 4(4):400–416.

Cavanaugh, J. C. and Blanchard-Fields, F. (1990). Adult Development. USA: Wadsworth Publishing

Company.

Cavanaugh, J. C. and Blanchard-Fields, F. (2014). Adult development and aging. Nelson Education.

Ceccaroni, L., Martınez, P., Hernandez, J. Z., and Verdaguer, X. (2005). Integratv-4all: an

interactive television for all. In Proceedings of the 1st Symposium on Ubiquitous Computing and

Ambient Intelligence (Granada, Spain.

Cesar, P. and Chorianopoulos, K. (2009). The evolution of tv systems, content, and users toward

interactivity. Foundations and Trends in Human-Computer Interaction, 2(4):373–95.

Bibliography 230

Chao, F.-L., Feng, C.-S., Fanjiang, B., and Sun, C.-L. (2017). Design jigsaw puzzle and app

for nostalgia-based support on elderly with dementia. In Awareness Science and Technology

(iCAST), 2017 IEEE 8th International Conference on, pages 284–289. IEEE.

Chen, K. and Chan, A. (2011). A review of technology acceptance by older adults. Gerontechnology,

10(1):1–12.

Chen, K. and Chan, A. H. S. (2014). Gerontechnology acceptance by elderly hong kong chinese: a

senior technology acceptance model (stam). Ergonomics, 57(5):635–652.

Chen, Y. (2009). Usability analysis on online social networks for the elderly. Technical report,

Helsinki University of Technology.

Cheng, S.-T., Chow, P. K., Song, Y.-Q., Edwin, C., Chan, A. C., Lee, T. M., and Lam, J. H. (2014).

Mental and physical activities delay cognitive decline in older persons with dementia. The

American Journal of Geriatric Psychiatry, 22(1):63–74.

Choi, S., Han, J., Lee, G., Lee, N., and Lee, W. (2011). Remotetouch: Touch-screen-like interaction

in the tv viewing environment. In Proceedings of the SIGCHI Conference on Human Factors in

Computing Systems, CHI ’11, pages 393–402, New York, NY, USA. ACM.

Cholerton, B., Baker, L. D., Montine, T. J., and Craft, S. (2016). Type 2 diabetes, cognition, and

dementia in older adults: Toward a precision health approach. Diabetes Spectrum, 29(4):210–

219.

Chorianopoulos, K. (2008). User interface design principles for interactive television applications.

Intl. Journal of Human–Computer Interaction, 24(6):556–573.

Clark, J. (1924). The ishihara test for color blindness. American Journal of Physiological Optics.

Coelho, J. and Duarte, C. (2011a). Building supportive multimodal user interfaces. Supportive

User Interfaces: SUI 2011, page 8.

Coelho, J. and Duarte, C. (2011b). The contribution of multimodal adaptation techniques to

the guide interface. In International Conference on Universal Access in Human-Computer

Interaction, pages 337–346. Springer.

Coelho, J. and Duarte, C. (2015). Socially networked or isolated? differentiating older adults and

the role of tablets and television. In Human-Computer Interaction, pages 129–146. Springer.

Coelho, J. and Duarte, C. (2016). A literature survey on older adults’ use of social network services

and social applications. Computers in Human Behavior, 58:187–205.

Coelho, J., Duarte, C., Biswas, P., and Langdon, P. (2011). Developing accessible tv applications.

In The proceedings of the 13th international ACM SIGACCESS conference on Computers and

accessibility, pages 131–138. ACM.

Coelho, J., Duarte, C., Feiteira, P., Costa, D., and Costa, D. (2012a). Building bridges between

elderly and tv application developers. In Proceedings of the 5th International Conference on

Advances in Computer-Human Interactions (ACHI 2012), Valencia, Spain.

Coelho, J., Duarte, C., Guerreiro, T., Feiteira, P., Costa, D., Costa, D., Neves, B., Alves, F., Biswas,

P., and Langdon, P. (2012b). Involving all stakeholders in the development of tv applications

for elderly. International Journal on Advances in Intelligent Systems Volume 5, Number 3 & 4,

2012.

Bibliography 231

Coelho, J., Guerreiro, T., and Duarte, C. (2013a). Designing tv interaction for the elderly–a case

study of the design for all approach. In A Multimodal End-2-End Approach to Accessible

Computing, pages 49–69. Springer.

Coelho, J., Guerreiro, T., Duarte, C., Biswas, P., Aslan, G., and Langdon, P. (2013b). Tv applications

for the elderly: Assessing the acceptance of adaptation and multimodality. In Proceedings of

the 6th International Conference on Advances in Computer-Human Interactions, Nice, France,

volume 24. Citeseer.

Coelho, J., Rito, F., and Duarte, C. (2017). “you, me & tv”—fighting social isolation of older

adults with facebook, tv and multimodality. International Journal of Human-Computer Studies,

98:38–50.

Coelho, J., Rito, F., Luz, N., and Duarte, C. (2015). Prototyping tv and tablet facebook interfaces

for older adults. In Human-Computer Interaction, pages 110–128. Springer.

Coleman, G. W., Gibson, L., Hanson, V. L., Bobrowicz, A., and McKay, A. (2010). Engaging the

disengaged: How do we design technology for digitally excluded older adults? In Proceedings

of the 8th ACM Conference on Designing Interactive Systems, pages 175–178. ACM.

Commission(ECFIN), E. (2008). The 2009 ageing report: Underlying assumptions and projection

methodologies for the EU-27 Member States (2007-2060). Office for Official Publications of the

European Communities.

Cooper, W. (2008). The interactive television user experience so far. In Proceedings of the 1st

International Conference on Designing Interactive User Experiences for TV and Video, UXTV

’08, pages 133–142, New York, NY, USA. ACM.

Cornejo, R., Tentori, M., and Favela, J. (2013). Enriching in-person encounters through social

media: A study on family connectedness for the elderly. Int. J. Hum.-Comput. Stud., 71(9):889–

899.

Cornwell, E. Y. and Waite, L. J. (2009). Measuring social isolation among older adults using

multiple indicators from the nshap study. Journals of Gerontology: Series B, 64B(suppl 1):i38–

i46.

Costa, D., Dias, R., and Duarte, C. (2014). A comparison of the accessibility of web applications in

tv and desktop. In BCS HCI.

Cota, M. P., Thomaschewski, J., Schrepp, M., and Goncalves, R. (2014). Efficient measurement

of the user experience. a portuguese version. Procedia Computer Science, 27(0):491 – 498.

DSAI2013.

Craik, F. I. and Jennings, J. M. (1992). Human memory.

Czaja, S. J., Guerrier, J. H., Nair, S. N., and Landauer, T. K. (1993). Computer communication as

an aid to independence for older adults. Behaviour and Information Technology, 12(4):197–207.

Dalton, D. S., Cruickshanks, K. J., Klein, B. E., Klein, R., Wiley, T. L., and Nondahl, D. M. (2003).

The impact of hearing loss on quality of life in older adults. The Gerontologist, 43(5):661–668.

Davis, A. (1991). Epidemiological profile of hearing impairments: the scale and nature of the

problem with special reference to the elderly. Acta Oto-Laryngologica, 111(sup476):23–31.

Bibliography 232

Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information

technology. MIS quarterly, pages 319–340.

de Bruin, E. D. (2012). Towards standardised evaluation tools. Age and ageing, 41(6):704–705.

de Morais, W. O. and Wickstrom, N. (2011). A serious computer game to assist tai chi training for

the elderly. In Serious Games and Applications for Health (SeGAH), 2011 IEEE 1st International

Conference on, pages 1–8. IEEE.

de Souza, C. M. and Silva, A. N. (2016). Aplicativos para smartphones e sua colaboracao na

capacidade funcional de idosos. Revista de Saude Digital e Tecnologias Educacionais-RESDITE,

1(1).

Derboven, J., Van Gils, M., and De Grooff, D. (2012). Designing for collaboration: a study in

intergenerational social game design. Universal Access in the Information Society, 11(1):57–65.

Douglas, A., Letts, L., and Liu, L. (2008). Review of cognitive assessments for older adults.

Physical & Occupational Therapy in Geriatrics, 26(4):13–43.

Duarte, C., Coelho, J., Feiteira, P., Costa, D., and Costa, D. (2011a). Eliciting interaction require-

ments for adaptive multimodal tv based applications. In International Conference on Universal

Access in Human-Computer Interaction, pages 42–50. Springer.

Duarte, C., Langdon, P., Jung, C., Coelho, J., Biswas, P., and Hamisu, P. (2011b). Guide: creating

accessible tv applications. Everyday Technology for Independence and Care, AAATE, pages

905–912.

ECFIN), E. C. D. (2008). The 2009 ageing report: Underlying assumptions and projection

methodologies for the EU-27 Member States (2007-2060). Office for Official Publications of the

European Communities.

Elder, H. A. (1970). On the feasibility of voice input to an on-line computer processing system.

Communications of the ACM, 13(6):339–346.

Enns, N. and MacKenzie, I. S. (1998). Touchpad-based remote control devices. ACM.

Eronen, L. et al. (2004). User centered design of new and novel products: case digital television.

Helsinki University of Technology.

Eurostat European Commission (2012). Population structure and ageing.

Far, I. K., Silveira, P., Casati, F., and Baez, M. (2012). Unifying platform for the physical, mental

and social well-being of the elderly. In Embedded and Multimedia Computing Technology and

Service, pages 385–392. Springer.

Farage, M. A., Miller, K. W., Ajayi, F., and Hutchins, D. (2012). Design principles to accommodate

older adults. Global journal of health science, 4(2):2.

Fisk, A. D., Rogers, W. A., Charness, N., Czaja, S. J., and Sharit, J. (2009). Designing for older

adults: Principles and creative human factors approaches. CRC press.

Flick, U. (2009). An introduction to qualitative research. Sage.

Fozard, J. L. and Gordon-Salant, S. (2001). Sensory and perceptual changes with aging. Handbook

of the psychology of aging, 5.

Bibliography 233

Furnas, G. W., Landauer, T. K., Gomez, L. M., and Dumais, S. T. (1987). The vocabulary problem

in human-system communication. Communications of the ACM, 30(11):964–971.

Gajos, K. Z., Wobbrock, J. O., and Weld, D. S. (2007). Automatically generating user interfaces

adapted to users’ motor and vision capabilities. In Proceedings of the 20th annual ACM

symposium on User interface software and technology, pages 231–240. ACM.

Garattini, C., Wherton, J., and Prendergast, D. (2012). Linking the lonely: An exploration of a

communication technology designed to support social interaction among older adults. Univers.

Access Inf. Soc., 11(2):211–222.

Garcıa-Vazquez, J. P., Rodrıguez, M. D., Andrade, A. G., and Bravo, J. (2011). Supporting the

strategies to improve elders’ medication compliance by providing ambient aids. Personal and

Ubiquitous Computing, 15(4):389–397.

Gaver, W., Boucher, A., Bowers, J., Blythe, M., Jarvis, N., Cameron, D., Kerridge, T., Wilkie, A.,

Phillips, R., and Wright, P. (2011). The photostroller: supporting diverse care home residents

in engaging with the world. In Proceedings of the SIGCHI Conference on Human Factors in

Computing Systems, CHI ’11, pages 1757–1766, New York, NY, USA. ACM.

Gawlinski, M. (2003). Interactive television production. Taylor & Francis.

Geerts, D., Cesar, P., and Bulterman, D. (2008). The implications of program genres for the design

of social television systems. In Proceedings of the 1st international conference on Designing

interactive user experiences for TV and video, pages 71–80. ACM.

Gerling, K. M., Schulte, F. P., and Masuch, M. (2011). Designing and evaluating digital games

for frail elderly persons. In Proceedings of the 8th international conference on advances in

computer entertainment technology, page 62. ACM.

Gibson, L., Moncur, W., Forbes, P., Arnott, J., Martin, C., and Bhachu, A. S. (2010). Designing

social networking sites for older adults. In Proc. BCS 2010, pages 186–194. British Computer

Society.

Gilbert, E. and Karahalios, K. (2009). Predicting tie strength with social media. In Proceedings

of the SIGCHI Conference on Human Factors in Computing Systems, CHI ’09, pages 211–220,

New York, NY, USA. ACM.

Giorgi, S., Talamo, A., and Mellini, B. (2011). The ”life frame”: responding to the elderly people’s

need of remembering. In CHI ’11 Extended Abstracts on Human Factors in Computing Systems,

CHI EA ’11, pages 1381–1386, New York, NY, USA. ACM.

GmbH, S. (2007). Cogniplus. technical report. Technical report.

Godard, N., Pecci, I., and Isokoski, P. (2013). Weslide: Gestural text entry for elderly users of

interactive television. In Proceedings of the 11th European Conference on Interactive TV and

Video, EuroITV ’13, pages 55–58, New York, NY, USA. ACM.

Goldman, N. and Narayanaswamy, K. (1992). Software evolution through iterative prototyping.

In Proceedings of the 14th international conference on Software engineering, pages 158–172.

ACM.

Gomes, G., Duarte, C., Coelho, J., and Matos, E. (2014). Designing a facebook interface for senior

users. The Scientific World Journal, 2014.

Bibliography 234

Grosinger, J., Vetere, F., and Fitzpatrick, G. (2012). Agile life: Addressing knowledge and social

motivations for active aging. In Proceedings of the 24th Australian Computer-Human Interaction

Conference, OzCHI ’12, pages 162–165, New York, NY, USA. ACM.

Harley, D., Howland, K., Harris, E., and Redlich, C. (2014). Online communities for older users:

What can we learn from local community interactions to create social sites that work for older

people. In Proceedings of the 28th International BCS Human Computer Interaction Conference

on HCI 2014 - Sand, Sea and Sky - Holiday HCI, BCS-HCI ’14, pages 42–51, UK. BCS.

Hauptmann, A. G. and McAvinney, P. (1993). Gestures with speech for graphic manipulation.

International Journal of Man-Machine Studies, 38(2):231–249.

Hawthorn, D. (2000). Possible implications of aging for interface designers. Interacting with

computers, 12(5):507–528.

Hayes, M., van Stolk-Cooke, K., and Muench, F. (2015). Understanding facebook use and the

psychological affects of use across generations. Computers in Human Behavior, 49(0):507 –

511.

Heart, T. and Kalderon, E. (2013). Older adults: are they ready to adopt health-related ict?

International journal of medical informatics, 82(11):e209–e231.

Hope, A., Schwaba, T., and Piper, A. M. (2014). Understanding digital and material social

communications for older adults. In Proc. CHI 2014, pages 3903–3912. ACM.

Iatrino, A. and Modeo, S. (2006). Text editing in digital terrestrial television: a comparison of three

interfaces. In Proceedings of EuroITV, volume 6, pages 224–241.

Ingmarsson, M., Dinka, D., and Zhai, S. (2004). Tnt: a numeric keypad based text input method. In

Proceedings of the SIGCHI conference on Human factors in computing systems, pages 639–646.

ACM.

Izadi, S., Brignull, H., Rodden, T., Rogers, Y., and Underwood, M. (2003). Dynamo: A public

interactive surface supporting the cooperative sharing and exchange of media. In Proceedings of

the 16th Annual ACM Symposium on User Interface Software and Technology, UIST ’03, pages

159–168, New York, NY, USA. ACM.

Jackson, G. R. and Owsley, C. (2003). Visual dysfunction, neurodegenerative diseases, and aging.

Neurologic clinics.

Jin, Z. X., Plocher, T., and Kiff, L. (2007). Touch screen user interfaces for older adults: button size

and spacing. In Proceedings of the 4th international conference on Universal access in human

computer interaction: coping with diversity, UAHCI’07, pages 933–941, Berlin, Heidelberg.

Springer-Verlag.

Joinson, A. N. (2008). Looking at, looking up or keeping up with people?: motives and use of

facebook. In Proc. CHI 2008, pages 1027–1036. ACM.

Judge, T. K., Neustaedter, C., Harrison, S., and Blose, A. (2011). Family portals: Connecting

families through a multifamily media space. In CHI ’11, pages 1205–1214, New York, NY, USA.

ACM.

Bibliography 235

Kanungo, T., Mount, D. M., Netanyahu, N. S., Piatko, C. D., Silverman, R., and Wu, A. Y. (2002).

An efficient k-means clustering algorithm: Analysis and implementation. IEEE transactions on

pattern analysis and machine intelligence, 24(7):881–892.

Karahasanovic, A., Brandtzaeg, P. B., Heim, J., Luders, M., Vermeir, L., Pierson, J., Lievens, B.,

Vanattenhoven, J., and Jans, G. (2009). Co-creation and user-generated content-elderly people’s

user requirements. Comput. Hum. Behav., 25(3):655–678.

Keates, S. and Robinson, P. (1998). The use of gestures in multimodal input. In Proceedings of the

third international ACM conference on Assistive technologies, pages 35–42. ACM.

Ketcham, C. J., Seidler, R. D., Van Gemmert, A. W., and Stelmach, G. E. (2002). Age-related

kinematic differences as influenced by task difficulty, target size, and movement amplitude. The

Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 57(1):P54–P64.

Kim, H., Monk, A., Wood, G., Blythe, M., Wallace, J., and Olivier, P. (2013). Timelypresent:

Connecting families across continents. Int. J. Hum.-Comput. Stud., 71(10):1003–1011.

Kim, K. S., Oh, S.-S., Ahn, J.-H., and Lee, S.-H. (2012). Development of a walking game for the

elderly using controllers of hand buttons and foot boards. In Computer Games (CGAMES), 2012

17th International Conference on, pages 158–161. IEEE.

Kim, S. and Garrison, G. (2009). Investigating mobile wireless technology adoption: An extension

of the technology acceptance model. Information Systems Frontiers, 11(3):323–333.

Kobayashi, M., Hiyama, A., Miura, T., Asakawa, C., Hirose, M., and Ifukube, T. (2011). Elderly

user evaluation of mobile touchscreen interactions. In IFIP Conference on Human-Computer

Interaction, pages 83–99. Springer.

Komine, K., Miyazaki, M., Yeun-Bae, K., and Uratani, N. (2004). A spoken dialogue interface

for tv operations based on data collected by using woz method. IEICE TRANSACTIONS on

Information and Systems, 87(6):1397–1404.

Konig, W. A., Gerken, J., Dierdorf, S., and Reiterer, H. (2009). Adaptive pointing–design and

evaluation of a precision enhancing technique for absolute pointing devices. In IFIP Conference

on Human-Computer Interaction, pages 658–671. Springer.

Kunert, T. (2009). User-centered interaction design patterns for interactive digital television

applications. Springer Science & Business Media.

Kurniawan, S. H. (2008). Ageing. In Web Accessibility, pages 47–58. Springer.

LAM, J. C. and LEE, M. K. (2006). Digital inclusiveness–longitudinal study of internet adoption

by older adults. Journal of Management Information Systems, 22(4):177–206.

Lane, F. and Snowdon, J. (1989). Memory and dementia: a longitudinal survey of suburban elderly.

Clinical and abnormal psychology. Amsterdam: Elsevier, pages 365–76.

Lang, C. and Barton, H. (2015). Just untag it: Exploring the management of undesirable facebook

photos. Computers in Human Behavior, 43(0):147 – 155.

Langdon, P., Aurisicchio, M., Clarkson, P., and Wallace, K. (2003). An integrated ethnographic

and empirical methodology in a study of knowledge searches in aerospace design. In DS 31:

Proceedings of ICED 03, the 14th International Conference on Engineering Design, Stockholm.

Bibliography 236

Larrabee, G. J. and Crook, T. H. (1994). Estimated prevalence of age-associated memory impairment

derived from standardized tests of memory function. International psychogeriatrics.

Laursen, B., Jensen, B. R., and Ratkevicius, A. (2001). Performance and muscle activity during

computer mouse tasks in young and elderly adults. European journal of applied physiology,

84(4):329–336.

Lawrence, E., Sax, C., Navarro, K. F., and Qiao, M. (2010). Interactive games to improve quality

of life for the elderly: Towards integration into a wsn monitoring system. In Proceedings of

the 2010 Second International Conference on eHealth, Telemedicine, and Social Medicine,

ETELEMED ’10, pages 106–112, Washington, DC, USA. IEEE Computer Society.

Lehtinen, V., Nasanen, J., and Sarvas, R. (2009). ”a little silly and empty-headed”: Older adults’

understandings of social networking sites. In Proc. BCS 2009, pages 45–54. British Computer

Society.

Lessiter, J., Freeman, J., Davis, R., and Dumbreck, A. (2004). Understanding dtt remote control

button labelling: a multi-method approach. In Proceedings, Second European conference on

Interactive Television, Brighton.

Lindley, S. (2012). Shades of lightweight: supporting cross-generational communication through

home messaging. Universal Access in the Information Society, 11(1):31–43.

Lindley, S. E., Harper, R., and Sellen, A. (2008). Designing for elders: exploring the complexity of

relationships in later life. In Proceedings of the 22nd British HCI Group Annual Conference

on People and Computers: Culture, Creativity, Interaction-Volume 1, pages 77–86. British

Computer Society.

Lindley, S. E., Harper, R., and Sellen, A. (2009). Desiring to be in touch in a changing communica-

tions landscape: attitudes of older adults. In Proc. CHI 2009, pages 1693–1702. ACM.

Louis, E. D. and Ferreira, J. J. (2010). How common is the most common adult movement disorder?

update on the worldwide prevalence of essential tremor. Mov Disord, 25(5):534–41.

Lull, J. (2014). Inside Family Viewing (Routledge Revivals): Ethnographic Research on Television’s

Audiences. Routledge.

Lutz, W., Sanderson, W., and Scherbov, S. (2008). The coming acceleration of global population

ageing. Nature, 451(7179):716–719.

MacKenzie, I. S. and Soukoreff, R. W. (2002). Text entry for mobile computing: Models and

methods, theory and practice. Human–Computer Interaction, 17(2-3):147–198.

Mankoff, J., Fait, H., and Juang, R. (2005). Evaluating accessibility by simulating the experiences

of users with vision or motor impairments. IBM Systems Journal, 44(3):505–517.

Mazzeo, R. S., Cavanagh, P., Evans, W. J., Fiatarone, M., Hagberg, J., McAuley, E., and Startzell,

J. (1998). Exercise and physical activity for older adults. Medicine and science in sports and

exercise, 30(6):992–1008.

McCall, C., Maynes, B., Zou, C. C., and Zhang, N. J. (2013). An automatic medication self-

management and monitoring system for independently living patients. Medical engineering &

physics, 35(4):505–514.

Bibliography 237

McGee, D. R., Cohen, P. R., and Oviatt, S. (1998). Confirmation in multimodal systems. In

Proceedings of the 36th Annual Meeting of the Association for Computational Linguistics

and 17th International Conference on Computational Linguistics-Volume 2, pages 823–829.

Association for Computational Linguistics.

McLaughlin, A. C., Rogers, W. A., and Fisk, A. D. (2009). Using direct and indirect input

devices: Attention demands and age-related differences. ACM Trans. Comput.-Hum. Interact.,

16(1):2:1–2:15.

Meza-Kubo, V., Moran, A. L., and Rodriguez, M. D. (2009). Intergenerational communication

systems in support for elder adults with cognitive decline. In 2009 3rd International Conference

on Pervasive Computing Technologies for Healthcare, pages 1–4. IEEE.

Michailidou, E., Parmaxi, A., and Zaphiris, P. (2015). Culture effects in online social support

for older people: perceptions and experience. Universal Access in the Information Society,

14(2):281–293.

Miles, M. B. and Huberman, A. M. (1994). Qualitative Data Analysis: An Expanded Sourcebook.

SAGE Publications.

Mitzner, T. L., Boron, J. B., Fausset, C. B., Adams, A. E., Charness, N., Czaja, S. J., Dijkstra, K.,

Fisk, A. D., Rogers, W. A., and Sharit, J. (2010). Older adults talk technology: Technology

usage and attitudes. Computers in human behavior, 26(6):1710–1721.

Morrell, R. W. (2001). Older adults, health information, and the World Wide Web. Psychology

Press.

Morris, M. E. (2005). Social networks as health feedback displays. IEEE Internet Computing,

9(5):29–37.

Mubin, O., Shahid, S., and Al Mahmud, A. (2008). Walk 2 win: towards designing a mobile

game for elderly’s social engagement. In Proceedings of the 22nd British HCI Group Annual

Conference on People and Computers: Culture, Creativity, Interaction-Volume 2, pages 11–14.

British Computer Society.

Myers, B. A., Bhatnagar, R., Nichols, J., Peck, C. H., Kong, D., Miller, R., and Long, A. C. (2002).

Interacting at a distance: Measuring the performance of laser pointers and other devices. In

Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI ’02,

pages 33–40, New York, NY, USA. ACM.

Nacke, L. E. (2010). Wiimote vs. controller: electroencephalographic measurement of affective

gameplay interaction. In Proceedings of the International Academic Conference on the Future

of Game Design and Technology, pages 159–166. ACM.

Neufeldt, C. (2009). Wii play with elderly people. Enhancing Interaction Spaces by Social Media

for the Elderly, 6(3):50–59.

Neves, B. B., Franz, R. L., Munteanu, C., Baecker, R., and Ngo, M. (2015). ”my hand doesn’t

listen to me!”: Adoption and evaluation of a communication technology for the ’oldest old’. In

Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems,

CHI ’15, pages 1593–1602, New York, NY, USA. ACM.

Bibliography 238

Nichols, T. A., Rogers, W. A., and Fisk, A. D. (2006). Design for aging. Handbook of Human

Factors and Ergonomics, Third Edition, pages 1418–1445.

Nielsen, J. (1995). 10 usability heuristics for user interface design. Nielsen Norman Group, 1(1).

Norman, D. (2013). The design of everyday things: Revised and expanded edition. Constellation.

Norval, C., Arnott, J. L., and Hanson, V. L. (2014). What’s on your mind?: Investigating recommen-

dations for inclusive social networking and older adults. In Proc. CHI 2014, pages 3923–3932.

ACM.

Oakley, I., Brewster, S., and Gray, P. (2001). Solving multi-target haptic problems in menu

interaction. In CHI’01 extended abstracts on Human factors in computing systems, pages

357–358. ACM.

Obrist, M., Bernhaupt, R., and Tscheligi, M. (2008). Interactive tv for the home: An ethnographic

study on users’ requirements and experiences. Intl. Journal of Human–Computer Interaction,

24(2):174–196.

Ostlund, B. (2010). Watching television in later life: a deeper understanding of tv viewing in the

homes of old people and in geriatric care contexts. Scandinavian journal of caring sciences,

24(2):233–243.

Oviatt, S. (1997). Multimodal interactive maps: Designing for human performance. Human-

computer interaction, 12(1):93–129.

Oviatt, S. (1999). Mutual disambiguation of recognition errors in a multimodel architecture. In

Proceedings of the SIGCHI conference on Human Factors in Computing Systems, pages 576–583.

ACM.

Oviatt, S. (2003). Multimodal interfaces. The human-computer interaction handbook: Fundamen-

tals, evolving technologies and emerging applications, 14:286–304.

Oviatt, S., DeAngeli, A., and Kuhn, K. (1997). Integration and synchronization of input modes dur-

ing multimodal human-computer interaction. In Referring Phenomena in a Multimedia Context

and their Computational Treatment, pages 1–13. Association for Computational Linguistics.

Pak, R. and McLaughlin, A. (2010). Designing displays for older adults. CRC PressI Llc.

Parra, M. A. (2013). Cognitive assessment in alzheimer’s disease. Advances in Alzheimer’s disease,

2(04):123.

Pew Research Center (2015a). The Demographics of Social Media Users.

Pew Research Center (2015b). Social Media Usage: 2005-2015.

Philips (2005). Philips motiva - remote patient management platform.

Phiriyapokanon, T. (2011). Is a big button interface enough for elderly users. Towards user interface

guidelines for elderly users. Sweden, Malardalen University. Thesis (Master of Computer

Engineer).

Plaza, I., Martin, L., Martin, S., and Medrano, C. (2011). Mobile applications in an aging society:

Status and trends. Journal of Systems and Software, 84(11):1977 – 1988.

Bibliography 239

Raffle, H., Ballagas, R., Revelle, G., Horii, H., Follmer, S., Go, J., Reardon, E., Mori, K., Kaye, J.,

and Spasojevic, M. (2010). Family story play: Reading with young children (and elmo) over a

distance. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems,

CHI ’10, pages 1583–1592, New York, NY, USA. ACM.

Ranganathan, V. K., Siemionow, V., Sahgal, V., and Yue, G. H. (2001). Effects of aging on hand

function. Journal of the American Geriatrics Society, 49(11):1478–1484.

Reeves, L. M., Lai, J., Larson, J. A., Oviatt, S., Balaji, T., Buisine, S., Collings, P., Cohen, P., Kraal,

B., Martin, J.-C., et al. (2004). Guidelines for multimodal user interface design. Communications

of the ACM, 47(1):57–59.

Reinecke, K. and Bernstein, A. (2011). Improving performance, perceived usability, and aesthetics

with culturally adaptive user interfaces. ACM Trans. Comput.-Hum. Interact., 18(2):8:1–8:29.

Rice, M., Newell, A., and Morgan, M. (2007). Forum theatre as a requirements gathering methodol-

ogy in the design of a home telecommunication system for older adults. Behaviour & Information

Technology, 26(4):323–331.

Roibas, A. C., Sala, R., Ahmad, S., and Radman, M. (2005). Beyond the remote control: Going

the extra mile to enhance itv access via mobile devices & humanizing navigation experience

for those with special needs. In Proceedings of the 3rd European Conference on Interactive

Television.

Romero, N., Markopoulos, P., Baren, J., Ruyter, B., Ijsselsteijn, W., and Farshchian, B. (2007).

Connecting the family with awareness systems. Personal Ubiquitous Comput., 11(4):299–312.

Routledge, C., Wildschut, T., Sedikides, C., and Juhl, J. (2013). Nostalgia as a resource for

psychological health and well-being. Social and Personality Psychology Compass, 7(11):808–

818.

Santana, P. C., Rodrıguez, M. D., Gonzalez, V. M., Castro, L. A., and Andrade, A. G. (2005).

Supporting emotional ties among mexican elders and their families living abroad. In CHI ’05

Extended Abstracts on Human Factors in Computing Systems, CHI EA ’05, pages 2099–2103,

New York, NY, USA. ACM.

Saxon, S. V., Etten, M. J., and Perkins, E. A. (2009). Physical change and aging. Springer

Publishing Company.

Sayago, S., Santos, P., Gonzalez, M., Arenas, M., and Lopez, L. (2007). Meeting educational needs

of the elderly in ict: two exploratory case studies. Crossroads, 14(2).

Scheibe, S. and Carstensen, L. L. (2010). Emotional aging: Recent findings and future trends. The

Journals of Gerontology Series B: Psychological Sciences and Social Sciences, page gbp132.

Schepers, J. and Wetzels, M. (2007). A meta-analysis of the technology acceptance model:

Investigating subjective norm and moderation effects. Information & Management, 44(1):90–

103.

Sengpiel, M. and Dittberner, D. (2008). The computer literacy scale (cls) for older adults-

development and validation. In Mensch & Computer, pages 7–16.

Shari, T. (2008). Physical impairment. In Harper, S. and Yesilada, Y., editors, Web accessibility: a

foundation for research. Springer.

Bibliography 240

Shneiderman, B. (2000). The limits of speech recognition. Communications of the ACM, 43(9):63–

65.

Shneiderman, B. (2005). Shneiderman’s eight golden rules of interface design. Retrieved July,

25:2009.

Sillanpaa, N., Alli, S., and Overmark, T. (2009). Easy-to-use social network service for people with

cognitive or speech and language impairments.

Silva, J. M., Mouttham, A., and El Saddik, A. (2009). Ubimeds: a mobile application to improve

accessibility and support medication adherence. In Proceedings of the 1st ACM SIGMM interna-

tional workshop on Media studies and implementations that help improving access to disabled

users, pages 71–78. ACM.

Silva, J. P. A. S. (2011). Design de interacao para TV Digital. PhD thesis, Universidade de Sao

Paulo.

Silveira, P., van de Langenberg, R., van het Reve, E., Daniel, F., Casati, F., and de Bruin, E. D.

(2013). Tablet-based strength-balance training to motivate and improve adherence to exercise in

independently living older people: a phase ii preclinical exploratory trial. Journal of medical

Internet research, 15(8):e159.

Smith, S. T., Talaei-Khoei, A., Ray, M., and Ray, P. (2009). Electronic games for aged care and

rehabilitation. In e-Health Networking, Applications and Services, 2009. Healthcom 2009. 11th

International Conference on, pages 42–47. IEEE.

Springett, M. V. and Griffiths, R. N. (2007). Accessibility of interactive television for users with

low vision: learning from the web. In European Conference on Interactive Television, pages

76–85. Springer.

Stanney, K., Samman, S., Reeves, L., Hale, K., Buff, W., Bowers, C., Goldiez, B., Nicholson,

D., and Lackey, S. (2004). A paradigm shift in interactive computing: Deriving multimodal

design principles from behavioral and neurological foundations. International Journal of Human-

Computer Interaction, 17(2):229–257.

Statista (2018). Most famous social network sites worldwide as of april 2018, ranked by number of

active users (in millions).

Steel, N., Huppert, F. A., McWilliams, B., and Melzer, D. (2004). 7. physical and cognitive

function.

Stephanidis, C., Paramythis, A., Sfyrakis, M., Stergiou, A., Maou, N., Leventis, A., Paparoulis,

G., and Karagiannidis, C. (1998). Adaptable and adaptive user interfaces for disabled users in

the avanti project. Intelligence in Services and Networks: Technology for Ubiquitous Telecom

Services, pages 153–166.

Stobel, C. and Blessing, L. (2010). Mobile device interaction gestures for older users. In Proc.

NordiCHI 2010, pages 793–796. ACM.

Strickland, D. and Bertoni, J. M. (2004). Parkinson’s prevalence estimated by a state registry.

Movement Disorders, 19(3):318–323.

Bibliography 241

Sundar, S. S., Oeldorf-Hirsch, A., Nussbaum, J., and Behr, R. (2011). Retirees on facebook: can

online social networking enhance their health and wellness? In CHI’11 extended abstracts on

human factors in computing systems, pages 2287–2292. ACM.

Taylor, S. J. and Bogdan, R. (1984). Introduction to qualitative research methods: The search for

meaning.

Tentori, M., Cornejo, R., and Favela, J. (2010). Sentient displays to connect elders with cognitive

disabilities to the digital era of social interaction. In International Workshop on Interactive

Systems in Healthcare, page 161.

Thomson Reuters, T. (2013). The thomson reuters impact factor@ONLINE.

Tiwari, N., Pandey, P. C., and Sharma, A. (2016). A smartphone app-based digital hearing

aid with sliding-band dynamic range compression. In Proc. 22th National Conference on

Communications, volume 1, page 6.

Trinh, V.-Q., Chung, G.-S., and Kim, H.-C. (2012). Improving the elder’s quality of life with

smart television based services. World of Academy of Science, Engineering and Technology,

67:572–575.

Tsekleves, E., Whitham, R., Kondo, K., and Hill, A. (2009). Bringing the television experience

to other media in the home: an ethnographic study. In Proceedings of the seventh european

conference on European interactive television conference, pages 201–210. ACM.

Turunen, M., Melto, A., Hella, J., Heimonen, T., Hakulinen, J., Makinen, E., Laivo, T., and Soronen,

H. (2009). User expectations and user experience with different modalities in a mobile phone

controlled home entertainment system. In Proceedings of the 11th International Conference on

Human-Computer Interaction with Mobile Devices and Services, page 31. ACM.

United Nations (2010). World Population Ageing: 2009. United Nations, New York.

Venkatesh, V. and Bala, H. (2008). Technology acceptance model 3 and a research agenda on

interventions. Decision sciences, 39(2):273–315.

Venkatesh, V. and Davis, F. D. (2000). A theoretical extension of the technology acceptance model:

Four longitudinal field studies. Management science, 46(2):186–204.

Venkatesh, V., Morris, M. G., Davis, G. B., and Davis, F. D. (2003). User acceptance of information

technology: Toward a unified view. MIS quarterly, pages 425–478.

Vetere, F., Davis, H., Gibbs, M., and Howard, S. (2009). The magic box and collage: Responding to

the challenge of distributed intergenerational play. Int. J. Hum.-Comput. Stud., 67(2):165–178.

Visu, A., Kalaiselvi, V., et al. (2011). Penpal-electronic pen aiding visually impaired in reading

and visualizing textual contents. In Technology for Education (T4E), 2011 IEEE International

Conference on, pages 171–176. IEEE.

Vroman, K. G., Arthanat, S., and Lysack, C. (2015). ?who over 65 is online?? older adults?

dispositions toward information communication technology. Computers in Human Behavior,

43(0):156 – 166.

Bibliography 242

Vutborg, R., Kjeldskov, J., Pedell, S., and Vetere, F. (2010). Family storytelling for grandparents and

grandchildren living apart. In Proceedings of the 6th Nordic Conference on Human-Computer

Interaction: Extending Boundaries, NordiCHI ’10, pages 531–540, New York, NY, USA. ACM.

Wallace, J., Thieme, A., Wood, G., Schofield, G., and Olivier, P. (2012). Enabling self, intimacy

and a sense of home in dementia: an enquiry into design in a hospital setting. In Proceedings of

the SIGCHI Conference on Human Factors in Computing Systems, pages 2629–2638. ACM.

Waycott, J., Vetere, F., Pedell, S., Kulik, L., Ozanne, E., Gruner, A., and Downs, J. (2013). Older

adults as digital content producers. In Proc. CHI 2013, pages 39–48. ACM.

Weber, R. P. (1990). Basic content analysis. Number 49. Sage.

Wickremaratchi, M. and Llewelyn, J. (2006). Effects of ageing on touch. Postgraduate medical

journal, 82(967):301–304.

Xie, B., Watkins, I., Golbeck, J., and Huang, M. (2012). Understanding and changing older adults

perceptions and learning of social media. Educational Gerontology, 38(4):282–296. PMID:

22639483.

Zapata, B. C., Fernandez-Aleman, J. L., Idri, A., and Toval, A. (2015). Empirical studies on usability

of mhealth apps: A systematic literature review. Journal of medical systems, 39(2):1–19.

Zaphiris, P., Kurniawan, S. H., and Ellis, R. D. (2002). Age related differences and the depth

vs. breadth tradeoff in hierarchical online information systems. In ERCIM Workshop on User

Interfaces for All, pages 23–42. Springer.

Zapirain, B. G., Zorrilla, A. M., and Larranaga, S. (2010). Psycho-stimulation for elderly people

using puzzle game. In 2010 2nd International IEEE Consumer Electronics Society’s Games

Innovations Conference, pages 1–8. IEEE.

Documents

Repositório da Universidade de Lisboa: Página principalrepositorio.ul.pt/bitstream/10451/36053/1/ulsd... · I will always be grateful to my dearest friends and LaSIGE colleagues