36
A Methodological Quest for A Methodological Quest for Studying Interactions in Studying Interactions in Advanced Video Advanced Video Conferencing Environments Conferencing Environments Dr Mary Allan & Prof David Thorns School of Sociology and Anthropology University of Canterbury NZ 4th International Conference on e-Social Science University of Manchester, UK 18 - 20 June 2008

Download Presentation

Embed Size (px)

Citation preview

A Methodological Quest for A Methodological Quest for Studying Interactions in Studying Interactions in

Advanced Video Advanced Video Conferencing EnvironmentsConferencing Environments

Dr Mary Allan & Prof David ThornsSchool of Sociology and Anthropology

University of Canterbury NZ

4th International Conference on e-Social Science University of Manchester, UK

18 - 20 June 2008

2Allan & Thorns June 08

The Powers at Work

Knowledge Economy = Collaboration

Sustainability = Reduce travel, Diminish Carbon Footprint

Design solution= Simulate FTF

Real Need

Users :Not real FTF

fuel and sustainability

0

20

40

60Soaring prices and carbon footprint

How to meet it?

3Allan & Thorns June 08

How is the Need Addressed?

• Design-Considerable resources have been invested in enhancing the design of videoconferencing tools

• Uptake- Numerous studies have attempted to discover the factors contributing to uptake of video conferencing and advanced systems such as Access Grid and telepresence tools

However…. • So far studies show that uptake is lower than

anticipated

4Allan & Thorns June 08

Pensive Note

Low uptake is challenging the hypothesis

“If you build a Better System ‘THEY’ will come”

5Allan & Thorns June 08

Prevalent Approaches to Diffusing Technology?

• Designers come up with a new product

• Serve people what designers have prepared

– Presume what users want

• Disseminate users’ survey asking questions which provide answers to questions raised by design teams

• Give users a pre prepared choice

– Offer a menu for users to choose from

6Allan & Thorns June 08

The Prevalent Approach

Build a better system so users will like it

But …..• What is a better system ? • Who decides what’s

better?• How do we know users

will like it?

7Allan & Thorns June 08

Traditional User Centred Approaches

• Traditional user-centred approaches are often found to fall short of:– Real engagement of designers and users– Attention to social and political aspects– Portrayal of complexity and representativeness – (Grudin & Pruitt, 2002)

• User centred design processes are time consuming (vredenburg et.al 2002)

Here we focus on users’ engagement

8Allan & Thorns June 08

Our proposed Approach

Let People Do The Choosing

9Allan & Thorns June 08

Aim of Study

Investigate the reasons underpinning uptake of Advanced video conferencing technologies through the discovery of the ways in which

usersusers, rather than designers construct the functionality of the technology

10Allan & Thorns June 08

Grounded User Centred Approach

• Identify what users think is important and construct as scheme of criteria to inform the design of videoconferencing

• We began our investigation by applying Glaser’s Grounded Theory Approach – a ‘free of preconception investigation’ (Glaser, 1992 )

11Allan & Thorns June 08

Implementing Our Approach

A case Study observing the population of users of an Advanced Video Conferencing tool – The Access Grid (AG) from within the research community in NZ universities. Participants included research students and staff

12Allan & Thorns June 08

The NZ Context

• The AG technology has been in use in NZ since 2005 as part of a national project the BRCSS1 which attempted to link social scientists and facilitate collaboration across the 8 universities in the country

• In 2007 the high speed education and research network KAREN2 was launched enabling enhanced use of advanced video conferencing tools

1.BRCSS- Building Research Capabilities in the Social Sciences

2.KAREN- kiwi Advanced Research and Education Network

13Allan & Thorns June 08

Background of Case Study

• The infrastructure and the technology were put in place, preparing the way for wide uptake,

and…• Triggered the need for researching the

uptake and the ways in which it may alter cross site meeting and collaboration practices

14Allan & Thorns June 08

Phase 1: Observations

• Observed 17 Access Grid sessions - cross site presentations, seminars, teaching and business meetings

• Manually annotated as many details we could capture in real time - limited bandwidth did not permit recording prior to the implementation of our high speed connection, KAREN

• Compilation of the notes from all observations revealed the following key themes:– Socio technical interactions- the way people interacted

with other people in and across nodes, and the way they related to the technology in their environment

– Group dynamics across different contexts– Group dynamics within and across nodes– Different modes of disseminating information

15Allan & Thorns June 08

Phase2- Users’ Survey

• The key themes identified in the observations provided the basis for formulating survey questions which were constructed in two levels:– Questions relating to actual AG

experience – Questions relating to perceived potential

and activities enabled by the technology

16Allan & Thorns June 08

Phase 2 cont’ - Survey Findings

•92% suggested that interactions on the AG are different from FTF •82.4% suggested that it is Possible to participate in AG the same way as in FTF9.6% gap between potential (possible) and actual experience suggesting that when asked about actual experience more participants view AG interactions as different to FTF, however, 82.4% see the potential

17Allan & Thorns June 08

What are FTF Identifiers:

• Synchronicity- the ability to engage in temporal turn taking interactions that enable the smooth alternation of speaker and listener who are co present (Bosch et al., 2004)

• Multimedia information incorporating Mehrabian’s ‘3V’s’ – verbal, vocal, and visual (Mehrabian, 1971)

• A sense of being with one another, a sense of presence (Heeter, 1992), and co-presence (Goffman, 1963; Schroeder, 2006)

• Intuitiveness-not requiring any complex or clunky means (Naumann, A., et.al 2007)

18Allan & Thorns June 08

FTF Identifiers and Survey

Socio-technical- relationship with the technology , familiarity

Intuitiveness

Multimedia information

Group dynamics – users constructing social spaces

Shared space and co-presence

Socio-technical interactions- users interacting with, and through the technology- turn taking, flow of conversation, facial cues, body language

Synchronicity

Observations/Survey Themes

FTF

19Allan & Thorns June 08

Phase 3- Juxtaposing

Total=89.7%

Clunkiness of technology, lack of flowIntuitiveness

Total=2733%

• AG does not allow participants to follow the gaze of their conversation partners

• Positioning of people facing screen rather than camera altered the interaction within and across nodes” facing the wall, not us

• Proximity to camera determined presence or absence• Sharing space- muting microphones was used to exclude

groups

A sense of being with one another, a sense of presence

Total=3138%

Presence of camera altered interactions View of self image, and the image of others was different

to that experienced in FTF Quality of sound sometimes inadequate Ability to discern body language and facial cue, not always

possible

Multimedia information

Total=1417%

Turn taking in AG requires different practices to those of FTF

Time lag affects turn taking

Synchronicity

CommentsAG aspects different to FTFFTF

20Allan & Thorns June 08

Phase 3cont’-Juxtaposing

Levels of difference in FTF identifiers

17

3833

9.7

05

10152025303540

AG disimilairty

Sync

hron

icity

mul

timed

iash

ared

spa

cein

tuiti

vene

ss

FTF

Rate of disimilarity FTF to AG

rate of dif ference in AG

21Allan & Thorns June 08

Phase 3 cont’-Juxtaposing

•In relation to the multimedia identifier participants commented: “although everyone can see each other and contribute you don’t get the same body language cues when someone wants to speak’.

22Allan & Thorns June 08

Phase 3 cont’-Juxtaposing

•In relation to shared space identifier participants commented: “Looking into camera to

speak meant that you were excluding those in the room with you (We

were all sitting facing the screen/camera)”

23Allan & Thorns June 08

Summary of Phase2 &3

• The findings of the survey showed users’ identification of the points of difference between perceptions/expectations , and experience of similarity of AG to FTF

• Juxtaposition of the differences identified against traditional FTF featured showed a clearer view of the obstacles of AG in becoming more like FTF

• The survey enabled measuring in detail specific dissimilarities which could be addressed, leading to the construction of criteria for the design of a systematic annotating tool

24Allan & Thorns June 08

Grounded users Centred Design -Overview

• Unstructured observations • Findings used to identify key areas• Key areas informed the design of a user survey constructed of two

levels:– Questions relating to actual AG experience – Questions relating to perceived potential and activities enabled by the

technology• Survey findings discovered a gap between experience and perceptions

and identified the features around which this gap evolved • Experienced activities were juxtaposed with actions traditionally

associated with FTF interactions to discover level of similarity or dissimilarity of experience in AVC. The purpose of this juxtaposition was to ascertain what is needed for AVC interactions to become more like FTF meetings,

• Findings were used for constructing a set of features to be systematically explored

• Features identified to be configured in the design of computerised annotating tools which will enable informed systematic annotation of observations

25Allan & Thorns June 08

Phase 4

Exploring annotation tools– The Memetic - incompatibility with

our system’s configuration – The Digital Replay System (DRS) –

useful as a quantitative tool for annotating micro-level actions and measuring the frequencies of these actions.

26Allan & Thorns June 08

Field Notes

• To import our AGVCR recording files to the DRS system we needed to convert them to a file format recognised by the DRS.

• Using Camtasia we captured desktop screens of AGVCR and converted them to QuickTime format (mov).

• DRS - useful for annotating micro-level actions and measuring their frequencies. For example tracking varying levels of participation from each node /or individual participant (Thorns et al., 2008)

27Allan & Thorns June 08

DRS Annotation Example

28Allan & Thorns June 08

Limitations so Far

• Using Camtasia limited the length of the recordings to 20 minutes segments to avoid crashing our computers

• The system was not very useful in annotating the more prominent features we have identified through our grounded investigation– Multimedia information - 38%– A sense of being with one another, a

sense of presence - 33%

29Allan & Thorns June 08

Summary of Trial

Build a better system so users will like it

But …..• What is a

better system ?

• Who decides what’s better?

• How do we know users will like it?

Build a better system through users’ eyesIdentify users’ criteria and design a grounded approach annotating system

30Allan & Thorns June 08

Summary

• Deploying the technology does not ensure uptake

• Traditional user centred designs are lacking and time consuming

• Effective Grounded approach can be achieved with annotation system developed enabling the analysis of large data sets

• Grounded annotation will provide systematic analysis of how technology is applied and constructed by users and so provide authentic information for further design of the technology

31Allan & Thorns June 08

Future Thoughts – Big Scale Grounded

Annotations

Develop a tool that will enable large scale observations to be annotated and analysed in a more effective and hopefully less time consuming way using grounded processes to inform its annotating schemes

32Allan & Thorns June 08

Enhancing the Design of Annotating Tools

• Easy synchronisation of recordings of AG and other AVC sessions– change- AG recording format or allow for format to

be read by annotating system)

• Speech recognition system to recognise repeating speakers – annotator identifies speaker first time and system

to recognise that speaker from then on

• Gaze tracking system able to work across cameras and follow gaze not only in physical node but across nodes – calculate how each camera is distorting the gaze to

arrive at who the person is actually looking at

33Allan & Thorns June 08

Grounded Annotation for Enhancing Design of

AVC

• Measure effectiveness of visual by measuring: (38% responses)– Rate of turn taking, – Levels of involvement of participants– Gaze tracking across sites

• Measure level presence in shared space by measuring:(33% responses)– Distances from cameras/ microphones– Number of times microphones muted– Positioning( measuring distances) of people in

relation to others in the physical node

34Allan & Thorns June 08

Sociological view

As social scientists we look NOT at what the technology can do( and how happy users are with it) but rather at what users actually do with the technology, which sometimes can be a little different from what designers thought they would do

35Allan & Thorns June 08

Final Note

Ask NOT what the technology can do for the user,

but what the user will do with the technology

Thank YouComments, Questions ….

Contact details:

[email protected]

[email protected]