Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Knowledge Technologies Institute
1
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
3 – Prototyping and Iterative Evaluations
Viktoria Pammer-Schindler March 15, 2016
Knowledge Technologies Institute
2
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Days and Topics March 1 Administrative Stuff, Intro to Designing Interactive Systems
March 8 Understanding Context of Use – Assignment 1 Handed Out
March 15 Prototyping and Iterative Evaluations – Assignment 1 Deadline (before/in lecture) – Assignment 2 Handed Out
April 12 Participatory Design (Theory and Workshop) – Assignment 2 Interviews with Tutor (throughout the day)
April 19/21 (maybe guest lecture on April 21, lecture on April 19 will be cancelled)
April 26 (10-13) Evaluations Workshop (Cognitive Walkthrough, Observation and post-hoc discussion of prototypes – Assignment 3) – Assignment 4 Handed Out
April 28 (9-14) Android Sensing / Context-Aware Interactive Systems Tutorial Day (different room)
May 3 Ubiquitous Computing, particularly Ubiquitous User Interfaces
May 10 Questions on Programming to Tutor
May 24 (10-13) Presentations 1 (Assignment 4)
May 31 (10-13) Presentations 2 (Assignment 4)
Knowledge Technologies Institute
3
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Recap
1. What do you need to plan for in an observation / semi-structured interview? a. Objective and entities of interest b. Cooperation with participants c. Data collection d. Interview guideline with key questions (for the interview)
Knowledge Technologies Institute
4
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Recap
2. What are the five perspectives of an activity that contextual design is interested in for work modelling? a. Sequence and intents (sequence model) b. Interaction and coordination (flow model) c. Rules and policies governing behaviour (cultural model) d. Physical space of activity (physical model) e. Artefacts of relevancde (artefact model)
Knowledge Technologies Institute
5
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Forward
• So, you observed and interviewed key users • … and have now some solution ideas • … which are so cool that you immediately want to
implement them and try them out.
• If you do this, you might spend months fine-tuning your functionality
• … then trying to sell this to your target users • … only to find out that for a long list of reasons users
don’t like your solution.
Knowledge Technologies Institute
6
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Today
Prototyping – Key Concepts … and a few tricks
Iterative Evaluations … we focus on early, informal evaluations that help you mature your solution ideas
Knowledge Technologies Institute
7
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Learning Goals After today’s lecture you should 1. Know about different kinds of prototypes 2. Know about different kinds of evaluations that are useful in UCD 3. Know and understand in principle the methods cognitive walkthrough
and heuristic evaluation (will be practiced in Assignment 3 after Easter)
After today’s assignment you should 1. Be able to represent the key problem (pain point) that your system will
solve in a storyboard 2. Be able to represent the measure of success for your system in a
storyboard 3. Be able to develop prototypes
Knowledge Technologies Institute
8
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
DESIGN
Analyse
Design/Implement Evaluate
Knowledge Technologies Institute
9
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Prototyping
Easier (for users + designers) to critique existing designs than to immediately think up a good one.
Early prototypes - low fidelity
– Many! – Different! – Throw away!
Evolutionary prototypes - from low to high fidelity – Slowly develop into a final product – Don’t start too early
Knowledge Technologies Institute
10
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Prototyping
Horizontal prototypes: Capture all of system/product, but not in depth
Vertical prototypes: Go in depth w.r.t. a specific
functionality / aspect, but don’t cover the whole system/product
Knowledge Technologies Institute
11
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Low-Fidelity Prototypes
Focus – Sequence of actions (both of user and system) – Content – “Insert” into earlier, visionary storyboards
Implementation: – Paper: Hand-drawn – Interactivity: Create multiple layers of “content” with sticky
notes, plastic overlays – Wireframes – more detailed issues like layout come into
focus
Knowledge Technologies Institute
12
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
http://alistapart.com/article/paperprototyping
Knowledge Technologies Institute
13
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
http://alistapart.com/article/paperprototyping
Knowledge Technologies Institute
14
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
http://www.paulolyslager.com/paper-prototyping-tool-participatory-design-research/
Knowledge Technologies Institute
15
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Tutorials / Manuals
– Write a tutorial before implementing the system – Explain and illustrate key steps – Can also walkthrough a tutorial with users to
check understanding / questions they would have before buying target system
Knowledge Technologies Institute
16
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Prototyping with a Computer
Simulate or animate parts of the system Depending on used technology may be the start of
evolutionary prototyping
– Take longer to implement than “no-coding” low-fidelity prototypes
– Stakeholders may already think “this is for real” – Fundamental flaws may be hard to detect – BUT useful for testing more complex functionality (e.g.,
Wizard of Oz evaluations)
Knowledge Technologies Institute
17
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
High-fidelity prototypes
For stakeholders, high-fidelity prototypes may be undistinguishable from “the real thing”
Knowledge Technologies Institute
18
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
So, what are high-fidelity prototypes missing?
– Full integration with other information systems – Convenience functionalities (e.g., import from
different filetypes…) – Working on multiple platforms
Assuming for all these that they are not key
functionalities integral to user oriented evaluations!
Knowledge Technologies Institute
19
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
How to use Prototypes
Create prototypes Evaluate prototypes
– a lot with users in the beginning of the process, to make sure that you got the problem, the task, the language etc. right (informal, interactive mixed design/evaluate processes)
– A lot with experts as your design matures, to capture usability problems – use experts to save users’ time, use experts to systematically capture design problems based on experts’ experience
– A lot with users again throughout the rest of the process to make sure that your development efforts have the right prioritisation
Knowledge Technologies Institute
20
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
EVALUATE
Analyse
Design/Implement Evaluate
Knowledge Technologies Institute
21
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Evaluation Goals
Does system match user needs? – Which needs to what extent (different kinds of
requirement on functionality or qualities of system) – Which users/stakeholders
Which effects does system have on stakeholders? – On user’s cognitive or work processes? – On organisational processes? – On key peformance indicators
Knowledge Technologies Institute
22
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Evaluation Goals
Formative evaluations – Identify directions for ongoing/future development
Summative evaluation – Make a stmt about “extent of match”
Knowledge Technologies Institute
23
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Types of Evaluations
Over time, evaluations get more formal and extensive. Expert Evaluations (no users)
– Save on evaluations by reviewing literature! – Cognitive Walkthrough – Heuristic Evaluations – Model-based Evaluations
Evaluating with Users
– Observations (Thinking aloud, protocol analysis, post-task walkthrough) – Experiments (most often: control group vs. intervention group) – Query-based (Interview, questionnaire)
This can be very powerful. Just grab a study colleague that does not go to this course, show him/her your prototype, give him/her a task, and observe/lislten to your friend trying to figure out what to do! (Read up in Nielsen’s “Thinking Aloud” – Link on Last Slide)
Knowledge Technologies Institute
24
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Cognitive Walkthrough
Goal: Figure out “learnability” – what prior knowledge do users need to have, over what can be explored within a system
You need
– A prototype system (whatever the quality/maturity) – A task that the user shall perform
• E.g.: Specify that you need to see your boss asap.
– A detailed list of actions the user needs to carry out to achieve the task with the prototype system
– An indication of what prior, relevant, experience the user has
Knowledge Technologies Institute
25
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Cognitive Walkthrough
At each action, ask 1. Does the effect of the user’s interaction with the system match
the goal of the action? 2. Does the user see that an action is available? (Recognition,
Match betw. System and real world, consistency) 3. Can the user recognize that the action is the right one?
(Recognition) 4. After action, can the user understand the system’s feedback?
(Visibility, Help in recovering from errors)
Knowledge Technologies Institute
26
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Example Cognitive Walkthrough
Knowledge Technologies Institute
27
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Heuristic Evaluation
Goal: Systematically figure out usability problems based on rules-of-thumb
You need
– A prototype system (whatever the quality/maturity) – Option: A task that the user shall perform
• E.g.: Specify that you need to see your boss asap.
– List of heuristics
Knowledge Technologies Institute
28
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Heuristics 1. Visibility of system status 2. Match between system and real world 3. User control and freedom 4. Consistency and standards 5. Error prevention 6. Recognition rather than recall 7. Flexibility and efficiency of use 8. Aesthetic and minimalistic design 9. Help users recognize, diagnose and recover from errors 10. Help and documentation
Knowledge Technologies Institute
29
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Example Heuristic Evaluation
Knowledge Technologies Institute
30
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Expert-based Evaluations
Use multiple experts Once you understand the methods
– Vary the questions in cognitive walkthrough • For instance, simplify to: Will users see what they need to do? If the
user does the right thing, will he know it? – Try to insert user actions into the walkthrough that correspond to
common “errors”, like pressing a button multiple times, the network connection failing, etc.
Knowledge Technologies Institute
31
V. Pammer-Schindler March 15, 2016 Sensors and User Models @TUG, Prototyping and Iterative Evaluations
Readings
Recommended Reading – Nielsen 1995, Ten Usability Heuristics for Design:
http://www.nngroup.com/articles/ten-usability-heuristics/
Further Readings – Nielsen 2012, Thinking Aloud:
http://www.nngroup.com/articles/thinking-aloud-the-1-usability-tool/
– Rohrer 2014, When to Use Which User Evaluation: http://www.nngroup.com/articles/which-ux-research-methods/