IS1825 Multimedia Development for Internet Applications Lecture 07: Usability Testing

1
IS1825 Multimedia Development for Internet
ApplicationsLecture 07 Usability Testing
Rob Gleasure R.Gleasure_at_ucc.ie http//corvus2.ucc.
ie/phd/rgleasure/index.html
2
IS1825

• Todays lecture
• Usability
• User Experience (UX)
• Testing

3
Usability

• Usability has been broken down into several
  components
• Learnability
• Usefulness (achieving goals)
• Satisfaction

Learning
Achieving goals
Satisfaction
4
Usability

• Dense, dense topic
• Usability strategies can be roughly considered as
  targeting either
• Conceptual models
• How can designers create a useful abstraction for
  users quickly and efficiently
• Ergonomics
• How can designers control the physical demands of
  users

5
Usability

• Conceptual-model approaches can be considered
  according to three main types
• Metaphor-based design
• Aligning to conceptual models of existing
  real-world systems with which users have
  experience
• Idiom-based design
• Aligning to conceptual models of systems of the
  same type with which users have experience
• Context-based design
• Aligning to conceptual models that users already
  possess of the design context

6
Learnability
7
Usability

• These strategies help learnability by allowing
  users a head start in understanding a system
• They help usefulness by allowing users to infer
  functionalities from this understanding
• They help satisfaction by creating expectations
  that are more likely to be met

8
User Experience (UX)

• In addition to this usage timeline, the field of
  user experience design (UX) has grown to capture
  important determinants of product design outside
  of this scope
• Product boxes, animations, imagery, screen
  transitions, etc.
• Several companies have thrived in this space,
  e.g. Bang Olufsen, Apple
• UX has steadily increased in importance for two
  main reasons
• Mass consumer markets have grown for IT products
  and non-specialists
• People use many systems because they just like
  them

9
7 Usability Testing Truths (Krug 2006)

• If you want a great system you have to test
• Testing one user is 100 better than testing none
• Testing one user early in the project is better
  then testing 50 near the end
• The importance of recruiting representative users
  is overrated
• The point of testing is not to prove or disprove
  something, its to inform your judgement
• Testing is an iterative process
• Nothing beats a live audience reaction
• Testing is essential if we want to improve our
  usability and UX

10
Usability Testing and Prototypes

• Some testing will be done on the features of the
  work-in-progress system itself (tracer bullets),
  however other testing can be done with prototypes
• Prototypes consist of anything from paper based
  representations to fully functional websites
• It allows stakeholders to interact with an
  envisioned product, gain experience in using it
  and explore uses for it
• Types of Prototyping
• Low-Fidelity Prototyping
• High-Fidelity Prototyping

11
Low-Fidelity Part 1
12
Low-Fidelity Part 2
13
Low-Fidelity Part 3
14
Types of Low-Fidelity Usability Prototypes

• Storyboarding
• Sketching
• Prototyping with Index Cards

15
Advantages of Low-fidelity

• They are simple to make
• They are cheap
• They can be quick to produce
• They are also easy to modify quickly to
  incorporate changes on the fly and retest,
• i.e. Is this what you mean?
• Minimal resistance to change

16
Disadvantages

• Limited error checking
• Navigational limitations
• Some users might find it difficult to
  conceptualise
• Very little UX feedback

17
High-Fidelity Prototyping

• Here materials that will be used in the final
  product are used
• The prototype will be designed to resemble the
  final product as closely as possible

18
Advantages

• Complete functionality, so people might
  understand it better
• Fully interactive
• Look and feel of the final product

19
Disadvantages

• Difficult to develop
• Time-consuming to create
• Not effective for requirements gathering

20
Testing the complete lifecycle

• For learnability
• Test the system with new users repeatedly
  forgetting to reset with new users from time to
  time is very common and leads to test-users
  behaving more like developers than true
  representative users
• For usability
• Try and form a hierarchical view of uses. The
  majority of features never get used, so make sure
  the priority items are right before moving on

21
Testing the complete lifecycle

• For satisfaction
• Use beta-tests to allow for more prolonged usage
  this lets you gauge users response over long
  periods of time and in less artificial
  environments (e.g. white coat syndrome)
• For UX
• Test the complete experience, not just the parts
  you think are important, e.g. the unboxing, the
  help dialogue, etc.
• Ask people how they feel in a way that really
  captures emotions and not just explanations

22
Issues with Testing

• Client involvement is a tricky balancing act. You
  absolutely need their feedback, however too much
  feedback can cause a number of problems
• Schedule bottleneck - progress is stifled waiting
  for clients to provide feedback
• Scope creep - the client keeps thinking of new
  requirements, the addition of which keeps adding
  massive delays. Alternatively they are not added,
  perhaps leaving the client feeling ignored

23
Issues with Testing (continued)

• Feedback fever - the client wants to make every
  decision, without necessarily understanding the
  technical repercussions
• Conflicting feedback - Jim from Marketing wanted
  to get involved in this project, so had a look
  over the complete site and wonders if it would be
  better as an iPhone app?

24
Want to read more?

• Norman, D. (2002). The Design of Everyday Things,
  Basic Books
• Cooper, A. (2007). About Face 3 The Essentials
  of Interaction Design
• Krug, S. (2005). Don't Make Me Think A Common
  Sense Approach to Web Usability, New Riders