Docstoc

Usability testing

Document Sample
Usability testing Powered By Docstoc
					Usability Testing
HiØ, Masterstudium Informatikk
Grensesnittdesign høsten 2006

 Gisle Andresen (gislea@hrp.no)
Forsker, Institutt for Energiteknik
Contents
1.   Introduction
2.   Usability measurement
3.   Data-collection techniques
4.   Evaluation of usability tests
User-centered design
•   A design approach driven by the
    needs and preferences of
    users in order to create effective
    interactive systems
•   UCD ensures Usability
              Why UCD?
              •          Negative
                         consequences of
                         not following UCD:
                         poor usability
                     •       reduced effectivity,
                     •       human errors,
                     •       frustration etc.


http://www.usabilitymustdie.com/usability_review_remote_control.htm
Why UCD?
•       Positive
        consequences of
        following UCD
        approach: good
        usability
    •    increased effectivity,
    •    Error tolerance,
    •    User satisfaction etc.
Why UCD?
•   A reaction against technology-
    centered design; i.e., design
    driven by what is technologically
    possible
•   left-over principle: we
    automate everything possible
    and leave the rest to the human
Key activities of UCD process
•   Identify user needs
•   Develop prototypes
•   Perform usability evaluations
Typical project model: Waterfall
       UCD and Waterfall
User need
  ident.         Prototyping

                               Evaluation
Usability evaluations
•   Inspection
•   Usability testing
Usability testing: characteristics
•   Several ”real” users participate
•   The users performs realistic tasks
•   Measure usability
•   Use/Problem descriptions
Why several real users?
•   Users differ from designers
•   Users differ from one another
Why realistic tasks?
•   Find usability problems of
    significance to task performance
Why measure?
•   Scientific reasons
•   Pragmatic reasons
                Scientic reasons

                  Everything
                 that exists,
                exists in some
                   amount




http://serendip.brynmawr.edu/exhibitions/Mind/Images/42.GIF
                Pragmatic reasons
                •          Define specific goals
                •          Easy to communicate
                           results
                •          Allows statistical
                           analysis
                •          High status


http://www.usabilitymustdie.com/askJakob.htm
Why Use-/problem descriptions?
•   Learn how the system is really used
•   Find causes of usability problems
Usability testing approaches (Preece)
•   Formative vs. Summative
•   Quick and dirty vs. Experimental
Usability testing approaches (Rubin)
•   Exploratory test
•   Assessment test
•   Validation test
•   Comparison test
What determines the approach?
•   Purpose of test
•   Stage in the design process
•   Practical limitations
              Usability measures

              Measure:
              to assign numbers
              to objects or attributes
              according to rules




http://www.psych.utoronto.ca/museum/hippchron.htm
Creating a usability measure
•   Select usability attribute
•   Operationalise criteria
•   Select data-collection technique
   Usability attributes

                                      Efficiency


Usefullness
                  Learnability

                                      Effectivity
        Satisfaction
                       Memorability
Select attribute
•   What attributes are most relevant?
•   Important to take into account
    characteristics of the users, tasks
    and context
Select attribute: Examples
•   Commercial: easy to learn
•   Entertainment: satisfaction
•   Seldom used: memorability
•   Innovative: usefullness
•   Safety relevant: effective
•   Complex: efficient
Operationalise criteria
•   Can the attribute be ”measured”?
•   Define a scale
•   Define acceptance or target criteria
Operationalise criteria: Example 1
•       Effective
    •     Number of tasks performed correctly
    •     Scale: Summarise the number of tasks
          performed and divide by number of
          possible tasks
    •     Criteria: All users must perform all
          important tasks
Operationalise criteria: Example 2
•       Efficient
    •     The effort users have to invest in
          performing the task
    •     Scale: the more effort, the less
          efficient
    •     Criteria: less effort than old product
               Operationalise criteria: Example 3




http://zing.ncsl.nist.gov/hfweb/proceedings/etgen-cantor/
Student tasks
•   Can you think of any potential
    usability problems with your product?
•   What is the most important usability
    attribute for your product?

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:10
posted:10/3/2011
language:Norwegian
pages:30