To be able to argue for a decision pro or contra a usability property one must be able to support the argument by appropriate measurements. A good
comprehensive overview about HMI testting can be found in Dumas and Fox 2008 [38]. Dumas and Fox stress the importance to view usability
testing as part of the overall design and engineering process. In that respect it differs from the traditional testing paradigm of psychology in
general as well as from the more specialized paradigm of cognitive psychology (cf. figure 4.1). They enhance furthermore the scope
of usability testing which usually is looking for functionality and usability by adding the growing
segment of gaming by introducing the the dimension of pleasure [38]:1145.
Figure 4.1:
Psychological and Usability Tests; Differences
|
The main difference between the usual psychological experimental set up and a HMI setup can be characterized as follows:
- Usability testing is primary located within the design phase of an engineering process; psychological tests are intended to measure rather
general properties of human persons based on representative samples (there is also some usability testing after the release of a completed system,
but this is a kind of a validation checking the conformance of the requirements with the completed system).
- While the number of test persons in usability testing can be small (1 - 5), do psychological test need 'sufficiently large' samples (usually
more than 10)
- 'Speaking aloud' and 'assisting' the test person during a usability test is possible, this is not possible during a psychological test
- The 'environment' for usability tests can be highly 'artificial' (fakes, mock-ups, simulations...) to enable the detection of 'flaws' in the
design; in psychological tests the restrictions for independent variables are stronger to be 'reproducible' and 'objective'
Subsections
Gerd Doeben-Henisch
2012-12-14