How this was tested
To create a framework capable of addressing user experience over time authors refer to Beauty and Goodness – the two distinct overall evaluative judgements of quality for interactive products as proposed by Hassenzahl (one of the authors). To measure how product experience changes over time and what influences such changes, authors develop a questionnaire listing three distinct quality aspects of interactive products: pragmatics, stimulation, and identification.
While Beauty and Goodness, being high level summary judgements, were measured as single items, each of the quality aspects consists of seven bi-polar pairs (semantic differential scales) of attributes:
The product chosen for this study was an Interactive TV set-top box with particular focus on uWand – new class of ‘remote touch’ technology from Philips. 10 individuals used this product at home for 4 weeks and were asked to fill in a questionnaire during the 1st week of use and at the end of the 4th week.
Results of this test showed that relationships between attributes of quality aspects actually changed over time.
During the first encounters stimulation (original, creative, new, innovative) rather than identification (classy) was a determinant of Beauty, while pragmatic (practical, direct, manageable, predictable, clear) quality attributes influenced judgements of product Goodness. However, after four weeks, stimulation lost its strong impact on Beauty judgements, while Goodness became more related to identification (professional, inclusive, valuable, integrating, brings me closer to people) and stimulation (creative, courageous, original) rather than pragmatic qualities.
What conclusions can be drawn from these results? Well, for one this research proved that selected user experience aspects actually change over time. However, unfortunately the sample size is too small to make any conclusive decisions regarding whether the right aspects (and their attributes) where chosen for study and how exactly these aspects change over time. To a large extent these are drawbacks generally inherent to research methods based on semantic differential scales.
So we merely use this post as an introduction into user experience measurement frameworks and will follow up with more alternative approaches showing variety of applicable and useful methods.