This paper presents a process the authors term "image-based experience sampling and reflection." The process combines three components, experience sampling, conjoint analysis and an image-based methodology.
The authors present a paper prototype of this methodology using a real-life case study. The technique uses experience sampling to gather information about a particular context at various times throughout the day. However, instead of prompting participants to answer a series of questions about their activities, the system records audio/video and later asks users to review the audio/video segments (via a PDA) and answer a series of questions, at their convenience. Participant responses to the questions are later analyzed using conjoint analysis in order to statistically analyze the data. The authors believe that this technique is less disruptive to users and provides more contextual information with the use of audio/video.
The example presented in that paper is that of a "virtual architect," a kitchen re-modeling application that helps users re-design their kitchen based on their actions and preferences. Various video cameras are installed in the kitchen with sensors in order to detect when someone enters the room. Video clips are then randomly captured throughout the day. Participants then use a PDA to review the images and provide feedback on questions about their actions and preferences. Given a large enough sample, these responses are then analyzed using a conjoint analysis.
The authors conclude that the results from the paper prototype testing indicate that this technique could prove valuable in other applications.
|
