Keep talking: an analysis of participant utterances gathered using two concurrent think-aloud methods

摘要

This paper presents the results of a study that compared two think-aloud styles: the classic approach and a relaxed think-aloud on the nature and number of participant utterances produced. Overall, ten categories of utterance were extracted from the verbal data ranging from categories that had a direct impact on usability problem analysis, to those which simply described procedural actions. There were no categories of utterance that were unique to either method. The interactive think-aloud led to the production of more utterances that could be directly used in usability problem analysis. Participants provided explanations, opinions and recommendations during classic think-aloud, even though they were not instructed to do so. This finding suggests that the social context of testing may override the classic instruction to think aloud.