ABSTRACT
Traditional research on design cognition typically employed artificial settings and quasi-realistic tasks. Studies examining individual designers have primarily employed questionnaires, qualitative interviews and diary self-reports. In sum, any methodology used to study design cognition should circumvent the memory distortions and omissions that limit the effectiveness of retrospective research methods. Moreover, in real-world design, the contextual setting is typically social and team-based. Protocol analysis studies that explore team-based interactions often utilise strangers, depriving designers of their prevailing social network and normal interaction partners. In vivo research requires domain knowledge of the vocabulary and structure of tasks so that the researcher can understand what is taking place. Reliability is important for any methodology used to study design cognition, but is particularly important in relation to in vivo data because of the high degree of contextual variance that can arise, as opposed to the relative contextual stability in experimental settings.
