Abstract
This study examined the role of self-generated explanations on performance in a process control microworld, extending previous cognitive science research into a new, applied domain. The experiment was conducted using DURESS II, an interactive, thermal-hydraulic process control simulation. During this one-month experiment, participants controlled the system under normal and fault conditions on a quasi-daily basis. Participants in the self-explanation (SE) group occasionally watched a replay of their own performance immediately after completing a trial, while the control group did not. In addition, the SE group was instructed to explain aloud the reasons for their control actions while watching the replay. The SE participants were divided post-hoc into “good” and “poor” groups according to several performance criteria. An analysis of the protocols produced during self-explanation revealed that “good” SE participants included considerably more words suggesting self-explanations in their protocols than did the “poor” SE participants. Thus, self-explanation was correlated with performance within the SE group.
Get full access to this article
View all access options for this article.
