Abstract
Psychology faces a measurement crisis, and mind-wandering research is not immune. The present study explored the construct validity of probed mind-wandering reports (i.e., reports of task-unrelated thought [TUT]) with a combined experimental and individual-differences approach. We examined laboratory data from over 1000 undergraduates at two U.S. institutions, who responded to one of four different thought-probe types across two cognitive tasks. We asked a fundamental measurement question: Do different probe types yield different results, either in terms of average reports (average TUT rates, TUT-report confidence ratings), or in terms of TUT-report associations, such as TUT rate or confidence stability across tasks, or between TUT reports and other consciousness-related constructs (retrospective mind-wandering ratings, executive-control performance, and broad questionnaire trait assessments of distractibility-restlessness and positive-constructive daydreaming)? Our primary analyses compared probes that asked subjects to report on different dimensions of experience: TUT-content probes asked about what they'd been mind-wandering about, TUT-intentionality probes asked about why they were mind-wandering, and TUT-depth probes asked about the extent (on a rating scale) of their mind-wandering. Our secondary analyses compared thought-content probes that did versus didn't offer an option to report performance-evaluative thoughts. Our findings provide some "good news"-that some mind-wandering findings are robust across probing methods-and some "bad news"-that some findings are not robust across methods and that some commonly used probing methods may not tell us what we think they do. Our results lead us to provisionally recommend content-report probes rather than intentionality- or depth-report probes for most mind-wandering research.