Disfluency prompts analytic thinking--but not always greater accuracy: response to Thompson et al. (2013).

In this issue of Cognition, Thompson and her colleagues challenge the results from a paper we published several years ago (Alter, Oppenheimer, Epley, & Eyre, 2007). That paper demonstrated that metacognitive difficulty or disfluency can trigger more analytical thinking as measured by accuracy on several reasoning tasks. In their experiments, Thompson et al. find evidence that people process information more deeply-but not necessarily more accurately-when they experience disfluency. These results are consistent with our original theorizing, but the authors misinterpret it as counter-evidence because they suggest that accuracy (and even confidence) is a measure of deeper processing rather than a contingent outcome of such processing. We further suggest that Thompson et al. err when they discriminate between "perceptual fluency" and "answer fluency," the former of which is an element of the latter. Thompson et al. advance research by adding reaction time as a measure of deeper cognitive processing, but we caution against misinterpreting the meaning of accuracy.