AMORY H. DANEK CAROLA SALVI Moment of Truth: Why Aha! Experiences are Correct ABSTRACT Having a sudden insight is often associated with inherent confidence, enough for Archimedes to run naked through the streets shouting “Eureka!”. Recent evidence demonstrates that public displays of enthusiasm, such as the ancient polymath’s, are actually supported by a higher likelihood of being correct. Keywords: insight, creativity, problem-solving, Aha! experience, confidence. The history of great discoveries is full of anecdotally successful insights, suggesting that solutions achieved via insight are more likely to be correct than those achieved without it. But is this sense of correct- ness actually justified by higher solution accuracy? Although research on insight problem-solving dates back nearly 100 years, this question has been addressed only recently. For decades, insight has been defined by the assumption that “insight problems” could be solved exclusively by insight, and therefore only accurate solutions were considered, based on small numbers of problems. A new operational definition of insight as self-reported Aha! experience, together with the use of large problem sets that could either trigger an Aha! experience or be solved without one, allowed researchers to escape this circular logic and to produce comparative data on the accuracy of solution types (Bowden, Jung-Beeman, Fleck & Kounios, 2005). Recent evidence converges on the following finding: when the solution to a problem comes to mind accompanied by a self-reported Aha! experience, it is more accurate compared with solutions where this feeling of epiphany is missing (Danek, Fraps, von M€ uller, Grothe & € Ollinger, 2014; Salvi, Bricolo, Kounios, Bowden & Beeman, 2016). This “accuracy effect” holds across several different task domains: Compound Remote Associates problems (CRAs), anagrams, rebus puzzles, line draw- ings (Salvi et al., 2016), and magic tricks (Danek et al., 2014; Hedne, Norman & Metcalfe, 2016), (Figure 1). Earlier, Metcalfe found that solution processes which felt “subjectively catastrophic” (i.e., came to mind sud- denly) were more often correct than those which did not (Metcalfe, 1986).In the same vein, Kounios and col- leagues demonstrated that those who tend to solve more problems via insight make more errors of omission (i.e., when people time out), whereas those who do not, make more errors of commission (i.e., incorrectsolu- tions) (Kounios et al., 2008). Salvi et al. (2016) directly approached this issue, demonstrating higher accuracy for insight solutions compared with non-insight solutions in four different experiments involving different tasks. This effect remained stable when potential confounds were excluded, for example, problems solved too quickly (considered as immediate recognition), or solved in the last 5 s (considered guesses, mostly yielded without a concurrent insight). Despite this converging evidence, it remains unclear exactly why Aha! solutions are more likely to be correct than solutions for which no Aha! is reported. Since problem-solvers do not receive feedback as to whether their solutions are correct, where does this intuitive sense of success come from? A possible explanation for this effect could be that solvers use confidence in correct solutions as a cue for reporting an insight. In the absence of feedback, indeed, correct solutions receive higher confidence rat- ings than incorrect solutions (Danek & Wiley, 2017). When accurate, solvers may feel highly confident about their solution and therefore retrospectively report having had an Aha! experience. At first glance, this argument seems to be supported because confidence highly correlates with Aha! experience ratings (Danek & Wiley, 2017; Webb, Little & Cropper, 2016), suggesting that the Aha! experience is rated in alignment with the level of confidence. Furthermore, confidence is a key dimension of this multi-dimensional experi- ence (Danek & Wiley, 2017). However, these two studies (Danek & Wiley, 2017; Webb et al., 2016) specifi- cally mentioned confidence in their instructions, possibly inflating the relation between confidence and Aha! ratings. Two following studies showed (Hedne et al., 2016; Salvi et al., 2016) that the accuracy effect persists even when the confidence confound is eliminated from instructions; i.e., when Aha! experiences were not explicitly defined as associated with confidence. Moreover, solvers sometimes feel confident also about The Journal of Creative Behavior, Vol. 0, Iss. 0, pp. 1–3 © 2018 Creative Education Foundation DOI: 10.1002/jocb.380 1