Анализ каракулей на полях диагностических бланков при решении теста на когнитивную рефлексию

One indirect piece of evidence supporting the existence of cognitive reflection as a switching mechanism between the first (intuitive) and second (analytical) thinking systems is the presence of characteristic margin notes or scribbles. The potential value of analyzing such marginalia has been noted by the author of the Cognitive Reflection Test (CRT), S. Frederick, as well as other researchers. This article presents both qualitative and quantitative analyses of marginal scribbles, conceptualized as by-products of in-person testing – spontaneous, unplanned visualizations of respondents’ reflective processes and task-solving mechanisms. The study addresses two empirical questions: (1) whether such scribbles serve as evidence supporting the hypothesis of cognitive reflection as a distinct facet of rationality, and (2) whether they instead reflect mathematical problem-solving skills. The sample included 452 students (aged 17 – 30) from universities and vocational institutions in Perm, Russia. All participants completed two sets of CRT items adapted from prior studies: S. Frederick’s original CRT-3, the extended CRT-7 (M. Toplak, R. West, K. Stanovich), and a verbal CRT version (K. Thomson, D. Oppenheimer). Scribbles were observed in 14.8% of the paper-based CRT forms. The highest frequency occurred with the classic CRT-3 tasks. Five categories of scribbles were identified: (1) correction of an incorrect intuitive answer to a correct reflective one without additional notes; (2) correction with written explanation; (3) visualizations of correct mathematical solutions; (4) incorrect calculations; and (5) uncertain responses. In most cases (77%), classic corrections from incorrect intuitive to correct reflective answers were not accompanied by any written calculations – indirectly supporting the cognitive reflection hypothesis. By contrast, mathematical calculations were equally associated with both correct reflective and incorrect intuitive responses, suggesting that a tendency toward calculation does not ensure correct answers to tasks designed with intuitive lures. Moreover, engaging the second thinking system without adequate mathematical skills may lead to incorrect responses, which might be mistakenly interpreted as intuitive. In conclusion, current evidence is insufficient to decisively confirm or refute either hypothesis. Both the cognitive reflection and mathematical skills hypotheses receive partial support. Nonetheless, the analysis of marginal scribbles in CRT forms appears to be a promising methodological tool for future research into the interaction between the two systems of thinking.