Unskilled and unaware? Differences in metacognitive awareness between high and low-ability students in STEM

Abstract

IntroductionMetacognition, or the ability to monitor and control one's cognitive processes, is critical for learning in self-regulated contexts, particularly in introductory STEM courses. The ability to accurately make predictions about one's ability and performance can determine the effectiveness in which students effectively prepare for exams and employ good study strategies. The Dunning-Kruger pattern, where low-performing individuals are more overconfident and less accurate at the ability to predict their performance than high-performing individuals, is robustly found in studies examining metacognitive monitoring. The extent to which the Dunning-Kruger pattern can be explained by the lack of metacognitive awareness is not yet established in the literature. In other words, it is unclear from prior work whether low-performing students are “unskilled and unaware” or simply “unskilled but subjectively aware.” In addition, arguments about whether this pattern is a psychological phenomenon or a statistical artifact of the measurement of metacognition can be found in the literature.MethodsStudents enrolled in three different physics courses made predictions about their exam scores immediately before and after taking each of the three exams in the course. Student predictions were compared to their exam scores to exam metacognitive accuracy. A new method for examining the cause of the Dunning-Kruger effect was tested by examining how students adjust their metacognitive predictions after taking exams.ResultsIn all contexts low-performing students were more overconfident and less accurate at making metacognitive predictions than high-performing students. In addition, these students were less able to efficiently adjust their metacognitive predictions after taking an exam.DiscussionThe results of the study provide evidence for the Dunning-Kruger effect being a psychological phenomenon. In addition, findings from this study align with the position that the skills needed to accurately monitor one's performance are the same as those needed for accurate performance in the first place, thus providing support for the “unskilled and unaware” hypothesis.