Cognition about cognition: Do scales from different fields assess metacognition alike?
PDF copy of article available by clicking here.
Abstract
Metacognition is a construct of long-lasting interest in multiple fields of research. Yet, exchange between fields has been limited, leaving it an open question to what extent this construct can be conceptualized as a general cognitive entity. We thus implemented a cross-disciplinary analysis investigating if self-report scales from four fields tap into the same underlying construct and give rise to a general factor of metacognition (M). In a preregistered online study (N = 661) and utilizing an analytical approach to mitigate overfitting, a systematic model comparison showed that a bifactor model including a general factor of metacognition performed best. This general factor explained 61 % of the systematic variance, suggesting that there exists an important general component of metacognition. We will discuss how the different subscales of the four scales relate to one another and to M, elaborate on a potential jingle-fallacy in metacognition research, and give recommendations on which subscales to use to best tap into M. In sum, our integrative approach contributes to a better understanding of metacognition and how to best measure it.
Comment: The finding of a factor-analysis based M factor only reflects a general statistical factor in the collection of measures. It does not reflect a real ability…just a summary index of variance. Same as the general factor of intelligence (g) does not reflect a real brain-based ability…it is just a statistical index.
