Abstract:

Metacognition is the cognition of cognition, including all aspects of cognition. The recent years have witnessed a rapid increase in research on the combination of metacognition and face cognition. Face cognition is often thought to be ‘special’ in some respects, due to its innateness, automaticity, neural specificity, and heredity. The combination of metacognition and face cognition can broaden the research of the two fields and explore whether the phenomena widely confirmed at the level of face cognition apply to its metacognitive level, and whether the metacognitive illusions found in many non-face fields also appear in the field of face cognition. In addition, face conveys important identity information and social information. The combination of metacognition and face cognition also has rich practical significance in social communication, clinical diagnosis, administration of justice, and other practical fields. The current research on metacognition of face cognition is still in its infancy. The term "metacognition" has been rarely directly mentioned, and the related concepts has been used instead. Moreover, most researchers have only focused on the study of the judgment of learning (JOL) in face memory. This paper aims to make an in-depth analysis of the current metacognitive research on face cognition, in order to provide more inspiring research ideas for scholars in the related fields, and promote the follow-up development of the empirical research on the combination of metacognition and face cognition studies.

This paper firstly systematically reviewed the domestic and foreign research on the cut-in points of the combination of metacognition and face cognition, and further summarized the main research methods. The cut-in points of recent studies on the combination of the two fields included the applicability of metacognitive illusions (Dunning-Kruger effect and egocentric bias) in face cognition and the applicability of face cognition phenomena (own-race effect and familiarity advantage) in metacognition. As for the main research methods of the research on the combination of metacognition and face cognition, according to different testing timings, methods have been mainly divided into three ways, including the prospective measurements (e.g., judgment of learning, JOL), concurrent measurements (e.g., feeling of knowing, FOK), and retrospective measurements (e.g., retrospective assessment questionnaires). According to different evaluation targets, methods can also be divided into the estimates of one’s own performance (e.g., self-assessment questionnaires) and the estimates of other people's performance (e.g., estimates of the number of people who answered correctly in the face cognition tasks). Each method has its own advantages and disadvantages, and it should be selected according to the metacognitive components that researchers aim to explore. Multiple methods can be combined based on different research aims.

Finally, four feasible directions for further theoretical and applied research were proposed. First, the previous research remains at the level of metacognitive monitoring. In the future, it can be expanded to the level of metacognitive control of face cognition, metacognitive performance of the subjective judgment of facial arousal and the cognitive neural mechanism of the two fields. Second, comparative studies can be conducted with non-face cognition fields to further explore whether the consistent metacognitive results across face and non-face fields are related to the domain general feature of metacognition or due to some possible personality traits. Thirdly, whether metacognitive ability of face cognition can be improved by improving the face cognition ability would also be an interesting direction for future studies to focus on. Fourthly, with the wide application of human face automatic recognition system, it is particularly important to expand the targets of the metacognitive evaluation from human beings to algorithms, based on the part of the need for manual review of machine learning face recognition results. This would provide new perspectives for understanding face cognition, and enhance its application values.

Key words: face cognition, metacognition, Dunning-Kruger effect, egocentric bias, own-race effect, familiarity advantage