Metacognitive monitoring and calibration in the vocal emotion recognition task

The influence of specific task demands on metacognition in socio-emotional processing, particularly emotion recognition, remains poorly understood. In this study, we aimed to address this gap by examining differences in metacognitive monitoring (MM) using self-rated confidence (SRC), metacognitive accuracy (MA), and objective vocal emotion recognition ability (ERA) across discrete emotions.

To this end, we used a categorical emotion recognition task. A hundred participants listened and were asked to identify one of six vocal emotions (anger, sadness, surprise, happiness, fear, and neutrality) in eighteen semantically neutral sentences. Immediately following each recognition response, participants provided a retrospective SRC.

A series of one-way repeated-measures ANOVAs revealed a significant main effect of emotion on MM, MA, and ERA. Post-hoc pairwise comparisons indicated no significant differences between specific emotions for MM. Conversely, significant differences between emotions were observed for both MA and ERA, which exhibited a similar pattern of variability. Notably, sadness demonstrated the highest calibration and accuracy, while fear showed the poorest performance on both measures.

Our findings suggest that the differences in MA between emotions are primarily attributed to the accuracy of ERA variations rather than the variations in MM. These results indicate that specific emotional content acts as a distinct task demand, influencing performance and metacognitive outcomes. Future research should explore the underlying mechanisms of MM in emotion recognition by comparing various MA indices across tasks and domains to provide a more comprehensive understanding of MA in the social domain.