Confidence judgments: The monitoring of object-level and same-level performance

The influential metacognitive framework of Nelson and Narens (1990) distinguishes between object-level and meta-level, with two metacognitive processes, monitoring and control, governing the interplay between them. Monitoring refers to the process by which the meta-level tracks the accuracy of object level-performance, whereas control refers to the processes by which the meta-level regulates object-level processes. In this study, I examine the prediction derived from Koriat’s (Psychological Review, 119, 80–113 2012a) self-consistency model (SCM) that when people indicate their confidence in the accuracy of their choice, their confidence actually monitors the likelihood that others will make the same choice better than the accuracy of that choice. This was shown to be the case for three levels of processing: choosing the correct option, predicting the choice made by others, and predicting the predictions made by others about the majority choice. The conditions under which object-level correspondence and same-level correspondence are aligned or diverge are discussed.