Testing Three Models of Calibration Accuracy

We compared three theoretical models, including one- and two-factor confirmatory factor analysis (CFA) solutions, as well as a three-level, hierarchical CFA solution in which a third-order general factor explained two second-order factors, which in turn explained variance in vocabulary, probabilities, and paper folding tests. Results supported the three-level model over the one- and two- factor models with respect to goodness of fit indices and fewest number of estimated parameters. The hierarchical solution was consistent with the hypothesis that calibration utilizes two different types of second-order processes (i.e., sensitivity and specificity) that rely on separate judgments of correct and incorrect performance that can be subsumed under a general third-order discrimination process (e.g., d').