Instability of individual differences in the association between confidence judgments and memory performance

Area of Science:

• Cognitive Psychology
• Neuroscience
• Psychology

Background:

• Significant individual variability exists in the link between memory accuracy and subjective confidence.
• Understanding the stability of this metacognitive relationship is crucial for reliable assessment.

Purpose of the Study:

• To investigate the test-retest and alternate-forms reliability of individual differences in the accuracy-confidence association.
• To determine if metacognitive accuracy estimates are stable within a single test session.

Main Methods:

• Experiment 1: College students completed recognition memory tasks, followed by alternate forms two weeks later, assessing confidence-accuracy correlations.
• Experiment 2: Elderly adults underwent similar recognition memory testing to replicate findings.
• Experiment 3: College students recalled general knowledge answers, rating confidence, to assess within-test stability.

Main Results:

• Low split-half and alternate-forms stability were observed in the relationship between confidence and recognition performance.
• These findings were consistent across both college students and elderly adults.
• Individual differences in confidence-accuracy association for recall were not stable within the same test.

Conclusions:

• Individual differences in metacognitive accuracy are not stable within or across different memory test formats.
• Current methods may not yield reliable estimates of an individual's metacognitive accuracy.
• Development of more stable assessment procedures for metacognitive accuracy is needed.