Priors and payoffs in confidence judgments.

Area of Science:

• Cognitive Psychology
• Neuroscience
• Decision Science

Background:

• Priors and payoffs influence perceptual decision-making.
• Understanding their impact on confidence judgments is crucial but limited.
• Normative models suggest priors should affect confidence, but payoffs should not.

Purpose of the Study:

• To experimentally investigate how human observers' confidence judgments are affected by priors and payoffs.
• To determine if confidence reflects the probability of being correct, as predicted by normative theories.
• To compare observed behavior against Signal Detection Theory models of perceptual and metacognitive decision-making.

Main Methods:

• Conducted an orientation-discrimination task with manipulated priors and payoffs.
• Probed both perceptual (response selection) and metacognitive (confidence) decision-making.
• Analyzed discrimination and confidence criteria using Signal Detection Theory models.

Main Results:

• No participants consistently followed the normative model for confidence judgments.
• A majority exhibited non-normative confidence behavior.
• Some ignored payoffs entirely, while others incorrectly incorporated them into confidence assessments.

Conclusions:

• Human confidence judgments in perceptual tasks frequently deviate from normative predictions.
• Metacognitive errors in incorporating priors and payoffs can occur.
• These deviations have implications for real-world decision-making where conditions may not be matched.