Integration of the ecological and error models of overconfidence using a multiple-trace memory model

Area of Science:

• Cognitive Psychology
• Decision Science
• Memory Research

Background:

• Overconfidence in probability judgments is a well-documented phenomenon.
• Existing models like Brunswikian and Thurstonian approaches offer partial explanations for overconfidence.
• Integrating memory processes offers a novel perspective on judgment calibration.

Purpose of the Study:

• To examine a memory-based account of probability judgment calibration.
• To describe the Minerva-Decision Making (MDM) model for understanding overconfidence.
• To investigate the impact of experience and encoding quality on calibration.

Main Methods:

• Utilized the Minerva-Decision Making (MDM) model, a multiple-trace memory model.
• Integrated ecological (Brunswikian) and error (Thurstonian) models of overconfidence within MDM.
• Conducted three experiments to test MDM's predictions regarding experience and encoding quality.

Main Results:

• Confirmed that increased experience decreases overconfidence.
• Confirmed that improved encoding quality decreases overconfidence.
• Demonstrated that both factors lead to lower output variance in the MDM model, improving calibration.

Conclusions:

• The MDM model provides a robust framework for understanding overconfidence through memory processes.
• Experience and encoding quality are key factors influencing the calibration of probability judgments.
• This memory-based approach offers significant implications for decision-making research.