Is Overconfidence a Trait? An Adversarial Collaboration

Area of Science:

• Psychology
• Behavioral Science

Background:

• The nature of overconfidence, specifically whether it is a psychological trait, remains a subject of scholarly debate.
• Previous research has faced challenges in isolating overconfidence from task-specific performance factors.

Purpose of the Study:

• To investigate whether overconfidence can be reliably measured as a psychological trait.
• To address the scholarly dispute regarding the fundamental nature of overconfidence.

Main Methods:

• An adversarial collaboration between two research teams with preregistered analyses and predictions.
• Development of novel tasks with ostensibly random performance to measure trait overconfidence without task-related confounds.
• Study involved 942 U.S. adults, with data collected via CloudConnect.

Main Results:

• Robust relationships in confidence were found across novel tasks, as indicated by confirmatory factor analyses and raw correlations.
• Findings suggest that individuals exhibit consistent beliefs about their performance capabilities even when performance is random.
• Evidence supports the existence of trait overconfidence.

Conclusions:

• The study provides support for the claim that overconfidence may function as a genuine psychological trait.
• Methodological innovation in measuring confidence on random-performance tasks offers a new avenue for trait assessment.