Beliefs and Bayesian reasoning.

Area of Science:

• Cognitive Psychology
• Decision Science
• Bayesian Inference

Background:

• Human reasoning often deviates from normative models.
• Beliefs about real-world probabilities may impact statistical judgment.
• Understanding factors influencing Bayesian reasoning is crucial for cognitive science.

Purpose of the Study:

• To investigate if real-world probability beliefs affect posterior probability judgments in Bayesian reasoning.
• To determine how the proximity of correct posterior probabilities to intuitive estimates influences reasoning accuracy.

Main Methods:

• An internet-based experiment was conducted with participants assessing posterior probabilities in Bayesian problems.
• The study manipulated the closeness of the correct posterior probability to participants' intuitive estimates.
• A computational model was used to analyze the relationship between intuitive estimates and calculated probabilities.

Main Results:

• Participant responses were significantly more accurate when the correct posterior probability was near their intuitive estimate.
• Judgments of posterior probability were substantially influenced by pre-existing beliefs about real-world probabilities.
• A weighted sum model accurately predicted responses, combining intuitive estimates and provided probabilities.

Conclusions:

• Cognitive biases related to real-world probability beliefs play a significant role in Bayesian reasoning.
• The alignment between intuitive and normative probabilities is key for accurate statistical inference.
• Future research should explore interventions to mitigate the influence of intuitive beliefs on Bayesian judgments.