Surprisingly rational: probability theory plus noise explains biases in judgment

Area of Science:

• Cognitive Psychology
• Decision Science
• Behavioral Economics

Background:

• Human probability judgments are often seen as biased, suggesting a departure from probability theory.
• This perspective has significantly influenced various fields, leading to the belief that people cannot reason with probabilities.

Purpose of the Study:

• To propose an alternative model for probability judgments, emphasizing random noise rather than heuristic-based biases.
• To demonstrate that apparent biases can arise from random variation within a theoretically sound reasoning process.

Main Methods:

• Analysis of data from two experimental studies on probability judgments.
• Comparison of observed judgments with predictions from a noise-based model of probabilistic reasoning.

Main Results:

• For certain probabilistic expressions, judgments closely align with probability theory when noise effects are minimal.
• Systematic deviations in other expressions are explained by the proposed noise model.
• The model accounts for established biases like conservatism, subadditivity, and conjunction/disjunction fallacies.

Conclusions:

• Human probability judgments are fundamentally based on probability theory.
• Systematic biases observed in these judgments are attributable to random noise in the cognitive process.
• This noise-based account offers a more nuanced understanding of probabilistic reasoning and its deviations.