Rational learning and information sampling: on the "naivety" assumption in sampling explanations of judgment biases

Area of Science:

• Cognitive Psychology
• Behavioral Economics
• Decision Science

Background:

• Judgment biases are often attributed to biased information processing.
• Alternative explanations suggest biases arise from non-representative information samples.
• Existing models often assume decision-makers are unaware of sample biases (naive).

Purpose of the Study:

• To investigate if biased judgments can occur without naive decision-making.
• To explore if awareness of sample generation and rational processing can still lead to errors.
• To provide a rational explanation for previously observed judgment biases.

Main Methods:

• Developed a rational analysis of Denrell's (2005) experience sampling model.
• Modeled scenarios where information search is 'interested' rather than 'disinterested'.
• Analyzed conditions under which rational sampling and processing yield systematic errors.

Main Results:

• Demonstrated that systematically biased judgments can emerge even with perfect information processing and awareness of sample biases.
• Proved that interested information search can lead to biased judgments, even under rational conditions.
• Showed that favoring outcomes with more accessible information is consistent with rational behavior.

Conclusions:

• The 'naivety' assumption is not necessary to explain systematic judgment biases.
• Rational decision-making, coupled with interested information search, can explain behaviors previously attributed to cognitive or motivational biases.
• Provides a rational framework for understanding phenomena like in-group bias and preference for popular alternatives.