Lossy encoding of distributions in judgment under uncertainty

Area of Science:

• Cognitive Psychology
• Decision Science
• Behavioral Economics

Background:

• Traditional models of judgment under uncertainty focus on normative ideals of accuracy.
• People's actual judgments often deviate from these normative ideals, suggesting alternative cognitive processes.

Purpose of the Study:

• To propose and test a new theory of judgment under uncertainty.
• To conceptualize judgment as lossy compression of probability distributions rather than direct probability expression.

Main Methods:

• Development of formal computational models based on the lossy compression theory.
• Conducting four experiments to test the predictive accuracy of these models against human judgment and interpretation of guesses.

Main Results:

• Computational models derived from the lossy compression theory accurately predicted human behavior in judgment tasks.
• The theory explains phenomena such as the aversion to vacuously-correct guesses and the conjunction fallacy.

Conclusions:

• Judgment under uncertainty can be better understood as efficient encoding of probability distributions.
• This lossy compression framework offers a novel perspective on cognitive biases and decision-making under uncertainty.