A Parameter-Free Elicitation of the Probability Weighting Function in Medical Decision Analysis

Abstract

An important reason why people violate expected utility theory is probability weighting. Previous studies on the probability weighting function typically assume a specific parametric form, exclude heterogeneity in individual preferences, and exclusively consider monetary decision making. This study presents a method to elicit the probability weighting function in rank-dependent expected utility theory that makes no prior assumptions about the functional form of the probability weighting function. We use both aggregate and individual subject data, thereby allowing for heterogeneity of individual preferences, and we examine probability weighting in a new domain, medical decision making. There is significant evidence of probability weighting both at the aggregate and at the individual subject level. The modal probability weighting function is inverse S-shaped, displaying both lower subadditivity and upper subadditivity. Probability weighting is in particular relevant at the boundaries of the unit interval. Compared to studies involving monetary outcomes, we generally find more elevation of the probability weighting function. The robustness of the empirical findings on probability weighting indicates its importance. Ignoring probability weighting in modeling decision under risk and in utility measurement is likely to lead to descriptively invalid theories and distorted elicitations.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 483 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.