New and Old Biases in Subjective Probability Distributions: Do They Exist and are They Affected by Elicitation Procedures.

Abstract

Past research indicates that people exhibit biases in assessing probability distributions on continuous variables. Three types of biases have been identified: too many true values falling into the extreme tails of the distributions, a displacement toward 50% for distributions assessed on percentages, and a general tendency to underestimate. This study explored the nature of these biases with particular emphasis on how they interact and how they are affected by the procedure used to elicit the distributions.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 1977
Accession Number
ADA048600

Entities

People

  • David A. Seaver
  • Tsuneko Fujii
  • Ward Edwards

Tags

Communities of Interest

  • Biomedical
  • Human Systems

DTIC Thesaurus Topics

  • Accuracy
  • Biological Sciences
  • California
  • Data Analysis
  • Department Of Defense
  • Engineering
  • Geography
  • Military Personnel
  • Military Research
  • Probability
  • Probability Distributions
  • Psychology
  • Public Policy
  • Social Problems
  • Social Sciences
  • Students
  • Systems Engineering

Readers

  • Regression Analysis.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.