A Theory of Diagnostic Inference: Contract Final Report,

Abstract

The essential aspects of diagnostic inference are that they are causal rather than correlational, backward rather than forward (one goes from effects to prior causes), concerned with a specific rather than the general case, and constructive (one can synthesize, enlarge, or otherwise develop new hypotheses). We have developed and tested models of two aspects of diagnostic infer-inference: judgments of the causal strength of a hypothesis/explanation, and how probabilistic assessments concerning the occurrence of a past event are made on the basis of (often conflicting) evidence received from less than perfectly reliable sources. The psychological rationales, quantitative formulations, and implications of these models are presented in this report together with a description of various experiments designed to test the models. We also discuss several commonalities between the two lines of research: the use of cognitive anchoring-and-adjustment strategies to cope with complex inference tasks; the constructive nature of diagnosis; the importance of surprise in inference; and the relation of the present work to normative standards of judgment and choice.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1983
Accession Number
ADA137032

Entities

People

  • H. J. Einhorn
  • R. M. Hogarth

Organizations

  • University of Chicago

Tags

Communities of Interest

  • Biomedical
  • C4I
  • Human Systems
  • Materials and Manufacturing Processes
  • Weapons Technologies

DTIC Thesaurus Topics

  • Air Force
  • Applied Psychology
  • Behavioral Sciences
  • Biological Sciences
  • Biomedical Research
  • Causal Reasoning
  • Engineering
  • Human Factors Engineering
  • Industrial Engineering
  • Information Science
  • Jet Propulsion
  • Military Research
  • Navy
  • Operations Research
  • Psychology
  • Reasoning
  • Systems Engineering

Readers

  • Artificial Intelligence
  • Theoretical Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference