A Theory of Diagnostic Inference. II. Judging Causality.

Abstract

Diagnosis involves inferring the process that generated a set of outcomes/results/symptoms. It is essentially causal rather than correlational, looks backward rather than forward, and is constructive in nature. A model of causal reasoning within diagnosis is presented that assumes judgments of causal strength to be a function of the strength of a hypothesis minus the strength of alternatives. However, such judgments are made conditionally on a background or field, analogous to the perception of figure against ground. Report demonstrates how changes in the causal background can lead to reversals in judgments of causal strength. The strength of an explanation is then considered as a function of cues to causality. These are multiple, probabilistic indicators that capture aspects of the data and content inherent in a causal relation. Results show that people trade-off content and data-based cues to causality and, judged causal strength decreases as a function of the strength of alternatives. Results and a theoretical model are discussed with respect to the importance of expectations in defining causal relevance, rules for combining the cues to causality, and replacing vs. disconfirming hypotheses. Finally, the normative implications of the theory are considered with respect to three trade-offs: the acquisition of causal knowledge vs. superstition; order-out-of-chaos vs. creativity; and imagination vs. uncertainty.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1982
Accession Number
ADA119781

Entities

People

  • Hillel J. Einhorn
  • Robin M. Hogarth

Organizations

  • University of Chicago

Tags

Communities of Interest

  • Biomedical
  • Human Systems
  • Weapons Technologies

DTIC Thesaurus Topics

  • Air Force
  • Analysis Of Variance
  • Applied Psychology
  • Behavioral Sciences
  • Causal Reasoning
  • Combinatorial Analysis
  • Human Factors Engineering
  • Information Science
  • Military Research
  • New York
  • Operations Research
  • Personality
  • Psychology
  • Reasoning
  • Statistics
  • Students
  • Systems Engineering

Fields of Study

  • Psychology

Readers

  • Artificial Intelligence
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference