Inexpert Calibration of Comprehension.

Abstract

Students with a wide range of coursework in physics or music theory read expositions in both domains. After reading, for each text students provided a judgment of confidence in ability to verify inferences based the central principle of the text. The primary dependent variable was calibration of comprehension, the degree of association between confidence and performance in the inference test. Two results of most interest were expertise in a domain was inversely related to calibration and subjects were well-calibrated across domains. Both of these results can be accommodated by a self-classification strategy: Confidence judgments are based on self-classification as expert or non-expert in the domain of the text, rather than an assessment of the degree to which the text was comprehended. Because self-classifications are not well differentiated within a domain, application of the strategy by experts produces poor calibration within a domain. Nonetheless, because self-classification is generally consistent with performance across domains, application of the strategy produces calibration across domains. Keywords: Comprehension; Meta-comprehension; Expert knowledge.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 01, 1986
Accession Number
ADA165701

Entities

People

  • Arthur M. Glenberg
  • William Epstein

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Applied Psychology
  • California
  • Cognitive Science
  • Computer Science
  • Contracts
  • Education
  • Educational Psychology
  • Engineering
  • Identification
  • Information Science
  • Judgment
  • Military Research
  • Procurement
  • Psychology
  • Social Sciences
  • Students
  • Universities

Fields of Study

  • Education

Readers

  • Artificial Intelligence
  • Regression Analysis.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Information Retrieval
  • AI & ML - Neural Networks