Expert Systems: Their Impact on Performance and Cognitive Strategies in Diagnostic Inference.

Abstract

This report covers a 1-year research effort designed to develop a laboratory sequential diagnostic inference task, and assess the impact of introducing a computerized expert-aid system on performance of that task. The report details the laboratory task that was developed to be performed on a microcomputer, the expert system that assisted in the inference task, and the results of an initial experiment where subject performance was tracked under both manual and expert-aided conditions. Assessment of subject performance included a number of both subjective and objective measures such as accuracy, time to perform the task, subjective certainty, cognitive strategies, etc. Results indicate that: a) the development of the laboratory inference task was successful, and the various task manipulations had a great impact on performance measures: b) subjects' performance substantially changed as a result of using the expert system: and c) subjects' strategies were identifiable and did not qualitatively change as a function of introducing the expert system. Implications of the findings for the design and utilization of expert systems are discussed, as well as directions for future research.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1987
Accession Number
ADA188518

Entities

People

  • Sallie E. Gordon

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Acquisition
  • Air Force
  • Air Force Facilities
  • Analysis Of Variance
  • Artificial Intelligence
  • Cognition
  • Computer Programs
  • Computers
  • Data Analysis
  • Databases
  • Engineering
  • Expert Systems
  • Human-Computer Interfaces
  • Human-Machine Systems
  • Multivariate Analysis
  • Psychology
  • Students

Readers

  • Artificial Intelligence
  • Instructional Design and Training Evaluation.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference