Functional Knowledge in Problem Solving

Abstract

An experiment compared solving operational and diagnostic problems after different instruction about a fictitious device. Solution of both kinds of problems was facilitated by instruction (1) that focused on functional relations among components of the device or (2) that focused on states of the individual components. For operational problems, this result contrasted with an earlier finding (Greeno & Berger, 1987) that only functional instruction facilitated inference and learning of operational procedures. In this study, component instruction included information about the states of switches and all participants saw a diagram of the device with information about connections between components. Both results are consistent with a characterization of relevant device-model knowledge by Kieras (1984) as including knowledge of device topology: connections between components and relations of connections to the controlling operations and indicators of the device. Comparison of information in the instructional conditions with planning nets for the operating procedures showed that functional instruction included needed information about connections between components and that component instruction included needed information about states of switches that determine connections between components.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 09, 1990
Accession Number
ADA228694

Entities

People

  • Daniel Berger
  • James G. Greeno

Organizations

  • Stanford University

Tags

DTIC Thesaurus Topics

  • Absorption
  • Artificial Intelligence
  • California
  • Classification
  • Cognitive Science
  • Electrical Plugs
  • Energy
  • Indicators
  • Plastic Explosives
  • Power Converters
  • Power Supplies
  • Procurement
  • Psychology
  • Security
  • Simulations
  • Solar Energy
  • Switches

Readers

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

Technology Areas

  • AI & ML