Back to search

Hosted by Utah State University Libraries

A Method for Knowledge Engineering in Clinical Decision Making

Abstract

dc:description.abstract

The purpose of this study was to validate the problem behavior evaluation section of an expert system computer program, Class.BD. Class.BD was developed to assist special education personnel in determining whether students qualify for special education services as behaviorally disordered/severely emotionally disturbed students. The subjects were six Utah who regularly individuals from the state of 1) work with behaviorally disordered/severely emotionally disturbed students and 2) participate in multidisciplinary assessment teams. Three of the subjects were special educators, and three were school psychologists. Specifically, this study investigated the impact of five behavioral factors on the subjects' ratings of the seriousness of problem behaviors. The five behavioral factors were 1) the severity or nature of the problem behavior, 2) the frequency with which the problem behavior occurs, 3) the duration over which the problem behavior has been occurring, 4) the generality of the problem behavior or the number of school environments the behavior occurs in, and 5) the percentage of the student's peers who engage in the same behavior. For each behavioral factor, three levels of that factor were determined: high, moderate, and low. Problem behavior descriptions were developed by the researcher, each of which presented the five behavioral factors at a predetermined combination of levels. Of 65 problem behavior descriptions, 33 described externalized problem behaviors and 32 described internalized problem behaviors. Subjects were asked to rate the seriousness of each problem behavior description on an 11 point scale, where l=mild and ll=severe. The results showed high levels of agreement among subjects on ratings of seriousness of problem behaviors. There was also high agreement between the subjects' ratings and ratings generated by the Class.BD expert system. Thus, Class.BD was validated. Further, the subjects gave highly similar ratings to descriptions of externalized and internalized problem behaviors. The results also indicated that the severity of the problem behaviors had the most impact on subjects' ratings. Subjects discriminated three levels of severity but only two levels of frequency, duration, generality, and percentage of peers. Finally, the results provided support for the use of analysis of variance as a viable method of knowledge engineering, i.e., extracting information about how experts make decisions. Its superiority over traditional interview methods is discussed.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy (PhD)
Level thesis:degree_level
Dissertation
Discipline thesis:degree_discipline
Psychology
Grantor dc:publisher
Hosted by Utah State University Libraries

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Giere, Sheila S.
Contributors dc:contributor
  • Richard Baer
  • Michael Bertoch
  • Joseph Ferrara

Subjects

dc:subject × 4

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
Utah State University
Base URL
digitalcommons.usu.edu/do/oai/
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Giere, Sheila S.. A Method for Knowledge Engineering in Clinical Decision Making. Dissertation thesis, Hosted by Utah State University Libraries, https://digitalcommons.usu.edu/etd/5989