Abstract
The authors describe the development and initial validation of a computer-based expert system, Mandate Consultant (Parry, 1985), designed to review the regulatory procedures for developing Individual Education Programs (IEP). The formative process involved three phases: (a) definition of need and proposed solution, (b) design of a prototype, and (c) progressive refinement through field-testing and revision cycles. The summative component included a two-phase experimental design for validating the accuracy of expert system output through comparisons with human experts. The findings indicated that the expert system-generated conclusions matched the conclusions of the “better” human experts, and were considered more appropriate than the conclusions of the majority of experts. Furthermore, “blinded” evaluators judged the expert system-generated conclusions as being equally acceptable as those produced by the “better” human experts, and more acceptable than those of the majority of experts.
Get full access to this article
View all access options for this article.
