Intelligent Computer Assisted Instruction (ICAI): Formative Evaluation of Two Systems.

摘要

This report reviews major components of an 18 month evaluation of Intelligent Computer Assisted Instruction (ICAI), and emerging field of Artificial Intelligence that draws on computer technologies and cognitive science in an attempt to build more powerful instructional programs. The primary goals of this effort were to develop an increased understanding of the state of the art of ICAI for the purposes of: (a) identifying strategies to enhance the general usefulness of ICAI technology for Army training problems, and (b) developing concepts for efficiently and effectively managing military ICAI projects. The approach taken to accomplish these goals was to intensively examine two selected ICAI systems using a formative evaluation methodology. The two systems selected were: (a) PROUST, a system designed by Soloway and Johnson for analyzing bugs in novice programmers' PASCAL programs, using a top-down approach which attempts to infer the intentions and plans of the programmer, and (b) WEST, a system designed by Burton and Brown to teach basic mathematics and strategic thinking skills, based on the premise that students can learn from their mistakes or 'bugs'.