An expert system model of organizational climate and performance

Abstract

Application of computer technology has greatly increased the manager's ability to make informed decisions about inanimate resources (e.g., money, materials, equipment, space and time). However, very little has been done to automate decisions involving human behavior because of the complexities involved. This research uses a third generation expert system development shell to create a prototype management consultant for behavioral issues. The frame-based, object-oriented expert system represents individuals and organizations in a decision support system. The expert system allows managers to make real time inquiries about the effect of changes in individual attitudes in specific organizations upon organizational performance. A survey questionnaire is developed to measure 133 individual attitudes. Selected organizational behavior and group dynamics findings are translated into 52 production rules. The rules are written as methods which are activated by the system following the structure of current behavioral models to predict performance. The system is validated by situational analysis. Individual attitudes are adjusted using fuzzy logic algorithms in 18 different situations, and the changes in calculated performance are compared with managers' predictions. Statistical analysis shows it is possible to predict changes in performance due to changes in attitude and circumstances.