Empirical Evidence for a Descriptive Model of Implementation

This article presents a model of factors associated with the successful implementation of a computerbased information system or OR/MS model. The model hypothesizes that two classes of variables, model quality and management support, influence attitudes and perceptions of a model or system. The research model also predicts that attitudes and perceptions, management support, system or model quality, decision style and situational and personal factors are related to successful implemen tation. tn general, the relationships in the model receive support from the data; however, further research is needed, particularly to provide more evidence on possible causal links in the model. Certain classes of variables are also difficult to relate to implementation success, including decision style and personal and situational factors. If the evidence for the model is considered sufficient to take action, implementation strategies should concentrate on several key variables. First, the quality of the model must be high, both in terms of logic and user interface. Favorable attitudes and perceptions should be stressed during development of a model or system. Management support should be encouraged and solicited and the implem.entor should try to consider different decision styles. Finally, personal and situational factors are likely to be related to success. The purpose of this article is to summarize the findings of nine empirical studies which furnish data to test hypotheses derived from a descriptive model of implementation. It is hoped that hypothesis testing and models of implementation such as those proposed in [14] ~along with empirical studies will provide a better understanding of the variables associated with successful implementation.