Strategies for Control of End-User Computing: Impacts on End Users

The prevalence of end-user computing has resulted in a need for more effective methods of controlling such computing activities. The present experiment examined the impact of two potential control strategies, behavior-based and outcome-based control, on end-user affective reactions. Results indicated that subjects in the behavior-based control condition reported greater levels of stress, less self-determination, and less process satisfaction than the outcome control group. Further analysis suggested that the detrimental effect of behavior-based control on process satisfaction could be explained through the mediating effect of stress. Implications of the results for the design of effective end-user control strategies are discussed. INTRODUCTION user (Alavi and Weiss, 1985-86; Davis and 01son, 1985). These potential benefits and the imEnd-user computing (EUC) refers to the capaprovements in the performance/price ratios of bility of users to have direct control of their own computer hardware and software have been the computing needs (Davis & Olson, 1985) includdriving force behind the EUC activities in oring the capability to develop their own applicaganizations. Despite its potential benefits and tions. These applications may be solely created the general enthusiasm for end-user computing by the end user or through the assistance of activities, there is a growing concern about the others (data processing professionals or other organizational risks and cost of these activities users). End-user computing is one of the most (Alavi and Weiss, 1985-86; Davis and Olson, important recent developments in corporate 1985). Consider the following vignettes: computing environments. It is a rapidly expaneling phenomenon and is of growing strateUn a large electronics company about half of the gic importance to many corporations (Benson, corporate computer resources were being con1983; EDP Analyzer, 1983; Gritty & Rockart, sumed by managers and staff who were develop1984), and it is estimated that four out of five ing and operating their own applications. Proadministrative and professional workers will be jection for support of EUC activities were comdirectly applying computing tools to support parable to the cost of an additional large their work by 1990 (Davis & Olson, 1985). mainframe computer every six months! Senior management became very concerned abut the Potential benefits attributed to EUC include encomputing expenditures in this area and hanced productivity, overcoming the shortage of demanded more controls and reassurance that data processing professionals, ease of use and the high EUC costs were, in fact, resulting in responsiveness, and overcoming implementation comparable benefits (Henderson and Treacy, problems by transferring this process to the end 1986).