Measuring General Computer Self-Efficacy: The Surprising Comparison of Three Instruments in Predicting Performance, Attitudes, and Usage

General computer self-efficacy is an individual’s perception of ability across multiple computer applications and is conceptually distinct from specific CSE, which targets a single activity. General CSE is currently measured using both general and global instruments, and in many studies, it has been a significant predictor of computing behaviors. But there have been weak or inconsistent results, which have confounded its value as a construct of interest. One reason for such results is a lack of research concerning instrument suitability; another is that current instruments do not adequately isolate the construct. This study introduces a new measure of general CSE, one calculated by summing application-specific CSEs (SGCSE). It then compares SGCSE with two current general measures, in predicting ability, attitudes, and usage. Results suggest that the choice of instrument significantly affects study results. SGCSE was best for performance predictions and the global measure was surprisingly robust.