A proposed model and measurement instrument for the formation of IS satisfaction: the case of end-user computing satisfaction

This paper presents a model that explicitly defines satisfaction and the antecedent factors that help form it. The model distinguishes between the notions of expectations and desires and argues that both have an impact on overall satisfaction in the form of the difference between priors and post hoc usage perceptions coupled with the individual’s evaluation of these discrepancies. These two types of satisfaction, in turn, will have both direct and multiplicative impact on overall satisfaction. Given this understanding, we highlight possible limitations in existing instruments and provide a solution for creating new measures that should overcome these limitations. A complete set of measures is provided in this paper for future empirical testing, which are general enough to allow researchers to create measures for other aspects related to IS satisfaction beyond those targeted in this paper. We employ our model within the context of the five satisfaction areas outlined by Doll and Torzadeh (1988). Building upon their initial set of perceptual measures, we examine whether there are indeed two types of discrepancy effects, whether there are multiplicative effects for each discrepancy, and whether there is a higher order interaction between the two discrepancy components. Partial least squares analyses of data consisting of over 200 instructional staff member at a large university related to their satisfaction with an online grading system are employed and the results presented at the conference.