Cross-validation is an important and often neglected step in the scientific process. Measurement models can vary across samples and must be tested and retested before they are accepted as valid. In a review of user satisfaction instruments, Klenke concludes that there is an appalling lack of effort to cross-validate MIS instruments and calls for efforts to retest the End-User Computing Satisfaction (EUCS) instrument using new data. Using different sampling methods and a new sample of 359 respondents, this study replicates an earlier confirmatory factor analysis of the EUCS instrument. This replication suggests that the EUCS instrument is robust (i.e., not affected by sampling methods) and can be used with confidence to evaluate information systems.
