Reliability Generalization of Perceived Ease of Use, Perceived Usefulness and Behavioral Intentions

In an effort to identify independence issues in our sample of coded studies, we searched on author names and sample size to identify studies that may have used the same subjects to measure the reliability of PEOU, PU, or BI, for the same technology. This review included multiple studies reported in one article and studies reported in different articles. After identifying studies with potential independence issues, we examined these studies in detail. If two or more studies used the same sample to measure the constructs of PEOU, PU, or BI, referencing the same technology (i.e., to evaluate acceptance of the same system), then only one of the studies was included in our sample. We retained or excluded duplicate studies based on the following criteria: (1) the reporting of Cronbach’s alphas was given priority over composite reliabilities, (2) the completeness of reporting for other construct reliability coefficients and study characteristics, and (3) precision in measurement reporting. If two or more studies used the same subjects but in different contexts (i.e., one study measured acceptance of a word processing application while another measured acceptance of a database application), then we retained all studies. Below we provide a list of studies that we examined for independence issues as well as specific details on whether we retained or excluded the studies and our exclusion criteria.