The Rasch calibrated person measures were subjected to a path analytic model specifying the hypothesized relationships among the UTAUT constructs. Data from Rasch person measures differ from raw data inputted directly from scale scores in that they have been converted into log odd units with linear interval metrics, thus enabling the use of path analysis – a CTT-based statistical method. Descriptive statistics showed that the Mardia’s normalized estimate of multivariate kurtosis (12.089) was higher than the cut-off value of 3, suggesting departure from the multivariate normal distribution of the data. Therefore, the maximum likelihood estimation with Bollen-Stine bootstrapping method (Bollen & Stine, 1992) was performed to evaluate the path model. Values of key goodness-of-fit