Hypotheses Results
The hypotheses were tested via a PLS approach to SEM. Structural regression modeling software, SmartPLS version 2.0 (Ringle, Wende, and Will 2005) was used for this analysis. This software was selected because PLS analysis has the ability to handle larger, more complex models with multiple latent variables and indicators. PLS analysis also accommodates nonnormally distributed data, which often occurs in behavioral studies (Chin, 1998). PLS analysis is, therefore, appropriate for this study, given the multiple relationships and manifestation variables employed in the theoretical model and the nonnormal distribution of the emotional value construct in the model.