A structural equation modeling (SEM) approach was
adopted in our data analysis, as it can simultaneously test
the structural and measurement models (Bagozzi and Yi,
1989), and it provides a more complete analysis for the
inter-relationships in a model (Fornell, 1982). The vari-
ance-based Partial Least Square (PLS) method was chosen
over covariance-based methods, such as LISREL, as it sup-
ports both exploratory and confirmatory research (Gefen
et al., 2000) and it can be applied to relatively small sample
sizes (Fornell and Bookstein, 1982). Chin (1998) and Gefen
et al. (2000) advise that the minimum sample size for a PLS
analysis should be the larger of (i) 10 times the number of
items for the most complex construct; or (ii) 10 times the