RESULTS
The analysis had three stages (Davis and Lang, 2011). The
first stage was to conduct an exploratory factor analysis to
distill the product items into groups. Second and third, the
analyses tested the proposed conceptual model with CFA
and SEM. The process was to conduct separate measurement
models for each latent variable and then to create the
structural model. The commonly used approach was
employed as we wanted to use an analysis method that not
only supported model refinement but also had the ability to
rigorously assess model fit across our ten models (Kline,
1998). Before data analysis was started, initial data screening
was carried out for missing values and outliers, and the
normality of the dataset was tested, resulting in a sample size
of n = 515. In our analysis, we utilised a combination of
Microsoft Excel 2007 software, PASW Statistics 19 and
Analysis of Moment Structures (AMOS 19) (Arbuckle, 2010)
for SEM.