Understanding consumer online shopping behaviour from the perspective of transaction costs

Table 5.14 Preliminary Data Analysis- Consumer’s Risk-Bearing Propensity (RISK) 
Step One - Exploratory 
Wording EFA 
 
Loadings 
RISK
1 
RISK
2
RISK
3
RISK
4 
RISK
5 
I am usually among the first to try
.71 
RISK1 
1.00 
I like to buy newer and different things 
.84 
RISK2 
.533 
1.00 
I like to try new and different places .87 
RISK3 .477 .713 1.00
I would not mind taking a risk 
.73 
RISK4 
.355 
.468 
.643 
1.00 
I 
enjoy 
buying 
unfamiliar 
brands 
.77 
RISK5 .495 .561 .555 .445 1.00 
Reliability 
0.85 
KMO 
.812 
Variance Explained 
62.28 
Bartlett’s 
.000 
Step Two – Confirmatory 
Indicators CFA 
 
Stand. Loadings 
Stand. 
Error 
Critical 
Ratio
AVE CR 
RISK1 .60 
-- 
-- 
.51 
.833 
RISK2 .87 
.076 
18.39
*
RISK3 .82 
.072 
18.32
*
RISK4 .55 
.061 
17.64
*
RISK5 .66 
.068 
13.67
*
Model Fit 
Statistic 
Model Fit 
Statistic 
χ
2
7.735 
TLI 
.986 
df
2 
CFI 
.997 
P .021 
RMSEA 
.055 
GFI .997 
RMR 
.020 
* 
p 
< .001
5.5.14 Preliminary Analysis – Perceived Enjoyment 
The perceived enjoyment was measured using five items (ENJ1 to ENJ5). These were 
subjected to the two-step preliminary data analysis process shown in Figure 5.1, the results of 
which are presented in Table 5.13. Evaluation of the correlation matrix through the KMO and 
Bartlett’s Test results in a high KMO statistic (.857) and a significant probability level 
(p< .001) for the Bartlett’s test. These results indicate that sufficient correlations were found 
within the correlation matrix for factor analysis to proceed. In addition, bivariate correlations 
were inspected and all coefficients fell within the acceptable range for factor analysis of .30 
and .90. EFA was then conducted which produced a single factor structure with strong item 

---

243
loading ranging from .80 to .90 and the variance explained was 75.12%. Cronbach’s alpha 
of .92 indicated good reliability of the scale. At this point, as all the items met the criteria of 
Step One, they were retained for CFA analysis in Step Two. 
In Step Two (refer Figure 5.1), CFA was conducted on the five items retained from Step One 
which resulted in all computed statistics indicating a very good fit of the measurement model. 
For instance, 
χ
2
/df (1.188) was less than 5, GFI
(.996
), TLI (.998) and CFI (.997) were 
above .90, RMSEA (.014) was less than .08 and RMR (.008) was less than .05. In addition, as 
shown in Table 5.15, the standardized factor loadings of the five items ranged from .73 to .96, 
all exceeding the preferable criterion of .70 (Hair
 et al.
2006). The AVE for this construct 
was .66, exceeding the recommended level of .50 (Hair
 et al.
2006). The CR was .907, 
exceeding the threshold of .70 (Nunnally 1978). Thus, items ENJ1 to ENJ5 were considered 
to provide a reliable and valid measurement scale for perceived enjoyment, and all five items 
were retained for constructing the overall measurement model. 
Having followed the sequential process of scale evaluation and scale verification as shown in 
Figure 5.1, all scale items retained for further analysis are argued to be both valid and reliable 
measures. The next analysis via AMOS 21.0 is to examine the overall measurement model 
which provides the basis for the hypotheses testing posed in this study. 

---

244