Understanding consumer online shopping behaviour from the perspective of transaction costs

Table 5.7 Preliminary Data Analysis- Reputation of Online Store (REP) 
Step One - Exploratory
 
Wording EFA 
 
Loadings 
REP1 REP2 REP3 REP4 
Excellent 
public 
image 
.85 REP1 
1.00 
Extremely committed to customer satisfaction 
.88 
REP2 
.728 1.00
Is known to be concerned about customers 
.90 
REP3 
.661 
.716 
1.00 
Has a good reputation compared to others
.86 
REP4 
.597 
.644 
.765 
1.00 
Reliability 
0.90 
KMO 
.810 
Variance Explained 
76.41 
Bartlett’s 
.000 
Step Two – Confirmatory 
Indicators CFA 
 
Stand. Loadings 
Stand. 
Error 
Critical 
Ratio
AVE CR 
REP1 .72 
-- 
-- 
.66 
.887 
REP2 .78 
.036 
29.36
*
REP3 .92 
.050 
25.36
*
REP4 .83 
.044 
24.34
*
Model Fit 
Statistic 
Model Fit 
Statistic 
χ
2
.111 
TLI 
0.981 
df
3 
CFI 
0.972 
P .847 
RMSEA 
.002 
GFI 0.975 
RMR 
.003 
* 
p 
< .001
 
5.5.7 Preliminary Analysis – Perceived Convenience 
The perceived convenience construct was measured using six items (CONV1 to CONV6). 
These were subjected to the two-step preliminary data analysis process shown in Figure 5.1, 
the results of which appear in Table 5.8. Evaluation of the correlation matrix through the 
KMO and Bartlett’s Test results in a high KMO statistic (.838) 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 loading ranging from .77 to .87 and the variance explained was 65.60%.

---

228
Cronbach’s alpha of .89 was computed indicating 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 six items retained from Step One. 
The results indicated that this measurement model poorly fit the data (
χ
2
/df = 57.72 > 5, GFI 
= .836 < .90, TLI = .856 < .90, CFI = .856 < .90, RMSEA = .243 < .08, and RMR 
=.170 > .05). In order to improve the model, indicators which were related to larger 
reductions of chi-square or problematic standard residuals (2.5 as cut-off) (Anderson and 
Gerbing 1988) were identified and eliminated one by one. After removing item CONV5, an 
acceptable model was achieved. All overall goodness-of-fit statistics were within acceptable 
ranges. For example, 
χ
2
/df (1.565) was less than 5, GFI (.999), TLI (.998) and CFI (.998) 
were above .90, RMSEA (.024) was less than .08 and RMR (.010) was less than .05. 
Parameter estimates of the final measurement model were inspected and no problematic 
occasions were found. Table 5.8 shows that the standardized factor loadings of the remaining 
five items ranged from .65 to .91. All exceeded the minimum standardized loading .50 with 
three of them above the ideal criterion of .70 (Hair
 et al.
2006). The AVE was .59, exceeding 
the recommended level of .50 (Hair
 et al.
2006). The CR was .878, exceeding the threshold 
of .70 (Nunnally 1978). Taken together, items CONV1, CONV2, CONV3, CONV4 and 
CONV6 were considered to constitute a reliable and valid measurement scale for perceived 
convenience,
 
and these four items were retained for constructing the overall measurement 
model. 

---

229