Understanding consumer online shopping behaviour from the perspective of transaction costs

 
5.9.1 Moderating Effect of Consumer’s Risk-bearing Propensity on Perceived TCs 

Online Purchase Behaviour 
To test the moderating effect of consumer’s risk-bearing propensity on the relationship 
between perceived consumer TCs and online purchase behaviour (H7a), a two-group 
structural model using the two groups of consumer risk-bearing propensity was established 
(Hair
 et al.
2006). Initially, the two-group structural model was tested allowing all of the 
hypothesized relationships, including the path of perceived TCs 

online purchase behaviour, 
to be estimated freely in both groups. Next, a second model was tested adding a constraint 
fixing the path of perceived TCs 

online purchase behaviour to be equal in both groups. 
The multigroup SEM fit results for each model are shown in Table 5.25. 
Table 5.25 Multigroup SEM Results of Testing the Effect of Consumer’s Risk-bearing 
Propensity on Perceived TCs 

Online Purchase Behaviour 
Totally Free
(TF) 
TCs

Online Purchase
Is Equal 
∆χ
2
χ
2
489.632 
493.194 
3.562 
df 198 
199 
1 
GFI .952 
.952 
CFI .967 
.967 
RMSEA .039 
.039 
The GFI and CFI were unchanged, and the RMSEA was the same for both models. However, 
the 
∆χ
2
= 3.562 with 1 
df
was insignificant (
∆χ
2
< 3.841, p > .10). Since the 
∆χ
2 
was 
insignificant, constraining the TCs 

online purchase behaviour to be equal between the two 

---

271
risk-bearing propensities groups did not hurt the model fit. Therefore, the model with equality 
constraint was supported instead of the totally free (TF) model. This result showed that the 
moderating effect of consumer’s risk-bearing propensity on the relationship between the TCs 
and online purchase behaviour was not supported. Thus, H7a was rejected. The conclusion is 
that the relationship between perceived TCs and online purchase behaviour was the same 
among either low risk-bearing propensity or high risk-bearing propensity online shoppers.