Examining User Perception of Third-Party Organization Credibility and Trust in an E-Retailer - End User Computing Challenges and Technologies: Emerging Tools and Applications - page 217

Figure 2. SEM model with standardized estimates

*p < .01
SMC in parentheses

X 2 = 134.063; p = .001
CMIN/DF = 1.43
IFI = .96
CFI = .96
TLI = .95

was applied that tested the measurement model
and then evaluated the structural model. The mea-
surement model included all the items underlying
the constructs of TPO credibility, purchase risk,
value, e-retailer trust, and purchase intentions.
The confirmatory factor analysis indicated an
acceptable fit to the data ( x 2 =132.17, p = .012)
with a goodness-of-fit (GFI) index of .863. An
acceptable fit of the model to the data provides
strong evidence of both discriminant and conver-
gent validity among the constructs (Anderson &
Gerbing, 1988).
The structural model containing the hypoth-
esized relationships among TPO credibility, pur-
chase risk, value, trust, and purchase intentions
then was tested. The model fit the data well ( x 2
= 134.063, p = .001). The fit measures (i.e., IFI
= .96; TLI = .95; CFI = .96) provide different
perspectives on the acceptability of the model
and indicate good support for the model with all
indices above the generally accepted mark of .90
(Hair et al., 1998).

All hypothesized paths are supported at the p
< .01 level of significance. Results indicate that
TPO credibility is positively related to both seal
value and trust in the e-retailer while negatively
related to purchase risk. Seal value, in turn, re-
lates to trusting attitudes in the e-retailer, while
purchase risk is negatively related to trust. As
expected, trust in the e-retailer is positively related
to purchase intentions.
The squared multiple correlation (SMC) is
equivalent to the more familiar R 2 and is an
indicator of the explanatory power of the model.
The SMC reveals the amount of variance of a
dependent variable that is explained by a set of
predictor variables. The explanatory power of the
model is fairly high, indicating that the model
explains 49.8% of the variance for trusting the
e-retailer as well as 31.1% of the variance in
purchase intentions.