Comparative fit analyses
A good test of model fit is to compare it against other competing nested models
(Hayduk, 1987; Sauer and Dick, 1993). The first alternative to be tested was a
model in which country was constrained to have no effect. The test was
accomplished by comparing the full model against a nested model in which the
parameters of country were set to equal zero. The result indicated that the full
model was significantly better: chi-square difference: (2 = 221.45 (df = 77)) ±
(2 = 115.60 (df = 75)) = (2 = 65.70 (df = 2); p < 0.005). The native country of
the respondents has a significant effect on how price affects their perceived
value and willingness to purchase.