Microsoft Word 2007 ichrie conference Proceedings Final-Final 06-06-07. doc

Convergent and Discriminant Validity
Convergent validity is evidenced if different indicators used to measure the same construct obtain strongly 
correlated scores. In SEM, convergent validity can be assessed by reviewing the t-test for the factor loadings 
(Hatcher, 1994). Here all factor loadings for indicators measuring the same construct were statistically significant, 
showing that all indicators effectively measure their corresponding construct (Anderson and Gerbing, 1988) and 
supporting convergent validity. Table 3 shows values of factor loadings for the modified model of 
experience 
(Model B). Observed variables specified to measure each of the constructs all have relatively high loadings 
(statistically significant at p < .05), ranging from .56 to .89, which is supporting evidence of convergent validity. 
Discriminant validity was examined by the pairwise correlations between factors obtained from the revised 
seven-factor model (27 items). As a rule of thumb, Kline (1998) suggests that each pairwise correlation between 
factors should not exceed .85. Estimated correlations between factors were not excessively high, ranging from .094 
to .603, and none of the pairs for the 95% confidence interval approach 1.00, thus providing support for discriminant 
validity (Anderson and Gerbing, 1988). 
Table 3 
Confirmatory Factor Analysis Results for Revised Model of Consumer Experience (N = 198) 
Factor 
loading 
t-value