Skip to main content
. 2017 Feb 21;12(2):e0170950. doi: 10.1371/journal.pone.0170950

Table 2. Goodness of fit indices for the hypothetical models tested and measurement invariance across gender and age.

Model χ2 df CFI TLI RMSEA (CI 90%) WRMR ΔCFI
1 factor 620.50 35 .95 .94 .07 (.07-.08) 2.27
1 factor + CE* 513.89 32 .96 .95 .07 (.06-.07) 2.01
Measurement Invariance for the one factor model
Gender
Male (n = 1287) 319.87 35 .95 .93 .08 (.07-.09) 1.62
Female (n = 1627) 369.26 35 .96 .94 .08 (.07-.08) 1.73
Configural invariance 638.74 73 .96 .95 .07 (.07-.08) 2.35
Strong invariance 778.85 89 .95 .94 .07 (.07-.08) 2.75 -.01
Age
13–15 years old (n = 1265) 293.53 35 .95 .94 .08 (.07-.09) 1.55
16–19 years old (n = 1649) 353.01 35 .96 .94 .07 (.07-.08) 1.71
Configural invariance 643.49 73 .95 .94 .07 (.07-.08) 2.35
Strong invariance 692.02 89 .95 .95 .07 (.07-.08) 2.54 -.01

Note. CE = Correlated Errors; χ2 = Chi square; df = degrees of freedom; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index; RMSEA = Root Mean Square Error of Approximation; CI = Confidence Interval; WRMR = Weighted Root Mean Square Residual; ΔCFI = Change in Comparative Fix Index.

*CE = 3 with 6; 3 with 7; 19 with 20.