Skip to main content
. 2008 May 19;5:29. doi: 10.1186/1479-5868-5-29

Table 2.

Tests for Factorial Invariance of the Four Models across Different Genders: Goodness-of-Fit Statistics

Model S-B χ2 Df CFIa Changes in CFIa RMSEA (90%CI)a SRMRa
Enjoyment-Competence Model
 1: Configual model 131.55 33 0.94 0.05(0.04–0.06) 0.04
 2: Measurement model 169.73 40 0.93 0.01 0.05(0.04–0.06) 0.04
 Partial measurement modelb 154.79 38 0.94 0.00 0.05(0.04–0.06) 0.04
 3. Structural model 163.66 39 0.93 0.01 0.05(0.04–0.06) 0.04

Outcome expectation Model
 1: Configual model 168.42 49 0.94 0.04(0.04–0.05) 0.04
 2: Measurement model 179.11 56 0.93 0.01 0.04(0.03–0.05) 0.04
 3. Structural model 181.28 57 0.93 0.01 0.04(0.03–0.05) 0.04

Social Support Model
 1: Configual model 237.92 74 0.98 0.04(0.03–0.04) 0.03
 2: Measurement model 248.82 81 0.97 0.01 0.04(0.03–0.04) 0.04
 3. Structural model 257.70 86 0.97 0.01 0.04(0.03–0.04) 0.04

Environment Model
 1: Configual model 69.76 21 0.96 0.04(0.03–0.05) 0.03
 2: Measurement model 74.60 25 0.96 0.00 0.04(0.03–0.05) 0.03
 3. Structural model 78.06 28 0.96 0.00 0.04(0.03–0.05) 0.03

Based on the univariate χ2 incremental values (p < 0.05)

a: The cut-off points of model fit: CFI greater than .90, Changes in CFI equal or less than 0.01, the point estimate of RMSEA less than 0.05 and the 90%CI of RMSEA less than 0.08, SRMR less than 0.08.

b: Equality constraints related to item Enjoy3 and item Comp3 were removed sequentially