Skip to main content
. 2023 Dec 5;14:1287274. doi: 10.3389/fpsyg.2023.1287274

Table 3.

Multiple-group invariance testing grouped by gender, age, and language skills.

Model across gender Compared Model χ2(df) ΔSBχ2 (Δdf) CFI RMSEA SRMR
A: Configural invariance 8.264 (10) 1.000 0.000 0.017
B: Full metric invariance A 16.899 (14) +9.321 (4) 0.993 0.028 0.044
C: Full scalar invariance B 25.149 (18) + 8.527 (4) 0.982 0.039 0.050
D: Full strict invariance C 28.123 (23) + 3.300 (5) 0.987 0.029 0.056
Model across age
A: Configural invariance 8.832 (10) 1.000 0.000 0.020
B: Full metric invariance A 11.731 (14) +2.830 (4) 1.000 0.000 0.030
C: Full scalar invariance B 17.924 (18) +6.755 (4) 1.000 0.000 0.035
D: Full strict invariance C 26.079 (23) +8.119 (5) 0.992 0.023 0.046
Model across language skills
A: Configural invariance 10.012 (15) 1.000 0.000 0.022
B: Full metric invariance A 13.024 (23) +2.642 (8) 1.000 0.000 0.031
C: Full scalar invariance B 32.791 (31) +21.175 (8)** 0.994 0.019 0.049
D: Partial scalar invariance (except PS) B 23.685 (29) +11.644 (6) 1.000 0.000 0.043
E: Partial strict invariance (except PS) D 31.039 (39) +7.399 (10) 1.000 0.000 0.054

PS, physical spontaneity; χ2, Satorra-Bentler scaled Chi-square; df, degrees of freedom; Δ, difference value; CFI, comparative fit index; RMSEA, root mean square error of approximation; SRMR, standardized root mean square residual; **p < 0.01.