Skip to main content
. 2022 May 6;13:795770. doi: 10.3389/fpsyg.2022.795770

TABLE 4.

Measurement invariance of the QEWB for Samples 1, 2, and 3.

Model χ2 df p CFI RMSEA Model comparison χ2 df p Δ CFI Δ RMSEA
Samples 1 and 2 (Model 5a)
Invariance Model 1 353.179 264 <0.001 0.966 0.029
Invariance Model 2A 485.643 332 <0.001 0.941 0.034 2A vs. 1 131.291 68 <0.001 −0.025 0.005
Invariance Model 2B 470.891 328 <0.001 0.945 0.033 2B vs. 1 114.807 64 0.000 −0.021 0.016
Invariance Model 2C 440.588 322 <0.001 0.954 0.030 2C vs. 1 86.199 58 0.010 −0.012 0.001
Invariance Model 2D 417.300 314 <0.001 0.960 0.029 2D vs. 1 64.553 50 0.081 −0.006 0.000
Invariance Model 3 No convergence
Samples 1, 2, and 3 (Model 5a, item 9 removed)
Invariance Model 1 472.511 348 <0.001 0.959 0.032
Invariance Model 2A 685.999 476 <0.001 0.930 0.035 2A vs. 1 206.723 128 <0.001 −0.029 0.003
Invariance Model 2B The first-order derivative product matrix, as well as the latent variable covariance matrix for Sample 3 was not positive definite

Samples 1 and 2 (Model 5a): Invariance Model 1 = configural invariance model; Invariance Model 2A = metric invariance model; Invariance Model 2B = partial metric invariance model with the factor loading of item 11 on the Effortful Engagement factor freely estimated in both groups; Invariance Model 2C = partial metric invariance model with the factor loadings of items 11 and 6 on the Effortful Engagement factor freely estimated in both groups; Invariance Model 2D = partial metric invariance model with the factor loadings of items 11 and 6 on the Effortful Engagement factor and item 8 on the Sense of Purpose factor freely estimated in both groups; Invariance Model 3 = partial scalar invariance model; Samples 1, 2, and 3 (Model 5a, item 9 removed): Invariance Model 1 = configural invariance model; Invariance Model 2A = metric invariance model; Model 2B = partial metric invariance model with the factor loading of item 3 on the Effortful Engagement factor freely estimated in all groups; χ2, Chi square; df, degrees of freedom; p, probability value; CFI, comparative fit index; TLI, Tucker–Lewis index; RMSEA, root mean square error of approximation; 90% CI, 90% confidence interval of the RMSEA; SRMR, standardised root mean square residual.