Skip to main content
. 2023 Jul 10;14:962651. doi: 10.3389/fpsyg.2023.962651

Table 3.

Fit indices from different models estimated.

Model fit indices
(df) χ2 p CFI RMSEA BIC AIC
Model 1: single factor (286) 1528.39 0.00 0.878 0.080 39,162 37,768
Model 2: random intercept (311) 1810.67 0.00 0.853 0.084 39,313 38,031
Model 3: timepoint-specific factors (285) 1318.64 0.00 0.899 0.073 38,967 37,568
Model 4: timepoint-specific random intercepts (309) 1617.36 0.00 0.872 0.079 39,136 37,845
Model 5: construct-specific factors (283) 1430.98 0.00 0.887 0.077 39,088 37,680
Model 6: construct- and timepoint-specific factors (271) 1204.92 0.00 0.908 0.071 38,947 37,484
Model 7: random intercept and timepoint-specific factors (284) 946.04 0.00 0.935 0.058 38,616 37,212
Model 8: random intercept and construct-specific factors No convergence
Model 9: random intercept with construct- and timepoint-specific factors No convergence
Post-hoc model specifications
Model 7a: with loadings and residual correlations set to 0 if non-significant (293) 955.11 0.00 0.935 0.057 38,566 37,203
Model 7b: metric invariance on the timepoint-specific factors (301) 959.55 0.00 0.935 0.057 38,522 37,195
Model 7c: metric invariance on the timepoint-specific factors and residual invariance (314) 990.01 0.00 0.934 0.056 38,485 37,217

Bolded represents the best fitting value of the 9 hypothesized models. a,b,crepresent post-hoc models based on the results of a previously hypothesized model and is not bolded for that reason.