Skip to main content
. 2017 May 23;8:711. doi: 10.3389/fpsyg.2017.00711

Table 5.

Analytic strategy: goodness of fit statistics for each model.

Model χ2 df p CFI TLI RMSEA [90%]
Model 1: Stability 90.220 36 <0.001 0.982 0.961 0.038 [0.028, 0.048]
Model 2: Bidirectional 47.238 35 0.081 0.996 0.991 0.018 [0.000, 0.030]
Model 3: Add early effects 35.932 33 0.333 0.999 0.998 0.009 [0.000, 0.025]
Model 4: Add adjustment 200.383 83 <0.001 0.969 0.937 0.037 [0.030, 0.043]
Model 5: Add covariates 270.749 145 <0.001 0.972 0.946 0.029 [0.023, 0.034]
Model 6: Gender differences 337.653 182 <0.001 0.958 0.921 0.028 [0.024, 0.033]
Model 7: Add covariates 498.163 324 <0.001 0.960 0.930 0.023 [0.019, 0.026]

Models 1 through 5 are nested within each other and represent the entire sample (n = 1,057). Models 6 and 7 tested a multiple group model for boys (n = 528) and girls (n = 529) using the same components as Models 4 and 5, respectively.