Table 3.
Fit indices for executive function CFA models
Fit Indices |
Comparison with 3 factor model |
|||||||||
---|---|---|---|---|---|---|---|---|---|---|
Models | χ2 | df | p-value | RMSEA (95% CI) | CFI | NNFI | AIC | Δχ2 | Δdf | p-value |
Three factors: Working Memory, Inhibition, Shifting | 1.09 | 1 | 0.2963 | 0.024 (0, 0.217) | 0.999 | 0.992 | 3973.381 | |||
Two factor models: | ||||||||||
Collapsing Working Memory & Inhibition | 1.58 | 3 | 0.6637 | 0 (0, 0.106) | 1.000 | 1.044 | 3971.519 | 0.4 9 | 2 | 0.7827 |
Collapsing Working Memory & Shifting | 19.37 | 3 | 0.0002 | 0.188 (0.114, 0.272) | 0.747 | 0.493 | 3987.660 | 18.279 | 2 | 0.0001 |
Collapsing Inhibition & Shifting | 24.25 | 3 | <0.0001 | 0.214 (0.141, 0.297) | 0.671 | 0.342 | 3992.540 | 23.159 | 2 | 0.0000 |
One factor: collapsing all three EF factors | 25.95 | 4 | <0.0001 | 0.189 (0.124, 0.261) | 0.660 | 0.490 | 3992.241 | 24.86 | 3 | 0.0000 |
Note. RMSEA = Root Mean Square Error of Approximation, CFI = Comparative Fit Index, NNFI = Non-Normed Fit Index/Tucker Lewis Index, AIC = Akaike Information Criterion. Accepted model is in bold.