TABLE 3.
Fit indices for the 4 models with and without invariance test.
| Npar | χ2 | dl | CFI | NNFI | RMSEA | [CI] | |
| Model with intrinsic motivation as a mediator | |||||||
| Model for the whole population | 82 | 84.16* | 38 | 0.98 | 0.96 | 0.02 | [0.01, 0.03] |
| 1-Configural model (sex) | 142 | 78.73* | 66 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 2-Saturation (S) between sexes | 138 | 81.30* | 70 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 3-(S) + saturation of identical items over time (ST) | 136 | 82.84* | 72 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 4-(S) + (ST) + intercepts (I) | 130 | 94.03* | 78 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 5-(S) + (ST) + (I) + residual errors (U) | 124 | 192.03* | 84 | 0.96 | 0.93 | 0.03 | [0.03, 0.04] |
| 5a-(S) + (ST) + (I) + residual errors (U) of item 2 and 3 are relaxed | 128 | 103.87* | 80 | 0.99 | 0.98 | 0.016 | [0.01, 0.03] |
| 6-(S) + (ST) + (I) + U + correlation of u of identical items over time (CU) | 125 | 105.16* | 83 | 0.99 | 0.99 | 0.016 | [0.00, 0.02] |
| 7-(S) + (ST) + (I) + U + (CU) + Var-cov (CV) | 98 | 158.81* | 110 | 0.98 | 0.97 | 0.02 | [0.01, 0.03] |
| 8-(S) + (ST) + (I) + U + (CU) + (CV) + Path | 89 | 202.03* | 119 | 0.97 | 0.96 | 0.03 | [0.02, 0.03] |
| Model with inattention as mediator | |||||||
| Model with the whole population | 82 | 50.10* | 37 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 1-Configural model (sex) | 142 | 77.05* | 66 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 2-Saturation (S) between sexes | 138 | 84.81* | 70 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 3-(S) + saturation of identical items over time (ST) | 136 | 85.82* | 72 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 4-(S) + (ST) + intercepts (I) | 130 | 90.09* | 78 | 0.99 | 0.99 | 0.01 | [0.00, 0.02] |
| 5- (S) + (ST) + (I) + residual errors (U) | 124 | 118.47* | 84 | 0.99 | 0.99 | 0.02 | [0.01, 0.03] |
| 6-(S) + (ST) + (I) + U + correlation of u of identical items over time (CU) | 121 | 118.94* | 87 | 0.99 | 0.99 | 0.03 | [0.01, 0.03] |
| 7-(S) + (ST) + (I) + U + (CU) + Var-cov (CV) | 94 | 176.76* | 114 | 0.99 | 0.99 | 0.03 | [0.02, 0.03] |
| 8-(S) + (ST) + (I) + U + (CU) + (CV) + Path | 85 | 234.15* | 123 | 0.98 | 0.98 | 0.03 | [0.02, 0.03] |
| Model with frequency of reading as mediator | |||||||
| Model with the whole population | 82 | 67.49* | 37 | 0.98 | 0.96 | 0.02 | [0.01, 0.03] |
| 1-Configural model (sex) | 142 | 104.27* | 66 | 0.98 | 0.96 | 0.02 | [0.01, 0.03] |
| 2-Saturation (S) between sexes | 138 | 112.20* | 70 | 0.98 | 0.95 | 0.02 | [0.02, 0.03] |
| 3-(S) + saturation of identical items over time (ST) | 136 | 117.47* | 72 | 0.98 | 0.95 | 0.02 | [0.02, 0.03] |
| 4-(S) + (ST) + intercepts (I) | 130 | 121.77* | 78 | 0.98 | 0.96 | 0.02 | [0.02, 0.03] |
| 5- (S) + (ST) + (I) + residual errors (U) | 124 | 126.39* | 84 | 0.98 | 0.96 | 0.02 | [0.01, 0.03] |
| 6-(S) + (ST) + (I) + U + correlation of u of identical items over time (CU) | 121 | 128.31* | 87 | 0.98 | 0.96 | 0.02 | [0.01, 0.03] |
| 7-(S) + (ST) + (I) + U + (CU) + Var-cov (CV) | 94 | 167.44* | 114 | 0.98 | 0.96 | 0.02 | [0.01, 0.03] |
| 8-(S) + (ST) + (I) + U + (CU) + (CV) + Path | 85 | 172.79* | 123 | 0.97 | 0.96 | 0.02 | [0.01, 0.03] |
| Model with leisure reading as mediator | |||||||
| Model with the whole population | 65 | 0.00* | 0 | 1.00 | 1.00 | 0.00 | [0.00, 0.00] |
| 1-Configural model (sex) | 108 | 0.00* | 0 | 1.00 | 1.00 | 0.00 | [0.00, 0.00] |
| 2-Var-cov (CV) | 81 | 44.52* | 27 | 0.98 | 0.95 | 0.02 | [0.01, 0.04] |
| 3-(CV) + Path | 72 | 58.82* | 36 | 0.97 | 0.95 | 0.02 | [0.01, 0.04] |
Npar is the number of parameters estimated; dl is the degree of freedom; CFI is the “Comparative Fit Index”; TLI is the “Tucker-Lewis Index” and RMSEA is the “Root Mean Square Error of Approximation.” *Means that statistically significant at the 5% level.