From: Determinants of access to the SARS-CoV-2 vaccine: a preliminary approach
Model adjustment criteria (index) | Definition | Acceptance level | Explication |
---|---|---|---|
Chi-square (CMIN) and Chi-square per degrees of freedom (CMIN/df) | Tests the H0 that the residual covariance estimate is equal to a matrix composed only by zeros. It considers the degrees of freedom (df) of χ2. The chi-square value is divided by the number of degrees of freedom to obtain an adjustment value for the model less sensitive to the sample size. | Table value χ2; values below 5 are tolerable. | Comparison of values obtained χ2 with the tabulated value given by df. |
Root square mean error of approximation (RMSEA) | Identify how well a model fits the population, not just the sample used for the estimation. | Ideal values up to 0.08; below 0.10 are tolerable | A value between 0.05 and 0.08 indicates a good approximation |
Normed Fit Index (NFI) | Compare the chi-square for the tested model against the chi-square for the baseline model assuming that the measured variables are completely independent. | 0 (has no adjustment) and 1 (perfect fit) | A value close to 0.90 or 0.95 reflects a good adjustment |
Comparative Fit Index (CFI) | Comparative adjustment index is considered to be more appropriate for the evaluation of models generated from samples considered small | It shows whether and to what extent the quality of adjustment of the proposed model is better than that of the base model. | A value close to 0.90 or 0.95 reflects a good adjustment |