Comparison of fit statistics for longitudinal models of self-rated health: National Longitudinal Study of Adolescent to Adult Health, Waves I–V (1994–2019)
| Model . | χ2 . | df . | BIC . | CFI . | TLI . | 1-RMSEA . | Notes . |
|---|---|---|---|---|---|---|---|
| Latent Time-Invariant With Freely Estimated Intercepts | 2,014.645 | 15 | 1,873.385 | 0.928 | 0.899 | 0.896 | Negative error variance for subjective health LV at Wave II |
| Autoregressive (AR) | 57.817 | 12 | −55.191 | 0.998 | 0.997 | 0.982 | — |
| Linear Growth Curve | 274.902 | 16 | 124.224 | 0.991 | 0.988 | 0.964 | — |
| Quadratic Growth Curve | 48.552 | 12 | −64.456 | 0.999 | 0.998 | 0.984 | Negative error variance for subjective health LV at Wave I |
| Freed Loading Growth Curve | 46.457 | 13 | −75.969 | 0.999 | 0.998 | 0.986 | Negative error variance for subjective health LV at Wave II |
| Model . | χ2 . | df . | BIC . | CFI . | TLI . | 1-RMSEA . | Notes . |
|---|---|---|---|---|---|---|---|
| Latent Time-Invariant With Freely Estimated Intercepts | 2,014.645 | 15 | 1,873.385 | 0.928 | 0.899 | 0.896 | Negative error variance for subjective health LV at Wave II |
| Autoregressive (AR) | 57.817 | 12 | −55.191 | 0.998 | 0.997 | 0.982 | — |
| Linear Growth Curve | 274.902 | 16 | 124.224 | 0.991 | 0.988 | 0.964 | — |
| Quadratic Growth Curve | 48.552 | 12 | −64.456 | 0.999 | 0.998 | 0.984 | Negative error variance for subjective health LV at Wave I |
| Freed Loading Growth Curve | 46.457 | 13 | −75.969 | 0.999 | 0.998 | 0.986 | Negative error variance for subjective health LV at Wave II |
Notes: N = 12,300. Results are based on the DWLS estimator with its missing data procedures, as described in Asparouhov and Muthén (2010). All models control for age at the first wave, and observed SRH variances are constrained to be equal.