Logarithmic Level 1 Models Testing Group and Group by Time Interactions for All Groups Using 5 Time Periods
| Model Statistic | Estimates |
|---|---|
| Overall model fit | |
| Fit statistic (−2RLL) | 290.92 |
| AIC statistic | 320.92 |
| Repeated measures error structure | |
| Autoregressive variance estimate | 0.04 (SE = 0.002, p < .001) |
| Autoregressive covariance estimate | 0.64 (SE = 0.24, p = .007) |
| Random coefficients variance matrix | |
| Intercept random variance | 0.09 (SE = 0.01, p < .001) |
| Intercept-change random covariance | −0.002 (SE = 0.003, p = .51) |
| Change parameter random variance | 0.01 (SE = 0.002, p < .001) |
| Random Coefficients (Level 1 Parameters) With AS MI Coding | |
|---|---|
| AS MI intercept estimate | 0.12 (SE = 0.04, p = .007) |
| AS MI change parameter estimate | −0.07 (SE = 0.02, p < .001) |
| Comparison of each group from AS MI Intercept | |
| AS NMI | −0.29 (SE = .08, p < .001) |
| SN MI | 0.10 (SE = .06, p = .09) |
| GP NMI | −0.48 (SE = .07, p < .001) |
| GP MI | −0.08 (SE = .08, p = .26) |
| Comparisons of AS MI with each group on difference in change parameter | |
| AS NMI | 0.01 (SE = 0.02, p = .60) |
| SN MI | 0.02 (SE = 0.02, p = .34) |
| GP NMI | 0.002 (SE = 0.03, p = .94) |
| GP MI | 0.02 (SE = 0.03, p = .45) |
| Random Coefficients (Level 1 Parameters) With AS NMI Coding | |
|---|---|
| AS NMI intercept estimate | −0.16 (SE = 0.04, p < .001) |
| AS NMI change parameter estimate | −0.06 (SE = 0.02, p = .001) |
| Comparison of each group from AS NMI intercept | |
| AS MI | 0.29 (SE = 0.06, p < .001) |
| SN MI | 0.39 (SE = 0.06, p < .001) |
| GP NMI | −0.20 (SE = 0.07, p = .005) |
| GP MI | 0.20 (SE = 0.08, p = .008) |
| Comparisons of AS NMI with each group on difference in change parameter | |
| AS NMI | −0.01 (SE = 0.02, p = .60) |
| SN MI | 0.01 (SE = 0.02, p = .68) |
| GP NMI | −0.01 (SE = 0.03, p = .69) |
| GP MI | 0.01 (SE = 0.03, p = .75) |
Estimating logarithmic model with five time points using all groups. Models are coded in two ways so that each AS group is compared with all other groups on intercept and change parameters. Model fit statistics and covariance structures are the same for each model. Only the random and fixed parameter estimates change for the two models.