Table 3 Impact of effect distribution. Based on 1000 simulated datasets with sample size of 500 and effect size of 0.4 for the logit link and effect size of 0.3 for the probit link. Impact of effect distribution on model performance in evaluating the total effect of the repeated independent variable on the repeated outcome.
From: Non-linear mixed models in the analysis of mediated longitudinal data with binary outcomes
Simulated Data | SEM | Scaled NLMM | ||||
|---|---|---|---|---|---|---|
Effect Distribution | Bias (%) | Coverage (%) | Power (%) | Bias (%) | Coverage (%) | Power (%) |
Logit Link Results | ||||||
Equal | -0.6 | 96 | 89 | -1.3 | 95 | 89 |
Direct | -0.6 | 95 | 91 | -1.0 | 95 | 91 |
Indirect | -0.2 | 95 | 90 | -1.4 | 95 | 89 |
Probit Link Results | ||||||
Equal | 8.8 | 95 | 79 | 0.2 | 94 | 88 |
Direct | 5.9 | 96 | 80 | -0.9 | 95 | 90 |
Indirect | 8.6 | 95 | 76 | 0.09 | 94 | 88 |