Multilevel mediation modeling in group-based intervention studies.
The writers present and assess a method to test for mediation in multilevel data sets that is formed when an intervention administered to intact groups is aimed at producing change in individual mediator and outcome variables. They employ simulated data of this form to conduct a comparison of ordin...
| Publicado en: | Evaluation Review Vol. 23; no. 4; pp. 418 - 445 |
|---|---|
| Autores principales: | , |
| Formato: | Artículo |
| Publicado: |
Sage Publications Inc.
August 1999
|
| Materias: | |
| Acceso en línea: | Ver este registro en EBSCOhost |
| Sumario: | The writers present and assess a method to test for mediation in multilevel data sets that is formed when an intervention administered to intact groups is aimed at producing change in individual mediator and outcome variables. They employ simulated data of this form to conduct a comparison of ordinary least squares (OLS) and two multilevel estimators of the mediated effect. They evaluate approximations of OLS and multilevel standard errors and make recommendations for optimal estimator choice, and they apply these methods to data from an existing substance use intervention. Their aim in doing so is to demonstrate the impact that multilevel mediation modeling can have on the conclusions drawn from real-world evaluation studies. |
|---|