Three-Level Models With Small Teacher-Level Sample Sizes

• In Event: Methodological Issues and Applications in Multilevel Modeling

Abstract

When interventions are implemented by teachers, researchers need to account for the nested structure in schools (i.e., students are nested within teachers nested within schools). Schools usually have a very limited number of teachers in each grade, which might pose potential challenges for study design and reporting. Through a simulation study, we show that the intraclass correlation coefficients (ICCs) at the teacher and school levels are prone to biased estimates in most experiments. However, the total variance estimation is largely unbiased. So, we recommend that researchers use the total variance, rather than school- level variance alone, to report standardized treatment effects for meta-analysis. We also discuss the implication for reporting empirical design parameters to design experiments involving teachers.

Authors

• Huibin Zhang, University of Tennessee
  Presenting Author

• Zuchao Shen, University of Georgia
  Presenting Author

• Walter L. Leite, University of Florida
  Presenting Author