This article contrasts two analytical methods for making cross-level inferences between individual-level factors and group-level factors in school-effectiveness research, the aptitude by treatment interaction design and hierarchical linear modeling. Although the aptitude by treatment interaction method is suitable for making cross-level inferences when the intraclass correlations are low, partitioning the interaction into within-and between-contexts components is recommended to discern if the interaction is due to confounding contextual effects.
Sheehan, Janet K. and Han, Tianqi
"Hierarchical Modeling Techniques to Analyze Contextual Effects: What Happened to the Aptitude by Treatment Design?,"
Mid-Western Educational Researcher: Vol. 9:
4, Article 3.
Available at: https://scholarworks.bgsu.edu/mwer/vol9/iss4/3