This response to Hoffman and Johnson’s comments on the authors’ research on multilevel models of the relationship of individual and school variables to school disorder concludes that Hoffman and Johnson present a rather narrow view of multilevel modeling that does not address important questions regarding theory and study design.
The authors also argue that errors and omissions further weaken Hoffman and Johnson’s arguments. Hoffman and Johnson argue that Welsh, Greene, and Jenkins ignore cross-level interactions and overstate the primacy of individual-level explanatory variables. However, their data on 1,000 schools seem to support the previous conclusions about the primacy of individual-level predictors. Hoffman and Jenkins identify some potential issues to consider in multilevel studies, but they do little to support, clarify, or address them. Their call to examine cross-level interactions and include larger samples of schools is legitimate but not novel. They can not credibly claim to have conducted a more methodologically rigorous or theoretically consistent study in any meaningful sense. Footnotes and 17 references
Downloads
Related Datasets
Similar Publications
- Visualization of partial bloody fingerprints on nonporous substrates using columnar thin films
- A Longitudinal Analysis of Risk and Protective Factors of Bias-Based Bullying Victimization Among Adolescents
- Persisting Concerns About Image Exposure Among Survivors of Image-Based Sexual Exploitation and Abuse in Childhood