Dear everyone,
In cross-classified (multilevel) models, the cross-classified variance components are generally assumed to be independent. Does anyone know of any (published or unpublished) discussion of this assumption or how deviations from independence may affect the results?
Best regards,
Are Skeie Hermansen
University of Oslo
In cross-classified (multilevel) models, the cross-classified variance components are generally assumed to be independent. Does anyone know of any (published or unpublished) discussion of this assumption or how deviations from independence may affect the results?
Best regards,
Are Skeie Hermansen
University of Oslo
