BackHome
Refresh: OffView Personal Schedule
Session Submission Summary

Problems With Interpretations of Multilevel Data: Extending Research Beyond Hierarchical Linear Modeling

Fri, April 4, 8:15 to 9:45am, Convention Center, Floor: 100 Level, 117

Session Type: Symposium

Abstract

Although Cronbach (1976) argued for the interpretation of construct meaning at the individual and cluster levels; to date, such approaches have seldom been implemented in applied contexts. We address this methodological shortcoming in four studies which use real data to demonstrate alternative multilevel data analytic approaches to derive more accurate and meaningful results from data analysis in the context of data nesting. We also extend research conducted using Hierarchical Linear Modeling (Raudenbush, & Bryk, 2002), which has been the primary approach to modeling nested data, to less frequently but similarly important, multilevel factor analysis and multilevel latent class models. We discuss implications of using these approaches on policy interventions and validity of inferences derived from survey data and test scores.

Sub Unit

Chair

Papers

Discussant