Doubly Latent Multilevel Structural Modeling: An Illustration of School Climate Influences on Student Outcomes

• In Event: Marriott Roundtable Session 41
  In Roundtable Session: Methodological Design and Illustrations

Abstract

Doubly latent multilevel structural equation models accommodate the limitations of traditional multilevel modeling frameworks by controlling for both measurement and sampling error, leading to bias-corrected estimates of indirect effect. We illustrate the usefulness of this approach in investigating the mediating role of student engagement in associations between teacher reports of school climate and student outcomes. The illustration is based on a statewide high school sample of 62,001 students from 315 schools and 11,112 teachers from 298 schools.

Authors

• Yuane Jia, Rutgers University

• Timothy R. Konold, University of Virginia

• Dewey G. Cornell, University of Virginia