Poster #24 - Extracting Trait Variance From Multiple Informant Sources in Applications of Multilevel Structural Models

• In Event: 79.052 - MTCC Poster Session 19
  In Poster Session: 79.052-2 - Educational Measurement, Psychometrics, and Assessment Poster Session

Abstract

The use of multiple informants is considered best practice in assessments that rely on rating based systems of measurement. We illustrate a framework for separating trait and non-trait informant effects across ratings obtained from different informant types for purposes of extracting trait factors that can be integrated into broader structural models of substantive interest. The procedure is illustrated through a multilevel application, but can be adapted to situations in which non-clustered data structures are present. The modeling approach is described in the context of student (N = 60,441) and teacher (N = 11,442) reports of school climate that were obtained from 298 high schools. Wherein, the resulting trait factors are evaluated in relation to a school level achievement factor.

Authors

• Timothy R. Konold, University of Virginia

• Dewey G. Cornell, University of Virginia

• Yuane Jia, Rutgers University

• Marisa Malone, University of Virginia