home - login - register

PERC 2016 Abstract Detail Page

Previous Page  |  New Search  |  Browse All

Abstract Title: Using structural equation modeling to test the physics identity framework
Abstract: Fully modeling the complex relationships between multiple factors in a theoretical framework requires connecting several statistical techniques. Structural equation modeling (SEM) combines confirmatory factor analysis (CFA), path analysis, and multiple regression. Rather than providing a form of exploratory data analysis, SEM is a confirmatory technique that involves testing and modifying a hypothesized model. Specific tests indicate ways to improve the model, and a variety of fit indices are used to assess the model fit. The implementation of SEM will be shown through the example of testing the physics identity theoretical framework. This includes performing the initial CFA to verify the identity dimensions and assessing model fit. Additionally, the method of modeling multiple groups will be demonstrated by comparing the physics identity model for males and for females. Examples of implementation will be shown in R. Furthermore, the types of research questions to which SEM is most applicable will be described.
Abstract Type: Symposium Poster
Parallel Session: Expanding Research Questions by Expanding Quantitative Methodologies

Author/Organizer Information

Primary Contact: Robynne M. Lock
Texas A&M University - Commerce
and Co-Presenter(s)
Zahra Hazari, Geoff Potvin, and Jennifer Cribbs