Date of Award
8-16-2024
Document Type
Open Access Dissertation
Department
Psychology
First Advisor
Christine DiStefano
Abstract
In structural equation modeling, relationships between exogenous and endogenous variables are often assumed to be linear though that is not a requirement. A recent trend in modeling latent interactions and moderations has stimulated further inquiry into nonlinear relationships and the analysis techniques associated with those types of models. However, rarely do investigators specify other potential nonlinear relationships such as quadratic relationships. This study explores the potential for quadratic relationships within Expectancy Value Theory when p the impact of high school students’ motivational beliefs on their pursuit of STEM-related activities. Using the High School Longitudinal Survey as a source of information, this investigation found evidence that there is a significant quadratic relationship between students’ science expectancy beliefs and their later STEM intentions. Additionally, results from this study provide insight to the dynamic nature of motivational beliefs as described in Situated Expectancy Value Theory. Finally, this study underscores the need for further study of the Latent Moderated Structures technique that is often used to assess nonlinear relationships to insure its appropriate and reliable use.
Rights
© 2024, Steven G Barth
Recommended Citation
Barth, S. G.(2024). Evaluating Nonlinear Relationships in Sem Models: An Example of Parameter Estimation and Stability Using Expectancy Value Theory. (Doctoral dissertation). Retrieved from https://scholarcommons.sc.edu/etd/7797