Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Date of Award
Master of Arts (MA)
Department of Graduate Psychology
Monica K. Erbacher
Although change scores are used in a variety of statistical methods (e.g., analysis of variance and regression), there is a lack of application of latent variable modeling methods to change scores. This thesis provides a detailed description of two latent variable modeling methods applied to change scores: factor analysis of change scores and change score factor mixture modeling. To illustrate advantages of these methods, both were applied to change score data from undergraduates. Students responded to sense of identity items during a university-wide assessment day on two occasions, once as incoming freshmen and again as second-semester sophomores. Change scores were computed by subtracting sophomore item responses from freshmen item responses. Factor analysis results indicated sense of identity change scores were best represented by two factors, change in sense of self and purpose and development of morals and beliefs. Factor mixture modeling results suggested two latent classes underlying these factors. The classes differed in both factor means and factor variances, which implied two possible change patterns associated with development of sense of identity. One class contained students who were stable on the two change score factors (i.e. changed minimally on sense of self and purpose and morals and beliefs) and the other class contained students who were fluid on one of the two factors. Classes were somewhat replicated with a second, independent sample, in that two classes were detected, but class means and variances diverged from those in the first sample. Results across the two methods provided insightful information about change processes of sense of identity, particularly how development of sense of identity is not the same across students. The applied example highlights the advantages of applying these methods to change scores. Implications of the two methods are further discussed throughout the thesis.
Ong, Thai Q., "Examining latent change classes: An application of factor mixture modeling to change scores" (2016). Masters Theses. 117.
Available for download on Wednesday, April 18, 2018