There has been some research in the use of propensity scores in the context of measurement error in the confounding variables; one recommended method is to generate estimates of the mis-measured covariate using a latent variable model, and to use those estimates (i.e., factor scores) in place of the covariate. I describe a simulation study designed to examine the performance of this method in the context of differential measurement error and propose a method based on moderated nonlinear factor analysis (MNLFA) to try to address known problems with standard methods. Although MNLFA improves effect estimation somewhat in the presence of differential measurement error relative to standard factor analysis methods, the greatest gains come from the nonstandard practice of including the treatment variable as an indicator in the scoring models. More research is required on the effects of model misspecification on the performance of these methods for causal inference applications. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]