Date of Award

1-1-2012

Document Type

Campus Access Dissertation

Department

Educational Studies

Sub-Department

Educational Psychology / Research

First Advisor

Michael A Seaman

Abstract

Propensity score matching is a relatively new technique used in observational studies to approximate data that have been randomly assigned to treatment. This technique assimilates the values of several covariates into a single propensity score that is used as a matching variable to create similar groups.

This dissertation comprises two separate but related studies. The first is a simulation wherein generated data are analyzed to determine if one of three propensity score matching techniques performs more effectively in reducing sample bias and if those methods more accurately detect a simulated effect size than does a traditional analysis of covariance (ANCOVA).

Share

COinS