How do Propensity Score Methods Measure up In the Presence of Measurement Error?

Authors

Patricia Rodríguez de Gil, Univeristy of South Florida
Aarti P. Bellara, Univeristy of South Florida
Rheta E. Lanehart, Univeristy of South Florida
Reginald S. Lee, University of South Florida
Eun Sook Kim, University of South FloridaFollow
Jeffrey D. Kromrey, Univeristy of South Florida

Document Type

Article

Publication Date

Fall 10-2015

Keywords

propensity score methods, measurement error, Monte Carlo simulation

Digital Object Identifier (DOI)

http://doi.org/10.1080/00273171.2015.1022643

Abstract

Considering that the absence of measurement error in research is a rare phenomenon and its effects can be dramatic, we examine the impact of measurement error on propensity score (PS) analysis used to minimize selection bias in behavioral and social nonexperimental studies. A Monte Carlo study was conducted to explore the effects of measurement error on treatment effect and balance estimates in PS analysis across seven different PS conditioning methods. In general, the results indicate that even low levels of measurement error in the covariates lead to substantial statistical bias in estimates of treatment effects and concomitant reduction in confidence interval coverage across all methods of conditioning on the PS.

Was this content written or created while at USF?

Yes

Citation / Publisher Attribution

Multivariate Behavioral Research, v. 50, issue 5, p. 520-532.

Scholar Commons Citation

Rodriguez de Gil, P., Bellara, A. P., Lanehart, R. E., Lee, R. S., Kim, E. S., & Kromrey, J. D. (2015). How do propensity score methods measure up in the presence of measurement error: A Monte Carlo study. Multivariate Behavioral Research, 50(5), 520-532.