Abstract Details
Activity Number:
|
370
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
|
Sponsor:
|
ENAR
|
Abstract - #309594 |
Title:
|
Using Regression Discontinuity Designs to Enhance Power in Propensity Score Analysis
|
Author(s):
|
T. Mark Beasley*+
|
Companies:
|
Univ of Alabama of Birmingham
|
Keywords:
|
Propensity Scores ;
Regression Discontinuity ;
Statistical Power
|
Abstract:
|
When the probability of treatment assignment is highly associated to a set of covariates, there is little between-group overlap in the distribution of Propensity Scores (PS), resulting in fewer PS matches. This effectively reduces the size of the matched sample and lowers statistical power. Regression Discontinuity Designs (RDDs) were developed to assess treatment effects when treatment assignment is perfectly associated with a single variable X, with a threshold that determines treatment assignment. One modification to RDDs involves a triage system where subjects with extremely high values of X receive treatment, subjects extremely low values of X are not treated, and a middle group is assigned based on randomization, availability, or ethical considerations. We propose using a set of covariates instead of a single variable to define the likelihood (i.e., propensity) of receiving the treatment and combining this with the triage modification of RDDs that allows a middle range of PS overlap for matching. Through a series of simulation studies, we will investigate the potential bias and enhancement of statistical power by combining PS and RDD methods, in poor PS matching conditions.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2013 program
|
2013 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.