| Session Title: Master Teacher Series: Practical Issues and Step-by-Step Guidelines for Using Propensity Scores |
| Demonstration Session 208 to be held in Pacific C on Thursday, Nov 3, 8:00 AM to 9:30 AM |
| Sponsored by the Quantitative Methods: Theory and Design TIG |
| Presenter(s): |
| M H Clark, University of Central Florida, mhclark@mail.ucf.edu |
| Vajeera Dorabawila, New York State Office of Children and Family Services, vajeera.dorabawila@ocfs.state.ny.us |
| Abstract: Quasi-experiments are excellent alternatives to true experiments when random assignment is not feasible. Unfortunately, causal conclusions cannot easily be made from results that are potentially biased. Some advances in statistics that attempt to reduce selection bias in quasi-experiments use propensity scores, the predicted probability that units will be in a particular treatment group. Because propensity score research is still relatively new, many applied social researchers are not familiar with the methods, applications and conditions under which propensity scores should be used. Therefore, the proposed demonstration will present an introduction to computing and applying propensity scores using SPSS and STATA. The demonstration will include: 1. introduction of propensity score adjustment methods; 2. how propensity scores can be used to make statistical adjustments using matching, stratifying, weighting and covariate adjustment; 3. examples using stratification and matching 4. a discussion of known limitations and problems when using propensity score adjustments. |