Enabling Propensity Score Analysis in Survival Analysis

Background and Purpose   Propensity score analysis (PSA) and survival analysis (SA) are widely applied by social work researchers. PSA offers an approach to program evaluation when randomized controlled trials are infeasible, unethical, or when researchers need to draw causal inferences from survey, census, administrative, or other types of observational data. SA enables researchers to address questions that have to do with whether and when an event of interest takes place. Despite ongoing and increasing use of PSA and SA methods, few attempts have been made to combine PSA and SA into a single approach. In practice, researchers often face the problem of insufficient sample size after conducting a nearest-neighbor-within-caliper matching and therefore fail to run SA that controls for selectivity. This workshop addresses the problem of selection bias in non-experimental data by describing statistical and methodological tools that will enable social work researchers to conduct PSA in survival analysis and, as a result, take advantage of their complementary strengths.

Contents The workshop covers the following topics: (a) examine the theoretical and methodological foundations of PSA and SA, including methodological challenges and solutions when combining the two approaches; (b) demonstrate two integrated models that incorporate specific PSA approaches (i.e., inverse probability of treatment weights [IPTW] estimator and propensity score subclassification [PSS]) with Cox proportional hazard model – both examples employed the panel data of the National Survey of Child and Adolescent Well-Being (NSCAW) to examine the connection between the timing of child maltreatment rereport and caregivers’ use of substance abuse services; (c) demonstrate the application of the same IPTW and PSS estimators to a parametric survival model (i.e., a piecewise exponential model); and (d) discuss the importance of advancing knowledge about the adverse consequences of ignoring selection bias in observational studies and the need to promote rigor in causal inferences.

Pedagogical Techniques   Guided by the Neyman-Rubin counterfactual framework, this workshop discusses statistical principles by using a PowerPoint presentation and then illustrates application examples by running IPTW and PSS in conjunction with SA with Stata.

Significance   The workshop will demonstrate how to add analytic rigor to the applications of SA and PSA.  This will be done for SA by merging it with techniques of PSA that enable taking account of selectivity.  The proposed approach will advance scientific knowledge about the importance of the strongly ignorable treatment assignment assumption embedded in studies of causal inference, and will promote rigor in social work research when this assumption is violated.