ASA Connect

A gentle reminder that the Boston Chapter of the American Statistical Association (BCASA) is hosting a traveling course

Guidelines for Using State-of-the-Art Methods to Estimate Propensity Score and Inverse Probability of Treatment Weights When Drawing Causal Inferences

Date: Saturday, November 16, 2019 (All day event)

Location: Cargill 097 (CG 097), Northeastern University, 45 Forsyth St, Boston, MA 02115

Map: https://goo.gl/maps/DxUdySJaJctyyNG58

Instructor: Beth Ann Griffin, RAND Corporation

Cost: Academic/nonprofit, BCASA Member, $30; Academic/nonprofit, Non-member $60; Industry, BCASA member $135; Industry, Non-member $165; 
Late registration (starting Nov 7) will incur an additional charge of $30. 
You can join the American Statistical Association and its Boston Chapter at this link https://www.amstat.org/ASA/Membership/Become-a-Member.aspx

Registration: https://bcasa2019propensity.eventbrite.com Registration includes course materials, light refreshments, and lunch.

Abstract:
Estimation of causal effects is a primary activity of many studies. Examples include testing whether a substance abuse treatment program is effective, whether an intervention improves the quality of mental health care, or whether incentives improve retention of military service members. Controlled, random-assignment experiments are the gold standard for estimating such effects. However, experiments are often infeasible, forcing analysts to rely on observational data in which treatment assignments are out of the control of the researchers.

This short course will provide an introduction to causal modeling using the potential outcomes framework and the use of propensity scores and weighting (i.e., propensity score or inverse probability of treatment weights) to estimate causal effects from observational data. It will also present step-by-step guidelines on how to estimate and perform diagnostic checks of the estimated weights for testing the relative effectiveness of two or more interventions. Attendees will gain hands-on experience estimating propensity score weights using boosted models in R, SAS and Stata; evaluating the quality of those weights; and using them to estimate intervention effects.

Additional topics (if time allows) can also include methods for conducting sensitivity analyses for unobserved confounding and estimation of the effects of time-varying treatments. Attendees should be familiar with linear and logistic regression; no knowledge of propensity scores is expected.

About the instructor:
Beth Ann Griffin is a senior statistician at the RAND Corporation Inference and is a member of the Pardee RAND Graduate School faculty. Her statistical research focuses on causal effects estimation when using observational data. Her substantive research has primarily fallen into three areas: (1) substance abuse treatment for adolescents, (2) military health, and (3) gun policy research. She is currently the principal investigator of two projects sponsored by the National Institute of Drug Abuse (NIDA), one which is focused on improving a promising health services research tool (the TWANG package) for estimating causal effects of treatment using propensity score weights (www.rand.org/statistics/twang) and the other which aims to develop well-operationalized, empirically-supported sequences of decision rules- known as "Adaptive Interventions" (AIs)-to provide guidance about substance-use services decisions for adolescent clients. Griffin also serves on the editorial board of the Annals of Applied Statistics, Statistics in Medicine and Observational Studies. She received her Ph.D. in biostatistics from Harvard University.