330.626.11 Estimating the Effects of Mental Health Interventions in Non-Experimental Settings
- Mental Health
- Summer Inst. term
- 1 credit
- Academic Year:
- 2017 - 2018
- East Baltimore
Experience with linear and logistic regression.
Discusses the importance of the careful design of non-experimental studies, and the role of propensity scores in that design, with the main goal of providing practical guidance on the use of propensity scores in mental health research. Covers the primary ways of using propensity scores to adjust for confounders when estimating the effect of a particular “cause” or “intervention,” including weighting, sub classification, and matching. Examines issues such as how to specify and estimate the propensity score model, selecting covariates to include in the model, and diagnostics. Draws examples from school-based prevention research, drug abuse and dependence, and non-randomized treatment trials, among others. Primarily emphasizes non-experimental studies; however, also discusses applications to randomized trials.
- Learning Objectives:
- Identify the characteristics of well-designed non-experimental studies
- Explain the role of propensity scores in non-experimental studies
- Distinguish between different propensity score approaches.
- Diagnose whether a propensity score approach has succeeded in balancing the groups
- Methods of Assessment:
Class participation: 30%; Take home project: 70%
- Instructor Consent:
No consent required
- Special Comments:
Students are required to complete readings prior to the start of the course. Take-home project will be due on July 22, 2016. A portion of the class participation grade will be based on responses to the reading done prior to class.