Instructor: Roderick Rose, Ph.D.
Research Assistant Professor
University of North Carolina at Chapel Hill
Social workers interventions should be evidence-based, meaning the effectiveness of the intervention to cause the desired change is supported by scientific evidence. The gold standard is to use randomized control trials. But what if randomization is not possible, ethical, or even desirable? Can we make plausible claims of causal relationships from observed statistical associations?
This workshop critically examines the three current major frameworks (Campbell, Rubin, & Pearl) for causal inference that provide the language, assumptions, and tools for causal inference; and the six major quasi-experimental methods that promote causal inference: instrumental variable estimation, regression discontinuity, propensity score analysis, fixed effects, difference-in-difference, and interrupted time series.
This workshop addresses an important need in research conducted in social work by helping researchers to answer the question, “what method should I use to evaluate or study an intervention given a set of prevailing study conditions?”
By the end of the workshop, participants will be conversant in the frameworks and will understand the strengths and weaknesses of each method, and obtain limited hands-on experience with each. Participants will be able to apply this knowledge in a practical sense in research and evaluation design, analysis, grant writing, and reviews.
Session 1: February 1, 2016 – April 30, 2016
Session 2: May 1, 2016 – July 31, 2016
Session 3: September 1, 2016 – November 30, 2016
To register for this online course, click here.