Bayesian Estimation of the Autocorrelation in Single-Case Designs: A Multilevel Extension

• In Event: Poster Session 7
  In Poster Session: Division D Section 2 Poster Session

Abstract

Single case designs (SCDs), which study a single participant over time under different treatment conditions, are commonly used in clinical and educational research to assess intervention effects. When analyzed together, SCDs can provide invaluable causal inferences when randomized experiments are not feasible and have the potential to contribute to evidence-based practice. The proposed paper describes a method for incorporating an autoregressive parameter into a multilevel Bayesian intervention model. Potential advantages of using a Bayesian analysis with SCD data, as well as a full description of the model and data, discussion of the results, limitations, and considerations to be mindful of when conducting a Bayesian analysis with SCD data, will be presented in this paper.

Authors

• Jonathan G. Boyajian, University of California, Merced

• Sarah Depaoli, University of California - Merced

• William R. Shadish, University of California, Merced