Introduction to Bayesian analysis for developmental researchers

Session Type: Professional Development Session

Abstract

Bayesian methods allow researchers to incorporate their knowledge of the domain into the analyses, providing a powerful tool for researchers to update past information with current data. Despite the effectiveness and wide application of Bayesian analyses, they are not commonly covered in many statistics courses. Bayesian methods are especially powerful for developmental research data, where data collection from children is costly and it is critical to use methods that let researchers make the most of the existing data. Further, as Bayesian analyses become more widespread, it is important to provide conceptual knowledge about these tools so that reviewers and consumers of journal articles are educated about their appropriate application.
Our workshop will be taught by Drs. Angeline Tsui, Michael Henry Tessler, and Michael C. Frank. We will offer theoretical and practical Bayesian statistical training in three modules. The three modules cover the fundamental concepts of Bayesian statistics and methods of doing Bayesan inference (i.e., regression and mixed-effect models). We will use examples and introduce the relevant R packages (e.g., brms). We’ll end with a discussion of how to design studies (e.g., sample size planning, stopping rules) under the Bayesian approach. By the end of the workshop, participants will have acquired sufficient knowledge to evaluate others’ Bayesian analyses and begin integrating this analytic approach into their own research.

Sub Unit

• Professional Development Sessions

Session Organizers

• Sin Mei Angeline Tsui, Stanford University

• Michael Henry Tessler, Massachusetts Institute of Technology

Participant

• Michael C. Frank, Stanford University