Sarah Depaoli (Associate Professor, Quantitative Psychology) made significant contributions to Bayesian statistics in Psychology through a series of recently published papers.
Together with Jitske Tiemensma (Assistant Professor, Health Psychology), a paper was published in Health Psychology Review exploring the benefits of Bayesian methods in health-based inquiries that are typically confined to small sample sizes (e.g., in stress-based research). A comment about the importance of this work was published in the same issue.
Two papers were published in the special issue on Bayesian statistics in Psychological Methods. The first paper outlines methods for avoiding the misuse of Bayesian statistics and improving replicability within the field. The second paper presents an extensive systematic review of over 1500 papers published in the last 25 years using Bayesian methods in Psychology. This paper highlights some important growth patterns for the use of Bayesian methods, as well as discouraging trends of misuse. In this paper, the authors call for a reconfiguration of how Bayesian methods are taught and used within Psychology.
The use of priors (through Bayesian statistics) for finite mixture models was explored in a paper published in Structural Equation Modeling: A Multidisciplinary Journal. This paper highlights the importance of the prior placed on the mixture class parameter. Further guidelines for reporting mixture model results were published in the same journal in a subsequent paper. Finally, a paper was recently accepted into Research in Human Development examining the specification of priors using a systematic review and expert opinions for multi-group longitudinal models where one group is exceptionally small.
Of note, 5 of these papers included graduate student co-authors (one as a first author).
