The current paper demonstrates the usefulness of Bayesian estimation with small samples. In Bayesian estimation, prior information can be included. Prior information is information about the model parameters that originates from sources other than the data at hand. These sources can be literature, experts, or general knowledge. Including prior information increases the precision of the posterior distribution. The posterior distribution reflects what values are likely given the current state of knowledge, irrespective of the size of the current sample. Null hypothesis significance testing (NHST), on the other hand, suffers from low power with small samples, which often renders non-significant p-values that are difficult to interpret. An issue that received little attention in previous research, however, is the proper acquisition of prior information. The current study provides a set of general guidelines for collecting prior knowledge and formalizing it in prior distributions. Moreover, the current study also demonstrates how prior knowledge can be acquired systematically with an empirical application about development of working memory in young heavy cannabis users and non-using peers. To collect prior information, meta-analyses, reviews, empirical papers, experts, and general knowledge were involved. The paper closes with a discussion that also warns against the misuse of prior information.

Zondervan-Zwijnenburg, M., Peeters, M., Depaoli, S., & van de Schoot, R. (2017). Where do priors come from? Applying guidelines to construct informative priors in small sample research. Research in Human Development, 14:4, 305-320.

We report much additional information at the Open Science Framework.

PhD Student

In her PhD project, Mariëlle focuses on including prior knowledge in statistical analyses (informative Bayesian research) and confronting prior knowledge with new data.

Margot Peeters
Margot studies longitudinal analyzing techniques to study the development of adolescents in the most efficient way. Techniques include latent transition, zero inflated, survival analysis, missing data analysis and Bayesian statistics.
Visit website
Sarah Depaoli
Assistant Professor at the University of California, Merced
Sarah’s research interests are largely focused on issues surrounding Bayesian estimation of latent variable models. She has a particular interest in estimation issues arising from nonlinear growth patterns over time. She is also interested in improving accuracy of uncovering unobserved (latent) groups of individuals. She is currently working with several students that are involved in research spanning a wide range of methodological topics .
Visit website