Considering K-12 learners' use of Bayesian methods

Bayesian statistical methods have become increasingly widely-used by statisticians, scientists, and engineers, yet their use has remained within the purview of professionals—and not learners, especially those at the K-12 level. At the same time, decades of developmental research indicates that children learn about the world in ways that align with Bayesian models of cognition. There is, then, the potential for K-12 learners’ use of Bayesian methods, but this potential has not been widely recognized. In this article, we have two goals. First, we review prior research on Bayesian methods and learning, focusing on past research in developmental science, and—after finding limited past research—carrying out a systematic review of research in science education, statistics education, and the learning sciences. Second, we offer specific, targeted areas where we think that Bayesian methods could have an impact on learning sciences research. We discuss future directions and our early-stage collaborative research.