Updated pre-print: Toward a Bayesian perspective for science learners

Joshua Rosenberg


With Marcus Kubsch, E.-J. Wagenmakers, and Mine Dogucu, I’ve updated our pre-print that argues that a Bayesian perspective can be used by science teachers and learners, especially when science learners are engaged in reasoning / arguing / analyzing and interpreting data in light of uncertainty.

The pre-print–Why and How a Bayesian Approach Supports Science Educators and Learners to Reason Under Scientific Uncertainty–is here: https://osf.io/aznyq/

In this version, we focused on steps to make a Bayesian approach practical. Part of this is heuristics that can guide learners’ involvement in several science practices:

We also created a Shiny app / widget that uses Bayes’ rule under the hood to transform prior information and new evidence into an estimate of how likely a belief is. That is available here: https://kubsch.shinyapps.io/Confidence_Updater/. It can - when worked/thought through - be thought of as a support similar to how the Claim, Evidence, Reasoning (CER) framework is commonly used in science teaching.

We welcome feedback on this work.