When has a learner truly learned?

This episode explores how statistics for n-of-1 designs can help practitioners make day-to-day decisions in applied behavior analysis (ABA). Mark Ramos describes how a personal concern about his daughter’s early intervention led him to ask a simple but important question: when does observed performance (for example, 4 of 5 vs. 8 of 10 correct trials) justify concluding a child has truly mastered a skill? Using Bayesian ideas, he shows how increasing trial counts changes the posterior probability that latent mastery meets a chosen threshold, and how that can be integrated into routine ABA records. He also describes a prototype app to give practitioners updated, individualized probabilities and emphasizes choosing statistical tools to fit the context.