Teaching of Confidence Intervals in Context

ABSTRACT We present a taxonomy of conceptual steps in learning about confidence intervals (CIs), and outline pedagogical approaches used in teaching the topic, as a guide for tertiary‐level educators. We describe a number of documented misconceptions of CIs relating to the interpretation of the confidence coefficient and of the interval itself. Educators should be aware of these, and aim to confront and prevent the misconceptions in their teaching. In our study of 150 tertiary students' commentary on a misunderstanding of a CI for a population mean, we found that a large fraction of students had difficulty with properly separating description and inference, when presented with a strongly contextualised example, devoid of formulae and mathematical framing. We suggest possible sources for this confusion and offer strategies that can be used in teaching to address it. This includes, for example, contrasting a CI with a prediction interval.