Much of statistical reasoning involves combining ideas of data and chance to carry out inferential procedures and interpret results. Underlying this reasoning is a conceptual understanding of some important ideas, such as probability, variability, samples, sampling distributions, and the normal distribution. Although statistics courses teach this content, students are often unable to fully integrate and apply these ideas when reasoning about real statistical problems or when interpreting the results of research investigations. This project will develop and disseminate materials that help students learn the core concepts underlying statistical inference. Specifically, we will create instructional modules that use simulation software to help students develop abstract concepts such as sampling distribution, confidence interval, p-values, and power. Based on our previous research, we will build the modules on a model of conceptual change, where students first make predictions and then test them, using the simulation software. These modules will also include assessment instruments to assess prerequisite conceptual knowledge and instruments to assess understanding of the statistical concepts underlying statistical inference. All materials (including our simulation software programs) will be made available to statistics educators on the World Wide Web.
