By Todd Iverson (Winona State University)
Information
Personal experience has demonstrated that simulations are an effective tool for enhancing learning in mathematical statistics. However, students often face challenges when developing the code for more complex simulations. For instance, power simulations over a parameter space require generating and summarizing samples from both the null distribution as well as from a distribution for each parameter space, which can lead to cumbersome and difficult-to-manage code. In response, we introduce purrrfect, an extension of the purrr library from the tidyverse, designed to streamline the process of creating and summarizing the experimental tableaux. This extension enables students to capture parametric simulations in a single, cohesive table, facilitating clear connections between key statistical concepts, e.g., the sample, the statistic, and the parameter. Although purrrfect remains a work in progress, this presentation shares insights from using the platform over the last two years to introduce simulations to approximately twenty undergraduate statistics majors in an upper-division mathematical statistics course at Winona State—a non-selective, four-year residential institution—highlighting its potential to simplify and enhance the learning process.