Teaching with Datasets that Recognize the Intersectional Perspectives of Students

Undergraduate student populations are becoming increasingly diverse, and certainly they represent more intersectional perspectives than the statisticians who are typically written about in undergraduate textbooks (Bayes, Fisher, Pearson, Gosset, …). Inspired by a wonderful classroom session sharing the work of David Blackwell (when discussing the Rao-Blackwell Theorem), I have searched for scholars from many different backgrounds whose work connects, in some way, to the undergraduate curriculum. The resulting database (at https://hardin47.github.io/CURV/) allows educators to bring in examples of statisticians whose work is connected to the undergraduate curriculum. I describe the database, how I've used it in class, and how others can contribute.  

Students working on interdisciplinary projects usually have enough knowledge to correctly apply statistical methodology. However, often they have not had much practice thinking carefully about the data and potential implications to inference due to population source, data collection methodology, and factors resulting from structural, social, and environmental determinants. When data is available, students commonly make the mistake of immediately embarking on conducting statistical tests or fitting statistical models. To translate statistical results into interdisciplinary conclusions, statisticians must take a step back to understand, respond to, and describe the broader context of the research study and data available. I will share our "3 Cs framework for teaching how to approach, implement, and present on research questions utilizing datasets from interdisciplinary collaborators" (Moore, Mehta) and highlight, as case studies, projects that I've utilized in my statistics consulting classes to engage students in thinking critically, including considerations of structural racism and health inequity, and how to contribute to the team based on their backgrounds and perspectives.  

In this talk, we explore two different datasets which were collected to study two common aspects of the college experience: group projects and final exam stress. Both datasets are accessible to students without requiring detailed subject-area knowledge, and both are closely related to student experiences in classes and on campus. In the first study, researchers investigated how group identities and diversity related to performance on projects in a large business course, while the second study assesses whether petting and interacting with animals can really help manage stress. Each study allows students to think carefully about data collection, identity, and ethics: How should "group diversity" be measured from demographic characteristics? What demographic information is appropriate to collect and share in a study, and how does data privacy conflict with reproducibility? Throughout the talk, we suggest discussion questions on the data and the original research papers, and give examples of how these data can be used to teach topics like paired-sample tests, interaction terms in linear regression, and multiple comparisons. 

A fundamental quest of statistics and data science is drawing robust conclusions about our world with incomplete data. This makes statistics and data science unique and applicable in almost every field of work or study. Teaching statistics and data science with real, diverse and relatable data reinforces the practicality and applicability of statistics. In fact, teaching statistics with real data with context and purpose (2016 GAISE report) is an effective teaching technique because data drives the motivation behind most statistical concepts. This technique enhances students' conceptual understanding of topics and helps them perceive the relevance of statistics and data science. As a consequence, we can capture students' attention and pique their interest in statistics and data science.
In this presentation, we will share examples of how to find, collect, or scrape real, diverse and relatable datasets to use in the introduction of fundamental topics in statistics and data science. Some of the real data used in this presentation are also relatable to students as they touch on issues relating to justice, equity, diversity, and inclusion. We will approach statistics and data science