Debating the Next BIG Thing in Teaching Statistics
Allan Rossman & Beth Chance, California Polytechnic State University-San Luis Obispo

Download Presentation (PowerPoint)

We engage in a series of debates about the next BIG thing in teaching undergraduate statistics. Inspired by David Moore's maxim that "nothing tunes the neurons like disagreement," we aim to stimulate thought and foster discussion and perhaps even provoke further debates among conference participants throughout USCOTS and afterward. We will present (at least) two sides for each of the following propositions:

• The next BIG thing will be the disappearance of textbooks from our courses.
• The next BIG thing will be the elimination of face-to-face contact among students and instructors.
• The next BIG thing will be the dropping the letters z and t from the introductory course.
• The next BIG thing will be students entering undergraduate courses with considerable knowledge of statistics.
• The next BIG thing will be research-based decisions about curriculum and pedagogy.
• The next BIG thing will be a topic to be chosen by popular vote in the pre-conference survey.
• The next BIG thing will be ... we reserve the right to think of new ideas and replace these ones until the conference begins!

Citizen Statisticians: Modern Statistics for Modern Students
Rob Gould, UCLA

Hotelling asked many important questions about the state of statistics education in his 1948 Annals paper, and in doing so, created a dichotomy between producers and consumers of statistics that still stands in today's statistics curricula. But because statistics is the science of data, our efforts would be better spent thinking of students in terms of producers and consumers of data. This is an appropriate shift of focus, because today's students bring something new to the classroom: immersive experience with data. Modern technology makes it easy for all students to produce, process, display and collect data. The dichotomy disappears, and rather than educate producers or consumers, we should instead prepare citizen statisticians: people who will interact with data in both formal and informal settings, in planned and surprising fashions, and in professional and personal contexts, throughout their lives.

Personalized Education - My Thirty Year Search for the Next BIG Thing
Dennis Pearl, The Ohio State University

I will describe my thirty-year pursuit of the next BIG thing in statistics education from my days as a shipping and receiving clerk at a dried fruit packing house; to multiple redesigns of our introductory statistics class; to working on building a national infrastructure for statistics education; to my recent adventure being held up at the Canadian border for being a statistics professor. A common theme keeps re-emerging: to effectively reach all learners we must provide a way to personalize each teacher's pedagogical options and each student's educational experience. My happy conclusion is that this dream is within reach: support for statistics education has never been higher; resources have never been more abundant, and the technology required for personalization is now ubiquitous.

Bayesian Statistics: "The Second Coming!"
Wayne Stewart, University of Auckland

The once commonly used Bayesian paradigm is making its way back and has the potential to re define modern statistics. The credibility, applicability and the richness of the archetype are self evident. Most of the agreed disadvantages of Bayesianism namely, priors and MCMC are two edged. The priors are a backdoor for information previously found and relevant to the study. MCMC, a mathematical tool, has the added value that functions of random variables can be easily summarized even when they are analytically intractable. Although the theory is well developed and reasonably straightforward for statisticians, the teaching of it still remains a challenge specially to undergraduate students with less mathematical knowledge and skills. How can we teach Bayesian statistics in ways that will actively facilitate the use of this incredibly powerful paradigm rather than procrastinating and watching the unexploited opportunities float by?

In this talk I will illustrate the effectiveness of Bayesian statistics and how it differs from classical statistics. I will also show some fascinating examples of the paradigm by a meaningful comparison of confidence intervals with Bayesian credible intervals to point out its interpretational simplicity and advantages. Bayesian estimates can be biased but will often have better frequentist mean squared errors. Hierarchical modeling which is easily accomplished in a Bayesian framework and difficult to perform within the classical paradigm is a natural way of pooling information to produce smaller interval estimates for parameters.

It Takes a Village: Future Directions for Statistics Education Research
Bob delMas, University of Minnesota

Download Presentation (PDF)

Arguably, the discipline of Statistics Education is a relatively new field that has emerged over the last three decades. While fairly young, it is a fairly broad field that has investigated effective methods for teaching statistics, students' statistical reasoning and thinking, the nature and cause of faulty statistical reasoning, non-cognitive outcomes and other factors that affect the learning of statistics. The field is witnessing the application of new technologies and the exploration of new content in the teaching of statistics. The growth of the field raises the need for statistics education researchers who are educated in a variety of areas related to conducting educational research. This talk will consider implications of the new pedagogies and content for research and the type of training needed to conduct research. Future statistics education researchers will need training in statistics, educational measurement, educational research methodologies, education and learning theory. As such, it will take a village of individuals trained in a variety of areas working together to move our understanding of statistics education forward. A collaborative model for graduate training within a supportive environment that promotes brainstorming, making mistakes, arguments, discussion, creative thinking and experimentation will be presented.