Teaching Introductory Statistics in the 2020s:
Multivariable Thinking, Data Fluency,
and Statistical Inference Beyond “P < 0.05”

A Virtual Workshop for Two-Year College Educators

Monday, June 29 – Tuesday, June 30
12:00 – 4:00 p.m. ET

This virtual workshop is free, but advance registration is required.


Register

This workshop will focus on ideas and activities for teaching introductory statistics in accordance with the American Statistical Association’s GAISE recommendations and “A World Beyond P < 0.05” initiative. Highlighted themes will include multivariable thinking, data fluency, misconceptions and limitations of p-values, and alternatives to p-values such as effect sizes.

Monday, June 29

12:00 p.m. – 12:10 p.m.   Overview (Allan Rossman)
12:10 p.m. – 1:10 p.m.   Developing Multivariable Thinking (Roxy Peck)
1:20 p.m. – 2:20 p.m.   Data Cleaning with Data Moves: Just a Spoonful of Sugar (Rob Gould)
2:30 p.m. – 3:30 p.m.   Multiple Variables, One Method, No Login (Danny Kaplan)
3:30 p.m. – 4:00 p.m.   Q&A (Roxy, Rob, and Danny)

Tuesday, June 30

12:00 p.m. – 12:10 p.m.   Overview (Allan Rossman)
12:10 p.m. – 1:10 p.m.   Building Student Intuition Through Simulation-Based Inference (Beth Chance)
1:20 p.m. – 2:20 p.m.   Be More Effective and Less Significant (Jeff Witmer)
2:30 p.m. – 3:30 p.m.   Evaluating Causal Evidence (Kari Lock Morgan)
3:30 p.m. – 4:00 p.m.   Q&A (Beth, Jeff, and Kari)


This workshop is being conducted on behalf of the Joint Committee of the American Statistical Association and the American Mathematical Association of Two-Year Colleges.

Please direct questions to Allan Rossman or Rebecca Nichols. A Zoom link for this virtual workshop will be sent to all registrants in advance.

Abstracts and Biographical Sketches

Developing Multivariable Thinking (Roxy Peck)
Recent guidelines for introductory statistics recommend providing students with opportunities to develop multivariable thinking. Because this recommendation is relatively new, many faculty members are uncertain about how to implement it. This session will explore activities that incorporate multivariable thinking, which can be integrated at various points during the course to complement traditional course content.

Roxy Peck is a professor emerita of statistics at Cal Poly – San Luis Obispo. She has received the American Statistical Association’s Founders Award in recognition of her contributions to K–12 and undergraduate statistics education, as well as the CAUSE/USCOTS Lifetime Achievement Award in Statistics Education. She has co-authored textbooks for introductory statistics and edited the essay collection Statistics: A Guide to the Unknown.

Data Cleaning with Data Moves: Just a Spoonful of Sugar (Rob Gould)
While it would be nice for every data set our students confront in their lives to come in a tidy comma-separated file, this is often not the case. Even when it is the case, the data are rarely ready for immediate analysis. While the phrase “data cleaning” may have the unfortunate tone of a tedious chore, it is actually an act of modeling that can be both fun and challenging. This session will provide practice with “data moves” for cleaning and preparing data that will widen our students’ access to data.

Rob Gould is a teaching professor in the UCLA Department of Statistics and current chair of the ASA/NCTM Joint Committee on K–12 Education in Statistics and Probability. He is co-author of an introductory statistics textbook and founder of the ASA’s DataFest competition. He is also the chief architect of the Mobilize Introduction to Data Science curriculum, a year-long course for high-school students. Gould has received the ASA’s Waller Distinguished Teaching Career Award and the CAUSE/USCOTS Lifetime Achievement Award.

Multiple Variables, One Method, No Login (Danny Kaplan)
I will introduce the StatPREP Little Apps that use real, rich, multivariate data while fitting easily into an existing curriculum, regardless of course text. I will also show how to introduce all the standard settings of statistical inference (e.g., difference of two proportions, difference of two means, regression, categorical counts) with a single, simple method that applies naturally to modeling with multiple explanatory variables. We’ll use the regression Little App, illuminated by the Compact Guide to Classical Inference, to see how inference can be done with two formulas and the magic number 2^2.

Danny Kaplan teaches statistics and applied math at Macalester College. He has written several textbooks about statistical modeling, data science, applied math, and computing and is a co-developer of the mosaic and ggformula R packages. He is a co-principal investigator for the StatPREP program and a recipient of the CAUSE/USCOTS Lifetime Achievement Award in Statistics Education.

Building Student Intuition Through Simulation-Based Inference (Beth Chance)
We explore how to use simulation-based inference (SBI) to introduce and reinforce students’ understanding of the concepts of significance testing and confidence intervals at the beginning, middle, or end of an introductory statistics course using freely-based online applets.

Beth Chance is professor of statistics at Cal Poly – San Luis Obispo. She has been involved with the AP Statistics program and ASA Section on Statistics (and Data Science) Education for many years. She is a co-author of introductory and intermediate statistics textbooks for algebra-based courses and an introductory statistics textbook for math and statistics majors. She has received national teaching awards from the American Statistical Association and the Mu Sigma Rho honorary society.

Be More Effective and Less Significant (Jeff Witmer)
The words “statistically significant” get in the way of understanding what a hypothesis test tells us. Moreover, there are important things a hypothesis test does not directly tell us, which is where effect size comes in. We’ll explore effect sizes for common inference situations.

Jeff Witmer is a professor at Oberlin College. He has written several textbooks and is currently editor of the Journal of Statistics Education.

Evaluating Causal Evidence (Kari Lock Morgan)
There are (at least) three reasons for why a sample difference may be observed between two groups: (a) a true causal relationship; (b) the groups differed to begin with; and (c) just random chance. To evaluate evidence for the causal claim (a), we’ll discuss ways to help introductory students reason about and assess evidence against the competing claims (b) and (c).

Kari Lock Morgan earned her PhD in statistics from Harvard University and is now an assistant professor of statistics at Penn State University. Her primary research interests are causal inference and statistics education. She is a co-author of an introductory statistics textbook and the recipient of the national Robert V. Hogg Award for Excellence in Teaching Introductory Statistics.


Register