Teaching Statistics and Quantitative Methods in the 21st Century

Joseph Lee Rodgers

• Publisher: Routledge
• Year: 2020
• eBook ISBN: 9780429810206
• Topic: Education
• Subtopic: Education Theory & Practice

This work, which provides a guide for revising and expanding statistical and quantitative methods pedagogy, is useful for novice and seasoned instructors at both undergraduate and graduate levels, inspiring them to use transformative approaches to train students as future researchers.

Is it time for a radical revision in our pedagogical orientation? How are we currently teaching introductory statistics and quantitative methods, and how should we teach them? What innovations are used, what is in development? This ground-breaking edited volume addresses these questions and more, providing cutting-edge guidance from highly accomplished teachers. Many current textbooks and syllabi differ in only superficial ways from those used 50 years ago, yet the field of quantitative methods—and its relationship to the research enterprise—has expanded in many important ways. A philosophical axiom underlying this book is that introductory teaching should prepare students to potentially enter more advanced quantitative methods training and ultimately to become accomplished researchers.

The reader is introduced to classroom innovation, and to both pragmatic and philosophical challenges to the status quo, motivating a broad revolution in how introductory statistics and quantitative methods are taught. Designed to update and renovate statistical pedagogy, this material will stimulate students, new instructors, and experienced teachers.

Chapter 1
Teaching Statistics and Research Methods in the 21st Century

An Introduction to 17 Chapters on Statistical Pedagogy, Curriculum, Philosophy, and Administration

Joseph Lee Rodgers

This edited book is about teaching statistics, in particular within social and behavioral science programs. Most of the chapter authors work in psychology departments or in quantitative methods programs in colleges of education. The 18 chapters in this book address the philosophical challenges of teaching statistics (and related quantitative methods) and propose innovations that can help improve our teaching of statistics. Ultimately, the whole book coheres to produce a call for reform. In this introductory chapter, I’ll first organize a community effort to share in the “improve our teaching” initiative. Following, I’ll present a summary and then define past efforts to stimulate teaching reform. Next, I’ll summarize the 17 substantive chapters. Finally, I’ll present 10 pieces of teaching advice, styled for teachers in the arena of statistics and quantitative methods.

The Rodgers Challenge

I begin this introductory chapter with a challenge. I’ll present the challenge and the reward—and then elaborate. The challenge is addressed to anyone reading this particular sentence, which is just one sentence out of a rather longer book. But to be transparent, the challenge is most applicable to those who teach statistics at the introductory level, and especially in behavioral or social science settings.

The challenge: Read the introduction of every single chapter in this edited book—17 chapters, not including the introductory chapter. Then choose at least three of the chapters and read them, carefully and completely. Once completed, send me an e-mail addressed to [email protected] or to [email protected], and let me know that you’ve completed the “Rodgers Challenge” (and please be sure to put “Rodgers Challenge” in your subject line). In your e-mail, let me know which chapters you read. Also, include any comments you may have about your reaction to the chapters: Did you like them? Did you disagree with certain features? Do you believe the chapters will inspire statistics teaching? Did they inspire you? Do you believe that statistics teaching in the behavioral and social sciences needs to be redefined and reformed? Does this book have the seeds to stimulate that reform? How will we know when such a reform has begun and when it will be completed? In your e-mail, feel free to ignore these questions, or address every one—it’s up to you.

The reward: To those who accept the Rodgers Challenge and contact me by e-mail, I’ll send you back a summary of what you and others have said about teaching statistics and the chapters in this book. I’ll include your e-mail in the summary, along with those of all others who respond. I’ll make a short statement to encourage interaction and discussion. In fact, I’ll expect and hope to be one of the discussion leaders. If your immediate response to this challenge is, “For reading three chapters and responding, all I get is a crummy e-mail summary,” then please feel free to opt out of the challenge. If what you say is, rather, “I think my own teaching will be enhanced and stimulated by interacting with others and with others’ opinions, and maybe I’ll really enjoy reading these three chapters,” then the Rodgers Challenge was designed, specifically, exactly, for you!

To belabor the point: Why is this challenge worth doing? If you engage in this exercise, you will be voting and connecting. You will be voting with your feet, indicating that teaching statistics is important and is worth doing well. You will be connecting with others who agree. That connection, alone, will be worth the small effort to join the Rodgers Challenge. (And also, let’s be clear—many of the chapters in this book are awesome and amazing—get ready to be impressed!)

I have one other quick comment about the Rodgers Challenge, and then I’ll move onto a more traditional introductory chapter. I’m not sure exactly when I’ll circulate the summary. Once the book is published, it’ll take some time for the potential audience of statistics teachers to find out about it, consider it, purchase it, and then eventually put “eyes on the introduction.” And then it’ll take some time to read those three chapters. But obviously, this is a dynamic process, one that ultimately has no end date. I expect to circulate summaries not just once but on a regular basis—maybe once every six months, for several years. And each will be as updated as you, the audience of responders to the Rodgers Challenge, provide material to support that process. So here’s the take-home message: Don’t worry that it’s too late. It isn’t. Whenever you read this introduction, all chapter introductions, and then at least three chapters, you’re on!!! If it’s been six weeks since you read about the challenge and you’ve now completed it, then great, send me an e-mail. It’s been six years since you first read about the challenge but you’ve only now completed it, no problem, send me an e-mail! And once on, unless you opt out, you’re a part of the circulation list in perpetuity. I look forward to hearing from you with great pleasure and anticipation!

Overall Summary

The book you hold in your hand (or that you see on the screen in front of you) is an effort by around two dozen practicing quantitative methodologists to improve our teaching craft as it applies to introductory statistics (and related quantitative methods). At least that’s the starting point to motivate this book. There’s more, though. I list three goals, explicitly:

• Goal # 1: We present several philosophical/administrative/organizational chapters that can help orient, improve, and motivate better teaching of introductory statistics and quantitative methods within the social and behavioral sciences. That’s the first major section.
• Goal #2: We present a number of chapters that carefully describe innovations, new methods, revised old methods, and general approaches to teaching statistics and quantitative methods to introductory students. That’s the second major section.
• Goal #3: The overall set of 18 chapters is designed to motivate a broad revolution in how statistics and quantitative methods are taught at the introductory level. That’s the broad purpose of the whole book.

Thus, if you’re a teacher and you have a teaching philosophy (or want to develop or improve one), study the papers in the first major section. If you’re a teacher and you are interested in “best practices,” new approaches, and how to teach well, study the papers in the second major section. If you feel that teaching statistics has become fairly rote and rather stagnant and needs to be revitalized, then this is the book for you. Please, in that case, study the whole book.

Teaching Reform

Several times in this book, you will see the assertion that the tables of contents of most introductory statistics textbooks have hardly changed in the past 50+ years. Similarly, syllabi for introductory statistics courses have been relatively fixed over the past several decades. Although the textbook TOCs and the syllabi have remained stable, the whole discipline of statistics and quantitative methods has exploded, with exciting developments and intellectual fireworks. When I began my teaching career in 1980, structural equation modeling was just being proposed. Multilevel modeling barely existed, with little formal development or coherence. Categorical data analysis, hazards modeling, mixture modeling, latent growth curve modeling, and modern Bayesian methods including Bayes factors were still in development or, in some cases, only on the horizon. One would assume that with the explosion of new methods, the foundation upon which those methods have been built would somehow reflect the overall growth of the field. But that assumption would be, mostly, incorrect.

A few efforts at statistical reform can be identified during the past half century. In the field of statistics, whole issues of core journals have been devoted to statistical pedagogy (e.g., see Horton and Hardin (2015), who introduce a special issue of The American Statistician devoted to “Statistics and the Undergraduate Curriculum”). The American Statistical Association has twice published standards for teaching statistics, in 2005 and 2016. The GAISE guidelines refer to the Guidelines for Assessment and Instruction in Statistics Education College Report. In commenting on the 2016 revised GAISE guidelines, Wood, Mocko, Everson, Horton, and Velleman (2018, pp. 53–54) noted the following changes since the 2005 report: First, “More students study statistics”; second, “The growth in available data has made the field of statistics more salient and provided rich opportunities to address important statistical questions”; third, “The discipline of ‘data science’ has emerged as a field that encompasses elements of statistics, computer science, and domain-specific knowledge”; fourth, “More powerful and affordable technology options have become widely available”; fifth, “alternative learning environments have become more common”; sixth, “Innovative ways to teach the logic of statistical inference have received increasing attention.”

I’m a quantitative psychologist, as are most of the contributing authors; many of us are also members of the American Statistical Association and participate in both quantitative psychology and statistics conferences and publish in both types of journals. The sense of reform that emerges from careful study of the preceding paragraph is not so evident in the statistics classes taught in the social and behavioral sciences. There are several causes and consequences. First, many of the teachers of social and behavioral science introductory statistics courses are not well trained in sophisticated quantitative methods. Many are, rather, substantive psychologists (or sociologists or economists, etc.) who step into the statistics course as a service to the department or as a way to enhance their own methodological skills. Unlike the authors of the chapters in this book, those teachers are unlikely to encounter the GAISE guidelines. Second, publishers and book authors have cornered the market on a long-term financial bonanza. Virtually all behavioral and social science departments require an introductory statistics course, creating demand for hundreds of thousands of textbook purchases each semester. The intellectual and financial momentum that emerges from that simple statement is both profound and nearly impossible to confront. Many introductory statistics textbook authors—like many of the introductory statistics teachers themselves—are not trained methodologists. (Most trained and practicing methodologists have been busy developing and evaluating the methods that I listed earlier in this chapter rather than writing textbooks.) The textbook authors are often teachers at small colleges and universities, ones that do not support an overall quantitative methods program, writing textbooks to respond to the large demand for such textbooks. It is not surprising that those textbooks tend to follow a relatively standard formula—one that has changed only slightly during the past 50+ years. Third, there do exist examples of “reform-oriented textbooks,” and without exception, those are written by highly trained methodologist and statisticians. A few examples include Judd, McClelland, and Ryan (2017), Abelson (1995), Maxwell, Delaney, and Kelley (2017), Cohen, Cohen, West, and Aiken (2003), and Freedman, Pisani, and Purves (2007). Notably, most of these texts were published a long time ago, and the dates given are for second, third, or even fourth editions. Also notably, none of those textbooks have been widely adopted for teaching introductory statistics courses. Nor were they intended for introductory teaching; most of those were written for more advanced courses.

In comparison, very few, if any, undergraduate statistics textbooks have been written to emphasize innovation or modern statistical methods. Perhaps of greatest concern, statistics textbooks in 2019 (when this introduction is being written) have no component that appears to link to more sophisticated quantitative methods. The assumption is that how statistics was taught in 1970 is just fine for how statistics should be taught in 2020. Garfield et al. (2011) further support the principle that the field needs much more than just minor adjustments in how we teach introductory statistics.

This edited book is, in a fundamental and very real sense, a corrective effort. Whether the goal of teaching reform emerges from efforts around 2020 or sometime in the distant future, the preceding paragraph should make clear that there are substantial challenges standing in front of the goal of improving and modernizing the enterprise of teaching statistics. But this edited book represents such an effort. The chapter authors have, without exception, two remarkable traits. First, they are outstanding methodologists. Second, they are committed and excellent teachers. They are exactly who should signal the directions toward which statistical pedagogy is heading. I turn now to a summary of the chapters in this book.

Chapter Summaries

Section One

Chapters 2 through 8 define the first major section, titled “Meta-Issues Related to Teaching: Curriculum, Content, Philosophy, and Supply–Demand Issues.” This section treats the discipline of teaching statistics. It may be surprising to some (though not to introductory statistics teachers) that statistics is a whole discipline and not just a set of procedures. Further, statistics does overlap into philosophy in fascinating ways.

Michael C. Edwards, in Chapter 2 (“The Role of Philosophy of Science When Teaching Statistics to Social Scientists: Two Constructivists Walk into a Bar (or Do They?”) brings philosophy-of-science issues to bear on teaching statistics and quantitative methods. He not only discusses how to do so, he has been “walking the walk” in his own teaching. Read his lively account of the challenges he has encountered in bringing formal philosophy of science concepts into statistical pedagogy.

A. T. Panter and colleagues, in Chapter 3 (“Optimizing Student Learning in Quantitative Courses”) treat a broader level than introductory statistics, as they embed statistics within STEM (science, technology, engineering, and mathematics) teaching. These authors have worked within their home institution to implement learning communities directed toward positive student outcomes. Further, the implementation was carefully evaluated in relation to those outcomes in the context of a randomized trial.

Leona S. Aiken, in Chapter 4 (“Not the What of Quantitative Training But the Who”), treats a critical issue for those interested in reform and innovation in statistical pedagogy: Who will deliver that reform and innovation? Drawing on several decades of careful data collection and analysis, she argues that modern hiring practices mitigate against progressive and sophisticated pedagogy, as much (even most) introductory statistics teaching is done by those who are relatively untrained in formal quantitative methods.

Jessica K. Flake and colleagues, in Chapter 5 (“Is Methods Research Moving Into Practice? The Critical Role of Quantitative Training”), treat the issue of how sophisticated quantitative methods are taught and then subsequently enter into routine use in applied settings. Using two subtopics (psychometrics and mediation analysis), they study the path from training to the production of “defensible psychological knowledge.” They document often ineffective applications, traceable to ineffective training.

Matthew S. Fritz, in Chapter 6 (“Singletons: Reevaluating Course Objectives When an Introductory Statistics Course Is a Student’s Only Statistics Course”), treats a relatively common though seldom considered problem. Many students only take a single statistics course; how does that inform our teaching? He argues that objectives and assessments should be adjusted in relation to this realization (and teachers should naturally encourage students to take more than just one statistics course).

Rachel T. Fouladi, in Chapter 7 (“When Statistics Assumptions Are Interesting Outcomes Instead of Nuisances—Looking Beyond the Mean”), treats a relatively sophisticated teaching issue, the assumptions underlying the conduct of statistics analysis. She quite correctly notes that assumptions are not ju...

Table of contents

• Cover
• Half Title
• Series Page
• Title Page
• Copyright Page
• Contents
• List of Contributors
• Preface
• Foreword
• 1 Teaching Statistics and Research Methods in the 21st Century: An Introduction to 17 Chapters on Statistical Pedagogy, Curriculum, Philosophy, and Administration
• SECTION I Meta-Issues Related to Teaching: Curriculum, Content, Philosophy, and Supply-Demand Issues
• SECTION II Modern Classroom Innovations in Teaching Statistics and Quantitative Methods
• Index