Introducing Statistical Software when Teaching Methods and Statistics

Teaching research methods and statistics is tough. There’s no way around it. You are tasked with communicating a wealth of complex ideas to students who often have a genuine fear of numbers and math. I have had students approach me on the very first day of class, worried that they would not be able to succeed simply because there is a mathematical component to the course. Computer software is supposed to help us with calculations, but the irony is that learning how to use software seems to be just an additional burden that many—students and teachers alike—are reluctant to take on. The question of whether, why, when, and how to introduce software when teaching statistics is a thorny issue, not to mention the “what” of which software to use. It is unlikely that I could answer all of these questions in a fashion that would satisfy most of you, in such a short Blog post. However, I think sharing my own perspective and strategies could be helpful, and perhaps encourage you to explore some new possibilities.

Personally, I believe that it is very important that students learn how to use statistical software. Some have argued that students should first gain a deep conceptual knowledge of statistical concepts, and that trying to teach software at the same time simply gets in the way of this goal. Although there is some definite, and obvious, truth to the fact that teaching software may take time away from teaching concepts, I believe it ignores an important benefit of teaching software: motivating students to learn more. Consider this, how can we best motivate students to learn the concepts behind statistics? In my experience, the best way to encourage someone to learn something difficult is to show them what they themselves can do with this knowledge. Once students see the potential of statistics, and its ability to do interesting things that are personally relevant, they will be far more motivated to understand it. To me, there is something almost magical about having a strong foundation in research methods and statistics. What I mean by that is that whenever some silly question pops into my head, I can very quickly delve into the literature and get a decent grasp of what the available evidence looks like, and the strength of this evidence. If I’m really lucky, I can even find some publicly available data online and run some quick analyses to get a tentative answer to the question I’m pondering. Introducing students to this magic, giving them the experience of what it is like to ask a question and get evidence of an answer by analyzing data, will teach them more about “why” they should learn statistics than any abstract statements about its utility. In other words, if we start by showing students the wonders that lie at the end—analyzing data to help answer an interesting question—we’ll have a much easier time motivating students to make this difficult journey.

A key to hooking students with the magic of statistics is to use these tools to answer a question that they find personally interesting. Starting your course with a brief survey that asks them about their interests will help you to identify the kinds of examples that are going to be the most engaging. Be it sports, movies, television, pets, videogames, or food, there exists plenty of public data out there that they can learn to analyze, to learn more about these topics. If you like, the first analyses you can have them do can be with user-friendly software, such as a spreadsheet program (e.g., Google Sheets, OpenOffic Calc, or Microsoft Excel). But I personally think that it would be most helpful for them to learn software that they could conceivably use to do more powerful analyses in the future, as soon as possible. My own preference is for R, a free and open-source statistical programming language that is used across a wide range of disciplines and has great popularity among those specializing in quantitative methods. But R is certainly daunting: it is much more akin to learning how to program in a computer language, than using more familiar forms of software with drop-down menus and a point-and-click interface. How to overcome this challenge? Even the most user-friendly tutorials on how to get started with R can be difficult for undergraduates, who often even struggle with the installation of R and its companion software, R Studio (which makes interfacing with R and its outputs a bit easier).

To help students—and instructors!—with this, I have created what I consider to be the gentlest on-ramp to learning R: Research Methods: Interactive Demonstrations in R at York (ReMInDeRY). These free, online tutorials are accessed solely through a web browser, and teach important concepts in research methods and statistics, while simultaneously introducing students to R. Students are able to run, edit, and write R code, and observe its output, all through the browser. There are currently three tutorials, and over the course of completing them they will learn the foundations of how R works, and even experience what it is like to answer a question by analyzing some real-world data (about pets in Toronto). Hopefully, this is enough of a taste to motivate them to learn more about R, with additional resources for doing so also provided on the ReMInDeRY site. These tutorials were all created with the {learnr} package, created by Barret Schloerke and colleagues at RStudio. So you can also use this package to make your own custom tutorials. At the site you can also find interactive in-class demonstrations for methods concepts, to help engage your students in-class as well as online.

Because teaching methods and statistics is tough, we as instructors are often reluctant to incorporate even more into our courses, as this would seem to make an already difficult task even harder. However, I would argue that helping your students experience the magic of what they can do with statistics, by analyzing real data to address an interesting question, will actually make your job easier. Having students that are motivated and excited to learn a new skill, because they’ve had a first-hand taste of what they can do with this skill, is a boon for any instructor.

Dr. Raymond A. Mar is a Professor of Psychology at York University. His research revolves around the real-world influence of imaginative experiences, about which you can learn more on his website.