Registration for Winter term is fast approaching, and I come bearing advice. It cuts across colleges and concentrations, BSs to BAs to BGSs. I suspect you’ll dislike it, but here goes: if you haven’t taken a course in statistics yet, make room for one. And more importantly, if you have taken an introductory statistics course, then make room for a second, more advanced one.

Sarah Royce

If someone had told me that during my undergraduate years, I would have asked them just which circle of hell they get their advice from. And then I would have found a new advisor. For whatever reason, there’s something inherently aversive about the idea of taking a course in statistics.

Maybe numbers in general just turn people off, but even among the numerical disciplines statistics inhabits a special realm of dislike. Calculus is no less arduous than statistics, but is somehow more admirable. It’s the math of Newton, after all. As for statistics – well, it famously lives somewhere amidst lies and damned lies.

But for all its bad connotations, statistics has a rather rare, quite valuable quality: external validity, meaning that it is useful even outside the boundaries of the ivory tower.

Pause there for a moment: useful outside of the classroom. Given that most of us will spend four or five years at a university and then 50 years or more somewhere else, anything that’s going to inform those other 50 plus years is worth investigating.

Considering that I’ve spent the past five years studying psychology, I find it more than a little ironic that my recommendation for a course with practical utility is statistics. I remember thinking at the beginning of my undergraduate career that studying human behavior was going to dramatically shape how I viewed the world, how I interacted with people and how I behaved. But even though I can tell you a lot about how people behave in a variety of laboratory and even naturalistic settings, it is rare that my psychological education shapes my day-to-day opinions or helps me to make an informed decision. My guess is that students in many other disciplines would say the same.

Shouldn’t we be learning how to make informed decisions outside of labs and lecture halls? How to weigh evidence and interpret the flood of information on television, on the web – in student newspapers, even. My past columns have tipped my hand, but I clearly think the answer is yes, and statistics, eternally chagrined, is a path to enlightened decision making.

On the bright side, I think that most undergraduates encounter a statistics course somewhere along the road to a bachelor’s degree. But that first course isn’t really enough – mostly spent learning formulas and the necessary software. Introductory statistics produces students with the knowledge that some findings are significant and others aren’t, and perhaps the ability to figure that out using some program or other. A good start but, like the introductory course in a new language, not really much more than a primer.

Unfortunately, this first course is commonly also the last; most area requirements entail only that initial course and sometimes not even that. At the University of Arkansas, where I did my undergraduate work, statistics cannot satisfy the mathematics core requirement for a bachelor’s degree (only college algebra, finite mathematics, and calculus can). The statistics course that I took was a degree requirement for my major and was a necessarily shallow foundation.

I’m not entirely sure what the situation is here at the University, but based on what I could gleam from the LSA website, it appears similar. The quantitative reasoning requirement sounds lovely, but a formal course in statistics is merely one of a number of courses that can satisfy this requirement. I doubt many people would choose it over a geoscience course.

I am entirely aware that there are only so many courses that you can cram in to a bachelor’s degree, and asking for two statistics classes as a core requirement is pretty unrealistic. Asking for one, though, is reasonable and in my opinion entirely worthwhile. So how can we make the one statistics course at the university level go further?

For starters, we can emphasize statistics more in high school. There’s no reason why calculus should be the end-all, be-all of high school mathematics, especially because statistics is going to be much more useful to most of these students in the long run. High school students should still have the option to take calculus, and those who are planning to study engineering or physics would do well to take lots of calculus as early as possible. But everyone else should have the option to take introductory statistics, and should be strongly encouraged to enroll. This foundational knowledge could be bootstrapped by universities, making a student’s one University course in statistics much more than a introduction.

Oh, won’t the University be surprised when statistics courses in the winter term are overflowing with students (some silently cursing my name)? Almost as surprised as you will be two or three years hence, when it turns out that a knowledge of probability theory and effect sizes is not just useful, but profoundly useful.


Jackson can be reached at edjacks@umich.edu.


Leave a comment

Your email address will not be published. Required fields are marked *