It's hard to love your Statistics course. Many students majoring in the social sciences are required to take it, but rarely do they enjoy the long list of tools that must be learned; and the curious way in which one must twist one’s mind to understand the concepts.

I took Statistics in the second millennium, which means that we didn’t use computers at all. We had to memorize concepts and definitions, and learn how we would use tools, assuming we ever had access to data and to the computer power (and/or the time) to actually use them.

I think this is a very unsatisfactory way to teach, period. In my course, students learn the concepts and learn how to use them pretty much simultaneously. They see, for instance, the meaning of “standard deviation” not just from seeing a picture in a book, drawn from a hypothetical (i.e., made up) data set, but from typing in the right commands to analyze data that they can actually see.

One of the great pitfalls is the use of fake data, which are nice, clean, and unrealistic. Many textbooks nowadays come with a CD (or a website) full of data drawn out of real examples. My sense is that this data still feels fake: it was collected by someone else, for some purpose that had nothing to do with my class.

So every time I have taught the class, I have had my students carry out an extensive survey of what they know best—the student body. They actually collect the data (which is worth the course in itself, I think), input it, and analyze it. Because they actually know what it is to be a student, they can develop sensible and fairly sophisticated hypotheses, which they test scientifically and present to their peers.

This highly hands-on way of teaching Statistics has an added benefit: upper-division classes, most often, deal with the application of the theoretical tools learned in the theory classes (mostly Intermediate Macro and Micro) to specific aspects of economic life (from industrial organization to developing countries).

Most practicing economists approach reality with a mix of theory and empirical analysis. Traditionally, undergraduates have been shut out of empirical analysis because it is too hard (it is) and because they don’t know how to apply the tools they’ve learned about in Statistics to the data. By giving students plenty of experience in the use of statistical analysis on data that they actually understand, our major makes it easy for them to understand reality a bit better.