Social Science Statistics Blog

When I was still teaching, one of my favorite courses was one in visualizing data (a la Tukey, Tufte, Cleveland). I had students download the Supplementary Homicide Reports from ICPSR and each take a state -- and analyze the data and write a report to the governor of that state concerning the nature and extent of homicide facing the state. As a start, I sent them to my article "Visualizing Homicide" in the Journal of Quantitative Criminology. In this way, everyone used the same techniques, but everyone had different data and therefore had a different report to write. Thus they combined analysis with the need to communicate the results (and prioritize the findings).

I got my B.S. in statistics a few years ago, and my school did not have a class in explaining statistics to non statisticians. But they did (do) have a lot of opportunities for Quantitative students to interact with non quantitative students.

They do have is a very large 200 level stats class that all humanities students are required to take. This class is taught by a proffessor who excells in communicating quantitative concepts in ways that creative writing and history majors can understand. And it requires a large number of teaching assistants, both graduate and undergraduate statistics students work with this class, and I think it was a valuable experience for all of us. We would spend an hour or more each week in lab working with history and psych students on understanding basic statistical concepts, and for me it really helped bridge the communication barrier. And grading their homework was certainly "educational."

In addition to the teaching type of situation there were a number of cross disciplinary classes that were available and sometimes required for quantitative students. Working on group projects with business majors really honed my skills in communicating statistical concepts, the limitations of what I could do and the results of any analysis to them. And I'm very grateful for all of this experience now that I work in the business world.

But I don't think you can force people to learn this sort of thing.

Some of the problem comes from the audience side as well. Often even those who should be able to develop a rudimentary grasp of statistics (i.e. college students) approach it as though the math is some sort of magic spell. What is worse, they're never taught how to look at the numbers and extract meaning from them. So, as a result, even if we're trying to speak clearly, our audience is convinced going in that they can't understand us.

Oops.

I think you're absolutely right- a focus on the mathematics is good, but such a focus to the exclusion of comprehension is a serious problem.