How About Some Homework On Correlation vs. Causation?

Dear Extra Credit:

Not so fast. Steve Shapiro's data on homework completed and quarterly grades [Extra Credit, Fairfax Extra, March 22] don't prove that doing homework results in better grades. The data show correlation, not causation. It could be that the ability to get better grades results in higher homework completion rates. The data cannot distinguish between cause and effect.

It may be that students who get better grades are more conscientious about their responsibilities.

There may be many reasons why students receiving poorer grades do less homework: It takes them too long; they have no interest in the subject; or they lack motivation and discipline.

Furthermore, a straight-line relationship between homework completed and quarterly grades explains only about 60 percent of the variation in the data.

The pitfalls in drawing conclusions using data from uncontrolled experiments are not widely recognized, even by the mathematically literate.

A controlled experiment would have to introduce some degree of randomness to the percentage of homework completed, maybe by assigning different amounts of homework to students.

Of course, few parents and teachers would agree to participate in such an experiment.

Ferd Neider

Vienna