The Confidence Gap: Part III


I have never understood why professors deliberately design their tests to have such low means. While some professors are simply clueless as to the actual difficulty of their tests, many intentionally write exams that most students can't complete. The only explanation I can think of it that an excessively low mean spreads the distribution of students at the top of the class, and allows professors to pick out the most brilliant of the students, the true "physics gods," as we called them.

C. Tests that everyone fails, or "teaching to the top"

In addition to unacknowledged differences in prior experience with the subject, what are some other factors that undermine student confidence? Some of these come under the general heading of "teaching to the top." That's the frequent and irritating habit physics professors have of focusing on the most brilliant students instead of the average ones.

One frequently observed irritation for introductory physics classes is the way professors write tests designed so that almost everyone does terribly. This does not usually affect the grading, since the professor will simply set the curve so that the average grade, however low it may be, translates to around a B. Still, even when students understand that their grade of 57 on a physics test is really a B, not a F, it can't help but be somewhat demoralizing. Being unable to solve a significant chunk of the problems on a test leaves students with the impression that they don't really "get" the subject, even if their final grade turns out, by a seemingly miraculous process, to be far higher than they think they deserve.

I have never understood why professors deliberately design their tests to have such low means. While some professors are simply clueless as to the actual difficulty of their tests, many intentionally write exams that most students can't complete. It's true that it is hard to distinguish among students' abilities if everyone scores in the 90s, but why wouldn't a mean of 75 or 80 work?

The only explanation I can think of it that an excessively low mean spreads the distribution of students at the top of the class, and allows professors to pick out the most brilliant of the students, the true ?physics gods,? as we called them. These are the ones who get 90s even when the class average is 50.

Now it's sort of understandable that physics professors, many of whom were presumably the physics gods of their undergraduate days, would get a kick out of identifying their future peers. But is it really necessary to do so at the introductory level? After all, most students in a lower level physics class are not ever going to become theoretical physicists, or even necessarily stay in physics at all. So why not wait until higher level courses, which are expected to be more challenging anyway, to identify those students who can go the full distance in the field? There seems no particular benefit to the top students that's worth demoralizing everyone else.

This issue is common to hard science courses in general, far more so than in the social sciences or humanities. This may be because it is simply more difficult to write, say, a history test that only the smartest can pass, without requiring students to answer completely irrelevant questions. But I suspect the real difference is that these areas have much less of a focus on native ability than the sciences, and it may well be that natural talent plays far less of a role in history than in math.

Still, physics and other science professors for lower level courses should be firmly reminded that their job is not to discover the next Einstein, but to teach at a level that serves the average students in the class.

Posted: Wed - April 16, 2003 at 06:25 PM      

©