Deflated grades prove equally unfair

Mike Singer

My colleagues, integrative biology professors David Hillis and Eric Pianka, object to grade inflation and have expressed so through firing lines that recently ran in The Daily Texan. Hillis notes that inflated grades are unfair to excellent students. Indeed. But deflated grades are unfair, too, when students receive lower grades for their work than is currently normal, making them appear to potential employers as less able than they are.

This wouldn’t matter if grading standards were similar among teachers. Now, with the appearance of, we can see how variable grades are among different sections of the same class. Grades for BIO 301M, which I traditionally teach, ranged from 7 percent, 10 percent and 15 percent As in the most sternly-graded classes, up to 38 percent, 39 percent and even 65 percent As at the other end of the scale. The proportion of Cs ranged from 2 percent and 8 percent up to 40 percent, while Ds and Fs ranged from 1 percent to 22 percent.

If students in these classes were performing equally, some of my students making Cs could have made As in a different BIO 301M class. Even more extreme, all of the C students and some of his D and F students in one class would have made As and Bs in a different class.

Can we tell whether students in these classes were equally able? I think we can. I have taught two BIO 301M classes back-to-back each fall for the past 10 years or more, and I give both classes the same true-false, computer-graded tests. Each year, I can compare the performances of two groups of students given the same lectures, the same text reading and the same tests. I decide in advance what percentage of students will get each grade (previously 30:30:30:10). When I apply this to each class separately, I observe that the test scores at each grade cutoff are never more than 1 percent apart between classes. Why is this? It is simple statistics: These classes are large enough samples of the student body that they are good representations of it. So, the differences in BIO 301M grade distribution that we see in reflect differences in professorial policy, not differences in student ability or effort. This is seriously unfair.

My policy can result in a student scoring 250/360 at end of semester making a B and a student scoring 249 making a C. The difference is meaningless, so it’s bad luck for Mr. or Ms. 249.

An alternative is to seek natural gaps in scores, as Pianka advocates for. But, while I observe that allocating As to 30 percent of the students in each of my classes results in near-identical test-score cutoffs between A and B grades, I also observe that the positions of gaps remain variable even in large classes taking the same test. If I used gaps, I’d get widely differing GPAs for my two classes, even though the students’ overall performance is the same.

Publication of GPAs must already be generating intense competition for access to “easy” classes. Winning these contests will have nothing to do with student academic ability, so unless we reform our grading policies grades will continue to be unfairly diverse. However, any kind of standardization could only apply to large classes, as small ones are genuinely diverse in student composition. If I teach five students and they all flub a test, I think to myself, “Oh dear, I have five under-performing students. Is there flu going around?” On the other hand, if I have 110 students and their mean test scores are low, I think, “Oh dear, either that was a very hard test or I haven’t been teaching effectively.” Whether my test was hard or my teaching ineffective, in neither case should I penalize the students, which is why I decide in advance what percentage of them will receive each grade and stick to it regardless of test scores.

Monetary exchange rates used to be arbitrarily confined to fluctuate within specified limits, a procedure colorfully known as the “snake in the tunnel.” Our students would benefit from such a snake in classes that are large enough and predictable enough to merit it. Could we professors be cajoled into harmonizing our grading policies? I suspect that the answer is a quote from the anonymous professor in Ph.D. Comics: “Don’t tell me what to do!”

Singer is a professor of integrative biology.