Volume CXXXIII, Number 13 January 25, 2002 f

Bowdoin grading practices: What we say vs. what we do
DAVID VAIL
FACULTY CONTRIBUTOR

To paraphrase Mark Twain, grading, like sausage making, should not be watched too closely. It's an inherently subjective and imprecise weighing of many aspects of student performance. It's highly quantified in some fields, inescapably qualitative in others. And it serves multiple purposes, with no consensus about their relative importance.

Other articles in this week's Orient highlight Bowdoin's grade inflation. This one documents a widespread deflation of standards: As and Bs received for work that falls below the standards spelled out in the College Catalogue. In 1998, following an earlier grading debate, the faculty set high standards for A and B work:

"A, the student has mastered the material of the course and has demonstrated exceptional critical skills and originality;
"B, the student has demonstrated a thorough and above average understanding of the material of the course."

Some of us supported this wording in the hope that it would counter grade inflation. It hasn't worked out that way. In 1997-98, 85 percent of all grades were As and Bs; last year the fraction was 87 percent. At Bowdoin as in Lake Wobegon, nearly everyone is above average - and getting better.

To get a handle on the apparent gap between grading rhetoric and practice, several Economics Department members surveyed faculty late last fall. 82 colleagues responded, just under half of the fall teaching staff. This article interprets their responses, without addressing basic survey research issues such as sample biases and statistical significance.

Based on 81 faculty members' practices, deflated standards-i.e. inflated grades-are nearly universal. 85 percent of respondents acknowledge that they have given As to students who did not fully meet the Catalogue description of A work; 80 percent say the same about Bs.

The responses indicate that deflated standards are not a rarity. For 40 percent of responding faculty, at least one in ten A students fall short of the Catalogue standard; for 20 percent of teachers, this holds for at least one-quarter of As. In the case of Bs, for 48 percent of respondents, at least one in ten B students falls short. And, for 31 percent of faculty, this holds for one-quarter or more of all Bs.

A few noteworthy patterns stand out in the responses. Although the accompanying tables show that arts and humanities faculty award the largest proportion of As and Bs, comparatively few of them (80 percent) say they have given As for below-standard performance; and, just 15 percent of the arts and humanities respondents give one-quarter or more such "below-standard-performance" As. In contrast, math and science grades are comparatively low, even though more math and science teachers (93 percent) give "undeserved" As and fully 37 percent give at least one-quarter such As. Social science grading falls in the middle. For below-standard Bs, there's a roughly similar pattern across academic divisions.

The big question is why nearly all Bowdoin teachers violate, and many routinely violate, "legislated" standards. Our open-ended questions seeking explanations elicited diverse, thought provoking, and often eloquent responses. Due to space limitations, I will indicate just the most frequent responses from those teachers who award at least 25 percent of As and Bs to students with below-standard work.

More than half of those who award many below-standard As stress that it is not reasonable to expect "exceptional creativity and originality" in some courses, particularly at the introductory level. From this perspective, the problem is with the Catalogue definition. However, one-third of this group articulate strategic motives for giving "easy" As. These rationales range from a frank desire for good course evaluations and high enrollments to the common view that it is not smart (or fair) to diverge too far from college-wide practice.
Among the many respondents who frequently give Bs for below-standard work, three-fourths sounded variations on the theme that B is Bowdoin's de facto norm for competent-"average"-performance. But here, too, one-third of the responses were colored by strategic thinking, akin to the logic for giving below-standard As.

The search for systematic differences across the academic ranks (non-tenure track, tenure track, tenured) turned up nothing dramatic or statistically significant. For example, these self-reported data do not support the hypothesis that tenure-track faculty "go easy" on students to bolster their enrollments and teaching evaluations.

What practical insights can we take from the responses?

Most basically, the disparities in grading practices among faculty and the divergence between Catalogue standards and common practice argue for a broad stakeholder dialogue?including students?about appropriate standards and how to apply them equitably and consistently, particularly across disciplines and course levels.

Specifically, in the name of intellectual and moral integrity, we should lower either grade distributions or the Catalogue standards (or both). The solution will not be simple. For instance, keeping current Catalogue standards but lowering grades would paradoxically require the biggest downward adjustment in math and science, which already have lower grades than the other divisions.
De facto, Bowdoin's grading has collapsed to a two point system. This spring the faculty will consider the marginal refinement of adding plus and minus grades. Since many faculty currently round grades upward and since Cs are widely viewed as punitive, pluses and minuses, in my view, would help- a little- to make grades better reflect differential performance and to enhance fairness.

Note: The survey results, including responses to open-ended questions, will be made available on a website.