Sunday, May 02, 2004
Grade Inflation
I have been thinking about Michael Bérubé's comments on grade inflation in response to Princeton University’s announcement that it would place quotas on A grades, limiting them to no more than the top 35% of students in any course. I am intrigued by the implications of Princeton’s actions, and of Bérubé’s solution. He proposes to take statistics from faculty grading patterns and factor these patterns into student grades. He outlines the system as follows:
Every professor, and every department, produces an average grade -- an average for the professor over her career and an average for the discipline over the decades. And if colleges really wanted to clamp down on grade inflation, they could whisk it away statistically, simply by factoring those averages into each student's G.P.A. Imagine that G.P.A.'s were calculated on a scale of 10 with the average grade, be it a B-minus or an A-minus, counted as a 5. The B-plus in chemical engineering, where the average grade is, say, C-plus, would be rewarded accordingly and assigned a value of 8; the B-plus in psychology, where the average grade might be just over B-plus, would be graded like an easy dive, adequately executed, and given a 4.7.
It should be pointed out that this would be a nightmare system for employers to read. A student who majored in chemical engineering with a GPA around 8 would look much the same as a student who majored in psychology and also had a GPA around 8. But the psychology student, having survived the ravages of forced grade deflation would be truly outstanding, whereas the chemical engineering student would merely have succeeded in a hard subject. Factoring subject matter would create such complexities as to render it uninterpretable.
But this raises a very serious point. In order for GPAs to have meaning, they must be transferable from subject to subject, but also from university to university. At my university, English courses have the lowest average grade of any subject area; whereas, to be honest, my impression is that many students choose English to avoid subjects that they think are harder. I should add that the English department services a large number of (again, to be honest) low-achieving education students, which skew the statistics somewhat. On average, I’d say that no more than 15% of my students in any class get As. On the other hand, at a very selective institution Princeton, surely a much higher percentage of As is a reflexion of the calibre of the student. What we really need, then is a way to factor this in. However, other countries have tried to enforce national standards with limited success. In the end, I believe institutional reputation serves the purpose.
Another question that arises out of this discussion is the meaning of the average grade. Does this mean the average for an individual class, the average over a number of semesters, or something else? I tend to view average as implying that a piece of work displays a certain skill set—whatever could be expected of a person with little exposure to the subject matter or to educated literary discourse who has a serious crack at the assignment. Such work tends to be somewhat lacking in the rhetorical and mechanical expectations of an educated treatment of the subject matter but shows a genuine, if not perfect, familiarity with the subject. It generally lacks much insight. Work that improves on this is above average (B range); work that improves seriously is excellent (A range). I generally award Ds to work that falls below this standard and Fs to work that is not turned in. The one exception is in my grammar courses, where students do so spectacularly badly that the scale for submitted work tips from three passing grade ranges (A, B, C) and one failing grade (D) to one with two failing grades (D and F). Student grades cluster in the B and C range for most of my classes but more in the C range for my grammar classes.
I think we need to look closely at the institutional and cultural pressures that govern grading policies in order to derive some clear notions of what an average student actually is.
Every professor, and every department, produces an average grade -- an average for the professor over her career and an average for the discipline over the decades. And if colleges really wanted to clamp down on grade inflation, they could whisk it away statistically, simply by factoring those averages into each student's G.P.A. Imagine that G.P.A.'s were calculated on a scale of 10 with the average grade, be it a B-minus or an A-minus, counted as a 5. The B-plus in chemical engineering, where the average grade is, say, C-plus, would be rewarded accordingly and assigned a value of 8; the B-plus in psychology, where the average grade might be just over B-plus, would be graded like an easy dive, adequately executed, and given a 4.7.
It should be pointed out that this would be a nightmare system for employers to read. A student who majored in chemical engineering with a GPA around 8 would look much the same as a student who majored in psychology and also had a GPA around 8. But the psychology student, having survived the ravages of forced grade deflation would be truly outstanding, whereas the chemical engineering student would merely have succeeded in a hard subject. Factoring subject matter would create such complexities as to render it uninterpretable.
But this raises a very serious point. In order for GPAs to have meaning, they must be transferable from subject to subject, but also from university to university. At my university, English courses have the lowest average grade of any subject area; whereas, to be honest, my impression is that many students choose English to avoid subjects that they think are harder. I should add that the English department services a large number of (again, to be honest) low-achieving education students, which skew the statistics somewhat. On average, I’d say that no more than 15% of my students in any class get As. On the other hand, at a very selective institution Princeton, surely a much higher percentage of As is a reflexion of the calibre of the student. What we really need, then is a way to factor this in. However, other countries have tried to enforce national standards with limited success. In the end, I believe institutional reputation serves the purpose.
Another question that arises out of this discussion is the meaning of the average grade. Does this mean the average for an individual class, the average over a number of semesters, or something else? I tend to view average as implying that a piece of work displays a certain skill set—whatever could be expected of a person with little exposure to the subject matter or to educated literary discourse who has a serious crack at the assignment. Such work tends to be somewhat lacking in the rhetorical and mechanical expectations of an educated treatment of the subject matter but shows a genuine, if not perfect, familiarity with the subject. It generally lacks much insight. Work that improves on this is above average (B range); work that improves seriously is excellent (A range). I generally award Ds to work that falls below this standard and Fs to work that is not turned in. The one exception is in my grammar courses, where students do so spectacularly badly that the scale for submitted work tips from three passing grade ranges (A, B, C) and one failing grade (D) to one with two failing grades (D and F). Student grades cluster in the B and C range for most of my classes but more in the C range for my grammar classes.
I think we need to look closely at the institutional and cultural pressures that govern grading policies in order to derive some clear notions of what an average student actually is.
Comments:
Post a Comment
 |
|