Proposal
Rationale, Terms and Trial Period for
Grade Normalization
At its September, 2005 meeting, the Law Faculty, acting on a unanimous recommendation of the Curriculum Committee, approved this proposal on a voluntary, trial basis for the current school year. The idea is to use this year for education and conversation about the plan, to familiarize us with the administrative problems it may engender, and to inform how its terms and structure should perhaps be changed if such a scheme is to be regularized as a mandatory system.
The idea of grade normalization is one the faculty has talked about on and off at VUSL for over fifteen years. A majority of law schools have some system of grade norming; thus, faculty who have taught at other law schools already have familiarity with the operation of such a system. The terms in the proposed plan are, by comparison to other schemes in effect at other law schools, quite timid-very large ranges of tolerance are provided. It is truly aimed only at sets of grades that can be called “outliers.”
Grade normalization plans generally have two main purposes: (1) to bring grades in all courses within some range so that a Professor's known grading pattern is not a predominant reason that courses are either taken or avoided; (2) to create fairness to students in multiple-section courses so that they are not disadvantaged by their assignment to one section rather than to another.
Brief Rationale
We begin with a rehearsal of some current pathologies which, in part, prompt interest in normalization. First, in watching registration patterns and talking to many students about their law schedule planning, it is clear that the known grading practices of faculty (including adjunct faculty) play a great role in student course selection. In some cases, students are chased away from courses which are elective but which their practice horizon suggests would be highly useful for them by what they see as harsh grading practices. More often, students are attracted to courses not all that useful to their educational goals because of known high grading practices. Such decisions have a variety of negative effects: overall rigor is reduced, especially when such courses can be enrolled in before meaningful retention checkpoints are reached;1 the student's long-range educational benefit is often sacrificed to short-term interests such as class rank, scholarship retention, reduction of effort, etc.; enrollment “gaming” forces the school to adopt increasingly harsh registration rules to combat the dislocating effects of the gaming.2
 1 The first checkpoint for a 1L is after two semesters but also includes summer courses if the student enrolls in summer school.
2 One example of many: in planning sections of Advanced Legal Research, Appellate Advocacy, Pretrial Skills, and Legal Drafting, the courses which fulfill the second-year “legal writing” requirement, we historically planned seats for the entire second-year class. Enough, right? Wrong. 3Ls (who had already of course fulfilled the requirement) began last year enrolling in those 2L drafting courses in which the professors (sometimes adjuncts, sometimes not) gave very high grades; 2Ls began signing up for two or three such courses! This led to tighter restrictions this year aimed at preventing the gaming without at the same time preventing legitimate educational interests. (For example, 3Ls were permitted to register before the 2Ls for Pretrial Skills but not for legal drafting.)
Wide disparities in grading create especially acute problems when they occur in multiple sections of the same required courses. For example, this past Fall Semester, in one required first-year course, the grades in two sections of the same first-year course differed by .33. This means that, on average, each student in Prof. X's class received a grade more than one “notch” lower (a “notch” is from B+ to B or from B to B- or from B- to C+, etc.) than each student in Prof. Y's class. As students were randomly distributed in these sections, students quickly perceive an unfairness in such situations, especially since grades and ranks carry consequences-not only for employment purposes after graduation but, indeed, immediately within VUSL for purposes of retention, scholarship retention or eligibility, selection to Law Review, etc. Moreover, it is quite clear that neither Prof. X nor Prof. Y have done anything improper. A normalization system would, if properly constructed, aid this more specific problem as well as the more general ones mentioned above.
Proposed Terms
We have looked at the normalization plans from a variety of law schools and have cobbled the following plan together from features of different plans striving for simplicity, the use of ranges wide enough to permit a good deal of individual faculty autonomy but narrow enough to avoid the pathologies listed above, and ease of administration. [Please note that an individual faculty member may either assign the grades as usual and the administration can check the grades for compliance with the system OR faculty members can simply turn in a ranking of students by raw score and the administration will assign grades at roughly the midpoint of the range and, depending on the professor's choice, either enter the grades as such or tender them back to the faculty member for further tweaking within the ranges.]
A brief note on method. We looked very carefully at our actual grading behavior in the recent past.3 We considered as important designing a scheme that, consistent with other stated goals, did the least violence to that practice. For a great many persons, the grades they have been submitting for years already comply. It is, indeed, the outliers who are affected by any scheme of “normalizing” or “standardizing.”
The plan sorts courses into categories. In each category, a range for the “mean” grade (arithmetic average grade) is provided together with upper and lower limits of percentages of grades within the “B range” (which includes B+, B, and B-). The peculiar elegance of this variation is that it virtually forces certain distributional patterns and prevents others without specifying percentage ranges for A's, B's, C's, D's, etc. Because the mean is determined and the B ranges is limited, a ”normalized” distribution occurs without further requirement.
3 Note that on the Registrar's section of the VUSL webpage, the previous four semesters of grade distributions are always available.
Some courses certainly should, because of their special pedagogical needs, be exempted from any “normalized” system. The following list is only a first impression of those courses which for various reasons seem to us to fall outside the rationale for normalization.
Proposed Exempt Courses 4
1L Legal Research
Seminar
Honors Sections
ASP Sections
Clinic
Externships [note: most graded S/U] 5
Independent Reading & Research
Current Representation
[Note: We would apply the special rule infra on multiple sections of the same required course to non-honors sections of Legal Research. Those sections appear to be highly normalized with each other historically anyway.]
First-Year Courses
For all first-year courses except exempt courses and Legal Writing, the mean grade shall fall between 2.40 and 2.90 [midpoint of 2.65] and the B range shall constitute from 35% to 55% of all grades. 6
Certain Required Second-Year Courses
For the courses in Constitutional Law II, Evidence, and Business Associations, the mean grade shall fall between 2.50 and 3.00 [midpoint of 2.75] and the B range shall constitute from 35% to 55% of all grades.
All Other Courses
For all other courses, including Legal Writing, the mean grade shall fall between 2.70 and 3.50 [midpoint of 3.1] and the B range shall constitute from 33% to 67% of all grades, unless [except for Legal Writing] the course (or section) has more than 40 students enrolled in which case the mean range shall be tightened to 2.75 to 3.25 [midpoint of 3.0].
Multiple Section Normalization
For all non-honors sections of required courses 7 taught more than once during an academic year, the highest and lowest means of such courses shall not differ by more than .25. 8
4 It may be that we should eventually consider exempting from the rules courses or sections with very low enrollment, say, fewer than 8.
5 The normalization rules, of course, have no impact on courses graded S/U. Students exercising the 3L S/U option in a course otherwise graded are not counted for purposes of any normalization rule.
6 In calculations, use the following equivalencies: A=4.0; A-=3.7; B+=3.3; B=3.0; B-=2.7; C+=2.3; C=2.0; C-=1.7; D+=1.3; D=1.0; D-=0.7; F=0.
7 “Required” in this context does not include courses which appear on a menu of courses which satisfy a requirement, such as the 2L writing requirement or the perspective requirement.
8 When required courses are taught during the summer session, grades therein should be normalized with the academic year in which the students are required to take such course. Thus, 1Ls who take Constitutional Law II or Evidence in the summer are normalized against their classmates who will be taking that course the following year. Business Associations and Legal Profession, on the other hand, which can be taken either before or after the second academic year, shall be normalized based on the past year.
Phasing by Trial Period
We propose that we use the 05-06 academic year as both a trial period for this plan (or what emerges from this plan) and a time for consultation with faculty, staff, and students as to its desirability, feasibility, and fairness. It will surely be useful to conduct some forums on it for various groups both to explain the system and to receive feedback on it. It is our belief that as understanding grows as to how this scheme does (and does not) operate, fear of it and resistance to it will diminish both among faculty and students. Certainly it is preferable to “adopt” the system and require compliance with it only when all voices have been heard, all rough places made plane, and as many as possible buy into its spirit and purpose, rather than to impose it prematurely on constituencies ignorant of both its rationale and its functioning.
We would imagine that, once a non-voluntary system were adopted, mechanisms for exceptions and deviations would be put in place. We can imagine two such devices now and suggest them for consideration. (1) A faculty member who wishes to reclassify her course (including making it exempt) before the course is offered at any given time, may do so upon a faculty vote upon a recommendation from the Curriculum Committee. The faculty member, of course, must state the reasons for the reclassification. (2) After the evaluation of student work in a given course is completed, the faculty member may request the Associate Dean for Academic Affairs for permission for slight deviation from any of the normalization standards for “unusual circumstances.”
9 A number of possibilities occur to us here: the section, while not an honors section, has an unusually disproportionate number of honors students in it; the course is unusually small and the rules are an impediment to fairness, not a basis for greater fairness [note: this becomes less of a problem if we exempt courses with fewer than 8 enrollees]; for whatever reason, the class generally performed with a remarkable degree of excellence or of dimness or of similarity, etc.
October, 2005
Curriculum Committee
Scrivener: Bruce Berner
|