Campuses across the country have been examining, often quite publicly, the state of grades and their meaning. Such studies have been prompted by a perception that grades have been rising over the past 30 years, and particularly during the past 10 years, and that there is a potential shift in the meaning of grades (A, B, etc.). The Educational Policy Committee examined this issue at the UNC-CH campus and presented a report to the Faculty Council at the end of the 1999-2000 academic year (Educational Policy Committee, Grade Inflation at UNC-Chapel Hill: A Report to the Faculty Council, 2000).

Several points drawn from that report help to frame the issues:

(1) There is at least anecdotal, and in many cases concrete, evidence that mean GPAs have been rising at a wide range of colleges and universities.

(2) Mean undergraduate GPAs have been rising similarly at UNC-CH, with two main periods of growth: late 1960s to 1975, and 1987 to present (see Figure 1 of the EPC report). The rise is from a mean of approximately 2.4 to a mean of 3.0.

(3) This increase in mean GPA is accounted for by a significant increase in the awarding of As and Bs and a significant decrease in the awarding of Cs and below.

The EPC concluded that there has been a real increase in grades at UNC-CH during the two periods mentioned above, and this increase is sufficient to warrant examination and possible action on the part of the Faculty Council. They identified several causes with emphasis on those indicating potential problems in grading practices. Finally, they offered a series of possible remedies, ranging from faculty discussion to financial incentives. The present Task Force on Grading Standards was charged with considering the EPC Committee report and extending the analysis to the larger issue of grading practices: how grades are assigned, what these grades mean as judgments of student performance, whether there has been an erosion of grading standards at UNC-CH, and what changes in patterns of grades may indicate.

The members of the Task Force on Grading Standards began their work with the articulation of a basic premise: a grade represents a faculty member's evaluation of a student's work. Thus, grading is the right and responsibility of individual faculty members working with their students in a disciplinary context, guided by their chair and dean.

By long-standing academic tradition, grading decisions are the responsibility of the individual teacher, and at least one court has found that teachers have First Amendment protection with respect to grading. In Parate v. Isibor, 868 F.2d 821 (6th Cir. 1989), the court held that "[B]ecause the assignment of a letter grade is symbolic communication intended to send a specific message to the student, the individual professor's communicative act is entitled to some measure of First Amendment protection. . . Although the individual professor does not escape the reasonable review of university officials in the assignment of grades, s/he should remain free to decide, according to his/her own professional judgment, what grades to assign and what grades not to assign. "

We note two meanings of "grade inflation," a term that recurs in casual conversations and in the EPC report. One is a descriptive term indicating that grades have been rising. It is clear that at least mean GPAs are being inflated in this descriptive sense. The second meaning is a more normative, or even pejorative, term indicating that the "value" of a grade has eroded (i.e. an A today does not mean as much as an A yesterday). Given this second meaning, the current report uses the term "grading standards." Our charge, then, has been to determine whether standards by which grades are assigned at UNC-CH have been well defined; whether these standards are being followed in grading practices on campus; and whether the standards need to be changed due either to shifts in the meaning of grades or to a lack of clarity in definitions.

To address the charge, the Grading Standards Committee considered the following questions:

(1) Has there been an historical increase in grades assigned at UNC-CH? Although it is clear from the EPC report that at least some measures of grades have increased, it is less clear whether this pattern holds when separate analyses are performed.

(2) If there has been an increase, what might have caused that increase?

(3) Is this cause a matter of concern?

(4) If there is a cause for concern, what are the possible solutions?

(5) What are the implications of these solutions?

It is likely that a variety of factors contribute to rising grade averages since grading is such a complex process. Some of these factors are benign, some may be malignant, but no single theory seems adequate to explain the phenomenon. For example, over the past 25 years, there have been a number of demographic and educational shifts that may affect student performance and grading practices:

(1) Students are taking longer to complete their degrees, and some students no doubt take lower course loads so they can focus their energies on fewer courses and earn higher grades. Many students also schedule their coursework in ways to maximize their chances for better grades, taking one or two difficult courses in a semester in which they also enroll in several less demanding courses.

(2) The gender mix of the undergraduate student body has changed over the last 25 years. The proportion of female undergraduates has grown steadily, and this year it is over 60%. Females have higher SAT scores and they generally perform better at academic tasks (See e.g. recent report on grades by members of sororities as compared to fraternities). Moreover, they tend to major in the departments in which grades have risen most dramatically.

(3) SAT scores for incoming UNC students have risen dramatically in the last 20 years (even allowing for the "recalibration" of scores by ETS), so it is possible that students are better prepared to do college work than their predecessors.

(4) Changes in pedagogy may also have an effect on grades. The lecture method, which dominated instruction 25 years ago, has given way to a variety of active learning approaches in many classrooms, especially in the humanities and social sciences. Empirical research has demonstrated the superiority of these methods for teaching higher order intellectual skills, and it is possible that students perform better because they are being taught better. (See, for example, McKeachie, W., Pintrich, P., Lin, Y., & Smith, D. (1986). Teaching and learning in the college classroom: A review of the research literature. Ann Arbor, University of Michigan, National Center for Research to Improve Post secondary Teaching and Learning.)
It is also possible that teachers who have changed their pedagogy and adopted active learning methods may persist in using grading approaches more appropriate to their previous goals. In these cases, student grades could rise because students can achieve these goals more easily via active learning.

(5) For several reasons, an increase in class size over the last 25 years may also have had an effect on grades. Grading in large classes is often shared with teaching assistants (or is entirely their responsibility). Inexperienced instructors have more difficulty with grading, and tend to give students "the benefit of the doubt." It is extremely difficult for a faculty member to insure that grading criteria and standards are uniform across all the As in a course. Also, the tendency in large classes to resort to more easily graded assignments and exams may inadvertently make these exercises easier for students (and hence yield higher grades).

(6) Over the last 25 years, it has become more commonplace for students simply to drop courses in which they are not doing well. Indeed, faculty members have been encouraged to "counsel students out" if they are in trouble. In essence, these practices cut off the bottom of the grade distribution, inevitably yielding a higher overall average.

(7) Every individual class is unique. Certain courses are designed to instruct students with higher grade point averages. One might expect, for example, that grades in honors courses will be higher than those in other kinds of courses, but even a more typical kind of course can have an exceptionally large number of excellent students in a given semester.

(8) Students must maintain a 2.0 GPA in order to stay enrolled at UNC. If one assumes a normal distribution in grades, then the ends of this distribution must be C (2.0) and A (4.0), not F (0.0) and A (4.0), and the median should be 3.0, not 2.0.

(9) Some faculty members have suggested that there is a link between rising grades and the use of student evaluations, but empirical studies fail to show any causative relationship. Several studies discount the likelihood of rising grades resulting from instructors trying to "buy" better student ratings of instruction. Still, many faculty members believe that there is widespread manipulation of grades. An extensive study involving 28,629 undergraduates over a year period found no relationship between frequency of evaluation and 2 indices of grade inflation: average end of term grades and student ratings of course workload.(See Franklin, Jennifer & Theall, Michael (1991) Grade inflation and student ratings: A closer look. Paper presented at the Annual Meeting of the American Educational Research Association (Chicago, IL, April 3-7))

Moreover, at research universities student evaluations play only a minor role in promotion, tenure, and merit pay decisions in most departments. The reward system weighs research productivity and grantsmanship far more than teaching, so it is especially unlikely that the advent of the Carolina Course Review had much of an impact on grading practices at UNC, anecdotal evidence notwithstanding.

(10) Data on grades at UNC (and other institutions) show a consistent pattern of higher grades in the humanities and social sciences than in the natural sciences and mathematics. Some investigators have suggested that the explanation for this pattern can be traced to teaching practices in these disciplines. Courses in the natural sciences tend to focus on facts and concepts, use lectures as the primary teaching method, and base course grades primarily on exams. Courses in the social sciences and humanities focus on the development of intellectual skills, use discussion as the primary teaching method, and base course grades on essays, exams, and other written work. Students in these courses receive frequent feedback on their performance so they can improve their work and tend to receive higher grades as a consequence.

(11) It is clear from empirical studies that grading practices vary from teacher to teacher and department to department. It is also true that most faculty members have no training in testing and measurement, and may therefore follow practices that (taken collectively) can affect the grade distribution across an entire institution. Their beliefs about the function of grades also play an important role in how they use grades in their courses. For example, some teachers believe that grades are an important motivational tool, so they use grades as incentives to encourage particular student behavior, such as class attendance or class participation. Grades for improvement or "effort" are also used to mold students' behavior and reward them for application to their studies. These practices may result in higher grades, but they may also promote better learning habits and more effective learning. Teachers who have adopted cooperative learning techniques in their courses may have higher grade distributions because grades are assigned collectively. However, since students help one another learn the subject matter, higher grade distributions may reflect better learning. Other teachers see the grading system as a way to sort students into ability categories and to identify the "best" students. In this case, teachers may use grading practices that insure a wide distribution of grades, such as asking questions on exams that only a few students can correctly answer or using the "best" student papers as the standard for performance (rather than basing performance standards on independent criteria). These practices tend to result in course grades that resemble a normal distribution and are relatively stable over time, but may not accurately reflect student learning in the course. Some teachers find testing and grading extremely difficult, and are not confident that their exams or assignments provide accurate assessments of student performance. When students perform much worse or much better than expected, they will "curve" the grades up or down to create an artificial distribution that they feel is more reasonable. This practice divorces grades from actual student performance and may result in grades that either overestimate or underestimate student achievement in the course.

These (and other) grading practices no doubt affect the undergraduate grade distribution and may be a significant cause of the overall rise in grades across the institution.

A trend toward higher grades at UNC has implications for faculty members, for the university as a whole, and for students.

For faculty members, a trend toward rising grades suggests that faculty members should evaluate their own expectations and evaluations of student work. On one hand, increasing numbers of As and Bs might mean that faculty members are no longer able to distinguish degrees of excellence in student performance. If this is the case, faculty should consider redefining or more clearly communicating their expectations to students in order to be sure that grading practices are meaningful and consistent. For example, a course syllabus should set forth clear, explicit criteria and standards for course grades. On the other hand, increasing numbers of As and Bs might mean that students are performing more conscientiously and fulfilling the requirements for high achievement in their courses. The decision as to whether course requirements are appropriate for course levels rests with the faculty member, in the context of his or her own department.

For the university as a whole, a trend toward rising grades suggests that the university should evaluate its position with respect to its peer institutions. On one hand, increasing numbers of As and Bs might mean that UNC is attracting students who are performing at a higher level than was the case in the past. The Admissions Office can evaluate the selectivity of its admissions decisions and the level of promise and preparation of admitted students; the Dean's Office can evaluate practices such as counseling and course drop policy that might have an impact on student performance and grade levels. On the other hand, the university might be concerned about its credibility in the wider academic world, and question whether its degrees have become devalued because more students are graduating with higher averages than was the case in the past. College deans can monitor the comparative standing of UNC grade levels with those of other schools and programs in UNC's peer group. A dean is in the position to determine whether his or her college, or a program within that college, is experiencing rising grades at a level that is unusual among peer institutions.

For students, the trend toward rising grades suggests that students should evaluate the uses to which their grades may be put after graduation. Grades record achievement, and they are also used as predictors of success for post-collegiate aspirations. According to Marcia Harris, Director of University Career Services, employers often use grades as a screening device by interviewers or recruiters, the most competitive of whom choose to consider only candidates whose grade point average is above a certain mark, sometimes 3.0, and sometimes 3.5. For graduate and professional schools, grades are part of the admissions mix, along with standardized tests, recommendations, and other demonstrations of skill (such as writing samples, scientific papers, or other projects). Students who are competing for postgraduate selection obviously hope their own grades will be high, but they also hope that employers and admissions officers appreciate the context in which those grades have been earned. Higher individual grades might advantage particular students, but higher aggregate grades might disadvantage them as well.

Because the increase in student grade point averages can be attributed to a wide variety of causes, as we have attempted to define, and because no two classes are ever exactly alike in the quality of students' work over the course of a semester, we find it difficult to assign in an all-inclusive way a single cause for higher grades among our students. Likewise, it is hard to determine a single way to address grading practices.

We acknowledge that it is possible that there may be isolated instances where an individual instructor exaggerates the quality of a student's work in grading. If such instances occur, it should be left to the individual units of the University (departments, curricula, programs, etc.) to ascertain if such exaggerations do, in fact, exist. Only within a given unit can the nature of the course and its goals, the pedagogy employed, and the nature of assigned work and examinations be evaluated in light of grades administered. As a result, our recommendations to the Faculty Council are directed toward the various educational units within the University. We propose that the Faculty Council forward to these units the following directives:

(l) The 2000 Educational Policy Committee Report describes the meaning of grades as follows:

"A": Outstanding mastery of course material

"B": Superior mastery of course material

"C": Adequate mastery of course material

"D": Mastery of course material that is unsatisfactory or poor

"F": Unsatisfactory mastery of course material

The faculty of each educational unit within the University should review in a formal meeting these standards in the light of the disciplines and the pedagogies by which they are taught. The goal should be to arrive at a consensus within that unit about what constitutes conformity with the standards described above. Within individual units, discussion of grading practices can lead to more consistency and greater attention to the importance of fair and well-defined grading standards. A thorough and thoughtful discussion of grading practices within every unit can strengthen faculty performance, and clearer definitions of grading standards will benefit every unit. An articulation of grading standards and criteria--expressed in every syllabus--will likewise benefit every student.

(2) For graduate teaching assistants and other part-time or temporary instructors within the units, such a review should occur on an annual basis. In addition, part of the ordinary review within the units of the teaching performance of graduate teaching assistants and other part-time instructors should include a review of the instructor¹s grading patterns.

(3) The chairs and heads of each unit should review the grading patterns of the individual faculty on an annual basis. Such evaluations of faculty should review grading patterns as they are established over several semesters. Faculty experiencing difficulty with grading may be referred to the Center for Teaching and Learning. Chairs might also consult the Center about workshops on grading for their entire faculty.

(4) Deans should require in the annual report of each educational unit of the University a general summary of grading patterns within that unit--not for the purpose of singling out the grading patterns of individual faculty members by name, but to insure that reviews of grading patterns within the unit occur on a regular basis.

(5) The Registrar should distribute annually to deans and chairs a summary of grades assigned within all units. Such disclosure can help members of the university community be sure that their own grading standards are in harmony with the expectations of other units. At the same time, UNC can position itself with regard to peer institutions, who may be experiencing rising grades at similar or greater rates. Employers and admissions officers can draw on the information to help contextualize the grades presented to them by UNC students.

Submitted by:

Douglas Crawford-Brown
Carol Fowler Durham
Wendy Haines
Matt Hamilton
Joy Kasson
George Lensing
Beverly Whitaker Long, Chair
Ed Neal
Annie Peirce