Developing Grading Criteria for Writing Assignments
Trying to decide the relative merits of a piece of writing can lead to a tangle of problems. Given a set of student essays, instructors frequently disagree, often vehemently, with one another’s assessments. Because we teachers have little opportunity to discuss grading practices with colleagues, we often develop personal criteria that can seem eccentric to others.
Professional writing teachers grant that the assessment of writing, like the assessment of any art, involves subjective judgments. But the situation is not entirely relative either, for objective standards for good writing can be formulated, and readers with different tastes can given the opportunity to confer and develop criteria assess writing samples with surprisingly high correlation. But the potential for wide disagreement about what constitutes good writing is a factor with which both students and teachers must contend.
Even though professors may confer and agree upon criteria for student essays, these criteria often vary from discipline to discipline (and of course from teacher to teacher), a phenomenon that often confuses students. Not only do styles vary widely across the disciplines, but there are also fundamental differences in the way arguments are structured and elaborated a problem students feel acutely as they move through their general education courses. To make matters more confusing for students, different teachers within the same discipline often value different kinds of writing. Some teachers want students to sound like professionals in the field. Others assign narratives, personal reflections, and other alternative assignments calling for voices other than the apprentice academic. Because of such a variety of expectations, instructors should describe their criteria for judging writing, and, whenever possible, provide samples of successful student papers from previous classes.
Criteria for writing are usually presented to students in one of two ways: analytically or holistically. The analytical method gives separate scores for each criterion for example, ideas, ten points; organization, ten points, sentence structure, five points whereas the holistic method gives one score that reflects the professor’s overall impression of the paper, considering all criteria at once. Many instructors prefer analytic scales because the breakdown of the grade into components, when combined with the instructor’s written comments, conveys detailed information about the teacher’s judgment of the essay. Some people object philosophically to analytic scoring, however, on the grounds that writing cannot be analyzed into component parts. Can ideas really be separated from organization, or clarity of expression from clarity of thought? Also, holistic grading is faster and so is often preferable when one’s main concern is rapidity of assessment rather than precision of feedback.
Both analytic and holistic scoring methods can also be classified two ways: general description methods and primary trait methods. Proponents of general description argue that criteria for writing can be stated in a general or universal way (good organization, graceful sentence structure, and so forth). Proponents of the primary trait method, however, argue that criteria must be stated specifically in terms of the given writing task. For example, the criteria for a music history paper detailing the early history of opera (including effective use of sources and a clear explanation of historical circumstances) would differ from those of a political science paper arguing that the electoral college should be abolished (including empirical data and the debate that surrounds the topic). Thus, a primary trait scale uses grading criteria keyed directly to the assignment.
A good way to refine one’s grading practices is to confer with colleagues about what constitutes excellent, good, satisfactory, and poor papers. A surefire way to simulate such conversations is to “staff-grade” with colleagues a set of essays written in response to an assignment within your discipline. One participant selects in advance four or five essays that seem to span the range of quality from excellent to poor, duplicates them for their colleagues, and uses them to initiate discussion. In developing criteria, instructors are advised to use a number scale that does not translate directly into letter grades. A six-point scale ranging from 6 (best) to 1 (worst) is most common. Using a numerical scale temporarily suspends the additional problem of variable standards for letter grades. Thus, a “hard grader” and an “easy grader” might agree that a particular essay rates a 4 on a six-point scale but disagree on how to translate that 4 into a letter grade. The hard instructor might give it a C+ and the easy instructor a B. Since standards for letter grades are a different issue from standards for ranking several pieces of writing, problems of devising criteria for writing are simplified if the two issues are separated, at least initially. After department members develop criteria for each gradation on their scoring scale, professors can discuss the relationship of scores and grades. A departmental “norming” session session every year or so can increase instructors’ sense of mutual agreement in their grading practices, and can help each professor feel confident that their grading standards are shared by their peers, especially if students question grades.
Assigning a letter grade to a piece of writing always poses a dilemma. Teachers who use analytic sales often add up each student’s total score, rank the papers, and translate scores into letter grades by establishing a curve or by setting point ranges for levels of grades. Other teachers, using a more holistic method, try to develop an interior sense of what an A, B, C, or D essay looks like. If possible, it is best to read through a set of papers quickly before marking them and assigning grades, trying to get a feel for the range of responses and sizing up what the best papers are like. In grading essay exams or short papers, many teachers develop schemes for not knowing who the authors are until the papers are graded, such as assigning a number to each student, having a cover sheet with both the students’ names and numbers, and then having the students write just their numbers on the remaining pages. Not knowing who wrote which essay eliminates any halo effect that might bias the grade.
When students know an instructor’s criteria for assigning grades, the quality of their final products will improve gratifyingly. It is satisfying indeed to see how well many undergraduates can write when they are engaged in their projects and follow the stages of the writing process through formulating a thesis and employing a process of revision. By setting high standards, by encouraging multiple drafts, and by encouraging students to seek help from one another through discussion and peer review (and, if necessary, from College writing centers) in short, by expecting excellence instructors can feel justified in applying rigorous criteria. Professors can also consider allowing students to improve a paper if they wish by rewriting it. The presence of grades should never override the more important emphasis on revision and improvement.
The point, then, of assigning writing across the curriculum is to engage students in the process of inquiry and active learning. Although one of our goals is to improve students’ communication skills, writing is more than communication; it is a means of learning, thinking, discovering, and seeing. When teachers give students good problems to think about and involve them actively in the process of solving these problems they are deepening students’ engagement with the subject matter and promoting their intellectual growth. Well-designed writing assignments give students continued practice in critical thinking. Teachers know when their approach is working: the performance of their students improves.