Why “Blind” Grading Makes Good Sense, and Should Be Used More Extensively Outside of the Context of Law School Exams

Updated:
Posted in: Education

Two activities that occupy a lot of time for law professors during the winter break are grading exams and visiting with family and friends —including high-school-age kids, and college students who are home for the holidays.  As I was doing both of these things (in turn, not simultaneously) over the last month, it struck me that the “blind” grading that is currently the norm in law school exams—in which exams have school-issued identification numbers but no names on them—is uncommon in American colleges (or at least at the research universities from which the recent college students I know and teach come from), and almost unheard of in high school.  In the space below, I briefly discuss the great advantages (and occasional disadvantages) of blind grading, and suggest that colleges and even high schools consider making wider use of it.

The Origins and Intuitive Appeal of Law School Blind Grading

Some legal historians have linked the prevalence of anonymous grading in law schools to the proliferation, beginning in the 1970s, of race-based affirmative action programs in law school admissions. Interestingly enough, some of these scholars suggest that blind grading was deployed not so much to prevent discrimination against (otherwise identifiable) minority-group students, but rather to prevent faculty members who support affirmative action from affording a grading preference in favor of minority students.  This counterintuitive notion points up the fact that purposeful or subconscious bias need not be driven by animus, but also can stem from positive feelings that a grader has towards an individual or group.  Indeed, blind grading may be just as important in preventing teachers from rewarding “favorites” as it is in precluding graders from punishing “troublemakers.”  (And in a world where all grades are “relative”—whether or not a class is graded on a curve—in the sense that GPAs are compared across the student body, boosting one student’s grade is really not so different than knocking another’s down.)

Nor are the characteristics that may generate grading bias limited to immutable characteristics like race, gender, sexual orientation, religion, etc.  Some of the most common grading bias may involve what I will call here an “expectation effect.”  Many of us can recall teachers for whom the first graded assignment or two in a course (say, English papers) would create a perception of student ability that could then be hard to dislodge, perhaps especially in the downward direction.  That is, getting an “A” on the first few assignments could create a presumption in the mind of the teacher that subsequent work would also be in the “A” range, even when it turned out to be of inferior quality. And that unconscious presumption could afford a student the benefit of significant doubt down the grading road.

Indeed, performance expectations that affect a teacher’s grading can be generated not just from group stereotypes or individual past performance, but also by family track records.  In K-12, when teachers often teach older and younger siblings within the same family, it is hard to avoid assessing a student’s work without memory of or reference to how well an older sibling performed in the not-so-distant past.

Some Non-Obvious Benefits of Blind Grading

Protecting students from biased assessments of their work and the work of their peers (with whom they are competing) may be the most obvious advantage of blind grading, but it is not the only one.  Protecting teachers from accusations of bias is also important.  (Reflect back on the affirmative-action connection to blind grading.  Absent blind grading, if minority students got weaker grades than other students, some faculty would be more open to charges of old-fashioned discrimination.  Designing tests that may be racially or culturally insensitive is one thing, but that is a different kind of accusation than singling out students of color for harsher evaluation.)

And blind grading also gives teachers (and students) more credibility when teachers want to endorse or support students who have done well in a class for admission to higher levels of education or for jobs.  A letter of recommendation discussing a student’s excellent performance on a blindly graded exam has particular credibility, especially when there might otherwise be a reason to suspect bias by a teacher in favor of that particular student.  Consider, for example, students whose parents are prominent alums or donors, or students of color at a school whose faculty is known for supporting diversity.  Absent success on blindly graded exams, these students’ excellent performances may get discounted by outside evaluators.

Possible, But Ultimately Not Powerful, Limitations On Or Drawbacks To Blind Grading

So what are the downsides of blind grading?  Some have suggested that blind grading creates an incentive to have fewer midterms and other graded assignments prior to a final exam, because blind grading means that, in the wake of every blindly graded assignment that is turned in, students have to be admonished to preserve the confidentiality of exam-taker identity until the exams are graded, which in turn chills student-faculty interaction.  I personally don’t find that to be a very powerful argument against blind grading; most students are savvy enough to be able to talk to faculty about the substance of a course, even before an assignment that has been turned in has been graded and returned, but to avoid saying anything that would betray whose exam was whose.

Another argument against blind grading is that this method makes it harder for a teacher to reward classroom participation (which, of course, is not generally anonymous.)  But teachers can be given the power to adjust an overall course grade to reflect classroom performance after the written work has been graded blindly, and those written-work grades have been recorded.  “Unblinding” teachers to take classroom participation into account only at that point preserves the benefits of blind grading, and also makes transparent just how much work the evaluation of classroom performance is doing in the overall grade.

Obviously, blind grading cannot be used in classes where each student’s projects are unique.  If every student is writing a paper on a different topic, and one as to which there have been consultations between the student and the teacher, then blind grading is impossible.   But most papers in high school and college, and certainly most exams at both levels, are ones in which all the students are working from the same prompts or questions.  This is true not just in social science classes, but also in math classes. Some might think that test answers in math-type exams are either “right” or “wrong,” so that grading subjectivity is beside the point, but “partial credit” can be a somewhat idiosyncratic and contested concept.

The mention of math classes does raise a possible problem with blind grading, however.  In a world where many papers and exams were hand-written (as they still may be in Math, Science and Engineering exams), the use of blind grading in small classes, especially where students have completed many assignments, might be a bit of a sham.  If the grader is able to discern, over time, what the handwriting of some students looks like, then the seemingly anonymous grading isn’t so anonymous at all.  And it may be worse to have “blind” grading that isn’t in fact blind than it is to have old-fashioned non-anonymous grading.  At least then we wouldn’t have a pretext of complete impartiality that can be used to cover up bias.  Potential bias is bad enough; potential bias in a system that holds itself out as one in which bias is impossible is worse still. But the vast majority of college and high school papers, and the majority of college papers and exams (outside the STEM fields at least) are now typed and/or involve a large number of students, so that handwriting identification is unlikely to be a problem.

Another limitation on blind grading might be that, until relatively recently, technology made the use of this method of evaluation a hassle.  But these days no one can really argue that it is technologically difficult for registrars at all levels to assign students ID numbers to be used (and changed) for each graded exam or paper.

At the end of the day, I can’t really think of any truly powerful reasons not to have blind grading used more broadly at lower education levels.  (After all, people in other disciplines that employ the scientific method well understand the importance of “blindness” in conducting research and evaluating results.) My guess as to why it is so intuitive for law professors to use blind grading (even in law schools that no longer are allowed to make use of affirmative action) is that law professors have been trained to think, and have instincts about, due process.  I would never be one to say that law schools get everything right when it comes to pedagogical matters, but blind grading is one area where I think that others could learn from our example.

Posted in: Education

Comments are closed.