What’s in a Grade? It Depends on Whom You Ask

How precisely can professors evaluate students’ work? That question was at the heart of a recent debate over whether to change the grading system at Eastern Washington University.

Eastern Washington had long awarded course grades on a 4.0 system, in which grades are given to the tenths decimal place, offering many more options for professors than the more-conventional letter system. Instead of awarding an A or an A-minus, for instance, professors might give a 3.7, 3.8, 3.9, or 4.0.

After a debate that played out largely along disciplinary lines, the Academic Senate voted this spring to switch to letter grades. The university expects to roll out the change in the fall of 2018.

This might sound like a minor development. But the debate led Eastern Washington’s faculty to grapple with questions that often go unexamined. How precise can grades be? How consistently do professors apply standards? And what purpose does a grade really serve?

Eastern Washington had considered switching to letter grades before. The latest debate was sparked by the arrival of a new provost, who wondered why the university recorded grades differently from most colleges. Decimal-point course grades are uncommon enough that they weren’t even included as a possible choice in a recent survey on grading conducted by the American Association of Collegiate Registrars and Admissions Officers for a forthcoming book. That said, the approach does have some regional credibility: It is also used by the University of Washington.

The decimal system was popular with some professors, particularly in the STEM disciplines of science, technology, engineering, and mathematics, who liked having the ability to make fine distinctions, and felt that the number and nature of assignments they gave could back them up.

One strong proponent of the decimal system is Brian Houser, a professor of physics. In his classes, Mr. Houser sets a total number of points that can be earned and subtracts them when students make “small mistakes.” Mr. Houser then converts the points students earned out of the total possible into a decimal course grade, which he thinks did a good job of capturing granular differences in their performance.

Mr. Houser says that professors in a computational field like his want to give students partial credit for their correct thinking on a problem even if they get the final answer wrong. The previous grading system also gives instructors a way to dock points if students arrive at the right number but leave off the units or mix up whether the figure is positive or negative.

It’s important that students learn to be careful about such things, Mr. Houser says. The decimal system allows him to give students a grade that’s lower — but not too much lower — because of the accumulation of such errors. “Small mistakes,” he says, “are magnified when you have a coarser grading system.”

That, he says, is unfair to students: “If I call them small mistakes, I think they should have a small impact on the grade.”

But some professors worried that Eastern Washington’s system was putting its top students at a disadvantage for graduate-school admissions. A student who earned straight As at some other college would graduate with a perfect 4.0. But an Eastern Washington student, the argument went, who earned 3.8s and 3.9s — which are presumably equivalent to As — would finish with a lower GPA. In admissions to a competitive graduate program, they thought, this could make a difference. And besides, not everyone was convinced that fine distinctions between grades were meaningful.

A Pretense of Precision?

Before determining how much precision is possible in a measurement, it helps to know what is being measured. But there is “no shared agreement of what goes into a grade,” says Natasha Jankowski, director of the National Institute for Learning Outcomes Assessment.

Sure, most professors would probably say that the grades they award reflect the extent to which students have learned the material. But grades often rest in part on student behavior — showing up to class on time, for instance — which muddies their meaning, Ms. Jankowski says.

Even if professors isolate student learning from student behavior, only so much precision is possible, Ms. Jankowski says. That’s true even on objective assignments, she says, since professors must use their judgment to interpret the thinking that led students to their answer.

At Eastern Washington, faculty members’ perceptions of precision were key to whether they liked the grading system. “It didn’t seem like one of those debates where anyone was wrong,” says Mark Baldwin, associate vice president for undergraduate policy and planning. Some faculty felt they could distinguish 3.1 work from 3.2 work, he says. Others, concentrated in the humanities and social sciences, did not.

Before coming to Eastern Washington, Kevin S. Decker, a philosophy professor, was used to letter grades. But he liked the precision possible under the decimal system, and sometimes also used decimal grades on the individual assignments in his classes. This would lead to questions from students about why they got a particular grade, which Mr. Decker welcomed. He found that decimal grades forced him to really justify why he was giving one grade and not another.

Lynn Briggs, dean of the university college and a professor of English, also found that the decimal system invited lots of pushback from students. But she was less certain that the grades she awarded under it held up to scrutiny. Ms. Briggs found herself explaining small differences in grades to students but thinking, “Frankly, if you were in a different class and I were comparing you to different students, it might have come down differently.” The grading system gave students the wrong idea about what her grades meant, Ms. Briggs says. “It always felt like a pretense,” she says, “to have it come down to the decimal.”

Terrel Rhodes, vice president for the office of quality, curriculum, and assessment at the Association of American Colleges & Universities, says that very fine distinctions between grades might make sense for some kinds of work, like tests of factual knowledge. But he doesn’t think they are a good fit for grading more-interpretive work that tends to require higher-order thinking, like an essay or a proof.

Letter grades — especially with the option to award plusses and minuses — allow a professor to make distinctions as fine as are called for, Mr. Rhodes says. “That’s probably as far as most unpacking of students’ performance can go.”

But having a sensible grading system is only a first step, Mr. Rhodes adds. The “power” of a grading system, whether decimals or letter grades, “rests on how broadly shared the expectations are for what each of those designations mean.”

What Grades Mean

It turns out that grading can be personal and idiosyncratic. This should come as no surprise to academics, says Linda B. Nilson, director emerita of the office of teaching effectiveness and innovation at Clemson University and author of a book laying out an alternative grading system. They know it from their own experience. “What happens when you send in a manuscript to a journal?,” she asks. Each reviewer has a different reaction. And that’s in a setting in which professionals are evaluating the work of another professional. There is much more variation in quality when it comes to the work of undergraduates, Ms. Nilson adds, which makes it that much harder to evaluate their work according to a consistent set of standards.

Jackie Coomes was deeply involved in the grading discussion at Eastern Washington. She is the immediate past president of the Faculty Organization, which includes all Eastern Washington faculty. In that position, she served as chair of the Academic Senate, the committee that voted on the grading change. Now she wishes professors would have talked longer about how grading works in their disciplines and the purpose grades serve.

Ms. Coomes, a professor of mathematics education, says that her students who would have earned a 3.8 or a 3.9 under the old system will now probably get an A-minus. That makes Ms. Coomes sound like a tough grader, until you consider that she has typically awarded a 4.0 to students who earn 93 out of 100 points.

An even bigger issue, though, is what kind of mastery professors believe the various available grades signify. Ms. Coomes recalls a conversation she once had with a colleague in which she explained what an A, B, and C mean to her. Her C was a student who could practice and memorize procedures but didn’t demonstrate conceptual understanding. To her colleague, that was a B.

Mr. Baldwin, the associate vice president, would also have liked to see the faculty discuss grades more. Some of their implications, he thinks, were overlooked. One thing he thinks the faculty could have thought more about is how grades are perceived by students, especially those without a family history of college enrollment. Some faculty may like the ability to make fine-grained distinctions. But to students, he says, that can lead to the impression that any mistake will be punished with a lower grade. The university’s academic advisers, he says, see the new system as an improvement.

So does Ms. Briggs. Some students may like getting granular feedback from their professors, she says. But fixating on precision might not be the best preparation for life after college. “For the rest of their lives,” Ms. Briggs says, “their work is going to be judged in a subjective way — even if they are doing objective tasks.”

Beckie Supiano writes about college affordability, the job market for new graduates, and professional schools, among other things. Follow her on Twitter @becksup, or drop her a line at beckie.supiano@chronicle.com.