Grades, the currency on campus
Could an economic perspective on grades help improve university teaching?
| Photo by Angus Fergusson.
Late can be costly in childcare. When parents arrive late to pick up their children, caregivers are forced to work extra hours – often with no compensation. One solution is to charge late parents a small fine to deter this behaviour. But when economists Uri Gneezy and Aldo Rustichini did just that – introducing a $3 late fee at a group of daycare centres in Haifa, Israel – it had the opposite effect: the late arrivals more than doubled.
In recent years, behavioural economists like Drs. Gneezy and Rustichini have begun to study how people respond to incentives using controlled experiments. Their work reveals a host of psychological effects: we’re risk averse, prone to choking when the stakes are high, and we over-value the present. Different types of incentives – financial, social or moral – can have different effects. As a result, policies often backfire.
In the daycare study, the late penalty failed because it took an important social cost – the desire not to impose on others – and replaced it with a financial cost that parents found quite affordable. Instructors often use a similar policy with grades, deducting five or ten percent of the mark from late assignments. Could this have the perverse effect of encouraging procrastination?
The comparison between grades and financial incentives is not far-fetched. After all, students work for grades, and they trade them for scholarships and a spot in graduate school. Grading practices affect which courses students take and how much they enjoy them.
So, do students treat grades like money? The question remains unanswered, but behavioural economists are starting to experiment in the classroom. In one recent study, Sally Sadoff and her colleagues at the University of Chicago offered trophies and cash to primary and secondary students if they improved their scores on regular standardized tests. Dr. Sadoff and her co-authors found that subtle differences in the reward scheme had a profound effect on student performance. For example, if students had to wait one extra month to receive the reward, its motivating effect was extinguished, suggesting that students discount the future heavily.
This is a serious problem in education, where incentives are almost always delayed. But there may be an upside. According to Dr. Sadoff, now an assistant professor at the University of California in San Diego, it could help us understand why some teaching practices work better than others. For instance, in another study she is looking at the effect of giving monthly financial rewards to high school students based on several measures of performance, including attendance and test scores. These frequent rewards have a lasting effect: the students in the monthly program outperform their peers even after the program ends. Dr. Sadoff says, “You learn more if you put in slow and steady effort.”
It’s a principle Leslie Reid took into account five years ago, when she embarked on a major shift in her teaching. Newly appointed to the Tamaratt Teaching Professorship at the University of Calgary, Dr. Reid began to think about how she could use grades to fuel the learning process more effectively. In her introductory geology course, Dr. Reid replaced one of the two high-stakes midterms with more frequent open-book learning activities. In the new assignments, her students work in groups, interacting with each other right in the lecture hall. These assignments are graded and returned one week later. With more opportunities for feedback, the students are more motivated to engage with the material. At the same time, Dr. Reid gains a better sense of where they stand. It has given her a positive perspective on grading as a mutual feedback exchange. “It can be a win-win,” she says.
Other insights from the economics literature are more surprising. With her collaborators, Dr. Sadoff has found that financial rewards are actually more effective when they are given to students before a test, under the condition that the money must be forfeited if the student fails to perform. This is loss aversion at work: people tend to place a higher value on items that they already own.
Baiting students with cash might not be possible in the university classroom, but there may be other ways to engage loss aversion. One possibility is a system of points with a base level allocated to each student on the first day of class (say 35, or 150). The points could then be used to track participation throughout the semester, with additions for positive contributions and deductions for missed classes. Then, at the end of the course, the final tally could be converted to a grade.
Behavioural economics might also explain why grading sometimes decreases student motivation: attaching an incentive sends a message that one is required. For instance, in a 1997 study researchers asked people in central Switzerland whether they would be willing to accept construction of a nuclear waste storage site in their community. When financial compensation was offered as part of the deal, it actually made people less likely to accept the proposal. The incentive backfired because it signaled that the potential downside would be large.
Similarly, when grades are attached to an activity, it can send a message that the task must be undesirable. Thus it could be especially dangerous to hand out marks for the most interesting tasks – such as attendance in elective courses. Dr. Reid also points out that when students know they’re being graded, it can discourage risk-taking and creativity. “I still philosophically wrestle with how much should be graded and how much shouldn’t be,” she admits.
Yet some grading is necessary. So how can instructors frame grades effectively?
One thing is certain: the evaluation system has to be clear. Changing the game partway through can cause students to perceive inequality – and that will almost certainly lead to problems. In his literature classes at the University of Toronto, associate professor Nick Mount reminds his students what different grades mean in terms of performance relative to the university population as a whole. He also tells students what grades don’t mean, since they are statistical entities subject to error. A 3M National Teaching Fellow, Dr. Mount stresses that instructors should reflect on their methods and talk openly with students about their reasoning. “If you do that,” he says, “they’ll come along for the ride every single time.”
Another helpful approach may be to build a chance for redemption into the grading scheme. Tom MacRae – another 3M Fellow and professor in the department of biology at Dalhousie University – has taught biochemistry and cell biology for more than 25 years. Early in his career, Dr. MacRae restructured the assignments in his courses so that students could resubmit written work after receiving feedback on their first attempt. This flexibility might tap into students’ desire for a fair incentive system; it’s also a way to help students hone crucial skills. “If you don’t give them the opportunity to redo things, you’re wasting your time,” he insists.
Larry Smith incorporates flexibility into his popular macroeconomics course in a slightly different way: he offers a short but challenging bonus assignment as a chance to earn extra grades. It’s difficult enough that only the most disciplined students attempt it, so it doesn’t create the problem of too much additional marking.
Mr. Smith, an adjunct associate professor of economics at the University of Waterloo, also takes the focus off grades by constantly reminding students about another incentive: the broad applicability of the skills they learn in his course on the job market. He even highlights the achievements of former students to drive the message home. This strategy seems to work. In his 30 years of teaching at Waterloo, only three students out of some 30,000 have ever appealed their mark.
And it’s not because his course is easy. While there is a perception among some faculty that awarding high grades will boost their ratings in student evaluations of their teaching, the research indicates that this is not necessarily the case. Studies done in the 1980s by George Howard and Scott Maxwell suggest that when grades and student satisfaction are both high, it’s more likely because students perform well in the courses they enjoy. Dissatisfaction occurs most often when students see no connection between grades and performance.
University of Ottawa psychologist Tracy Vaillancourt recently demonstrated this in an experimental study testing how students respond to an essay grade from an unfamiliar professor figure. When the students received negative feedback that was not based on performance, they retaliated with poor ratings of the professor.
Erin Chown, a recent graduate of Queen’s University, confirms that the psychological rewards of grades are not straightforward. “Getting a good mark from a prof you respect in a course you’re interested in means a lot more than getting a good mark in a course that’s just easy,” she says. “Even a mid-range mark from a tough prof in a great course would mean more.”
For Dr. Reid from U of Calgary, the next frontier is to allow students to build their own assessment path. Some faculty have found innovative ways to do this. In his political theory course at the University of Michigan, Mika LaVaque-Manty has students choose from a set of optional assignments, such as writing an essay or developing a blog, and the students also choose how those assignments will be weighted in their final grade. In his syllabus, Dr. LaVaque-Manty frames this evaluation scheme as a video game with basic required levels (like reading and attendance) followed by optional quests where students can choose their own path.
Most instructors would think twice before using a self-designed evaluation scheme in an introductory course, but Dr. LaVaque-Manty has found that this game-based approach works well in a large class of mostly first- and second-year students. In his view, one of the crucial aspects is that performance is tracked with points, and students receive points for everything they do. “Nothing will pull them down,” he says, so they stop thinking in terms of the traditional subtractive model of grading.
Experimental economist Sally Sadoff cautions that when designing incentives, it’s important to keep in mind that trade-offs are often involved. For example, loss aversion can be an effective motivator, but it may not be optimal for well-being and satisfaction. “Sometimes these things work in different directions,” she says, and the results can be difficult to predict without experimental research.
In the end, instructors who wield grades should try to understand how they work, since incentives have the power to help or hinder learning. “We create and foster these anxieties in students because ultimately it’s good for business,” says U of T’s Dr. Mount, “But that’s not really what they’re there for, and that’s not what I’m there for.”
Roslyn Dakin is a doctoral candidate in biology at Queen’s University and a freelance journalist.
Other stories that might be of interest: