Wednesday, September 11, 2013

Cheating: Authoritarian versus design approaches

One surprising conclusion I've reached about students is that many of the behaviors they engage in which seem irrational to us often look very rational from their point of view. Take reading assigned material, for instance. We know full well that many students don't read regularly (and a few don't read at all). No doubt this is shortsighted, but from the average student's point of view, not reading can look pretty rational in the circumstances from a cost-benefit point of view. If you lack the background knowledge to complete the reading, have been given little help in preparing to read, have few strategies for dealing with tough texts, find reading alternately boring or anxiety-producing, cannot discern a connection between reading (or reading carefully) and subsequent academic performance, and can nevertheless pull decent grades without reading regularly, then ... well, why read?

For we instructors, seeing our students as (admittedly flawed) rational actors can make us better teachers. In the case of reading, say, we can make reading a more rational strategy by not 'covering' the reading in lectures, designing assignments that reward careful reading, filling in gaps in students' background knowledge when necessary, and so on.

This observation — that how we teach and the learning environments we create make certain student behaviors rational from their point of view — was in my mind as read this interview with James Lang, the author of a recent book on academic cheating.

Lang puts his main point very nicely:
Cheating is an inappropriate response to a learning environment that’s not working for the student.  Both sides of that sentence are important. It’s inappropriate, which means that we have to hold the student accountable for the dishonest action, and ensure that we maintain high standards of academic integrity.  But it’s equally true that something in that learning environment doesn’t seem to be working for that student. He might see the course as a curricular requirement that means nothing to him; he might be confused by the assignment or see it as busywork; he might see himself as not having the knowledge or skills he needs to complete the assignment. 

Yes, there are students who are genuinely dishonest, who will cheat no matter the context. But most students believe cheating is wrong, and (call me Pollyanna-ish), I think the overwhelming majority of students who cheat would prefer not to. They'd prefer to produce competent, even excellent, work. My observation has been that students cheat when they cannot effectively navigate the learning environments they find themselves in. So cheating is a moral failing. But it's not usually a moral failing in the sense of being evil. It's more like a moral failing borne of desperation, a moral failing that often piggybacks on a litany of prior non-moral failings. As Lang points out, students mostly cheat because they have previously failed to plan, failed to seek help, failed to gauge their own level of understanding properly (thereby succumbing to the all-too-common illusion of comprehension), or failed to be honest about their own shortcomings and lack of preparation or motivation.

Neither I, nor Lang, are advocating we let cheaters off the hook. Rather, Lang's work suggests that certain anti-cheating strategies may be counterproductive. A good bit of what I've heard suggested to forestall cheating adopts what we might call an authoritarian approach. On this approach, we try to make it harder to cheat, make cheating easier to identify, and assign harsh penalties. The authoritarian approach treats cheating as the classroom equivalent of a criminal justice problem. Might the authoritarian approach succeed in reducing cheating? Sure. But it's tiring, frankly, to play this cat-and-mouse game (I set up an exam that's cheat-proof, students identify a way to cheat, etc.), And I wonder if there aren't better approaches, ones that focus on the motivations behind cheating in the first place.

I'd be interested in exploring a non-authoritarian approach, one that, while holding cheaters to firm institutional and ethical standards, tries to design learning environments that make cheating unattractive, unnecessary, and irrational.  Lang hints at a couple of ways we might do this: 
The more choices and control you can give to students over how they will demonstrate their learning to you, the more you nudge them toward mastery learning. By contrast, if you force all students to jump through the same six hoops, you are sending the message that what matters are the hoops, not the learning.
Some students cheat because they have poor metacognition — that is, they have an inaccurate picture of their own understanding of the course material. Typically they overestimate their understanding, which leads them to underprepare. ...Without question, the best means of improving student metacognition is with frequent, low-stakes assessments. Whatever you are going to ask students to do on their graded assessments, give them the opportunity to try smaller, low-stakes versions in class or on homework assignments before they have to ramp up and try for the grade.
So I'd be interested here in a conversation about what teaching techniques, etc., we might use to make cheating an irrational strategy, not because of the penalties, etc., associated with it but because cheating is perceived as unnecessary for performing well academically. Ideas, anyone?


  1. Hooray for the so-called authoritarian approach. Why coddle cheaters, even if they are academically inept: incapable of understanding a college level text or just unwilling to prepare for discussions and exams? Why lower the 'stakes,' i.e. standards to accommodate 'students' who either don't belong in college or won't make an effort at learning? Someone inclined to cheat for ANY reason is a bad actor and deserves to be punished, not only so that he learns a valuable lesson but so our institutional standards are upheld. Don't try telling me to be sympathetic in such as case, as if we were dealing with the academic equivalent of the child who takes food from a store in the midst of a famine.

  2. Robert, lower-stakes assignments aren't lower-standards assignments. They are short, focus on a single skill, repeatable, and often "graded" as a 0 or a 1. These assignments are used to build up a series of skills, whose combination is required in order to succeed at a high-stakes assignment like a paper or an essay exam, which is typically graded in a more fine-grained way (F to A). This is preferable to a practice in which we ask students to acquire skills that we do not teach them and do not give them an opportunity to learn, and then pile all those skills into one major complex assignment, and make that a sizable portion of the grade. Anyone interested in learning would be demoralized by that.


If you wish to use your name and don't have a blogger profile, please mark Name/URL in the list below. You can of course opt for Anonymous, but please keep in mind that multiple anonymous comments on a post are difficult to follow. Thanks!