Cheating on MOOCs

This is the first blog post I’ve done as a homework assignment.  For the teacher in my new Canvas class in Understanding Cheating in Online Courses (Bernard Bull, Assistant Vice President of Academics and Associate Professor of Educational Design & Technology at Concordia), has asked those of us who blog to write something about their experiences seeing (or participating) in cheating online.

I’ve touched on the subject of cheating on a MOOC previously when I mentioned one of the key security features of a MOOC course: their worthlessness (or more specifically, their current lack of significant financial value vs. their extremely high value as a means of self-propelled education).

For if the primary value one gets from a MOOC is the self-improvement that comes from independent learning, cheating is not only inappropriate, but imbecilic.

Think about it for a moment.  At least for now, the only thing I’m going to get out of the hours I’ve put into my Coursera Einstein class or my edX Justice class (not to mention the Udacity Statistics class I have every intention of finishing before the end of June) is an understanding of Einstein’s Theory of Special Relativity, comprehension of ethical philosophy, or knowledge of statistical methods.  So if I chose to cheat on any of the assignments associated with those courses, I might “pass” in terms of getting a piece of paper that says I finished the course, but would have failed to become a person who understands these subjects.

The whole notion of the cheater only cheating him- or herself does not necessarily apply to courses where something more than learning is at stake. For when you are taking a class for credit towards something that carries external value (in terms of dollars or prestige), then in addition to cheating yourself, you are also cheating other students to whom you are being being compared work-wise, as well as putting the integrity of the whole system at risk.

But during a period when MOOCs still carry uncertain external value, behaving dishonestly makes no practical sense (especially since students who can’t keep up with the work have other alternatives – such as simply dropping the course – an option that carries little to no social stigma).

So far, I’ve only experienced one incident that might be considered an example of MOOC cheating.  In my Coursera Modernism and Postmodernism class (where grades were based on peer reviewed essays) one of the last essays I was grading (in which we were supposed to look at authors we had read in the context of the work of the contemporary philosopher Richard Rorty) was clearly written for an earlier assignment (one which asked us to talk about earlier writers and thinkers in the context of Freud).

Now it’s entirely possible that the Coursera peer-review system accidentally delivered the wrong essay to me, or that someone carelessly copied and pasted their Freud essay into the field when they thought they were submitting their Rorty assignment.  But if this was done intentionally, the student was probably hoping that a well-written essay (albeit on the wrong topic) would have been given a low but passing grade (especially by a reviewer who may not have recognized it as something the student would have likely written weeks earlier).  In short, they were hoping to game the peer-review system.

Most of the discussion in my Cheating class centers on plagiarism and the various technical tools that can be used to curb it.  But in the MOOC environment, most of us peer-graders do not have access to such tools.  So it’s possible that lots of students were lifting their essays from somewhere when putting together their written assignments.  Which means only obvious slip ups (like the Freud-Rorty switcharoo  – presuming it was intentional and not a human or technical error) stand a chance of being caught.

The fact that only one of the fifteen courses I’ve started since January requires submission of written work points out another security measure against MOOC cheating: a lack of things to actually cheat on.

For example, by the time I’ve finished HarvardX’s Justice course, I will have been required to answer approximately 75-100 relatively easy multiple-choice questions, and respond to approximately 30 prompts on the discussion boards.  And while I’ve made an effort to put 30-60 minutes into my prompted responses, far less time submitting a few quick sentences would have also been accepted.

In other words, the limited number of challenging assessments and other graded assignments built into many (albeit not all) current MOOC classes means there are just not that many corners one needs to cut in order to pass the course.

Now I expect all of this to change once (1) MOOC developers begin to focus on improving the quality and rigor of assessment (rather than consider them an afterthought): and (2) MOOCs begin to gain some external value (either as an alternative to expensive college credits or as credentials recognized and valued by employers).

For at that point, cheating on MOOCs will no longer be irrational, but simply immoral.  Which means that discussions of MOOCs for credit and assessment quality and security need to go hand in hand if the whole system is going to represent anything other than a curious experiment that failed by forgetting to take into account human desire to get something for nothing.

2 Responses to Cheating on MOOCs

  1. Morgan Creighton May 25, 2013 at 4:19 am #

    Thanks for sharing your experiences. I’ve taken about a half dozen MOOCs, mostly from Coursera, and there are two things about cheating that surprise me.

    The first surprise is that so many of the students don’t seem to care at all that others are cheating. I suppose that if one already has a degree and is well credentialed, then certificates don’t matter. So who cares whether someone is devaluing the value of the certificate by cheating. If it’s already worthless, so the reasoning must go, what difference does cheating make?

    Of course, there are a great many people for whom a “real” degree will remain unattainable. And tolerating cheating is a disservice to them. I’m a little disappointed that I haven’t seen a lot of empathy for folks in that position.

    The second surprise is that obvious examples of what I would consider cheating are not seen as such by many, many other students. For example, in several of the programming classes I’ve taken, students will overtly and even proudly post test code they’ve written, and use tests shared by others.

    Well, to me, that’s blatant cheating. In programming, the real work is crafting the tests. Any knucklehead can bang out a program if given a test suite for it. The real learning that goes on is in transforming the prose of the assignment description into an executable specification — test code — that drives the program design.

    But, I didn’t seem to find many that shared that view, so perhaps my thinking is old-fashioned.

  2. Paul Morris May 26, 2013 at 5:32 am #

    Sadly, I have to agree with both the main article and the earlier reply. I have taken (or am currently studying) ten MOOC courses this year and have been aware but not overly concerned about the opportunities to cheat which are inherent in a trust based system. Some students were rather too free in posting specific guidance for graded assessments and some students were far too ready to ask for help before, apparently, making any attempt to tackle assignments but most were inclined to view such behaviour benignly as mere ill-judged excess of enthusiasm.

    I have lately come to be more concerned (or, more accurately, disappointed) at the behaviour of my fellow-students. I recently finished the second part of the edX hosted Berkeley Stat2X sequence presented by Ani Adhikari (an excellent and warmly recommended three-part course) and in the post-exam wash up was reviewing some points which had not been 100% clear. You can imagine that I was unhappy to find that the entire text of the exam questions had been posted online within minutes of the exam opening with pleas for help in solving the problems. Many, presumably, well-intentioned readers came along to show their expertise by providing the required information. Where answers were contradictory (and there was evidently some mischievous posting of deliberately wrong answers) the pleas became increasingly frantic; ‘Please, I must have the right answers soon. the deadline is only an hour away!!!!’.

    On another course, Coursera’s Interactive Programming in Python, widespread cheating became apparent in the peer-assessed coding projects if only because the same unusual error was apparent in a large number of offerings. On inspection and following discussion on the forum it became apparent that many had simply blindly copied a program offered by an ex-student, complete with errors. 40% of the students I assessed had used exactly the same copies code. Some other peer assessors had even higher proportions; one reported 80% as using the same code. My concern in this is not only to wonder why students would bother but also that, without the very characteristic errors, I certainly wouldn’t have noticed any problem.

    Cheating in a for-credit situation while reprehensible is at least understandable. When my own students cheat (or collaborate ‘too closely’) it is usually easy to spot as I know what to expect from each. Cheating in a no-credit learning environment is more difficult to fathom and, given machine grading or peer assessment, far more difficult to identify.

Leave a Reply