Dealing with cheating

At the APS Distance Learning conference last weekend there was a session on dealing with cheating. We heard from Dave Pritchard from MIT and Gerd Kortemeyer from Michigan State. Both of them have run courses using online homework (the MSU one is a fully online course) and they had all kinds of data about how students use those systems. I found myself uncomfortable in that session and I wanted to jot some of my thoughts down before I forgot about them.

First, I thought it was interesting that they both pointed out that peer copying is the biggest problem. I’ve always felt that finding online solutions is a bigger problem, so I was interested to hear what they had to say. They showed timing data showing how many students were able to get problems right within seconds of seeing them, drawing a conclusion that they were using the results of a peer. The argument goes that if they’re using other sources, they’d have to spend time reading the problem and going to hunt for solutions, which would take longer than the few seconds they took to complete the problem. What wasn’t clear to me was whether you could analyze the people who took much longer to get an inordinate number of correct answers on the first try.

Second, I was struck by the futility of the “whack-a-mole” approach to deal with cheating. They do this, we do that, repeat. I started thinking about some of the more authentic assessment methods I’ve heard of, and I realized that one of the biggest problems with their mass adoption is their scale-ability. Take my standards-based grading with voice approach: I watch screencasts of students solving problems, assessing their confidence and approach as much as what’s on the page. It comes a close second in authenticity to oral exams, but, while it scales a little better than orals, it can’t scale easily past the 40-person class I’ll have this fall.

Then I started thinking about the impressive project/problem-based learning that so many of my high school teacher friends do. I think that teachers like that would have been a little bewildered by the session, as they don’t worry about things like peer copying, online solutions, and, to a certain extent at least, student motivation. Those students are given an opportunity to do things like launch a weather balloon with a cell phone attached and they jump to it, being involved in the brainstorming, design, execution, analysis, and communication about the project.

That reminds me of the conversations that we have so often in the Global Physics Department, about what the best high school preparation for a student would be. We almost always land on “deep, not broad” because we college teachers have recognized that taking a student who’s seen how science works can be helped with their content deficit better than students who’ve seen everything at a surface/homework level can be helped learn what science really is.

So what can be done with these huge introductory college courses, especially those that use online homework? Well, whack-a-mole is one way. Another is something like standards-based assessment, where homework is not graded, but is rather for learning. A third approach is to try to envision a project-based approach where students are encouraged (possibly in teams) to work toward something that’s easy to motivate and assess, but doesn’t have the cheating enforcement so prevalent. Now, admittedly I don’t teach such huge courses, but I guess I’m just expressing how depressed I was in those sessions. We kept hearing about students who would find ways to defeat the system, and then we’d see statistics showing how it would hurt their learning. THEN we’d learn how, if the researchers shared their findings with the students, MORE cheating took place, as, basically, the sharing simply introduced more students to the tools to cheat with. Ugh. Blah. Give me an oral exam. Or a weather balloon!

So, what do you think? Here are some choices to seed the conversation:

  • Students aren’t motivated to learn, so we have to get them to jump through these homework hoops. Whack-a-mole is where it’s at.
  • What do you mean with whack-a-mole? I love that game at the carnival. Another ugly mole pops up, and I whack it. I love being innovative in how I can predict where that next mole will be, and I know if I whack the one on the left, the next one will be on the right. Fun!
  • Standards based grading, especially with voice, just can’t scale. So lucky you that you have small classes, it just won’t work for me.
  • I think giving points for out-of-class work is unethical. Students can pay for cramster.com‘s premium package that’ll guarantee an eight hour turn around for even teacher-written homework, so the rich kids can get those points for free and the poor ones are left behind.
  • I don’t care if students cheat on homework, they’ll always fail my test if they do so. I give points for homework just to get them to do it.
  • I find the students who help their peers get their homework right are the ones who become the strong majors, and they get practice teaching this way, which is the best way to learn.
  • Other thoughts?
About these ads

About Andy "SuperFly" Rundquist

Associate professor of physics at Hamline.
This entry was posted in glodal physics department, sbar, sbg, screencasting, teaching. Bookmark the permalink.

5 Responses to Dealing with cheating

  1. drmagoo says:

    This is something I’ve thought a lot about. I think homework has value, in that it’s practice needed to develop skills, skills that can’t be developed just by coming to class. There has to be some credit associated with doing it, or too many students won’t do it. If I have to drag the students kicking and screaming through the semester, I will. Of course, I’d prefer they engage willingly, and many do, but not all.

    I very much like the explanations I’ve heard from Ken Heller and his group about comparing learning problem solving to learning a sport, or music, or something that requires daily practice. At the end of the day, I think that the problem solving skills, NOT the physics content knowledge, is what I hope my students will hang onto for years after class. Those skills require practice. I assign small homework assignments for each class period, rather than bigger weekly ones, to give them reason to do daily practice (or at least three times a week).

    I do try to incentivize trying over getting the right answer, by giving them opportunities to get what I call “mulligans,” where they get to choose a few times a semester to just get credit for an assignment that they struggle on. The idea is that I’d rather they tried, get a 0/3, and go through the learning process than cheat out of pressure to get the points.

    I’ve actually noticed cheating much more in upper-lever classes, where I used to see students copying problems from Griffiths’ solution manuals. Since then, I’ve used HW that can’t be found in a book (or at least, not identified with any book), and that’s gone away almost entirely.

    I would love to do more screencasts, but they’re just not practical. I have almost zero time during the week where I can put on headphones and listen to what they’re saying without paying attention to what’s around me – I either have my office door open, in which case I’m interrupted regularly, or I’m at home, and I have kids awake at all hours of the day that I can’t ignore. I may try again more when they’re older and I can get some alone time.

  2. bwfrank says:

    This comment isn’t about cheating per se, but it’s related in the sense of how to get students to practice. We don’t assign homework, but students are expected to work on problems from the book. My plan for the fall is to try to incentivize such practice, because many students don’t practice enough and some not at all. Over the past three semesters in intro physics, I have been using standards-based quizzes rather than “reading quizzes”, to incentivize mastery. With that focus, I have noticed that some students never read or practice problems and just willy nilly take quizzes over and over hoping they’ll pass eventually. So my problem isn’t about cheating, but it’s the same dilemma. Students aren’t learning and practicing as much as they should. So this year, I plan on using homework as “ticket to assess”. It’ll be something like students must show evidence of working on three problems related to the standard–one that was easy for them, one that was of medium difficulty, and one problem that was challenging enough so that they really struggled to solve or could not solve it. [Answers to all questions are given, problems typically increase in difficulty, and my standards typically align with chapters, so this seems feasible]. The homework won’t be graded, and students could certainly still cheat by copying down someone’s else’s effort or they could engage superficially in the homework just to have chance to assess. My plan is to have students somehow report back to me each week which problems they did–and I’ll be able to look at which problems students found easy, medium, hard (and solvable), and hard (and not solvable). I haven’t figured out how I will *collect* or *check* homework engagement, and am happy to hear advice.

    • Andy "SuperFly" Rundquist says:

      This reminds me of how Frank, Sam Shah, and others have a reassess application, where the students have to document the work they’ve done to improve their understanding before being given a new problem to work on.

  3. Joss Ives says:

    The issues of cheating and not wanting to have to bribe students with marks to do homework are what have led me to replacing homework with a practice quiz + frequent quiz model. The big thing here is that the quizzes MUST have the stuff on them that you think are important for them to be practicing. So if the problem solving in end-of-chapter problems is important to you, then those types of problems need to be on the practice quizzes and actual quizzes so that they can completely see how the practice they are (or should be) putting in will potentially be paying off.

    • Andy "SuperFly" Rundquist says:

      I agree about how the assessment needs to match what you care about. I feel that SBG has helped me do that better than my old assessment paradigm.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s