Yet Another Survey Shows Cheating is More Common Online
Plus, cheating is up at Cal Poly. Plus, Pearson wants a Chegg detector. Plus, buckle up.
Issue 73
Research: Cheating in Online Exams is Nearly Double the Rate of On-site Exams
New - and in my view, great - research shows deeply pervasive cheating and a very strong connection between cheating behaviors and online exams. Stefan Jankea and Änne Petersen of the University of Mannheim, Tanja M. Fritz and Martin Daumiller of University of Augsburg and Selma C. Rudert of the University of Koblenz and Landau - all in Germany - are the authors.
Some background - it’s a survey of German college students with a sample of more than 1,600 - undergrad and grad - across a mix of institution types.
What I find great about the survey is it asked students to report how often they’d engaged in specific behaviors without specific reference to whether those behaviors were cheating. That’s important because we know that students rationalize misconduct and do not, when asked, acknowledge it as cheating. Moreover, this survey asked about a robust 19 different behaviors on a more nuanced 7-point scale. The survey also isolates online tests versus on-site tests, as well as primary mode of instruction.
Finally, the survey was distributed through institutions, using student e-mail addresses. And while this may disincentivize responses regarding misconduct, I find it more reliable than external, commercial survey samples that ask “students.”
And with the usual disclaimers about self-reported misconduct being under-reported, which the authors of the report acknowledge, the research found:
Overall, only 4.9% of the participants reported that they had not engaged in any of the investigated behaviors during the summer semester 2020 at all, whereas 12.4% indicated that they had engaged in only one of the behaviors. In contrast, the vast majority of participants (82.7%) indicated that they had engaged in multiple behaviors during the time in question
Strawberry popsicle stick - only 5% did not engage in behavior that most would consider to be academic dishonesty.
Even removing the five most innocuous of behaviors - things such as making up an excuse to ask for a deadline extension - the results were still stunning:
Still, only 28.6% of the students indicated that they had not engaged in any of the remaining, more severe, behaviors during the past semester and another 18.2% indicated that they had engaged in only one of the behaviors. In contrast, the majority of the participants (53.2%) indicated that they had engaged in multiple remaining behaviors during the time in question
Entire truckload of strawberry popsicle sticks - more than half of students admitted engaging in the “more severe” cheating behaviors.
Further, the authors say there is a definite, quantifiable link between cheating and online testing:
31.7% of all students who had written on-site exams indicated that they had used unallowed assistance and/or engaged in direct exchange with other students during the assessment. For online testing, the number of persons engaging in these behaviors was almost twice as high with 61.4% reporting that they had engaged in either behavior.
And:
the rate of cheating in exams during this time depended on the examination mode, with online exams being more prone to the use of unpermitted assistance then on-site exams.
The authors do make efforts to link their findings, or limit them, to “ad hoc” online methods and lack of experience and preparation during online, pandemic-induced courses. Though it’s not clear whether their research pinpointed any specific pandemic effect or simply uncovered alarmingly high rates of misconduct and the existing link between online assessment and cheating.
It’s Time to End the Debate - Cheating is More Common in Online Classes and Exams
Given this new research from Germany, it’s time to end the debate as to whether classes and exams given online are more prone to cheating. They are. And I feel like we should know that.
Cribbing from the background in the Jankea, Petersen, Fritz, Daumiller and Rudert research:
Empirically, the number of studies comparing the frequency of cheating between on-site and online exams is limited to less than a dozen studies. The results of these studies are inconclusive, with [three] demonstrating higher cheating rates and [two] others showing steady or [two] even lower rates.
The thing is, while that’s true, it’s also not. At least one of the studies showing “steady” rates, doesn’t. It says:
The results showed that students felt they were almost four times more likely to be dishonest in on-line classes than live classes (42.2% to 10.2%) and that their classmates were over five times more likely to cheat (61.0% to 11.5%)
And the two that showed lower rates are from 2006 and 2009 respectively. And age isn’t their only limitation. Let’s just say they’re not great.
Even so, that’s what we knew then. What we know now has been consistent and overwhelming. There is a direct, significant link between cheating and education environments that are online. So much so that it’s not news anymore.
Here are just a few examples from the past year alone.
From Issue 36 of “The Cheat Sheet” - a May survey of college students:
Compared to when classes were primarily held in person, is cheating:
More common: 38%
Less common: 10%
About the same: 13%
Not sure: 38%
From Issue 58 of “The Cheat Sheet”:
By a whopping margin of 54% to 8%, instructors and faculty in the University of California system said that the “Amount of academic dishonesty (cheating) on tests and homework assignments” was either “much higher” or “higher” during remote instruction compared to “in-person” methods.
Issue 37: Network security provider Cisco says visits to cheating sites went up 4x in correlation to the move to online learning last year.
Issue 27: 79% of students at California high school say they've cheated more during distance learning.
Issue 72: Asked if they “started cheating more” during online education, a robust 57.8% of students said yes. Only 24% said no.
That does not include the dozens of school leaders who, in explaining the spikes in cheating cases at their schools, have directly pinned online or remote learning. Dozens.
Just last week, in Issue 71, Chris Bryson at the University of Arkansas said,
The increase [in cheating] is all related to exams and online exams specifically
I know some will make the case that online and pandemic online are not the same. And that’s fine. But there’s no denying that cheating is up and it’s being driven, to a very large extent, by classes and tests being online.
Pearson is Looking for an Academic Integrity Analyst
I spotted this fascinating nugget on Twitter, from Brendan DeCoster (@therealdecoster).
Pearson, it seems, wants someone to find and flag when “students take content from Pearson Virtual School” or other assessments and “post them to outside websites to obtain answers.”
Brendan is right - this is an ad for a “Chegg Detector.”
For background, see Issue 57.
Cal Poly Reports Major Cheating Increases
The student paper at California Polytechnic State University - Cal Poly - reports on an uptick in academic misconduct noting:
online test-taking appears to have blurred the line for some students about academic integrity
And that:
With the majority of classes offered virtually in the 2020-2021 academic year, [the school] saw an increase in academic integrity violations at Cal Poly. A total of 474 cases of academic integrity violations were reported
That compares, the paper says, with 359 cases the previous year.
Both years represent a “significant increase” over past levels - 183 cases in 2018-19, 101 cases in 2017-18 and 182 in 2016-17. Averaging those three, that’s 155 - making the 2020-21 caseload a more than threefold increase.
The paper says a school spokesperson:
was hesitant to draw a correlation between an online format and an increase in academic integrity violations due to the nature of unprecedented times
Though the same spokesperson also says:
he saw many cases of prohibited collaboration during test-taking with an online format.
The article also quotes a Cal Poly sophomore regarding a “lack of accountability when it came to testing in online classes.” She said,
“The professor isn’t there to watch you take the test,” [the student] said. “They can’t control what you look at. They can’t control if you’re really texting someone in the middle [of] class.”
Continuing,
“I know people who would like put their answers or put their notes on the screen [in front of them] to like look at the camera but they’re actually looking behind them,” Dutta said. “People would find ways to cheat that system.”
Widespread Cheating on Online Driving Exams
The Albany (NY) newspaper has a story on “widespread cheating” on driving tests given online. Here’s the story’s opening paragraph:
State motor vehicle workers say potentially thousands of individuals may have illicitly obtained New York driver's permits since the Department of Motor Vehicles launched a program last year allowing written tests to be taken online, leading to widespread cheating.
Further:
The 50-question online driver's test normally takes about 45 minutes to an hour to complete … But DMV workers said in many instances individuals who take the online tests are completing them — often with perfect scores — in less than seven minutes.
Cheating on online tests? You don’t say. On driving tests? Buckle up.
In the next “The Cheat Sheet” - I will try very hard to get to that “research” on proctoring out of Australia. Plus, other things that have been in the pipeline including a very odd “take down” notice - from Chegg. Plus, more cheating coverage.
To subscribe and share: