Anecdotal Evidence: 10 Points for Cheating

When you hear the words “anecdotal evidence” do you think someone is trying to turn a single data point into a theory? We share your skepticism.

But hear us out on this, because this is something we keep hearing from customers. Put simply: when students take online exams in non-proctored environments (e.g. at home) using a standard browser, their average score is 10 percentage points higher than when LockDown Browser and Respondus Monitor are used with the exam. For instance, if the class average for a non-proctored online exam is 85%, the score will drop to 75% if students are required to use LockDown Browser (which locks down the computer itself) and Respondus Monitor (which records students during the exam with a webcam).

There aren’t formal studies to back this up. Not yet. But customers share these anecdotes with us so often that there’s likely something to it. And it’s not as if this should come as a surprise -- the whole reason instructors use Respondus Monitor is to deter cheating.

Instructors will sometimes report a difference of 8 percentage points, sometimes it’s 15. But the most consistent number we hear is 10 percentage points.

One instructor took her analysis a step further (her university prefers we don’t use their name, so we aren’t). The first exam in her course was non-proctored and the students used a standard browser – scores were comparable to past terms. The three remaining exams required the use of Respondus Monitor. Across the three exams, the average score dropped about 10 percentage points when compared to previous terms. But as she looked at the data further, she saw that the drop in scores wasn’t evenly distributed across the class. Students with the highest scores on the first exam maintained similar scores throughout the term. But many students who scored in the middle and lower range for that first exam saw their scores plummet when Respondus Monitor was required. For example, a student with a score of 82% on the first exam might drop to 65% or lower on the second exam.

There’s another interesting bit to her analysis. A few of her students did moderately well on the first exam, terribly on the second exam, and then moderately well on the remaining two exams. One can only speculate, but these students may have cheated during the first exam, were unprepared for the second exam where Respondus Monitor was required for the first time, and then came better prepared for their remaining exams.

Perhaps you’ve experienced similar results when using LockDown Browser and Respondus Monitor. If so, we’re interested in what you’ve found. And if you’ve done a more formal study on the use of these applications, we’d like to hear about those results too.