Dartmouth Abandons Controversial Online Cheating Investigation at Medical School (seattletimes.com) 38
Dartmouth's Geisel medical school is dropping its investigation into alleged online cheating, the New York Times reports:
In March, Dartmouth charged 17 students with cheating based on a review of certain online-activity data on Canvas — a popular learning-management system where professors post assignments and students submit their work — during remote exams. The school quickly dropped seven of the cases after at least two students argued that administrators had mistaken automated Canvas activity for human cheating. Now Dartmouth is also dropping allegations against the remaining 10 students, some of whom faced expulsion, suspension, course failures and misconduct marks on their academic records that could have derailed their medical careers.
"I have decided to dismiss all the honor code charges," Duane Compton, dean of the medical school, said in an email to the Geisel community Wednesday evening, adding that the students' academic records would not be affected. "I have apologized to the students for what they have been through."
Dartmouth's decision to dismiss the charges followed a software review by The New York Times, which found that students' devices could automatically generate Canvas activity data even when no one was using them. Dartmouth's practices were condemned by some alumni along with some faculty at other medical schools.
A Dartmouth spokesman said the school could not comment further on the dropping of the charges for privacy reasons.
"The moral of the current story is clear," argued the Times reporter on Twitter.
"Colleges that use surveillance tech can end up erroneously accusing some of their best students."
"I have decided to dismiss all the honor code charges," Duane Compton, dean of the medical school, said in an email to the Geisel community Wednesday evening, adding that the students' academic records would not be affected. "I have apologized to the students for what they have been through."
Dartmouth's decision to dismiss the charges followed a software review by The New York Times, which found that students' devices could automatically generate Canvas activity data even when no one was using them. Dartmouth's practices were condemned by some alumni along with some faculty at other medical schools.
A Dartmouth spokesman said the school could not comment further on the dropping of the charges for privacy reasons.
"The moral of the current story is clear," argued the Times reporter on Twitter.
"Colleges that use surveillance tech can end up erroneously accusing some of their best students."
In person exams are the only way. (Score:3)
Re: (Score:3, Insightful)
Just ditch exams. They are a crap way if measuring ability anyway, especially for estimating job performance.
We have better methods, let's use them.
Re: (Score:3)
I wouldn't let a doctor who hadn't got extensive experience and had his work evaluated and monitored by other doctors near my brain.
Exams are just one step for doctors, they have to go through years of training and evaluation in hospitals before doing complex surgeries themselves.
Re: (Score:3)
Exams are just one step for doctors
Sure. Exams are a necessary but not sufficient test of competence.
But that contradicts your earlier statement that we should "ditch exams".
Re: (Score:1)
Re: (Score:2)
I don't think they are all that necessary. I'd rather have the doctor be able to answer questions on work placement than in an exam hall. Memorizing everything the night before is not the same as being able to recall and use that knowledge in their daily working life.
The exam is at best a check to see if they are ready to move on to the work experience part, and a poor one at that.
Re: (Score:2)
We have better methods
Such as?
Re: (Score:3)
Well, yes and no. We have much better methods, but they a) require really good teachers and b) take more time and effort.
Re: (Score:3)
Just ditch exams. They are a crap way if measuring ability anyway, especially for estimating job performance.
We have better methods, let's use them.
What's the alternative to school exams? Job performance evaluations are only for people who are offered jobs. The school exams help to filter out people who get the privilege of interviewing for a job. It would be arguably more accurate to evaluate every student using on-job performance (internships for all students?), but that's not practical. Absent exams, what else filters out students? Homework? Projects? All these things are essentially the same as exams in terms of assigning a single score for
Re: (Score:2)
Coursework, or projects if you like. The advantage is that they are not dependent on performance on one specific day with one specific set of questions that purely by chance the student may or may not have studied for. Also exams are not much like actual work, where as coursework is.
Several projects over the course of 2-3 years is a much better metric.
Re: (Score:2)
Just ditch exams. They are a crap way if measuring ability anyway, especially for estimating job performance.
We have better methods, let's use them.
Modular exams certainly are; I think year-end exams might be more useful but university is all about getting the best pass-rate in your sector so you can get more students and hence more money, so rigour is long, long, long gone. When we recruit now we don't even look at applicant's education any more. We give them a preliminary basic (and I mean basic) maths and English multiple-choice test and if they score better than a randomly-pecking pigeon would we take a look.
About 80% of applicants don't beat the p
Re:In person exams are the only way. (Score:4, Interesting)
No.
Re: (Score:2)
Can't blame the accused here, as the system is obviously stupid. I am blaming the colleges that think this is a good idea.
Re: (Score:2)
If Dartmouth didn't buy the software in the first place, non of this would have happened, they would have caught exactly the same number of cheaters (none), but everyone would be happier.
Re: (Score:2)
Will agree with you in terms of always having cheaters?
I had an instructor in grad school who was in a CONSTANT arms race with these clowns.
TBH I could honestly give a rip about these folks, as they are the ones that won't be able to land a decent job after graduation. What did tick me off however was the controls this instructor had to put into place to try and nail these schmucks also meant that honest folks got roped in (assigned seating, he gave you the stink eye if you had to use the toilet during the
Beware of gathering Data (Score:5, Insightful)
Collecting data in one thing.
Properly analyzing it is another.
Put another way, if you look for something long enough and hard enough, you'll find it. Even when it's not there.
Re: (Score:3)
It doesn't sound like they were looking too hard for misconduct. The problem is they were looking for evidence of misconduct in a data set they did not understand.
If you don't understand your data, you can't draw any valid conclusions from it. And if you know how to use a dataset for X, that doesn't mean you know how to use it for Y.
Re: (Score:2)
Indeed. Makes me wonder whether one should study there. They seem to have some rather severe fundamental incompetence.
Re: (Score:2)
All it takes is one administrator, then once the witch hunt gets started it takes on a life of its own.
Re: (Score:2)
Indeed. Does not speak well for the rest of the organization though. Like they have coasted along too long on their reputation and the rot has now set in deeply.
back to test basics (Score:2)
Damages for Suffering (Score:5, Insightful)
Imagine facing expulsion from medical school. You don't just pack up your bags and go to a different medical school. It doesn't work that way. You've basically been told that what you been working for for 8-12 years is over, your future is going to be something else, and probably, "if you confess this will be easier on you."
That is some serious fucking trauma, based on a shitty software package. Dartmouth is lucky that there weren't any super-anxious kids in this group who thought that "the easy way out" would be better.
I'll be shocked if there's not a lawsuit for intentional or negligent infliction of mental duress.
Back in the day Dartmouth wrote most of its own software and everybody operated on the honor code. If you wanted to take your test book outside and sit under a tree and work through your Calculus exam, that was up to you.
It would be nice to see a little bit of Academia return to the College on the Hill. Maybe even teach Blackstone's Formulation - there are ethicists and philosophers right across the Green.
-D'95
Re: (Score:2)
It's medical school. It's four years, three in some places, just like any other undergraduate degree.
Getting expelled because your school administration are dumbasses would suck, regardless of what degree you happen to be pursuing.
Re: (Score:2)
Re: (Score:2)
In addition to all the problems you mention, many med school rejects also have $250k or more in debt with no job to pay it off.
Blotch on their reputation? (Score:3)
I think this is likely as a result of the pushback they got from all this that ended up tarnishing their reputation, and likely that some potential students are now looking at other medical schools instead.
Bring anything but a live friend (Score:5, Insightful)
Yes, we need to keep the med exam cheaters (Score:1)
Otherwise, where would be get our vital supply of pharmacy benefit managers? Without them, drug prices might be set by the free market and pharma CEOs would starve to death.
Surely an honor code (Score:3, Insightful)
Just Do Open Note Exams or Projects (Score:3)
I'm a grad student (pursuing PhD) and my classes used no cheat detection or surveillance (they even told us to turn our cameras off for bandwidth reasons; some students weren't necessarily even in the country at the time) and dealt with things in two pretty obvious ways:
1. Semester projects.
2. Open Note tests.
Now, this is probably easier for some classes than others; at grad level they can throw you a bunch of previously unanalyzed data and tell you to do analysis on it, and there's not going to be anything online you can just copy the way you could in say an English Literature class. But for tests doing open notes makes more sense. Especially compared to the Canvas detection methods used here, which are not going to catch (say) someone who just downloaded all the stuff that was on Canvas already.
Is that really the moral of the story? (Score:1)
Or is the moral rather that if you get caught cheating, you should hire some good tech people to point out extremely hypothetical ways in which your device could set off cheat alerts by itself?
the true moral (Score:2)
Times reporter tweets:
"Colleges that use surveillance tech can end up erroneously accusing some of their best students."
Dartmouth DID NOT use surveillance tech. They used logs from a program that is used to access course materials. They were clueless as to how those logs get generated. An over-zealous prof noticed the logs and started a panic among idiot administrators.
So, the true moral is that NYT reporters, like most so-called "journalists" are clueless but claim to know all.
re: (Score:1)