Online university testing

[Mosby]

Seems like some profs aren't digging it, lol.

Link to tweet

[unblock]

1. i have no problem with this as long as the "correct" answer could only reasonably be arrived at

by getting it from the cheating website.

if the "you cheated" answer could plausibly be arrived at by honest but incorrect work, then i have a huge problem with it.

[ProfessorGAC]

2. That's A Good Point

Exams in Advanced Organic Mechanisms were basically several problems of:
A + B <> C + D. Then 50 blank lines.
1 point for each correct step, 0 for mistakes. But, -1 if a step included something that could not happen.
I can see this being done on a test like this. Intentionally adding "can't happen" steps, but could reasonably been included by the test taker as a legit error.
Doesn't really prove cheating. Just a lack of understanding. Of course, a 600 level course for people also pursuing PhDs shouldn't have that many folks making that error. Getting stuck & leaving out a step is more likely!

[unblock]

3. right. to prove cheating, it really has to be a "unique" sort of error

mapmakers prove cheating (competitors copying their maps) by putting in inconsequential but clear errors in their own maps, such as putting a squiggle in a road that's actually straight. it's really hard for a competitor who shows up with the exact same error that they actually went to that exact same road and somehow mistook a straight road for a bendy one.

so if the cheating website answers "2 + 2 = ?" with a bizarre number like 16.37, and the cheater turns up with exactly 16.37, that's remarkably compelling evidence of cheating.

but if the question is "x*x = 4, what is x?" and the cheating website says 2, well, haha, the correct answer is +2 *or* -2. but that's hardly evidence of cheating; that's in fact a common mistake, or perhaps even a miscommunication of how technically the question was meant to be answered.

surely the real question and "answer" lies somewhere in between, but a lot depends on the exact question and how innocently plausible the "cheating" answer is.

[BGBD]

4. then again

is it ethical to bait a class into "cheating" by putting an impossible question on a test?

How many of these folks would have cheated had the question been one they could have actually solved? It took he and his TAs to craft the question, would a class of undergrads be expected to recognize it as impossible?

He said he was suspicious when the average of the class was 6 points higher than normal. Does that mean it was from cheating or just being in a more comfortable atmosphere to take the test?

[unblock]

6. i'm not sure exactly what is meant by "fundamentally flawed"

i agree, there's room for this to be problematic if he wasted a ton of time and/or caused much agitas among the honest students who couldn't figure out that question, effectively sending them on a wild goose chase of sorts.

it would have been cleaner if he simply asked an honest but hard question and supplied a very incorrect answer on the cheating website (again, one that would be most improbably arrived at by any means other than seeing it on the cheating website).

the "flawed" question approach would be fine if the test made it clear what one was expected to do with such a question. e.g., if the instructions made it clear that "there is no answer" may be the correct answer.

[BGBD]

8. Depending on the type of class

You can always fall back on the old "show your work." Even in a more qualitative course, "Why?" can go a long way.

[WhiskeyGrinder]

5. Man, some professors are dicks.