> Stupid assignments encourage students to cheat. Make them interesting and this problem will go away.
and yet
> Usually TAs would get a cheating report for reused bits of code and things that would solve an exercise with techniques far away from students knowledge or forbidden functions. TAs would ask you questions about your code and trigger cheating review if you could not explain why you wrote it this way. It was usually effective for detecting people that didn't wrote their own exercises.
So both courses have cheating, one possibly has less cheating, and that one has a more effective detection procedure, but cheating happens on both all the same. The problem does not go away with a better course but is - based on your anecdote and not from the presumably better view of the TAs - lessened somewhat.
Perhaps the same course could be run in California and would still attract the same level of cheating as the checkbox style one. Maybe it's simply that a course that appears more difficult to cheat on or is more difficult to cheat on attracts less cheating.
> The problem does not go away with a better course but is - based on your anecdote and not from the presumably better view of the TAs - lessened somewhat.
I was a TA in France, and did a bit of TA-ing unofficially in the US. So, my view is based on both experience, TA and Student.
While there was some cheating happening on both sides, it was practically non-existent in France after a few weeks of school. The abilities required to cheat and get away with it were extremely hard to master, and usually would carry much more risks than just complete the exercise. Successful cheating were, sometimes, rewarded if the problem solved was worthy enough. You could even get clearance from the professor in advance (not for copying code from StackOverflow, of course, but rather for using forbidden functions or libraries).
Finally, most important projects were done in groups of 2 to 5-6. Individual students putting a group at risk deliberately, would get caught extremely easily and carry a far worse sentence than just a bad grade: they would be excluded from other students groups and left to do their projects alone or with dropouts students. Finding back you way after this is very very difficult.
> Finally, most important projects were done in groups of 2 to 5-6
Much less reason to cheat in group work too. It's much easier to just put in minimal effort and let someone else in the group do the work. That's just another form of cheating.
In Germany we say "Team" in teamwork is an acronym for "Toll Ein Anderer Machts" (great someone else does the work) and especially in university there is a lot of truth to it. Especially so when there is a big difference in motivation in regards to the final grade.
Not to say there aren't a lot of upsides to group work. But stopping people that don't wanna engage with the subject isn't one of them.
> Perhaps the same course could be run in California and would still attract the same level of cheating as the checkbox style one.
Actually, that's not a hypothetical scenario: School 42, an offshoot of the school GP described with the same cursus, has antennas in both Paris and California.
I wonder if the cheating rates differ between the two.
and yet
> Usually TAs would get a cheating report for reused bits of code and things that would solve an exercise with techniques far away from students knowledge or forbidden functions. TAs would ask you questions about your code and trigger cheating review if you could not explain why you wrote it this way. It was usually effective for detecting people that didn't wrote their own exercises.
So both courses have cheating, one possibly has less cheating, and that one has a more effective detection procedure, but cheating happens on both all the same. The problem does not go away with a better course but is - based on your anecdote and not from the presumably better view of the TAs - lessened somewhat.
Perhaps the same course could be run in California and would still attract the same level of cheating as the checkbox style one. Maybe it's simply that a course that appears more difficult to cheat on or is more difficult to cheat on attracts less cheating.