We employed the method of mixing the order of questions, not giving question titles, and having a pool of subtly different questions (a negation or a different constant slipped in) in order to vastly increase the amount of necessary communication to collaborate.
We had the students turn on their webcams just to discourage them sitting in the same room, which basically would negate our former efforts. Neither did we flag any behavior nor did we record anything or used some automatic software.
I believe that worked quite nicely. A few cheaters were easily spotted when they had the wrong constants in their answer.
I understand that not every course and exam style is suitable for that. But for our databases course, it was applicable.
I think that this was not much more invasive than a traditional exam. Anything more, like watching the screen, or having audio, would make me feel uncomfortable.
We had the students turn on their webcams just to discourage them sitting in the same room, which basically would negate our former efforts. Neither did we flag any behavior nor did we record anything or used some automatic software.
I believe that worked quite nicely. A few cheaters were easily spotted when they had the wrong constants in their answer.
I understand that not every course and exam style is suitable for that. But for our databases course, it was applicable.
I think that this was not much more invasive than a traditional exam. Anything more, like watching the screen, or having audio, would make me feel uncomfortable.