Cheating in exams
Cheating is a thing in exams, and your view completely changes when you move from a student side to become an instructor. It is not easy to ignore. Cheating is very easy to observe but very painful to ignore. In an examination hall, the moment a student starts behaving as If he is thinking too profoundly or starts drawing hyperplanes in the air or starts looking towards roof thinking in a particular way – you can quickly notice the overacting. Then there are always students with wider-angle and high zoom eyes. The thing is, as an instructor, you don’t want to punish honest students. There are still honest students (majority) in class who try honestly, sit alone in exams themselves cheating punishes such students most. Their evaluation becomes unfair if people solve papers jointly. So, I had to come up with a solution this time. It is easy when things can be quantified. It becomes solvable. So I did evaluate assignments using Machine learning techniques and came with a penalty if assignments were too similar to each other. The observations were too good to believe initially. For example, the students who were in the red zone remained silent and never challenged. Some of the students who were in the yellow zone questioned the penalty but also admitted taking help and mentioned that they solved independently but with some help.
This is how similarity looked for a class of 40 students.