Techniques for detecting and deterring cheating in home exams in programming

The Covid-19 pandemic led to increased use of home exams, with a perceived increase of cheating. Assessment integrity is a key challenge for higher education. Apart from remote proctoring, what other mitigations may be possible against cheating in home exams, and specifically for programming courses with huge classes? The paper presents our approaches to mitigate cheating, for CS1 based on questions with subtly different variants, for CS2 based on plagiarism detection and timestamps - in sufficient detail that others could use a similar approach. These two approaches can be partially effective against collaboration, but less so against contract cheating where help is acquired from an outside third party. Hence, towards the end of the paper we also outline possible approaches to mitigate such cheating, without or in addition to remote proctoring.