> There's an incentive for a dishonest student to cheat because it's significantly easier than actually doing the work. That's only true if cheating gives you

> There's an incentive for a dishonest student to cheat because it's significantly easier than actually doing the work. That's only true if cheating gives you the same benefit as actually doing the work. In many fields (especially science/technology/engineering/math), I would argue that the cheater doesn't capture much benefit. It's pretty easy to verify their abilities and see that they don't know what they're doing. In these areas, even where the credential is necessary it is not sufficient. On the other hand, there are cases where it seems that a cheater is just as capable of doing the work. Where that's true, it calls into question the wisdom of using such credentials in the first place, since mastery of the material ends up being uncorrelated with performance. > but students don't know best. True, but that's only a problem if we assume learning is a one-shot thing. If learning is a continuous, iterative process then there's plenty of chances for the learner to self-correct based on real-world experience. That's one reason I went off on my "agenda" about why education in general needs to change.