There is an interesting finding from behavioural economy on the subject of “strict evaluations” and it does not apply to students but to teachers and parents.
It was discovered during the training of pilots. The trainers were of the opinion that reprimand works better than praise. They discovered repeatedly over the years when training landing approaches that the students usually completed a better attempt after reprimand and a worse attempt after praise. They therefore reprimanded often and praised rarely.
This can be explained quite easily by means of probability: if the quality of the attempts fluctuates, after a praised attempt in the best 10 percent comes a 90 percent worse attempt. And after a reprimanded attempt in the worst 10 percent comes a 90 percent better attempt.
The “learning” effect for the teacher: after a reprimand, the student apparently makes an effort, whereas after praise evidently becomes negligent.
This even applies if praise has a positive effect in reality (and it has, says all empirical data). If one praises the best 10 percent of attempts and the motivated student is doubly likely to complete a good attempt, they will still be worse in 80 percent of the cases.
At some point, therefore, one reaches the “finding” that strict grades are a preparation for the seriousness of life. However, that is a gut feeling, not empirical. That is how we now make educational policy.