If you're not familiar with the terminology basically it means that instead of deciding "You need 45/50 to get an A, 38/50 to get a B, 31/50 to get a C etc. etc." they decide "The top 10% o candidates gets an A, the next 15% gets a B etc. etc" (those numbers are completely made up but you get the Idea). This generally improves fairness because if the test is slightly harder or easier than expected then it doesn't really matter, everyone should still be similar relative to each other and so will get the grade they should have.
This system is used in the UK, but is there something similar in the US? I know that less depends on test performance but you do have tests which are graded so I'm interested to hear about it.
Submitted June 08, 2020 at 01:15AM by python_megapixel https://ift.tt/3f5LS3j
No hay comentarios:
Publicar un comentario