Originally Posted by Bostonian
AP exam scores are reported on a 1-5 scale and are intended to be comparable to the A-F grade scale (5 = A, 3 = C). The College Board grades each exam to get a raw score and then decides where to set the thresholds for a scaled score of 2, 3, 4, and 5. The ceiling could be lifted simply by reporting the raw scores as well as the 1-5 scaled scores.

A problem with raw scores is that they aren't consistent from year to year. My knowledge all comes from talking to graders and head graders from the 90's, but I assume it is still similar. Keeping grades between tables(literal folding tables) consistent in a given session was a huge issue. Trying to normalize this across different tests over multiple years just wasn't possible. Raw scores in different years just meant different things. The tests just weren't normalized that way.

A common exit exam that covers a broader swath of knowledge that seniors take is less subject to this problem than APs taken over a 4 year span.

The second problem is that raw scores on a common exit exam is useless if the ceiling isn't high enough. In the UK the various combined A level math exams are harder than the Calc BC exam. The STEP exams are harder than A levels. Going on the standard belief for above level tests, as pushed by TIP/CTY etc, that each of these tests should generate its own normal curve for appropriate populations suggests that SATs should stratify the top 25-30% of kids; multiple APs should stratify the top ~10%; if you want to accurately sort the top 5% you need harder tests.

I am ambivalent. I would like the option of more meritocracy but I also fear the results and collateral damage in our hyper competetive society...