Let me add to my post above.
IF you made the assumption that NCEES starts with a cut score of 56/80 (which they never said) and then adjusts their subsequent cutscores based on equating (which they do say), you could possibly derive some correlation that would tell you the cut scores for past exams.
IF you had enough data (which I don't believe we have) and knowledge in data mining you could possibly devise a neural network or something like that to predict the cutscore for the current and future exams. I still don't believe it would be much better than guessing.
Thats exactly what i was trying to state.
But the flaws would be the assumption that A) they start with 56/80 B)It would not be depth specific .
I believe you do have enough data.
You definitely would come close to predicting the cut scores of past exams and future ones within a certain +/-.
Any ***** who doesn't think that is possible, must not of faired to well in Engineering Statistics.
BUT I WOULDN'T WAIST MY TIME DOING SUCH A THING. Because know one person has a clue as to how many questions they got correct, so such an effort would be futile to ones benefit.
