Predict the Scale

[Praetor]

In '03 December was easier curve than June, roughly equal to October.
In '02 December was easier than both June and October (much easier than June)
In '01 December was more difficult than June and October (though very close)
In 2000 December was almost identical to the others

I think the best case scenario here is June's curve, where 85 was a 167.

[borderlaw]

Same as October's I would guess.

Anyone know if we receive a free point to start at 101?

[Praetor]

We don't, that's part of the reason why October's scale was so brutal.

[tludy]

I felt the Oct04 test was easier. So I'm not very optimistic. I cant imagine anything better than 89=170.

[Matthew_24_24]

89=170 is about right.

October curve was insane, so i dont think it will be repeated.  However, I do think it will be a harsh curve.  One point less harsh sounds just about perfect.

Matt

[calibos11]

The October test almost always has the harshest curve, b/c the largest number of ppl are taking the test translating into a greater number of highly proficient testtakers.

[bluetooth]

Although I didn't take this Dec. test, I would guess that the scale would be equal to or slightly easier than that of Oct.

[Praetor]

The October test almost always has the harshest curve, b/c the largest number of ppl are taking the test translating into a greater number of highly proficient testtakers.

Not in '02 or '03, as I said earlier the June scales were the harshest in those years. Still, I think we have reason to be slightly optimistic about the scale. The 100 questions again kinda pisses me off though.

[24601]

The sample of test takers should not affect the curve.  The curve is decided before the test is given.  That is why they have experimental sections. 

I didn't take the test, so I can't opine about the curve.

[rosemaryblue]

Are you sure? I thought experimentals were given to try out new questions that  could possibly appear (in similar form but obviously different q) in the future and that this was determined by seeing if the CURVE for a particular experimental question/section was typical of the test's curve...