>The spreadsheet now gives two AP scores -- one using the original >"subtract 1/4 pt for wrong answers" scheme, and once using the new >scoring conversion templates.
I just want to remind folks that a simple conversion of a score ... assuming the correction was added back ... makes an assumption that IF the student had been doing the test under a non correction for guessing direction, they would have responded exactly the same way. This assumes that the strategy one would use given advice about how to respond IF a formula were used AND if it were not used ... are the same. If students really take the correction for guessing directions to heart, then the strategies under both conditions should not be the same. How much difference there might really be is a matter of empirical investigation.
Some work would suggest that a mere adding back of the correction is perhaps not enough ... if you really believe that following a strategy when a correction is applied ... hurts those with partial knowledge on items.