I had a question about how AP readers would grade an AP Statistics question. The story below is how Twitter helped me get the answer, in great detail, within a couple of hours. Twitter: you are awesome. (AP Stats exam, in some ways you are less awesome. Some of your obsessions make sense to me. But “Name the test or lose 12.5-25% of your score!” is harsh, man. The test is pretty obvious from context.)
#statschat #apstat readers, would this probably get 3.5/4 (PEEE) as I marked it as an AP problem? Still exploring intricacies. pic.twitter.com/1tfiswDlBs
— David Griswold (@DavidGriswoldHH) March 6, 2017
@DavidGriswoldHH given that p is clearly defined in the conclusion, I'm ok (but not thrilled) with the lack of definition at the start.
— Bob Lochel (@bobloch) March 6, 2017
@DavidGriswoldHH yep, missed that. So this becomes EPEE – no bump up to a 4 given here
— Bob Lochel (@bobloch) March 6, 2017
@DavidGriswoldHH bump from 3.5 to 4 is usually quite hard. 2.5 to 3 or 1.5 to 2 more likely if we see evidence.
— Bob Lochel (@bobloch) March 6, 2017
@DavidGriswoldHH @bobloch conditions check is not correct. Should use np0 and nq0, not n(phat) and n(qhat)
— Jonathan Osters (@callmejosters) March 6, 2017
@callmejosters @bobloch I haven't clarified that well enough with them and need to do so so was generous there.
— David Griswold (@DavidGriswoldHH) March 6, 2017
@callmejosters @DavidGriswoldHH @bobloch Would not round up either because communication on conditions is minimal. Random ✓ means what?
— Amy Hogan (@alittlestats) March 6, 2017
@alittlestats @callmejosters @bobloch So I think conclusion is AP would give PPEE. B/c of my bad, PEEE is reasonable. AP would be 3/4.
— David Griswold (@DavidGriswoldHH) March 6, 2017