This document last updated: 3 May 2012.

In general, the mock paper seems fine, other than C1 being too long for Section C and D6 being mildly problematic in some respects. D6 is quite a lot longer than D5, but it's not unusual for such variation to exist.

C1 (b)(i) 1/2 (by intuition or a detailed calculation) (ii) 5/11 (by detailed calcluation, confirmed by Excel simulation)

C2 (a) k=3/20 (b) X: 9/20, 6/20, 5/20 Y: 5/20, 8/20, 7/20 (c) 1/4, 3/8, 3/8 (d) 75/20 (e) No

C4 (b) 0.815+/-0.004

D5 (a) sample Z=-1.73, critical Z=-1.645 reject H0 (b) sample Z=-3.35 reject H0 (c) (+not+ assuming variances are equal - question unclear) 63.7C is 146.66F - if you round this to 146.7 to fit the precision of the 146.9 value, you get sample Z=0.85; if you do not round it, you get sample Z=1.014. Regardless, critical Z=1.96 do not reject H0

D6 (b) sample t=1.37 critical t=1.714 do not reject H0 where H0 is "no dependency" (c) 0.83? (calculation seems simplistic to me but I see no alternative) (d)(i) sample t=2.87 critical t=1.717 reject H0 where H0 is "no link" (ii) R=0: 5.22 R=1: 16.0 (calculation seems invalid to me but I see no alternative)

Ian Rudy ()