s2 OCR how did u find it??

so for the last question, some of us accepted Ho, while others rejected it. is anyone 110% of what the anwser was, because this question was worth 10 marks?

Well with the binomial you have in the context you can do it with 1- (Px=0 +px=1 + px=2) so why add in an extra approximation?

Well with the binomial you have in the context you can do it with 1- (Px=0 +px=1 + px=2) so why add in an extra approximation?

Because you can't work out a binomail with n=50, the tables don't go that far, so you have to approximate

Hi, the answer was definately to accept the null hypothesis, the probability was about 0.04 which is > 0.025

lol

50C0 P^50 etc................. too many calculations, much better to approximate using Poisson.

Well the binomial you use is an approx to the normal from before so you are using another approx like they ask in the q, and its only 3 calculations which isnt too bad. It doesnt look like theres anything wrong with using the binomial or the poisson so will they have to give marks for either method or is one of them just wrong?

Well firstly the binomial we use in part B) is refering to a different situation than part A), in sense part A) was all about getting the p value for B), so its not a double approximation or any thing like that. As far as marks go, i don't see how you could of used the binomial anyway, unless you spent 30mins working out the binomial equations....

It was just 1 - ((P x=0) + (P x=1) + (P x=2)) so just the 3 binomials to play with... I hope

ok then...explain to me how you did just p(x=2) under X~B(50,0.021)....

Use standard binomial equation 50C2 x (0.021)^2 x (1-0.021)^48

yeah i can see where your coming from...although i'm still sure that we were supposed to approximate using poisson.....lol i'm sure one of us is right

Oh yeah, the question did ask us to use an approximation....so the poisson method is correct..i think