S2 Edexcel Tuesday 17th January SOLUTIONS IN FIRST POST

They may sometimes say it, though they sometimes don't. I'm doing the June 2006 paper right now and they haven't told me to use a 2-tailed test. You often have to deduce it from what I've seen.


I think you can either see if P(X<=p) < 0.05, or P(X >= p) > 0.95. Maybe someone can explain it clearer...

anybody know the s2 june 2011 grade boundaries?

June 2006 Q4(b). They say that since the result is significant, then the rate of breakdowns occurring has decreased. But why? If the mean was 5 in 4 weeks, and 11 breakdown happened, how is this evidence that it's decreased??

Must just be an error/typo in the mark scheme as clearly the rate has increased.

you know you've done too many questions in a short space of time when you notice the same review exercise 2 question with a tomato grower called 'Dhriti' in one of the past papers.

Haha, I noticed that too. I looked at the Review Exercise questions and I noticed I had done almost all of them in past papers! Except the Pat/Lyn one. I just can't wait until this exam is over! Don't have to worry about any more exams until 30th... S3 seems more interesting, I'd like to go back to learning it but revision for S2 has been getting in the way.

yeah i know, that pat/lyn one was a killer though, had to do it twice to understand it haha. i'm doing s3 in june instead of fp3, better chance of getting an A* plus im going into actuarial science so. I've got 4 more exams :/ D1, C4, M2 and sociology, gonna die this monthhhhhhh

Ah, cool. Sorry to hear about having so many exams. I'm doing S3 in June alongside FP3 and... quite a few other modules... but hoping I can get a better chance at an A* by doing more modules.

Hmm, I haven't decided for sure what I'm going to apply for yet. Probably maths/physicsy type degree... will have to wait until next year.

I had 7 last jan and 11 last june so it's all good i'm trying to get them all out of the way now so I only have S3/FP3 in June plus STEP/AEA for warwick which will be...challenging you taking a gap year?
also, when they ask questions about approximations i.e.: binomial to poisson/normal, do we have to give specific values? i.e.: np large, or np larger than 10?

100 ums - 75
90 ums - 67
a - 59
b - 50
c - 41
d - 33
e - 25

thanks mate, shame i cant rep you since I have done too much recently apparently!

how is s3 btw?

Haha it's okay, your rep would be kind of wasted anyway since I appear to be stuck at -3 rep... been like that for ages now even though I'm getting 10 positive rep a day, yet every time I get a neg it goes down...

S3 I remember being interesting! First chapter is a doddle because it's a continuation of S2, but the Chi-squared distribution stuff, confidence intervals, etc is where it starts to get pretty cool.

can someone tell if I'm correct

Approximations:

Binomial with poisson-(n large p small (np<10) so E(X)=Var (X). One mark scheme gave specific values of n>50 and p<0.2 but i havent found this in any other markscheme.)
Binomial with normal- (n large, p close to 0.5 np and nq>5)
poisson with normal- (lambda greater than 10)

any help is much appreciated!

Nice to hear that S3 is interesting. I'm self teaching and originally had planned to do M2/FP3 as my second year units but after actually quite enjoying S2 I'm seriously considering S3. Is the content more similar to S2 or S1 in nature though? Because S1 I found really quite boring with the whole venn diagrams and conditional probability stuff.

can someone tell if I'm correct

Approximations:

Binomial with poisson-(n large p small (np&lt;10) so E(X)=Var (X). One mark scheme gave specific values of n&gt;50 and p&lt;0.2 but i havent found this in any other markscheme.)
Binomial with normal- (n large, p close to 0.5 np and nq&gt;5)
poisson with normal- (lambda greater than 10)

any help is much appreciated!

Seems correct. It simplifys it a lot if you remember:
Binomial can be approximated by both normal and poisson distribution.
Poisson can ony be approximated by normal.

The only time you'll have to consider which to choose is if you have a binomial distribution. And then it's just:
- Poisson if np &lt; 10
- Normal if p is close to 0.5

From what I've seen in past papers, the questions make it very obvious which to use e.g approximate X~B(100, 0.02) or X~B(50, 0.5).