The Student Room Group

S2 Edexcel Tuesday 17th January SOLUTIONS IN FIRST POST

Here are the solutions.

A tricky paper I thought but not too difficult. Probably around 60-62 for an A.

Before I get loads of abuse for doing question 2 as a one-tailed-test let me say that I did do it as a two-tailed-test originally and then when checking back through I changed my mind and changed it.

The question says no better than tossing a coin. Now I know that some people are arguing that that could mean worse as well as better but I think it makes more sense to say better. I mean if it was worse then you would be saying that the weather forecast had less probability than a guess. I think it makes more sense in the context of the question to do a one-tailed-test.

If you've done a two-tailed-test then you'll probably lose 2 marks although I'm not 100%

There was a very similar question on Jan 11 about a student cheating on a test. In that case you would lose 1 mark for using a two tailed test

FIRST MISTAKE! LAST QUESTION P(X<=12) = 0.9362 SO YOU DON'T REJECT H0
(edited 12 years ago)

Scroll to see replies

Reply 1
Thanks Gibbo. When are you planning to upload the solutions? Could you do it at midnight on Wednesday? :tongue:
Reply 2
Thanks for this. Still not too confident about the definitions... I think I can get 90+ though. The thing is, the boundaries are sometimes quite high (Jan 2011 had 72/75 for A* and 69/75 for an A), but when this does happen I tend to get full marks or close to full marks anyway because it is an easier paper. I would like to get 100 in S2 because I am probably not going to get that kind of mark in FP2 or FP3.
Reply 3
Original post by safmaster
Thanks Gibbo. When are you planning to upload the solutions? Could you do it at midnight on Wednesday? :tongue:


I would like that too, but I think the rules say that it has to be done at 4:30 AM the next day, or something... (in the middle of the night)

Original post by Zhy
Thanks for this. Still not too confident about the definitions... I think I can get 90+ though. The thing is, the boundaries are sometimes quite high (Jan 2011 had 72/75 for A* and 69/75 for an A), but when this does happen I tend to get full marks or close to full marks anyway because it is an easier paper. I would like to get 100 in S2 because I am probably not going to get that kind of mark in FP2 or FP3.


I'm just doing AS further maths and we're doing m1, s2 and fp1. And I think content-wise this is the easiest modulke. Just need to stop making stupid mistakes.
Reply 5
Oh dear, I'm terrible at S2.. :frown:

If you've got a poisson distribution, how do you decide between approximating to binomial and normal? I never know which one I'm meant to approximate to..
Reply 6
Original post by Zhy
I would like that too, but I think the rules say that it has to be done at 4:30 AM the next day, or something... (in the middle of the night)


No, I'm quite sure that's for PM examinations. S2 is an AM exam so discussion can begin at midnight.
Original post by alexsasg

Original post by alexsasg
Oh dear, I'm terrible at S2.. :frown:

If you've got a poisson distribution, how do you decide between approximating to binomial and normal? I never know which one I'm meant to approximate to..


Just look at the question, high marks for P to N as you have to do continuity correction, graphs etc. 4 or 5 marks for P to B as it's much easier.
Reply 8
Original post by areebmazhar
Just look at the question, high marks for P to N as you have to do continuity correction, graphs etc. 4 or 5 marks for P to B as it's much easier.


I'm confused, approximating Poisson using a Binomial? I've never seen them ask that, I've always seen it as the other way around. How would you approximate a Poisson with a Binomial? I suppose if you're given X ~ Po(2) or something, then np = 2, then I'm assuming they give you some other information?

Original post by Zhy
I'm confused, approximating Poisson using a Binomial? I've never seen them ask that, I've always seen it as the other way around. How would you approximate a Poisson with a Binomial? I suppose if you're given X ~ Po(2) or something, then np = 2, then I'm assuming they give you some other information?


Yeah, I misread the question. Poisson only approximates to normal
Reply 10
When approximating to a Normal, how should you write the distribution?

If mean and variance are both 70, should you say X ~ N(70, 70) or X ~ N(70, √70²)?

They're the same thing, but I've seen the latter more common in the mark schemes...
Reply 11
Oh :colondollar: my bad. Sorry. And thanks! So if Poisson can only be approximated to normal..

how do you what to approximate binomial to? Any help is greatly appreciated. :smile:
Original post by alexsasg

Original post by alexsasg
Oh :colondollar: my bad. Sorry. And thanks! So if Poisson can only be approximated to normal..

how do you what to approximate binomial to? Any help is greatly appreciated. :smile:


Just look at my previous post but replace poisson with binomial.
Reply 13
Original post by alexsasg

Original post by alexsasg
Oh :colondollar: my bad. Sorry. And thanks! So if Poisson can only be approximated to normal..

how do you what to approximate binomial to? Any help is greatly appreciated. :smile:


A lot of the time, the question itself will tell you what to approx to. But when it says "use an suitable approximation" for a binomial of X~B(n,p) , if n is large and p is small, and if np 10 go with Poisson. otherwise, its gonna be a Normal approx. Its worked pretty well so far for me xD
Reply 14
If n is large and p looks sort of small, BUT np and nq are both slightly bigger than 5, what should you do?
Reply 15
On S2 January 2007, question 3d it is X - B(50, 0.45) which can be easily found from the tables so why do you have to do normal approximation? Thanks
Reply 16
Anyone?
Reply 17
Original post by Mathssss
Anyone?


In normal cases, you are right, you can use tables. But the question explicitly says to "use a suitable approximation", hence you need to approximate the binomial to a normal.

Also: n is large, p is close to 0.5 and np>10, which are the conditions under which you should approximate a binomial to a normal.
Reply 18
Original post by safmaster
In normal cases, you are right, you can use tables. But the question explicitly says to "use a suitable approximation", hence you need to approximate the binomial to a normal.

Also: n is large, p is close to 0.5 and np>10, which are the conditions under which you should approximate a binomial to a normal.


Thanks for the reply, so if the question hadn't have said 'use a suitable approximation' would just doing Binomial have been right?
Reply 19
Original post by Mathssss
Thanks for the reply, so if the question hadn't have said 'use a suitable approximation' would just doing Binomial have been right?


Yes, for sure.

Quick Reply

Latest