S2 18th January 2013

Original post by brittanna

Oh. I just looked through the table, starting at n=5 and writing down the corresponding probabilities.

Original post by uniphysics

Sorry, but that's kinda what they didn't want you to do :frown:

That's why they gave a value which isn't in the tables, although I used the tables too to get an idea of what sort of value it would be. I was at suicidal levels by that point :P

Taking logs would have been the expected method of doing it, that's the way I did it

BUT if you wrote down the P(X=0) for 7 and for 8 and showed 7 was > 0.01 and 8 was <0.01, I'm quite certain you'll still get the marks

Reply 581

chignesh10

15

Original post by uniphysics

Sorry, but that's kinda what they didn't want you to do :frown:

That's why they gave a value which isn't in the tables, although I used the tables too to get an idea of what sort of value it would be. I was at suicidal levels by that point :P

You can do it anyway you want. as long as you get the right answer

Reply 582

JJMills

For the last question on the mosquito section where it was talking about the number of days until 10% were left I had t-1 in my equation for some reason so I got 34 days.

Can anyone tell if this is correct or if it was just 33 days?

Reply 583

notmewill

Are there solutions for this paper anywhere? :smile:

Reply 584

BobLoblawLawBlog

16

Original post by Mr Tough

Taking logs would have been the expected method of doing it, that's the way I did it

BUT if you wrote down the P(X=0) for 7 and for 8 and showed 7 was > 0.01 and 8 was <0.01, I'm quite certain you'll still get the marks

Yeah, I did that. I just did it for 5 and 6 as well.

Reply 585

Economi

9

Does anyone have the marks for each question?

Reply 586

samwalker1994

Original post by soempty

5a was the hardest 1 marker ever, spend on it more time than on any other question and still got it wrong.

Do you people used one-tail or two-tails for her voting claim? I started off as one tail as it was logical, but decided to leave it as two-tailed test because question just asks to find out whether he claim is correct, which kind of asking for a two-tailed test...

The rest was fine I think. Hopefully grade boundaries will be very easy, because paper is indeed quite hard compared to the majority (bearing in mind I have done all pastpapers since 2005 for revision).

I think it was one tail, it was unclear but it was implied that you only need to work out p<0.45. The part (b) did it as one tailed and when you get the answer it was something like 5.4% and they normally like to make it close to the tail where possible.

Also, as long as you did it correctly I don't think it makes a difference as 5.4% is greater than both 5% and 2.5% so both answers should lead to the same outcome.

Reply 587

Original post by chignesh10

You can do it anyway you want. as long as you get the right answer

You can IF you get the right answer :P

If the value you're looking for is in the tables then you kind of can't :P

Reply 588

Zebrastripe

Original post by uniphysics

If you're aiming for an A, you need a total of 480UMS in whichever 6 modules you're doing. So yes, you can afford to do badly on this one presuming you do relatively well in the others :smile:

thank you, yeah i need an A so badly to meet my uni offers. i wasn't really well enough for this exam so didn't get much last minute revision done but am prepared to slave away hard for core 3 and core 4. so wish me luck!

Reply 589

Original post by Mr Tough

Taking logs would have been the expected method of doing it, that's the way I did it

BUT if you wrote down the P(X=0) for 7 and for 8 and showed 7 was > 0.01 and 8 was <0.01, I'm quite certain you'll still get the marks

ahhh, yes, fair enough

Reply 590

dan94adibi

14

Original post by Mr Tough

It should be 0.221 (to the 3dp/sf they asked for)

they said the P(X=0) = 0.05

and X~B(12,p)

so (1-p)^12=0.05

1-p = twelfth root of 0.05 (type this into your calc)

then do 1-the number on your calc, leaves you with 0.2209..., 0.221 to 3sf

Yup thats it. I wasn't sure of the right figure but thats the way to do it.

Reply 591

Original post by soempty

5a was the hardest 1 marker ever, spend on it more time than on any other question and still got it wrong.

Do you people used one-tail or two-tails for her voting claim? I started off as one tail as it was logical, but decided to leave it as two-tailed test because question just asks to find out whether he claim is correct, which kind of asking for a two-tailed test...

The rest was fine I think. Hopefully grade boundaries will be very easy, because paper is indeed quite hard compared to the majority (bearing in mind I have done all pastpapers since 2005 for revision).

I think all you needed to say was f(x) = 1 - F(x) but I'm not sure

I used a two tail test, but I think you would be able to do either because the value they gave you to test was at the lower tail so the upper tail wasn't really needed.

Did you find the critical region? I only did that because they asked for the definition so I thought it was just following up from that :P

Yhhh grade boundaries are likely to be like June 2011 I think.

Reply 592

JJMills

Original post by Mr Tough

It should be 0.221 (to the 3dp/sf they asked for)

they said the P(X=0) = 0.05

and X~B(12,p)

so (1-p)^12=0.05

1-p = twelfth root of 0.05 (type this into your calc)

then do 1-the number on your calc, leaves you with 0.2209..., 0.221 to 3sf

That's what I got.

My method for working out p was somewhat long winded because I completely forgot about roots!

I had p = 1 - e^(ln(0.05)/12)!

Reply 593

Original post by DJMayes

I really enjoyed sitting this paper – it was a departure from the usual predictability of S2 papers and dabbled in a variety of fresh ideas which made the exam a lot more enjoyable than previous years. I also think it constituted one of the more challenging S2 papers, and should have relatively low grade boundaries (For S2). The mosquito question is likely to cause some disagreements, and the definition of hypothesis could hopefully go the same way as significance and lower the 100% boundary.

My answers:

Spoiler

Yep I agree with them all, my definition of hypothesis is slightly different, including the phrase "population parameter" which one of my teachers said should get the marks...its quite easy to define it generally, but not sure what the specific phrase edexcel would have wanted

To everyone on this thread asking about grade boundaries, I think you should remember that this paper is mainly sat by further mathematicians, so the grade boundaries will be high.

Jan 2012 AND 2011 S2 - 69/75 = 80 UMS

Just goes to show there will be a lot of people out there with high marks, so don't expect the A boundary to be about 60, in all likelihood, it will be about 65 or so, and the 100 will be anywhere between 73 and 75.

Still saying that, a lot of the best students in my school have lost a lot of marks, so it wouldn't be surprising to have the grade boundaries for A at roughly 65.

The one in q6, I'm pretty certain it's a two tailed, not only because there were only two hypothesis qs on the paper and the other one was a 1-tailer, but the question didn't reference looking for an increase/decrease on her prediction p=0.45, it just asked to test if it were true or not. In any case, in real life, you make your hypothesis first, then do the test (in this case, to find x=5)...you cannot say that because x=5 we are testing lower, because this implies you make your hypothesis after the test, or the question told you to test p<0.45 (which it didn't)....therefore it's a 2-tailed test. If you've used a 1-tailed, it's lucky because you'll only drop 1 mark as the conclusion was still the same

Reply 594

Original post by Zebrastripe

thank you, yeah i need an A so badly to meet my uni offers. i wasn't really well enough for this exam so didn't get much last minute revision done but am prepared to slave away hard for core 3 and core 4. so wish me luck!

hahaa good luck :tongue:

you'll be fine :P

Reply 595

chignesh10

15

Original post by uniphysics

You can IF you get the right answer :P

If the value you're looking for is in the tables then you kind of can't :P

No, that means half the questions we do cannot be right if we use the tables

Reply 596

BobLoblawLawBlog

16

Would this get the mark for the definition of a hypothesis:

A statement made about the value of a population parameter that you want to test

Reply 597

naimaasif

Original post by Skaterkid

Anyone started revision for it?

does anyone have an unofficial mark scheme for the exam?

Reply 598

Original post by uniphysics

I think all you needed to say was f(x) = 1 - F(x) but I'm not sure

I used a two tail test, but I think you would be able to do either because the value they gave you to test was at the lower tail so the upper tail wasn't really needed.

Did you find the critical region? I only did that because they asked for the definition so I thought it was just following up from that :P

Yhhh grade boundaries are likely to be like June 2011 I think.

5a. They said in the question, the probability that the mosquito lives longer than t days, is given by 225/(t+15)^2

this implies that P(T>t) = 225/(t+15)^2

P(T>t)=1-P(T<=t)
= 1-F(t)

hence 225/(t+15)^2=1-F(t)

rearrange to give you the c.d.f when t is greater than or equal to 0

Reply 599