Ocr (mei) s2 exam - 22/06/11

Reply 20

Rakesh26

Original post by Ian.

As long as you defined your units, they can't take marks from you :smile:

Ooh okay..I can't remember if I did :tongue:

Ah well, can't do anything about it now :smile:

I'm pretty sure that for both the hypothesis tests on the paper - both came about to be insignificant (Insufficient evidence to reject the null hypothesis etc).. anyone else get the same? :smile:

Reply 21

poiuy

10

hopefully. only 1 person out of 9 of us did, and if on the markscheme you have to calculate the s.d, i think too many people wouldn't, so they may just allow us the marks. It was pretty sneaky.

Reply 22

sebbyws

For drawing the two Normal Curves for the apples.... did you do a normal curve with Mean 210.5 in the middle, and for the cropped apples, draw another curve to the left with mean 185??

Reply 23

supercal35

with the poisson approximation , did anyone get a really small number like 2x10^-16, i wasnt too sure what they wer asking for , i read it as what is the probality of one or no boxes out of a hundred having less than 220g for all 3 apples . is this correct ?

(edited 12 years ago)

Reply 24

supercal35

Original post by sebbyws

For drawing the two Normal Curves for the apples.... did you do a normal curve with Mean 210.5 in the middle, and for the cropped apples, draw another curve to the left with mean 185??

thts what i did :smile:

Reply 25

xChelsea

OP

Original post by sebbyws

For drawing the two Normal Curves for the apples.... did you do a normal curve with Mean 210.5 in the middle, and for the cropped apples, draw another curve to the left with mean 185??

Yeah i did the same but i had the 210.5 one was lower than the 185 one as the standard deviation was larger meaning the spread of the data was larger.

Reply 26

Beth1234

11

Original post by Rakesh26

Ooh okay..I can't remember if I did :tongue:

Ah well, can't do anything about it now :smile:

I'm pretty sure that for both the hypothesis tests on the paper - both came about to be insignificant (Insufficient evidence to reject the null hypothesis etc).. anyone else get the same? :smile:

I believe I got that... did you find there was quite a big difference between the critical value and the test statistic for the last question? I did...

Also - did everyone do Binomial for question 3 (the bit about 100 packets), and then approximate it with Poisson? I did that... saw it was five marks, and was slightly surprised. It didn't seem very much work.

Reply 27

Beth1234

11

Original post by supercal35

with the poisson approximation , did anyone get a really small number like 2x10^-16, i wasnt too sure what they wer asking for , i read it as what is the probality of one or no boxes out of a hundred having less than 220g for all 3 apples . is this correct ?

Yeah, I did that - except I didn't get something that small! Do you still rememeber what you got for the bit before? (the bit where you had work out the probability of three apples all being at least 220g or something like that)

Reply 28

Lewis246

Loved it

(well, as far as maths exams go...) :redface:

First part of question 4 threw me a bit though, although I got it right in the end, and I doubt I picked up all 9 marks on the last part of Q4, as I only used half a page of working.

Can't see myself getting full marks on the question where you have to sketch the normal curves as well, as I'm crap at that...

But aside from that it was great!

Reply 29

xChelsea

OP

Original post by Beth1234

Do you still rememeber what you got for the bit before? (the bit where you had work out the probability of three apples all being at least 220g or something like that)

I think i got 0.01882 for 3)ii)

Reply 30

Evangelica

12

Original post by Rakesh26

I got y = 0.058x + 0.78...

Damn. I'm terrible at knowing how many decimal places to put my answer to :frown:

Now I feel stupid :frown:

Hopefully they give method marks...

I got what you did except I got y=o.5833 + 0.78 I think, just more sensitivity.

Reply 31

Evangelica

12

Original post by xChelsea

I think i got 0.01882 for 3)ii)

I got that! Oh dear I think I only did it to 4dp 0.0188. I hope I don't lose marks for that :smile:

Reply 32

Beth1234

11

Original post by xChelsea

I think i got 0.01882 for 3)ii)

I got 0.0115, or something like that... and I have absolutely no idea if I am right! All I know is that it made a great deal of sense at the time :confused:

(edited 12 years ago)

Reply 33

xChelsea

OP

Original post by Beth1234

I got 0.115, or something like that... and I have absolutely no idea if I am right! All I know is that it made a great deal of sense at the time :confused:

I don't know.
i got 0.266 for 3)i) and then i cubed this answer to get 0.01882.

Reply 34

supercal35

Original post by Beth1234

Yeah, I did that - except I didn't get something that small! Do you still rememeber what you got for the bit before? (the bit where you had work out the probability of three apples all being at least 220g or something like that)

yeah i think i got np as 36 point something so the probability i got was 0.36something (not good at remembering decimals )

Reply 35

Beth1234

11

Original post by supercal35

yeah i think i got np as 36 point something so the probability i got was 0.36something (not good at remembering decimals )

I'm afraid that can't really be right... I'm sure you can only approximate a binomial distribution with a Poisson distribution if np is not too large (as a rule of thumb, less than 5, I believe. Certainly not more than 10).

(edited 12 years ago)

Reply 36

Beth1234

11

Original post by xChelsea

I don't know.
i got 0.266 for 3)i) and then i cubed this answer to get 0.01882.

I did that same method.

Actually, now I'm playing around with the calculator - I must have typed 0.226 into the calculator...

Oh well. One mark lost. It didn't change anything much in the next part - I was still able to do a Poisson distribution. That's fine... just don't let me find anything else! :redface:

Reply 37

sebbyws

Original post by Beth1234

I'm afraid that can't really be right... I'm sure you can only approximate a binomial distribution with a Poisson distribution if np is not too large (as a rule of thumb, less than 5).

you may approximate the binomial using poission providing that N is LARGE and P is SMALL.

it is when you approximate the binomial using a normal distribution in which P has to be close to 0.5.

Reply 38

Ian.

2

Original post by sebbyws

For drawing the two Normal Curves for the apples.... did you do a normal curve with Mean 210.5 in the middle, and for the cropped apples, draw another curve to the left with mean 185??

Original post by supercal35

thts what i did :smile:

You have to be careful, there were 4 marks available here. They would want you to express how spread out the data for each distribution curve was. One was wider, but shallower than the other one, which was thinner and taller.

Original post by Beth1234

I'm afraid that can't really be right... I'm sure you can only approximate a binomial distribution with a Poisson distribution if np is not too large (as a rule of thumb, less than 5, I believe. Certainly not more than 10).

I disagree here. Binomial can be approximated by the poisson and the normal; Poisson can only be approximated by the normal, and the normal can't be approximated (in S2). You can approximate the binomial with the normal by setting your mean as np, and your standard deviation as np(1-p).

(edited 12 years ago)

Reply 39

Rakesh26

Original post by Beth1234

I believe I got that... did you find there was quite a big difference between the critical value and the test statistic for the last question? I did...

Also - did everyone do Binomial for question 3 (the bit about 100 packets), and then approximate it with Poisson? I did that... saw it was five marks, and was slightly surprised. It didn't seem very much work.

Hmm on the last question, I think I got '-1.504' as the test statistic, which isn't that far from the critical value (-1.645 I think). But I'm not 100% sure :s-smilie: