The Student Room Group

Edexcel S3 - Wednesday 25th May AM 2016

Scroll to see replies

Does anyone have any advice for the sampling method questions? Can never seem to get them right :colonhash:

January 2006 Question 1 springs to mind:
S3 January 2006 Q1.png
Reply 261
Original post by Zacken
No, you need to decide whether the test is one tailed or two tailed from the context.

So just a made up example,
Claim is that as book pages rise, the number of borrows decrease.

If the question was not to test the claim but to see if there was positive correlation between the 2 variables I'm sure you'd test p <0 ? as your ranks calculated are assuming the claim is correct.
Original post by Rkai01
So just a made up example,
Claim is that as book pages rise, the number of borrows decrease.

If the question was not to test the claim but to see if there was positive correlation between the 2 variables I'm sure you'd test p <0 ? as your ranks calculated are assuming the claim is correct.


Yeah, you test p < 0 because it's a one tailed test.
Original post by Euclidean
Does anyone have any advice for the sampling method questions? Can never seem to get them right :colonhash:

January 2006 Question 1 springs to mind:
S3 January 2006 Q1.png


What went wrong in that one / what have you tried?
Reply 264
For June 2012 question 3d. Surely it's implying the garage changes the brakes every 20000 so technically they are changing them too late as it is supposed to be within the interval. Why is it the other way around?
Original post by SeanFM
What went wrong in that one / what have you tried?


Let me type up what I wrote:

15(30)+150=15(40)=60015(30) + 150 = 15(40) = 600 pupils (first B1 mark)

The headteacher should assign each boy with a number in ascending order from 1-300 inclusive, likewise they should assign each girl a number in ascending order from 1-300 inclusive. They should use random number tables to select 20 boys and 20 girls ignoring repeat numbers between 1-300. Each selection of 20 boys and 20 girls should be combined to form a strata of size 40 which has the same split of 50/50 boys and girls as the population. 15 of these strata can be formed from the population. The teacher should then sample from strata.


S3 January 2006 Q1 MS.png

I gave myself 1/7 for that answer.

What I guess I didn't realise was that the strata need representatives from each class to make it representative of the population. But in short I think I just don't understand how to answer the questions in the way they want me to.

I normally make a note of things I don't get right on past papers on the front, and from the 6 or so I've done sampling questions like these are a recurring topic :redface:
Original post by Euclidean
Let me type up what I wrote:



S3 January 2006 Q1 MS.png

I gave myself 1/7 for that answer.

What I guess I didn't realise was that the strata need representatives from each class to make it representative of the population. But in short I think I just don't understand how to answer the questions in the way they want me to.

I normally make a note of things I don't get right on past papers on the front, and from the 6 or so I've done sampling questions like these are a recurring topic :redface:


I think 1/7 would be a bit harsh, I'm not sure what the examiner would have to say.

I think you just need to look at the 6 questions again and see what there needs to be learnt. I can't remember exactly what they ask, but I guess the general way is to create a 'class' for each distinct object. Eg a girl and a boy in class 1, a girl and a boy in class 2 are different classes, all the way up to 6th form.
Original post by Rkai01
For June 2012 question 3d. Surely it's implying the garage changes the brakes every 20000 so technically they are changing them too late as it is supposed to be within the interval. Why is it the other way around?


Please link the paper in the future if you can - I felt particularly lazy and didn't want to find it myself.

This is a very, very, very nasty question. I would have thought and written what you did, that 20,000 is too late to change it as seen by the confidence interval.

However, you have to read that sentence in another way. You could read it as: Brake pads are supposed to be changed every 20,000 miles. (The made bit means created with the intention of it lasting for 20,000 miles on average).

Then you have to think about what the xi's actually are - the actual times that the brake pads are replaced. These times aren't necessarily when the brake pads are useless or stop working, it's just when they are changed. They are just being 'safe' and replacing them earlier than how long they actually would last for, which is 20,000 on average.

If that's confusing, I could try rewording it. It is a very nasty question.
Reply 268
are you guys ready for the upcoming s3 exam?
I can't seem to understand S3 June 2009 question 2, I got 20.4 and 19.8 but the answer is 19.1 and 21.1. Please help.

The question:
The heights of a random sample of 10 imported orchids are measured. The mean height ofthe sample is found to be 20.1 cm. The heights of the orchids are normally distributed.
Given that the population standard deviation is 0.5 cm,
(a) estimate limits between which 95% of the heights of the orchids lie
Original post by astarsian
I can't seem to understand S3 June 2009 question 2, I got 20.4 and 19.8 but the answer is 19.1 and 21.1. Please help.

The question:
The heights of a random sample of 10 imported orchids are measured. The mean height ofthe sample is found to be 20.1 cm. The heights of the orchids are normally distributed.
Given that the population standard deviation is 0.5 cm,
(a) estimate limits between which 95% of the heights of the orchids lie


They want the limits between which 95% of all the heights of the orchids lie, not just the mean height lies.

So it's 20.1 +- 1.96 * 0.5, there's not /sqrt(10). That's only if you want the interval that the mean lies in.
Original post by Zacken
They want the limits between which 95% of all the heights of the orchids lie, not just the mean height lies.

So it's 20.1 +- 1.96 * 0.5, there's not /sqrt(10). That's only if you want the interval that the mean lies in.


oh okay, is there a formula for working out limits?
Please can anyone help with this question, how do you know to take less than or equal to 6 as 5 and greater than or equal to 12 as 12, and how do you work out the mean being there are gaps in the classess, do you use the midpoint of each class? Thanks :smile:image.jpg
Original post by astarsian
oh okay, is there a formula for working out limits?

If finding the 95% confidence inteval for the mean the formula is x bar plus minus ( 1.96 multiplied by the standard error) and if finding the 95% confidence interval for the mean the formula is x bar plus minus ( 1.96 multiplied by the standard deviation)
Hope that helps :smile:
Original post by economicss
Please can anyone help with this question, how do you know to take less than or equal to 6 as 5 and greater than or equal to 12 as 12, and how do you work out the mean being there are gaps in the classess, do you use the midpoint of each class? Thanks :smile:


Yea, you use the midpoints to get the mean.
Original post by Zacken
Yea, you use the midpoints to get the mean.


Thank you :smile: is it just arbitrary that they took less than or equal to 6 as being 5, as I thought you would use 6? Thanks!
Original post by economicss
Thank you :smile: is it just arbitrary that they took less than or equal to 6 as being 5, as I thought you would use 6? Thanks!


I would have used 6 as well, but I don't think it really matters what you label it as as long as you get the expected frequency and observed frequency correct; the textbook has a section on this, have you read it?
Original post by Zacken
I would have used 6 as well, but I don't think it really matters what you label it as as long as you get the expected frequency and observed frequency correct; the textbook has a section on this, have you read it?

Thank you, erm I hope so, don't suppose you have a page number I could refer to please? :smile: Thanks
Original post by economicss
Thank you, erm I hope so, don't suppose you have a page number I could refer to please? :smile: Thanks


Page 84-87.
Original post by Zacken
Page 84-87.

Thank you!

Quick Reply

Latest