The Student Room Group

Official Edexcel S3 thread Wednesday 20 May 2015 Morning [6691]

Scroll to see replies

Since the S3 exam is coming up, please feel free to put down any queries or confusion that you might have over the course.. I will try my best to answer it :wink:

S3 is a beast and hopefully we can all nail it !
Original post by DCMed96
P( |L-T|<0.1) basically means L-T can either be less than 0.1 or greater than -0.1 ...
you can approach this by writing this as P(-0.1<L-T<0.1), then you move the T to get P(-0.1+T<L<0.1+T) and then you move on the Z ....

I hope you can get it... it is quite annoying sometimes


I get how to get the answer, thanks! but I just wanna understand the reasoning behind it (lol, i know that's annoying question but it just bugs me when I don't understand something fully and im doing it mindlessly!)

(let Y=L-T, to make it easier to explain) can you explain why exactly the the modulus of Y is between -0.1 and 0.1? (i hoping Im making sense rn :P)
Original post by mmms95
I get how to get the answer, thanks! but I just wanna understand the reasoning behind it (lol, i know that's annoying question but it just bugs me when I don't understand something fully and im doing it mindlessly!)

(let Y=L-T, to make it easier to explain) can you explain why exactly the the modulus of Y is between -0.1 and 0.1? (i hoping Im making sense rn :P)


Dont you understand of significance of modulus?? brush up your C3 stuff....
modulus is making anything inside positive.... so inside could be either positive or negative..
Original post by DCMed96
Dont you understand of significance of modulus?? brush up your C3 stuff....
modulus is making anything inside positive.... so inside could be either positive or negative..


Yes I know the significance of the modulus :P anyways, I got it! (drew a diagram to understand) but thanks anyways :smile:


it's to do with confidence intervals, mean will be within confidence interval...
when you have to rank data, for spearman's rank correlation coefficient, will it always be ranking from highest to lowest? (for both sets of data?)
Chapter 1 - Not to bad (when you have a modulus bit then it gets hard!)
Chapter 2 - Sampling techniques is a killer!
Chapter 3 - Relatively ok
Chapter 4 - Hardest thing for me, always make stupid mistakes
Chapter 5 - Easy peasy

Also, is it worth doing the 2002 papers? I wasn't sure due to the different specification. But then again, stats is stats so it is all the same. Any guidance?

Also can some one quote me so i don't lose the thread? thanks
Original post by mmms95
when you have to rank data, for spearman's rank correlation coefficient, will it always be ranking from highest to lowest? (for both sets of data?)


An example
Row A, 1 is high 7 is low
Row B, 1 is low, 7 is high

The above wouldn't work - Both sets have to be ranked in the same way. It doesn't matter in the order of ranking due to the square. However, The way of ranking has to be consistent across both sets.

Eg
Rank 7 - rank 1 = 6 6^2 = 36

vs

Rank 1 - Rank 7 = -6 (-6)^2 = 36
Original post by mmms95
it's to do with confidence intervals, mean will be within confidence interval...

No it's not
Original post by getback339
No it's not


It is for the second part, this questions is really similar to June 2009 7b. It will be be;

(mu estimate) - (1.96 x (100/root of n))=20

the first mark is show the 'at least' 95%
Original post by tazza ma razza
An example
Row A, 1 is high 7 is low
Row B, 1 is low, 7 is high

The above wouldn't work - Both sets have to be ranked in the same way. It doesn't matter in the order of ranking due to the square. However, The way of ranking has to be consistent across both sets.

Eg
Rank 7 - rank 1 = 6 6^2 = 36

vs

Rank 1 - Rank 7 = -6 (-6)^2 = 36


Yeah, I know, I was asking if you, for example, having two sets of numerical data A/B, would you rank the highest number in A as one, and the highest number in B as 1 too? That's the way I always do it, just wanted to clarify if it was right, or we have to do according to the question?
Original post by mmms95
Yeah, I know, I was asking if you, for example, having two sets of numerical data A/B, would you rank the highest number in A as one, and the highest number in B as 1 too? That's the way I always do it, just wanted to clarify if it was right, or we have to do according to the question?


ahaha sorry! If it is not specified, then yeah you would do that ( i do it that way as well! ) If specified or in context ( eg times, heights, temperatures etc) then do it as it is in the question
Original post by tazza ma razza
ahaha sorry! If it is not specified, then yeah you would do that ( i do it that way as well! ) If specified or in context ( eg times, heights, temperatures etc) then do it as it is in the question


Alright thank you!! :smile:


Posted from TSR Mobile
I ****ing hate the definitions.
Original post by simonli2575
I ****ing hate the definitions.


How many are there that we actually need to know for the exam??

I hate the different methods on sampling etc...
Original post by DCMed96
How many are there that we actually need to know for the exam??

I hate the different methods on sampling etc...

Simple random sampling, lottery sampling, Quota sampling, Stratified sampling, systematic sampling, and secondary and primary data are all that I currently remember (Though not their definitions).
Is there any January papers?
Original post by Damien_Dalgaard
Is there any January papers?


there is one in 2006 i think, but s3 is usually taken place is summer only
Original post by DCMed96
there is one in 2006 i think, but s3 is usually taken place is summer only


Any reason how come :smile:

Just for more practice, I will put links in the OP to papers etc.

Quick Reply

Latest

Trending

Trending