The Student Room Group

AS Edexcel S1 - (5th June 2015) Exam Thread

Scroll to see replies

Does anyone have a list of coding questions that have come up in past exam papers?
I haven't touched upon coding yet... :redface:
will percentiles come up? because im not really sure how to do them..

also when the question asks which average is more suitable how do you answer that?
Original post by .catherine.
will percentiles come up? because im not really sure how to do them..

also when the question asks which average is more suitable how do you answer that?


Its in the specification so as with everything else which features in the spec, there is a chance it will come up. Same method as using interpolation to find quartiles but you're trying the find the (p/100)(n)th piece of data where p is the Pth percentile. So if they ask for the 60th percentile - multiple p/100 by the sum of frequencies and interpolate as you do for quartiles.

Use quartiles/IQR for skewed data as mean and S.d is effected by skew and extreme values.
Original post by Bustamove
When drawing a box plot with outliers, do you end the "whiskers" at the outlier limits or do you end it at the next value which is not outside the outlier?

are both of these acceptable?

When I say "whiskers", I don't know the exact word for it but it's the points where the highest value and the lowest value is pointed out when there is no outliers
(Hope this makes sense)

help?


For this I've been taught to do the outlier boundary if there's an outlier on that side of the box and whisker otherwise just use your highest/lowest values


Posted from TSR Mobile
For rounding numbers...how many sf/dp should we round?
Also is there a fixed way to round the value of money?like a whole number?
Reply 385
I find Probability so hard and I've done all the Past paper Qs relating to them anyone know what else can be done Cuz I get stuck on them so so much
Original post by iman.k
I find Probability so hard and I've done all the Past paper Qs relating to them anyone know what else can be done Cuz I get stuck on them so so much


what exactly are you stuck on, if you've done all the past papers you should be ready to go?
Hey, can someone tell me why you'd use Swl/Sll and not Swl//Sww in this question?



Thank you!!

Also, does anyone know where I can get a good histogram explanation? I SUCK at histograms (esp. telling the width and height of a bar).
Original post by Hai_Ann
Does anyone have a list of coding questions that have come up in past exam papers?
I haven't touched upon coding yet... :redface:


that isn't part of Edexcel stats
Original post by Hai_Ann
Does anyone have a list of coding questions that have come up in past exam papers?
I haven't touched upon coding yet... :redface:


Original post by Virolite
that isn't part of Edexcel stats


I think coding is in edexcel S1... there was a coding question in june 2014 paper, question 2... nothing really major though... There's also one on jan 2011 question 4.... and I'm sure there's a few more here and there in the edexcel papers..
Original post by Hai_Ann
Does anyone have a list of coding questions that have come up in past exam papers?
I haven't touched upon coding yet... :redface:


There's coding in the June 2014 (R) paper


Posted from TSR Mobile
Original post by кяя
Also, does anyone know where I can get a good histogram explanation? I SUCK at histograms (esp. telling the width and height of a bar).


You want to compare class width to the width of a bar to find the width of another classes' bar, then use "f = FD x CW" to work out the height as FD is common to the height of both bars (if that makes any sense)
Original post by кяя
Hey, can someone tell me why you'd use Swl/Sll and not Swl//Sww in this question?.


Original post by Virolite
you would need to use Swl and Sww to find b in the equation of the regression line


For that question, you use Swl/Sll to find b instead of Swl/Sww because the regression line is w=a+bl
if you look at bl, it has to be Sll instead of Sww because l is the function and b is just the coefficient.... if it was bw, then you use Sww instead of Sll

think of it like this:
Equation for regression line y=a+bx

so therefore, to find b, it's b=Sxy/Sxx
Sxx is used instead of Syy because x is the function and b is just the coefficient in the regression line

I hope this makes sense :smile:
(edited 8 years ago)
Original post by Bustamove
I think coding is in edexcel S1... there was a coding question in june 2014 paper, question 2... nothing really major though... There's also one on jan 2011 question 4.... and I'm sure there's a few more here and there in the edexcel papers..


Maybe in the international papers but it's not in the standard stats spec
Original post by Bustamove
I think coding is in edexcel S1... there was a coding question in june 2014 paper, question 2... nothing really major though... There's also one on jan 2011 question 4.... and I'm sure there's a few more here and there in the edexcel papers..


that's just using the expected mean, not necessarily coding as such
Original post by Bustamove
For that question, you use Swl/Sll to find b instead of Swl/Sww because the regression line is w=a+bl
if you look at bl, it has to be Sll instead of Sww because l is the function and b is just the coefficient.... if it was bw, then you use Sww instead of Sll

think of it like this:
Equation for regression line y=a+bx

so therefore, to find b, it's b=Sxy/Sxx
Sxx is used instead of Syy because x is the function and b is just the coefficient in the regression line

I hope this makes sense :smile:


Ahh, that makes sense now. Thanks!

So if the equation was, for example, x=a+by it would be Sxy/Syy?
Original post by Virolite
that's just using the expected mean, not necessarily coding as such


yea, I haven't found anything which is massive in terms of coding, only small minor questions such as those, but it's still there... gonna learn it just in case :wink:
(edited 8 years ago)
Original post by Virolite
Maybe in the international papers but it's not in the standard stats spec


There IS coding in the Edexcel specification. It is under Measures of location mean, median, mode: "Data may be discrete, continuous,grouped or ungrouped.Understanding and use of coding."
Original post by кяя
Ahh, that makes sense now. Thanks!

So if the equation was, for example, x=a+by it would be Sxy/Syy?


technically yes... but it will never be written as x=a+by as y tends to be the thing that the regression line tries to work out.

They'll usually give you the regression line in the question e.g. m=a+bn...
and then what b is the coefficient to, you can base it on that :smile:
Original post by Virolite
that's just using the expected mean, not necessarily coding as such


June 2014:

2. The mark, x, scored by each student who sat a statistics examination is coded usingy = 1.4x 20The coded marks have mean 60.8 and standard deviation 6.60Find the mean and the standard deviation of x.

Quick Reply

Latest

Trending

Trending