The Student Room Group

Ocr mei s3 june 2015

Scroll to see replies

I got 90 for n and the last question pissed me off I tried everything but still got the smallest of probabilities. You know you're doing something wrong when you get 0.000000000000000000000000000008 or similarly ridiculous
Original post by chizz1889
I will do an unofficial mark scheme after I've drove home if u guys want!


Sound good man
Original post by chizz1889
Had a sudden realisation which saved me on the last question, got it to be distributed ~N(0,9.75). I got the sample size to be 90. What about you?


Spill the beans mate
Reply 23
For the last question for the variance I did 1/4 and 1/9 of the variances in order to get the "mean" variance?

Posted from TSR Mobile
Last question i messed up, got n=90
wat did evry1 get for the tests
got 0.77 something for another part of the last qs
Here's my answers that I can remember :
1. a) i) cant remember the ratios but it was a simple question just make sure they add up to 500 afterwards
ii) I put by gender and also by year group
iii) adv of staratified sample is that the sample contains data from each strata in the same ratio as it exists in the population so is more representative.
b) i)the underlying distribution is unknown.
ii) the population must be symmetrically distributed about the population median.
iii) W-=21, critical value =17. Not significant.
2.i) integrated function between 0&5 =1 k=2/35
ii) show that
iii) make sure you group 0<x<2 chi squared =7.552 CV = 6.251 , significant etc
3 degrees of freedom, result was significant etc
iv) significantly less than expected in 0<x<2. Significantly more than expected in 3<x<4.
3. i) 5% chance of rejecting H0 when it is true.
ii) sample is random, differences must have underlying normally distributed population. (Variance is unknown?)
iii) CV & TS were 2.something, TS>CV, result was significant etc
iv) 10% t9 = 1.833 d mean (4.93 +- 1.833*6.3105/root10)
4.i) in repeated samples , 95% of confidence interval constructed in the same way will contain the true population mean
ii) I got n=90 and mean = 46.2999 (46.3)
iii) X~N(86.1,25.9) then did less than 90 - 0.7783
iv) Y~N(3.5,25.9) then did more than 5 - 0.3841
v) G~N(0,9.75) then did more than 3 - 0.1683

BAR ANY ERRORS, IT's COMPLETE :biggrin:
(edited 8 years ago)
Original post by Dan205
For the last question for the variance I did 1/4 and 1/9 of the variances in order to get the "mean" variance?

Posted from TSR Mobile

Close it was 1/2 and 1/3. Will post a pic in a second why
Reply 28
Original post by chizz1889
Close it was 1/2 and 1/3. Will post a pic in a second why


Okay cheers :smile:

Posted from TSR Mobile
Original post by Dan205
For the last question for the variance I did 1/4 and 1/9 of the variances in order to get the "mean" variance?

Posted from TSR Mobile

image.jpg
Reply 30
Original post by chizz1889
image.jpg


Yeah I got that! :smile: cause 1/9 is 1/3 squared and 1/4 is 1/2 squared :tongue:

Posted from TSR Mobile
Anyone have idea of confidence interval? Did the question ask for differences or just one row?
Original post by Dan205
Yeah I got that! :smile: cause 1/9 is 1/3 squared and 1/4 is 1/2 squared :tongue:

Posted from TSR Mobile


Yeah then you had to add the variances of each so it ends up being 1/2 & 1/3 😀
Original post by chizz1889
Here's my answers that I can remember :
1. a) i) CANT REMEMBER
ii) CANT REMEMBER
b) i) there's a significant difference after the drive so this test eliminates these differences (I had no idea)
ii) the population must be symmetrically distributed about the population median.
iii) W-=21, critical value =17. Not significant.
2.i) integrated function between 0&5 =1 k=2/35
ii) show that
iii) make sure you group 0<x<2
3 degrees of freedom, result was significant etc
iv) significantly less than expected in 0<x<2. Significantly more than expected in 3<x<4.
3. i) 5% chance of rejecting H0 when it is true.
ii) variance must be unkwown, sample is small, differences must have underlying normally distributed population
iii) result was significant etc
iv) CANT REMEMBER
4.i) in repeated samples , 95% of confidence interval constructed in the same way will contain the true population mean
ii) I got n=90 and mean = 46.2999
iii) X~N(86.1,25.9) then did less than 90
iv) Y~N(3.5,25.9) then did more than 5
v) G~N(0,9.75) then did more than 3



Feel free to help by adding new to the parts I can't remember or make corrections ☺️

for first qs i think one of the numbers i got was 118 and another was 124
for the second question strata like gender, rural/urban, type of house
for chi squared got around 7.something for test statistic and 6. soemthing for critical value
for qs 3 2.262 was crit value i think and test statisticc was a bit bigger 2.7? cant remeber, for CI use k=1.833

for qs4 could you please explain last part, would appreciate it
Can anybody remember what Q1A was (all of it) and also Q3iv?
Reply 35
Original post by chizz1889
Yeah then you had to add the variances of each so it ends up being 1/2 & 1/3 😀


Cool, I'm happy now :tongue:

Posted from TSR Mobile
Original post by mathsexam2013
for first qs i think one of the numbers i got was 118 and another was 124
for the second question strata like gender, rural/urban, type of house
for chi squared got around 7.something for test statistic and 6. soemthing for critical value
for qs 3 2.262 was crit value i think and test statisticc was a bit bigger 2.7? cant remeber, for CI use k=1.833

for qs4 could you please explain last part, would appreciate it


Oh yes! Cheers and look above I have shown in a pic why the variance is 9.75 😊
Original post by chizz1889
Oh yes! Cheers and look above I have shown in a pic why the variance is 9.75 😊


thanks for that, were your answers same as the ones i posted?
Reply 38
What are we thinking for grade boundaries?
yeah they were

Quick Reply

Latest

Trending

Trending