Hey there! Sign in to join this conversationNew here? Join for free
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    Anyone taking this? Looking forward to it. Never done an exam with just four questions before. Wonder if time is an issue or not.
    Offline

    0
    ReputationRep:
    I've got it after D1, all of the past papers I've tried were pretty easy for it so hopefully it'll be okay. Hopefully they won't do like they did for Core 3 this year and make the paper much harder than any of the past ones.
    Offline

    0
    ReputationRep:
    The exam structure of S2 is a lot more straightforward than that of C3. there isnt much they can test us on. hopefully it wont be unusual
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    Feel sorry for anybody doing S1, S2 and D1 on the same day. probably the three most boring exams.
    Offline

    2
    ReputationRep:
    I've got this and S1 tommorow morning. Done all the past papers before but not really do any stats in about 3 weeks, planning on doing as many papers as possible this afternoon/tonight to jog my memory.
    Offline

    2
    ReputationRep:
    Was pretty good IMO, finished with bags of time left and ended up leaving about 20mins early and was happy with all my answers . Was weird having to draw the scatter diagram, accidently did the axes the wrong way around at first but re-did it at the end Also had a couple of issues with not being sure how to get the correct values from the normal tables for phi(0.766666) or something, so I reckon I may have dropped a couple of marks there if I got it wrong, but I reckon I'll still have gotten an A*.
    Offline

    1
    ReputationRep:
    Seemed pretty straight forward. For a value such as 0.76666666666 youd have to linearly extrapolate between the two points either side (0.76 and 0.77)
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    Wasn't too bad at all. Drawing a scattergram was pretty stupid. I also made a thread asking about what to do for values like 0.7666 and apparently you can just round it to 3 and markschemes are pretty nice about it.
    Offline

    0
    ReputationRep:
    Seemed an OK test, no harder/easier than the past papers.

    One question, on the final part of the laptops question (2000 laptops < 50 faulty or something) were we supposed to use Normal approximation? I thought for Normal approximation p is supposed to be "not too close to 0 or 1" but p=0.02 seems pretty close to 0 :confused:
    Offline

    8
    ReputationRep:
    Firstly, on the "lowest score that the top 10% would achieve" question, did you have to round that or leave it as a decimal. I think it was something like 70.04 which I made 71 since it had to be greater, but it didn't say integer scores in the question so I don't know.

    On the very last part of Q4 (Normal mean test), did anyone get a Z value beyond those given on the tables (~3.9)? I did and said that would allow me to reject H0 anyway, but if that's wrong then how did you arrive at your answer?

    On the regression line, I realised at the end that I had got it wrong and corrected it by crossing out the wrong work, but didn't have time to use the correct value in the follow-through parts. Would I still get the follow-through marks?
    Offline

    0
    ReputationRep:
    On the very last part of Q4 (Normal mean test), did anyone get a Z value beyond those given on the tables (~3.9)? I did and said that would allow me to reject H0 anyway, but if that's wrong then how did you arrive at your answer?
    Yeah, I got -3.9
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    With the hypothises test you could end up with -3.9 and use the test statistic way. Then use the inverse table values for the 5% level or whatever it was couldn't you? You reject it (and by miles) I think. And I think you accept the Chi-squared test (by miles again) as well. Anybody agree?

    And yeah, I was a bit thrown by using a normal approximation with p = 0.02. I used it anyway and said it's a bit weird at the start. Also strange how (I think and hope) all the normal distrubtions question required continuity corrections. I think I got the mark needed to be 60, by the way.

    Can't remember the equation of the regession line. But A was the independent and T was dependent variable. It was like t = 0.285a + 623.321 or something, wasn't it?
    Offline

    8
    ReputationRep:
    (Original post by Tallon)
    With the hypothises test you could end up with -3.9 and use the test statistic way. Then use the inverse table values for the 5% level or whatever it was couldn't you? You reject it (and by miles) I think. And I think you accept the Chi-squared test (by miles again) as well. Anybody agree?
    Yes, to both.

    And yeah, I was a bit thrown by using a normal approximation with p = 0.02. I used it anyway and said it's a bit weird at the start. Also strange how (I think and hope) all the normal distrubtions question required continuity corrections. I think I got the mark needed to be 60, by the way.
    I'm sure it was normal, Poisson had too high a lambda to allow you to get probablities. Maybe since n is really really high, p is allowed to be small anyway?

    Can't remember the equation of the regession line. But A was the independent and T was dependent variable. It was like t = 0.285a + 623.321 or something, wasn't it?
    Yes. I checked it with my graphical calculator, just not in time to correct the other parts.

    Did you round it up to an integer (61) for the 'lowest mark gained by the top10%'? Or leave it as like 60.064?
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    (Original post by Game_boy)
    Yes, to both.



    I'm sure it was normal, Poisson had too high a lambda to allow you to get probablities. Maybe since n is really really high, p is allowed to be small anyway?



    Yes. I checked it with my graphical calculator, just not in time to correct the other parts.

    Did you round it up to an integer (61) for the 'lowest mark gained by the top10%'? Or leave it as like 60.064?

    did you get 60 as well? I thought you got 70? Or can't remember.

    I think I put down 61. However, thinking about it now, I think it's 60.
    Offline

    2
    ReputationRep:
    (Original post by dogie)
    Seemed pretty straight forward. For a value such as 0.76666666666 youd have to linearly extrapolate between the two points either side (0.76 and 0.77)
    That's what I did do, I just worry it'll be 0.0001 off ah well no point worrying about it

    (Original post by MarkS2)
    Seemed an OK test, no harder/easier than the past papers.

    One question, on the final part of the laptops question (2000 laptops < 50 faulty or something) were we supposed to use Normal approximation? I thought for Normal approximation p is supposed to be "not too close to 0 or 1" but p=0.02 seems pretty close to 0 :confused:
    I used Normal as np was too large for Poisson. I think too close to 0 is only if it's really really small, i.e like 0.0001.

    (Original post by Game_boy)
    Firstly, on the "lowest score that the top 10% would achieve" question, did you have to round that or leave it as a decimal. I think it was something like 70.04 which I made 71 since it had to be greater, but it didn't say integer scores in the question so I don't know.

    On the very last part of Q4 (Normal mean test), did anyone get a Z value beyond those given on the tables (~3.9)? I did and said that would allow me to reject H0 anyway, but if that's wrong then how did you arrive at your answer?

    On the regression line, I realised at the end that I had got it wrong and corrected it by crossing out the wrong work, but didn't have time to use the correct value in the follow-through parts. Would I still get the follow-through marks?
    Said nothing about integer scores so I left it.

    Yeah my z value was like -3.9 but I do it the test statistic way so rejected H0 by miles.

    Maybe, maybe not.

    (Original post by Tallon)
    With the hypothises test you could end up with -3.9 and use the test statistic way. Then use the inverse table values for the 5% level or whatever it was couldn't you? You reject it (and by miles) I think. And I think you accept the Chi-squared test (by miles again) as well. Anybody agree?

    And yeah, I was a bit thrown by using a normal approximation with p = 0.02. I used it anyway and said it's a bit weird at the start. Also strange how (I think and hope) all the normal distrubtions question required continuity corrections. I think I got the mark needed to be 60, by the way.

    Can't remember the equation of the regession line. But A was the independent and T was dependent variable. It was like t = 0.285a + 623.321 or something, wasn't it?
    I agree with your answers to the hypothesis testing

    Yes the normal distribution one did need a continuity correction. I can't remember my mark needed :\

    I got the same on the dependent/independent variable stuff and said it's because the altitude is chosen but the time to take off depends on the altitude. the linear regression line was t = 0.258a + 623.23 (3sf on a coefficient, 2dp on constant), I still have the data stored in my calculator.
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    ugh... you guys and your fancy calculators.....

    Proper mathletes, right?



    by the way, I'm pretty sure the question said the uniform score was rounded to the nearest integer? I'm sure it was 60 now, even though I forget the exact details of the question.
    Offline

    8
    ReputationRep:
    (Original post by Tallon)
    did you get 60 as well? I thought you got 70? Or can't remember.

    I think I put down 61. However, thinking about it now, I think it's 60.
    I can't remember if 60 or 70. I know it was xx.0xx, and I rounded that up.
    If it was meant to be an integer, you were certainly meant to round it up as it was score > xx.0xx so the lowest integer that meets that would be above that.
    Offline

    8
    ReputationRep:
    (Original post by Tallon)
    ugh... you guys and your fancy calculators.....

    Proper mathletes, right?
    We get them for doing FP2; there aren't many of use doing Further Maths. We can then take them into our normal exams since we have them already.
    • PS Helper
    • Thread Starter
    Offline

    4
    ReputationRep:
    PS Helper
    (Original post by Game_boy)
    I can't remember if 60 or 70. I know it was xx.0xx, and I rounded that up.
    If it was meant to be an integer, you were certainly meant to round it up as it was score > xx.0xx so the lowest integer that meets that would be above that.


    That's what I thought at the time.
    But if you round it up it means slightly less than the top 10% get that mark, whereas if you round it down just more than 10% get that mark.

    So which did they want :s?
    Offline

    0
    ReputationRep:
    i rounded to 60.1...damn it...

    i got negative for the last one and i think accepted H0 for chi squared (too stressful to think after the horrible bio paper)

    overall...i think i was a nice paper....i came out and was like: "thank god there wasn't any commenting on contribution of chi squared"
 
 
 
  • See more of what you like on The Student Room

    You can personalise what you see on TSR. Tell us a little about yourself to get started.

  • Poll
    Brexit voters: Do you stand by your vote?
  • See more of what you like on The Student Room

    You can personalise what you see on TSR. Tell us a little about yourself to get started.

  • The Student Room, Get Revising and Marked by Teachers are trading names of The Student Room Group Ltd.

    Register Number: 04666380 (England and Wales), VAT No. 806 8067 22 Registered Office: International House, Queens Road, Brighton, BN1 3XE

    Write a reply...
    Reply
    Hide
    Reputation gems: You get these gems as you gain rep from other members for making good contributions and giving helpful advice.