Hey there! Sign in to join this conversationNew here? Join for free
    Offline

    3
    ReputationRep:
    What accuracy level are you supposed to go to when comparing expected and observed frequencies when the expected freq is a long decimal?

    Also if expected frequencies come out as a fraction can they be left like this? Sometimes if you do this you get an answer more accurate than the mark scheme - would this lose a mark?

    Thanks
    Offline

    3
    ReputationRep:
    Any point in researching into the actual concepts behind chi-squared distribution?
    Offline

    22
    ReputationRep:
    (Original post by economicss)
    So if a question asks what assumptions have been made, am I right in thinking it could never be that we've assumed the population is normally distributed with the central limit theorem? Not sure on my wording there, thanks
    Yep.
    • Very Important Poster
    • Thread Starter
    Offline

    21
    ReputationRep:
    Very Important Poster
    (Original post by econam)
    What accuracy level are you supposed to go to when comparing expected and observed frequencies when the expected freq is a long decimal?

    Also if expected frequencies come out as a fraction can they be left like this? Sometimes if you do this you get an answer more accurate than the mark scheme - would this lose a mark?

    Thanks
    I stuck to 4 d.p until the end and then in mark schemes, they seem to have answers which round to 2 or 3 d.p :dontknow: I don't think you'd lose a mark for that - and fractions would probably okay till the end.
    Offline

    3
    ReputationRep:
    (Original post by Euclidean)
    To be honest, aside from some combinations of variables trouble I had yesterday (squaring not squaring etc), I got all of S3 on my first run-through in like 5 hours. Confidence intervals just sort of make sense if you think about the concept.

    Error we can assume is normally distributed. It's value depends on how big of a sample we take, so bigger sample means smaller error. The interval depends on what percentage (for 90% this is 5% either side) we don't want to concern ourselves with on either side of the true mean so we can use normal tables to find that corresponding Z value (as error is normally distributed).

    That's probably a patchy explanation in terms of the actual statistics but it's how I reasoned with the idea.
    You mentioned to had done solomon papers, any hard ones?
    Offline

    2
    ReputationRep:
    (Original post by physicsmaths)
    Yeh you get the percentage 1- it and half it and look in the samll P(Z>z) tables at the back. Then do
    x+-numbers(sd/sqrtn)


    Posted from TSR Mobile
    Makese sense thanks
    (Original post by Katiee224)
    Quick question, if we get tied ranks when calculating spearmans rank correlation coefficient, are we meant to abandon it and use the pmcc instead?

    Does anyone know what we're meant to do
    No edexcel feed you and make the test SRCC or PMCC obvious. This would be an advantage/disadvantage type
    Offline

    7
    ReputationRep:
    anyone have a link to the 2015 papers?
    Offline

    2
    ReputationRep:
    will the dof when combining cells always still be 1, similar to a dice example where the last value is not independent?
    Offline

    18
    ReputationRep:
    (Original post by Zacken)
    Yep.
    I thought the assumtion was that we assume S_a=omega_a etc. Ive onky done a few and this is always what they say for somereason.


    Posted from TSR Mobile
    Offline

    22
    ReputationRep:
    (Original post by physicsmaths)
    I thought the assumtion was that we assume S_a=omega_a etc. Ive onky done a few and this is always what they say for somereason.


    Posted from TSR Mobile
    Yeah, that is the assumption. Although it's called sigma not omega.
    Offline

    18
    ReputationRep:
    (Original post by Zacken)
    Yeah, that is the assumption. Although it's called sigma not omega.
    Oh yeh, too much mechanics starting to get confused lol
    Offline

    3
    ReputationRep:
    (Original post by Zacken)
    Yep.
    Thanks
    Offline

    6
    ReputationRep:
    (Original post by 1asdfghjkl1)
    anyone have a link to the 2015 papers?
    http://qualifications.pearson.com/co...e_20150520.pdf

    The MS is on the edexcel website
    Offline

    2
    ReputationRep:
    Anyone seen or done any past paper questions with continuity corrections that they can point me towards? Or even book exercise questions for that matter. Thanks.
    • Very Important Poster
    • Thread Starter
    Offline

    21
    ReputationRep:
    Very Important Poster
    (Original post by paradoxequation)
    Anyone seen or done any past paper questions with continuity corrections that they can point me towards? Or even book exercise questions for that matter. Thanks.
    S2 probably has more of them, when you have questions asking you to approximate discrete distributions (i.e poisson/binomial) by the normal distribution. You just observe what the number is and the inequality, and make the number more appropriate (i.e if you're looing for ... <5 then the greatest value you're looking for is less than or equal to 4.5.

    (Original post by Nikhilm)
    Any point in researching into the actual concepts behind chi-squared distribution?
    I don't.. think so
    Offline

    3
    ReputationRep:
    [QUOTE=SeanFM;65024425]S2 probably has more of them, when you have questions asking you to approximate discrete distributions (i.e poisson/binomial) by the normal distribution. You just observe what the number is and the inequality, and make the number more appropriate (i.e if you're looing for ... <5 then the greatest value you're looking for is less than or equal to 4.5.

    5.5
    Offline

    2
    ReputationRep:
    How do you solve the questions such as the part b one in the attached solution? I get the LHS, but I'm always a bit unclear on the steps from how they go from < 0.01 to < -2.3632. I understand they are using the percentage points but how exactly do you get there? Thanks.
    Offline

    2
    ReputationRep:
    Sorry, forgot to attach...
    Attached Images
     
    • Very Important Poster
    • Thread Starter
    Offline

    21
    ReputationRep:
    Very Important Poster
    (Original post by Nikhilm)

    5.5
    If you're looking for less than 5 in a discrete distribution, then the most it can be in a continuous distribution is 4.5. If it is less than or equal to 5, then the most it can be is 5.5.
    Offline

    3
    ReputationRep:
    http://www.solutionbanks.com/solutio...l%20S3/CH5.pdf exercise 5C Q7

    Surely they've done this wrong. You're suppose to put A, B, C .... at the top and then rank correspondingly?
 
 
 
  • 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
    Would you like to hibernate through the winter months?
  • 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

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