Scribd Logo
Upload

Welcome to Scribd!

  • 97 views

    Edexcel S2 - June 2001 - Solutions

    Uploaded by

    shangelivia
    This document contains solutions to exam questions involving probability distributions and statistical tests. Key points: 1) Traffic flow data can be sampled from a list of dates/times, and is likely modeled by the Poisson distribution. 2) Examples are given of calculating probabilities for various binomial and Poisson distributions. 3) A binomial test is used to test if the proportion of gold beads has changed, and the null hypothesis is not rejected. 4) Examples use the binomial, normal, and cumulative binomial distributions to calculate probabilities. 5) More examples use the Poisson distribution to calculate probabilities for lightbulb usage. A one-tailed Poisson test is used to test if a new measure has reduced lightbulb damage.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Edexcel S2 - June 2001 - Solutions

    Uploaded by

    shangelivia
    97 views3 pages
    This document contains solutions to exam questions involving probability distributions and statistical tests. Key points: 1) Traffic flow data can be sampled from a list of dates/times, and is likely modeled by the Poisson distribution. 2) Examples are given of calculating probabilities for various binomial and Poisson distributions. 3) A binomial test is used to test if the proportion of gold beads has changed, and the null hypothesis is not rejected. 4) Examples use the binomial, normal, and cumulative binomial distributions to calculate probabilities. 5) More examples use the Poisson distribution to calculate probabilities for lightbulb usage. A one-tailed Poisson test is used to test if a new measure has reduced lightbulb damage.

    Original Title

    S2 2001-06 MS

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document contains solutions to exam questions involving probability distributions and statistical tests. Key points: 1) Traffic flow data can be sampled from a list of dates/times, and is likely modeled by the Poisson distribution. 2) Examples are given of calculating probabilities for various binomial and Poisson distributions. 3) A binomial test is used to test if the proportion of gold beads has changed, and the null hypothesis is not rejected. 4) Examples use the binomial, normal, and cumulative binomial distributions to calculate probabilities. 5) More examples use the Poisson distribution to calculate probabilities for lightbulb usage. A one-tailed Poisson test is used to test if a new measure has reduced lightbulb damage.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    97 views3 pages

    Edexcel S2 - June 2001 - Solutions

    Uploaded by

    shangelivia
    This document contains solutions to exam questions involving probability distributions and statistical tests. Key points: 1) Traffic flow data can be sampled from a list of dates/times, and is likely modeled by the Poisson distribution. 2) Examples are given of calculating probabilities for various binomial and Poisson distributions. 3) A binomial test is used to test if the proportion of gold beads has changed, and the null hypothesis is not rejected. 4) Examples use the binomial, normal, and cumulative binomial distributions to calculate probabilities. 5) More examples use the Poisson distribution to calculate probabilities for lightbulb usage. A one-tailed Poisson test is used to test if a new measure has reduced lightbulb damage.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 3

    Edexcel S2 – June 2001 – Solutions

    1ai) As it is a small village, a census can be undertaken. The sampling frame


    would be the electoral register, or any other suitable list.

    aii) For collecting traffic flow details, a sample survey will need to be used. The
    sampling frame would be a list of days/dates/times in which traffic could be
    counted.

    b) The statistic could be “the number of cars passing a point in ten minutes”
    and this is likely to be modelled by the Poisson Distribution.

    2a) Let X = “the number of accidents in the next month” Then X ~ Po(0.9)

    P ( X = 0) = e−0.9 0.407

    b) Let Y = “the number of accidents in the next six months”, then Y~Po(5.4)

    e −5.4 (5.4) 2
    P (Y = 2) = = 0.06659
    2!

    c) If M = “number of months without accidents”,


    then from a) above M ~ Bin(4, 0.407)

    4
    P ( M = 0) =   (0.407) 0 (0.593)
    = 2 0.350
    2

    3) H0 : p = 1 H1 : p ≠ 1
    4 4

    We require a two tailed test of binomial proportion at the 5% level.


    Assuming Ho to be correct, then X ~ Bin(20, 0.25)

    From tables, critical region defined as X ≤ 1, X ≥9

    We are told exactly 2 are gold. This is not within the critical region, and so
    we do not reject Ho. There is insufficient evidence at this level to suggest the
    proportion of gold beads has changed.

    4a) Let X = “t number of letters that are marked first class” then X ~ Bin(10, 0.2)

    P ( X ≥−3) =≤=
    1− P( X= 2) 1 0.6778 0.3222 (from Cum. Bin. Tables)

    b) P ( X < 2) =≤=P( X 1) 0.3758 (from Cum. Bin. Tables)


    c) Using the normal approximation, F ~ N(14, 11.2)

     12.5 − 14 
    P ( F ≤ 12)≤ = =P( F
    ≤ 12.5) PZ 
     11.2 
    − = P ( Z ≤ 0.4482)
    = 0.3264

    d) We assume that all letters are independent of each other.

    5a) Let X = “number of lightbulbs required in a week”, then X ~ Po (2)

    P ( X = 4) =≤−P( X≤= 4) −P( X = 3) 0.9473 0.8571 0.0902 (from tables)

    b) P ( X >−5) =≤=
    1− P( X= 4) 1 0.9834 0.0166 (from tables)

    c) Let Y = “number of lightbulbs required in 3 weeks), then Y ~ Po(6)

    P (Y ≤ 5) = 0.4457 (from tables)

    d) H 0 : λ = 8; H1 : λ < 8

    We require a one tailed test of Poisson mean at the 5% significance level.


    Assuming Ho true, X ~ Po(8)

    From tables, critical region is X ≤ 3

    We are told 3 bulbs are requested. This is within the critical region, so we
    reject Ho in favour of H1 concluding there is evidence to suggest the new
    measures are reducing damage.

    6a) pdf given by differentiation

    1
     (−3x + 12≤x )
    2
    1 x 4
    f ( x) =  27
     0 otherwise

    b) Mode is given by the maximum, ie differential = 0

    −6 x=+ 12= 0; x 2 Mode = 2

    c)

    1 4
    4 4
     1  1 3 4 2
    d) µ = ∫ x  +(−−3 x 2 +=12 x  dx ∫ x x dx
    1 
    27  1
    9 9
    4
     1 4 3
    =  − +x 4 x
     36 27 1
      256 256  1 4 
     = − + −− + 
      36 27  36 27 
    9
    =
    4

    1
    e) F(2.25) = (−=(2.25)
    − 3 + 6(2.25) 2 5) 0.517
    27

    f) F(mean) from e) above is greater than 0.5, hence greater than the median.

    1
    F(2) = (−(2)3=
    −+ 6(2) 2 5) 0.407 ie less than the median
    27

    Hence mode < median < mean

    7a) P(T < 0.2) = 0.2

    b) by symmetry E(T) = 0.5

    1 1
    1  1 1 1 1
    c) E (T 2 ) = ∫ t 2=dt =  t 3  Hence Var(T) = − =
    0  3 0 3 3 4 12

    d) Let X = “number of children stopping < 0.2” the X ~ Bin(20, 0.2)

    P ( X ≤ 4) = 0.6296 (from tables)

    e) I would expect the mean to remain at 0.5, but the variance to decrease as the
    children become more accurate.

    0.2
    0.2
    f) P (T < 0.2)== ∫ =4t dt
    0
     2t 2 
    0
    0.08

    g) Let Y = “number of players stopping < 0.2” then Y ~ Bin(75, 0.08)

    As n is large, and p very small, the Poisson approximation is used. Y ~ Po(6)

    P (Y >−7) ≤=
    =−1 P(=Y 6) 1 0.7440 0.256

    End of Paper

    You might also like