Choosing and critiquing statistical models

Choosing and critiquing statistical models Test

Spec N3 · AQA A-Level Maths: Subtopic test

Studying this course? Keep your progress in one place.

Two ways to do this test

Question by question

One question at a time. Each answer is marked the moment you hand it in, so you learn as you go. Drago helps if you get stuck. All 10 questions, or just 5.

Exam mode

The whole paper in one go, timed like the real exam. Nothing is marked until you hand it in, then you get every mark explained.

10 questions, 27 marks

AQA A-Level Maths

Choosing and critiquing statistical models

Total 27 marks

Name

Class

Date

  1. 1
    A teacher models the number XX of pupils in a class of 3030 who walk to school by the binomial distribution X∼B(30,0.4)X\sim B(30,0.4).
    (a)
    Which of the following is an assumption needed for this binomial model?
    [1 mark]
    • AEach pupil's journey time is Normally distributed
    • BThe probability of walking is different for each pupil
    • CWhether one pupil walks is independent of whether any other pupil walks
    • DThe number of pupils in the class is not fixed
    Show the mark scheme

    Mark scheme

    Answer: C
    (b)
    Pupils in this class often decide in friendship groups, so whether a pupil walks affects whether their friends walk. Which condition for the binomial model is not satisfied?
    [1 mark]
    • AThe trials are independent
    • BThe number of trials is fixed
    • CThere are only two outcomes for each pupil
    • DThe probability of success is 0.40.4 for every pupil
    Show the mark scheme

    Mark scheme

    Answer: A
    (c)
    The pupils in the class live at very different distances from the school. Explain why B(30,0.4)B(30,0.4) may not be a suitable model.
    [2 marks]
    Show the mark scheme

    Mark scheme

    B1: the probability of walking is not the same (constant) for every pupil
    R1: in context, e.g. pupils who live nearer are more likely to walk than those who live further away

    Accept: the trials do not all have the same probability of success

    Do not accept: the pupils are not independent, without reference to distance

    Total for question 1: 4 marks

  2. 2
    A researcher records the monthly household income XX pounds in a city. The sample mean is 24002400, the sample median is 19001900 and the sample standard deviation is 18001800. She models XX by the Normal distribution N(2400,18002)N(2400,1800^2).
    (a)
    Which feature of the sample data most strongly suggests that the Normal model may not be appropriate?
    [1 mark]
    • AThe mean is positive
    • BThe standard deviation is smaller than the mean
    • CIncomes are measured in pounds
    • DThe mean is greater than the median, suggesting positive skew
    Show the mark scheme

    Mark scheme

    Answer: D
    (b)
    Using the model N(2400,18002)N(2400,1800^2), the proportion of households predicted to have a negative income is closest to
    [1 mark]
    • A0.0%0.0\%
    • B9.1%9.1\%
    • C50.0%50.0\%
    • D90.9%90.9\%
    Show the mark scheme

    Mark scheme

    Answer: B
    (c)
    State, with a reason, whether the Normal model is appropriate for monthly household income.
    [2 marks]
    Show the mark scheme

    Mark scheme

    B1: not appropriate
    R1: a valid reason using the data or the model, e.g. the mean (
    24002400) is greater than the median (19001900), so the data are skewed; or the model predicts about 9%9\% negative incomes, which is impossible
    Accept: the distribution is not symmetrical

    Do not accept: a reason with no reference to this data or model

    Total for question 2: 4 marks

  3. 3
    A call centre claims that 70%70\% of calls are answered within 2020 seconds. A manager records the next 1010 calls received on a Monday morning and models the number XX answered within 2020 seconds by X∼B(10,0.7)X\sim B(10,0.7).
    (a)
    State the distribution of XX and find the probability that at least 88 of the 1010 calls are answered within 2020 seconds.
    [3 marks]
    Show the mark scheme

    Mark scheme

    B1: X∼B(10,0.7)X\sim B(10,0.7) stated (given in the question, or seen in the working)
    M1:
    P(X≥8)=1−P(X≤7)P(X\ge8)=1-P(X\le7), or P(8)+P(9)+P(10)P(8)+P(9)+P(10) with the correct parameters
    A1:
    0.3830.383 (0.3828…0.3828\ldots)
    (b)
    The manager's sample is the next 1010 calls on a Monday morning, when the call centre is unusually busy. Give two reasons, referring to this context, why the binomial model may not be appropriate.
    [4 marks]
    Show the mark scheme

    Mark scheme

    B1: the trials may not be independent
    R1: in context, e.g. when the centre is busy a queue builds up, so a slow answer to one call makes the next call more likely to be answered slowly

    B1: the probability of success may not be constant (or
    0.70.7 may not apply)
    R1: in context, e.g. Monday morning is busier than usual, so the probability of answering within
    2020 seconds is probably lower than the overall 70%70\%
    Accept: the sample is not random or not representative of all calls

    Total for question 3: 7 marks

  4. 4
    A lake contains a large population of a species of fish. A biologist measures a random sample of 200200 fish and finds that the sample mean length is 3131 cm and the sample standard deviation is 44 cm. She proposes the model L∼N(31,42)L\sim N(31,4^2) for the length LL cm of a fish.
    (a)
    Of the 200200 fish in the sample, 2727 are longer than 3838 cm.
    (i) Use the model to find the probability that a fish is longer than
    3838 cm.
    (ii) Find the number of fish longer than
    3838 cm that the model predicts in a sample of 200200.
    (iii) Compare the prediction with the observed number and comment on whether the model is suitable.
    [6 marks]
    Show the mark scheme

    Mark scheme

    M1: (i) z=38−314=1.75z=\frac{38-31}{4}=1.75
    A1: (i)
    0.04010.0401 (0.04006…0.04006\ldots)
    M1: (ii)
    200×0.0401200\times0.0401
    A1: (ii)
    8.08.0 (8.01…8.01\ldots)
    R1: (iii) the observed number (
    2727) is more than three times the predicted number (88)
    R1: (iii) the model underestimates the number of long fish, which suggests the lengths are not symmetrical (positively skewed), so the Normal model is not suitable

    Accept:
    88
    (b)
    Separately, a small tank holds 5050 fish, of which 1010 are undersized. A fisherman removes 1010 fish at random, one after another, without replacement. He models the number of undersized fish removed by B(10,0.2)B(10,0.2).
    (i) Using this model, find the probability that none of the
    1010 fish is undersized.
    (ii) Find the probability that the first two fish removed are both not undersized, using the actual numbers of fish in the tank, and compare it with the value given by the binomial model.

    (iii) Explain why the binomial model is not exact here, and state a change to the situation that would make it a good model.
    [6 marks]
    Show the mark scheme

    Mark scheme

    M1: (i) 0.8100.8^{10}
    A1: (i)
    0.1070.107 (0.1074…0.1074\ldots)
    M1: (ii)
    4050×3949\frac{40}{50}\times\frac{39}{49}
    A1: (ii)
    0.6370.637 (156245\frac{156}{245}), compared with 0.82=0.640.8^2=0.64 from the model
    R1: (iii) fish are removed without replacement, so the probability of an undersized fish changes with each removal and the trials are not independent

    R1: (iii) a much larger tank (or replacing each fish after it is checked) would make the probability almost constant

    Accept: sampling from a large population

    Total for question 4: 12 marks

End of questions

Subtopic tests