6. A biased coin and a fair coin are thrown repeatedly.
The probability that the biased coin shows heads is \(\dfrac{1}{3}\)
Let \(B\) represent the number of throws of the biased coin until it shows heads for the first time. Let \(F\) represent the number of throws of the fair coin until it shows heads for the first time.
(a) Find
(i) \(\mathrm{P}(B = 3)\) (2)
(ii) \(\mathrm{P}(2 \leqslant F \leqslant 8)\) (3)
Each day Chris invites his friend Shivani to play a game with these coins. One player’s score will be \(X\) and the other player’s score will be \(Y\), where
\[X = 4F \quad \text{and} \quad Y = \frac{B^2}{2}\]
Chris and Shivani will play a large number of games and the winner will be the player with the higher total score.
Before their first game, Chris allows Shivani to choose whether her score for every game will be \(X\) or her score for every game will be \(Y\)
(b) State, explaining your reasoning and showing any calculations you have made, which choice Shivani should make. (5)
Shivani should choose \(X\) as it has a higher expected score
A1ft
3.2a
(5)
(13 marks)
Notes
M1: For attempting to calculate or consider expected values/means of \(X\) and \(Y\) Implied by sight of 8 and 4.5
B1: for \(\mathrm{E}(X) = 8\)
M1: Using \(\mathrm{E}(B^2) = \mathrm{Var}(B) + [\mathrm{E}(B)]^2\) or \(\mathrm{E}(B^2) = \mathrm{G}'_B(1) + \mathrm{G}''_B(1)\) with a correct PGF given for \(B\) \(\mathrm{E}(B^2) = 15\) is M1
A1: for \(\mathrm{E}(Y) = 7.5\) or equivalent
A1ft: for interpreting the outcome of their calculations in terms of the problem. Must clearly choose \(X\) or \(Y\) and refer to its expectation/average being higher Allow reference to expectation as a comparison e.g. \(8 \gt 7.5\) ft their \(\mathrm{E}(X)\) and \(\mathrm{E}(Y)\) provided at least one of \(\mathrm{E}(X)\) or \(\mathrm{E}(Y)\) correct dependent on 1st M1 being awarded e.g. candidates can score M1B1M0A0A1 or M1B0M1A1A1 or M1B1M1A0A1 NB: Accept expectations given as totals. E.g. \(8n\) and \(7.5n\), but not \(7.5n^2\) etc
M1: Realising \(T = 7\), 11 (and 14) are needed If extra incorrect totals are stated, then M0
M1: Attempting \(\mathrm{P}(T = 7)\) and \(\mathrm{P}(T = 11)\) (at least one correct or both with missing ×2) May be embedded in the calculation for \(\mathrm{E}(T)\), eg \(7 \times 2 \times 0.6 \times 0.3 + 11 \times 2 \times 0.6 \times 0.1 + \ldots\)
M1: Attempting \(\mathrm{E}(T)\) for their values with at least 2 non-zero products for two of the \(T\) values correct or correct ft eg \(7 \times 0.18 + 7 \times 0.18\) only counts as 1 product Must be for their totals, simply calculating \(\mathrm{E}(X) = 2 \times 0.6 + 5 \times 0.3 + 9 \times 0.1\) here is M0
A1: 4.68 oe Correct answer with no obvious incorrect working scores 4 out of 4
Mark scheme (d)
Scheme
Marks
AO
\(Y \sim \mathrm{B}(150, 0.06)\)
M1
3.3
\(\approx \mathrm{Po}(9)\)
M1
1.1b
\(\mathrm{P}(Y = 4) \approx 0.0337\)
A1
3.4
(3)
(12 marks)
Notes
M1: Selecting the correct binomial model (may be implied by sight of \(\mathrm{Po}(9)\))
M1: Writing or using \(\mathrm{Po}(9)\) allow ft \(\mathrm{Po}(np)\) from their stated binomial distribution
7. The probability of winning a prize when playing a single game of Pento is \(\dfrac{1}{5}\)
When more than one game is played the games are independent. Sam plays 20 games.
(a) Find the probability that Sam wins 4 or more prizes. (2)
Tessa plays a series of games.
(b) Find the probability that Tessa wins her 4th prize on her 20th game. (2)
Rama invites Sam and Tessa to play some new games of Pento. They must pay Rama £1 for each game they play but Rama will pay them £2 for the first time they win a prize, £4 for the second time and £\((2w)\) when they win their \(w\)th prize \((w \gt 2)\)
Sam decides to play \(n\) games of Pento with Rama.
(c) Show that Sam’s expected profit is £\(\dfrac{1}{25}\left(n^2 - 16n\right)\) (6)
Given that Sam chose \(n = 15\)
(d) find the probability that Sam does not make a loss. (4)
Tessa agrees to play Pento with Rama. She will play games until she wins \(r\) prizes and then she will stop.
(e) Find, in terms of \(r\), Tessa’s expected profit. (4)
Mark scheme (a)
Scheme
Marks
AO
[\(X\) = no. of prizes Sam wins] \(\quad X \sim \mathrm{B}(20, 0.2)\)
\(\mathrm{E}(S) = 0.2n\) and \(\mathrm{Var}(S) = 0.2 \times 0.8n\)
M1
3.1b
\(\mathrm{E}(S^2) = 0.16n + 0.04n^2\) or \(\dfrac{1}{25}(n^2 + 4n)\)
A1
1.1b
(*) So expected profit for Sam is \(\dfrac{1}{25}(n^2 + 4n) + \dfrac{1}{5}n - n = \dfrac{1}{25}\left(n^2 - 16n\right)\)
A1cso
3.2a
(6)
Notes
1st M1 Correct start to problem - sight or use of \(\mathrm{B}(n, 0.2)\). May be implied by \(\mathrm{E}(S) = 0.2n\)
2nd M1 use of AP formula with \(a = d = 2\), or 2×AP formula with \(a = d = 1\), or equivalent NB: must be working in another variable, AP formulae cannot be in terms of \(n\)
1st A1 \(S^2 + S - n\) or equivalent, must be in a form from which expectation can be found
3rd M1 for use of \(\mathrm{E}(S) = 0.2n\) and \(\mathrm{Var}(S) = 0.16n\) (must be labelled or used as variance) (corrected from the printed mark scheme, which has \(\mathrm{Var}(S^2) = 0.16n\))
2nd A1 for correct value for \(\mathrm{E}(S^2)\)
(*) 3rd A1 for a correct solution only, pulling together everything to get given answer
Mark scheme (d)
Scheme
Marks
AO
Using profit expression: Require \(\mathrm{P}(S^2 + S - n \geqslant 0)\) Using a listing method: Indicates that 4 wins is first non-loss
M1
3.1b
Profit expression: Solving quadratic, leading to \(S = \ldots\) Listing: \(\mathrm{P}(S \geqslant 4)\)
M1
2.1
\(\mathrm{P}(S \geqslant 4)\) where \(S \sim \mathrm{B}(15, 0.2)\)
M1
1.1b
\(= 0.35183\ldots\) awrt 0.352
A1
1.1b
(4)
Notes
1st M1 for using a suitable prob statement or using a listing method to indicate first non-loss \(S\)
2nd M1 for solving the inequality or using a listing method to reach \(\mathrm{P}(S \geqslant 4)\) NB: award first two M marks for \(\mathrm{P}(S \geqslant 4)\) provided it does not come from incorrect working
3rd M1 for attempting \(\mathrm{P}(S \geqslant 4)\) with \(\mathrm{B}(15, 0.2)\)
A1 for awrt 0.352 NB: solutions stemming from finding values of \(n\) gain no marks
Mark scheme (e)
Scheme
Marks
AO
\(T\) = game on which Tessa wins her \(r\)th prize \(\quad T \sim \mathrm{negB}(r, 0.2)\) or \(r(r + 1)\)
[Use of sum of probs = 1 implies \(2a + b = 0.45\)] \(b = \underline{\mathbf{0.09}}\)
A1ft
1.1b
(4)
Notes
1st M1 for a correct attempt at \(\mathrm{E}(X^2)\) (at least 3 correct non-zero products and addition) Missing brackets around –2 and –1 is M0 unless recovered
2nd dM1 (dep on 1st M1) for use of 3.9 = their \(\mathrm{E}(X^2) - [\mathrm{E}(X)]^2\) ft their \(\mathrm{E}(X) = 0.4\)
1st A1 for \(a = 0.18\) o.e.
2nd A1 (dep on 1st M1 only) for \(b = 0.09\) o.e. or their \(b = 0.45 - 2 \times \text{“}a\text{”}\) (provided both \(a\) and \(b\) are probabilities)
1st M1 for identifying at least 2 cases e.g. \(X_1 = 3, X_2 \geqslant 1\) counts as 2 cases (ignore extras including any incorrect pairs identified) implied by at least two correct products of probs. or correct ft products of probs.
2nd M1 for a correct numerical expression for the probability ft their “0.18”
1st M1 for attempting \(\mathrm{E}(X^2)\) – at least 3 correct non-zero terms
2nd M1 Use of \(\mathrm{Var}(X) = \mathrm{E}(X^2) - [\mathrm{E}(X)]^2\) to form an expression in \(k\) with \(\sqrt{\mathrm{Var}(X)} = \mathrm{E}(X)\) ft their \(10k\) and their \(4k\) Must use a consistent expression for \(\mathrm{E}(X)\) in both side of their equation
1st A1 for a correct equation for \(k\) (e.g. 2TQ) or \(k = \dfrac{5}{16}\) or 0.3125. They may ignore or reject solution of \(k = 0\) Allow for \(r = 0.349\) or better
3rd M1 for attempt at an equation in \(r\) and \(k\) using sum of probs. At least 4 terms correct in terms of \(k\) or numerically using their value of \(k\)
2nd A1 for \(\dfrac{67}{192}\) or exact equivalents e.g. \(0.348958\dot{3}\) Correct exact answer implies full marks
M1: Realising all the different combinations 7 or more can be scored from 2 games. (no need for arrangements) Implied by \((0.4^2)\) and \((0.4 \times 0.3)\) and \((0.4 \times 0.25)\)
M1: Fully correct method.
M1: For multiplying "their (c)" with "their \(\mathrm{P}(x_5 + x_6 \geqslant 7)\)" providing at least 2 combinations are used to find \(\mathrm{P}(x_5 + x_6 \geqslant 7)\)
dM1: Dependent on 3rd M1 being awarded for using or writing \(\mathrm{B}(3, \text{“their } \mathrm{P}(x_1 + x_2 + x_3 + x_4 + x_5 + x_6 \geqslant 27)\text{”})\) \((1 - \text{“}0.0384\text{”})^3\) or
M1: For writing or using \(1 - \mathrm{P}(Y = 0)\) eg \(1 - (1 - \text{“}0.0384\text{”})^3\)
A1cso: awrt 0.111 from correct working
NB (d) 1st 3 marks (corrected from the printed mark scheme: this note is headed “NB (b)” there, but it is about the first three marks of part (d))
1st M1 At least 4 values correct for (\(X^2\) and \(2 - 3X\)) or for (\(X^2 - 2\) and \(-3X\)) or \(X^2 + 3X\) or \(X^2 + 3X - 2\) (o.e.) Allow for solving equation with one sign error
1st A1ft All correct or correct ft with their \(b\) but must have \(b \gt 5\) (accurate to 1 sf) Allow solving equation to get awrt \(-3.6\) and awrt 0.56 or \(\frac{-3 \pm \sqrt{17}}{2}\) (ft their \(b \gt 5\)) If there are omissions but no errors in the lists of values then if 2nd M1 and 2nd A1 are scored then the 1st M1 and 1st A1 can be given by implication.
2nd M1 For identifying the correct values of \(X\) required i.e. \(X = -1\) and \(X = 0\)
2nd A1 0.35 NB It is possible to score M0A0M1A1 here if their table of values is incorrect Correct answer with no incorrect working seen scores 4/4 (Allow correct use of their \(b \gt 5\))
4. Members of a photographic group may enter a maximum of 5 photographs into a members only competition. Past experience has shown that the number of photographs, \(N\), entered by a member follows the probability distribution shown below.
\(n\)
0
1
2
3
4
5
\(\mathrm{P}(N = n)\)
\(a\)
0.2
0.05
0.25
\(b\)
\(c\)
Given that \(\mathrm{E}(4N + 2) = 14.8\) and \(\mathrm{P}(N = 5 \mid N \gt 2) = \dfrac{1}{2}\)
(a) show that \(\mathrm{Var}(N) = 2.76\) (6)
The group decided to charge a 50p entry fee for the first photograph entered and then 20p for each extra photograph entered into the competition up to a maximum of £1 per person. Thus a member who enters 3 photographs pays 90p and a member who enters 4 or 5 photographs just pays £1
Assuming that the probability distribution for the number of photographs entered by a member is unchanged,
(b) calculate the expected entry fee per member. (3)
Bai suggests that, as the mean and variance are close, a Poisson distribution could be used to model the number of photographs entered by a member next year.
(c) State a limitation of the Poisson distribution in this case. (1)
Mark scheme (a)
Scheme
Marks
AO
\(4\mathrm{E}(N) + 2 = 14.8\) or \(\mathrm{E}(N) = 3.2\)
M1
3.1a
\(0.2 + 0.1 + 0.75 + 4b + 5c = 3.2\)
M1
1.1b
\(\dfrac{c}{0.25 + b + c} = 0.5\) or \(0.25 = c - b\)
M1: Attempt at both required equations with at least one term in \(k\) and one term in \(m\) correct
A1: Correct equation using \(\Sigma p = 1\)
A1: Correct equation using \(\Sigma px = 3.8\)
dM1: (dep on 1st M1) Solving simultaneously (may be implied by one correct value found)
A1: both values correct (may be implied by correct answer)
M1: Attempt to find \(\mathrm{E}(X^2)\) using their value of \(k\) and their value of \(m\) with at least 3 correct products or correct ft products Note: \(\mathrm{E}(X^2) = 6k + 7.5m\)
7. A spinner can land on red or blue. When the spinner is spun, there is a probability of \(\dfrac{1}{3}\) that it lands on blue. The spinner is spun repeatedly.
The random variable \(B\) represents the number of the spin when the spinner first lands on blue.
(a) Find
(i) \(\mathrm{P}(B = 4)\)
(ii) \(\mathrm{P}(B \leqslant 5)\)
(4)
(b) Find \(\mathrm{E}(B^2)\) (3)
Steve invites Tamara to play a game with this spinner.
Tamara must choose a colour, either red or blue.
Steve will spin the spinner repeatedly until the spinner first lands on the colour Tamara has chosen. The random variable \(X\) represents the number of the spin when this occurs.
If Tamara chooses red, her score is \(\mathrm{e}^X\)
If Tamara chooses blue, her score is \(X^2\)
(c) State, giving your reasons and showing any calculations you have made, which colour you would recommend that Tamara chooses. (5)
\(= \dfrac{2\mathrm{e}}{3} \times \dfrac{1}{1 - \frac{\mathrm{e}}{3}}\) or \(\dfrac{2\mathrm{e}}{3 - \mathrm{e}}\)
A1
1.1b
\(\mathrm{E}(\mathrm{e}^X) = 19.297\ldots\) {\(\gt 15 = \mathrm{E}(B^2)\)} so Tamara should choose red since it has the greater expected score
A1
2.2a
(5)
(12 marks)
Notes
1st M1 for choosing a suitable geometric model (sight of \(\mathrm{Geo}\left(\frac{2}{3}\right)\) or at least 3 correct probabilities)
2nd M1 for realising the need for appropriate expected value and using \(\mathrm{E}(\mathrm{g}(X))\) [Need sum and \(\mathrm{f}(x)\)] NB simply finding \(\mathrm{e}^{\mathrm{E}(X)} = \mathrm{e}^{1.5} =\) awrt 4.48 is M0 and probably no more marks.
3rd M1 for a suitable strategy to turn the expression into a sum that can be found
1st A1 for correct use of sum to infinity of geometric series
2nd A1 for interpreting the outcome of the calculations in terms of a solution to the problem must choose red and see the awrt 19.3 (and allow ft of their \(\mathrm{E}(B^2) \lt 19\))
Sum of probabilities \(= 1\) gives: \(2q + r = \tfrac{16}{30}\) (o.e.)
M1
1.1b
Solve: \(24r - 26q = \tfrac{7}{15}\) and \(r + 2q = \tfrac{8}{15}\) e.g. \(37r = \tfrac{111}{15}\)
dM1
1.1b
So \(r = \tfrac{1}{5}\) and \(q = \tfrac{1}{6}\)
A1
1.1b
(7)
Notes
1st M1 for realising the need to find \(\mathrm{E}(X)\) – a correct attempt with at least 3 correct terms
1st A1 for the correct expression (needn’t be simplified at this stage)
2nd M1 for a correct attempt at \(\mathrm{E}(X^3)\) with at least 3 correct terms seen Treat no \(\tfrac{7}{30}\) terms as one correct term
2nd A1 for \(64r - 19q\) (must be simplified) or for \(24r - 26q = \tfrac{7}{15}\)
3rd M1 for using sum of probabilities = 1 to form an equation in \(q\) and \(r\) (needn’t be simplified) Must be correct or clearly state that \(\Sigma\)probs = 1 being attempted with only one slip
4th dM1 for solving their 2 linear equations in \(q\) and \(r\) (dep on 3rd M1 and 1stor 2nd M1) Must see correct method to reduce to a linear equation in one variable
3rd A1 for \(r = \tfrac{1}{5}\) and \(q = \tfrac{1}{6}\) or any exact equivalents (dep on 2 correct equations seen)
1st M1 for 1st stage towards solving the inequality (factorising the cubic)
1st A1 for solving the inequality
2nd M1 for identifying the values of \(X\) required i.e. \(-1\) and 4
2nd A1ft for \(\tfrac{13}{30}\) or exact equivalent e.g. \(0.4\dot{3}\) (Allow ft of “their \(r\)” + \(\tfrac{7}{30}\)) (corrected from the printed mark scheme: this note prints \(\tfrac{13}{20}\), but the answer in the scheme is \(\tfrac{13}{30} = 0.4\dot{3}\))
Alternative
\(X\)
\(-3\)
\(-1\)
1
2
4
\(X^3\)
\(-27\)
\(-1\)
1
8
64
\(X^2 + 6X\)
\(-9\)
\(-5\)
7
16
40
Table 1st M1 for at least 4 correct values for \(X^3\) and \(X^2 + 6X\) (must be labelled)
1st A1 for all 10 correct values. [NB Can score M1A0M1A1ft in (d)]
2nd M1 for use of CLT – must use \(\overline{X}\) and normal or sight of \(\mathrm{N}\left(\text{“}3\text{”}, \sqrt{\dfrac{\text{“}2.6\text{”}}{80}}^{\,2}\right)\) with any letter
2nd A1ft for a correct mean and variance, ft their 3 and their 2.6 This M1A1ft may be implied by sight of correct st. dev. used in a standardisation leading to \(\mathrm{P}(Z \gt 1.39)\) Must see correct use of \(Z\) NB \(\dfrac{2.6}{80} = 0.0325\) and \(\sqrt{\dfrac{2.6}{80}} = 0.18027\ldots\) so allow e.g. \(\mathrm{N}(3, \text{awrt } (0.180)^2)\)
3rd A1 for using the normal model to find probability awrt 0.0828
ALT Use of \(\sum X\) (If see clear attempt at \(\mathrm{P}(\Sigma X \gt 260)\) condone \(\mathrm{P}(\Sigma X \gt 260.5)\)) then:
2nd M1 for \(\Sigma X \sim \mathrm{N}(\ldots)\) or any letter \(\sim \mathrm{N}\left(\text{“}240\text{”}, \sqrt{\text{“}2.6\text{”} \times 80}^{\,2}\right)\)
2nd A1ft for mean \(= \text{“}3\text{”} \times 80 = 240\) and variance \(= \text{“}2.6\text{”} \times 80 = 208\)
May see \(\mathrm{P}(\Sigma X \gt 260.5) = 0.077597\ldots\) but it will only score 2nd M1 2nd A1ft and 3rd A0
3. A fair six-sided black die has faces numbered 1, 2, 2, 3, 3 and 4
The random variable \(B\) represents the score when the black die is rolled.
(a) Write down the value of \(\mathrm{E}(B)\) (1)
A white die has 6 faces numbered 1, 1, 2, 4, 5 and \(c\) where \(c \gt 5\) The discrete random variable \(W\) represents the score when the white die is rolled and has probability distribution given by
\(w\)
1
2
4
5
\(c\)
\(\mathrm{P}(W = w)\)
\(a + b\)
\(a\)
0.3
\(a\)
\(b\)
Greg and Nilaya play a game with these dice.
Greg throws the black die and Nilaya throws the white die. Greg wins the game if he scores at least two more than Nilaya, otherwise Greg loses.
The probability of Greg winning the game is \(\dfrac{1}{6}\)
(b) Find the value of \(a\) and the value of \(b\) Show your working clearly. (5)
The random variable \(X = 2W - 5\)
Given that \(\mathrm{E}(X) = 2.6\)
(c) find the exact value of \(\mathrm{Var}(X)\) (6)
Mark scheme (a)
Scheme
Marks
AO
2.5
B1
1.1b
(1)
Notes
B1: \(\tfrac{5}{2}\) or 2.5
Mark scheme (b)
Scheme
Marks
AO
A complete strategy to find a value for \(a\) and a value for \(b\).
dM1
3.1b
Greg
Nilaya
4
1
4
2
3
1
M1
2.1
\(\tfrac{1}{6}(2a + b) + \tfrac{1}{3}(a + b) = \tfrac{1}{6}\) oe \(4a + 3b = 1\) oe
M1
1.1b
\(3a + 2b = 0.7\) oe
M1
1.1b
\(a = 0.1 \qquad b = 0.2\)
A1
1.1b
(5)
Notes
dM1: Dependent on 3rd and 4th Method marks being awarded. For a complete strategy to find a value of \(a\) and a value of \(b\). Need 2 independent equations in \(a\) and \(b\), one equation must be prob = 1/6 and the other \(3a + 2b = 0.7\) oe and an attempt to solve. For an attempt we require a method to eliminate one variable leading to a value for \(a\) and \(b\), or correct values.
M1: For using the given contextual information to list 3 different combinations for Greg to win. Implied by \(4a + 3b = 1\) oe
M1: For using \(\mathrm{P}(g) \times \mathrm{P}(n)\) for each combination identified as a win for Greg \(= \tfrac{1}{6}\) It must be a linear equation in \(a\) and \(b\) with 2/3 terms on the LHS, at least one of which must be correct and equal to 1/6
\(\mathrm{E}(W^2) = 30a + c^2 b + b + 4.8\) or \(\text{“}0.3\text{”} + 4 \times \text{“}0.1\text{”} + 16 \times 0.3 + 25 \times \text{“}0.1\text{”} + \text{“}8\text{”}^2 \times 0.2\)
M1
1.1b
\(\mathrm{E}(W^2) = 20.8\)
\(\mathrm{Var}(W) = \text{“}20.8\text{”} - \text{“}3.8\text{”}^2\) or 6.36
M1
1.1a
\(\mathrm{Var}(X) = 25.44\)
A1ft
1.2
(6)
(12 marks)
Notes
M1: For translating the given mathematical context into an expression for \(\mathrm{E}(W)\) May be implied by a correct equation or \(\mathrm{E}(W) = 3.8\)
M1: For use of \(\sum w\mathrm{P}(W = w)\ [= 3.8]\) If algebraic then at least 2 terms must be correct, if numerical at least 3 terms correct ft their values of \(a\) and \(b\). This must be seen in part (c) [NB: \(16a + 2b + 2cb + 2.4 - 5 = 2.6\) oe would get M1M1]
A1: cao
M1: For use of \(\sum w^2\mathrm{P}(W = w)\). If algebraic then at least 2 terms must be correct, if numerical at least 3 terms correct ft their values of \(a\) and \(b\).
M1: For use of \(\mathrm{Var}(W) = \mathrm{E}(W^2) - [\mathrm{E}(W)]^2\)
A1ft: \(4 \times \text{“their Var}(W)\text{”}\) ft their \(\mathrm{Var}(W)\) provided \(a\), \(b\) and \(\mathrm{Var}(W)\) are > 0 and \(c \gt 5\)
Alternative for (c). allow a mix of methods
Scheme
Marks
[\(\mathrm{E}(X) =\)] \(-3 \times (a + b) - 1 \times a + 0.9 + 5 \times a + (2c - 5) \times b\)
M1
\(-3 \times (a + b) - 1 \times a + 0.9 + 5 \times a + (2c - 5) \times b = 2.6\)
M1
\(c = 8\)
A1
Values of \(X\) \(-3, -1, 3, 5, 2c - 5\)
M1
\(\mathrm{E}(X^2) = 9 \times (a + b) + 1 \times a + 2.7 + 25 \times a + (2c - 5)^2 \times b\) \(= 32.2\)
6. The independent random variables \(X_1\) and \(X_2\) are each distributed \(\mathrm{B}(n, p)\), where \(n \gt 1\) An unbiased estimator for \(p\) is given by
\[\hat{p} = \frac{aX_1 + bX_2}{n}\]
where \(a\) and \(b\) are constants.
[You may assume that if \(X_1\) and \(X_2\) are independent then \(\mathrm{E}(X_1X_2) = \mathrm{E}(X_1)\mathrm{E}(X_2)\)]
(a) Show that \(a + b = 1\) (2)
(b) Show that \(\mathrm{Var}(\hat{p}) = \dfrac{\left(2a^2 - 2a + 1\right)p(1 - p)}{n}\) (4)
(c) Hence, justifying your answer, determine the value of \(a\) and the value of \(b\) for which \(\hat{p}\) has minimum variance. (5)
(d)
(i) Show that \(\hat{p}^2\) is a biased estimator for \(p^2\)
(ii) Show that the bias \(\to 0\) as \(n \to \infty\) (5)
(e) By considering \(\mathrm{E}[X_1(X_1 - 1)]\) find an unbiased estimator for \(p^2\) (3)
Mark scheme (a)
Scheme
Marks
\(\mathrm{E}\left(\dfrac{aX_1 + bX_2}{n}\right) = \dfrac{anp + bnp}{n} = ap + bp = (a + b)p\)
M1
\(a + b = 1\ *\)
A1* cso
(2)
Notes
M1 Using \(\dfrac{a\mathrm{E}(X_1) + b\mathrm{E}(X_2)}{n}\) and subst \(\mathrm{E}(X_1) = np\) and \(\mathrm{E}(X_2) = np\)
Acso* Answer given. Need \(p(a + b) = p\) and statement \(a + b = 1\) and no errors
M1 Using \(\dfrac{a^2\mathrm{Var}(X_1) + b^2\mathrm{Var}(X_2)}{n^2}\) and subst \(\mathrm{Var}(X_1) = np(1 - p)\) – may be implied by \(\dfrac{1}{n^2}\left(a^2np(1 - p) + b^2np(1 - p)\right)\)
A1 correct answer in any form
M1d dep on 1st M1 Subst \(b = 1 - a\)
A1cso* method must be shown and no errors.
Mark scheme (c)
Scheme
Marks
Min value when \(\dfrac{(4a - 2)p(1 - p)}{n} = 0\) \(\Rightarrow 4a - 2 = 0\)
M1A1
\(a = \dfrac{1}{2},\ \ b = \dfrac{1}{2}\)
A1A1ft
\(\dfrac{\mathrm{d}^2\mathrm{Var}(\hat{p})}{\mathrm{d}a^2} = \dfrac{4p(1 - p)}{n} \gt 0\) or \(\because\) quadratic with positive \(x^2\) \(\therefore\) minimum point or sketch
B1
(5)
Notes
M1 \(\dfrac{\mathrm{d}}{\mathrm{d}a}(\mathrm{Var})\) (must differentiate with respect \(a\)) or attempt to complete the square
6. When a tree seed is planted the probability of it germinating is \(p\). A random sample of size \(n\) is taken and the number of tree seeds, \(X\), which germinate is recorded.
(a)
(i) Show that \(\hat{p}_1 = \dfrac{X}{n}\) is an unbiased estimator of \(p\).
(ii) Find the variance of \(\hat{p}_1\). (4)
A second sample of size \(m\) is taken and the number of tree seeds, \(Y\), which germinate is recorded.
Given that \(\hat{p}_2 = \dfrac{Y}{m}\) and that \(\hat{p}_3 = a(3\hat{p}_1 + 2\hat{p}_2)\) is an unbiased estimator of \(p\),
(b) show that
(i) \(a = \dfrac{1}{5}\),
(ii) \(\mathrm{Var}(\hat{p}_3) = \dfrac{p(1-p)}{25}\left(\dfrac{9}{n} + \dfrac{4}{m}\right)\). (6)
(c) Find the range of values of \(\dfrac{n}{m}\) for which \[\mathrm{Var}(\hat{p}_3) \lt \mathrm{Var}(\hat{p}_1) \text{ and } \mathrm{Var}(\hat{p}_3) \lt \mathrm{Var}(\hat{p}_2)\] (3)
(d) Given that \(n = 20\) and \(m = 60\), explain which of \(\hat{p}_1\), \(\hat{p}_2\) or \(\hat{p}_3\) is the best estimator. (3)
(i) M1 For either \(3a\,\mathrm{E}(\hat{p}_1) + 2a\,\mathrm{E}(\hat{p}_2)\) or \(3ap + 2ap\) M1 Putting their \(\mathrm{E}(\hat{p}_3) = p\)
(ii) M1 for \(\dfrac{9}{25}\mathrm{Var}(\hat{p}_1) + \dfrac{4}{25}\mathrm{Var}(\hat{p}_2)\) M1d for substituting (aii) for \(\mathrm{Var}(\hat{p}_1)\) and (aii) with \(m\) instead of \(n\) for \(\mathrm{Var}(\hat{p}_2)\) A1 cso
M1 Putting \(\mathrm{Var}(\hat{p}_3) \lt\) their \(\mathrm{Var}(\hat{p}_1)\) leading to an inequality of the form \(\dfrac{n}{m} \lt a\) or \(\dfrac{n}{m} \gt a\) where a is a constant. M1 Putting \(\mathrm{Var}(\hat{p}_3) \lt\) their \(\mathrm{Var}(\hat{p}_2)\) leading to an inequality of the form \(\dfrac{n}{m} \gt a\) or \(\dfrac{n}{m} \lt a\) where a is a constant.
Or since \(\dfrac{1}{3}\) is not in the range \(\dfrac{9}{21} \lt \dfrac{n}{m} \lt 4\) \(\mathrm{Var}(\hat{p}_3)\) is not the smallest variance. \(\mathrm{Var}(\hat{p}_1) = 0.05\,p(1-p)\) \(\mathrm{Var}(\hat{p}_2) = 0.0167\,p(1-p)\)
M1
Therefore \(\hat{p}_2\); is the best estimator as it has the smallest variance
A1ft; A1ft
(3)
(16 marks)
Notes
1/3 is not in their range in part(c) M1 attempt to find all 3 variances or eliminating \(\mathrm{Var}(\hat{p}_3)\) with reason and finding the other 2 variances. A1ft correct estimator chosen. A1ft correct supporting reason from correct working for their var formulae
SC if 1/3 is in their range in part(c) they may get B1 for stating \(\hat{p}_3\) B1dependent on the previous B being awarded- stating smallest variance award first two marks on epen.