Exercise 2A

e −2.5 × 2.53
1 a P(X= 3)=
3!
= 0.213763 =  0.2138 (4 d.p.)

1 − P( X 1) =
b P(X > 1) = 1 − P( X = 0) − P( X =1)
−2.5 0 −2.5 1
e × 2.5 e × 2.5
1−
= −
0! 1!
1 − 0.08208 − 0.20521 =
= 0.7127 (4 d.p.)

c P(1 < X 3) =P( X = 2) + P( X = 3)

e × 2.5 e × 2.53
−2.5 2 −2.5
= +
2! 3!
= 0.25651 + 0.21376 = 0.4703 (4 d.p.)

e −3.1 × 3.14
2 a P(X= 4)
=
4!
= 0.173347
=  0.1733 (4 d.p.)

b P(X 2) =
1 − P( X < 2) = 1 − P( X = 0) − P( X =1)
−3.1 0 −3.1 1
e × 3.1 e × 3.1
=1− −
0! 1!
=1 − 0.045049 − 0.139652 = 0.8153 (4 d.p.)

c P(1X 4) =
P( X =+
1) P( X =+
2) P( X =+
3) P( X =
4)
 3.11 3.12 3.13 3.14 
= e −3.1  + + + 
 1! 2! 3! 4! 
= 0.045049 × (3.1 + 4.805 + 4.96516 + 3.84800
= ) 0.7531 (4 d.p.)

e −4.2 × 4.22
3 a P(X= 2)
=
2!
= 0.13226
=  0.1323 (4 d.p.)

b P(X 3) =
P( X =+
0) P( X =+
1) P( X =+
2) P( X =
3)
 4.20 4.21 4.22 4.23 
= e −4.2  + + + 
 0! 1! 2! 3! 
= 0.0149955 × (1 + 4.2 + 8.82 =
+ 12.384) 0.3954 (4 d.p.)

c P(3X 5) =
P( X =
3) + P( X =
4) + P( X =
5)
 4.23 4.24 4.25 
= e −4.2  + + 
 3! 4! 5! 
= 0.0149955 × (12.384 + 12.9654 + 10.8909=
) 0.5429 (4 d.p.)

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 1
 e −0.84 × 0.841
4 a P(X= 1)=
1!
= 0.362638 =  0.3626 (4 d.p.)

b P(X 1) = 1 − P( X < 1) = 1 − P( X =0)

−0.84 0
e × 0.84
= 1−
0!
=1 − 0.431710 = 0.5683 (4 d.p.)

c P(1 < X 3) =P( X =2) + P( X =

3)
 0.842 0.843 
−0.84
= e  + 
 2! 3! 
= 0.43171 × (0.3528 + 0.098784)= 0.1950 (4 d.p.)

λ2 λ3
5 P ( X= 2=
) e−λ and P ( X= 3=
) e−λ
2! 3!
λ 2
λ 3
If P(X = 2) = P(X = 3) then = so λ = 3
2! 3!

λ4 λ2
6 P ( X= 4=
) e −λ
and P ( X= 2=
) e −λ

4! 2!
λ 4
λ2
If P(X = 4) = 3 × P(X = 2) then so λ2 = 36 and therefore λ = 6
= 3×
4! 2!
Reject the negative root because the Poisson parameter must be positive.

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 2
Exercise 2B
Note that because you are required to use tables in this exercise, when calculating the difference between
two values (for example in question 2b), you would obtain a slightly different answer on your calculator;
this is because the table contains values already rounded to 4 d.p. and so you may introduce inaccuracy
at the fourth decimal place when combining values.

1 a Use tables with λ = 5.5

P(X 3) = 0.2017

b Use tables with λ = 5.5

P(X 6) = 1 − P(X 5) = 1 − 0.5289 =
0.4711

c Use tables with λ = 5.5

P(3= X 7) P(X 7) − P(X  2)
= 0.8095 − 0.0884 = 0.7211

2 a Use tables with λ = 10

P(X 8) = 1 − P(X 7) =1 − 0.2202 =
0.7798

b Use tables with λ = 10

P(7 =X 12) P(X 12) − P(X 6)
= 0.7916 − 0.1301 = 0.6615

c Use tables with λ = 10

P(4 < X <= 9) P(X 8) − P(X 4)
= 0.3328 − 0.0293 = 0.3035

3 a Use tables with λ = 3.5

P(X 2) = 1 − P(X 1) = 1 − 0.1359 =
0.8641

b Use tables with λ = 3.5

P(3= X 6) P(X 6) − P(X 2)
= 0.9347 − 0.3208 = 0.6139

c Use tables with λ = 3.5

P(2 < X 5)= P(X 5) − P(X 2)
= 0.8576 − 0.3208 = 0.5368

4 a Use tables with λ = 4.5

P(X 5) = 1 − P(X 4) = 1 − 0.5321 =
0.4679

b Use tables with λ = 4.5

P(3 < X 5)= P(X 5) − P(X 3)
= 0.7029 − 0.3423 = 0.3606

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 1
4 c Use tables with λ = 4.5
P(1X <= 7) P(X 6) − P(X 0)
= 0.8311 − 0.0111 = 0.8200

5 a Use tables with λ = 8

P(X a ) = 0.3134
so a = 6

b Use tables with λ = 8

P(X b) = 0.7166
so b = 9

c Use tables with λ = 8

P(X <= c) P( X c =
− 1) 0.0996
so c − 1 =4
c=5

d Use tables with λ = 8

1 P( X d ) =0.8088
P(X > d ) =−
so P( X d ) =1 − 0.8088 =
0.1912
so d = 5

6 a Use tables with λ = 3.5

P(X a ) = 0.8576
so a = 5

b Use tables with λ = 3.5

P(X > b) =1 − P( X b) =0.6792
so P( X b) = 1 − 0.6792 =
0.3208
so b = 2

c Use tables with λ = 3.5

P(X c)0.95
This is true for all values of c > 6

d Use tables with λ = 3.5

P(X > d )0.005
So 1 − P( X d )0.005
P( X d )1 − 0.005
P( X d )0.995
This is true for all values of d > 8

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 2
Poisson distributions 2C
1 a i The probability that there are exactly 4 requests for replacement light bulbs is P(X = 4).
As X  Po(3)
e −3 × 34
P( X= 4)
= = 0.1680 (4 d.p.)
4!

ii The probability that there are more than 5 requests is P(X > 5).
Find this from the tables using λ = 3
P( X > 5) =1 − P( X 5) =1 − 0.9161 =0.0839

b Let Y be the number of requests in a fortnight, so use the tables with λ = 6 .

i As Y  Po(6)
e −6 × 66
P(Y= 6)
= = 0.1606 (4 d.p.)
6!

ii Use the tables with λ = 6

P( X 4) = 0.2851

2 a Weeds must grow independently of the presence of other weeds. They must grow at a constant
average density so that the mean number in any area of the field is proportional to the area.

b Let X be the number of weeds in a random 4 m2 area of the field. In this model,
X  Po(4 ×1.3), i.e. X  Po(5.2) , so λ = 5.2 .
As the tables in the textbook do not give values of the Poisson cumulative distribution function for
λ = 5.2 and the required probability P( X2) must be found using a calculator.
P( X 2) = 0.1088

c Let Y be the number of weeds in a random 5 m2 area of the field. In this model,
X  Po(5 ×1.3), i.e. X  Po(6.5) , so λ = 6.5 and the tables can be used to find the required value.
P( X > 8) =1 − P( X 8) =1 − 0.7916 =0.2084

3 a Detection occurs at a constant mean rate of 2.5. So a suitable model is to let X be the number of
faulty components detected in a hour, with X  Po(2.5) .

b It is assumed that faulty components are found independently of each other and that the detection
is evenly spread throughout each hour (so that the mean rate of detection in k hours is 2.5k for all
positive values of k).

c As X  Po(2.5)
e −2.5 × 2.52
P( X= 2)
= = 0.2565 (4 d.p.)
2!

d Let Y be the number of faulty components detected in a 3-hour period, so Y  Po(7.5).

Use the tables with λ = 7.5
P( X 6) = 1 − P( X 5) =
1 − 0.2414 =
0.7586

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 1
3 e Let Z be the number of faulty components detected in a 4-hour period, so Z  Po(10) .
Use the tables with λ = 10
P( X 7) = 1 − P( X 6) =1 − 0.1301 =
0.8699

4 a Let X be the number of telephone calls in a 20-minute interval, so X  Po(5) .

e −5 × 54
i P( X= 4)
= = 0.1755 (4 d.p.)
4!

ii Use the tables with λ = 5

P( X > 8) =1 − P( X 8) =1 − 0.9319 =0.0681

b Let Y be the number of telephone calls in a 30-minute interval, so Y  Po(7.5) .

i Use the tables with λ = 7.5

P( X 6) = 1 − P( X 5) =
1 − 0.2414 =
0.7586

ii Use the tables with λ = 7.5

P( X 10) = 0.8622

5 a Let X be the number of cars crossing in any given minute, so X  Po(3), λ = 3.

P( X > 5) =1 − P( X 5) =1 − 0.9161 =0.0839

b Let Y be the number of cars crossing in any given 2-minute period, so Y  Po(6), λ = 6.
P( X 3) = 0.1512

6 Let X be the number of customers arriving for breakfast between 10:00 am and 10:20 am.
As 20 minutes is 5 × 4 minutes, the model is X  Po(5).

a Use the tables with λ = 5

P( X 2) = 0.1247

b Use the tables with λ = 5

P( X > 10) = 1 − P( X 10) =
1 − 0.9863 =
0.0137

7 a Let X be the number of houses the estate agent sells in a week, so X  Po(1.8). As λ = 1.8, all
answers must be found using a calculator.

e −1.8 ×1.80
i P( X= 0)
= = 0.1653 (4 d.p.)
0!

e −1.8 ×1.83
ii P( X= 3)= = 0.1607 (4 d.p.)
3!

iii P( X 3) =
1 − P( X 2) =
1 − 0.7306 =
0.2694 (4 d.p.)

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 2
7 b Let Y be the number of weeks, over a period of 4 weeks, in which the estate agent meets her target.
As the probability that the agent meets her target is P( X 3) = 0.2694 (from part aiii), the model is
Y ~ B(4, 0.2694).
 4 1 3
P(Y = 1) =   (0.2694) (1 − 0.2694) = 0.4202 (4 d.p.)
1
 

8 a Let X be the number of patients arriving during a 30-minute period, so X  Po(2.5).

e −2.5 × 2.54
i P( X= 4)
= = 0.1336 (4 d.p.)
4!

ii Use the tables with λ = 2.5

P( X 3) = 1 − P( X 2) =
1 − 0.5438 =
0.4562

b If the next patient arrives before 11:15 am then there must be at least one patient in the 15-minute
period between 11:00 am and 11:15 am.
Let Y be the number of patients arriving during a 15-minute period, so Y  Po(1.25). As λ = 1.25,
the solution must be found using a calculator.
e −1.25 ×1.250
P( X  1) =
1 − P( X ==
0) 1 − 1 − 0.2865 =
= 0.7135 (4 d.p.)
0!

9 a Let X be the number of times the elevator breaks down in one week, Y  Po(0.75). As λ = 0.75,
the solutions must be found using a calculator.

e −0.75 × 0.750
i P( X 1) =
1 − P( X ==
0) 1 − 1 − 0.4724 =
= 0.5276 (4 d.p.)
0!

e −0.75 × 0.752
ii P( X= 2)
= = 0.1329 (4 d.p.)
2!

b If the elevator breaking down can be modelled using a Poisson distribution, then each breakdown
occurs independently of any previous history. So the probability of at least one breakdown in the
next week will be P( X 1) = 0.5276 , as given in part ai.

10 a Let X be the number of flaws in a 50 m length of material, so X  Po(1.5).

e −1.5 ×1.53
P( X= 3)= = 0.1255 (4 d.p.)
3!

b Let Y be the number of flaws in a 200 m length of material, so Y  Po(6).

Use the tables with λ = 6
P( X <
= 4) P( X= 3) 0.1512

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 3
10 c Let A be the number of rolls in a random sample of 5 which have fewer than 4 flaws. As
P( X < 4) = 0.1512 (from part b), the model is A  B(5, 0.1512).
5 5
1 −   ( 0.1512 ) (1 − 0.1512 ) −   ( 0.1512 ) (1 − 0.1512 )
1 4 0 5
P( A2) =
1  0
= 0.1670

11 a Let X be the number of chocolate chips in a biscuit, so X  Po(5).

Use the tables with λ = 5
P( X =
< 3) P( X= 2) 0.1247

b Let Y be the number of biscuits in a pack of 6 which contain fewer than 3 chocolate chips. As
P( X < 3) = 0.1247 (from part a), the model is Y  B(6, 0.1247).
6 3 3
P(Y = 3) =  (0.1247) (1 − 0.1247) = 0.0260
3

12 a Let X be the number of requests for buses on a Sunday in summer, so X  Po(5).

Use the tables with λ = 5
P( X <
= 4) P( X= 3) 0.2650

b Let n be the number of buses that the company must have to be 99% sure it can fulfil all requests;
so P( X n)0.99.
λ = 5 , P( X 10) 0.9863,
From the tables with = = P( X 11) 0.9945
So the company needs 11 buses to be 99% sure it can fulfil all requests.

13 a Let X be the number of boats hired in a 30-minute period, so X  Po(4.5).

Use the tables with λ = 4.5
P( X 6) = 1 − P( X 5) =
1 − 0.7029 = 0.2971

b Let Y be the number of boats hired in a 20-minute period, so Y  Po(3).

Use the tables with λ = 3
P(Y > 8) =1 − P(Y 8) =1 − 0.9962 =0.0038
So the probability that more than 8 boats are requested is 0.38%, which is less than 1%.

c Let n be the number of boats that the company must have to be 99% sure it can meet all demands
in a 30-minute period; so P( X n)0.99.
From the tables with λ= = 4.5 , P( X 9) 0.9829,
= P( X 10) 0.9933
So the company needs 10 boats to be 99% sure it can fulfil all requests over the hire period.

14 a Let X be the number of breakdowns in a randomly chosen week, so X  Po(1.5).

Use the tables with λ = 1.5
P( X 2) = 0.8088

b Let Y be the number of breakdowns in a randomly chosen 2-week period, so Y  Po(3).

Use the tables with λ = 3
P(Y 5) = 1 − P(Y 4) =1 − 0.8153 =
0.1847

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 4
14 c Let A be the number of breakdowns in a randomly chosen 6-week period, so A  Po(9).
Let n be the least number of breakdowns so that P( X > n)0.05
P( X > n) =−1 P( X n) ⇒ P( X n) =− 1 P( X > n)
So find n such that P( X n)0.95
From the tables with λ == 4.5 , P( X 13) 0.9261,
= P( X 14) 0.9585
So the smallest value of n is 14.

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 5
Exercise 2D
1 a X + Y ~ Po(3.3 + 2.7)
X + Y ~ Po(6)
e −6 × 65
P( X + Y = 5) = = 0.1606 (4 d.p.)
5!

b Use the tables with λ = 6

P( X + Y 7) = 0.7440

c Use the tables with λ = 6

P( X + Y > 4) =1 − P( X + Y 4) =1 − 0.2851 =0.7149

2 a A + B ~ Po(7.5)
e −7.5 × 7.57
P( A + B = 7) = = 0.1465 (4 d.p.)
7!

b Use the tables with λ = 7.5

P( A + B5) = 0.2414

c Use the tables with λ = 7.5

P( A + B > 9) =1 − P( A + B9) =1 − 0.7764 =0.2236

3 a As X and Y are independent

P( X = 2 and Y = 2) =P( X =2) × P(Y =2)
e −2.5 × 2.52 e −3.5 × 3.52
= ×
2! 2!
= 0.25652 × 0.18496 = 0.0474 (4 d.p.)

b As X and Y are independent:

P( X > 2 and Y > 2)= P( X > 2) × P(Y > 2)
(1 − P( X 2) )(1 − P(Y2) )
=
= (1 – 0.5438)(1 − 0.3208)
= 0.4562 × 0.6792 = 0.3099

c X + Y ~ Po(2.5 + 3.5), so X + Y ~ Po(6)

e −6 × 65
P( X + Y = 5) = = 0.1606 (4 d.p.)
5!

d Use the tables with λ = 6

P( X + Y 4) = 0.2851

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 1
4 a X ~ Po(3), Y ~ Po(5). As X and Y are independent:
P( X 3 and=
Y 3) P( X 3) × P(Y 3)
(1 − P( X 2) )(1 − P(Y2) )
=
= (1 – 0.4232)(1 − 0.1247)
= 0.5768 × 0.8753 = 0.5049

b X + Y ~ Po(8). Use the tables with λ = 8

P( X + Y 6) =
0.3134

5 a Let X be the number of cars and Y be the number of lorries passing in a 15-second period. So for a
15-second period X ~ Po(6), Y ~ Po(2).

i Assuming X and Y are independent:

P( X 4 and=
Y 4) P( X 4) × P(Y 4)
(1 − P( X 3) )(1 − P(Y3) )
=
= (1 – 0.1512)(1 − 0.8571)
= 0.8488 × 0.1429 = 0.1213

ii X + Y ~ Po(8). Use the tables with λ = 8

P( X + Y 9) =
0.7166

b It is assumed that the number of each vehicle type passing by follows a Poisson distribution (a
constant mean rate over a set period does not necessarily imply a Poisson distribution within that
period; for example, a set of traffic lights with a one minute cycle would allow a constant per-
minute rate without the per-15-second rate being consistent).
Another assumption is that the numbers of cars passing and trucks passing are independent (which
would not be the case if there is traffic congestion, where the rates of each would be affected by a
common external factor, and would in any case be unfeasible at large values where total road
space would become a restriction, so that many lorries passing might preclude many cars also
passing).

6 Let A and B be the number of taxis ordered by companies A and B (respectively) on a given day. So
A ~ Po(1.25) and B ~ Po(0.75), and A and B are independent.

e −1.25 ×1.252
a P( A
= 2)
= = 0.2238 (4 d.p.)
2!

b A + B ~ Po(2)
e −2 × 22
P( A + B = 2) = = 0.2707 (4 d.p.)
2!

c Let C be the total number of taxis ordered by the two companies in a given 5-day week.
C ~ Po(10). Use the tables with λ = 10
= P(C=
P(C < 10) 9) 0.4579

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 2
7 Let C and D be the number of times machines C and D (respectively) break down in a 12-week
period. So C ~ Po(1.2) and D ~ Po(0.6), and C and D are independent.
a As λ = 1.2, use a calculator to find the value.
P(C1) =
1 − P(C =
0)
e −1.2 ×1.20
= 1−
0!
1 − 0.3012 =
= 0.6988 (4 d.p.)

b P(C1 and=
D1) P(C1) × P( D1)
 e −0.6 × 0.60 
= 0.6988 1 − 
 0! 
= 0.6988(1 – 0.5488)
= 0.6988 × 0.4512 = 0.3153 (4 d.p.)

c C + D ~ Po(1.8). Use a calculator to find the required value.

e −1.8 ×1.83
P(C + D = 3) = = 0.1607 (4 d.p.)
3!

8 a Let A be the total number of calls received in a 4-minute period. A rate of 3 calls in 5 minutes is
equivalent to 2.4 calls in 4 minutes. So A  Po(2.4).
e −2.4 × 2.43
P( A= 3)= = 0.2090 (4 d.p.)
3!

b Let B be the total number of calls received in a 2-minute period. A rate of 3 calls in 5 minutes is
equivalent to 1.2 calls in 2 minutes. So B  Po(1.2).
P( B2) =
1 − P( B1) =
1 − P( B =−
0) P( B =
1)
e −1.2 ×1.20 e −1.2 ×1.21
1−
= −
0! 1!
1 − 0.30119 − 0.36143 =
= 0.3374 (4 d.p.)

c Let C be the total number of calls received in a 10-minute period. A rate of 3 calls in 5 minutes is
equivalent to 6 calls in 10 minutes. So C  Po(6). Use the tables with λ = 6
P(C5) = 0.4457

9 Let A, B and C be the number of times the ground, first-floor and second-floor photocopiers
(respectively) break down in a given week. So A  Po(0.4), B ~ Po(0.2) and C ~ Po(0.8), and assume
A, B and C are independent.

a As A, B and C are independent:

P( A =1 and B = 1 and C =1) =P( A =×
1) P( B =×
1) P(C =
1)
e −0.4 × 0.41 e −0.2 × 0.21 e −0.8 × 0.81
= × ×
1! 1! 1!
= 0.26813 × 0.16375 × 0.35946
= 0.0158 (4 d.p.)

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 3
9 b A + B + C ~ Po(1.4). Use a calculator to find the required value.
P(A + B + C1) =1 − P( A + B + C =0)
e −1.4 ×1.40
= 1−
0!
=1 − 0.2466 = 0.7534 (4 d.p.)

e −1.4 ×1.42
c P( A + B + C = 2) = = 0.2417 (4 d.p.)
2!

10 a Let A, B and C be the number of personal, business and advertising emails (respectively) arriving
in a given 30-minute period. Assume the number of each type of email arrives independently. So
A  Po(0.9), B ~ Po(1.85) and C ~ Po(0.75), and A, B and C are independent.
P( A1 and B1 and C1) = (1 − P( A = 0) ) × (1 − P( B = 0) ) × (1 − P(C = 0) )
 e −0.9 × 0.90   e −1.85 ×1.850   e −0.75 × 0.750 
= 1 −  × 1 −  × 1 − 
 0!   0!   0! 
= (1 − 0.4066) × (1 − 0.1572) × (1 − 0.4724)
= 0.5934 × 0.8428 × 0.5276
= 0.2639 (4 d.p.)

b Let D be the total number of emails received in an 8-hour period.

Total hourly email rate: 1.8 + 3.7 + 1.5 = 7, so D  Po(56).
1 − P( D50) =
P( D > 50) = 1 − 0.2343 =
0.7657 (4 d.p.)

c Let X be the number of days out of a 5-day working week on which the director receives more
0.7657 (from
than 50 emails. As the probability of receiving more than 50 emails is P( D > 50) =
part b), the model is X ~ B(5, 0.7657).
5 2 3
P( X = 2) =   (0.7657) (1 − 0.7657) = 0.0754 (4 d.p.)
2
 

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 4
Challenge
a If Q = 0 then because X and Y cannot take negative values, X = Y = 0.
P(Q= 0)
= P(X= 0 and Y= 0)
=P(X =0) × P(Y =0) (independence)
−λ −µ
= e ×e
= e−(λ + µ )

Alternatively, by addition of Poisson distributions X + Y  Po(λ + µ )

e − ( λ + µ ) × (λ + µ ) 0
So P(Q = 0) = P( X + Y = 0) =
0!
Which, as (λ + µ=0
) 1 and =
0! 1, gives
= e−(λ + µ )
P(Q= 0)

b If Q = 1 then ( X , Y ) = (0,1) or (1, 0)

P(Q= 1)= P ( ( X= 0 and Y= 1) or ( X= 1 and Y= 0 ) )
=P(X =0 and Y =+
1) P(X =
1 and Y =
0) (mutually exclusive)
=P(X =0) × P(Y =+
1) P(X =×
1) P(Y = 0) (independent)
 −λ e− µ µ1   e−λ λ1 − µ 
=e × + ×e 
 1!   1! 
−(λ +µ )
= (λ + µ )e

Alternatively, as X + Y  Po(λ + µ )
e − ( λ + µ ) × (λ + µ )1
So P(Q = 1) = P( X + Y = 1) =
1!
Which, as (λ + µ ) =
1
λ + µ and 1! = 1, gives
P(Q= 1)= (λ + µ ) e − ( λ + µ )

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 5
Exercise 2E

1 a

b Both the sample mean and the sample variance equal 1.43 correct to three significant figures. The
mean is close in value to the variance, which supports the choice of a Poisson distribution as a
model.

c Estimate λ = 1.4 from the data, so

2 a Let X be the number of cars passing the checkpoint in a 5-minute period.

b Both the sample mean and the sample variance equal 3.6 correct to one decimal place. The mean is
close in value to the variance, which supports the choice of a Poisson distribution as a model.

c Estimate λ = 3.6 from the data, so . Required value must be found using a calculator:

d From the data, the relative frequency of obtaining no more than 2 cars in a 5-minute period is:

which is close to the value calculated in part c and so further supports the choice of model.

3 a Let X be the number of flaws in a piece of cloth.

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 1
3 b Both the sample mean and the sample variance equal 2.9 correct to one decimal place. The fact
that the mean is close in value to the variance supports the choice of a Poisson distribution as a
model for flaws in a piece of cloth.

c The sample of only 120 pieces of cloth will be unreliable for estimating a proportion (especially of
extreme values) for a much larger sample. The fact that all of the 120 pieces sampled had fewer
than 8 flaws should not be taken as meaning that it is impossible, just rare, for a piece of cloth to
have 8 or more flaws and scaling up to give an estimate of 0 out of 10 000 would not be advisable.

d Estimate λ = 2.9 from the data, so Required value must be found using a calculator:

So there would be an estimated 99 pieces with at least 8 flaws.

Challenge

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 2
 Challenge (continued)
Var(
= X ) E( X 2 ) − (E( X )) 2 definition of Var( X )
= ∑ x P ( X= x ) − (E( X ))
2 2
definition of E( X 2 )
∞
e−λ λ x
= ∑ x2
x =0 x!
− λ2 given distribution
∞
e−λ λ x
0 + ∑ x2
= − λ2 splitting off first term
x =1 x !
∞
( x 2 − x + x )λ x
e−λ ∑ − λ2 taking out factor e − λ and rearranging
x =1 x!
∞
 x ( x − 1)λ x xλ x 
= e−λ ∑  + −λ
2
expanding
x =1  x ! x ! 
 ∞ λ x −2 ∞
λ x −1  2
= e−λ  λ 2 ∑ + λ∑  − λ cancelling common factors summation terms
=  x 2= ( x − 2 )! x 1 ( x − 1)! 
 2 ∞ λs ∞
λr  2
=e λ ∑ + λ∑  − λ
−λ
relabelling with s =x − 2, r =x −1
=  s 0=s! r 0 r! 

λr
= e − λ ( λ 2e λ + λ e λ ) − λ 2 known expansion eλ = ∑
r!
=λ

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 3
Chapter Review 2
1 a Let X be the number of accidents in a month. Assume a Poisson distribution.
So

b Let Y be the number of accidents in a 3-month period. So

c Let A be the number of accident-free months in a sample of six months. So from

part a

2 a X must have a constant mean rate so that the mean number of misprints in a sample is proportional
to the number of chapters. Misprints must occur independently of one another and be distinct (i.e.
can be counted singly in the text).

b The model is . Find P(X ⩽ 1) using a calculator:

c Let Y be the number of misprints in two randomly chosen chapters. So

As , the required value can be found from the tables in the textbook:

3 and

4 a The event (receipt of an email) has a constant mean rate through time. Emails are received singly
and independently of each other.

b i Let X be the number of emails received in a 10-minute period. Assume a Poisson distribution,
so

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 1
4 b ii Using the tables:

5 a A binomial distribution may be approximated by a Poisson distribution when n is

large and p is small, and typically np ⩽ 10.

c
Using the tables:

6 so
From the tables, for ,
For λ = 8 ,
So the largest integer value for λ satisfying the given condition is λ = 7

7 Let X be the number of fish caught in a 2-hour period. So . Use this to find the probability
that the angler catches at least 5 fish on a 2-hour fishing trip.

Then let Y be the number of trips in which the angle catches at least 5 fish from a sample of 5 trips.
So using the probability for catching at least 5 fish on a single trip,

8 a Let X be the number of cherries in a cake. So

ii Using the tables:

b Let Y be the number of cherries in 4 cakes. So

Using the tables:

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 2
8 c Let A be the number of packets containing more than 12 cherries in a sample of 8 packets. So,
using the result from part b,

9 a Let X be the number of cars sold in a week. A plausible model for number of cars sold in a week
would be .
It may be supposed that sales are independent of each other and occur singly (assuming the
salesman does not supply businesses); the constant mean rate is consistent with a Poisson model.

c Let Y be the number of weeks in which the salesman sells exactly 5 cars, in a sample of 4
consecutive weeks. So, using the result from part b,

10 Assuming a Poisson distribution for each random variable, with A and B being the number of letters
received by Aisha and Biyu in a given day (respectively): then and so

c Let Y be the number of days on which they receive a total of 3 letters between them, from a sample
of 5 days. So, using the result from part b,

11 Let X be the number of desktops sold in a day and Y the number of laptops sold in a day. Assuming
Poisson distributions for both variables and that sales of desktops and laptops are independent, then
and so

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 3
11 b

c Let A be the combined total of computer sales in a two-day period. So and the required
value can be found from the tables:

12 a Assuming that accidents can be modelled using a Poisson distribution (so that the mean number of
accidents in a given period of time will be proportional to the length of time), let X be the number
of accidents in a 6-month period. So , and from the tables:

b Let Y be the number of accidents in a single month. So

c Let A be the number of months in which there is at least one accident, out of a sample of 6 months.
From part b,

13 a Assume a Poisson distribution for breakdowns. Let X be the number of breakdowns in a single

month, so

b Let Y be the number of months in which there are at least 2 breakdowns.

4
P(Y= 3)=   (0.1443)3 (0.8557)=
1
0.0103 (4 d.p.)
3

14 a Visits can be counted singly; assuming visits are independent and at a constant average rate, they
may be modelled by a Poisson distribution; the rate of 240 per hour would then scale to a mean
rate of 4 in any given minute. Let X be the number of visits in a single minute. So

c Let Y be the number of visits in a two-minute period. So

Using the tables:

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 4
15 a Let X be the number of policies sold in the week.

=x
∑
=
fx 10 × 0 + 23 ×1 + 35 × 2 + 33 × 3 + 24 × 4 + 14 × 5 + 7 × 6 + 3 × 7 + 1× 8
= 2.86
∑f 10 + 23 + 35 + 33 + 24 + 14 + 7 + 3 + 1

b The values of mean and variance are the same, to 2 significant figures, which would support the
validity of a Poisson model.

Challenge
a Assuming the number of planes landing in a given period of time can be modelled by a Poisson
distribution, let X be the number of planes landing between 2 pm and 2:30 pm and let Y be the
number of planes landing between 2.30 pm and 3 pm. Then , and

P(A and B)
The solution uses the formula for conditional probability: P(A | B) =
P(B)
= P(X 5= and Y 5)
P(X = 5 | X + Y = 10) =
P(X + Y =10)
e −7.5 × 7.55 e −7.5 × 7.55
×
= 5! 5!
−15 10
e ×15
10!
10
10! 7.510 10 × 9 × 8 × 7 × 6  1  3 × 2 × 7 × 6 252 63
= 2
× 10 = ×  = = 10 = 8
(5!) 15 5× 4 × 3× 2 2 210 2 2
63
= = 0.2461 (4 d.p.)
256

P(X =
8 and Y =
2) + P(X = 9 and Y =+1) P(X =
10 and Y =
0)
=
P(X + Y =10)
e −7.5 × 7.58 e −7.5 × 7.52 e −7.5 × 7.59 e −7.5 × 7.51 e −7.5 × 7.510 e −7.5 × 7.50
× × ×
= 8! 2! + 9! 9! + 10! 0!
e −15 ×1510 e −15 ×1510 e −15 ×1510
10! 10! 10!
10
7.5  1 1 1  1
= 10
× 10! ×  + +  = 10 × ( 45 + 10 + 1)
15  2 × 8! 9! 10!  2
7
= = 0.0547 (4 d.p.)
128

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 5
Challenge (continued)
An alternative approach to this problem is to treat the 10 landings as each independently having a 0.5
chance of being in the first or second half of the hour and modelling them as a binomial. Let A be the
number of landings in the first half hour (2 pm to 2:30 pm) and . To answer part a, find
P(A = 5); to answer part b, find P(A > 7).

© Pearson Education Ltd 2019. Copying permitted for purchasing institution only. This material is not copyright free. 6