S2 Discrete Distributions - Poisson
S2 Discrete Distributions - Poisson
S2 Discrete Distributions - Poisson
com
1. Bhim and Joe play each other at badminton and for each game, independently of all others, the
probability that Bhim loses is 0.2
Bhim attends coaching sessions for 2 months. After completing the coaching, the probability
that he loses each game, independently of all others, is 0.05
(c) Calculate the mean and variance for the number of these 60 games that Bhim loses.
(2)
(d) Using a suitable approximation calculate the probability that Bhim loses more than 4
games.
(3)
(Total 10 marks)
2. A company has a large number of regular users logging onto its website. On average 4 users
every hour fail to connect to the company’s website at their first attempt.
(a) Explain why the Poisson distribution may be a suitable model in this case.
(1)
The company suffered from a virus infecting its computer system. During this infection it was
found that the number of users failing to connect at their first attempt, over a 12 hour period,
was 60.
(c) Using a suitable approximation, test whether or not the mean number of users per hour
who failed to connect at their first attempt had increased. Use a 5% level of significance
and state your hypotheses clearly.
(9)
(Total 15 marks)
3. A robot is programmed to build cars on a production line. The robot breaks down at random at a
rate of once every 20 hours.
(a) Find the probability that it will work continuously for 5 hours without a breakdown.
(3)
(d) Write down the probability that the robot will break down in the following 8 hour period.
Give a reason for your answer.
(2)
(Total 10 marks)
4. A café serves breakfast every morning. Customers arrive for breakfast at random at a rate of 1
every 6 minutes.
(a) fewer than 9 customers arrive for breakfast on a Monday morning between 10 am and 11
am.
(3)
(b) Using a suitable approximation, estimate the probability that more than 50 customers
arrive for breakfast next Tuesday.
(6)
(Total 9 marks)
5. An administrator makes errors in her typing randomly at a rate of 3 errors every 1000 words.
(a) In a document of 2000 words find the probability that the administrator makes 4 or more
errors.
(3)
The administrator is given an 8000 word report to type and she is told that the report will only
be accepted if there are 20 or fewer errors.
(b) Use a suitable approximation to calculate the probability that the report is accepted.
(7)
(Total 10 marks)
6. A cloth manufacturer knows that faults occur randomly in the production process at a rate of 2
every 15 metres.
A retailer buys a large amount of this cloth and sells it in pieces of length x metres. He chooses
x so that the probability of no faults in a piece is 0.80
(c) Write down an equation for x and show that x = 1.7 to 2 significant figures.
(4)
The retailer sells 1200 of these pieces of cloth. He makes a profit of 60p on each piece of cloth
that does not contain a fault but a loss of £1.50 on any pieces that do contain faults.
7. A botanist is studying the distribution of daisies in a field. The field is divided into a number of
equal sized squares. The mean number of daisies per square is assumed to be 3. The daisies are
distributed randomly throughout the field.
The botanist decides to count the number of daisies, x, in each of 80 randomly selected squares
within the field. The results are summarised below
∑ x = 295 ∑x 2
= 1386
(c) Calculate the mean and the variance of the number of daisies per square for the 80
squares. Give your answers to 2 decimal places.
(3)
(d) Explain how the answers from part (c) support the choice of a Poisson distribution as a
model.
(1)
(e) Using your mean from part (c), estimate the probability that exactly 4 daisies will be
found in a randomly selected square.
(2)
(Total 11 marks)
8. Each cell of a certain animal contains 11000 genes. It is known that each gene has a probability
0.0005 of being damaged.
(a) Suggest a suitable model for the distribution of the number of damaged genes in the cell.
(2)
(b) Find the mean and variance of the number of damaged genes in the cell.
(2)
(c) Using a suitable approximation, find the probability that there are at most 2 damaged
genes in the cell.
(4)
(Total 8 marks)
(a) Give two reasons to support the use of a Poisson distribution as a suitable model for the
number of calls per hour handled by the agent.
(2)
(b) Find the probability that in any randomly selected 15 minute interval the agent handles
The agent received some training to increase the number of calls handled per hour. During a
randomly selected 30 minute interval after the training the agent handles 14 calls.
(c) Test, at the 5% level of significance, whether or not there is evidence to support the
suggestion that the rate at which the agent handles calls has increased. State your
hypotheses clearly.
(6)
(Total 13 marks)
10. (a) State two conditions under which a Poisson distribution is a suitable model to use in
statistical work.
(2)
(b) Find the probability that in a randomly chosen 60 minute period there will be
The number of other vehicles, other than cars, passing the observation point in a 60 minute
interval is modelled by a Poisson distribution with mean 12.
(c) Find the probability that exactly 1 vehicle, of any type, passes the observation point in a
10 minute period.
(4)
(Total 11 marks)
(a) Suggest a suitable model for the number of faulty components detected per hour.
(1)
(b) Describe, in the context of this question, two assumptions you have made in part (a) for
this model to be suitable.
(2)
(c) Find the probability of 2 faulty components being detected in a 1 hour period.
(2)
(d) Find the probability of at least one faulty component being detected in a 3 hour period.
(3)
(Total 8 marks)
12. (a) Write down the conditions under which the Poisson distribution may be used as an
approximation to the Binomial distribution.
(2)
A call centre routes incoming telephone calls to agents who have specialist knowledge to deal
with the call. The probability of the caller being connected to the wrong agent is 0.01
(b) Find the probability that 2 consecutive calls will be connected to the wrong agent.
(2)
(c) Find the probability that more than 1 call in 5 consecutive calls are connected to the
wrong agent.
(3)
(d) Find the mean and variance of the number of wrongly connected calls.
(3)
(e) Use a Poisson approximation to find, to 3 decimal places, the probability that more than 6
calls each day are connected to the wrong agent.
(2)
(Total 12 marks)
The random variable K has a binomial distribution with parameters n = 25, p = 0.27.
14. (a) State the condition under which the normal distribution may be used as an approximation
to the Poisson distribution.
(1)
(b) Explain why a continuity correction must be incorporated when using the normal
distribution as an approximation to the Poisson distribution.
(1)
A company has yachts that can only be hired for a week at a time. All hiring starts on a
Saturday.
During the winter the mean number of yachts hired per week is 5.
(c) Calculate the probability that fewer than 3 yachts are hired on a particular Saturday in
winter.
(2)
During the summer the mean number of yachts hired per week increases to 25.
The company has only 30 yachts for hire.
(d) Using a suitable approximation find the probability that the demand for yachts cannot be
met on a particular Saturday in the summer.
(6)
(e) Estimate the number of Saturdays in the summer that the company will not be able to
meet the demand for yachts.
(2)
(Total 12 marks)
(a) Suggest a suitable model to represent the number of properties sold in a randomly chosen
week. Give two reasons to support your model.
(3)
(b) Find the probability that in any randomly chosen week the estate agent sells exactly
5 properties.
(2)
(c) Using a suitable approximation find the probability that during a 24 week period the
estate agent sells more than 181 properties.
(6)
(Total 11 marks)
16. Breakdowns occur on a particular machine at random at a mean rate of 1.25 per week.
(a) Find the probability that fewer than 3 breakdowns occurred in a randomly chosen week.
(4)
Over a 4 week period the machine was monitored. During this time there were 11 breakdowns.
(b) Test, at the 5% level of significance, whether or not there is evidence that the rate of
breakdowns has changed over this period. State your hypotheses clearly.
(7)
(Total 11 marks)
17. A manufacturer produces large quantities of coloured mugs. It is known from previous records
that 6% of the production will be green.
(a) Define a suitable distribution to model the number of green mugs in this sample.
(1)
(b) Find the probability that there were exactly 3 green mugs in the sample.
(3)
(c) Find the probability that there were between 10 and 13 (inclusive) green mugs in this
sample, using
18. Accidents on a particular stretch of motorway occur at an average rate of 1.5 per week.
(a) Write down a suitable model to represent the number of accidents per week on this stretch
of motorway.
(1)
(c) there is at least one accident per week for 3 consecutive weeks,
(3)
20. The random variable X is the number of misprints per page in the first draft of a novel.
(a) State two conditions under which a Poisson distribution is a suitable model for X.
(2)
The number of misprints per page has a Poisson distribution with mean 2.5. Find the probability
that
(c) the total number of misprints on 2 randomly chosen pages is more than 7.
(3)
(d) Using a suitable approximation find, to 2 decimal places, the probability that the chapter
will contain less than 40 misprints.
(7)
(Total 14 marks)
21. In a manufacturing process, 2% of the articles produced are defective. A batch of 200 articles is
selected.
(a) Giving a justification for your choice, use a suitable approximation to estimate the
probability that there are exactly 5 defective articles.
(5)
(b) Estimate the probability that there are less than 5 defective articles.
(2)
(Total 7 marks)
Find
23. From company records, a manager knows that the probability that a defective article is produced
by a particular production line is 0.032.
(b) Using a suitable approximation, find the probability that fewer than 4 of them are
defective.
(4)
(c) Using a suitable approximation, find the probability that more than 42 are defective.
(6)
(Total 13 marks)
24. Over a long period of time, accidents happened on a stretch of road at random at a rate of 3 per
month.
At a later date, a speed restriction was introduced on this stretch of road. During a randomly
chosen month only one accident occurred.
(c) Test, at the 5% level of significance, whether or not there is evidence to support the claim
that this speed restriction reduced the mean number of road accidents occurring per
month.
(4)
The speed restriction was kept on this road. Over a two-year period, 55 accidents occurred.
(d) Test, at the 5% level of significance, whether or not there is now evidence that this speed
restriction reduced the mean number of road accidents occurring per month.
(7)
(Total 16 marks)
25. (a) State two conditions under which a random variable can be modelled by a binomial
distribution.
(2)
In the production of a certain electronic component it is found that 10% are defective.
(b) Write down a suitable model for the distribution of defective components in a batch.
(1)
(e) Find the mean and the variance of the defective components in a batch.
(2)
A supplier buys 100 components. The supplier will receive a refund if there are more than 15
defective components.
(f) Using a suitable approximation, find the probability that the supplier will receive a
refund.
(4)
(Total 13 marks)
26. (a) Explain what you understand by a critical region of a test statistic.
(2)
The number of breakdowns per day in a large fleet of hire cars has a Poisson distribution with
1
mean .
7
(b) Find the probability that on a particular day there are fewer than 2 breakdowns.
(3)
(c) Find the probability that during a 14-day period there are at most 4 breakdowns.
(3)
The cars are maintained at a garage. The garage introduced a weekly check to try to decrease the
number of cars that break down. In a randomly selected 28-day period after the checks are
introduced, only 1 hire car broke down.
(d) Test, at the 5% level of significance, whether or not the mean number of breakdowns has
decreased. State your hypotheses clearly.
(7)
(Total 15 marks)
27. Minor defects occur in a particular make of carpet at a mean rate of 0.05 per m2.
(a) Suggest a suitable model for the distribution of the number of defects in this make of
carpet. Give a reason for your answer.
(3)
A carpet fitter has a contract to fit this carpet in a small hotel. The hotel foyer requires 30 m2 of
this carpet. Find the probability that the foyer carpet contains
The carpet fitter orders a total of 355 m2 of the carpet for the whole hotel.
(d) Using a suitable approximation, find the probability that this total area of carpet contains
22 or more defects.
(6)
(Total 14 marks)
28. The random variable R has the binomial distribution B(12, 0.35).
The random variable S has the Poisson distribution with mean 2.71.
(6)
(Total 14 marks)
29. (a) Write down two conditions needed to be able to approximate the binomial distribution by
the Poisson distribution.
(2)
A researcher has suggested that 1 in 150 people is likely to catch a particular virus.
Assuming that a person catching the virus is independent of any other person catching it,
(b) find the probability that in a random sample of 12 people, exactly 2 of them catch the
virus.
(4)
(c) Estimate the probability that in a random sample of 1200 people fewer than 7 catch the
virus.
(4)
(Total 10 marks)
30. Vehicles pass a particular point on a road at a rate of 51 vehicles per hour.
(a) Give two reasons to support the use of the Poisson distribution as a suitable model for the
number of vehicles passing this point.
(2)
After the introduction of a roundabout some distance away from this point it is suggested that
the number of vehicles passing it has decreased. During a randomly selected 10 minute interval
4 vehicles pass the point.
(d) Test, at the 5% level of significance, whether or not there is evidence to support the
suggestion that the number of vehicles has decreased. State your hypotheses clearly.
(6)
(Total 13 marks)
31. (a) Write down the condition needed to approximate a Poisson distribution by a Normal
distribution.
(1)
32. A doctor expects to see, on average, 1 patient per week with a particular disease.
(a) Suggest a suitable model for the distribution of the number of times per week that the
doctor sees a patient with the disease. Give a reason for your answer.
(3)
(b) Using your model, find the probability that the doctor sees more than 3 patients with the
disease in a 4 week period.
(4)
The doctor decides to send information to his patients to try to reduce the number of patients he
sees with the disease. In the first 6 weeks after the information is sent out, the doctor sees
2 patients with the disease.
(c) Test, at the 5% level of significance, whether or not there is reason to believe that sending
the information has reduced the number of times the doctor sees patients with the disease.
State your hypotheses clearly.
(6)
Medical research into the nature of the disease discovers that it can be passed from one patient
to another.
(d) Explain whether or not this research supports your choice of model. Give a reason for
your answer.
(2)
(Total 15 marks)
33. A botanist suggests that the number of a particular variety of weed growing in a meadow can be
modelled by a Poisson distribution.
(a) Write down two conditions that must apply for this model to be applicable.
(2)
Assuming this model and a mean of 0.7 weeds per m2, find
(b) the probability that in a randomly chosen plot of size 4 m2 there will be fewer than 3 of
these weeds.
(4)
(c) Using a suitable approximation, find the probability that in a plot of 100 m2 there will be
more than 66 of these weeds.
(6)
(Total 12 marks)
34. The continuous random variable X represents the error, in mm, made when a machine cuts
piping to a target length. The distribution of X is rectangular over the interval [–5.0, 5.0].
Find
(c) Find the probability that more than half of them are within 1.5 cm of the target length.
(3)
If X < − 4.2, the length of piping cannot be used. At the end of each day the supervisor checks a
random sample of 60 lengths of piping.
(d) Use a suitable approximation to estimate the probability that no more than 2 of these
lengths of piping cannot be used.
(5)
(Total 11 marks)
35. On a typical weekday morning customers arrive at a village post office independently and at
a rate of 3 per 10 minute period.
(b) no more than 7 customers arrive between 11.00 a.m. and 11.30 a.m.
(3)
The period from 11.00 a.m. to 11.30 a.m. next Tuesday morning will be divided into 6 periods
of 5 minutes each.
(c) Find the probability that no customers arrive in at most one of these periods.
(6)
(d) Using a suitable approximation, estimate the probability that more than 49 customers
arrive at the post office next Wednesday morning.
(7)
(Total 18 marks)
1. (a) Let X be the random variable the number of games Bhim loses.
X ~ B(9, 0.2) B1
9!
( 0.2 ) ( 0.8)
3 6
P(X ≤ 3) – P(X ≤ 2) = 0.9144 – 0.7382 or M1
3!6!
= 0.1762 = 0.1762
awrt 0.176 A1 3
Note
B1 – writing or use of B(9, 0.2)
9!
M1 for writing/ using P(X ≤ 3) – P(X ≤ 2) or (p)3 (1 – p)6
3!6!
A1 awrt 0.176
Special case : Use of Po(1.8)
e –1.8 1.83
can get B1 M1 A0 – B1 if written B(9, 0.2), M1 for
3!
or awrt to 0.161
If B(9, 0.2) is not seen then the only mark available for using Poisson is M1.
(b) can get M1 A0 – M1 for writing or using P(X ≤ 4) or may be implied
by awrt 0.964
57
(c) Mean = 3 variance = 2.85, B1 B1 2
20
Note
B1 3
B1 2.85, or exact equivalent
(b) Po (8) B1
Note
B1 Writing or using Po(8) in (i) or (ii)
Note
B1 both hypotheses correct. Must use λ or μ
M1 identifying normal
A1 using or seeing mean and variance of 48
These first two marks may be given if the following are seen in the
standardisation formula : 48 and 48 or awrt 6.93
M1 for attempting a continuity correction (Method 1: 60 ± 0.5 /
Method 2: x ± 0.5)
M1 for standardising using their mean and their standard deviation and
using either Method 1 [59.5, 60 or 60.5. accept ± z.] Method 2 [(x ± 0.5)
and equal to a ± z value)
59.5 – 48 x – 0.5 – 48
A1 correct z value awrt ±1.66 or ± , or =1.6449
48 48
A1 awrt 3 sig fig in range 0.0484 – 0.0485, awrt 59.9
M1 for “reject H0” or “significant” maybe implied by “correct
contextual comment”
If one tail hypotheses given follow through “their prob”
and 0.05, p < 0.5
If two tail hypotheses given follow through “their prob”
with 0.025, p < 0.5
If one tail hypotheses given follow through “their prob”
and 0.95, p > 0.5
If two tail hypotheses given follow through “their prob”
with 0.975, p > 0.5
If no H1 given they get M0
A1 ft correct contextual statement followed through from their prob and
H1. need the words number of failed connections/log ons has increased o.e.
Allow “there are more failed connections”
NB A correct contextual statement alone followed through
from their prob and H1 gets M1 A1
[15]
3. (a) Y ~ Po(0.25) B1
P(Y = 0) = e–0.25 M1
= 0.7788 A1 3
Note
B1 for seeing or using Po(0.25)
M1 for finding P(Y = 0) either by e–a, where a
is positive (a needn’t equal their λ) or using
tables if their value of λ is in them
Beware common Binomial error using, p = 0.05
gives 0.7738 but scores B0 M0 A0
A1 awrt 0.779
(b) X ~ Po(0.4) B1
P(Robot will break down) = 1 – P(X = 0)
= 1 –e–0.4 M1
= 1 –0.067032
= 0.3297 A1 3
Note
B1 for stating or a clear use of Po(0.4) in part (b) or (c)
M1 for writing or finding 1 – P(X =0)
A1 awrt 0.33
e –0.4 (0.4 )
2
(c) P(X = 2) M1
2
= 0.0536 A1 2
Note
e – λ λ2
M1 for finding P(X =2) e.g with their value of λ in
2!
or if their λ is in the table for writing
P(X ≤ 2) – P(X ≤ 1)
A1 awrt 0.0536
4. (a) X ~ Po(10) B1
P(X < 9) =P(X ≤ 8) M1
= 0.3328 A1 3
Note
B1 for using Po(10)
M1 for attempting to find P(X ≤ 8): useful values
P(X ≤ 9) is 0.4579(M0), using Po(6) gives 0.8472, (M1).
1
A1 awrt 0.333 but do not accept
3
(b) Y ~ Po(40) M1 A1
Y is approximately N(40,40)
P(Y > 50) = 1 – P(Y ≤ 50) M1
50.5 – 40
= 1 – P Z < M1
40
= 1 – P(Z < 1.660..) A1
= 1 –0.9515
= 0.0485 A1 6
N.B. Calculator gives 0.048437.
Poisson gives 0.0526 (but scores nothing)
Note
1st M1 for identifying the normal approximation
1st A1 for [mean = 24] and [sd = 24 or var = 24]
These first two marks may be given if the
following are seen in the standardisation formula :
24
or awrt 4.90
24
e –2 × 2 4
6. (a) X ~Po(2) P( X = 4) = = 0.0902 awrt 0.09 M1
4!
A1 2
Note
M1 for use of Po(2) may be implied
A1 awrt 0.09
(b) Y ~Po(8) B1
P(Y >10) = 1– P(Y ≤ 10) = 1 – 0.8159
= 0.18411… awrt 0.184 M1A1 3
Note
B1 for Po(8) seen or used
M1 for 1 – P(Y ≤ 10) oe
A1 awrt 0.184
35 3 6
(b) P (X ≤ 6) – P (X ≤ 4) =0.9665 – 0.8153 e–3 + M1
5! 6!
= 0.1512 A1 2
(c) μ = 3.69 B1
2
1386 295
Var (X) = – M1
80 80
(d) For a Poisson model, Mean = Variance ; For these data 3.69 ≈ 3.73
⇒ Poisson model B1 1
e –3.6875 3.6875 4
(e) = 0.193 allow their mean or var M1
4!
Awrt 0.193 or 0.194 A1 ft 2
[11]
B1 both . Must be one tail test. They may use λ or µ and either 9
or 18 and match H0 and H1
M1 Po (9) may be implied by them using it in their calculations.
M1 attempt to find P(X ≥ 14) eg 1 – P(X < 13) or 1 – P(X < 14)
A1 correct probability or CR
To get the next2 marks the null hypothesis must state or
imply that (λ) = 9 or 18
M1 for a correct statement based on their probability or
critical region or a correct contextualised statement that implies that.
A1. This depends on their M1 being awarded for accepting H0.
Conclusion in context. Must have calls per hour has not increased.
Or the rate of calls has not increased.
Any statement that has the word calls in and implies the rate not increasing
e.g. no evidence that the rate of calls handled has increased
Saying the number of calls has not increased gains A0 as it does not imply rate
NB this is an A mark on EPEN
They may also attempt to find P(X < 14) = 0.9261 and compare with 0.95
[13]
(b) Let X be the random variable the number of cars passing the
observation point.
Po(6) B1
e −6 6 4
(i) P(X ≤ 4) – P(X ≤ 3) = 0.2851 – 0.1512 or M1
4!
= 0.1339 A1
64 63 62 6
(ii) 1 – P(X ≤ 4) = 1 – 0.2851 or 1 – e − 6 + + + + 1 M1
4! 3! 2! 1!
= 0.7149 A1 5
B1 Using Po(6) in (i) or (ii)
e −λ λ4
M1 Attempting to find P(X ≤ 4) – P(X ≤ 3) or
4!
A1 awrt 0.134
M1 Attempting to find 1 – P(X ≤ 4)
A1 awrt 0.715
e −1.5 (1.5) 2
(c) P(X = 2) = P(X ≤ 2) – P(X ≤ 1) or M1
2
= 0.8088 – 0.5578
= 0.251 awrt 0.251 A1 2
(e) X ~ Po(10)
P(X > 6) = 1 – P(X ≤ 6) M1
= 1 – 0.1301
= 0.8699 awrt 0.870 A1 2
[12]
7 5 e −7
(b) P(X = 5) = P(X ≤ 5) − P(X ≤ 4) or M1
5!
= 0.3007 − 0.1730
= 0.1277 A1 2
awrt 0.128
P (X = 3) = 10C3(0.06)3(0.94)7 M1
10
C3(p)3(l – p)7
0.016808.... A1 3
awrt 0.0168
= P(0.75 ≤ z ≤ 2.26) A1
awrt 0.75 and 2.26
= 0.2147 A1 6
awrt 0.214 or 0.215
[13]
18. (a) Let X be the random variable the no. of accidents per week
X ~Po(1.5) need poisson and λ must be in part (a) B1 1
−1.5 eµ µ 2
P(X = 2) = e 1.5
2
(b) or P(X ≤ 2) – P(X ≤ 1) M1
2 2
= 0.2510 awrt 0.251 A1 2
z = 7.5 − 3 = 2.62
2.94
p(x >7) = 1 – p(x < 7.5)
= 1 – 0.9953
= 0.0047
e −4 4 5
P(X = 5) = M1
5!
P(X ≤ 5) – P(X ≤ 4)
= 0.1563 A1 5
(c) P(T = 5) = 0 B1 1
cao
[4]
Aliter (d)
p = 0.0262 < 0.05
AWRT 0.026 equn to –1.6449
[16]
26. (a) A range of values of a test statistic such that if a value of the test statistic
obtained from a particular sample lies in the critical region,
then the null hypothesis is rejected (or equivalent). B1B1 2
(d) X ~ P(17.75) B1
Implied
X ~ N(17.75, 17.75) B1
Normal, 17.75
21.5 − 17.75
P(X ≥ 22) = P Z > M1M1
17.75
Standardise, accept 22 or ±0.5
= –P(Z > 0.89) A1
awrt 0.89
= 0.1867 A1 6
[15]
(c) H0 : λ = 6 B1
H1 : λ < 6 B1
P(X ≤ 2) = 0.0620 α = 0.05 ⇒ critical region X ≤ 1 M1 A1
0.0620 > 0.05 2 not in critical region M1
The number of patients with the disease seen
by the doctor has not been reduced A1 6
(d) This does not support the model as the disease will occur in
outbreaks; the patients seen by the doctor are unlikely to
be independent of each other/don’t occur singly B1; B1 2
[15]
33. (a) Weeds grow independently, singly, randomly and at a constant rate
(weeds/m2) any 2 B1 B1 2
(b) Let X represent the number of weeds/m2
X ∼ Po(0.7), so in 4 m2, λ = 4 × 0.7 = 2.8 B1
P(Y < 3) = P(Y = 0) + P(Y = 1) + P(Y = 2) M1
2.8 2
= e−2.8 1 + 2.8 + A1
2
= 0.46945 A1 4
34.
f(x)
1
10
49.5 − 63
= P Z ≥
63
standardising M1
= P(Z ≥ −1.7008) A1
= 0.9554 (tables)
(accept awrt 0.955 or 0.956) A17
[18]
1. This question was well answered by the majority of candidates with many scoring full marks.
There were, of course, candidates who failed to score full marks. This was usually the result of
inaccurate details, rather than lack of knowledge. In particular, manipulation of inequalities
requires concentration and attention to detail. In part (a) the most common error seen was
using P( X = 3) = P( X ≤ 4) − P( X ≤ 3) .
Parts (b) and (c) were usually correct. The most common error was to find P(X ≤ 3) rather than
P(X ≤ 4) in part (b). A minority of candidates used the Normal as their approximation in part
(d). The simple rule “n is large, p is small: use Poisson” clearly applies in this case.
2. The majority of candidates were familiar with the technical terms in part (a), but failed to
establish any context.
Part (b) was a useful source of marks for a large proportion of the candidates. The only
problems were occasional errors in detail. In part (i) a few did not spot the change in time scale
and used Po(4) rather than Po(8). Some were confused by the wording and calculated P(X = 8)
rather than P(X = 0). The main source of error for (ii) was to find 1 – P(X ≤ 4) instead of
1 – P(X ≤ 3).
In part (c) the Normal distribution was a well-rehearsed routine for many candidates with many
candidates concluding the question with a clear statement in context.
The main errors were
• Some other letter (or none) in place of λ or µ
• Incorrect Normal distribution: e.g. N(60, 60)
• Omission of (or an incorrect) continuity correction
• Using 48 instead of 60
• Calculation errors
A minority of candidates who used the wrong distribution (usually Poisson) were still able to
earn the final two marks in the many cases when clear working was shown. This question was
generally well done with many candidates scoring full marks.
3. Although there were a minority of candidates who were unable to identify the correct
distribution to use the majority of candidates achieved full marks to parts (a) (b) and (c). Part (d)
seemed to cause substantial difficulty. In part (a) the majority of candidates identified that a
Poisson (rather than the Binomial) distribution was appropriate but some calculated the
parameter as 2.5 or 4 rather than 0.4. A few used Po(1) and calculated P(Y = 5).
In part (b) and part (c) the most common error was to use Po(2.5). The majority of candidates
were able to work out P(X >1) and P(X = 2) using the correct Poisson formula. Many thought
that their answer to part (c) was the correct solution while others used or multiplied their
answers to both (b) and (c). Whether stating a correct or incorrect solution only a minority used
the statistical term “independence” as the reason for their answer.
5. Part (a) was answered well with the majority of candidates gaining full marks.
Part (b) was also a good source of marks for a large majority of the candidates. Common errors
included using 23.9…for variance and 19.5 instead of 20.5. A sizeable minority of candidates
used 21.5 after applying the continuity correction. A few candidates had correct working up to
the very end when they failed to find the correct probability by not subtracting the tables’
probability from 1.
6. Parts (a) and (b) were completed successfully by most candidates. The most common errors
seen were using the wrong Poisson parameter or identifying the incorrect probability in part (b).
Part (c) proved to be a good discriminator with only those with good mathematical skills able to
attain all the marks for this part of the question. Few candidates used the method given on the
mark scheme and chose to use natural logarithms instead. Whilst this is an accepted method this
knowledge is not expected at S2 and full marks were gained by the most able candidates using
the given method.
In part (d) a few candidates seemed confused by this with some using 1.7 or 2/15 as a
probability rather than the 0.8 given in the question, and far too many seemed unable to use 60p
and £1.50 correctly when calculating the profit.
7. This question proved to be a good start to the paper for a majority of the candidates. There were
many responses seen which earned full marks.
The most common errors in parts (a) and (b) concerned the routine manipulation of inequalities.
In part (a) 1 – P ( X ≤ 1) was often seen and in (b), while most candidates agreed that
P(5 ≤ X ≤ 6 ) was the required probability, with many then choosing the standard technique of
P( X ≤ 6 ) – P( X ≤ 4 ) , there were candidates who proceeded with a variety of methods. Incorrect
expressions such as P( X ≤ 6 ) – P( X ≤ 4 ) were seen not infrequently. A correct but inefficient
method which was commonly used included:
P( X = 5) + P( X = 6 ) = (P( X ≤ 6 ) – P( X ≤ 5)) + (P( X ≤ 5) – P( X ≤ 4 ))
Part (c) was poorly answered. There were a significant minority of candidates who obtained a
‘correct’ answer for the mean in part (c), but who nevertheless lost the mark because their
answer was not written, as instructed, correct to 2 decimal places. Many candidates were unable
to calculate the variance. There were a variety of incorrect formulae used.
The general response to (d) was good, although many candidates simply gave the response that
is appropriate for a more frequent type of question on the Poisson distribution requiring
comment: (“singly/independently/randomly/constant rate”).
Part (e) was particularly well done. Even the minority who struggled, or even omitted, some of
the earlier parts of the question were able to gain both marks in part (e).
8. This question was generally answered well. A few candidates put the Poisson for (a) and then
used Variance = Mean to got 5.5 for the variance. Some candidates rounded incorrectly giving
an answer of 5.49 for the variance.
Part (c) was generally answered correctly although a minority of candidates used the normal
approximation – most used 2.5 in their standardisation and so got 1 mark out of the 4.
9. Part(a) was done well generally although a reference to ‘calls’ was not made by a few
candidates. Weak candidates talked about the Poisson needing large numbers and others seemed
to not understand what was required at all; writing ‘quick and easy’. Part (b) was done correctly
by the majority of candidates. A few did not use Po(4.5) in (i) and a number used P(X > 8) = 1 –
P(X ≤ 7) in (ii).In part (c) weaker candidates did not use λ for their hypotheses nor did they use
Po(9). Some hypotheses had λ = 3.5 and λ ≥ 3.5. Two tail tests were often suggested. Most
candidates got to 0.0739 and only a few candidates used the critical value route. Only the able
candidates got the interpretation of the significance test correct. Weak candidates generally only
considered whether it was significant or not, with mixed success. They rarely managed to
interpret correctly in context.
10. Most candidates were able to attempt part (b) successfully as these were fairly standard
calculations. However, when required to apply Poisson probabilities to a problem it was only
the better candidates who attained any marks in part (c).
(a) A sizeable proportion of candidates, whilst having learnt the conditions for a Poisson
distribution, failed to realise that this applied to the events occurring. There were
references to ‘trials’ and ‘things’ in some solutions offered.
(b) (i) Most candidates recognised Po(6) and were able to answer this successfully, either
from the tables or by calculation. Common errors were using incorrect values from
the table or calculating the exact value incorrectly.
(ii) Although many candidates attained full marks for this part, some were unable to
express ‘at least 5’ correctly as an inequality and used P(X ≤ 5).
(c) Many of the successful candidates used a Po(3) to give a correct solution. Of the rest most
candidates failed to realise that there were two ways for exactly one vehicle to pass the
point and so only performed one calculation. This was often for P(1 car) and P(1 other
vehicle), which were then added together. However there were some candidates who
gained the correct answer via this method.
11. This question was quite well done. Most candidates were aware of the conditions for a Poisson
distribution but many lost marks as they did not answer the question in context, although the
examination question actually specified this. A few did not realise that independent and random
were the same condition. Cumulative probability tables were well used in part (c) and there
were many accurate solutions using the Poisson formula.
Most candidates used a mean of 4.5 in part (d) and there were many accurate results.
12. This question was accessible to most candidates. Parts (a), (d) and (e) were generally well
answered with a small proportion of candidates using Po(10) in part (d) and thus quoting that
the mean and variance were the same. Part (b) caused a few problems as some used Po(0.1) and
found P(X = 2) or used X ~ B(2, 0.1). In part (c) some candidates still did not correctly use
P(X > 1) = 1 – P(X < 1). A similar mistake occurred in part (e) where they used
P(X > 6) = 1 – P(X ≤ 6).
Candidates need to be reminded of the rubric on the front of the question paper. It does say
‘appropriate degrees of accuracy’. Many rounded too early and did not realise that an answer to
1sf is not accurate enough. Answers of 0.001 were common in part (c)
13. This was usually completely correct with very few errors.
14. It was disappointing to find that a large number of candidates failed to attain both of the first 2
marks available. These were often the only marks lost by some, since the majority of candidates
achieved most or all marks. In part (d) most candidates did attempt an approximation, although
a minority calculated an exact binomial. Again, the common errors were to fail to use a
continuity correction and the standard deviation when using the approximation and then not
using the 1 – Φ(z). The simple calculation of 16 x the answer to part (d) was performed
correctly by the majority of candidates attempting this part of the question. A common error
was to attempt a binomial probability.
15. The majority of candidates knew the conditions for the Poisson distribution but many did not
get the marks because they failed to put them into context. As in many previous series, it was
very common for candidates to repeat at least some of these conditions parrot-fashion preceded
by “events occur” or “it occurs”. Other common errors were listing randomness and
independence as separate reasons and citing the fixed time period and lack of an upper limit as
reasons. Quite a number failed to mention the parameter for the distribution. The majority of
candidates answered part (b) correctly. Most candidates answered part (c) correctly. Where
marks were lost it was usually through failing to use a continuity correction rather than applying
it wrongly.
16. Most candidates answered Part (a) correctly. A small number of candidates calculated the
probability for less than or equal to 3 although a minority thought that dividing by 0! in P(X = 0)
gave zero. In part (b) carrying out the hypothesis test was more challenging though there was
clear evidence that candidates had been prepared for this type of question. However, using p
instead of λ or µ, when stating the hypotheses, was often seen and incorrectly stating H1 as λ >
1.25 or 5 also lost marks. Many candidates calculated P(X ≤ 11) instead of looking at P(X ≥ 11).
A diagram would have helped them or the use of the phrase “a result as or more extreme than
that obtained”. Those who used the critical region approach made more errors. Some candidates
correctly calculated the probability and compared it with 0.025 but were then unsure of the
implications for the hypotheses. A few candidates used a 2-tailed hypothesis but then used 0.05
rather than 0.025 in their comparison. Most candidates gave their conclusions in context.
18. Many candidates did well on this question and gained all 9 marks.
In part (a) nearly all candidates realised the Poisson distribution was appropriate but not all
stated the parameter of 1.5. Often this was stated in part b which did not gain the marks.
In part (c) many candidates used a Poisson distribution with a parameter of 4.5 rather than
cubing the probability of X ≥ 1 from a Poisson 1.5.
Part (d) was done very well by all candidates; the main error being the statement
P(X > 4) = 1 – P(X ≤ 3)
19. Most candidates correctly used a Po(3) distribution although a significant minority attempted to
used a Normal distribution. The most common error was using P(X > 7) = 1 – P(X ≤ 6)
20. Many candidates did not achieve any marks in part (a) as they failed to give conditions in
context, events being the most common error seen. Part (b) was done well and if candidates lost
a mark then it was usually the final mark due to accuracy. A common answer was 0.082. Part
(c) was generally completed satisfactorily, but there were a number of candidates who struggled
with the inequalities. A common error was P(Y > 7) = 1 – P(T ≤ 6). Diagrams were again in
evidence; these candidates were perhaps the least likely to make mistakes with the inequalities.
Some candidates calculated a probability using P(2.5) and then squared their answer. The
overall response to part (d) was good. However, only a minority scored full marks. Most
candidates failed to implement the instruction to write their answer “to 2 decimal places”. There
were other errors; omission of the continuity correction or the wrong version (40.5), confusion
between variance and standard deviation, and problems dealing with a negative z-value.
21. The overall response was good, with a large number of candidates scoring at least five out of the
seven marks. However, a small number of candidates chose to ignore the instructions to “use a
suitable approximation”. Most candidates were familiar with the conditions required for the
Poisson approximation to the Binomial in part (a). However, a small number of candidates
quoted this correct reason, but used this as justification for a Normal approximation. Not all
candidates using the Poisson distribution earned the second mark. A final answer of 0.156 was
fairly common, particularly amongst those who had used the Poisson formula rather than the
cumulative tables.
The response to part (b) was excellent. There were a large number of perfect answers. A small
number of candidates ignored the instruction to approximate and continued the use the Binomial
22. Candidates knew how to answer parts (a) and (b) but many did not work to sufficient accuracy.
If they used their calculator instead of the tables they were expected to give their answer to the
same accuracy as the tables. Too many of them did not read part(c) carefully enough. The
random variable T was defined to be normally distributed and thus P(T =5) = 0.
23. This question was a good source of marks for many of the candidates, with many of them
gaining full marks. For those that did not gain full marks, the common errors were premature
approximation; wrong interpretation of ‘fewer than 4’;ignoring the continuity correction and in
part (c) using a Poisson approximation and then a normal approximation to this Poisson
approximation.
24. For those candidates that could interpret ‘more than 4 accidents occurred’ correctly parts (a) and
(b) were a good source of marks. Part (b) was often well answered and many candidates gained
full marks. In part (c) incorrect hypotheses and ignoring the continuity correction were the
common errors coupled with poor use of the appropriate significance test. Candidates need to
have a simple algorithm at their fingertips to deal with tests of significance.
25. This was a good source of marks for a large majority of candidates. Many were able to write
down two conditions for a Binomial distribution, with some writing down all four conditions for
good measure. However some candidates muddled up the concepts of trial, event and outcome.
The response of ‘a fixed number of events’ was given no credit. Parts (b), (c), (d) and (e) were
usually well answered. In part (f), successful candidates either applied a Po (10) or a N(10, 9)
approximation. Some of the candidates who used the N(10, 9) approximation did not apply the
correct continuity correction.
26. In part (a) many candidates struggled to explain the concept of a critical region, although some
gave a correct definition as the range of values where the null hypothesis is rejected. Many
correct solutions were seen for parts (b) and (c). However the weaker candidates were not able
to translate the concept ‘at most 4 breakdowns’ to the correct inequality. In part (d), as with Q3,
many candidates successfully performed the required hypothesis test using a probability
method. Again, there was a sizeable number of candidates who incorrectly found P(X=1) and
compared this probability with the significance level. Again, a minority of candidates decided to
approach this question using the critical region strategy. Marks were lost if candidates did not
give evidence of their chosen critical region.
27. This was a well answered question and high marks were frequently being scored. In part (a),
many candidates chose the correct Poisson model and gave a correct reason in the light of the
problem posed. Many correct solutions were seen in parts (b) and (c). In part (d), many
candidates found the correct Normal approximation to the Poisson distribution. The most
common error was the incorrect application of the continuity correction with weaker candidates
generally losing the final two accuracy marks. In part (a), a small minority of candidates
incorrectly chose a Binomial model and applied this model throughout the question, thereby
losing a considerable number of marks.
28. The whole question was generally very well answered, by far the most common error was the
use of a continuity correction in part (c). A small number of candidates didn't realise that part
(b) required the sum of two probabilities.
29. This was a well answered question with many candidates scoring full marks. In part (a), many
candidates realised the conditions of a large value of n and a small value of p when
approximating the Binomial Distribution by the Poisson distribution. One common error in part
(b) was for candidates to apply the Poisson approximation when the number of trials was only
twelve, even though these candidates were able to write down the appropriate conditions in part
(a).
30. Candidates were able to express two conditions for a Poisson distribution in context with
vehicles passing by a particular point on the road. Many candidates then answered part (b) and
(c) correctly. In part (d) a majority of candidates was able to give a full solution by either using
a probability or critical region approach to their hypothesis test.
31. This question was generally well answered with many candidates gaining full marks. In part (b)
a minority of candidates standardized correctly but then found the incorrect area.
32. This question was well answered with many candidates gaining 14 or 15 marks. In part (a) most
candidates spotted the Poisson distribution but few related their reason to the context of the
question. In part (d) the majority of candidates realized that the Poisson was no longer
appropriate, but some failed to give a reason which related the breakdown of one of the Poisson
conditions to the context of the question.
33. Most candidates scored well on this question, but too many lost marks in part (a) by not giving
their answer in context. Too many candidates lost a mark by not answering part (b) to the
required level of accuracy. Candidates were obviously more at ease in part (c) where they were
able to leave the answer as they found it from tables.