4.17 · AHL · HL onlyPaper 1Paper 2Paper 3

Poisson distribution

In a minute

A counts events in a fixed stretch of time or space, when they happen independently at a steady average rate. Its mean and variance are both mm, your GDC gives its probabilities, and independent Poisson counts add to another Poisson count.

You'll be able to

  • Decide whether a Poisson model fits, write X∼Po(m)X \sim \mathrm{Po}(m), and state its assumptions in context.
  • Use the mean and variance of Po(m)\mathrm{Po}(m), both equal to mm, without proving them, including to judge whether data look Poisson.
  • Find Poisson probabilities on your GDC, after scaling the mean to the interval.
  • Add independent Poisson counts by adding their means.
  • Choose between normal, binomial and Poisson models, and combine them in multi-step questions.
Before you start3 quick questions on what this builds on
Check yourself
Events AA and BB are independent, with P(A)=0.45\mathrm{P}(A) = 0.45 and P(B)=0.6\mathrm{P}(B) = 0.6. Find P(A∩B)\mathrm{P}(A \cap B).
Check yourself
Y∼B(12,0.35)Y \sim \mathrm{B}(12, 0.35). Find P(Y≥3)\mathrm{P}(Y \ge 3). Give your answer to 3 s.f.
Check yourself
A WW has Var(W)=3\mathrm{Var}(W) = 3. Find Var(4W+7)\mathrm{Var}(4W + 7).

Revise first: 4.6 Probability diagrams, rules and independence, 4.7 Discrete random variables and expected value, 4.8 Binomial distribution, 4.9 Normal distribution and 4.14 Linear combinations and unbiased estimates

When a Poisson model fits

How many orders reach a website in an hour? How many potholes are on a kilometre of road? Each counts events in a fixed interval of time or space. There is no fixed number of trials, so the count has no upper limit.

Independent means one event doesn't make another more or less likely. Uniform average rate means the rate stays the same throughout the interval you care about: a café's lunchtime rush and quiet afternoon have different rates.

XX is a , and the single number mm fixes its whole distribution. It is the mean, and it is also the . Both results are in the formula booklet, and you don't need to prove them.

In the formula booklet
E(X)=m\mathrm{E}(X) = m
In the formula booklet
Var(X)=m\mathrm{Var}(X) = m
Worked exampleDuring its opening hours, an online bookshop receives orders at an average rate of 3.83.8 per hour. Let XX be the number of orders in a randomly chosen hour. (a) Write down the distribution of XX, and state two assumptions you have made, in context. (b) Find the standard deviation of XX.

Try it first, then reveal one step at a time.

  1. (a) XX counts events in one hour, with mean 3.83.8. So X∼Po(3.8)X \sim \mathrm{Po}(3.8).

With data, compare the sample mean with the of the variance, sn−12s_{n-1}^{2}, from 4.14. If they are close, a Poisson model is worth trying. If they are far apart, it probably isn't suitable.

Worked exampleA proofreader counts the typing errors on each of 5050 pages. Number of errors: 0,1,2,3,4,50, 1, 2, 3, 4, 5. Number of pages: 10,17,12,7,3,110, 17, 12, 7, 3, 1. (a) Find the mean and the unbiased estimate of the variance, to 3 s.f. (b) Explain why a Poisson model may be suitable.

Try it first, then reveal one step at a time.

  1. (a) Enter the errors as data and the pages as frequencies, then run one-variable statistics. Check that n=50n = 50.
Check yourself
X∼Po(m)X \sim \mathrm{Po}(m) has standard deviation 1.51.5. Find E(X)\mathrm{E}(X).
Check yourself
A café records how many customers walk in during each minute over several lunchtimes. The sample mean is 2.12.1 and the unbiased estimate of the variance is 6.86.8. Which comment is best?

Poisson probabilities on your GDC

Poisson pdf gives P(X=k)\mathrm{P}(X = k). Poisson cdf gives P(X≤k)\mathrm{P}(X \le k). Both need only the mean mm.

XX takes only whole numbers, so the rules from 4.8 still work: "fewer than 44" is X≤3X \le 3, and P(X≥k)=1−P(X≤k−1)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1).

Worked exampleA web server logs errors at an average rate of 2.72.7 per day, following a Poisson distribution. Find the probability of exactly 44 errors on a given day. Give your answer to 3 s.f.

Try it first, then reveal one step at a time.

  1. Let XX be the number of errors in a day, so X∼Po(2.7)X \sim \mathrm{Po}(2.7). You want P(X=4)\mathrm{P}(X = 4).
Bar chart of the Poisson distribution with mean 2.7, for 0 to 10 errors in a day. The bars rise to a peak at 2, then shrink towards zero but never stop. The bar at 4 is highlighted, and a dashed line marks the mean 2.7. 00.050.10.150.20.25 012345678910 … number of errors in a day, x mean m = 2.7 P(X = 4) P(X = x)
Po(2.7)\mathrm{Po}(2.7): the highlighted bar is P(X=4)\mathrm{P}(X = 4). The bars carry on past 1010, ever smaller.

Because the rate is uniform, the mean grows in proportion to the interval. A mean of mm per hour becomes mtmt in tt hours.

Worked exampleAmbulances arrive at a hospital at an average rate of 0.850.85 every 1010 minutes, following a Poisson distribution. (a) Find the probability that at least 55 arrive in one hour. Give your answer to 3 s.f. (b) Comment on one assumption of your model.

Try it first, then reveal one step at a time.

  1. (a) One hour is 6×106 \times 10 minutes, so the mean is 6×0.85=5.16 \times 0.85 = 5.1, and Y∼Po(5.1)Y \sim \mathrm{Po}(5.1).
Check yourself
X∼Po(2.2)X \sim \mathrm{Po}(2.2). Find P(X<3)\mathrm{P}(X < 3).
Check yourself
Potholes on a country road occur at an average rate of 0.450.45 per kilometre, following a Poisson distribution. A cycle route uses 88 km of the road. Find the probability that the route has at most 11 pothole. Give your answer to 3 s.f.

Adding independent Poisson counts

A help-desk inbox gets emails from staff and from customers. If each stream arrives independently at a steady rate, so does the combined stream, at the two rates added together.

This is also why the mean scales with the interval: two independent hours, each Po(m)\mathrm{Po}(m), add to Po(2m)\mathrm{Po}(2m).

Worked exampleA cycle shop repairs punctures at an average rate of 1.91.9 per day and brakes at 1.41.4 per day. Each follows a Poisson distribution, independently of the other. Find the probability of more than 55 repairs in total on one day. Give your answer to 3 s.f.

Try it first, then reveal one step at a time.

  1. Let P∼Po(1.9)P \sim \mathrm{Po}(1.9) and B∼Po(1.4)B \sim \mathrm{Po}(1.4). They are independent, so T=P+B∼Po(3.3)T = P + B \sim \mathrm{Po}(3.3).
Worked exampleA library takes back books at its desk at an average rate of 4.64.6 per hour and through its drop box at 2.52.5 per hour. Both follow Poisson distributions, independently. Find the probability that fewer than 1212 books are returned in total in 22 hours. Give your answer to 3 s.f.

Try it first, then reveal one step at a time.

  1. In one hour the total is Po(4.6+2.5)=Po(7.1)\mathrm{Po}(4.6 + 2.5) = \mathrm{Po}(7.1).
Check yourself
X∼Po(1.7)X \sim \mathrm{Po}(1.7) and Y∼Po(3.1)Y \sim \mathrm{Po}(3.1) are independent. Find P(X+Y=4)\mathrm{P}(X + Y = 4). Give your answer to 3 s.f.
Check yourself
A taxi firm has two phone lines. The numbers of calls to line A and to line B in an hour are independent, and each is Poisson. Which of these is also Poisson?

Normal, binomial or Poisson?

Exam questions often leave you to choose the model. Ask what the random variable is.

1. A continuous measurement, such as a mass or time, with a symmetric bell-shaped spread: a , from 4.9.

2. Successes in a fixed number nn of independent trials with the same probability pp: a . The count can't exceed nn.

3. Events in an interval of time or space, independent and at a uniform average rate, with no upper limit: a Poisson distribution.

Worked exampleChoose a normal, binomial or Poisson model for each, with a reason. (a) The mass of a bag of rice from a packing machine. (b) The number of seeds that germinate out of 3030, each germinating with probability 0.80.8, independently. (c) The number of bicycles passing a counter in 55 minutes on a weekday morning.

Try it first, then reveal one step at a time.

  1. (a) A mass is a continuous measurement that varies symmetrically about the target. Normal.

Some questions use two models in turn. A Poisson probability for one interval becomes the pp of a binomial across several independent intervals.

Worked exampleOn a boat trip, the number of dolphins seen follows a Poisson distribution with mean 1.31.3. A company runs 1010 independent trips in a week. Find the probability that at least one dolphin is seen on exactly 88 of the 1010 trips. Give your answer to 3 s.f.

Try it first, then reveal one step at a time.

  1. For one trip, D∼Po(1.3)D \sim \mathrm{Po}(1.3), and P(D≥1)=1−P(D=0)=0.727468…\mathrm{P}(D \ge 1) = 1 - \mathrm{P}(D = 0) = 0.727468\ldots
Check yourself
Each screw in a box of 5050 is faulty with probability 0.020.02, independently. Which model suits XX, the number of faulty screws in the box?
Check yourself
A ticket machine jams at an average rate of 0.60.6 times per day, following a Poisson distribution. Days are independent. Find the probability that in 77 days the machine jams on at most 22 of the days. Give your answer to 3 s.f.

Common mistakes

  • Using the mean for the wrong interval. Scale mm to the question's interval first, and write the new distribution down.
  • Writing P(X≥k)=1−P(X≤k)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k), which drops X=kX = k. Use 1−P(X≤k−1)1 - \mathrm{P}(X \le k - 1).
  • Saying "the mean and variance differ" to explain why a situation may not be Poisson when no data are given. Give a reason from the context: events that cluster, or a rate that changes.
  • Rounding a Poisson probability before using it as pp in a binomial. Store the full value and round only the final answer.

Exam tips

  • Find a Poisson probability: write X∼Po(m)X \sim \mathrm{Po}(m) for the right interval and the probability statement, then the GDC value to 3 s.f.
  • Explain why a Poisson model suits a situation: independence or a uniform average rate, in context. From data: the sample mean and variance are about equal.
  • A switch from "in an interval" to "on how many of the days" usually means a binomial, with the Poisson probability as pp.

Key takeaways

Learn thisX∼Po(m)X \sim \mathrm{Po}(m) models the number of events in a fixed interval of time or space when the events happen independently at a uniform average rate of mm per interval. XX can be 0,1,2,…0, 1, 2, \ldots with no upper limit.
In bookletE(X)=m\mathrm{E}(X) = m
In bookletVar(X)=m\mathrm{Var}(X) = m
Learn thisBecause the rate is uniform, a Poisson mean of mm per unit interval becomes mtmt for an interval tt times as long.
Learn thisIf X∼Po(m1)X \sim \mathrm{Po}(m_{1}) and Y∼Po(m2)Y \sim \mathrm{Po}(m_{2}) are independent, then X+Y∼Po(m1+m2)X + Y \sim \mathrm{Po}(m_{1} + m_{2}). Multiples such as 2X2X and differences such as X−YX - Y are not Poisson.
Learn thisChoosing a model: a continuous measurement with a symmetric bell-shaped spread is normal; successes in a fixed number of independent trials with constant pp are binomial; independent events in an interval at a uniform average rate, with no upper limit, are Poisson.

Practise this

19 past-paper questions on Poisson distribution in the AI HL bank, each with a mark scheme and a worked explanation.

Practise Poisson distribution →