4.8 · SLPaper 1Paper 2

Binomial distribution

In a minute

A counts the successes in a fixed number of , each with the same chance of success. You write X∼B(n,p)X \sim \mathrm{B}(n, p), find its probabilities on your GDC, and use E(X)=np\mathrm{E}(X) = np for the number of successes to expect.

You'll be able to

  • Decide whether a situation fits a , and write it as X∼B(n,p)X \sim \mathrm{B}(n, p).
  • Find P(X=r)\mathrm{P}(X = r) by hand for small cases and with your GDC for the rest.
  • Turn "at most", "at least", "fewer than", "more than" and "between" into the right GDC calculation.
  • Find the mean, and of X∼B(n,p)X \sim \mathrm{B}(n, p), and work back from them to nn and pp.
  • Use npnp as the expected number of successes, the same idea as the expected number of occurrences.
Before you start2 quick questions on what this builds on
Check yourself
Find 6C2{}^{6}C_{2}.
Check yourself
A YY takes the values 00, 11 and 22 with probabilities 0.50.5, 0.30.3 and 0.20.2. Find E(Y)\mathrm{E}(Y).

Revise first: 1.9 Binomial theorem and 4.7 Discrete random variables and expected value

When is a situation binomial?

Plant 1010 seeds. Each one grows with probability 0.60.6, whatever the others do. The number that grow could be anything from 00 to 1010. A binomial distribution gives the probability of each.

Each seed is a with two results. The result you count is a success (it grows); the other is a failure. "Success" is only a label: it could be a faulty part.

Independent means the result of one trial doesn't change the chances for any other. You met independent events in 4.6. Read X∼B(n,p)X \sim \mathrm{B}(n, p) as "XX is binomial with nn trials and success probability pp".

Worked exampleA fair six-sided die is rolled 2020 times. XX is the number of rolls that show a 11 or a 22. Explain why XX can be modelled by a binomial distribution, and state the values of nn and pp.

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

  1. Each roll is a success (a 11 or a 22) or a failure, and the number of rolls is fixed: n=20n = 20.
Worked exampleA bag holds 77 green and 33 yellow counters. Four counters are taken out one at a time, and XX is the number of green ones. Is XX binomial when (a) each counter is put back before the next is taken; (b) the counters are not put back? If it is, state nn and pp.

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

  1. (a) Every draw is from the same 1010 counters, so P(green)=710=0.7\mathrm{P}(\text{green}) = \frac{7}{10} = 0.7 each time. The 44 draws are independent, so X∼B(4,0.7)X \sim \mathrm{B}(4, 0.7).
Check yourself
A card is drawn from a standard pack of 5252 cards, noted and put back. This is done 1010 times. XX is the number of hearts drawn, and X∼B(10,p)X \sim \mathrm{B}(10, p). Find pp.
Check yourself
Eight sweets are taken one at a time, without replacement, from a jar of 1212 lemon and 1818 orange sweets. XX is the number of lemon sweets taken. Which binomial condition fails?

The probability of exactly r successes

Before you use a GDC, see where binomial probabilities come from. Build one small case by hand.

Worked exampleA spinner lands on red with probability 0.30.3. It is spun 44 times. Find the probability that it lands on red exactly twice.

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

  1. Let XX be the number of reds, so X∼B(4,0.3)X \sim \mathrm{B}(4, 0.3). Write R for red and N for not red, with P(N)=1−0.3=0.7\mathrm{P}(\text{N}) = 1 - 0.3 = 0.7.

The same reasoning works for any nn and rr. Each order with rr successes and n−rn - r failures has probability pr(1−p)n−rp^{r}(1-p)^{n-r}. There are nCr{}^{n}C_{r} such orders: the from 1.9.

Learn this
P(X=r)=nCr pr(1−p)n−r,r=0,1,…,n\mathrm{P}(X = r) = {}^{n}C_{r}\, p^{r}(1-p)^{n-r}, \quad r = 0, 1, \ldots, n

This formula is not in the formula booklet. Most binomial probabilities are found on a GDC, but the formula is quick for small cases. Two special cases: P(X=0)=(1−p)n\mathrm{P}(X = 0) = (1-p)^{n} and P(X=n)=pn\mathrm{P}(X = n) = p^{n}.

Worked exampleSeeds of a certain plant grow with probability 0.850.85, independently of each other. Rosa plants 1818 seeds. Use your calculator to find the probability that exactly 1515 of them grow. Give your answer to 3 s.f.

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

  1. Let XX be the number of seeds that grow. There are 1818 independent trials with p=0.85p = 0.85, so X∼B(18,0.85)X \sim \mathrm{B}(18, 0.85).
Check yourself
X∼B(4,14)X \sim \mathrm{B}\left(4, \frac{1}{4}\right). Find P(X=3)\mathrm{P}(X = 3) exactly.
Check yourself
In a town, 22%22\% of households own a dog. A random sample of 1616 households is chosen. Find the probability that exactly 44 of them own a dog. Give your answer to 3 s.f.

At most, at least and between

Most questions ask about a range of values, such as "at most 33" or "at least 66". Your GDC's binomial cdf adds P(X=0)\mathrm{P}(X = 0) up to P(X=k)\mathrm{P}(X = k) for you. So it gives P(X≤k)\mathrm{P}(X \le k) in one step.

Everything else comes from P(X≤k)\mathrm{P}(X \le k) and the rule. XX only takes whole numbers, so "fewer than 44" means X≤3X \le 3, and "more than 44" means X≥5X \ge 5.

Learn this
at most kP(X≤k)fewer than kP(X≤k−1)at least kP(X≥k)=1−P(X≤k−1)more than kP(X>k)=1−P(X≤k)a to b inclusiveP(a≤X≤b)=P(X≤b)−P(X≤a−1)\begin{array}{ll} \text{at most } k & \mathrm{P}(X \le k) \\ \text{fewer than } k & \mathrm{P}(X \le k - 1) \\ \text{at least } k & \mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) \\ \text{more than } k & \mathrm{P}(X > k) = 1 - \mathrm{P}(X \le k) \\ a \text{ to } b \text{ inclusive} & \mathrm{P}(a \le X \le b) = \mathrm{P}(X \le b) - \mathrm{P}(X \le a - 1) \end{array}
Worked exampleOn any school day, the probability that Ella's bus is late is 0.20.2, of other days. Find the probability that the bus is late at least once in 33 days.

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

  1. Let XX be the number of late days, so X∼B(3,0.2)X \sim \mathrm{B}(3, 0.2).

Move kk and switch the event in the chart below. Watch which bars are shaded. "At least 66" starts at the bar for 66, so its complement stops at 55.

Binomial probabilitiesDrag the slider
0.000.100.200.3002468101214X ~ B(15, 0.25)shaded: P(X ≥ 6)
15
0.25

6

Probability = 0.1484 (P(X ≥ 6))

Mean np = 3.75, variance np(1 − p) = 2.81

Each bar is P(X = k) = C(n, k) pᵏ (1 − p)ⁿ⁻ᵏ. The shaded bars make up the event, and adding their heights gives its probability.
Worked exampleA quiz has 1515 multiple-choice questions, each with 44 options and one correct answer. Tom guesses every answer at random. Use your calculator to find the probability that he gets at least 66 correct. Give your answer to 3 s.f.

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

  1. Let XX be the number of correct answers. Each guess is right with probability 14=0.25\frac{1}{4} = 0.25, independently. So X∼B(15,0.25)X \sim \mathrm{B}(15, 0.25).
Check yourself
X∼B(10,0.3)X \sim \mathrm{B}(10, 0.3). Find P(X<4)\mathrm{P}(X < 4).
Check yourself
An archer hits the gold ring with probability 0.650.65 on each arrow, independently. She shoots 2020 arrows. Find the probability that she hits the gold between 1010 and 1515 times inclusive. Give your answer to 3 s.f.
Check yourself
A darts player hits the bullseye with probability 0.150.15 on each throw, independently. She throws 1010 darts. Given that she hits at least one bullseye, find the probability that she hits exactly one. Give your answer to 3 s.f.
Check yourself
Each bulb in a string of party lights is faulty with probability 0.060.06, independently of the others. Find the least number of bulbs nn for which the probability that at least one bulb is faulty is greater than 0.950.95.

Mean, variance and expected number

Roll a fair die 6060 times. In 4.5 you expected 60×16=1060 \times \frac{1}{6} = 10 sixes. The number of sixes is X∼B(60,16)X \sim \mathrm{B}\left(60, \frac{1}{6}\right), and that expected number is just npnp.

In the formula booklet
E(X)=np\mathrm{E}(X) = np

E(X)\mathrm{E}(X) is the mean of XX: the average number of successes over many repeats of all nn trials. Like any expected number, it need not be a whole number.

The measures how widely XX spreads around its mean. A binomial distribution has its own formula for it. Its square root is the , σ\sigma.

In the formula booklet
Var(X)=np(1−p)\mathrm{Var}(X) = np(1-p)

Both formulas are in the formula booklet, and you don't need to prove them. As 1−p<11 - p < 1, the variance is always smaller than the mean: a quick check.

Worked exampleX∼B(50,0.12)X \sim \mathrm{B}(50, 0.12). Find the mean and the variance of XX.

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

  1. Here n=50n = 50 and p=0.12p = 0.12, so 1−p=0.881 - p = 0.88.
Worked exampleA factory finds that 4%4\% of its phone screens have a defect. A batch of 250250 screens is checked. (a) Find the expected number of screens with a defect. (b) Find the probability that more than the expected number have a defect. Give your answer to 3 s.f.

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

  1. Let XX be the number of screens with a defect. Assume each screen is faulty with probability 0.040.04, . Then X∼B(250,0.04)X \sim \mathrm{B}(250, 0.04).
Worked exampleX∼B(n,p)X \sim \mathrm{B}(n, p), with E(X)=12\mathrm{E}(X) = 12 and Var(X)=8.4\mathrm{Var}(X) = 8.4. Find nn and pp.

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

  1. Write both facts as equations: np=12np = 12 and np(1−p)=8.4np(1-p) = 8.4.
Check yourself
X∼B(80,0.25)X \sim \mathrm{B}(80, 0.25). Find Var(X)\mathrm{Var}(X).
Check yourself
The number of sixes in nn rolls of a fair die is X∼B(n,16)X \sim \mathrm{B}\left(n, \frac{1}{6}\right). The expected number of sixes is 1414. Find nn.

Common mistakes

  • Writing P(X≥k)=1−P(X≤k)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k). That removes X=kX = k as well. Use 1−P(X≤k−1)1 - \mathrm{P}(X \le k - 1).
  • Reading "more than 66" as X≥6X \ge 6. For a whole-number XX, X>6X > 6 means X≥7X \ge 7.
  • Leaving out nCr{}^{n}C_{r} in a binomial probability by hand. Without it you have found the probability of one order only.
  • Using the binomial for "the first success is on trial kk". That is one fixed order, k−1k - 1 failures then a success, so its probability is (1−p)k−1p(1-p)^{k-1}p.
  • Swapping pp and 1−p1 - p. Decide what a success is and write down its probability pp first.
  • Rounding a GDC value, such as P(X≤10)\mathrm{P}(X \le 10), before using it again. Carry the full value and round only the final answer.

Exam tips

  • Find a binomial probability: write the distribution, X∼B(n,p)X \sim \mathrm{B}(n, p), and the probability statement, such as P(X≥6)=1−P(X≤5)\mathrm{P}(X \ge 6) = 1 - \mathrm{P}(X \le 5), before the GDC value.
  • Paper 1 has no GDC, so binomial work there uses the formulas: npnp, np(1−p)np(1-p), or a small case such as P(X=0)=(1−p)n\mathrm{P}(X = 0) = (1-p)^{n}.
  • Give probabilities to 3 s.f. unless the question says otherwise, and never round a value you will use again.
  • Find the expected number: write npnp with the numbers in, then the value. Don't round it to a whole number.

Key takeaways

Learn thisP(X=r)=nCr pr(1−p)n−r,r=0,1,…,n\mathrm{P}(X = r) = {}^{n}C_{r}\, p^{r}(1-p)^{n-r}, \quad r = 0, 1, \ldots, n
Learn thisat most kP(X≤k)fewer than kP(X≤k−1)at least kP(X≥k)=1−P(X≤k−1)more than kP(X>k)=1−P(X≤k)a to b inclusiveP(a≤X≤b)=P(X≤b)−P(X≤a−1)\begin{array}{ll} \text{at most } k & \mathrm{P}(X \le k) \\ \text{fewer than } k & \mathrm{P}(X \le k - 1) \\ \text{at least } k & \mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) \\ \text{more than } k & \mathrm{P}(X > k) = 1 - \mathrm{P}(X \le k) \\ a \text{ to } b \text{ inclusive} & \mathrm{P}(a \le X \le b) = \mathrm{P}(X \le b) - \mathrm{P}(X \le a - 1) \end{array}
In bookletE(X)=np\mathrm{E}(X) = np
In bookletVar(X)=np(1−p)\mathrm{Var}(X) = np(1-p)
Learn thisX∼B(n,p)X \sim \mathrm{B}(n, p) counts successes in a fixed number nn of independent trials, each with two outcomes and the same success probability pp.
Learn thisOn your GDC, binomial pdf gives P(X=r)\mathrm{P}(X = r) and binomial cdf gives P(X≤k)\mathrm{P}(X \le k).
Learn thisP(X≥k)=1−P(X≤k−1)\mathrm{P}(X \ge k) = 1 - \mathrm{P}(X \le k - 1) and P(X>k)=1−P(X≤k)\mathrm{P}(X > k) = 1 - \mathrm{P}(X \le k).
Learn thisnpnp is the expected number of successes. The standard deviation is np(1−p)\sqrt{np(1-p)}.

Practise this

18 past-paper questions on 4.8 in the AA HL bank, each with a mark scheme and a worked explanation.

Practise 4.8 →