4.8 The binomial distribution

4.8 The binomial distribution

IB Maths: Analysis and Approaches HL: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • X∼B(n,p)X\sim B(n,p) needs a fixed nn, two outcomes, a constant pp and independent trials.
  • Use the GDC's binomial pdf for P(X=x)P(X=x) and binomial cdf for P(X≤x)P(X\le x).
  • 'At least kk' means 1−P(X≤k−1)1-P(X\le k-1).
  • E(X)=np\text{E}(X)=np, Var(X)=np(1−p)\text{Var}(X)=np(1-p), σ=np(1−p)\sigma=\sqrt{np(1-p)}.
  • For an unknown nn, use the complement with logarithms, or a GDC table with two consecutive values.

When is a binomial model appropriate?

When XX counts successes in a fixed number of independent trials, each with the same probability of success.

XX follows a binomial distribution, X∼B(n,p)X\sim B(n,p), when all four conditions hold: a fixed number of trials nn; two outcomes, success or failure; the same probability of success pp on every trial; and independent trials.

For example, the correct answers when 12 four-option questions are guessed is B(12,0.25)B(12,0.25). Taking 5 cards from a pack without replacement and counting hearts is not binomial, because pp changes.

Always define the variable and state its distribution. When criticising a model, explain in context why independence or a constant pp might fail.

00.050.10.150.20.250123456Number of correct answersProbability
The distribution B(12, 0.25), values 0 to 6 (the mean is 3)

Which of these is NOT a binomial situation?

Finding binomial probabilities

Use the calculator: pdf for P(X=x)P(X=x), cdf for P(X≤x)P(X\le x), and translate the words carefully.

On a GDC, binomial pdf gives P(X=x)P(X=x) and binomial cdf gives P(X≤x)P(X\le x). The formula P(X=r)=(nr)pr(1−p)n−rP(X=r)=\binom{n}{r}p^r(1-p)^{n-r} is in the formula booklet, but in an exam you use the calculator.

XX takes whole-number values only. 'Between 3 and 7 inclusive' is P(X≤7)−P(X≤2)P(X\le7)-P(X\le2).

Write down the probability statement, such as P(X≥12)=1−P(X≤11)P(X\ge12)=1-P(X\le11), before the number. The method mark is for this line.

00.050.10.150.20.250.0010.00980.00980.001012345678910xP(X = x)
X ~ B(10, 0.5): P(X ≤ 4) is the total of the bars for 0 to 4. The calculator gives this as cdf.

At most 5

  • P(X≤5)P(X\le5)

Fewer than 5

  • P(X≤4)P(X\le4)

At least 5

  • 1−P(X≤4)1-P(X\le4)

More than 5

  • 1−P(X≤5)1-P(X\le5)

Worked example

X∼B(10,0.5)X\sim B(10,0.5). Find P(X≥7)P(X\ge7).

Which expression is 'fewer than 5'?

Mean and variance

The mean is npnp and the variance is np(1−p)np(1-p); both are in the formula booklet.

The standard deviation is np(1−p)\sqrt{np(1-p)}. You do not need to prove these results.

The mean is the long-run average number of successes, so it need not be a whole number: 186 ticket holders turning up with probability 0.95 gives a mean of 176.7176.7.

The most likely value (the mode) is the xx with the largest P(X=x)P(X=x); find it from a GDC table. It is usually close to the mean.

For X∼B(20,0.85)X\sim B(20,0.85): E(X)=17\text{E}(X)=17, Var(X)=2.55\text{Var}(X)=2.55, σ=1.60\sigma=1.60.

00.050.10.150.20.250.0046121314151617181920xP(X = x)
X ~ B(20, 0.85): mean np = 17, and the most likely value (17) sits at the peak. Values for x = 12 to 20.
  • MeanE(X)=np\text{E}(X)=np
  • VarianceVar(X)=np(1−p)\text{Var}(X)=np(1-p)

X∼B(40,0.3)X\sim B(40,0.3). What is Var(X)\text{Var}(X)?

Problems where n is unknown

Use the complement with logarithms, or a GDC table showing two consecutive values of nn.

Some questions fix a probability and ask for the least or greatest nn.

Complement. 'At least one miss' is the complement of 'no misses', so P(at least one miss)=1−pnP(\text{at least one miss})=1-p^n.

Trial and improvement. When the condition involves a cumulative probability, such as P(X>180)<0.05P(X>180)<0.05 with X∼B(n,0.95)X\sim B(n,0.95), use a GDC table in nn. Always show two consecutive values, for example n=185n=185 gives 0.04330.0433 and n=186n=186 gives 0.09290.0929. One value alone does not show the boundary.

5101520253035400.20.40.60.81xy0.95n = 291 − 0.9ⁿ
Illustrative: with p = 0.9, P(at least one miss) = 1 − 0.9ⁿ first exceeds 0.95 at n = 29.

Complement and logarithms

  • P(at least one)=1−pnP(\text{at least one})=1-p^n
  • Solve the inequality for n

Trial and improvement

  • Cumulative probability condition
  • GDC table: two consecutive values

Worked example

Find the least nn for which 1−(0.78)n>0.991-(0.78)^n>0.99.

Find the least nn for which 1−(0.9)n>0.951-(0.9)^n>0.95.

Combining binomial models

A probability from one model can be the success probability of a second model.

If a flight is overbooked with probability 0.04330.0433 on any day, and days are independent, then the number of overbooked days in a week is Y∼B(7,0.0433)Y\sim B(7,0.0433). So P(Y≥1)=1−(1−0.0433)7=0.266P(Y\ge1)=1-(1-0.0433)^7=0.266.

Keep full calculator accuracy for the first probability, and round only the final answer. Store intermediate probabilities in the calculator memory.

  1. 1

    First model

    find a probability, e.g. overbooked on one day

  2. 2

    Use it as p

    Y∼B(7,0.0433)Y\sim B(7,0.0433) for a week

  3. 3

    Answer the question

    round only at the end

A two-stage model

Y∼B(5,0.1)Y\sim B(5,0.1). Find P(Y≥1)P(Y\ge1) to 3 significant figures.

Try an exam question

A fair coin is tossed 10 times. Let XX be the number of heads. (a) State the distribution of XX. (b) Find P(X≥7)P(X\ge7). (c) Find Var(X)\text{Var}(X).

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Find E(X)\mathrm{E}(X).

1 mark, marked against the mark scheme

Revision notes