The binomial distribution

The binomial distribution

Spec 4.1 · Edexcel A-Level Maths: 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 number of trials, two outcomes, constant pp and independent trials.
  • P(X=r)=(nr)pr(1−p)n−rP(X=r)=\binom nr p^r(1-p)^{n-r}.
  • Use the cumulative function for P(X≤r)P(X\le r): P(X≥r)=1−P(X≤r−1)P(X\ge r)=1-P(X\le r-1) and P(X>r)=1−P(X≤r)P(X>r)=1-P(X\le r).
  • “At least one” is 1−(1−p)n1-(1-p)^n.
  • Justify a model by naming the conditions in context, and say when they may fail.

The binomial model

When there is a fixed number of independent trials, each with two outcomes and a constant probability.

A binomial distribution X∼B(n,p)X\sim B(n,p) models the number of successes in nn trials when the four conditions hold.

Examples: correct guesses in a 10-question test with four options (p=0.25p=0.25), or defective bolts in a sample of 20 (p=0.05p=0.05). State the distribution with its parameters and justify it by naming the conditions.

  1. 1

    Fixed number

    There is a fixed number n of trials.

  2. 2

    Two outcomes

    Each trial is a success or a failure.

  3. 3

    Constant p

    The probability of success is constant.

  4. 4

    Independent

    The trials are independent.

Conditions for a binomial model

A test has 10 questions, each with 4 options, all guessed. Which distribution models the number correct?

Individual probabilities

Use P(X=r)=(nr)pr(1−p)n−rP(X=r)=\binom nr p^r(1-p)^{n-r}, where (nr)\binom nr counts the ways of choosing the successes.

Use the calculator’s binomial probability function for P(X=r)P(X=r), and check you have not left out the (nr)\binom nr factor.

The chart shows B(10,0.25)B(10,0.25): the most likely value is r=2r=2 and the probabilities fall away quickly.

00.050.10.150.20.250.30.00310.0004012345678910rP(X = r)
Probabilities for B(10, 0.25) (to 4 d.p.).
  • P(X=r)P(X=r)(nr)pr(1−p)n−r\binom nr p^r(1-p)^{n-r}

Worked example

X∼B(10,0.25)X\sim B(10,0.25). Find P(X=3)P(X=3).

X∼B(5,0.5)X\sim B(5,0.5). What is P(X=2)P(X=2)?

Cumulative probabilities

Convert to P(X≤r)P(X\le r), remembering XX takes whole numbers only.

The calculator’s cumulative probability function gives P(X≤r)P(X\le r). “At least one” is quickest as 1−P(X=0)=1−(1−p)n1-P(X=0)=1-(1-p)^n.

00.050.10.150.20.250.0090.00140.0001012345678910rP(X = r)
P(X = r) for X ~ B(10, 0.3): add bars from the left for P(X ≤ r)

Want

  • P(X<r)P(X<r)
  • P(X≥r)P(X\ge r)
  • P(X>r)P(X>r)
  • P(a≤X≤b)P(a\le X\le b)

Write as

  • P(X≤r−1)P(X\le r-1)
  • 1−P(X≤r−1)1-P(X\le r-1)
  • 1−P(X≤r)1-P(X\le r)
  • P(X≤b)−P(X≤a−1)P(X\le b)-P(X\le a-1)

Worked example

X∼B(10,0.25)X\sim B(10,0.25). Find P(X≥4)P(X\ge4).

Which is equal to P(X>6)P(X>6) for X∼B(10,0.25)X\sim B(10,0.25)?

Two-stage and inequality problems

Sometimes a binomial probability becomes the pp of a second binomial; for the least nn, use logarithms.

Dividing by a negative number (such as log⁡0.4\log0.4) reverses the inequality sign. Also find the largest kk with P(X≥k)>0.9P(X\ge k)>0.9 by evaluating cumulative probabilities for neighbouring values.

Check an inequality answer by testing nn and n−1n-1: 0.45=0.01020.4^5=0.0102 is too big, 0.46=0.00410.4^6=0.0041 works.

123456789100.20.40.60.81xyn = 5: 0.9898n = 6: 0.99591 − 0.4ⁿ
P(at least one success) = 1 − 0.4ⁿ against n, with p = 0.6: it first exceeds 0.99 at n = 6
  • At least one1−(1−p)n1-(1-p)^n

Worked example

A sample is accepted if X≤2X\le2, with q=P(X≤2)=0.9245q=P(X\le2)=0.9245. Find the probability that exactly 4 of 5 samples are accepted.

Worked example

Find the least nn for which 1−0.4n>0.991-0.4^n>0.99.

Find the least nn with 1−0.5n>0.951-0.5^n>0.95.

Validity of the model

When trials share a cause or the probability varies, answers are unreliable.

Binomial answers are only reliable if the conditions hold. Comment in context: ‘independence is unlikely because …’ and say what difference it would make.

  1. 1

    Fixed number of trials n

    decided in advance

  2. 2

    Two outcomes

    success or failure

  3. 3

    Independent trials

    one result does not affect the next

  4. 4

    Constant probability p

    the same on every trial

The four conditions a binomial model needs

Independence

How it may fail:
Trials share a cause
Example:
Students in one class, consecutive throws affected by fatigue

Constant p

How it may fail:
Probability varies between trials
Example:
Changes with time or conditions

Why might a binomial model fail for the number of students in one class who pass a test?

Try an exam question

X∼B(12,0.3)X\sim B(12,0.3). Find P(X=4)P(X=4) and P(X≥3)P(X\ge3).

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

Find P(X=3)P(X=3).

1 mark, marked against the mark scheme

Revision notes