HOSP 1207 (Business Stats) Learning Centre

The Binomial Distribution

In a statistical experiment, a series of measurements are obtained. Each measurement

is a random outcome that could take any value x, and is called a random variable. A
random variable can be either discrete (countable number of values or whole numbers)
or continuous (any number on a continuous scale).

An example of a discrete random variable is the number of guests who paid with a
credit card at a hotel. An example of a continuous random variable is the amount of time
guests spend at a table in a restaurant for a meal.

The collection of all the possible outcomes of an experiment and the probability of those
outcomes is called a probability distribution. For discrete random variables, each
outcome (e.g. 7 customers); can be assigned a probability for continuous random
variables, probabilities can only be assigned to an interval or range of values (e.g. 1.5 –
2.5 hours).

The probability distribution of a discrete random variable has two characteristics:

(1) EACH distinct value of the random variable has an assigned probability, and
(2) all the assigned probabilities for the random variables must add up to one.

Example: Mr. Brown developed a test to measure students’ ability to pay attention
during a lecture on stats. He tested a group of 460 students. The possible scores were
0, 1, 2, 3, 4, 5 with 0 indicating the lowest ability to pay attention. The test results are
shown below. We read this table as: “the probability that a randomly chosen student will
have a score of 2 is 0.24 or 24”.

Score # of Probability Notice that each score value (x) has a distinct
(x) Students P(x)
probability and the sum of all the probabilities
0 5 0.01
is 1.
1 32 0.07
2 110 0.24 If we wanted to know the probability of a
3 240 0.52 randomly chosen subject scoring either a 4 or
4 60 0.13 5 on the test, we add the probabilities:
5 13 0.03 P(4 or 5) = P(4) + P(5) = 0.13 + 0.03 = 0.16

To find the mean (or expected value) and standard deviation of a population that is
made up of discrete random variables, use the following formulas:
μ (population mean) = Σ [x · P(x)]
σ (standard deviation) = Σ [(x − μ)² · P(x)]

© 2013 Vancouver Community College Learning Centre. AuthoredbybyEmily

Darren Rigby
Simpson
Student review only. May not be reproduced for classes. & Emily Simpson
The binomial experiment is a particular type of discrete random variable experiment.
The outcomes of a binomial experiment form a binomial distribution.
Conditions of a binomial experiment:
• The experiment consists of a fixed number of independent trials, n.
• Each trial has only two possible outcomes, success (S) and a failure (F).
• the probability of a success in every trial is the same, called p. The probability of
failure is called q, and p + q = 1.
Let’s look at an experiment to better understand these conditions. Carnival-goers spin a
wheel with 24 spaces, four of which are marked with a star. A player gets three spins
and wins a prize if all three spins land on a star. This experiment is a series of trials:
repeating the same action to see what happens.
Condition 1: The number of trials stays the same and each trial must be independent of
the others. If two people play the game, they each get exactly three spins, and the result
of one spin does not affect the result of another.
Condition 2: Each trial has only two possible outcomes – success or failure. In the
experiment, a space with a star is a success and a space without a star is a failure.
Condition 3: The probability of success in each trial is the same. The probability of
stopping on a star is 4⁄24 = 1⁄6 so the probability of a success (p) is always the same.
Similarly the probability of failure q, is always the same since q = 1 – p = 1 – 1/6 = 5/6.
A binomial experiment is technically sampling with replacement, since each trial is
independent of the next trial.

What about a case where this is not true? For example, a bag of stones is filled with 5
green and 3 red stones, and you draw three stones out sequentially. The probability of
getting a red (or green) stone changes with each draw. On the first draw, your
probability of a red stone is 3/8. If you get a red stone, the probability of a red stone on
the second draw is 2/7 and so on. This is called sampling without replacement. Since
the probability of success changes between trials, it does not fulfil the condition of
independent trials and so is not a binomial experiment.

If we know an experiment can be described as a binomial distribution, we can calculate

the probability of exactly “r” successes out of “n” trials using the following formula:
!
,
! !
The Cn,r function is on your calculator. If you don’t know how to use it, see the tutor.

Example: Consider the carnival game. Find the probability that:

(a) exactly 1 star occurs in 3 spins.
(b) there are either 2 stars or 3 stars in 5 spins.
Solution: Let r represent the number of stars (successes) we want.
(a) Here, n = 3 and p = 1⁄6, and we want P(r = 1).

© 2013 Vancouver Community College Learning Centre.

Student review only. May not be reproduced for classes. 2
 !
1
! !

= 25⁄78 ≈ 0.32
(b) Here, n = 5 and p = 1⁄6. We want P(r = 2 or 3). Since the two events in the
question are disjoint — we can’t get exactly 2 stars and exactly 3 stars at the same time
— we can calculate their probabilities separately and add them.
P(r = 2 or 3) = P(r = 2) + P(r = 3)
! !
=
! ! ! !

= 125⁄648 ≈ 0.19
Over weeks of watching people spin the wheel at the carnival, the person running the
game will know about how many stars each person is likely to spin, and how far from
this average people are likely to get. We can calculate the mean and the spread for the
binomial distribution using two formulas that are easier than those shown on page 1.
If x is a binomially distributed variable, then μ = np and σ = np(1 − p) .
For the carnival game, x is the number of stars a player spins. For x, n = 3 and p = 1⁄6,
so μ is 3 · 1⁄6 = 0.5 and σ = 3 ⋅ 16 ⋅ 5 6 = 0.6455.

EXERCISES
A. Determine whether each of these statistical experiments can be modelled by the
binomial distribution. If they can, define the statistical variable X and determine the
values of n and p for the distribution. If they cannot, state which of the three conditions
from the first page on this handout is broken .
1) A tutor in the Learning Centre wants to know whether the way a student faces is an
indicator of whether that student will come for help. In one week the tutor tracks one
randomly chosen student per day who sits facing the tutor desk, and finds that over a
series of one-week experiments, those students ask for help 60% of the time.
2) You grab five jelly beans out of a jar, hoping to get a red one. On the manufacturer’s
website, it says that 23% of the jelly beans it sells in every package are red.
3) A child learns to hit a baseball by practicing. At the beginning of the day, he can’t hit
at all, but by the end, he has a batting average of 0.100. An experiment consists of 3
consecutive pitches from this day.
4) A coin is weighted so that it is unfair. It comes up heads 98% of the time rather than
50% of the time. You challenge a friend to a bet saying, “I’ll flip this coin, best two out of
three. Heads I win, tails I lose.”
B. For each of the situations in Part A that was modeled by the binomial distribution,
calculate the mean and standard deviation for the distribution.

© 2013 Vancouver Community College Learning Centre.

Student review only. May not be reproduced for classes. 3
C. For the distributions below, calculate P(r), µ and σ.
1) n = 3, p = 0.75, r = 2 3) n = 20, p = 0.05, r = 1
2) n = 7, p = 0.24, r = 3 4) n = 8, p = 0.125, r = 0

D. The outcomes of all the experiments below are binomially distributed.

1. A quality control technician examines cartons of a dozen eggs for broken eggs.
She finds that the probability of finding a broken egg is 0.0075. What is the probability
that a given carton has two broken eggs in it?
2. A yearlong survey shows that when a waiter works Friday nights, the probability
of receiving at least a 15% tip on a cheque is 0.65. If a waiter staffs 20 tables during a
Friday night shift, what is the probability of receiving less than a 15% tip on five of the
tables?
3. Two percent of the world's population have naturally red hair - caused by a
mutation in the gene mc1r. Suppose we choose 15 people at random from the
population and record their hair color. (a) What is the probability that three of these
people have red hair? (b) In a sample of 20,000 people, how many people are expected
to have red hair?
4. A seasoned darts player throws three darts at the target. He hits the “20”
section of the dartboard 75% of the time. (a) Find the probability that he hits the “20”
section at least once in 3 darts. (b) In a game of throwing 6 darts, what is the expected
number of darts that land in the “20” section?

SOLUTIONS
A. (1) yes; X = the number of students who ask for help; n = 5, p = 0.60 (2) no; there is
drawing without replacement, so the probability of success changes. 3. no; the
probability changes, and the trials are not independent because the child is learning and
improving as he practices. 4. yes; X = the number of heads; n = 3, p = 0.98. The coin
is unfair, but it’s consistently unfair.
B. 1. μ = 3, σ = 1.1 4. μ = 2.94, σ = 0.24
C. 1. P(2) = 0.4219 µ = 2.25, σ = 0.75 2. P(3) = 0.1614 µ = 1.68, σ = 1.13
3. P(1) = 0.3774 µ = 1, σ = 0.9747 4. P(0) = 0.3436 µ = 1, σ = 0.9354
D. 1. n = 12, p = 0.0075, P(2) = 0.0034
2. n = 20, p = 0.35, P(5) = 0.1272
3. (a) n = 15, p = 0.02, P(3) = 0.0029 (b) µ = np = (20,000)(0.02) = 400 people
4. (a) n = 3, p = 0.75, P(r ≥ 1) = 1 − P(r = 0) = 1 − 0.016= 0.984
(b) µ = np = (6)(0.75) = 4.5 darts

© 2013 Vancouver Community College Learning Centre.

Student review only. May not be reproduced for classes. 4