Discrete Probability Distributions

A discrete probability distribution gives the probability of every possible value of a discrete
random variable.

1. Binomial Probability Distribution

2. Hyper geometric Probability Distribution
3. Poisson Distribution
4. Geometric Distribution
5. Multinomial Distribution

Binomial Probability Distribution

Binomial Experiment: The experiment which consists of independent trials, each trial having
only two possible outcomes. If the probability of each outcome remains the same throughout
the trials, then such trials are called Bernoulli trials and the experiment having n Bernoulli
trials is called a binomial experiment.

In other words; an experiment is called a binomial, if it possesses the following properties:

1. The outcome of each trial may be classified into one of two categories: Success (S) or
Failure (F)
2. The probability of Success, denoted by P, remains constant for all trials.
3. The successive trials are all independent.
4. The experiment is repeated a fixed number of times, say n.

When X denotes the number of Successes in n trials of binomial experiment, it is called a

binomial random variable and probability distribution is called the Binomial Probability
Distribution.

When the binomial random variable X assumes a value x , the binomial probability distribution
is given by:

 n
f ( x)  P( X  x)    P x q n x , x = 0,1,2,3 … , n
 x
Where q = 1 – p (the probability that will not occur/ probability of failure on each trial)

The binomial probability distribution has two parameters n and p and is denoted by:

b( x ; n, p) ; (read as binomial probability of x having n trials & probability p for success)

(Page No: 1) By: Muhammad Memon

Question: Derive a formula that gives the probability of x successes in n trials for a binomial
experiment.

Solution: The experiment has n trials , each trial may result S or F, therefore total number
Successes are (n+1), i.e. 0, 1, 2, 3, … , n

The Sample Space has 2n possible sample points or outcomes.

Now, consider the probability of zero (0) Success i.e. P(X = 0) = P(FF…F), because in case of
zero Success, every trial results in F and the event consists of a sequence of n F’s i.e. { FF … F}

Since in each trial P(S) = p and P(F) =q and trials are independent, so we apply the multiplicative
law of probability for independent events and obtain

P(X = 0) = P(F) P(F) … P(F)

=qq…q
= qn

Now consider the probability of one Success i.e. (X=1)

In this case, one trial results in S and remaining (n-1) trials result in Fs. This event can occur in
several different sequences i.e. P(X=1) = P(SFF … F) or P(FSFF … F) & so on.

P(X = 1) = P(S) P(F) P(F) … P(F)

=pqq…q
= pqn-1
 n
The number of mutually exclusive sequences in which one S and (n-1) Fs occur is  
1 
Therefore the probability of exactly one Success for all possible sequences is:

 n
P(X = 1) =   pqn-1
1 
In general, when X = x , the probability of a sequence that has exactly x Successes
and (n - x ) failures is p x qn-1 and different sequences in which x Successes in n trials for a
binomial experiment is:
 n
P(X = x ) =   p x qn- x for x = 0, 1, 2, … ,n
 x
This is the formula for binomial probability distribution having n trials & probability p for
Success.

n
 n  x n- x n
The sum of probabilities i.e.    p q
x 0  x 
=1 or 
x 0
b( x ; n,p) = 1

It is to be noted that the probability of r or fewer Successes is given by:

r
 n
P(X  r) =   x  p x qn- x
x 0  

Where P(X  r) is the cumulative binomial distribution of X.

(Page No: 2) By: Muhammad Memon

Question: A fair coin is tossed 6 times. Find the probabilities of obtaining various number of
heads.

Question: Let X have a binomial distribution with n=4 and p=1/3 .

Find P(X = 1) , P(X = 3/2), P (X = 3), P(X ≤ 2)

Question: Twenty percent of all clothing items ordered from a mail-order are returned within a
week of receipt. Suppose six orders are received. Determine the probability
distribution of the number of return.

Question: A test consisting of 20 true-false questions is taken by a student who is totally

unprepared and therefore decides to make purely random choices between the true
and false answers.

1. Calculate the probability that the student will give exactly five correct answers.
2. Suppose at least 8 correct answers are required to pass the test. What is the
probability that the student will pass the test?

Question: The experience of a house-agent indicates that he can provide suitable accommodation
for 75 percent of the clients who come to him. If on a particular occasion 6 clients
approach him independently. Calculate the probability that
a) Less than 4 clients
b) Exactly 4 clients
c) At least 5 clients
will get suitable accommodation.

Binomial Frequency Distribution:

If the binomial probability distribution is multiplied by N, the number of experiments, the

resulting distribution is known as the binomial frequency distribution.

Thus the expected frequency of x Successes in N experiments is:

 n
N.   p x qn- x for x = 0, 1, 2, … ,n
 x
It should be noted that n independent trials constitute one experiment.

(Page No: 3) By: Muhammad Memon

Example: Six dice are thrown 729 times. How many times do you expect at least three dice to
show a five or six?

Solution: The probability of getting 5 or 6 with one die is p = 2/6

 six dice are thrown and there are 729 sets,
 the binomial frequency distribution is given by:
6 x
6  1  2 
x

729      
 x  3  3 
Hence the expected number of times at least 3 dice showing 5 or 6

 6  6  1  x  2  6 x 
= 729  x  3   3  
 3      
 6  1 3  2 3  6  1  4  2  2  6  1 5  2   6  1  6  2  0 
= 729                        
 3  3   3   4  3   3   5  3   3   6  3   3  

= 36 160  60  12  1  233
729

Properties of the Binomial Distribution:

(i) The mean number of Successes

(ii) The variance of the number of the Successes

The mean number of Successes is np i.e. µ= np

Similarly the mean number of failures is nq

The variance of the number of Successes is npq i.e.  2 = npq ,

then the standard deviation is npq i.e. = npq

Example: In an examination, 24 candidates appear. If the probability of passing the subject

be 1 3 , find the mean and dispersion of the distribution.

Solution: Here n = 24, p = 1

3 so that q = 1 – p = 2
3

Mean = np
Mean = 24 x 1
3 =8
And  = npq = 24  13  2 3  5.33  2.3

(Page No: 4) By: Muhammad Memon

Fitting a Binomial Distribution to observed Data:

Fitting a binomial distribution to given data consists in finding the values of p and n (the two
parameters that completely determine a binomial distribution) and is calculating the probabilities
as well as expected frequencies for x = 0, 1, 2, 3, … , n

Question: Fit a binomial distribution to the following data, obtained by tossing a biased- coin 5
times:

Number of Heads 0 1 2 3 4 5 Total

Frequency 12 56 74 39 18 1 200

Solution: First we find value of mean

_
x
fx i i

0  56  148  117  72  5 398
  1.99
f i 200 20
_
x  np
1.99 = 5p or p= 1.99  0.398 then q = 1 – 0.398 = 0.602
5

Hence the fitted binomial distribution is given by:

5 
b( x ; 5, 0.398) =  0.398 0.602
x 5 x

 x
The probabilities and frequencies are calculated as below:

No. of Heads Probability f(X) Expected

(X) Frequency
5
 0.398 0.602  0.07907
0 5
0 15.8
 0
 5
 0.398 0.602  0.26136
1 4
1 52.5
1 
5
 0.398 0.602  0.34559
2 3
2 69.1
 2
 5
 0.398 0.602  0..22847
3 2

3  
3 45.7

5
 0.398 0.602  0.07553
4
4 15.1
 4
 5
 0.398 0.602  0.00998
5 0
5 2.0
 5
Total 1.00000 200.0

The expected frequencies are obtained by multiplying each of the probabilities by 200.

(Page No: 5) By: Muhammad Memon