Course: Probability and Statistics

Instructor: Adnan Aslam

Adnan Aslam Course: Probability and Statistics 1 / 15

Sample proportion
• Suppose a proportion p of population favor a candidate A for
president.
• In a random sample of size n drawn from the population, a certain
proportion p̂ of the sample will favor candidate A
• and the collection of all such proportions defines a random variable P̂,
called the sample proportion.

Adnan Aslam Course: Probability and Statistics 2 / 15

Sampling Distribution of Sample Proportion
Population proportion
A population proportion p may be identified with the population mean, where the
mean is obtained from the units whose possible values are either 0’s or 1’s. In
other words, let
• Yi = 1, if the ith unit possesses the characteristic of interest,
• = 0, if the ith unit does not possesses the characteristic of interest.
Then the mean is
N
1 X
µ = Yi
N
i=1
Number of units having the characteristic of interest X
= =
Total number of units in the population N

where X represents the number of units having the characteristics of interest.

Thus the mean is simply the proportion of 1;s in the population and we write p
for µ, meaning proportion (usually called the proportion of success).
Adnan Aslam Course: Probability and Statistics 3 / 15
Sampling distribution of Sample proportion
• The sample proportion P̂ is defined as
n
1X X
P̂ = Y = =
n n
i=1
P
It is interesting to note the X = Yi is a binomial random variable
and the binomial parameter p is being called a proportion of success
here.
• The sample proportion P̂ has different values in different samples.
• It is obviously a random variable and has a probability distribution.
• This probability distribution of the proportions of all possible random
samples of size n, is called the sampling distribution of P̂.

Adnan Aslam Course: Probability and Statistics 4 / 15

Properties of sampling distribution of proportion
The sampling distribution of P̂ has the following important properties:
• The mean of the sampling distribution of proportions, denoted by µp̂
is equal to the population proportion p, that is µp̂ = p.
• The standard deviation of the sampling distribution of proportions,
called the standard error of P̂ and denoted by σp̂ , is given as
r
pq
σp̂ =
n
when the sampling is performed with replacement, or
r
pq N − n
σp̂ = .
n N −1
when the sampling is done without replacement from a finite
population and here q = 1 − p.

Adnan Aslam Course: Probability and Statistics 5 / 15

Properties of sampling distribution of proportion
• The sampling distribution of P̂ is the binomial distribution. However,
for sufficiently large sample sizes, the sampling distribution of P̂ is
approximately normal. As a rule of thumb, the sampling distribution
of P will be approximately normal whenever both np and nq are equal
to or greater than 5.
• We use a continuity correction of ± 12 , whenever we consider the
normal approximation to the binomial distribution. Now, we need to
1
use a continuity correction of ± 2n as P̂ = Xn .

Adnan Aslam Course: Probability and Statistics 6 / 15

Example
A population consists of N = 6 values 1, 3, 6, 8, 9 and 12. Draw all
possible samples of size n = 3 without replacement from the population
and find the proportion of even numbers in the samples. Construct the
sampling distribution of sample proportions and verify that
(a). µp̂ = p,
pq N−n
(b). Var (P̂) = n . N−1 ,
where q = 1 − p; P̂ and p are sample and population proportions
respectively.

Solution.
• The number of possiblesamples
 of size n = 3 that could be selected
6
without replacement is = 20.
3
• Let P̂ represent the proportion of even numbers in the sample, then
the 20 possible samples and the proportion of even numbers are given
as follows:
Adnan Aslam Course: Probability and Statistics 7 / 15
 Sample No. Sample Data Sample Proportion P̂
1 1,3,6 1/3
2 1,3,8 1/3
3 1,3,9 0
4 1,3,12 1/3
5 1,6,8 2/3
6 1,6,9 1/3
7 1,6,12 2/3
8 1,8,9 1/3
9 1,8,12 2/3
10 1,9,12 1/3
11 3,6,8 2/3
12 3,6,9 1/3
13 3,6,12 2/3
14 3,8,9 1/3
15 3,8,12 2/3
16 3,9,12 1/3
17 6,8,9 2/3
18 6,8,12 1
19 6,9,12 2/3
20 8,9,12 2/3

Adnan Aslam Course: Probability and Statistics 8 / 15

The sampling distribution of sample proportion is given below:

P̂ No of samples Probability f (P̂) p̂f (p̂) p̂ 2 f (p̂)

0 1 1/20 0 0
1/3 9 9/20 3/20 1/20
2/3 9 9/20 6/20 4/20
1 1 1/20 1/20 1/20
P
20 1 10/20 6/20

Now
X 10
µp̂ = p̂f (p̂) = = 0.5,
20
X hX i2
σp̂2 = p̂ 2 f (p̂) − p̂f (p̂)
 2
6 10 1
= − = = 0.05.
20 20 20

Adnan Aslam Course: Probability and Statistics 9 / 15

• To verify the given relations, we first calculate the population
proportion p and the population variance pq. Thus
• p=X N , where X represents the number of even numbers.
• 3
= 6 = 0.5, and
• σ2 = pq = (0.5)(0.5) = 0.25.
• Therefore µp̂ = 0.5 = p, and

pq N − n 0.25 6 − 3
. = . = 0.05 = Var (P̂)
n N −1 3 6−1
• hence the result

Adnan Aslam Course: Probability and Statistics 10 / 15

Example
Ten percent of the 1-kilogram boxes of sugar in a large warehouse are
underweight. Suppose a retailer buys a random sample of 144 of these
boxes. What is the probability that at least 5 per cent of the sample boxes
will be underweight?

Solution. Here the statistic is the sample proportion P̂.

• Given:
• Population proportion= 100
10
= 0.1
• Sample size=n = 144

Adnan Aslam Course: Probability and Statistics 11 / 15

Example
Ten percent of the 1-kilogram boxes of sugar in a large warehouse are
underweight. Suppose a retailer buys a random sample of 144 of these
boxes. What is the probability that at least 5 per cent of the sample boxes
will be underweight?

Solution. Here the statistic is the sample proportion P̂.

• Given:
• Population proportion= 100
10
= 0.1
• Sample size=n = 144
• Required P(p̂ ≥ 0.05)

Adnan Aslam Course: Probability and Statistics 11 / 15

Example
Ten percent of the 1-kilogram boxes of sugar in a large warehouse are
underweight. Suppose a retailer buys a random sample of 144 of these
boxes. What is the probability that at least 5 per cent of the sample boxes
will be underweight?

Solution. Here the statistic is the sample proportion P̂.

• Given:
• Population proportion= 100
10
= 0.1
• Sample size=n = 144
• Required P(p̂ ≥ 0.05)
• The sample size is large enough to assume that the sample proportion
P̂ is approximately normally distributed with mean

µp̂ = p = 0.1
r r
pq (0.1)(0.9)
σp̂ = = = 0.025
n 144
Adnan Aslam Course: Probability and Statistics 11 / 15
• We are asked to find the probability that the sample proportion of the
underweight boxes is equal to greater than 5%, i.e., we require

Adnan Aslam Course: Probability and Statistics 12 / 15

• We are asked to find the probability that the sample proportion of the
underweight boxes is equal to greater than 5%, i.e., we require
1
P(p̂ ≥ 0.05) ⇒ P(p̂ ≥ 0.05 − ) continuity correction
2n
1
= P(p̂ ≥ 0.05 − ) = P(p̂ ≥ 0.0465)
2(144)
p̂ − 0.10 0.0465 − 0.10
= P( ≥ )
0.025 0.025
= P(Z ≥ −2.14)
= 0.9838(From standard normal table)

Adnan Aslam Course: Probability and Statistics 12 / 15

Sampling Distribution of Difference between proportions
• Suppose there are two binomial populations with proportions of
successes p1 and p2 respectively.
• Let independent random samples of sizes n1 and n2 be drawn from
the respective populations
• and the differences p̂1 − p̂2 between the proportions of all possible
pairs of be computed.
• Then a probability distribution of the differences P̂1 − P̂2 can be
obtained.
• Such a probability distribution is called the sampling distribution of
the differences between the proportions P̂1 − P̂2 .

Adnan Aslam Course: Probability and Statistics 13 / 15

Properties of the sampling distribution of the differences between the
proportions P̂1 − P̂2
• The mean of the sampling distribution of P̂1 − P̂2 denoted by µp̂1 −p̂2
is equal to the difference between the population proportions, that is

µp̂1 −p̂2 = p1 − p2

• The standard deviation of the sampling distribution of P̂1 − P̂2 , (i.e.

the standard error of P̂1 − P̂2 ) denoted by σP̂1 −P̂2 is given by
r
p1 q1 p2 q2
σP̂1 −P̂2 = + .
n1 n2

• The sampling distribution of P̂1 − P̂2 is approximately normal for

sufficiently large sample sizes.

Adnan Aslam Course: Probability and Statistics 14 / 15

Example
Two random samples of sizes n1 = 40 and n2 = 45 are drawn from a
binomial population with p = 0.60. What is the probability that
−0.15 < p̂1 − p̂2 < +0.15?

Solution.
• Given
• p1 = p2 = p = 0.60
• n1 = 40 and n2 = 45.
• Required What is the probability that −0.15 < p̂1 − p̂2 < +0.15?

Adnan Aslam Course: Probability and Statistics 15 / 15