UNIT-5

Testing of Hypothesis

Prepared by : A. Anilkumar , Asst.Professor

 Chapter 1 : Tests of Hypothesis ( Large Sample Tests)

Statistical Hypothesis:
A Statement about the nature of the population that may be either right or wrong, it is
often stated in terms of a population parameter.

Testing of Hypothesis:
The Systematic Procedure for deciding whether to accept / reject the given hypothesis
is called Testing of Hypothesis or Testing of Significance

Prepared by : A. Anilkumar , Asst.Professor

Procedure of Testing Hypothesis:

STEP 1 : we write the Null Hypothesis

Null Hypothesis : A Hypothesis which asserts that there is no Significant Difference between the Statistic
and population parameter is called a Null Hypothesis. It is denoted by 𝐻0
It is of the form 𝑯𝟎 : 𝝁 = 𝝁𝟎

STEP 2 : We Write the Alternate Hypothesis

Alternate Hypothesis : Any Hypothesis Which Contradicts the Null Hypothesis is called an Alternative
Hypothesis. It is Denoted by 𝐻1

Alternate Hypothesis Could be of the form

i. 𝑯𝟏 ∶ 𝝁 ≠ 𝝁𝟎 ( Two Tailed Alternate Hypothesis, it is in either 𝜇 < 𝜇0 or 𝜇 > 𝜇0 )
ii. 𝑯𝟏 ∶ 𝝁 < 𝝁𝟎 ( Left Tailed Alternate Hypothesis)
iii. 𝑯𝟏 ∶ 𝝁 > 𝝁𝟎 ( Right Tailed Alternate Hypothesis)

Prepared by : A. Anilkumar , Asst.Professor

STEP 3: Specification of level of significance

Level of Significance: It is the confidence level with which we rejects or accepts the Null Hypothesis and
it is denoted by 𝜶.

LOS generally specified before a test Procedure so that the results obtained may not influence our
decision
we can take 𝜶 can be either 1% , 5% , or 10% . If we did not identify the 𝛼 value in the
Problem then we take usually 𝜶 = 5%

STEP 4 : Identification of Test Statistic

i. If 𝑛 > 30, then we can use Z test

ii. If 𝑛 ≤ 30 , then we can use t Test , where n = sample size

Prepared by : A. Anilkumar , Asst.Professor

STEP 5 : Find the Critical Region / Rejection Region Value

Critical Region: A region Corresponding to a Statistic in the sample S Which Leads to Rejection 𝐻0 is
called Critical Region.

𝜇 ≠ 𝜇0

𝜇 > 𝜇0

𝜇 < 𝜇0

Prepared by : A. Anilkumar , Asst. Professor

STEP 6 : Write The Conclusion by the Following Points.

i. First We Compare the Computed Value of the test Statistic 𝒁𝒄𝒂𝒍 with the Critical Value 𝒁𝜶 at
given level of significance.
ii. If 𝒁𝒄𝒂𝒍 < 𝒁𝜶 then we accept the Null Hypothesis.
iii. If 𝒁𝒄𝒂𝒍 > 𝒁𝜶 then we reject the Null Hypothesis.

Errors of Sampling:
Type I : Reject Null Hypothesis, when it is True.
Type II : Accept Null Hypothesis, when it is False.

Under Large Sample Tests, We Have Four Important Tests to Test the Significance. They Are

i. Testing Of Significance For Single Mean

ii. Testing Of Significance For Difference of Means
iii. Testing Of Significance For Single Proportion
iv. Testing Of Significance For Difference of Proportions

Prepared by : A. Anilkumar , Asst. Professor

i. Testing Of Significance For Single Mean:

ഥ−𝝁
𝒙
In this Concept, we Use the Test Statistic z= 𝝈 , Where n = Sample Size , 𝑥ҧ = Sample Mean
𝒏
𝜇 = Population Mean , 𝜎 = Population S.D

ഥ−𝝁
𝒙
If the Population S.D is not Known, then we can use z= 𝑺 , Where S = Sample S.D
𝒏

Note :
i. We Reject Null Hypothesis 𝐻0 when 𝑧 > 3 without mentioning any level of significance
𝝈
ii. The Values 𝒙 ± 𝟏. 𝟗𝟔 𝒏 are called 95% Confidence limits for the mean of the population Corresponding to the
given sample.
𝝈
iii. The Values 𝒙 ± 𝟐. 𝟑𝟑 are called 98% Confidence limits for the mean of the population Corresponding to the
𝒏
given sample.
𝝈
iv. The Values 𝒙 ± 𝟐. 𝟓𝟖 are called 99% Confidence limits for the mean of the population Corresponding to the
𝒏
given sample.

Prepared by : A. Anilkumar , Asst. Professor

ii. Testing Of Significance For Difference of Means:

𝒙𝟏 −𝒙𝟐
In this Concept, We Use the Test Statistic 𝒁 =
𝝈𝟐 𝝈𝟐
+
𝒏𝟏 𝒏𝟐

Where, 𝑥1 = mean of sample 1

𝑥2 = mean of sample 2
𝜎 = S.D. of Population
𝑛1 = Size of sample 1
𝑛2 = Size of sample 2

𝒙𝟏 −𝒙𝟐
Note: If the Population S.D is not Known, then we can use 𝒁 = ,
𝑺𝟏 𝟐 𝑺𝟐 𝟐
+
𝒏𝟏 𝒏𝟐

Where 𝑆1 = S. D of Sample 1 ; 𝑆2 = S. D of Sample 2

Prepared by : A. Anilkumar , Asst. Professor

iii. Testing Of Significance For Single Proportion:

𝒑−𝑷
In this Concept, we use the test Statistic 𝒛 = ,
𝑷𝑸
𝒏
Where, p = Sample Proportion ; P = Population Proportion, Q = 1- P

Prepared by : A. Anilkumar , Asst. Professor

Example : In a Random Sample of 125 cool drinkers, 68 said they prefer Thumsup to Pepsi. Test
the Hypothesis 𝑃 = 0.5 against the alternative hypothesis 𝑃 > 0.5
Since Calculated value of z > Tabulated Value of z , we reject the null hypothesis and conclude that
majority of men in the city are smokers
iv. Testing Of Significance For Difference of Proportions

a) When the Population Proportions 𝑷𝟏 and 𝑷𝟐 are Known

𝒑𝟏 −𝒑𝟐
In this Case, we use the test Statistic 𝒛 =
𝑷𝟏 𝑸𝟏 𝑷𝟐 𝑸𝟐
𝒏
+ 𝒏
𝟏 𝟐
Where, 𝑝1 = Sample 1 Proportion
𝑝2 = Sample 2 Proportion
𝑃1 = Population 1 Proportion
𝑃2 = Population 2 Proportion
𝑛1 = Size of sample 1
𝑛2 = Size of sample 2
𝑄1 = 1 − 𝑃1 ; 𝑄2 = 1 − 𝑃2

b) When the Population Proportions 𝑷𝟏 and 𝑷𝟐 are not Known but sample proportions 𝒑𝟏 and 𝒑𝟐 are
known
𝒑𝟏 −𝒑𝟐
In this Case, we use the test Statistic 𝒛 = 𝒑 𝒒 𝒑 𝒒 ,
𝟏 𝟏+ 𝟐 𝟐
𝒏 𝟏 𝒏 𝟐
Where, 𝑝1 = Sample 1 Proportion
𝑝2 = Sample 2 Proportion , 𝑛1 = Size of sample 1 , 𝑛2 = Size of sample 2 , 𝑞1 = 1 − 𝑝1 ; 𝑞2 = 1 − 𝑝2
c) Method of Pooling :
In This Method, we follow this Procedure.
 Chapter 2 : Tests of Significance (for Small Samples)

Degrees Of Freedom:
The Number of independent variates which make up the statistic is known as the degree of
freedom and it is denoted by 𝝑

In this Concept, we are dealing the problems with the sample size(n) < 30. So, we can use
the following methods to test the hypothesis . They are
1. t – Distribution
2. F – Distribution
3. 𝜒 2 Chi − Square Distribution
t – Distribution (or) Student’s t – Distribution :

It is Used for testing of hypothesis when the sample size is small and population S.D
(𝜎) is not known.

t – Test for Single Mean:

ഥ−𝝁
𝒙
i) In this Concept, We Use t= 𝒔 , Where Where n = Sample Size , 𝑥ҧ = Sample Mean , 𝜇 = Population
𝒏−𝟏
Mean , 𝑠 = Sample S.D
( we Use this formula when sample S.D is Directly given in the problem)

ഥ−𝝁
𝒙
ii) We Use t= 𝑺 , When Sample Variance(𝑆 2 ) is given.
𝒏
𝒙 𝟐
σ 𝒙𝒊 −ഥ
Where, Sample Vaiance (𝑆 2 ) =
𝒏−𝟏
𝒙 𝟐
σ 𝒙𝒊 −ഥ
From this S =
𝒏−𝟏
In this concept, we use degrees of Freedom(d.f) formula = n− 1 for finding t- table value for comparing
with t - calculated value
We Use t- Distribution table
while testing of hypothesis.

Note :
Suppose
In this case, if Alternate
Hypothesis is Two tailed test
(𝜇 ≠ 𝜇0 )
and 𝛼 = 0.05 𝑔𝑖𝑣𝑒𝑛, while
calculating t-table value we
𝛼
consider 2 = 0.025
Note :
1. 𝑡1−𝛼 = −𝑡𝛼
2. 𝑡−𝛼 = −𝑡𝛼
t – Test for Difference of means :
when we identify two samples and their variances in the problem, we use the below formula.

ഥ−ഥ
𝒙 𝒚 𝒙 𝟐 +σ 𝒚𝒊 −ഥ
σ 𝒙𝒊 −ഥ 𝒚 𝟐 𝒏𝟏 𝒔𝟏 𝟐 +𝒏𝟐 𝒔𝟐 𝟐
t= , Where 𝑺 = =
𝑺.
𝟏
+
𝟏 𝒏𝟏 +𝒏𝟐 −𝟐 𝒏𝟏 +𝒏𝟐 −𝟐
𝒏𝟏 𝒏𝟐

In this concept, we use degrees of Freedom(d.f) formula = 𝒏𝟏 + 𝒏𝟐 − 2 for finding t - table value for
comparing with t - calculated value.

To test the whether the two population means are equal , we can use this concept
In this Concept, always we take
1. Null Hypothesis : 𝝁𝟏 = 𝝁𝟐
2. Alternate Hypothesis : 𝝁𝟏 ≠ 𝝁𝟐
F – Distribution( Test for Equality of two Population Variances) :

To test the hypothesis that two population variances 𝜎1 2 𝑎𝑛𝑑 𝜎2 2 equal , we can
use this concept.
in this concept, we take always
1. Null Hypothesis : 𝝈𝟏 𝟐 = 𝝈𝟐 𝟐
2. Alternate Hypothesis : 𝝈𝟏 𝟐 ≠ 𝝈𝟐 𝟐
𝑺𝟏 𝟐
3. Test Statistic : F = 𝟐 if 𝑺𝟏 > 𝑺𝟐
𝑺
𝟐
(Or)
𝑺𝟐 𝟐
F= if 𝑺𝟐 > 𝑺𝟏
𝑺𝟏 𝟐

𝟐 𝒏𝟏 𝒔𝟏 𝟐 𝒙 𝟐
σ 𝒙𝒊 −ഥ 𝟐 𝒏𝟐 𝒔𝟐 𝟐 𝒚 𝟐
σ 𝒚𝒊 −ഥ
Where, 𝑺𝟏 = = and 𝑺𝟐 = = , where 𝑠1 2 𝑎𝑛𝑑 𝑠2 2 are variances
𝒏𝟏 −𝟏 𝒏𝟏 −𝟏 𝒏𝟐 −𝟏 𝒏𝟐 −𝟏
of sample 1 and sample 2 respectively.

In this Concept, we use degrees of freedom formula = (𝒏𝟏 −1, 𝒏𝟐 −1) to compare F table value with F
Calculated Value
We Use this
F-Distribution table
while comparing F
calculated Value when
𝜶 = 𝟎. 𝟎𝟓
We Use this
F-Distribution table while
comparing F calculated
Value when 𝜶 = 𝟎. 𝟎𝟏