Testing of Hypothesis

• Population

• Sample

• Parameter

• Statistic

• Standard Error – Standard deviation of the sampling

distribution of a statistic.
Statistical Measure Parameter Statistic

Mean 𝜇 𝑋
Median 𝑀 𝑚
Variance 𝜎2 𝓈2
Standard deviation 𝜎 𝓈
Proportion 𝑃 𝑝
Correlation Coefficient 𝜌 𝑟
Regression Coefficient 𝛽 𝑏
Test of Significance
(i) The deviation between the observed sample statistic and the
hypothetical parameter value.
(ii) The deviation between two independent sample statistics: is
significant or might be attributed to chance or the fluctuations of
sampling.

Large Sample Tests (𝒏 > 𝟑𝟎): Normal Test of significance.

Small Sample Tests(𝒏 ≤ 𝟑𝟎): t-test, F-test and Chi-square Test.

Null Hypothesis (𝑯𝟎 ): The hypothesis which is tested for possible rejection
under the assumption that it is true.
Ex: 𝑯𝟎 : 𝝁 = 𝝁𝟎 ; 𝜇 − population mean

Alternative Hypothesis (𝑯𝟏 ): Any hypothesis which is complementary to the

null hypothesis.

Ex: 𝑯𝟏 : 𝝁 ≠ 𝝁𝟎 (𝝁 > 𝝁𝟎 or 𝝁 < 𝝁𝟎 ) (Two-tailed alternative)

𝑯𝟏 : 𝝁 > 𝝁𝟎 (Right-tailed alternative)
𝑯𝟏 : 𝝁 < 𝝁𝟎 (Left-tailed alternative)
Type – I Error : Reject 𝑯𝟎 when it is true.
𝑃 Reject 𝑯𝟎 𝑯𝟎 } = 𝜶

Type – II Error : Accept 𝑯𝟎 when it is wrong. i.e., accept 𝑯𝟎 when 𝑯𝟏 is true.

𝑃 Accept 𝑯𝟎 𝑯𝟏 } = 𝛃

Critical Region : A region in the sample space S which amounts to

rejection of 𝑯𝟎 is termed as critical region of rejection.
Level of Significance: The probability ′𝜶′ that a random value of the
statistic ′𝒕′ belongs to the critical region is known
as the level of significance. (𝟏%, 𝟓% 𝑎𝑛𝑑 𝟏𝟎%)

Degree of Freedom: The number of independent observations, on which a

test is based, is known as d.f. of the test statistic.

𝒕−𝑬(𝒕)
Test Statistic: 𝒁= ~ 𝑵(𝟎, 𝟏) asymptotically as 𝒏 → ∞
𝑺.𝑬.(𝒕)
Critical Values or Significant Values: The critical value of the test
statistic at level of significance ′𝜶′ for two-tailed test is given by:

𝑃 𝑍 > 𝑧𝛼 = 𝛼

𝑃 𝑍 > 𝑧𝛼 + 𝑃 𝑍 < −𝑧𝛼 = 𝛼

𝑃 𝑍 > 𝑧𝛼 + 𝑃 𝑍 > 𝑧𝛼 = 𝛼 (By Symmetry)

2 𝑃 𝑍 > 𝑧𝛼 = 𝛼

𝑃 𝑍 > 𝑧𝛼 = 𝛼 2
In case of one-tailed test, the critical value 𝒛𝜶 is determined as:

For Right tailed test: 𝑷 𝒁 > 𝒛𝜶 = 𝜶

For Left tailed test : 𝑷 𝒁 < −𝒛𝜶 = 𝜶

The critical value of 𝒁 at commonly used level of significance are:

Level of Significance (𝜶)

Critical Value
𝟏% (0.01) 𝟓% (0.05) 𝟏𝟎% (0.10)
(𝒛𝜶 ) Two-tailed 0.005 Two-tailed 0.025 Two-tailed 0.05
One-tailed 0.01 One-tailed 0.05 One-tailed 0.10

Two-tailed Test 2.58 1.96 1.645

Right-tailed Test 2.33 1.645 1.28

Left-tailed Test -2.33 -1.645 -1.28

Procedure for Testing of Hypothesis
1. Set up the Null Hypothesis (𝑯𝟎 ).
2. Set up the Alternative Hypothesis (𝑯𝟏 ).
3. Choose the appropriate Level of Significance (𝜶).
𝒕−𝑬(𝒕)
4. Compute the Test Statistic: 𝒁 = , under 𝐻0
𝑺.𝑬.(𝒕)
5. Conclusion. We compare the computed value of 𝐙 in step 4 with
the significant value (tabulated value) 𝐳𝛂 at the
given level of significance.
If 𝒁𝒄𝒂𝒍 < |𝒛𝜶 | -- accept the null hypothesis 𝑯𝟎 .
If 𝒁𝒄𝒂𝒍 > |𝒛𝜶 | -- reject the null hypothesis 𝑯𝟎 ,
i.e., accept the alternative hypothesis 𝑯𝟏 .
Test for Single Mean
The test statistic for testing 𝑯𝟎 : 𝝁 = 𝝁𝟎 𝝁- population mean
vs
the alternative hypothesis: 𝑯𝟏 : 𝝁 ≠ 𝝁𝟎 (𝝁 > 𝝁𝟎 or 𝝁 < 𝝁𝟎 )
(Two-tailed alternative) or
𝑯𝟏 : 𝝁 > 𝝁𝟎 (right-tailed alternative)
𝑯𝟏 : 𝝁 < 𝝁𝟎 (left-tailed alternative)

𝒙−𝝁
Is given by the test statistic: 𝒁= ~ 𝑵(𝟎, 𝟏) ; 𝒙 - sample mean
𝝈 𝒏
(i) If the population s.d. 𝝈 is unknown then the estimate of 𝝈 = 𝓼 ,
𝒙−𝝁
the sample s.d. is used for test statistic. i.e., 𝒁=
𝓼 𝒏
The distribution of 𝑍0 when 𝐻0 : 𝜇 = 𝜇0 is true,
with the critical region for
(a) The two-sided alternative 𝐻1 : 𝜇 ≠ 𝜇0
(b) The one-sided alternative 𝐻1 : 𝜇 > 𝜇0
(c) The one-sided alternative 𝐻1 : 𝜇 < 𝜇0
Example 1
The burning rate of rocket propellant is being studied. Specifications require
that the mean burning rate must be 140 cm/s and standard deviation of the
burning rate is approximately 5 cm/s. The experimenter decides to test the
mean burning rate is 140 cm/s or not? A random sample of thirty-five
specimens are tested and sample mean rate obtained is 𝒙 = 𝟏𝟒𝟏. 𝟐𝟓 cm/s.
Test the significance at 𝜶 = 𝟎. 𝟎𝟓.
Solution:
Given, Sample mean: 𝒙 = 141.25 cm/s Population Mean : 𝝁 = 𝟏𝟒𝟎 cm/s
Sample size : n = 35 Population Standard deviation : 𝝈 = 𝟓 cm/s

1. Null Hypothesis : 𝑯𝟎 : 𝝁 = 𝟏𝟒𝟎

2. Alternative hypothesis: 𝑯𝟏 : 𝝁 ≠ 𝟏𝟒𝟎 (Two-tailed test)

3. Level of Significance: 𝜶 = 0.05 and 𝒛𝜶/𝟐 = 𝒛𝟎.𝟎𝟓/𝟐 = 𝒛𝟎.𝟎𝟐𝟓 = 𝟏. 𝟗𝟔

𝒙−𝝁 𝟏𝟒𝟏.𝟐𝟓 −𝟏𝟒𝟎

4. Test statistic is given by: 𝒁= = = 𝟏.479
𝝈 𝒏 𝟓 𝟑𝟓

𝒁𝒄𝒂𝒍 < |𝒛𝜶 | ⟹ 1.479 < 1.96

5. Accept the null hypothesis 𝑯𝟎 : 𝝁 = 𝟏𝟒𝟎
P-value: 𝑃 −𝟏𝟒𝟏. 𝟐𝟓 ≤ 𝑿 ≤ 141.25

141.25−140 𝒙 − 𝝁𝟎 141.25−140
= 𝑃 − ≤ ≤
5/ 35 𝜎 𝒏 5/ 35

= 𝑃 𝑍 ≥ 1.479

= 2 𝑃 𝑍 ≥ 1.479 ∵ 𝑍~𝑁(0, 1)

p-value = 0.14

Comparison: 0.14 > 0.05

Accept the null hypothesis 𝑯𝟎 : 𝝁 = 𝟏𝟒𝟎 at 𝜶 = 𝟎. 𝟎𝟓 l.o.s.

Example 2
All cigarettes presently on the market have an average nicotine content of at
least 1.6 mg per cigarette. A firm that produces cigarettes claims that it has
discovered a new way to cure tobacco leaves that will result in the average
nicotine content of a cigarette being less than 1.6 mg.

To test this claim, a sample of 40 of the firm’s cigarettes were analyzed. If it

is known that the standard deviation of a cigarette’s nicotine content is 𝟎. 𝟖 𝒎𝒈,
what conclusions can be drawn, at the 5 percent level of significance, if the
average nicotine content of the 20 cigarettes is 1.54?
Solution: Given, Sample mean: 𝒙 = 1.54 Population Mean : 𝝁 =1.6
Sample size : n = 40 Population S.D. : 𝝈 = 𝑖𝑠 𝑘𝑛𝑤𝑜𝑛 = 0.8
1. Null Hypothesis : 𝑯𝟎 : 𝝁 ≤ 𝟏. 𝟔

2. Alternative hypothesis: 𝑯𝟏 : 𝝁 > 𝟏. 𝟔 (Right-tailed test)

3. Level of Significance (𝜶) = 5% and 𝒛𝟎.𝟎𝟓 = 𝟏. 𝟔𝟒

𝒙 − 𝝁𝟎 𝟏.𝟓𝟒 −𝟏.𝟔
4. Test statistic is given by: 𝒁= = = −𝟎. 𝟒𝟕𝟒𝟑
𝜎 𝒏 𝟎.𝟖 𝟒𝟎

𝒁𝒄𝒂𝒍 < |𝒛𝜶 | ⟹ 0.47 < 1.64

5. Accept the Null hypothesis 𝑯𝟎 : 𝝁 ≤ 𝟏. 𝟔

P-value: 𝑃 𝑿 > 1.54

𝒙 − 𝝁𝟎 𝟏.𝟓𝟒 −𝟏.𝟔
=𝑃 >
𝜎 𝒏 𝟎.𝟖 𝟒𝟎

= 𝑃 𝒁 > − 0.474 𝑍~𝑁(0, 1)

= 𝑃 𝒁 < 0.474 ( Symmetry)

= 0.68
Example 3
An insurance agent has claimed that the average age of policy holders who
insure through him is less than the average for all agents, which is 30.5 years.
A random sample of 100 policy holders who had insured through him gave the
following age distribution:

Age 16-20 21-25 26-30 31-35 36-40

No. of
12 22 20 30 16
Persons

Test the claim at 5% level of significance.

 Confidence Interval for Single mean
If 𝑿 is the sample mean of a random sample of size 𝒏 from a normal population
with known variance 𝝈𝟐 , a 𝟏𝟎𝟎(𝟏 − 𝛂)% C.I. on 𝝁 when 𝝈 – is known is
given by

𝝈 𝝈
𝑿− 𝒁𝜶/𝟐 ≤μ ≤ 𝑿+ 𝒁𝜶/𝟐 - two tailed test
𝒏 𝒏

where 𝒁𝜶/𝟐 is the upper 𝟏𝟎𝟎𝜶/𝟐 percentage point of the standard normal
distribution.
100 1 − 𝛼 % C.I. for 𝝁 for upper confidence:

𝝈
𝜇 ≤ 𝑿+ 𝒁𝜶
𝒏

100 1 − 𝛼 % C.I. for 𝝁 for lower confidence is :

𝝈
𝑿 − 𝒁𝜶 ≤ 𝝁
𝒏
Example
ASTM Standard E23 defines standard test methods for notched bar impact testing of metallic materials. The
Charpy V-notch (CVN) technique measures impact energy and is often used to determine whether or not a
material experiences a ductile-to-brittle transition with decreasing temperature.
Ten measurements of impact energy (𝑱) on specimens of A238 steel cut at 60ºC are as follows: 64.1,
64.7, 64.5, 64.6, 64.5, 64.3, 64.6, 64.8, 64.2, and 64.3.
Assume that impact energy is normally distributed with 𝝈 = 𝟏𝑱. Find the 95% CI for , the mean
impact energy.
Solution: Since, 𝒛𝜶/𝟐 = 𝒛𝟎.𝟎𝟓/𝟐 = 𝒛𝟎.𝟎𝟐𝟓 = 𝟏. 𝟗𝟔 , 𝑛 = 10, 𝜎 = 1 and 𝑋 = 64.46
The resulting 95% confidence interval is,

𝝈 𝝈
𝑿 − 𝒛𝜶/𝟐 ≤ 𝝁 ≤ 𝑿 + 𝒛𝜶/𝟐
𝒏 𝒏
1 1
64.46 − 1.96 ≤ 𝜇 ≤ 64.46 + 1.96 ⇒ 63.84 ≤ 𝜇 ≤ 65.08
10 10

That is, based on the sample data, a range of highly plausible values for mean impact energy
for A238 steel at 60°C is 𝟔𝟑. 𝟖𝟒𝑱 ≤ 𝝁 ≤ 𝟔𝟓. 𝟎𝟖𝑱.
Test for Difference of Two Population Means
Let 𝑿𝟏 be the mean of a sample of size 𝒏𝟏 from a population with mean 𝝁𝟏 and
variance 𝝈𝟐𝟏 and 𝑿𝟐 be the mean of a sample of size 𝒏𝟐 from a population with
mean 𝝁𝟐 and variance 𝝈𝟐𝟐 ,
𝜎12 𝜎22
𝑋1 ~ 𝑁 𝜇1 , and 𝑋2 ~ 𝑁 𝜇2 ,
𝑛1 𝑛2
Also ( 𝑥1 − 𝑥2 ) being the difference of two independent normal variates is also a
normal variate.
𝜎12 𝜎22
𝑋1 − 𝑋2 ~ 𝑁 𝜇1 − 𝜇2 , +
𝑛1 𝑛2
Then the test statistic:
𝑋1 −𝑋2 −𝐸 𝑋1 −𝑋2
𝑍= ~ 𝑁(0, 1)
𝑆.𝐸. 𝑋1 −𝑋2
The test statistic for testing 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
( 𝝁𝟏 − 𝝁𝟐 = 𝟎 )

vs

the alternative hypothesis: 𝑯𝟏 : 𝝁𝟏 ≠ 𝝁𝟐 (𝝁𝟏 > 𝝁𝟐 or 𝝁𝟏 < 𝝁𝟐 )

(Two-tailed alternative) or
𝑯𝟏 : 𝝁𝟏 > 𝝁𝟐 (right-tailed alternative)

𝑯𝟏 : 𝝁𝟏 < 𝝁𝟐 (left-tailed alternative)

 𝑿1 −𝑿2
Test Statistic: 𝒁= ~ 𝑁(0, 1)
𝜎2
1 𝜎2
2
+
𝑛1 𝑛2

Remarks:
1. If 𝝈 – is known and 𝜎12 = 𝜎22 = 𝜎 2 i.e., the sample have been
drawn from the populations with common Standard Deviation (𝝈),

𝑿1 −𝑿2
𝒁= 1 1
~ 𝑁(0, 1)
𝝈 +
𝑛1 𝑛2
2. If 𝜎 - is not known, then the unbiased estimator of 𝜎 2 is given by:
𝑛1 𝓈12 + 𝑛2 𝓈22
𝜎2 =
𝑛1 +𝑛2

3. If 𝜎12 ≠ 𝜎22 and 𝜎1 and 𝜎2 are not known,

𝑿1 −𝑿2
𝒁= ~ 𝑁(0, 1)
2
𝓈1 2
𝓈2
+
𝑛1 𝑛2
Example 7
TVS Company wanted to test the mileage of its two wheelers with that of other brands. A
random sample of 125 TVS make gave a mileage of 90 km. A random sample of 150 two
wheelers of all other brands gave a mileage of 80 km. It is known that the standard deviation
of both TVS Company and all other brands was 12 km. If significance is 5%, do TVS
vehicles give a better mileage?
Solution:
Aim: To test the average mileage of TVS two-wheelers with that of other brands is equal or more.
Sample mean -1 = 𝑋1 = 90 𝑘𝑚 , Sample size-1 = 𝑛1 = 125
Sample mean -2 = 𝑋2 = 80 𝑘𝑚 , Sample size-2 = 𝑛2 = 150
Population standard deviation = 𝜎1 = 𝜎2 = 𝜎 = 12
1. Null Hypothesis : 𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
2. Alternative hypothesis: 𝑯𝟏 : 𝝁𝟏 > 𝝁𝟐 (Right-tailed test)
3. Level of Significance = 𝜶 = 5% and 𝒛𝜶 = 𝟏. 𝟔𝟒𝟓
𝑿1 −𝑿2 𝟗𝟎 − 𝟖𝟎
4. Test statistic is given by: 𝒁= 1 1
= 1 1
= 6.88
𝝈 𝑛1
+𝑛 𝟏𝟐 +
2 125 150

𝒁𝒄𝒂𝒍 > |𝒛𝜶 | ⟹ 6.88 > 1.645

5. Accept the Alternative hypothesis 𝑯𝟏 : 𝝁𝟏 > 𝝁𝟐

Conclusion: the average mileage of TVS two wheelers is more than that of all other brands.
Example 8
The average hourly wage of a sample of 150 workers in a plant ‘A’ was
Rs. 2.56 with a s.d. of Rs. 1.08. The average hourly wage of a sample
of 200 workers in a plant ‘B’ was Rs. 2.87 with a s.d. of Rs. 1.28. Can
an applicant safety assume that the hourly wages paid by plant ‘B’ are
higher than those paid by a plant ‘A’?
Test for Single Proportion
𝒑 − 𝑬(𝒑) 𝒑 − 𝑷
𝒁= = ~ 𝑵(𝟎, 𝟏)
𝑺. 𝑬. (𝒑) 𝑃𝑄
𝑛

The test statistic for testing 𝑯𝟎 : 𝑷 = 𝑷𝟎

vs
the alternative hypothesis: 𝑯𝟏 : 𝑷 ≠ 𝑷𝟎 (𝑷 > 𝑷𝟎 or 𝑷 < 𝑷𝟎 )
(Two-tailed alternative) or

𝑯𝟏 : 𝑷 > 𝑷𝟎 (right-tailed alternative)

𝑯𝟏 : 𝑷 < 𝑷𝟎 (left-tailed alternative)

Example 5
Hindustan Lever Ltd. Company expects that more than 30% of the households
in Kolkata city will consume its product if they manufacture a new face cream.
A random sample of 500 households from the city is surveyed, 163 are
favorable in manufacturing the product. Examine whether the expectation of the
company would be met at 5% level.
Solution: Given, Population Proportion : 𝑷 = 𝟎. 𝟑 ; 𝑄 = 1 − 𝑃 = 0.7
𝑋 = No. of customers favourable in manufacturing the new product
𝑿 𝟏𝟔𝟑
Sample proportion: 𝒑 = = = 0.326 Sample size : n = 500
𝒏 𝟓𝟎𝟎
Aim : To test the HLL Company’s manufacture of a new product of face cream will be
consumed by 30% of the households in KOLKATA or more.

1. Null Hypothesis : 𝑯𝟎 : 𝑷 ≥ 𝟎. 𝟑
2. Alternative hypothesis: 𝑯𝟏 : 𝑷 < 𝟎. 𝟑 (Left-tailed test)

3. Level of Significance (𝜶) = 5% and 𝒛𝜶 = −𝟏. 𝟔𝟒𝟓

𝒑−𝑷 𝟎.𝟑𝟐𝟔 − 𝟎.𝟑

4. Test statistic is given by: 𝒁 = = = 1.26
𝑃𝑄 𝟎.𝟑×𝟎.𝟕
𝑛 𝟓𝟎𝟎

𝒁𝒄𝒂𝒍 < |𝒛𝜶 | ⟹ 1.26 < 1.645

5. Accept the Null hypothesis 𝑯𝟎 : 𝑷 ≥ 𝟎. 𝟑
i.e., the HLL Company’s manufacture of a new product of face cream will be consumed by 30% of the
households in Kolkata.
Example 6
Twenty people were attacked by a disease and only 18 survived. Will you
reject the hypothesis that the survival rate, if attacked by this disease, is 85%
in favour of the hypothesis that it is more, at 5% level of significance.
 Test for Significance of Difference of Two
Population Proportions
Aim: To test the two population proportions 𝑷𝟏 and 𝑷𝟐 be equal, based on two
random samples. i.e., to investigate the significance of the difference
between the two sample proportions 𝒑𝟏 and 𝒑𝟐 .

Source: Draw two independent random samples, 𝑛1 from the population-1 with
sample proportion 𝑝1 and 𝑛2 from the population-2 with sample
proportion 𝑝2
The test statistic for testing 𝑯𝟎 : 𝑷𝟏 = 𝑷𝟐 (𝑷𝟏 − 𝑷𝟐 = 𝟎 )

vs

the alternative hypothesis: 𝑯𝟏 : 𝑷𝟏 ≠ 𝑷𝟐 (𝑷𝟏 > 𝑷𝟐 or 𝑷𝟏 < 𝑷𝟐 )

(Two-tailed alternative) or
𝑯𝟏 : 𝑷𝟏 > 𝑷𝟐 (right-tailed alternative)
𝑯𝟏 : 𝑷𝟏 < 𝑷𝟐 (left-tailed alternative)
 𝑝1 −𝑝2 − (𝑃1 −𝑃2 )
The test statistic is : 𝒁= 𝟏 𝟏
~ 𝑁(0, 1)
𝑷𝑸 +
𝒏𝟏 𝒏𝟐

𝑝1 − 𝑝2
𝒁= 𝟏 𝟏
𝑷𝑸 +
𝒏𝟏 𝒏𝟐

𝑛1 𝑝1 +𝑛2 𝑝2 𝑋1 +𝑋2
Where 𝑃 = =
𝑛1 +𝑛2 𝑛1 +𝑛2
Example 10
Before an increase in excise duty on tea, 800 persons out of sample of
1000 persons were found to be tea drinkers. After an increase in duty, 800
people were tea drinkers in a sample of 1200 people. Using standard error
of proportion, state whether there is a significant decrease in the
consumption of tea after the increase in excise duty?
Solution: Aim: To test whether decrease in the consumption of tea after increase in excise duty.
𝑛1 = 1000 , 𝑛2 = 1200
800
𝑝1 = Sample proportion of tea drinkers before increase in excise duty = = 0.8
1000
800
𝑝2 = Sample proportion of tea drinkers after increase in excise duty = = 0.67
1200
𝑛1 𝑝1 +𝑛2 𝑝2 800+800 16 6
𝑃= = = and 𝑄 =1−𝑃 =
𝑛1 +𝑛2 1000+1200 22 22

1. Null Hypothesis : 𝑯𝟎 : 𝑷𝟏 = 𝑷𝟐
2. Alternative hypothesis: 𝑯𝟏 : 𝑷𝟏 > 𝑷𝟐 (Right-tailed test)
3. Level of Significance = 𝜶 = 5% and 1% ⇒ 𝒛𝟎.𝟎𝟓 = 𝟏. 𝟔𝟒𝟓 , 𝒛𝟎.𝟎𝟏 = 𝟐. 𝟑𝟑
𝑝1 − 𝑝2 𝟎.𝟖 − 𝟎.𝟔𝟕
4. Test statistic is given by: 𝒁 = = = 6.842
𝟏 𝟏 16 6 𝟏 𝟏
𝑷𝑸 𝒏𝟏
+𝒏 ×
22 22 𝟏𝟎𝟎𝟎
+𝟏𝟐𝟎𝟎
𝟐

𝒁𝒄𝒂𝒍 > |𝒛𝟎.𝟎𝟓 | ⟹ 6.842 > 1.645 and 𝒁𝒄𝒂𝒍 > |𝒛𝟎.𝟎𝟏 | ⟹ 6.842 > 2.33
It is highly significant at both 5% and 1% level of significance.
5. Accept the Alternative hypothesis 𝑯 𝟏 : 𝑷𝟏 > 𝑷𝟐
Conclusion: There is a significance decrease in the consumption of tea after increase in excise duty.
Example 11
In a large city A, 20% of a random sample of 900 school children had
defective eye-sight. In other city B, 15% of a random sample of 1600
children had the same defect. Is this difference between the two proportions
significant ?