CIVL101

Lecture-29

Test of significance of large samples

Z-Statistic
Unit 6: Applications of Probability and Statistics

Learning Outcomes:

To calculate Z-Statistic for testing of hypothesis

(Test of significance of large samples)

 Z-Statistic

For a large sample, if 𝑡 is any statistic, then it follows a normal distribution.

The corresponding population parameter is mean = 𝐸(𝑡) and standard

deviation = 𝑆𝐸(𝑡).

𝑡−𝐸(𝑡)
For large samples: 𝑍 =
𝑆𝐸(𝑡)

Which is called the test statistic(Z-statistic).

Case Test of significance of SE = 𝝈∗ Statistic-Z
difference between:
1. Sample mean 𝑥ҧ and 𝜎/ 𝑛 𝑥ҧ − 𝜇
𝑧=
population mean 𝜇 𝜎∗
2. Means of two samples 𝑥1 𝑥1 − 𝑥2
𝜎2 𝜎2 𝑧=
and 𝑥2 + 𝜎∗
𝑛1 𝑛2

3. Sample standard deviation 𝜎/ 2𝑛 𝑠−𝜎

𝑧=
𝑠 and population 𝜎∗
standard deviation 𝜎
Case Test of significance of SE = 𝝈∗ Statistic-Z
difference between:
4. Samples standard 𝑠1 − 𝑠2
𝜎12 𝜎22 𝑧=
deviations 𝑠1 , 𝑠2 + 𝜎∗
2𝑛1 2𝑛2

5. Sample proportion 𝑝 and 𝑃𝑄/𝑛 𝑝−𝑃

𝑧=
Population proportion 𝑃 𝜎∗

6. Two sample proportions 𝑝1 − 𝑝2

𝑃𝑄 𝑃𝑄 𝑧=
𝑝1 , 𝑝2 + 𝜎∗
𝑛1 𝑛2
Problem 1. In a large lot of electric bulbs, the mean life and standard
deviation of the bulbs are 360 hours and 90 hours respectively. A sample of
625 bulbs is chosen. It is found that mean life and standard deviation of the
bulbs in the sample are 355 hours and 90 hours respectively. Can we
conclude that the sample is drawn from the given population? Test at 5 %
level of significance. [ At 5 % level 𝑍 = 1.96].

Solution. Define Null Hypothesis 𝐻0 : The sample is drawn from the

population with mean 𝜇 = 360 hours and S.D. 𝜎 = 90 hours.

Alternate Hypothesis 𝐻1 : 𝜇 ≠ 360 hours.

We shall use the two-tailed test.

From the given data: 𝑥ҧ = 355, 𝜇 = 360, 𝑛 = 625, 𝜎 = 90.

Define the test statistics as:

ҧ
𝑥−𝜇
𝑍= where 𝜎 ∗ = 𝜎/ 𝑛
𝜎∗

355 − 360
⟹𝑍= = −1.389
90
625
Since, 𝑍 = 1.389 < 1.96 at 5% level of significance.

So, we accept the null hypothesis that the sample is drawn from the population.
 Polling Quiz
In a large lot of electric bulbs, the mean life and standard deviation of the bulbs are
240 hours and 90 hours respectively. A sample of 400 bulbs is chosen. It is found
that mean life and standard deviation of the bulbs in the sample are 250 hours and
90 hours respectively. The value of Z- statistic will be:

(A) 2.222

(B) 1.111

(C) 3.333

(D) 2.112
Problem 2. The sizes and means of two independent random samples are
400, 225; 3.5 and 3.0 respectively. Can we conclude that the samples are
drawn from the same population with standard deviation 1.5 [ At 1 % level
𝑍 = 3].

Solution. Define Null Hypothesis 𝐻0 : 𝜇1 = 𝜇2

That is the samples are drawn from the same population.

Alternate Hypothesis 𝐻1 : 𝜇1 ≠ 𝜇2

We shall use the two-tailed test.

From the given data: 𝑛1 = 400, 𝑥1 = 3.5, 𝑛2 = 225, 𝑥1 = 3.0, 𝜎 = 1.5

Define the test statistics as:

𝑥1 −𝑥2 𝜎2 𝜎2
𝑧= where 𝜎 ∗ = +
𝜎∗ 𝑛1 𝑛2

3.5 − 3.0
⟹𝑍= = 4.0
1 1
1.5 +
400 225
Since, 𝑍 = 4 > 3 at 1% level of significance.

So, we reject the null hypothesis that the sample are drawn from the same
population.
Problem 3. The number of students in a class is 100. The average marks scored
by 64 boys is 66 with standard deviation of 10 while the average marks scored
by 36 girls is 70 with standard deviation of 8. Test at 2 % level of significance
whether the girls performed better than boys. [ At 2 % level 𝑍 = 2.33].

Solution. Define Null Hypothesis 𝐻0 : 𝜇1 = 𝜇2

That is boys and girls performed equally well.

Alternate Hypothesis 𝐻1 : 𝜇1 > 𝜇2 that is girls performed well

We shall use the right-tailed test.

Here: 𝑛1 = 36, 𝑥1 = 70, 𝜎1 = 8, 𝑛2 = 64, 𝑥1 = 66, 𝜎2 = 10

Define the test statistics as:

𝑥1 −𝑥2 𝜎12 𝜎22

𝑧= where 𝜎 ∗ = +
𝜎∗ 𝑛1 𝑛2

70 − 66
⟹𝑍= = 2.189
64 100
+
36 64
Since, 𝑍 = 2.189 < 2.33 at 2% level of significance.

So, we accept the null hypothesis that boys and girls performed equally
well. We can not conclude that the girls performed better.
 Polling Quiz
The mean height of 80 boys, who participated in the athletic competition in a
college was 167cm with standard deviation of 9cm. The mean height of
remaining 160 boys who did not participate was 163cm with standard
deviation of 10cm. The value of Z- statistic will be:

(A) 2.361

(B) 3.621

(C) 3.126

(D) 2.335
Problem 4. In a random sample of 200 people in a city, 108 like to
purchase imported watches and remaining like to purchase local watches.
Can we conclude that both the imported and local watches are popular in
the city? Test at 2 % level of significance. [ At 2 % level 𝑍 = 2.33].

Solution. We have 𝑛 = 200.

Sample proportion of people who would like to purchase imported

108
watches, 𝑝 = = 0.54
200

Define Null Hypothesis 𝐻0 : Both imported and local watches are popular.

Alternate Hypothesis 𝐻1 : 𝑃 ≠ 0.5 (Two-tailed test)

Define the test statistics as:

𝑝−𝑃
𝑧= where 𝜎 ∗ = 𝑃𝑄/𝑛
𝜎∗

0.54 − 0.5
⟹𝑍= = 1.13
(0.5)(0.5)/200

Since, 𝑍 = 1.13 < 2.33 at 2% level of significance.

So, we accept the null hypothesis that both imported and local watches are
popular.
Problem 5. A manufacturer of an electronic item finds the standard
deviation of the life of the items to be 60 hours. The manufacturer wants to
adopt a new process for producing the same items, which improves the life
time of the items. A random sample of 200 items produced by the new
process is selected and it was found that the standard deviation is 52 hours.
Should the manufacturer adopt the new process? Test at 2 % level of
significance. [ At 2 % level 𝑍 = 2.055].

Solution. We have 𝑛 = 200, 𝑠 = 52, 𝜎 = 60.

Define Null Hypothesis 𝐻0 : 𝑠 = 𝜎.

Alternate Hypothesis 𝐻1 : 𝑠 < 𝜎 (Left-tailed test)

Define the test statistics as:
𝑠−𝜎
𝑧= where 𝜎 ∗ = 𝜎/ 2𝑛
𝜎∗

52 − 60
⟹𝑍= = −2.667
60
400
Since, 𝑍 = 2.667 > 2.055 at 2% level of significance.

So, we reject the null hypothesis and accept the alternative hypothesis. The
manufacturer is justified in adopting the new process.
Problem 6. Before a big increase in the price of petrol, 400 persons out of a
sample of 1000 persons were found to purchase big sized cars. After the
increase in the price of petrol, 280 persons out of a sample of 800 persons
were found to purchase big sized cars. Find whether there is a significant
decrease in the purchase of big cars. Test at 5 % level of significance.

[ At 5 % level 𝑍 = 1.645].
400 280
Solution. We have 𝑛1 = 1000, 𝑝1 = = 0.4, 𝑛2 = 800, 𝑝2 = = 0.35
1000 800

Define Null Hypothesis 𝐻0 : 𝑝1 = 𝑝2 (No significant decrease)

Alternate Hypothesis 𝐻1 : 𝑝1 > 𝑝2 (Right-tailed test)

 400 280
We have 𝑛1 = 1000, 𝑝1 = = 0.4, 𝑛2 = 800, 𝑝2 = = 0.35
1000 800

Since the population proportion is not given, so we approximate:

𝑛1 𝑝1 + 𝑛2 𝑝2 400 + 280
𝑃= = = 0.3788
𝑛1 + 𝑛2 1800
𝑄 = 1 − 𝑃 = 1 − 0.3788 = 0.6222

Define the test statistics as:

𝑝1 −𝑝2 𝑃𝑄 𝑃𝑄
𝑧= where 𝜎∗ = +
𝜎∗ 𝑛1 𝑛2
 0.4 − 0.35
⟹𝑍= = 2.17
1 1
(0.3778)(0.6222) +
1000 800
Since, 𝑍 = 2.17 > 1.645 at 5% level of significance.

So, we reject the null hypothesis and accept the alternative hypothesis. We
conclude that there is a significant decrease in the purchase of big sized
cars .