3.

1 Sampling

Introduction:

Entire group of individuals under study.

Population
Example: Set of all students in the college.
Quantity associated with population like mean(𝜇), SD(𝜎).
Parameter
Example: Mean weight of students in the college.
A small part of the population.
Sample
Example: Set of randomly selected 50 students from the college.
Quantity associated with sample like mean(𝑥̅ ), SD(s).
Statistic
Example: Mean weight of 50 students from the college.
Sample size The number of units in the sample.
Large sample Sample size 𝑛 ≥ 30.
Small sample Sample size 𝑛 < 30.
Sampling
distribution The frequency distribution of means of different samples
of mean
standard The standard deviation of the sampling distribution of mean.
error of mean
Notation: 𝑆𝐸(𝑥̅ ).
Precision The reciprocal of the standard error

Dr. Narasimhan G, RNSIT 1

Testing of hypothesis and level of significance:

Statistical hypothesis Some assumption (statement) about the population based on

sample information which may or may not be true.
Testing a hypothesis A process to decide whether to accept or reject the hypothesis.
Null hypothesis (H0) A statistical hypothesis which we formulate to check whether it
can be rejected.
Alternative hypothesis (H1) The negation of the null hypothesis.
Type I error Rejecting H0 when it is true.
Level of significance (𝜶) Probability of type I error.
Type II error Accepting H0 when it is false.
Power of the test (𝜷) Probability of type II error.

True False

Accept H0 Correct decision Type II error

Reject H0 Type I error Correct decision

Dr. Narasimhan G, RNSIT 2

Confidence limits and confidence intervals:

Critical region The region in which the calculated sample value falling is rejected.
Acceptance region The region in which the calculated sample value falling is accepted.
Level of significance The probability of calculated sample value falling in the critical region
(or) rejection region.
The limits of the critical region. Critical value splits the region in to
Critical values
acceptance region and critical region. These are pre-assigned values.
An interval which is likely to contain the calculated sample value.
confidence interval
Example: In BP measurement, (80, 120) is the confidence interval.
The limits of the confidence interval.
confidence limits
Example: In BP measurement, 80 and 120 are confidence limits.
confidence coefficient The probability that the confidence interval contains the calculated
(𝟏 − 𝜶) sample value.

Simple sampling attributes:

An attribute means quality or characteristic such as drinking, smoking, disease, etc. An attribute
may be marked by its presence (K) or absence (not K) in a member of given population. The
sampling of attributes may be regarded as the selection of samples from population whose
members posses the attribute K or not K. The presence of K is the success and its absence a
failure. Suppose we draw a simple sample of size 𝑛 items, it follows binomial distribution and
hence the mean of this distribution is 𝑛𝑝 and standard deviation of this distribution is √𝑛𝑝𝑞.

Dr. Narasimhan G, RNSIT 3

 3.2 Test of significance for large samples

Introduction:
How to find standard error? 𝑆. 𝐸 (𝑥̅ ) = √𝑠 2 /𝑛 if 𝑠 is known

𝑆. 𝐸 (𝑥̅ ) = √𝜎 2 /𝑛 if 𝜎 is known

𝑆. 𝐸 (𝑝) = √𝑝𝑞/𝑛 if 𝑝 is known

𝑆. 𝐸 (𝑃) = √𝑃𝑄/𝑛 if 𝑃 is known

How to find confidence interval?

𝑥̅ ± 3[𝑆𝐸(𝑥̅ )] if 𝑥̅ is known.
𝑃 ± 3[𝑆𝐸(𝑃)] if 𝑃 is known.
𝑝 ± 3[𝑆𝐸(𝑝)] if 𝑝 is known.

How to find confidence interval at 𝜶 level of significance?

𝑥̅ ± 𝑧𝛼 [𝑆𝐸(𝑥̅ )] if 𝑥̅ is known.
2

𝑃 ± 𝑧𝛼 [𝑆𝐸(𝑃)] if 𝑃 is known.
2

𝑝 ± 𝑧𝛼 [𝑆𝐸(𝑝)] if 𝑝 is known.
2

What is critical value?

5% 1%
𝑍𝛼 1.96 2.58
𝑍𝛼/2 1.64 2.33

Working rule:
❖ Write the null hypothesis 𝐻0 : 𝜇 = 𝜇0 (𝑜𝑟) 𝑃 = 𝑃0
❖ Find calculated value:
𝑥̅ −𝜇
|𝑧| = | | , if 𝑥̅ is known,
𝑆.𝐸(𝑥̅ )
𝑝−𝑃
|𝑧| = | | , if 𝑝 is known.
𝑆.𝐸(𝑝)

❖ Find the critical value using the above table.

❖ If calculated value < critical value, accept H0. H0 is the conclusion.
If calculated value > critical value reject H0. H1 is the conclusion.

Dr. Narasimhan G, RNSIT 4

1. A sample of 900 members is found to have a mean of 3.4 cm. Can it be reasonably
regarded as a truly random sample from a large population with mean 3.25 cm and
SD 1.61 cm.
Since 𝑛 = 900, this is the large sample. Apply z test.
By data, 𝑥̅ = 3.4, 𝜇 = 3.25, 𝜎 = 1.61
𝐻0 : 𝜇 = 3.25, Sample is taken from the population with mean 3.25
To find: Calculated value

𝜎2 1.612
𝑆. 𝐸(𝑥̅ ) = √ =√ = 0.0537
𝑛 900
𝑥̅ − 𝜇 3.4 − 3.25
|𝑧| = | |=| | = 2.8
𝑆. 𝐸(𝑥̅ ) 0.0537
Therefore, calculated value of 𝑧 = 2.8
To find: Critical value
At 𝛼 = 0.05, critical value of 𝑧 = 1.96
Conclusion:
Since calculated value > critical value, Reject H0.
Therefore, sample is not taken from the population with mean 3.25

Dr. Narasimhan G, RNSIT 5

2. A coin was tossed 400 times and the head turned up 216 times. Test the hypothesis
that the coin is unbiased at 5% level of significance. [𝒛𝜶 = 𝟏. 𝟗𝟔].
𝟐

Since 𝑛 = 400, this is the large sample. Apply z test.

216 1
By data, 𝑝 = 400 and 𝑃 = 2 = 0.5

𝐻0 : 𝑃 = 0.5 ,The coin is unbiased.

To find: Calculated value

𝑃𝑄 (0.5)(0.5)
𝑆𝐸(𝑃) = √ =√ = 0.025
𝑛 400
216 1
𝑝−𝑃 −
|𝑧| = | | = |400 2| = 1.6
𝑆𝐸(𝑃) 0.025

Therefore, calculated value of 𝑧 = 1.6

To find: Critical value
At 𝛼 = 0.05, critical value of 𝑧 = 1.96
Conclusion:
Since calculated value < critical value, Accept H0.
Therefore, the coin is unbiased at 5% level of significance.

Dr. Narasimhan G, RNSIT 6

3. A die was thrown 9000 times and a throw of 5 or 6 was obtained 3240 times. On the
assumption of random throwing, do the data indicate an unbiased die? (𝜶 = 𝟎. 𝟎𝟏)
[𝒛𝜶 = 𝟐. 𝟓𝟖]
𝟐
Since 𝑛 = 9000, this is the large sample and apply z test.
3240 2 1
By data, 𝑝 = 9000 and 𝑃 = 6 = 3
1
𝐻0 : 𝑃 = 3 ,The die is unbiased.

To find: Calculated value

𝑃𝑄 (1/3)(2/3)
𝑆𝐸(𝑃) = √ =√ = 0.005
𝑛 9000
3240 1
𝑝−𝑃 −
|𝑧| = | | = |9000 3| = 5.33
𝑆𝐸(𝑃) 0.005

Therefore, calculated value of 𝑧 = 5.33

To find: Critical value
At 𝛼 = 0.01, critical value of 𝑧 = 2.58
Conclusion:
Since calculated value > critical value, Reject H0.
Therefore, the die is biased at 1% level of significance.

Dr. Narasimhan G, RNSIT 7

4. In 324 throws of a die, an odd number turned up 181 times. Is it reasonable to think
that ay 1% level of significance the die is an unbiased one? [𝒛𝜶 = 𝟐. 𝟓𝟖]
𝟐
Since 𝑛 = 324, this is the large sample and apply z test.
181 3 1
By data, 𝑝 = 324 and 𝑃 = 6 = 2
1
𝐻0 : 𝑃 = 3 ,The die is unbiased.

To find: Calculated value

𝑃𝑄 (1/2)(1/2)
𝑆𝐸(𝑃) = √ 𝑛 = √ = 0.0278
324

181 1
𝑝−𝑃 −
|𝑧| = | | = |324 2| = 2.1084
𝑆𝐸(𝑃) 0.0278

Therefore, calculated value of 𝑧 = 2.1084

To find: Critical value
At 1% level of significance, critical value of 𝑧 = 2.58
Conclusion:
Since calculated value < critical value, accept H0.
Therefore, the die is unbiased at 1% level of significance.

Note: The die is biased at 5% level of significance. (∵ critical value of 𝑧 = 1.96)

Dr. Narasimhan G, RNSIT 8

5. The mean of a certain normal population is equal to the standard error of the mean
of the samples of 100 from that distribution. Find the probability that the mean of
the sample of 25 from the distribution will be negative.
To find: 𝜇
By data, 𝜇 = 𝑆𝐸 (𝑥̅ ) when 𝑛 = 100.
𝜎2 𝜎2
𝜇 = 𝑆𝐸(𝑥̅ ) = √ 𝑛 = √100
𝜎
Therefore, 𝜇 = 10 ---- (1)

To find: 𝑆𝐸(𝑥̅ ) when 𝑛 = 25.

𝜎2 𝜎2
𝑆𝐸(𝑥̅ ) = √ 𝑛 = √25
𝜎
Therefore, 𝑆𝐸(𝑥̅ ) = ----- (2)
5

To find: 𝑃(𝑥̅ < 0)

𝑥̅ −𝜇 0−𝜇
𝑃(𝑥̅ < 0) = 𝑃 (𝑆.𝐸(𝑥̅ ) < 𝑆.𝐸(𝑥̅ ))
−𝜇
= 𝑃 (𝑧 < 𝑆.𝐸(𝑥̅ ))

Substituting (1) and (2),

𝜎/10
𝑃(𝑥̅ < 0) = 𝑃 (𝑧 < − )
𝜎/5
1
= 𝑃 (𝑧 < − 2)

= 0.3085

Dr. Narasimhan G, RNSIT 9

6. If a mean breaking strength of copper wire is 575 lbs with a standard deviation 8.3
lbs. How large a sample must be used in order that there be one chance in 100 that
the mean breaking strength of the sample is less than 572 lbs. (𝒁𝜶 = 𝟐. 𝟑𝟑)

By data, 𝑥̅ = 572, 𝜇 = 575, 𝜎 = 8.3.

𝐻0 : 𝜇 = 575, 𝑚ean breaking strength of copper wire is 575 lbs.
To find: Calculated value
𝜎2 8.32
𝑆. 𝐸(𝑥̅ ) = √ 𝑛 = √ 𝑛

𝑥̅ −𝜇 572−575
|𝑧| = | |=| |
𝑆.𝐸(𝑥̅ ) 2
√8.3
𝑛

Given: Critical value is 2.33

To find: 𝑛 such that 𝜇 < 572
𝜇 < 572
𝐻0 is rejected.
Calculated value > Critical value.

572 − 575
|| || > 2.33
2
√8.3
𝑛
On simplifying, 𝑛 > 41.56
Therefore, 𝑛 = 42.

Dr. Narasimhan G, RNSIT 10

 Note: Confidence interval is given by 𝑝 ± 𝑧𝛼 [𝑆𝐸(𝑝)] if 𝑝 is known.
2

7. An unbiased coin is thrown n times. It is desired that the relative frequency of the
appearance of heads should lie between 0.49 and 0.51. Find the smallest value of n
that will ensure this result with 90% confidence. (𝒁𝜶 = 𝟏. 𝟔𝟒𝟓)
𝟐

By data, 𝑝 = 0.5, 𝑞 = 0.5 𝑎𝑛𝑑 𝑧𝛼/2 = 1.645 ---- (1)

To find: Standard error of proportion
𝑝𝑞 1
𝑆. 𝐸(𝑝) = √ 𝑛 = 2√𝑛 --------------------- (2)

By data, 90% confidence interval = (0.49, 0.51)

𝑝 ± 𝑍𝛼 . 𝑆𝐸(𝑝) = (0.49, 0.51)
2

(𝑝 − 𝑍𝛼 . 𝑆𝐸(𝑝), 𝑝 + 𝑍𝛼 . 𝑆𝐸(𝑝)) = (0.49, 0.51)

2 2

Substituting (1) and (2),

1.645 1.645
(0.5 − , 0.5 + ) = (0.49, 0.51)
2√𝑛 2√𝑛
Therefore, the smallest value of n is given by
1.645
0.5 + = 0.51
2 √𝑛
1
1.645 2 = 0.01
√𝑛

𝑛 = 6765

Dr. Narasimhan G, RNSIT 11

8. A random sample of 500 pineapples was taken from a large consignment and 65 were
found to be bad. Show that the standard error of the proportion of bad ones in a
sample of this size is 0.015 and deduce that the percentage of bad pineapples in the
consignment almost certainly lies between 8.5 and 17.5.
65
By data, 𝑝 = 500 = 0.13

𝑝𝑞 0.13×0.87
𝑆. 𝐸(𝑝) = √ 𝑛 = √ = 0.015
500

Required confidence interval= (𝑝 − 3[𝑆𝐸(𝑝)], 𝑝 + 3[𝑆𝐸(𝑝)])

= (0.13 − 3(0.015), 0.13 + 3(0.015))
= (0.085, 0.175)
= (8.5%, 17.5%)

9. In a locality containing 18000 families, a sample of 840 families was selected at

random. Of these 840 families, 206 families were found to have a monthly income of ₹
25,000 or less.It is desired to estimate how many out of 18,000 families have a monthly
income of ₹ 25,000 or less. Within what limits would you place your estimate?

206
By data, 𝑝 = 840 = 0.2452

𝑝𝑞 (0.2452)(0.7548)
𝑆𝐸(𝑝) = √ =√ = 0.0148
𝑛 840

Required confidence limits = 𝑝 ± 3[𝑆𝐸(𝑝)]

= 𝑝 − 3[𝑆𝐸(𝑝)] 𝑎𝑛𝑑 𝑝 + 3[𝑆𝐸(𝑝)]
= 0.2452 − 3(0.0148) 𝑎𝑛𝑑 0.2452 + 3(0.0148)
= 0.2452 − 0.0444 𝑎𝑛𝑑 0.2452 + 0.0444
= 0.2008 𝑎𝑛𝑑 0.2896

Dr. Narasimhan G, RNSIT 12

 3.3 Comparison of large samples

How to find standard error?

𝑠12 𝑠22
√ + , 𝐼𝑓 𝑠1 , 𝑠2 𝑎𝑟𝑒 𝑘𝑛𝑜𝑤𝑛
𝑛1 𝑛2

𝜎12 𝜎22
𝑆𝐸(𝑥
̅̅̅1 − ̅̅̅)
𝑥2 √ + , 𝐼𝑓 𝜎1 , 𝜎2 𝑎𝑟𝑒 𝑘𝑛𝑜𝑤𝑛
𝑛1 𝑛2

1 1
𝜎√ + , 𝐼𝑓 𝜎 𝑖𝑠 𝑘𝑛𝑜𝑤𝑛
𝑛1 𝑛2

𝑃1 𝑄1 𝑃2 𝑄2
√ + , 𝐼𝑓 𝑃1 , 𝑃2 𝑎𝑟𝑒 𝑘𝑛𝑜𝑤𝑛
𝑛1 𝑛2
𝑆𝐸(𝑝1 − 𝑝2 )
1 1
√𝑃𝑄 ( + ) , 𝐼𝑓 𝑝1 , 𝑝2 𝑎𝑟𝑒 𝑘𝑛𝑜𝑤𝑛,
𝑛1 𝑛2
𝑛1 𝑝1 + 𝑛2 𝑝2
𝑤ℎ𝑒𝑟𝑒, 𝑃 =
𝑛1 + 𝑛2
Working rule:
❖ Write the null hypothesis 𝐻0 : 𝜇1 = 𝜇2 (𝑜𝑟) 𝑃1 = 𝑃2 .
❖ Find calculated value
̅𝑥̅̅̅−𝑥
̅̅̅̅
|𝑆𝐸(𝑥1̅̅̅̅−𝑥2̅̅̅̅)| , if ̅̅̅,
𝑥1 ̅̅̅
𝑥2 are known
1 2
|𝑧| = { 𝑝 −𝑝
1 2
|𝑆𝐸(𝑝 −𝑝 )| , if ̅̅̅,
𝑥1 ̅̅̅
𝑥2 are not known
1 2

❖ Find the critical value using the table.

❖ If calculated value < critical value, accept H0. H0 is the conclusion.
❖ If calculated value > critical value reject H0. H1 is the conclusion.

Dr. Narasimhan G, RNSIT 13

1. The means of samples of sizes 1000 and 2000 are 67.5 and 68.0 cms respectively. Can
the samples be regarded as drawn from the same population of SD 2.5 cm?
[ 𝒛𝜶 (𝟎. 𝟎𝟓) = 𝟏. 𝟗𝟔]
𝟐

Since samples sizes are 𝑛1 = 1000, 𝑛2 = 2000, apply z test.

By data, ̅̅̅
𝑥1 = 67.5, ̅̅̅
𝑥2 = 68.0 𝑎𝑛𝑑 𝜎 = 2.5
H0: 𝝁𝟏 = 𝝁𝟐 , Both the samples are drawn from the same population.
To find: Calculated value

1 1
𝑆𝐸(𝑥
̅̅̅1 – ̅̅̅̅)
𝑥2 = 𝜎√ +
𝑛1 𝑛2

1 1
= 2.5√1000 + 2000 = 0.0968

̅̅̅1 – ̅̅̅̅
𝑥 𝑥2
|𝑧| = | |
𝑆𝐸(𝑥 ̅̅̅1 – ̅̅̅̅)
𝑥2
67.5−68.0 0.5
=| | = 0.0968 = 5.16
0.0968

Therefore, calculated value of 𝑧 = 5.16

To find: Critical value
A𝑡 𝛼 = 0.05, critical value of 𝑧 = 1.96
Conclusion:
Since calculated value > critical value, reject 𝐻0 .
Therefore, Both the samples are not drawn from the same population.

Dr. Narasimhan G, RNSIT 14

2. A sample of height of 6400 soldiers has a mean of 67.85 inches and a standard
deviation of 2.56 inches while a sample of height of 1600 sailors has a mean of 68.55
inches and a SD of 2.52 inches. Does the data indicate that the sailors are on an
average taller than soldiers? Use 0.05 level of significance. [𝒛𝜶 = 𝟏. 𝟔𝟓]

Since samples sizes are 𝑛1 = 6400, 𝑛2 = 1600, apply z test.

By data, ̅̅̅
𝑥1 = 67.85, 𝑥
̅̅̅2 = 68.55, 𝑠1 = 2.56, 𝑠2 = 2.52
H0: 𝝁𝟏 = 𝝁𝟐 , The sailors are not taller than soldiers.
To find: Calculated value

𝑠12 𝑠22
𝑆𝐸(𝑥 𝑥2 = √
̅̅̅1 – ̅̅̅̅) +
𝑛1 𝑛2

2.562 2.522
= √ 6400 + 1600

= 0.0707
̅̅̅1 – ̅̅̅̅
𝑥 𝑥2
|𝑧| = | |
𝑆𝐸(𝑥 ̅̅̅1 – ̅̅̅̅)
𝑥2
67.85−68.55 0.7
=| | = 0.0707 = 9.9
0.0707

Therefore, calculated value = 9.9

To find: Critical value
At 𝛼 = 0.05, critical value = 1.65
Conclusion:
Since calculated value > critical value, reject 𝐻0 .
Therefore, the sailors are taller than soldiers at 0.05 level of significance.

Dr. Narasimhan G, RNSIT 15

3. A sample of 100 electric bulbs produced by manufacturer A showed a mean lifetime of 1190
hours and a standard deviation of 90 hours. A sample of 75 bulbs produced by manufacturer
B showed a mean lifetime of 1230 hours with a standard deviation of 120 hours. Is there a
difference between the mean lifetime of two brands at significant level of 0.05? (𝒁𝜶/𝟐 = 𝟏. 𝟗𝟔)

Since samples sizes are 𝑛1 = 100, 𝑛2 = 75, apply z test.

By data, ̅̅̅
𝑥1 = 1190, ̅̅̅
𝑥2 = 1230, 𝑠1 = 90, 𝑠2 = 120
H0: 𝝁𝟏 = 𝝁𝟐 , There is no difference between the mean lifetime of two brands.
To find: Calculated value

𝑠12 𝑠22
𝑆𝐸(𝑥 𝑥2 = √ +
̅̅̅1 – ̅̅̅̅)
𝑛1 𝑛2

902 1202
= √100 + = 16.5227
75

̅̅̅1 – ̅̅̅̅
𝑥 𝑥2
|𝑧| = | |
𝑆𝐸(𝑥 ̅̅̅1 – ̅̅̅̅)
𝑥2
1190−1230
=| | = 2.4209
16.5227

To find: Critical value

Therefore, calculated value = 2.4209
At 𝛼 = 0.05, critical value = 1.96
Conclusion:
Since calculated value > critical value, reject 𝐻0 .
Therefore, there is a difference between the mean lifetime of two brands at
significant level of 0.05.

Dr. Narasimhan G, RNSIT 16

4. One type of aircraft is found to develop engine trouble in 5 flights out of a total of 100 and
another type in 7 flights out of a total of 200 flights. Is there a significant difference in the two
types concerned so far as engine defects are concerned? (𝒁𝜶/𝟐 = 𝟏. 𝟗𝟔)

Since samples sizes are 𝑛1 = 100 𝑎𝑛𝑑 𝑛2 = 200, apply z test.

5 7
By data, 𝑝1 = 100 , 𝑝2 = 200 , 𝛼 = 0.05

𝐻0 : 𝑃1 = 𝑃2 , There is no significant difference in the two types concerned so far as engine defects
are concerned.
To find: Calculated value
𝑛1 𝑝1 + 𝑛2 𝑝2
𝑃=
𝑛1 + 𝑛2
100(0.05)+200(0.035)
= 100+200
= 0.04

1 1
𝑆𝐸(𝑝1 − 𝑝2 ) = √𝑃𝑄 ( + )
𝑛1 𝑛2

1 1
= √0.04 × 0.96 × (100 + 200)

= 0.024
𝑝1 − 𝑝2
|𝑧| = | |
𝑆𝐸(𝑝1 − 𝑝2 )
0.015
= 0.024 = 0.625

Therefore, calculated value = 0.625

To find: Critical value
At 𝛼 = 0.05, critical value = 1.96
Conclusion:
Since calculated value < critical value, Accept 𝐻0 .
Therefore, there is no significant difference in the two types concerned so far as
engine defects are concerned.

Dr. Narasimhan G, RNSIT 17

5. A machine produces 16 imperfect articles in a sample of 500. After the machine is
overhauled, it produces 3 imperfect articles in a batch of 100. Has the machine been
improved? [𝒛𝜶 = 𝟏. 𝟔𝟓]

Since samples sizes are 𝑛1 = 500 𝑎𝑛𝑑 𝑛2 = 100, apply z test.

16 3
By data, 𝑝1 = 500 = 0.032, 𝑝2 = 100 = 0.03

H0: 𝑷𝟏 = 𝑷𝟐 , The machine has not been improved.

To find: Calculated value
𝑛1 𝑝1 + 𝑛2 𝑝2
𝑃=
𝑛1 + 𝑛2
500(0.032) + 100(0.03)
= = 0.0317
500 + 100

1 1
𝑆𝐸(𝑝1 − 𝑝2 ) = √𝑃𝑄 ( + )
𝑛1 𝑛2

1 1
= √0.0317 × 0.9683 × ( + )
500 100

= 0.0192
𝑝1 − 𝑝2
|𝑧| = | |
𝑆𝐸(𝑝1 − 𝑝2 )
0.032−0.03 0.002
=| | = 0.0192 = 0.1042
0.0192

Therefore, calculated value = 0.1042

To find: Critical value
At 𝛼 = 0.05, critical value = 1.65
Conclusion:
Since calculated value < critical value, Accept H0.
Therefore, the machine has not been improved.

Dr. Narasimhan G, RNSIT 18

6. In a city A 𝟐𝟎% of a random sample of 900 schoolboys had a certain slight physical defect. In
another city B 𝟏𝟖. 𝟓% of a random sample of 1600 schoolboys had the same defect. Is the
difference between the proportions significant?
Since samples sizes are 𝑛1 = 900 𝑎𝑛𝑑 𝑛2 = 1600, apply z test.
By data, 𝑝1 = 20% = 0.2, 𝑝2 = 18.5% = 0.185
H0: 𝑷𝟏 = 𝑷𝟐 , the difference between the proportions in not significant.
To find: Calculated value
𝑛1 𝑝1 + 𝑛2 𝑝2
𝑃=
𝑛1 + 𝑛2
900(0.2) + 1600(0.185)
= = 0.1904
900 + 1600

1 1
𝑆𝐸(𝑝1 − 𝑝2 ) = √𝑃𝑄 ( + )
𝑛1 𝑛2

1 1
= √0.19 × 0.81 × (900 + 1600) = 0.0163

𝑝1 − 𝑝2
|𝑧| = | |
𝑆𝐸(𝑝1 − 𝑝2 )
0.2 − 0.185 0.015
=| |= = 0.92
0.0163 0.0163

Therefore, calculated value = 0.92

To find: Critical value

At 𝛼 = 0.05, critical value = 1.96
Conclusion:
Since calculated value < critical value, Accept H0.
Therefore, there is no significant difference between the proportions.

Dr. Narasimhan G, RNSIT 19

7. In two large populations there are 30% and 25% respectively of fair-haired people. Is this
difference likely to be hidden in samples of 1200 and 900 respectively from the two
populations?
Since samples sizes are 𝑛1 = 1200 𝑎𝑛𝑑 𝑛2 = 900, apply z test.
By data, 𝑃1 = 30% = 0.3, 𝑃2 = 25% = 0.25
H0: 𝑷𝟏 = 𝑷𝟐 , the difference between the proportions in not significant.
To find: Calculated value

𝑃1 𝑄1 𝑃2 𝑄2
𝑆𝐸(𝑃1 − 𝑃2 ) = √ +
𝑛1 𝑛2

(0.3)(0.7) (0.25)(0.75)
=√ +
1200 900
= 0.0196
𝑃1 − 𝑃2
|𝑧| = | |
𝑆𝐸(𝑃1 − 𝑃2 )

0.3 − 0.25 0.05

=| |= = 2.5510
0.0196 0.0196

Therefore, calculated value = 2.5510

To find: Critical value
At 𝛼 = 0.05, critical value = 1.96
Conclusion:
Since calculated value > critical value, Reject 𝐻0 .
Therefore, this difference is unlikely to be hidden in samples of 1200 and 900
respectively from the two populations.

Dr. Narasimhan G, RNSIT 20