UNIT-III

Statistical Estimation Theory

Unbiased Estimates:
If the mean of the sampling distribution of a statistic equals the corresponding
population parameter, the statistic is called an unbiased estimator of the parameter;
otherwise, it is called a biased estimator. The corresponding values of such statistics are
called unbiased or biased estimates, respectively.
Example:
The mean of the sampling distribution of means 𝜇𝑋̅ is 𝜇, the population mean. Hence the
sample mean 𝑋̅ is an unbiased estimate of the population mean 𝜇.
Efficient Estimates:
If the sampling distributions of two statistics have the same mean, then the statistic
with the smaller variance is called an efficient estimator of the mean, while the other statistic
is called an inefficient estimator.
Point Estimates & Interval Estimates:
An estimate of a population parameter given by a single number is called a
point estimate of the parameter.
An estimate of a population parameter given by two numbers between which the parameter
may be considered to lie is called an interval estimate of the parameter.
Confidence-Interval Estimates of Population Parameters:
Let 𝜇𝑆 𝑎𝑛𝑑 𝜎𝑆 be the mean and standard deviation respectively, of the sampling
distribution of a statistic 𝑆. Then if the sampling distribution of 𝑆 is approximately normal
we can expect to find an actual sample statistic 𝑆 lying in the intervals 𝜇𝑠 − 𝜎𝑠 𝑡𝑜 𝜇𝑠 + 𝜎𝑠 ,
𝜇𝑠 − 2𝜎𝑠 𝑡𝑜 𝜇𝑠 + 2𝜎𝑠 and 𝜇𝑠 − 3𝜎𝑠 𝑡𝑜 𝜇𝑠 + 3𝜎𝑠 about 68.27%, 95.45%, 𝑎𝑛𝑑 99.73% of
the time, respectively.
𝜇𝑠 ± 𝜎𝑠 , 𝜇𝑠 ± 2𝜎𝑠 𝑎𝑛𝑑 𝜇𝑠 ± 3𝜎𝑠 are called as 68.27%, 95.45%, 𝑎𝑛𝑑 99.73% confidence
interval for estimating 𝜇𝑠
1) Confidence Intervals for Means:
If the statistic 𝑆 is the sample mean 𝑋̅, then the 95% 𝑎𝑛𝑑 99% confidence
limits for estimating the population mean 𝜇, are given by 𝑋̅ ± 1.96𝜎𝑋̅ 𝑎𝑛𝑑 𝑋̅ ± 2.58𝜎𝑋̅
respectively. More generally, the confidence limits are given by
𝑋̅ ± 𝑧𝑐 𝜎𝑋̅
Where, 𝑧𝑐 depends on the particular level of confidence
Also,
𝜎
𝜎𝑋̅ = …(infinite or with replacement population)
√𝑁

𝜎 𝑁𝑃 −𝑁
And 𝜎𝑋̅ = √ …(finite or without replacement population)
√𝑁 𝑁 𝑃 −1

2) Confidence Intervals for Proportions:

If the statistic 𝑆 is the proportion of ‘‘successes’’ in a sample of size 𝑁 drawn from a
binomial population in which 𝑝 is the proportion of successes (i.e. the probability of
success), then the confidence limits for 𝑝 are given by 𝑃 ± 𝑧𝑐 𝜎𝑃
Where, 𝑃 is the proportion of successes in the sample of size 𝑁.
If sampling is either from an infinite population or with replacement then the confidence
limits for the population proportion is given by
𝑝𝑞
𝑃 ± 𝑧𝑐 √
𝑁
And for sampling without replacement from a population of finite size 𝑁𝑃 it is

𝑝𝑞 𝑁𝑃 − 𝑁
𝑃 ± 𝑧𝑐 √ √
𝑁 𝑁𝑃 − 1

Qn) Suppose that height of 100 male students in XYZ university represent a random sample
of the heights of all 1546 students in the university with sample mean 𝑋̅ = 67.45 𝑎𝑛𝑑 𝜎 =
2.93
Find the (a) 95% and (b) 99% confidence intervals for estimating the mean height.
Solution:
𝜎 2.93
(a) The 95% confidence limits are 𝑋̅ ± 1.96 = 67.45 ± 1.96
√𝑁 √100

Thus the 95% confidence interval for the population mean 𝜇 is 66.88 𝑡𝑜 68.02 𝑖𝑛𝑐ℎ𝑒𝑠,
which can be denoted by 66.88 < 𝜇 < 68.02
We can therefore say that the probability that the population mean height lies between
66.88 𝑎𝑛𝑑 68.02 𝑖𝑛𝑐ℎ𝑒𝑠 is about 95%, 𝑜𝑟 0.95.
In symbols we write Pr{66.88 < 𝜇 < 68.02} = 0.95
This is equivalent to saying that we are 95% confident that the population mean (or true
mean) lies between 66.88 𝑎𝑛𝑑 68.02 𝑖𝑛𝑐ℎ𝑒𝑠.
(b) The 99% confidence limits are
 𝜎 2.93
𝑋̅ ± 2.58 = 67.45 ± 2.58
√𝑁 √100

Thus the 99% confidence interval for the population mean 𝜇 is 66.69 𝑡𝑜 68.21 𝑖𝑛𝑐ℎ𝑒𝑠,
which can be denoted by 66.69 < 𝜇 < 68.21
We can therefore say that the probability that the population mean height lies between
66.69 𝑎𝑛𝑑 68.21 𝑖𝑛𝑐ℎ𝑒𝑠 is about 99% 𝑜𝑟 0.99
In symbols we write Pr{66.69 < 𝜇 < 68.21} = 0.99
This is equivalent to saying that we are 99% confident that the population mean (or true
mean) lies between 66.69 𝑎𝑛𝑑 68.21 𝑖𝑛𝑐ℎ𝑒𝑠.
Qn) In measuring reaction time, a psychologist estimates that the standard deviation is 0.05
seconds. How large a sample of measurements must he take in order to be (a) 95% and (b)
99% confident that the error of his estimate will not exceed 0.01 𝑠𝑒𝑐𝑜𝑛𝑑𝑠?
Solution:
𝜎
(a) The 95% confidence limits are 𝑋̅ ± 1.96
√𝑁
𝜎
Where 1.96 is the error of the estimate.
√𝑁
𝜎
This error will be less than 0.01 if 1.96 < 0.01
√𝑁
0.05
⇒ 1.96 < 0.01
√𝑁
0.05
⇒ 1.96 < √𝑁
0.01

⇒ 9.8 < √𝑁
⇒ 96.04 < 𝑁 ⇒ Thus we can be 95% confident that the error of the estimate will be less
than 0.01 𝑠𝑒𝑐𝑜𝑛𝑑𝑠 if 𝑁 is 97 or larger.
(b) H.W.
Qn) A random sample of 50 mathematics grades out of a total of 200 showed a mean of 75
and a standard deviation of 10.
(a) What are the 95% confidence limits for estimates of the mean of the 200 grades?
(b) With what degree of confidence could we say that the mean of all 200 grades is 75 ± 1?
Solution:
(a) Since the population size is not very large compared with the sample size, the 95%
confidence limits are

𝜎 𝑁𝑃 − 𝑁 10 200 − 50
𝑋̅ ± 1.96 √ = 75 ± 1.96 √ = 75 ± 2.4
√𝑁 𝑁𝑃 − 1 √50 200 − 1
(b) The confidence limits can be represented by

𝜎 𝑁𝑃 − 𝑁
𝑋̅ ± 𝑧𝑐 √ = 75 ± 1.23𝑧𝑐
√𝑁 𝑁𝑃 − 1

Since this must equal, 75 ± 1

we have 1.23𝑧𝑐 = 1 ⇒𝑧𝑐 = 0.81
The area under the normal curve from 𝑧 = 0 𝑡𝑜 𝑧 = 0.81 𝑖𝑠 0.2910
Hence the required degree of confidence is 2(0.2910) = 0.582 𝑜𝑟 58.2%
Qn) A sample population of 100 voters chosen at random from all voters in a given district
indicated that 55% of them were in favour of a particular candidate.
Find the (a) 95%, (b) 99%, and (c) 99.73% confidence limits for the proportion of all the
voters in favour of this candidate.
Solution:
Here, 𝑝 = 0.55 ⇒ 𝑞 = 0.45
(a) The 95% confidence limits for the population are
𝑝𝑞
𝑃 ± 1.96𝜎𝑃 = 𝑃 ± 1.96√
𝑁

(0.55)(0.45)
= 0.55 ± 1.96√
100

= 0.55 ± 0.10
(b) The 99% confidence limits for the population are
𝑝𝑞
𝑃 ± 2.58𝜎𝑃 = 𝑃 ± 2.58√
𝑁

(0.55)(0.45)
= 0.55 ± 2.58√
100

= 0.55 ± 0.13
(c) The 99.73% confidence limits for the population are
𝑝𝑞
𝑃 ± 3𝜎𝑃 = 𝑃 ± 3√
𝑁

(0.55)(0.45)
= 0.55 ± 3√
100

= 0.55 ± 0.15
Confidence Intervals for Standard Deviations
The confidence limits for the standard deviation of a normally distributed population,
as estimated from a sample with standard deviation 𝑠, are given by
𝜎
𝑠 ± 𝑧𝑐 𝜎𝑠 = 𝑠 ± 𝑧𝑐
√2𝑁
Probable Error:
The 50% confidence limits of the population parameters corresponding to a statistic 𝑆
are given by 𝑆 ± 0.6745𝜎𝑆 .
The quantity 0.6745𝜎𝑆 is known as the probable error of the estimate.
Qn) The standard deviation of the breaking strengths of 100 cables tested by a company was
180 𝑙𝑏. Find the (a) 95%, (b) 99%, and (c) 99.73% confidence limits for the standard
deviation of all cables produced by the company.
Solution:
Here, 𝜎 = 180, 𝑁 = 100
𝜎 180
(a) The 95% confidence limits are 𝑠 ± 1.96 = 180 ± 1.96 = 180 ± 24.9
√2𝑁 √2×100

(b) H.W. (c) H.W.

Qn) The voltages of 50 batteries of the same type have a mean of 18.2 𝑣𝑜𝑙𝑡𝑠 (𝑉) and a
standard deviation of 0.5 𝑉. Find (a) the probable error of the mean and (b) the 50%
confidence limits.
Solution:
𝜎 0.5
(a) Probable error of the mean= 0.6745𝜎𝑋̅ = 0.6745 = 0.6745 = 0.048
√𝑁 √50

(b) The 50% confidence limits are 18.2 ± 0.048 𝑉

Statistical Decision Theory
Population:
Any group under statistical study is called as a population.
Sample:
Any subset of the population is called as a sample. The number of elements in a
sample is called as the sample size and is denoted by 𝑁.
If 𝑁 ≥ 30 then it is called as ‘Large sample’
If 𝑁 < 30 then it is called as ‘Small sample’
Parameter:
Any function of the population values is called as a parameter. We consider the
following parameters:
1. 𝜇: Population Mean
2. 𝜎: Population standard deviation(s.d.)
3. 𝑝: population proportion
Statistic:
Any function of the sample value is called a statistic. We consider the following
statistic.
𝑋̅: Sample Mean
𝑠: Sample standard deviation(s.d.)
𝑃: Sample Proportion
Sampling Distribution & Standard Error:
Every statistic is a random variable. The probability distribution of a statistic is called
as its sampling distribution. The standard deviation of a statistic is called as its standard
error.
Statistical Decisions:
The decisions made about the population on the basis of information from sample is
called as ‘Statistical Decisions’.
Ex: To decide whether a new medicine is effective in controlling a particular disease.
Statistical Hypothesis:
Any assumption or guess or statement regarding the population is called as a
‘Statistical Hypothesis’. The statistical hypothesis may or may not be true. Generally the
statistical hypothesis is about probability distribution of the population.
Null Hypothesis:
 A statistical hypothesis which is formulated for the purpose of rejecting or nullifying it
is called as ‘A null hypothesis’. A null hypothesis is denoted by 𝐻0 (H not). Generally 𝐻0 is
based on some past data or record.
Example:
If we want to decide whether a given coin is loaded, we formulate the hypothesis that the
coin is fair (i.e.,𝑝 = 0.5, where p is the probability of heads).
Similarly, if we want to decide whether one procedure is better than another, we formulate
the hypothesis that there is no difference between the procedures (i.e., any observed
differences are due merely to fluctuations in sampling from the same population). Such
hypotheses are often called null hypotheses and are denoted by 𝐻0
Alternate Hypothesis:
Any hypothesis which differs from the given null hypothesis is called as an ‘alternate
hypothesis’. An alternate hypothesis is denoted by 𝐻1 (H one).
Example:
If one hypothesis is 𝑝 = 0.5, alternative hypotheses might be 𝑝 = 0.7 𝑜𝑟 𝑝 ≠ 0.5
Test of Hypothesis:
A procedure which enables us to determine whether observed sample differs
significantly from results expected and thus help us to decide whether to accept or reject
hypothesis is called a ‘Test of Hypothesis’ or ‘Decision Rule’.
Type-I Error and Type-II Error:
When we reject a hypothesis when it should be accepted, we will say that Type-I error
has been made.
When we accept a hypothesis when it should be rejected we say that Type-II error has been
made.
In order for decision rules (or tests of hypotheses) to be good, they must be designed so as to
minimize errors of decision. This is not a simple matter, because for any given sample size,
an attempt to decrease one type of error is generally accompanied by an increase in the other
type of error. In practice, one type of error may be more serious than the other, and so a
compromise should be reached in favour of limiting the more serious error. The only way to
reduce both types of error is to increase the sample size, which may or may not be possible.
Level of Significance:
The maximum probability with which we are ready to risk Type-I error is called the
level of significance, or ‘Significance level’.
This probability is denoted by 𝛼 and is specified before a sample is drawn.
A significance level of 0.05(5%) 𝑜𝑟 0.01(1%) is common.
The significance level is 5% it means that there are 5 chances in 100 that we would reject a
hypothesis when actually it should be accepted. This means we are 95% confident that the
decision taken is right.
Two-tailed and one-tailed tests:
To test a hypothesis, we develop a statistic and with the help of its value we take a
decision whether to reject or accept a hypothesis. Suppose we denote such statistics by S. If
we are interested in extreme values of S on both the sides of the distribution, then it is called
as ‘Two tailed test’. If hypothesis is ≠ type then it is ‘Two tailed Test’.
If we are interested in extreme values of S on only one side of the distribution, then it is
called as ‘One tailed Test’. If the hypothesis is > 𝑜𝑟 < type then it is a ‘One Tailed Test’.

Special Test:
Mean:
Here 𝑆 = 𝑋̅, the sample mean
𝜇𝑆 = 𝜇𝑋̅ = 𝜇, the population mean
𝜎
𝜎𝑆 = 𝜎𝑋̅ = , where σ is the population standard deviation and N is sample size
√𝑁
𝑋̅−𝜇
The 𝑧 score is given by 𝑧 = 𝜎
√𝑁

Proportions
Here 𝑆 = 𝑃, the proportion of successes in a sample
𝜇𝑆 = 𝜇𝑃 = 𝑝, Where p is the population proportion of successes
𝑝𝑞
𝜎𝑆 = 𝜎𝑃 = √ , where 𝑞 = 1 − 𝑝 and N is sample size
𝑁
𝑃−𝑝
The 𝑧 score is given by 𝑧 = 𝑝𝑞
√𝑁

Test of Mean using Normal distribution:

Qn) The breaking strength of cables produced by a manufacturer have a mean of 1800 𝑙𝑏
and a standard deviation of 100𝑙𝑏. By a new technique in the manufacturing process, it is
claimed that the breaking strength can be increased. To test this claim, a sample of 50 cables
is tested and it is found that the mean breaking strength is 1850𝑙𝑏. Can we support the claim
at the 0.01 significance level?
Solution:
𝐻0 : 𝜇 = 1800𝑙𝑏 and there is really no change in breaking strength.
𝐻1 : 𝜇 > 1800𝑙𝑏 and there is a change in breaking strength.
One tailed Test
𝜎 = 100, 𝑁 = 50, 𝑋̅ = 1850, 𝑙. 𝑜. 𝑠. 𝛼 = 0.01
𝑋̅−𝜇 1850−1800
𝑍= 𝜎 = 100 = 3.5355 > 2.33
√𝑁 √50

Reject 𝐻0
By new technique breaking strength has increased.
Qn) On an examination given to students at a large number of different schools, the mean
grade was 74.5 and standard deviation was 8.0. At one particular school where 200 student
took the examination, the mean grade was 75.9. Discuss the significance of this result at he
0.05 level from the view point of
a. One tailed test
b. Two tailed test
Solution:
a) 𝐻0 : Performance of school is same as the population.
And 𝐻1 :Performance of school is better than the population.
Given 𝜎 = 8, 𝑁 = 200, 𝑋̅ = 75.9, 𝑙. 𝑜. 𝑠. 𝛼 = 0.05
𝑋̅−𝜇 75.9−74.5
𝑍= 𝜎 = 8 = 2.4748 > 1.645
√𝑁 √200

Reject 𝐻0 .
∴ Performance of school is better than the population.
b) Two-tailed test
𝐻0 : 𝜇 = 74.5
𝐻1 : 𝜇 ≠ 74.5
𝑙. 𝑜. 𝑠. 𝛼 = 0.05
𝑍 = 2.4748 > 1.96

Reject 𝐻0
Performance of school differs from the population.
Test of proportion using Normal Distribution:
Qn) The manufacturer of a patent medicine claims that it is 90% effective in relieving an
allergy for a period of 8 hours. In a sample of 200 people who had the allergy, the medicine
provided relief for 160 people. Determine whether the manufacturer’s claim is legitimate at
5% level of significance.
Solution:
𝐻0 : 𝑝 = 90% = 0.9(Manufacturer’s claim is valid)
𝐻1 : 𝑝 < 0.9 (Manufacturer’s claim is not valid)
One-tailed test 𝑙. 𝑜. 𝑠. = 𝛼 = 0.05(𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 1.645)
Since, 𝑝 = 0.90 ⇒ 𝑞 = 1 − 𝑝 ⇒ 𝑞 = 0.10
Given 𝑁 = 200
160
⇒𝑃 = = 0.8
200
𝑃−𝑝 0.8−0.9
𝑍= 𝑝𝑞
= = −4.7140 < −1.645
(0.9)(0.1)
√𝑁 √
200

∴ Reject 𝐻0
∴ Manufacturer’s claim is not valid.
Qn) A pair of dice is tossed 100 times and it is observed that 23 times sum of numbers
appearing on uppermost faces is 7. Test the hypothesis that the dice are fair by using a two-
tailed test at 5% significance level.
Solution:
We know that, Number of elements in sample space i.e. 𝑛(𝑆) = 36
Let 𝐴 be the event of getting 7 as sum⇒ 𝑛(𝐴) = 6
1
𝐻0 : 𝑝 = (i.e. the dice are fair)
6
1
𝐻0 : 𝑝 ≠ (dice are not fair)
6

Two-tailed test 𝑙. 𝑜. 𝑠 = 𝛼 = 0.05(table value=1.96)

1 5
𝑞 =1−𝑝 =1− =
6 6
23
𝑁 = 100, 𝑃 = = 0.23
100
1
𝑃−𝑝 0.23−
6
𝑍= 𝑝𝑞
= = 1.69, 1.69 lies between ±1.96
√𝑁 1 5
√6×6
100

Accept 𝐻0 ⇒ Dice are fair

Test involving difference of Means using Normal Distribution:
Qn) An examination given to two classes consisting of 40 and 50 students respectively. In
the first class, the mean grade was 74 with a standard deviation of 8. In the second class, the
mean grade was 78 with a standard deviation of 7. Is there a significance difference between
the performance of two classes at 0.05 level?
Solution:
𝐻0 : 𝜇1 = 𝜇2
𝐻1 : 𝜇1 ≠ 𝜇2
Two-tailed test (𝑙. 𝑜. 𝑠. 𝛼 = 0.05, 𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 = 1.96)
𝑁1 = 40, 𝑋̅1 = 74, 𝜎1 = 8 and 𝑁2 = 50, 𝑋̅2 = 78, 𝜎2 = 7
(𝑋̅1 −𝑋̅2 )−(𝜇1 −𝜇2 ) (74−78)−0
𝑍= = 2 2
= −2.490 < −1.96
𝜎2 𝜎2 √8 +7
√ 1+ 2 40 50
𝑁1 𝑁2
Reject 𝐻0
There is a significant difference in the performance of the two classes.
Qn) A sample of 100 electric light bulbs produced by manufacturer A showed a mean
lifetime of 1190hrs with a standard deviation of 90hrs. A sample of 75 bulbs produced by
manufacturer B showed a mean lifetime of 1230hrs with a standard deviation of 120hrs. Is
there a difference between the mean lifetime of the two brands at significance level of 0.01?
Solution:
𝐻0 : 𝜇1 = 𝜇2
𝐻1 : 𝜇1 ≠ 𝜇2
Two tailed test (𝑙. 𝑜. 𝑠. 𝛼 = 0.01, 𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 = 2.58)
𝑁1 = 100, 𝑋̅1 = 1190, 𝜎1 = 90, 𝑁2 = 75, 𝑋̅2 = 1230, 𝜎2 = 120
(𝑋̅1 −𝑋̅2 )−(𝜇1 −𝜇2 ) (1190−1230)−0
𝑍= = 2 2
= −2.4209
𝜎2 𝜎2 √90 +120
√ 1+ 2 100 75
𝑁1 𝑁2

Which lies between ±2.58

Accept 𝐻0
There is no significant difference in the bulbs of A and B.
Test involving difference of Proportions using Normal Distributions:
Qn) Two groups 𝐴 𝑎𝑛𝑑 𝐵 consist of 100 people each who have a disease. A serum is given
to group A but not to group B, otherwise the two groups are treated identically. It is found
that in group 𝐴 𝑎𝑛𝑑 𝐵, 75 𝑎𝑛𝑑 65 people respectively, recover from the disease. Test the
hypothesis that the serum helps to cure the disease at 1% level.
Solution:
𝐻0 : 𝑝1 = 𝑝2
𝐻1 : 𝑝1 > 𝑝2
On tailed test(𝑙. 𝑜. 𝑠. 𝛼 = 0.01, 𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 = 2.33)
75 65
𝑁1 = 100, 𝑃1 = = 0.75, 𝑁2 = 100, 𝑃2 = = 0.65
100 100
𝑁1 𝑃1 +𝑁2 𝑃2 (100×0.75)+(100×0.65)
𝑝= = = 0.7 ⇒ 𝑞 = 1 − 𝑝 = 0.3
𝑁1 +𝑁2 100+100

(𝑃1 − 𝑃2 ) − (𝑝1 − 𝑝2 ) (0.75 − 0.65) − 0

𝑍= = = 1.5430 < 2.33
1 1
√𝑝𝑞 ( + ) √0.7 × 0.3( 1 + 1 )
𝑁1 𝑁2 100 100

Accept 𝐻0
Serum doesn’t help to cure the disease.
Qn) Random samples of 200 bolts manufactured by machine A and of 100 bolts
manufactured by machine B showed 19 and 5 defective bolts respectively. Test the
hypothesis that
a. The two machines are showing different qualities of performance
b. Machine B is performing better than A
Use 5% significance level.
Solution:
Consider proportion of defective bolts
19 5
𝑁1 = 200, 𝑃1 = = 0.095, 𝑁2 = 100, 𝑃2 = = 0.05
200 100
𝑁1 𝑃1 +𝑁2 𝑃2 (200×0.095)+(100×0.05)
𝑝= = = 0.08
𝑁1 +𝑁2 200+100

⇒ 𝑞 = 1 − 𝑝 = 0.92
(𝑃1 − 𝑃2 ) − (𝑝1 − 𝑝2 ) (0.095 − 0.05) − 0
𝑍= = = 1.3543
1 1
√𝑝𝑞 ( + ) √0.08 × 0.92 × ( 1 + 1 )
𝑁1 𝑁2 200 100
a. 𝐻0 : 𝑝1 = 𝑝2
𝐻1 : 𝑝1 ≠ 𝑝2
Two tailed test(𝑙. 𝑜. 𝑠. 𝛼 = 5%, 𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 = 1.96)
1.3543 lies between ±1.96
⇒ Accept 𝐻0
There is no significant difference in the performance.
b. 𝐻0 : 𝑝1 = 𝑝2
𝐻1 : 𝑝1 > 𝑝2
One tailed test(𝑙. 𝑜. 𝑠. 𝛼 = 5%, 𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 = 1.645)
1.3543 < 1.645

⇒ Accept 𝐻0
There is no significant difference in the performance.