Lecture- 8

(Quality Management)

Estimation of Population Mean, Sample Size and

Analysis of Variance
Dr. Ali Nesar Khan
Adjunct Faculty

8.1

Estimation
Estimation is the calculated approximation of a result which is usable
when input data are incomplete/uncertain or extremely difficult to
gather/procure. In mathematics, approximation or estimation typically
means to find upper and/or lower bounds of quantity or database that
cannot readily be computed readily.
Estimation of Population Mean from a Sample Mean
When a sample mean is used as point estimate of population some error
can be expected. Calculating the error, one can find the highest and
lowest values of the mean of the population. In that case,
Estimated population mean = Confidence interval for unknown
mean and unknown standard deviation = Sample mean margin of
error = Sample mean Limit of confidence level
Margin of error/limit of confidence level, m = Se. t =(t

)
n

Where, Se =Standard error & t is upper critical value for t

distribution with degree of freedom n-1.
Estimate of population mean, = x m = x t
= x (value of SE

n
X

value of t)

Example-4: Say 12 pieces of dyed fabrics were selected randomly from

dyeing-printing finishing mills in different times within a shift to
determine the dimensional stability (say shrinkage). The unit was

coloring knit fabrics using winch machines. The test results are as
follows:
4.12%

4.60%

4.40%

4.50%

4.83%

5.00%

5.10%

4.17%

4.51%

4.52%

4.59%

4.22%

From these test results, estimate the mean shrinkage value of the all the
fabrics processed in that shift.

x = 4.55

= 0.34
SE

= 0.34/12 =0.34/3.16=0.098 (in %)

n

This value of 0.098 is to be treated with t values for degree of freedom

12-1 = 11.
For 5% probability of error i.e., 100-5 =95% confidence limit, the value of
t against 11 is determined to be 2.201 from t table (as shown in
annexure).
Limit of confidence interval = value of SE XValue of t = 2.201*0.098 = 0.22
The mean of the population
= Sample mean Limit of confidence interval
= x (value of SE XValue of t)
= 4.55 0.22
In other words, the estimated mean of the population will lie within the
range 4.33to 4.77.

Example-5: Say 10 bobbins were tested from 5 ring frames (each with
506 spindles) to determine the lea strengths (in Ibs) and the results were
as shown below:
61, 57.5, 59, 62, 58.5, 62, 59, 57, 63, 62.5

From these test results, estimate the mean strength and standard
deviation of the population (i.e., of 2,530 bobbins in ring frame)

x = 60.15 Ibs

= 2.2

SE

= 2.2/3.16 = 0.70 Ibs

n

This value of 0.70 is to be treated with t values for degree of freedom 101 = 9.
For 5% probability of error i.e., 100-5 =95% confidence limit, the value of
t against 9 is determined to be 2.26 from t table (as shown in annexure).
The mean of the population
= Sample mean (value of SEx XValue of t)
=60.15 (2.26*0.7) = 60.15 1.58
In other words, the mean of the population will lie within the range
58.57Ibs to 61.73 Ibs.
Now 95% confidence limit or 5% probability of error corresponds to t
value 1.96. In that case, the value of confidence interval for standard
deviation = SE*1.96 = 0.70*1.96 = 1.73.
The estimated value of standard deviation of the population
= SE*1.96 =2.2 1.73
i.e., standard deviation of the population will range from 0.47 to 3.93 .

8.2

Difference between the Means of Two Large Samples

Example-6: 2 Units A & B are spinning same count of yarn using same
machines and same cotton following same spinning plan. The test results
are shown below:
Unit A

Unit B

Mean Strength

60 Ibs

64 Ibs

Standard deviation

4 Ibs

5 Ibs

Sample Tested

100

80

Estimate difference in mean count between these 2 samples?

The estimate of standard deviation of the two samples
2

SD ( )

(n1 1) 1 (n 2 1) 2
n1 n 2 2

(100 1)4 2 (80 1)5 2

100 80 2

99 * 4 2 79 * 5 2
100 80 2

4351
178

= 4.94
The sample is large- together forming 180.
At 95% confidence interval i.e., the value of population mean is given by

x 1 - x 2 1.96 ((1/n1+1/n2) = 64-60 1.96*4.94 (1/100+1/80) =40.22

The interval is between 3.78 to 4.22 Ibs.
The difference of mean strength between A & B mills is 3.78 to 4.22 Ibs
at 95% confidence limit.

8.3

Sample Size & Number of Tests to be conducted

The most frequently asked question concerning sampling in the industry
and industrial research is, "What should be the sample size and how
many test one should carry out?" The answer to this question is
influenced by a number of factors, including the purpose of the study,
population size, the risk of selecting a "bad" sample, and the allowable
sampling error. In addition to the purpose of the study and population
size, three criteria usually will need to be specified to determine the
appropriate sample size: the level of precision/accuracy of test
results, the level of confidence or risk, and the degree of variability
in the attributes being measured or tested.
A probability of 5% or 95% confidence is mostly used in textile testing,
QC activities and researches. ASTM recommends determining the sample
size with the help of the following formula,
N = (1.96/E)2
Where, E is error level
is the standard deviation
Example-7: A knitting factory with all its 25 knit machines of same
specification (same gauge and diameter) was producing cotton fabric of
same construction using 100% cotton yarn of Ne 28. In order to check
proper function of knitting machines, a gramage test (weight in gram per
square meter) was conducted by taking 5 meters of fabrics from 5 knit
machines of with the following results:
161g, 190g,174g, 188g, 193g
Was the selection of sample correct?

= 181.2
Standard Deviation, = 13.44

Standard error, S E
=13.44/5 =13.44/2.24 =6.0
n
Mean Count, x

The value of t at 95% confidence level = 2.77

Population mean (confidence interval) = 181.22.77*6 = 181.2 16.62,
Now this limit of error (or limit of confidence) i.e., 16.62 corresponds to
100*16.62/181.2=9.17% of the mean.

In textile testing the error is not allowed over 5%. So let us first calculate
this value of error limit or confidence limit for 5%, which corresponds to
error level =5*16.62 /9.17 = 9.06
So, at 5% probability error the value of E= 9.06
Number of sample or sample size,
N = (1.96* /E)2
= (1.96*13.44/9.06)2
=8.46
Therefore, the number of samples to be tested is 9.

8.4

Use of Variances for Significance Test

In order to test significance between 2 sets of variables by variance, the
followings are to be done:

determine variances of each population of variables;

calculate variance ratio with the formula (F = Variance expected

to be greater/Variance Expected to be lower)

compare value of variance ratio (also called F-Ratio) with v1/v2

table specified for corresponding significance limit;

Evaluate significance of difference and acceptability level in terms

of significance level.

Example-8: Two jigger machines (Say A & B) in an woven dyeing unit

were coloring same grey fabrics at identical process condition (same
recipes and same shed condition). It was observed for a week whether
they were working with same performance or not. Samples (say 25 from
A and 31 from B) were tested to determine shed percent. The test results
showed standard deviations of 7 percent and 5 percent respectively for A
& B with a slight variation in mean shade percent (say 2% to 5%) among
them. Now how can one be sure about the equal performance of those 2
machines in terms of shade development?
We are to calculate first the variance ratio of the samples such as:
Variance A = (SDA)2 =
Variance B= (SDB)2 =
Variance or F Ratio,

2
2

=72 = 49,
52 = 25

F = Variance expected to be greater/Variance Expected to be lower

=49/25 = 1.96
Now this value has to be consulted with Table V1/V2
V1 = 25-1 =24, V2 = 31-1=30
At 5% significance level of probability error, V1=24 & V2 =30, the value of
F ratio is 1.89
At 1% significance level at V1=24 & V2 =30, the value of F ratio is 2.47
2.47>1.96>1.89
So, the difference of work of machines A & B is significant at 5%
significance level.
8.5

Analysis of Variance
The analysis of variance is a technique for splitting up the total variation
observed in the data into different components as per classification of the
data.
Example-8: Four looms A,B,C and D were used to produce a certain kind
of cotton fabric. Four (4) samples from each looms (say by 100 square
meters) were selected from the outputs of each looms at random, and the
number of flaws found in each of 100 square meters were as shown
below:
A

14

20

12

22

11

10

18

25

12

23

Is there a significant difference in the performance of the 4

machines?
If there is no significant difference in the performance in the performance
of the 4 machines i.e.
Ho:1= 2= 3= 4
This called null hypothesis.
A

8
9

6
8

14
12

20
22

11

10

18

25

12

23

A =40

B=28

C=53

D=90

Ai 2 =82+92+112+122
=410

Bi2
=62+82+102+42
=216

Ci2=
142+122+182+92
=745

Di2=
202+222+252+232
=2038

Grand Total G = A+B+C+D =40+28+53+90 =211

Raw Sum of Square (RSS) = Ai 2+Bi2+Ci2+Di2= 410+216+745+2038
= 3409
Correction Factor (CF) = Square of grand total/Frequency = G2/n =
(211)2/16
=2782.56
Sum of Square of total (SST) = RSS-CF = 3409-2782.56 =626.44
Treatment with Sum of Square (TSS)= Average of Sum of Square-CF
= (A 2+B2+C2+D2)/4-2782.56
= (402+282+532+902)/4 -2782.56
= 13293/4-2782.56 =540.69

Error with sum of square (ESS) = SST-TSS = 626.44-540.59 =85.85

ANOVA Table
Source of
Variation

Sum of
Square

Treatment

540.69

Degree of
Freedom
(DF)
K =4-1=3

Error

85.85

N =16-4 =12

Ratio (MSS)
= SS/DF

Variance Ratio
=MSSt/MSSe

Table Value

180.23
7.15

F
=180.23/7.15
= 25.21

5.95 (at 1% )
& 3.49 (at
5% )
Obtained
from table
2.11a b

25.21>5.95>3.49
The table value for F(2.11b) at 1% level of significance is 5.95. The
calculated value of variance ratio is 25.21 which is greater than the table
value; this also rejects the null hypothesis and concludes that there is
significant difference in the performance of the 4 machines at 1%
significance level of F.
The value for variance ratio F (2.11a) at 5% level of significance is 3.49.
The calculated value of variance ratio is 25.21, which is rather greater

than the table value; this rejects the null hypothesis and concludes that
there is significant difference in the performance of the 4 machines at 5%
significance level of F.
Table-1: Significance limit of t

Degrees of freedom

Probability
5 per cent

1 per cent

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

12.706
4.303
3.182
2.776
2.571
2.447
2.365
2.306
2.262
2.228
2.201
2.179
2.160
2.145
2.131
2.120
2.110
2.101
2.093
2.086
2.080
2.074
2.069
2.064
2.060
2.056
2.052
2.048
2.045
2.042

63.657
9.925
5.841
4.604
4.032
3.707
3.499
3.355
3.250
3.169
3.106
3.055
3.012
2.977
2.947
2.921
2.898
2.878
2.861
2.845
2.831
2.819
2.807
2.797
2.87
2.779
2.771
2.763
2.756
2.750

40
60
120
00

2.021
2.000
1.980
1.960

2.704
2.660
2.617
2.576

(Abridged from Table III of Statistical Tables of Biological, Agricultural, and Medical Resarch (R.A.Fisher and F.
Yates) (Oliuer and Boyd, Edinburgh). By courtesy of the Authors and Publishers.)

Table-2: Variance or F ratios (5 percent significance limits of F)

V1
1
2
3
4
5
6
8
12
24
v2

00

1
2
3
4
5

161.4
18.51
10.13
7.17
6.61

199.5
19.00
9.55
6.94
5.79

215.7
19.16
9.28
6.59
5.41

224.6
19.26
9.12
6.39
5.19

230.2
19.30
9.01
6.26
5.05

234.0
19.33
8.94
6.16
4.95

238.9
19.37
8.84
6.04
4.82

243.9
19.41
8.74
5.91
4.68

249.0
19.45
8.64
5.77
4.53

254.3
19.50
8.83
5.63
4.36

6
7
8
9
10

5.99
5.59
5.32
5.12
4.96

5.14
4.74
4.46
4.26
4.10

4.76
4.35
4.07
3.86
3 .71

4.53
4.12
3.84
3.63
3.48

4.39
3.97
3.69
3.48
3.33

4.28
3.87
3.58
3.37
3.22

4.15
3.73
3.44
3.23
3.07

4.00
3.57
3.28
3.07
2.91

3.84
3.41
3.12
2.90
2.74

3.67
3.23
2.93
2.71
2.54

11
12
13
14
15

4.84
4.75
4.67
4.60
4.54

3.98
3.88
3.80
3.74
3.68

3.59
3.49
3.41
3.34
3.29

3.36
3.26
3.18
3.11
3.06

3.20
3.11
3.02
2.96
2.90

3.09
3.00
2.92
2.85
2.79

2.95
2.85
2.77
2.70
2.64

2.79
2.69
2.60
2.53
2.48

2.61
2.50
2.42
2.35
2.29

2.40
2.30
2.21
2.13
2.07

16
17
18
19
20

4.49
4.45
4.41
4.38
4.35

3.63
3.59
3.55
3.52
3.49

3.24
3.20
3.16
3.13
3.10

3.01
2.96
2.93
2.90
2.87

2.85
2.81
2.77
2.74
2.71

2.74
2.70
2.66
2.63
2.60

2.59
2.55
2.51
2.48
2.45

2.42
2.38
2.34
2.31
2.28

2.24
2.19
2.15
2.11
2.08

2.01
1.96
1.92
1.88
1.84

21
22
23
24
25

4.32
4.30
4.28
4.26
4.24

3.47
3.44
3.42
3.40
3.38

3.07
3.05
3.03
3.01
2.99

2.84
2.82
2.80
2.78
2.76

2.68
2.66
2.64
2.62
2.60

2.57
2.55
2.53
2.51
2.49

2.42
2.40
2.38
2.36
2.34

2.25
2.23
2.20
2.18
2.16

2.05
2.03
2.00
1.98
1.96

1.81
1.78
1.76
1.73
1.71

26
27
28
29
30

4.22
4.21
4.20
4.18
4.17

3.37
3.35
3.34
3.33
3.32

2.98
2.96
2.95
2.93
2.92

2.74
2.73
2.71
2.70
2.69

2.59
2.57
2.56
2.54
2.53

2.47
2.46
2.44
2.43
2.42

2.32
2.30
2.29
2.28
2.27

2.15
2.13
2.12
2.10
2.09

1.95
1.93
1.91
1.90
1.89

1.69
1.67
1.65
1.64
1.62

40
60
120
00

4.08
4.00
3.92
3.84

3.23
3.15
3.07
2.99

2.84
2.76
2.68
2.60

2.61
2.52
2.45
2.37

2.45
2.37
2.29
2.21

2.34
2.25
2.17
2.09

2.18
2.10
2.02
1.94

2.00
1.92
1.83
1.75

1.79
1.70
1.61
1.52

1.51
1.39
1.25
1.00

(Abridged from Table III of Statistical Tables of Biological, Agricultural, and Medical Resarch (R.A.Fisher and F.
Yates) (Oliuer and Boyd, Edinburgh).

Table-3: Variance or F ratios (1percent significance limits of F)

V2/v1
1
2
3
4
5

1
4052
98.49
34.12
21.20
16.26

2
4999
99.01
30.81
18.00
13.27

3
5403
99.17
29.46
16.69
12.06

4
5625
99.25
28.71
15.98
11.39

5
5764
99.30
28.24
15.52
10.97

6
5859
99.93
27.91
15.21
10.67

8
5981
99.36
27.49
14.80
10.27

12
6106
99.42
27.05
14.37
9.89

24
6234
99.46
26.60
13.93
9.47

00
6366
99.50
26.12
13.46
9.02

6
7
8
9
10

13.74
12.25
11.26
10.56
10.04

10.92
9.55
8.65
8.02
7.56

9.78
8.45.
7.59
6.99
6.55

9.15
7.85
7.01
6.42
5.99

8.75
7.46
6.63
6.06
5.64

8.47
7.19
6.37
5.80
5.39

8.10
6.84
6.03
5.47
5.06

7.72
6.47
5.67
5.11
4.71

7.31
6.07
5.28
4.73
4.33

6.88
5.65
4.86
4.31
3.91

11
12
13
14
15

9.65
9.33
9.07
8.86
8.68

7.20
6.93
6.70
6.51
6.36

6.22
5.95
5.74
5.56
5.42

5.67
5.41
5.20
5.03
4.89

5.32
5.06
4.86
4.69
4.56

5.07
4.82
4.62
4.46
4.32

4.74
4.50
4.30
4.14
4.00

4.40
4.16
3.96
3.80
3.67

4.02
3.78
3.59
3.43
3.29

3.60
3.36
3.16
3.00
2.87

16
17
18
19
20

8.53
8.40
8.28
8.18
8.10

6.23
6.11
6.01
5.93
5.85

5.29
5.18
5.09
5.01
4.94

4.77
4.67
4.58
4.50
4.43

4.44
4.34
4.25
4.17
4.10

4.20
4.10
4.01
3.94
3.87

3.89
3.79
3.71
3.63
3.56

3.55
3.45
3.37
3.30
3.23

3.18
3.08
3.00
2.92
2.86

2.75
2.65
2.57
2.49
2.42

21
22
23
24
25

8.02
7.94
7.88
7.82
7.77

5.78
5.72
5.66
5.61
5.57

4.87
4.82
4.76
4.72
4.68

4.37
4.31
4.26
4.22
4.18

4.04
3.99
3.94
3.90
3.36

3.81
3.76
3.71
3.67
3.63

3.51
3.45
3.41
3.36
3.32

3.17
3.12
3.07
3.03
2.99

2.80
2.75
2.70
2.66
2.62

2.36
2.31
2.26
2.21
2.17

26
27
28
29
30

7.72
7.68
7.64
7.60
7.56

5.53
5.49
5.45
5.42
5.39

4.64
4.60
4.57
4.54
4.51

4.74
4.11
4.07
4.04
4.02

3.82
3.78
3.75
3.73
3.70

3.59
3.56
3.53
3.50
3.47

3.29
3.26
3.23
3.20
3.17

2.96
2.93
2.90
2.87
2.84

2.58
2.55
2.52
2.49
2.47

2.13
2.10
2.06
2.03
2.01

40
60
120
00

7.31
7.08
6.85
6.64

5.18
4.98
4.79
4.60

4.31
4.13
3.95
3.78

3.83
3.65
3.48
3.32

3.51
3.34
3.17
3.02

3.29
3.12
2.96
2.80

2.99
2.82
2.66
2.51

2.66
2.50
2.34
2.18

2.29
2.12
1.95
1.79

1.80
1.60
1.38
1.00