Hypothesis Testing - II

S. Devi Yamini

March 27, 2018

Module 6 March 27, 2018 1 / 65

Overview

1 Introduction

Module 6 March 27, 2018 2 / 65

Overview

1 Introduction

2 Small Sample tests

Test of single mean
Test of difference of means
F-test (Test of equality of variance)

Module 6 March 27, 2018 2 / 65

Overview

1 Introduction

2 Small Sample tests

Test of single mean
Test of difference of means
F-test (Test of equality of variance)

3 Chi-square Test
Test for Independence
Test for goodness of fit

Module 6 March 27, 2018 2 / 65

Overview

1 Introduction

2 Small Sample tests

Test of single mean
Test of difference of means
F-test (Test of equality of variance)

3 Chi-square Test
Test for Independence
Test for goodness of fit

4 Design of Experiments
One-way ANOVA
Two-way ANOVA

Module 6 March 27, 2018 2 / 65

Small sample tests

1. If the population is normally distributed and σ is known (OR) if

σ is unknown and n ≥ 30 then we can apply Z test (standard
normal distribution).

2. If the population is normally distributed, σ is unknown, and

n < 30, then we apply t -test (Student’s t distribution).

Module 6 March 27, 2018 3 / 65

Small sample tests

1. If the population is normally distributed and σ is known (OR) if

σ is unknown and n ≥ 30 then we can apply Z test (standard
normal distribution).

2. If the population is normally distributed, σ is unknown, and

n < 30, then we apply t -test (Student’s t distribution).

Student’s t-distribution
The p.d.f of the t-distribution is

Γ( (r +2 1) ) 1
f (t ) = √ r
π r Γ( 2 ) (1 + t 2 ) (r +2 1)
r

with r degrees of freedom (the number of independent values or

quantities which can be assigned to a statistical distribution).
Module 6 March 27, 2018 3 / 65
Test of single mean

Null Hyp H0 : µ = µ0

Module 6 March 27, 2018 4 / 65

Test of single mean

Null Hyp H0 : µ = µ0

Test statistic :

x̄ − µ
t =
√S
n

follows t-distribution with n − 1 degrees of freedom.

Module 6 March 27, 2018 4 / 65

Test of single mean

Here,
(xi − x̄ )2
P
2
S =
n −1
is an unbiased estimator of population standard deviation σ 2 .

Module 6 March 27, 2018 5 / 65

Test of single mean

Here,
(xi − x̄ )2
P
2
S =
n −1
is an unbiased estimator of population standard deviation σ 2 .
The relation between S and s (sample standard deviation) is
r
n
S = s( )
n −1

Module 6 March 27, 2018 5 / 65

Test of single mean

Here,
(xi − x̄ )2
P
2
S =
n −1
is an unbiased estimator of population standard deviation σ 2 .
The relation between S and s (sample standard deviation) is
r
n
S = s( )
n −1

Standard error = √Sn

1 − α confidence limits for the mean are

S S
(x̄ − t α ,n −1 √ , x̄ + t α ,n −1 √ )
2 n 2 n

Module 6 March 27, 2018 5 / 65

T-table
Integre Technical Publishing Co., Inc. Moore/McCabe November 16, 2007 1:29 p.m. moore page T-11

Tables T-11
•

Table entry for p and C is Probability p

the critical value t ∗ with
probability p lying to its
right and probability C lying
between −t ∗ and t ∗ . t*

TABLE D
t distribution critical values
Upper-tail probability p

df .25 .20 .15 .10 .05 .025 .02 .01 .005 .0025 .001 .0005

1 1.000 1.376 1.963 3.078 6.314 12.71 15.89 31.82 63.66 127.3 318.3 636.6
2 0.816 1.061 1.386 1.886 2.920 4.303 4.849 6.965 9.925 14.09 22.33 31.60
3 0.765 0.978 1.250 1.638 2.353 3.182 3.482 4.541 5.841 7.453 10.21 12.92
4 0.741 0.941 1.190 1.533 2.132 2.776 2.999 3.747 4.604 5.598 7.173 8.610
5 0.727 0.920 1.156 1.476 2.015 2.571 2.757 3.365 4.032 4.773 5.893 6.869
6 0.718 0.906 1.134 1.440 1.943 2.447 2.612 3.143 3.707 4.317 5.208 5.959
7 0.711 0.896 1.119 1.415 1.895 2.365 2.517 2.998 3.499 4.029 4.785 5.408
8 0.706 0.889 1.108 1.397 1.860 2.306 2.449 2.896 3.355 3.833 4.501 5.041
9 0.703 0.883 1.100 1.383 1.833 2.262 2.398 2.821 3.250 3.690 4.297 4.781
10 0.700 0.879 1.093 1.372 1.812 2.228 2.359 2.764 3.169 3.581 4.144 4.587
11 0.697 0.876 1.088 1.363 1.796 2.201 2.328 2.718 3.106 3.497 4.025 4.437
12 0.695 0.873 1.083 1.356 1.782 2.179 2.303 2.681 3.055 3.428 3.930 4.318
13 0.694 0.870 1.079 1.350 1.771 2.160 2.282 2.650 3.012 3.372 3.852 4.221
14 0.692 0.868 1.076 1.345 1.761 2.145 2.264 2.624 2.977 3.326 3.787 4.140
15 0.691 0.866 1.074 1.341 1.753 2.131 2.249 2.602 2.947 3.286 3.733 4.073
16 0.690 0.865 1.071 1.337 1.746 2.120 2.235 2.583 2.921 3.252 3.686 4.015
17 0.689 0.863 1.069 1.333 1.740 2.110 2.224 2.567 2.898 3.222 3.646 3.965
18 0.688 0.862 1.067 1.330 1.734 2.101 2.214 2.552 2.878 3.197 3.611 3.922
19 0.688 0.861 1.066 1.328 1.729 2.093 2.205 2.539 2.861 3.174 3.579 3.883
20 0.687 0.860 1.064 1.325 1.725 2.086 2.197 2.528 2.845 3.153 3.552 3.850
21 0.686 0.859 1.063 1.323 1.721 2.080 2.189 2.518 2.831 3.135 3.527 3.819
22 0.686 0.858 1.061 1.321 1.717 2.074 2.183 2.508 2.819 3.119 3.505 3.792
23 0.685 0.858 1.060 1.319 1.714 2.069 2.177 2.500 2.807 3.104 3.485 3.768
24 0.685 0.857 1.059 1.318 1.711 2.064 2.172 2.492 2.797 3.091 3.467 3.745
25 0.684 0.856 1.058 1.316 1.708 2.060 2.167 2.485 2.787 3.078 3.450 3.725
26 0.684 0.856 1.058 1.315 1.706 2.056 2.162 2.479 2.779 3.067 3.435 3.707
27 0.684 0.855 1.057 1.314 1.703 2.052 2.158 2.473 2.771 3.057 3.421 3.690
28 0.683 0.855 1.056 1.313 1.701 2.048 2.154 2.467 2.763 3.047 3.408 3.674
29 0.683 0.854 1.055 1.311 1.699 2.045 2.150 2.462 2.756 3.038 3.396 3.659
30 0.683 0.854 1.055 1.310 1.697 2.042 2.147 2.457 2.750 3.030 3.385 3.646
40 0.681 0.851 1.050 1.303 1.684 2.021 2.123 2.423 2.704 2.971 3.307 3.551
50 0.679 0.849 1.047 1.299 Module 6
1.676 2.009 2.109 2.403 2.678 2.937 3.261 3.496 March 27, 2018 6 / 65
Problems

1. A random sample of 16 households is taken from a large block of

flats, and shows that household expenditure on food is 42 dollars
per week, with a standard deviation of 10 dollars. Assuming that
household expenditure on food is normally distributed, find the 95%
confidence interval for the population mean.

Module 6 March 27, 2018 7 / 65

Problems

1. A random sample of 16 households is taken from a large block of

flats, and shows that household expenditure on food is 42 dollars
per week, with a standard deviation of 10 dollars. Assuming that
household expenditure on food is normally distributed, find the 95%
confidence interval for the population mean.

Solution:
Given that n = 16, x̄ = 42, s = 10, α = 0.05

Module 6 March 27, 2018 7 / 65

Problems

1. A random sample of 16 households is taken from a large block of

flats, and shows that household expenditure on food is 42 dollars
per week, with a standard deviation of 10 dollars. Assuming that
household expenditure on food is normally distributed, find the 95%
confidence interval for the population mean.

Solution:
Given that n = 16, x̄ = 42, s = 10, α = 0.05
Since n < 30, we use t-distribution.
r r
n 16
S = s( ) = 10( ) = 10.33
n −1 15
α
2
= 0.025, degrees of freedom = n − 1 = 15.

Module 6 March 27, 2018 7 / 65

Test of single mean

Solution contd.
From the table for t-distribution, t0.025,15 = 2.1314.
Thus the 95% confidene interval for population mean µ is

S S
(x̄ − t0.025,n −1 √ , x̄ + t0.025,n −1 √ )
n n

10.33 10.33
(42 − 2.1314( ), 42 + 2.1314( ))
4 4
(36.5, 47.5)

Module 6 March 27, 2018 8 / 65

Test of single mean

2. The heights of 10 males of a given locality are found to be

70,67,62,68,61,68,70,64,64,66 inches. Is it reasonable to believe
that the average height is greater than 64 inches?

Module 6 March 27, 2018 9 / 65

Test of single mean

2. The heights of 10 males of a given locality are found to be

70,67,62,68,61,68,70,64,64,66 inches. Is it reasonable to believe
that the average height is greater than 64 inches?

Solution
Given n = 10

Module 6 March 27, 2018 9 / 65

Test of single mean

2. The heights of 10 males of a given locality are found to be

70,67,62,68,61,68,70,64,64,66 inches. Is it reasonable to believe
that the average height is greater than 64 inches?

Solution
Given n = 10 P
(xi −x̄ )2
x̄ = 66, S 2 = n −1
= 10

Module 6 March 27, 2018 9 / 65

Test of single mean

2. The heights of 10 males of a given locality are found to be

70,67,62,68,61,68,70,64,64,66 inches. Is it reasonable to believe
that the average height is greater than 64 inches?

Solution
Given n = 10 P
(xi −x̄ )2
x̄ = 66, S 2 = n −1
= 10
1. H0 : µ = 64 against H1 : µ > 64 (Right tailed test)
2. Level of significance α = 0.05

Module 6 March 27, 2018 9 / 65

Test of single mean

solution contd.
3. Test statistic:
Since population standard deviation is not known and n < 30,
we use t-test.

x̄ − µ 66 − 64
t = = √ =2
√S √10
n 10

follows t-distribution with 9 degrees of freedom.

4. Rejection region:
α = 0.05, Critical value is tα,9 = 1.833
The critical region is t ≥ 1.833.
Since Cal t > tα,9 , we reject H0 .
5. Conclusion:
There is sufficient evidence to believe that the average height
is greater than 64 inches.
Module 6 March 27, 2018 10 / 65
Test of difference of means

Null Hyp H0 : µ1 − µ2 = d

Module 6 March 27, 2018 11 / 65

Test of difference of means

Null Hyp H0 : µ1 − µ2 = d

Test statistic :

(x¯1 − x¯2 ) − (µ1 − µ2 )

t = q
S n1 + n1
1 2

follows t-distribution with n1 + n2 − 2 degrees of freedom.

Module 6 March 27, 2018 11 / 65

Test of difference of means

Here,
(x1i − x¯1 )2 + (x2i − x¯2 )2
P P
2
S =
n1 + n2 − 2
OR
n1 s12 + n2 s22
S2 =
n1 + n2 − 2
Problems:
1. Samples of two types of electric light bulbs were tested for
length of life and the following data were obtained.
Type I: n1 = 8, x¯1 = 1234 hours, s1 = 36 hours
Type II: n2 = 7, x¯2 = 1036 hours, s2 = 40 hours
Is the differnce in mean sufficient to warrant that Type I is superior
than Type II regarding the length of life?

Module 6 March 27, 2018 12 / 65

Problems

Solution
Given n1 = 8, n2 = 7, x¯1 = 1234, x¯2 = 1036.
1. H0 : µ1 = µ2 aginst H1 : µ1 > µ2
2. Level of significance α = 0.05
3. Test statistic:
Since population standard deviation is unknown and n < 30, we
apply t-distribution with n1 + n2 − 2 = 13 degrees of freedom.

(x¯1 − x¯2 ) − (µ1 − µ2 )

t = q = 9.3925
S n1 + n1
1 2

follows t-distribution with 13 degrees of freedom.

Module 6 March 27, 2018 13 / 65

Problems

Solution contd
4. Critical region:
tα,13 = 1.77 and critical region is t > 1.77
Since Cal t > 1.77, we reject H0 .
5. Conclusion:
There is a statistical evidence that Type I is superior to Type II.

Module 6 March 27, 2018 14 / 65

F-distribution

F-distribution is used to test the equality of the variances of two

populations from which two samples have been drawn.

Module 6 March 27, 2018 15 / 65

F-distribution

F-distribution is used to test the equality of the variances of two

populations from which two samples have been drawn.

H0 : σ12 = σ22

Test Statistic:
S12
F =
S22
(x1i −x¯1 )2 (x2i −x¯2 )2
P P
where S12 = n1 −1
and S22 = n2 −1
.
(Note: The larger among S12 and S22 will be the numerator)

Here F follows F-distribution with (n1 − 1, n2 − 1) degrees of

freedom.
The critical value is F(n1 −1,n2 −1)

Module 6 March 27, 2018 15 / 65

Test of Variances
Two samples of 6 and 7 items have the following values for a
variable.
Sample 1: 39,41,42,42,44,40
Sample 2: 40,42,39,45,38,39,40
Do the sample variances differ significantly?

Module 6 March 27, 2018 16 / 65

Test of Variances
Two samples of 6 and 7 items have the following values for a
variable.
Sample 1: 39,41,42,42,44,40
Sample 2: 40,42,39,45,38,39,40
Do the sample variances differ significantly?

Solution
n1 = 6, n2 = 7, x¯1 = 41.33, x¯2 = 40.43

Module 6 March 27, 2018 16 / 65

Test of Variances
Two samples of 6 and 7 items have the following values for a
variable.
Sample 1: 39,41,42,42,44,40
Sample 2: 40,42,39,45,38,39,40
Do the sample variances differ significantly?

Solution
n1 = 6, n2 = 7, x¯1 = 41.33, x¯2 = 40.43
1. H0 : σ12 = σ22 (There is no significant difference between the
variances)
H1 : σ12 6= σ22
2. Level of significance α = 0.05
3. Test statistic:
S12
F =
S22

Module 6 March 27, 2018 16 / 65

Test of variances - problem

solution contd.
(x1i −x¯1 )2 (x2i −x¯2 )2
P P
where S12 = n1 −1
= 3.06668 and S22 = n2 −1
= 5.61905.

Module 6 March 27, 2018 17 / 65

Test of variances - problem

solution contd.
(x −x¯ )2 (x2i −x¯2 )2
P P
where S12 = 1i
n1 −1
1
= 3.06668 and S22 = n2 −1
= 5.61905.
Hence, F = 1.8323.

Module 6 March 27, 2018 17 / 65

Test of variances - problem

solution contd.
(x −x¯ )2 (x −x¯ )2
P P
where S12 = 1i
n1 −1
1
= 3.06668 and S22 = 2i
n2 −1
2
= 5.61905.
Hence, F = 1.8323.
From the table of F for (5,6) degrees of freedom, F(6,5) = 4.95.
The critical region is F > 4.95

Module 6 March 27, 2018 17 / 65

Test of variances - problem

solution contd.
(x −x¯ )2 (x −x¯ )2
P P
where S12 = 1i
n1 −1
1
= 3.06668 and S22 = 2i
n2 −1
2
= 5.61905.
Hence, F = 1.8323.
From the table of F for (5,6) degrees of freedom, F(6,5) = 4.95.
The critical region is F > 4.95
Since cal F < 4.95, we accept H0 .

Module 6 March 27, 2018 17 / 65

Test of variances - problem

solution contd.
(x −x¯ )2 (x −x¯ )2
P P
where S12 = 1i
n1 −1
1
= 3.06668 and S22 = 2i
n2 −1
2
= 5.61905.
Hence, F = 1.8323.
From the table of F for (5,6) degrees of freedom, F(6,5) = 4.95.
The critical region is F > 4.95
Since cal F < 4.95, we accept H0 .
5. Conclusion:
There is no significant difference between the population
variances.

Module 6 March 27, 2018 17 / 65

Try these problems

1. The price of a popular tennis racket at a national chain store is

179 dollars. Ria bought five of the same racket at an online auction
site for the following prices: 155, 179,175,175,161. Assuming that
the auction prices of rackets are normally distributed, determine
whether there is sufficient evidence in the sample, at the 5% level
of significance, to conclude that the average price of the racket is
less than 179 dollars if purchased at an online auction.
(Hint: n = 5. Calculate x̄ and S from the data. H0 : µ = 179 against
H1 : µ < 179. Apply t-test.)

Module 6 March 27, 2018 18 / 65

Problems

2. Find the rejection region for each hypothesis test based on the
information given. The population is normally distributed.
(a) H0 : µ = 27 Against H1 : µ < 27, α = 0.05, n = 12, σ = 2.2 (Hint:
Since σ is given and population is normally distributed, we apply
Z -test. )

(b) H0 : µ = 52 Against H1 : µ 6= 52,α = 0.05, n = 6, σ is unknown.

(Hint: Since σ is unknown and n < 30, we apply t -test)

(c) H0 : µ = −105 Against H1 : µ > −105, α = 0.10, n = 24, σ is

unknown. (Hint: Since σ is unknown and n < 30, we apply t -test)

Module 6 March 27, 2018 19 / 65

Try these
3. An economist wishes to determine whether people are driving
less than in the past. In one region of the country, the number of
miles driven per household per year in the past was 18.59 thousand
miles. A sample of 15 households produced a sample mean of 16.23
thousand miles for the last year, with sample standard deviation
4.06 thousand miles. Assuming a normal distribution of household
driving distances per year, perform the relevant test at the 5% level
of significance.

4. Two random samples gave the following results:

Sample 1: n1 = 10, x¯1 = 15, (x1i − x¯1 )2 = 90 (sum of squared
P

deviations from the mean)

Sample 2: n2 = 12, x¯2 = 14, (x2i − x¯2 )2 = 108
P

Test whether the samples come from the same normal population at
5% significance level. (Hint: We need to test for (i) population mean
and (ii) population variance. H0 : σ12 = σ22 and µ1 = µ2 . First apply
F-test and then apply t-test) Module 6 March 27, 2018 20 / 65
Chi-square distribution

The sum of k independent squared standard normal variables is a

Chi-square random variable with k degrees of freedom. That is,
χ2 = Z12 + Z22 + . . . + Zk2

Module 6 March 27, 2018 21 / 65

Chi-square distribution

The sum of k independent squared standard normal variables is a

Chi-square random variable with k degrees of freedom. That is,
χ2 = Z12 + Z22 + . . . + Zk2

The curve is non symmetrical and skewed to the right

The curve differs for each degrees of freedom

Module 6 March 27, 2018 21 / 65

Chi-square distribution

The sum of k independent squared standard normal variables is a

Chi-square random variable with k degrees of freedom. That is,
χ2 = Z12 + Z22 + . . . + Zk2

The curve is non symmetrical and skewed to the right

The curve differs for each degrees of freedom

Applications
- Goodness of fit
- Test for independence

Module 6 March 27, 2018 21 / 65

Test for independence

Procedure
1 Formulate the null and alternate hypothesis:
H0 : Two variables are independent
H1 : Two variables are not independent

Module 6 March 27, 2018 22 / 65

Test for independence

Procedure
1 Formulate the null and alternate hypothesis:
H0 : Two variables are independent
H1 : Two variables are not independent
2 Calculate the Expected frequencies
E = (row total)(column total) / sample size
Note that each expected value must be greater than or equal to
5 for the chi square test to be valid

Module 6 March 27, 2018 22 / 65

Test for independence

Procedure
1 Formulate the null and alternate hypothesis:
H0 : Two variables are independent
H1 : Two variables are not independent
2 Calculate the Expected frequencies
E = (row total)(column total) / sample size
Note that each expected value must be greater than or equal to
5 for the chi square test to be valid
3 Calculate the test statistic:
X (O − E )2
χ2 =
E

Module 6 March 27, 2018 22 / 65

Test for independence

4 Find the degrees of freedom

df = (r − 1)(c − 1)

where r is the number of rows and c is the number of columns

Module 6 March 27, 2018 23 / 65

Test for independence

4 Find the degrees of freedom

df = (r − 1)(c − 1)

where r is the number of rows and c is the number of columns

5 Calculate the critical value (cv) at the given LOS

Module 6 March 27, 2018 23 / 65

Test for independence

4 Find the degrees of freedom

df = (r − 1)(c − 1)

where r is the number of rows and c is the number of columns

5 Calculate the critical value (cv) at the given LOS
6 Conclusion: If χ2 < cv, then accept H0 (Variables are
independent)
else reject H0 (Variables are not independent)

Module 6 March 27, 2018 23 / 65

Problems

1. The side effects of a new drug are being tested against a placebo.
A simple random sample of 565 patients yields the results below. At
a significance level α = 0.10, is there enough evidence to conclude
that the treatment is independent of the side effect of nausea?

Result Drug Placebo Total

Nausea 36 13 49
No Nausea 254 262 516
Total 290 275 565

Module 6 March 27, 2018 24 / 65

Problems

1. The side effects of a new drug are being tested against a placebo.
A simple random sample of 565 patients yields the results below. At
a significance level α = 0.10, is there enough evidence to conclude
that the treatment is independent of the side effect of nausea?

Result Drug Placebo Total

Nausea 36 13 49
No Nausea 254 262 516
Total 290 275 565

Solution
H0 : The treatment and the response are independent.
H1 : The treatment and the response are dependent.

Module 6 March 27, 2018 24 / 65

Problems

1. The side effects of a new drug are being tested against a placebo.
A simple random sample of 565 patients yields the results below. At
a significance level α = 0.10, is there enough evidence to conclude
that the treatment is independent of the side effect of nausea?

Result Drug Placebo Total

Nausea 36 13 49
No Nausea 254 262 516
Total 290 275 565

Solution
H0 : The treatment and the response are independent.
H1 : The treatment and the response are dependent.
α = 0.10

Module 6 March 27, 2018 24 / 65

Problems

solution contd.
Expected frequencies

Result Drug Placebo Total

Nausea 25.15 23.85 49
No Nausea 264.85 251.15 516
Total 290 275 565

Module 6 March 27, 2018 25 / 65

Problems

solution contd.
Expected frequencies

Result Drug Placebo Total

Nausea 25.15 23.85 49
No Nausea 264.85 251.15 516
Total 290 275 565

degrees of freedom df = (2 − 1)(2 − 1) = 1

Module 6 March 27, 2018 25 / 65

Problems

solution contd.
Expected frequencies

Result Drug Placebo Total

Nausea 25.15 23.85 49
No Nausea 264.85 251.15 516
Total 290 275 565

degrees of freedom df = (2 − 1)(2 − 1) = 1

Test statistic: χ2 = 10.53

Module 6 March 27, 2018 25 / 65

Problems

solution contd.
Expected frequencies

Result Drug Placebo Total

Nausea 25.15 23.85 49
No Nausea 264.85 251.15 516
Total 290 275 565

degrees of freedom df = (2 − 1)(2 − 1) = 1

Test statistic: χ2 = 10.53
Critical value is 2.71 (df = 1, α = 0.10)

Module 6 March 27, 2018 25 / 65

Problems

solution contd.
Expected frequencies

Result Drug Placebo Total

Nausea 25.15 23.85 49
No Nausea 264.85 251.15 516
Total 290 275 565

degrees of freedom df = (2 − 1)(2 − 1) = 1

Test statistic: χ2 = 10.53
Critical value is 2.71 (df = 1, α = 0.10)
Since χ2 > 2.71, there is enough evidence to reject H0 . Hence,
there is a relation between the treatment and response.

Module 6 March 27, 2018 25 / 65

Problems

2. Suppose the undergraduate degrees are BA, BE, BBA, and

several others. There are three possible majors for the MBA
students which are accounting, finance, and marketing. Can the
statistician conclude that the undergraduate degree affects the
choice of major from the given table?

UG/ MBA Accounting Finance Marketing Total

BA 31 13 16 60
BE 8 16 7 31
BBA 12 10 17 39
Other 10 5 7 22
Total 61 44 47 152

Module 6 March 27, 2018 26 / 65

Solution

1 H0 : The undergraduate degree and MBA major are independent

H1 : The undergraduate degree and MBA major are dependent
2 Expected frequencies:
UG/MBA Accounting Finance Marketing Total
BA 24.08 17.37 18.55 60
BE 12.44 8.97 9.59 31
BBA 15.65 11.29 12.06 39
Other 8.83 6.37 6.8 22
Total 61 44 47 152

Module 6 March 27, 2018 27 / 65

Solution contd.

3 Test statistic:
X (O − E )2
χ2 =
E
2
(31 − 24.08) (13 − 17.37)2 (7 − 6.8)2
χ2 = + + ... + = 14.7
24.08 17.37 6.8

Module 6 March 27, 2018 28 / 65

Solution contd.

3 Test statistic:
X (O − E )2
χ2 =
E
2
(31 − 24.08) (13 − 17.37)2 (7 − 6.8)2
χ2 = + + ... + = 14.7
24.08 17.37 6.8

4 degrees of freedom df = (r − 1)(c − 1) = (4 − 1)(3 − 1) = 6

Module 6 March 27, 2018 28 / 65

Solution contd.

3 Test statistic:
X (O − E )2
χ2 =
E
2
(31 − 24.08) (13 − 17.37)2 (7 − 6.8)2
χ2 = + + ... + = 14.7
24.08 17.37 6.8

4 degrees of freedom df = (r − 1)(c − 1) = (4 − 1)(3 − 1) = 6

5 α = 0.05, Critical value χ2(0.05,6) = 12.59

Module 6 March 27, 2018 28 / 65

Solution contd.

3 Test statistic:
X (O − E )2
χ2 =
E
2
(31 − 24.08) (13 − 17.37)2 (7 − 6.8)2
χ2 = + + ... + = 14.7
24.08 17.37 6.8

4 degrees of freedom df = (r − 1)(c − 1) = (4 − 1)(3 − 1) = 6

5 α = 0.05, Critical value χ2(0.05,6) = 12.59
The critical region is χ2 > 12.59.
Since calculated χ2 > 12.59, we reject H0 .
6 Conclusion: There is sufficient evidence to the claim that the
undergraduate degree and the MBA major are related.

Module 6 March 27, 2018 28 / 65

Problems
3. The operations manager of a company that manufactures tyres
wants to determine whether there are any differences in the quality
of workmanship among the three daily shifts. She randomly selects
496 tyres and carefully inspects them. Each tyre is either classified
as perfect, satisfactory, or defective, and the shift that produced it is
also recorded. The two categorical variables of interest are : shift
and condition of the tyre produced. Do these data provide sufficient
evidence at 5% significance level to infer that there are differences
in quality among the three shifts?

Perfect Satisfactory Defective Total

Shift 1 106 114 11 231
Shift 2 67 70 16 153
Shift 3 37 65 10 112
Total 210 249 37 496

Module 6 March 27, 2018 29 / 65

Problems

4. Various countries are compared using two variables- composition

of economy and growth band as shown in the table.

High growth Medium growth Low growth

Predominant 20 25 5
agriculture
Predominant 40 5 6
manufacturing
Predominant 5 55 20
services

Test whether the predominant function in an economy has an

impact on the growth of the economy.

Module 6 March 27, 2018 30 / 65

Goodness of fit

Chi-square test enables us to check whether the given data fits the
theoretical distributions such as Binomial, Poisson, Normal, etc.

Module 6 March 27, 2018 31 / 65

Goodness of fit

Chi-square test enables us to check whether the given data fits the
theoretical distributions such as Binomial, Poisson, Normal, etc.
Formulate null and alternate hypothesis
Calculate the expected frequencies
Level of significance
Test statistic:
X (O − E )2
χ2 =
E
degrees of freedom df = k − 1 where k represents the number
of categories
Calculate the critical value (cv) at the given LOS

Module 6 March 27, 2018 31 / 65

Goodness of fit

Chi-square test enables us to check whether the given data fits the
theoretical distributions such as Binomial, Poisson, Normal, etc.
Formulate null and alternate hypothesis
Calculate the expected frequencies
Level of significance
Test statistic:
X (O − E )2
χ2 =
E
degrees of freedom df = k − 1 where k represents the number
of categories
Calculate the critical value (cv) at the given LOS
Conclusion: If χ2 < cv, then accept H0
else reject H0

Module 6 March 27, 2018 31 / 65

Problems

1. The number of defects per unit in a sample of 330 units of a

manufactured product was found as follows:

No. of defects: 0 1 2 3 4
No. of units: 214 92 20 3 1

Fit a Poisson distribution to the data and test for goodness of fit.

Module 6 March 27, 2018 32 / 65

Problems

1. The number of defects per unit in a sample of 330 units of a

manufactured product was found as follows:

No. of defects: 0 1 2 3 4
No. of units: 214 92 20 3 1

Fit a Poisson distribution to the data and test for goodness of fit.

Solution
H0 : The data fits the Poisson distribution
H1 : The data doesnot fit the Poisson distribution

Module 6 March 27, 2018 32 / 65

Problems

1. The number of defects per unit in a sample of 330 units of a

manufactured product was found as follows:

No. of defects: 0 1 2 3 4
No. of units: 214 92 20 3 1

Fit a Poisson distribution to the data and test for goodness of fit.

Solution
H0 : The data fits the Poisson distribution
H1 : The dataPdoesnot fit the Poisson distribution
fx
Mean = λ = P f = 145 330
= 0.439
Expected frequencies:
−λ 0
P (X = 0) = e 0!λ = 0.645
Similarly calculate P (X = 1), P (X = 2), P (X = 3), P (X = 4)

Module 6 March 27, 2018 32 / 65

Problems

Solution contd.
X 0 1 2 3 4
O 214 92 20 3 1
E 212.75 93.4 20.5 3 0.35

Test statistic:
(O − E )2
χ2 = = 0.0292
E

Module 6 March 27, 2018 33 / 65

Problems

Solution contd.
X 0 1 2 3 4
O 214 92 20 3 1
E 212.75 93.4 20.5 3 0.35

Test statistic:
(O − E )2
χ2 = = 0.0292
E
Degrees of freedom df = k − 1 = (5 − 3) − 1 = 1

Module 6 March 27, 2018 33 / 65

Problems

Solution contd.
X 0 1 2 3 4
O 214 92 20 3 1
E 212.75 93.4 20.5 3 0.35

Test statistic:
(O − E )2
χ2 = = 0.0292
E
Degrees of freedom df = k − 1 = (5 − 3) − 1 = 1
Critical value is χ20.05,1 = 3.84

Module 6 March 27, 2018 33 / 65

Problems

Solution contd.
X 0 1 2 3 4
O 214 92 20 3 1
E 212.75 93.4 20.5 3 0.35

Test statistic:
(O − E )2
χ2 = = 0.0292
E
Degrees of freedom df = k − 1 = (5 − 3) − 1 = 1
Critical value is χ20.05,1 = 3.84
Critical region is χ2 > 3.84.

Module 6 March 27, 2018 33 / 65

Problems

Solution contd.
X 0 1 2 3 4
O 214 92 20 3 1
E 212.75 93.4 20.5 3 0.35

Test statistic:
(O − E )2
χ2 = = 0.0292
E
Degrees of freedom df = k − 1 = (5 − 3) − 1 = 1
Critical value is χ20.05,1 = 3.84
Critical region is χ2 > 3.84.
Since calculated χ2 < 3.84, we accept H0 . Hence the given data fits
well with Poisson distribution.

Module 6 March 27, 2018 33 / 65

Problems

1. An experiment is conducted in which a die is rolled 240 times.

The outcomes are in the table below. At a significance level
α = 0.05, is there enough evidence to support the hypothesis that
the die is unbiased?

Outcome 1 2 3 4 5 6
Frequency 34 44 30 46 51 35

Module 6 March 27, 2018 34 / 65

Problems

1. An experiment is conducted in which a die is rolled 240 times.

The outcomes are in the table below. At a significance level
α = 0.05, is there enough evidence to support the hypothesis that
the die is unbiased?

Outcome 1 2 3 4 5 6
Frequency 34 44 30 46 51 35

Solution
H0 : The die is unbiased
H1 : The die is biased

Module 6 March 27, 2018 34 / 65

Problems

Solution contd.
240
Expected frequencies: E = 6
= 40 where Total frequency is 240.

Module 6 March 27, 2018 35 / 65

Problems

Solution contd.
240
Expected frequencies: E = 6
= 40 where Total frequency is 240.

Outcome 1 2 3 4 5 6
Frequency 40 40 40 40 40 40

Module 6 March 27, 2018 35 / 65

Problems

Solution contd.
240
Expected frequencies: E = 6
= 40 where Total frequency is 240.

Outcome 1 2 3 4 5 6
Frequency 40 40 40 40 40 40

(O −E )2
Outcome O E (O − E ) (O − E )2 E
1 34 40 -6 36 0.9
2 44 40 4 16 0.4
3 30 40 -10 100 2.5
4 46 40 6 36 0.9
5 51 40 11 121 3.025
6 35 40 -5 25 0.625

Module 6 March 27, 2018 35 / 65

Solution contd.
X (O − E )2
χ2 = = 8.35
E

Module 6 March 27, 2018 36 / 65

Solution contd.
X (O − E )2
χ2 = = 8.35
E
Degrees of freedom df = k − 1 = 6 − 1 = 5

Module 6 March 27, 2018 36 / 65

Solution contd.
X (O − E )2
χ2 = = 8.35
E
Degrees of freedom df = k − 1 = 6 − 1 = 5

Critical value at α = 0.05 with 5 degrees of freedom is 11.07

Module 6 March 27, 2018 36 / 65

Solution contd.
X (O − E )2
χ2 = = 8.35
E
Degrees of freedom df = k − 1 = 6 − 1 = 5

Critical value at α = 0.05 with 5 degrees of freedom is 11.07

Since Calculated χ2 < 11.07, we accept H0 .

Conclusion: There is no sufficient evidence that the die is biased OR

The die is unbiased.

Module 6 March 27, 2018 36 / 65

Problems
2. A sample analysis of examination results of 500 students was
made. It was found that 220 students had failed, 170 had secured a
third class, 90 were placed in the second class and 20 got a first
class. Are these figures commensurate with the general
examination result which is the ratio 4 : 3 : 2 : 1 for various
categories respectively?

Module 6 March 27, 2018 37 / 65

Problems
2. A sample analysis of examination results of 500 students was
made. It was found that 220 students had failed, 170 had secured a
third class, 90 were placed in the second class and 20 got a first
class. Are these figures commensurate with the general
examination result which is the ratio 4 : 3 : 2 : 1 for various
categories respectively? (Expected frequencies are 200,150,100,50.
χ2 = 23.667, critical value is 7.81, reject H0 .)

Module 6 March 27, 2018 37 / 65

Problems
2. A sample analysis of examination results of 500 students was
made. It was found that 220 students had failed, 170 had secured a
third class, 90 were placed in the second class and 20 got a first
class. Are these figures commensurate with the general
examination result which is the ratio 4 : 3 : 2 : 1 for various
categories respectively? (Expected frequencies are 200,150,100,50.
χ2 = 23.667, critical value is 7.81, reject H0 .)
3. The demand for a particular spare part in a factory was found to
vary from day to day. In a sample study the following information
was obtained:
Days Mon Tue Wed Thurs Fri Sat
Demand 1124 1125 1110 1120 1126 1115
Test the hypothesis that the number of parts demanded does not
depend on the day of the week.

Module 6 March 27, 2018 37 / 65

Problems
2. A sample analysis of examination results of 500 students was
made. It was found that 220 students had failed, 170 had secured a
third class, 90 were placed in the second class and 20 got a first
class. Are these figures commensurate with the general
examination result which is the ratio 4 : 3 : 2 : 1 for various
categories respectively? (Expected frequencies are 200,150,100,50.
χ2 = 23.667, critical value is 7.81, reject H0 .)
3. The demand for a particular spare part in a factory was found to
vary from day to day. In a sample study the following information
was obtained:
Days Mon Tue Wed Thurs Fri Sat
Demand 1124 1125 1110 1120 1126 1115
Test the hypothesis that the number of parts demanded does not
depend on the day of the week.
(Expected frequencies are 1120. χ2 = 0.179, critical value is 11.07,
accept H0 ) Module 6 March 27, 2018 37 / 65
Design of Experiments

ANOVA - Analysis of Variance

Module 6 March 27, 2018 38 / 65

Design of Experiments

ANOVA - Analysis of Variance

Why ANOVA?
When comparing means across two samples - we use Z-test or
t-test

Module 6 March 27, 2018 38 / 65

Design of Experiments

ANOVA - Analysis of Variance

Why ANOVA?
When comparing means across two samples - we use Z-test or
t-test
If more than two samples are test for their means, we use
ANOVA

Module 6 March 27, 2018 38 / 65

Design of Experiments

ANOVA - Analysis of Variance

Why ANOVA?
When comparing means across two samples - we use Z-test or
t-test
If more than two samples are test for their means, we use
ANOVA

We study,
Completely Randomised Design or One-way ANOVA
Randomized Block Design or Two-way ANOVA
Latin Square Design or Three-way ANOVA

Module 6 March 27, 2018 38 / 65

CSD or One-way ANOVA

The observations are independent.

H0 : µ1 = µ2 = µ3

H1 : At least there is one difference among the means.

Between group variance

F = Within group variance

with degrees of freedom (k − 1, N − k ) where k denotes the number

of groups and N denotes the sample size.

Module 6 March 27, 2018 39 / 65

CSD or One-way ANOVA

Problem 1: Compare the means of these groups

I II III
1 2 2
2 4 3
5 2 4

Module 6 March 27, 2018 40 / 65

CSD or One-way ANOVA

Problem 1: Compare the means of these groups

I II III
1 2 2
2 4 3
5 2 4

Solution:
1: H0 : µ1 = µ2 = µ3
H1 : At least there is one difference among the means.
α = 0.05

Module 6 March 27, 2018 40 / 65

CSD or One-way ANOVA

Problem 1: Compare the means of these groups

I II III
1 2 2
2 4 3
5 2 4

Solution:
1: H0 : µ1 = µ2 = µ3
H1 : At least there is one difference among the means.
α = 0.05
2: Degrees of freedom
DFBet = k − 1 = 3 − 1 = 2,

Module 6 March 27, 2018 40 / 65

CSD or One-way ANOVA

Problem 1: Compare the means of these groups

I II III
1 2 2
2 4 3
5 2 4

Solution:
1: H0 : µ1 = µ2 = µ3
H1 : At least there is one difference among the means.
α = 0.05
2: Degrees of freedom
DFBet = k − 1 = 3 − 1 = 2, DFWithin = N − k = 9 − 3 = 6

Module 6 March 27, 2018 40 / 65

CSD or One-way ANOVA

Problem 1: Compare the means of these groups

I II III
1 2 2
2 4 3
5 2 4

Solution:
1: H0 : µ1 = µ2 = µ3
H1 : At least there is one difference among the means.
α = 0.05
2: Degrees of freedom
DFBet = k − 1 = 3 − 1 = 2, DFWithin = N − k = 9 − 3 = 6
F(2,6) = 5.14

Module 6 March 27, 2018 40 / 65

One-way ANOVA

3. XX
G = xij = 25

Module 6 March 27, 2018 41 / 65

One-way ANOVA

3. XX
G = xij = 25

G2
Correction Factor C .F . = N
= 69.444

Sum of squares total

P P 2 G2
SST = xij − N = 83 − 69.444 = 13.556

Module 6 March 27, 2018 41 / 65

One-way ANOVA

3. XX
G = xij = 25

G2
Correction Factor C .F . = N
= 69.444

Sum of squares total

P P 2 G2
SST = xij − N = 83 − 69.444 = 13.556

Sum of squares between

P Ti2 G 2 2 2 2
SSB = ri
− N = 83 + 83 + 93 − 69.444
SSB = 69.667 − 69.444 = 0.223

Module 6 March 27, 2018 41 / 65

One-way ANOVA

3. XX
G = xij = 25

G2
Correction Factor C .F . = N
= 69.444

Sum of squares total

P P 2 G2
SST = xij − N = 83 − 69.444 = 13.556

Sum of squares between

P Ti2 G 2 2 2 2
SSB = ri
− N = 83 + 83 + 93 − 69.444
SSB = 69.667 − 69.444 = 0.223

Sum of squares within

SSW = SST − SSB = 13.556 − 0.223 = 13.333

Module 6 March 27, 2018 41 / 65

One-way ANOVA

4. Mean sum of squares

SSB .223
MSB = = = 0.115
DFBetween 2

Module 6 March 27, 2018 42 / 65

One-way ANOVA

4. Mean sum of squares

SSB .223
MSB = = = 0.115
DFBetween 2

SSW 13.333
MSW = = = 2.222
DFWithin 6

Module 6 March 27, 2018 42 / 65

One-way ANOVA

4. Mean sum of squares

SSB .223
MSB = = = 0.115
DFBetween 2

SSW 13.333
MSW = = = 2.222
DFWithin 6

5. ANOVA table:
Source SS DF MS F
MSB
Between groups 0.223 2 0.115 MSW
= 0.0517
Within groups 13.333 6 2.222

Module 6 March 27, 2018 42 / 65

One-way ANOVA

4. Mean sum of squares

SSB .223
MSB = = = 0.115
DFBetween 2

SSW 13.333
MSW = = = 2.222
DFWithin 6

5. ANOVA table:
Source SS DF MS F
MSB
Between groups 0.223 2 0.115 MSW
= 0.0517
Within groups 13.333 6 2.222
Since Cal F < FCritical = 5.14, we accept H0 .
There is no significant difference between the means of the
group.
Module 6 March 27, 2018 42 / 65
Problems

2. A random sample is selected from each of three makes of ropes

and their breaking strength (in pounds) are measured with the
following results:

Group A 70 72 75 80 83
Group B 100 110 108 112 113 120 107
Group C 60 65 57 84 87 73

Test whether the breaking strength of the ropes differs significantly.

Module 6 March 27, 2018 43 / 65

Problems

2. A random sample is selected from each of three makes of ropes

and their breaking strength (in pounds) are measured with the
following results:

Group A 70 72 75 80 83
Group B 100 110 108 112 113 120 107
Group C 60 65 57 84 87 73

Test whether the breaking strength of the ropes differs significantly.

SST = 6964.44, SSB = 5838.44, SSW = 1126, F = 38.89, F(2,15) =
3.68, Reject H0

Module 6 March 27, 2018 43 / 65

Problems

3. An experiment with 10 plots and 3 treatments gave the following

results:

Plot no. 1 2 3 4 5 6 7 8 9 10
Treatment A B C A C C A B A B
Yield 5 4 3 7 5 1 3 4 1 7

Test whether the treatments differs significantly.

Module 6 March 27, 2018 44 / 65

Problems

3. An experiment with 10 plots and 3 treatments gave the following

results:

Plot no. 1 2 3 4 5 6 7 8 9 10
Treatment A B C A C C A B A B
Yield 5 4 3 7 5 1 3 4 1 7

Test whether the treatments differs significantly.

Solution
A 5 7 1 3
B 4 4 7
C 3 1 5

Module 6 March 27, 2018 44 / 65

Problems

3. An experiment with 10 plots and 3 treatments gave the following

results:

Plot no. 1 2 3 4 5 6 7 8 9 10
Treatment A B C A C C A B A B
Yield 5 4 3 7 5 1 3 4 1 7

Test whether the treatments differs significantly.

Solution
A 5 7 1 3
B 4 4 7
C 3 1 5

SST = 40, SSB = 6, SSW = 34, F = 1.619, F(2,7) = 4.74, Accept H0

Module 6 March 27, 2018 44 / 65

RBD or Two-Way ANOVA (without replications

In a One-way ANOVA, we select the random sample for each

group or column
A Two-way ANOVA allows us to "account for variation" at the
ROW level due to some other factor or grouping
By adding blocks or factors to the ROWS, we can reduce the
overall ERROR or WITHIN variance
Now we have 4 types of Sum of Squares or Sources of
Variation: (i) TOTAL (ii) COLUMNS or GROUPS (iii) ROWS or
BLOCKS (iv) ERROR or WITHIN
Note that SST=SSC+SSR+SSE

Module 6 March 27, 2018 45 / 65

RBD or Two-way ANOVA without
replication-Example

Module 6 March 27, 2018 46 / 65

RBD or Two-Way ANOVA

Step 1:
Null hypothesis: There is no significant difference in the means of
Columns (Groups) as well as Rows (Blocks). That is,
H01 : µ1 = µ2 = . . . = µc (Columns)
H02 : µ1 = µ2 = . . . = µr (Rows)
Alternate Hypothesis: There is at least one mean in the Columns
which differs from others. Also there is at least one mean in the
Rows which differs from the others.

Module 6 March 27, 2018 47 / 65

RBD or Two-Way ANOVA

Step 1:
Null hypothesis: There is no significant difference in the means of
Columns (Groups) as well as Rows (Blocks). That is,
H01 : µ1 = µ2 = . . . = µc (Columns)
H02 : µ1 = µ2 = . . . = µr (Rows)
Alternate Hypothesis: There is at least one mean in the Columns
which differs from others. Also there is at least one mean in the
Rows which differs from the others.

Step 2:
Degrees of Freedom: DFColumns = c − 1, DFRows = r − 1,
DFError = (c − 1)(r − 1).

Module 6 March 27, 2018 47 / 65

RBD or Two-Way ANOVA

Step 1:
Null hypothesis: There is no significant difference in the means of
Columns (Groups) as well as Rows (Blocks). That is,
H01 : µ1 = µ2 = . . . = µc (Columns)
H02 : µ1 = µ2 = . . . = µr (Rows)
Alternate Hypothesis: There is at least one mean in the Columns
which differs from others. Also there is at least one mean in the
Rows which differs from the others.

Step 2:
Degrees of Freedom: DFColumns = c − 1, DFRows = r − 1,
DFError = (c − 1)(r − 1).
Compute the Critical values F(c −1,(c −1)(r −1)) and F(r −1,(c −1)(r −1)) .

Module 6 March 27, 2018 47 / 65

 PP
Step 3: G = xij ;

Module 6 March 27, 2018 48 / 65

 G2
PP
Step 3: G = xij ; Correction Factor C .F . = N

Sum of squares total

PP 2
SST = xij − C .F .

Module 6 March 27, 2018 48 / 65

 G2
PP
Step 3: G = xij ; Correction Factor C .F . = N

Sum of squares total

PP 2
SST = xij − C .F .

Sum of squares
P Cj2
SSC = cj
− C .F .
where Cj represent the column sum of j th column and cj represent
the number of observations in the j th column.

Module 6 March 27, 2018 48 / 65

 G2
PP
Step 3: G = xij ; Correction Factor C .F . = N

Sum of squares total

PP 2
SST = xij − C .F .

Sum of squares
P Cj2
SSC = cj
− C .F .
where Cj represent the column sum of j th column and cj represent
the number of observations in the j th column.
P R2
SSR = i
ri
− C .F .
where Ri represent the row sum of i th row and ri represent the
number of observations in the i th row.

Module 6 March 27, 2018 48 / 65

 G2
PP
Step 3: G = xij ; Correction Factor C .F . = N

Sum of squares total

PP 2
SST = xij − C .F .

Sum of squares
P Cj2
SSC = cj
− C .F .
where Cj represent the column sum of j th column and cj represent
the number of observations in the j th column.
P R2
SSR = i
ri
− C .F .
where Ri represent the row sum of i th row and ri represent the
number of observations in the i th row.

SSE = SST − SSR − SSC

Module 6 March 27, 2018 48 / 65

Step 4:

Source SS DF MS F
SSC
Columns (Groups) SSC c −1 c −1
F1 = MSC
MSE
SSR MSR
Rows (Blocks) SSR r −1 r −1
F2 = MSE
(or) rows
SSE
Within or Error SSE (c − 1)(r − 1) (c −1)(r −1)

Module 6 March 27, 2018 49 / 65

RBD or Two-way ANOVA

Problem 1: The following data represent the number of units

produced per day by different workers using 4 different types of
machines.

Worker / Machines I II III IV

1 44 38 47 36
2 46 40 52 43
3 34 36 44 32
4 43 38 46 33
5 38 42 49 39

1. Test whether the five men differ with respect to mean

productivity
2. Test whether the mean productivity is same for the four different
machine types.

Module 6 March 27, 2018 50 / 65

Solution

1 Null hypothesis H0 : (a) The mean productivity is same for the

four different machines. (b) Five workers do not differ with
respect to mean productivity.
Alternate hypothesis H1 : (a) The mean productivity differs for
at least one machine. (b) The mean productivity differs for at
least a worker.

Module 6 March 27, 2018 51 / 65

Solution

1 Null hypothesis H0 : (a) The mean productivity is same for the

four different machines. (b) Five workers do not differ with
respect to mean productivity.
Alternate hypothesis H1 : (a) The mean productivity differs for
at least one machine. (b) The mean productivity differs for at
least a worker.

2 Degrees of Freedom: DFColumns = 4 − 1 = 3,

DFRows = 5 − 1 = 4, DFError = 3 × 4 = 12. The critical values
are F(3,12) = 3.49 and F(4,12) = 3.26

Module 6 March 27, 2018 51 / 65

Solution

3 Calculate the deviation for all observations with respect to

some origin, say 40.
Worker / Machines I II III IV Total
1 4 -2 7 -4 5
2 6 0 12 3 21
3 -6 -4 4 -8 -14
4 3 -2 6 -7 0
5 -2 2 9 -1 8
Total 5 -6 38 -17 20

Module 6 March 27, 2018 52 / 65

Solution

3 Calculate the deviation for all observations with respect to

some origin, say 40.
Worker / Machines I II III IV Total
1 4 -2 7 -4 5
2 6 0 12 3 21
3 -6 -4 4 -8 -14
4 3 -2 6 -7 0
5 -2 2 9 -1 8
Total 5 -6 38 -17 20
G2
G = 20; N = 20; C .F . = N
= 20

Module 6 March 27, 2018 52 / 65

Solution

PP
SST = xij − C .F . = 594 − 20 = 574

Module 6 March 27, 2018 53 / 65

Solution

PP
SST = xij − C .F . = 594 − 20 = 574
P Cj2
SSC = cj
− C .F . = 181.5 − 20 = 161.5

Module 6 March 27, 2018 53 / 65

Solution

PP
SST = xij − C .F . = 594 − 20 = 574
P Cj2
SSC = cj
− C .F . = 181.5 − 20 = 161.5
P Ri2
SSR = ri
− C .F . = 358.8 − 20 = 338.8

Module 6 March 27, 2018 53 / 65

Solution

PP
SST = xij − C .F . = 594 − 20 = 574
P Cj2
SSC = cj
− C .F . = 181.5 − 20 = 161.5
P Ri2
SSR = ri
− C .F . = 358.8 − 20 = 338.8
SSE = SST − SSC − SSR = 73.7

Module 6 March 27, 2018 53 / 65

Solution

Source SS DF MS F
Columns (Groups) 338.8 3 112.93 F1 = MSC
MSE
= 18.393
Rows (Blocks) 161.5 4 40.375 F2 = MSR
MSE
= 6.576
(or) rows
Within or Error 73.7 12 6.14

Module 6 March 27, 2018 54 / 65

Solution

Source SS DF MS F
Columns (Groups) 338.8 3 112.93 F1 = MSC
MSE
= 18.393
Rows (Blocks) 161.5 4 40.375 F2 = MSR
MSE
= 6.576
(or) rows
Within or Error 73.7 12 6.14

Since F1 > 3.49, we reject the corresponding H0 . That is, the mean
productivity differs for at least one worker.
Since F2 > 3.26, we reject the corresponding H0 . That is, the fiver
workers differ with respect to the mean productivity.

Module 6 March 27, 2018 54 / 65

Try these

2. Four kinds of fertilizer f1 , f2 , f3 and f4 are used to study the yield

of beans. The soil is divided into 3 blocks, each containing 4
homogeneous plots. The yields in kilograms per plot and the
corresponding treatments are as follows:

Block 1 f1 = 42.7, f3 = 48.5, f4 = 32.8, f2 = 39.3

Block 2 f3 = 50.9, f1 = 50, f2 = 38, f4 = 40.2
Block 3 f4 = 51.1, f2 = 46.3, f1 = 51.9, f2 = 53.5

Conduct an analysis of variance at the 0.05 level of significance

using the randomized block model.

Module 6 March 27, 2018 55 / 65

Try these

3. Three varieties of potatoes are being compared for yield. The

experiment is conducted by assigning each variety at random to one
of 3 equal size plots at each of 4 different locations. The following
yields for varieties A, B, and C, in 100 kilograms per plot, were
recorded:

Location 1 B:13 A:18 C:12

Location 2 C:21 A:20 B:23
Location 3 C:9 B:12 A:14
Location 4 A:11 C:10 B:17

Perform a two-way analysis of variance to test at 5% level of

significance.

Module 6 March 27, 2018 56 / 65

Latin Square Design or Three-Way ANOVA

In addition to rows and columns, we need to consider an extra

factor known as Treatments. Every treatment occurs only once in
each row and in each column. Such a layout is known as Latin
square design. For eg. if we are interested in studying the effects of
n types of fertilizers on a yield of a certain variety of wheat, we
conduct the experiment on a square field with n 2 plots of equal area
and associate treatments with different fertilizers; row and column
effects with variations in fertility of soil.

Module 6 March 27, 2018 57 / 65

Procedure

Step 1:
Null hypothesis: There is no significant difference in the means of
Columns (Groups) , Rows (Blocks), and Treatments Alternate
Hypothesis: There is at least one mean in the Columns which
differs from others. Also there is at least one mean in the Rows
which differs from the others. Similarly for Treatments

Step 2:
Degrees of Freedom: DFColumns = n − 1, DFRows = n − 1,
DFTreatments = n − 1, DFError = (n − 1)(n − 2).
Compute the Critical value F(n −1,(n −1)(n −2))

Module 6 March 27, 2018 58 / 65

 G2
PP
Step 3: G = xij ; Correction Factor C .F . = N

Sum of squares total

PP 2
SST = xij − C .F .

Sum of squares
P Cj2
SSC = n
− C .F .
where Cj represent the column sum of j th column.
P R2
SSR = n
i
− C .F .
where Ri represent the row sum of i th row.
P T2
SSTr = n
i
− C .F .
where Ti represent the Treatment sum of i th treatment.

SSE = SST − SSR − SSC − SSTr

Module 6 March 27, 2018 59 / 65

Step 4:

Source SS DF MS F
SSC
Columns (Groups) SSC n −1 n −1
F1 = MSC
MSE
SSR MSR
Rows (Blocks) SSR n −1 n −1
F2 = MSE
(or) rows
SSTr MSTr
Treatments SSTr n −1 n −1
F3 = MSE
SSE
Within or Error SSE (n − 1)(n − 2) (n −1)(n −2)

Module 6 March 27, 2018 60 / 65

Problems

1. Analyze the variance in the Latin square of yields (in Kgs) of

paddy where P, Q, R, S denote the different methods of cultivation

S 122 P 121 R 123 Q 122

Q 124 R 123 P 122 S 125
P 120 Q 119 S 120 R 121
R 122 S 123 Q 121 P 122

Step 1:
Null hypothesis: There is no significant difference in the means of
Columns (Groups) , Rows (Blocks), and Treatments (methods of
cultivation)
Alternate Hypothesis: There is at least one mean in the Columns
which differs from others. Also there is at least one mean in the
Rows which differs from the others. Similarly for Treatments

Module 6 March 27, 2018 61 / 65

Step 2:
n=4
Degrees of Freedom: DFColumns = 3, DFRows = 3, DFTreatments = 3,
DFError = 6.
Critical value F(3,6) = 4.76
Step 3:
Calculate the deviation about the origin as 120

S2 P1 R3 Q2
Q4 R3 P2 S5
P0 Q −1 S0 R1
R2 S3 Q1 P2

Module 6 March 27, 2018 62 / 65

Treatment sum: P = 5, Q = 6, R = 9, S = 10

G = 30; N = 16
G2
C.F. = = 56.25
N
PP 2
SST = xij − C .F . = 35.75
P Cj2
SSC = n
− C .F . = 2.75
P Ri2
SSR = n
− C .F . = 24.75
P Ti2 52 62 92 102
SSTr = n
− C .F . = 4
+ 4
+ 4
+ 4
− 56.25 = 4.25
SSE = SST − SSR − SSC − SSTr = 4

Module 6 March 27, 2018 63 / 65

 Source SS DF MS F
0.917
Columns (Groups) 2.75 3 0.917 F1 = 0.667
= 1.375
8.25
Rows (Blocks) 24.75 3 8.25 F2 = 0.667
= 12.36
(or) rows
Methods 4.25 3 1.417 F3 = 10..417
667
= 2.124
Within or Error SSE 6 0.667

Comparing F1 , F2 , F3 with Critical F , we accept H0 (columns),

accept H0 (Methods or treatments), Reject H0 (Rows).

Module 6 March 27, 2018 64 / 65

Digital Assignment - 3
1. Perform a Latin square analysis for the design
C 25 B 23 A 20 D 20
A 19 D 19 C 21 B 18
B 19 A 14 D 17 C 20
D 17 C 20 B 21 A 15
2. Three varieties of potatoes are being compared for yield. The
experiment is conducted by assigning each variety at random to one
of 3 equal size plots at each of 4 different locations. The following
yields for varieties A, B, and C, in 100 kilograms per plot, were
recorded:
Location 1 B:13 A:18 C:12
Location 2 C:21 A:20 B:23
Location 3 C:9 B:12 A:14
Location 4 A:11 C:10 B:17
Perform a three-way analysis ofModule
variance
6
to test at 5% March
level of
27, 2018 65 / 65