MODULE-5

ANALYSIS OF VARIANCE
 Analysis of Variance

• Analysis of Variance is a technique developed

by R.A.Fisher
• It is used to test for the significance of the
difference among more than two sample
means
• To make inferences about whether such
samples are drawn from the populations
having the same mean.
 Analysis of Variance

Analysis of Variance is a method of splitting the total

variation of a data into constituent parts which measures
different sources of variances.
The total variation is split into two components:
(a) Variation within the subgroups of samples
(b) Variation between the subgroups of the samples
 Assumptions
• Each of the samples is a simple random
sample.
• Populations from which the samples are
selected is normally distributed .
• Each of the samples is independent of
the other samples.
• Each one of the populations has the
same variance and identical means.
• The effect of various components are
additive.
Classification of ANOVA

One way ANOVA

Two way ANOVA

 One way ANOVA

One way classification model is

designed to study the effect of one
factor in an experiment.
 Example

• The following table gives the yields on 15 sample fields under

three varieties of seeds (viz. A, B, C):
Yields
A B C
5 3 10
6 5 13
8 2 7
1 10 13
5 0 17

• Test at 5% level of significance

 Solution

• Hypothesis

H 0 : 1   2
2 2

H1 :  1   2
2 2
 Calculation of mean of each sample and
grand average

X 1  (5  6  8  1  5) / 5  25 / 5  5
X 2  (3  5  2  10  0) / 5  20 / 5  4
X 3  (10  13  7  13  17) / 5  60 / 5  12
X  (5  4  12) / 3  7
Calculation of sum of squares between the
samples (SSB)


A  X1  X 
2

B  X2  X 
2

C  X3  X 
2

(5  7) 2  4 (4-7)^2 =9 (12-7)^2=25

(5  7) 2  4

(5  7) 2  4

(5  7) 2  4

(5  7) 2  4
TOTAL= 20 45 125

SSB= 20+45+125 = 190

 Calculation of sum of squares within the
samples (SSW)


A  X1  X1 
2

B  X2  X2 
2

C  X3  X3 
2

(5-5)^2=0

16

TOTAL= 26 58 56

SSW=26+58+56 = 140
 Contd….

• Prepare ANOVA table as follows:

Source of Sum of Degree of Mean Computed Table value
Variation Squares freedom Squares value of F of F

Between SSB = 190 c-1 = 3-1 = 2 MSB = F= 3.88

Samples SSB/c-1 = MSB/MSW=
190/2 = 95 95/11.67 =
8.14
Within SSW = 140 n-c = 15-3 = MSW=
Samples 12 SSW/n-c =
140/ 12 =
11.67
Total SST= 330 n-1= 15-1
=14
 INTERPRETATION

• Since the computed value of F is greater than

the table value of F, we reject the null
hypothesis and conclude that there is
significant difference in the variances. Hence,
the average yield of land under different
varieties of seed show significant differences.
 Two way Classification Model
Two way classification model is designed to study the effects of
two factors simultaneously in the same experiment.
ANOVA Table
S.V Sum of Degree of Mean Squares Variance Ratio
Squares freedom
Between SSC c-1 MSC = SSC/c-1 F1 = Greater
Columns variance/ Smaller
Variance

Within Samples SSR r-1 MSR = SSR/r-1 F2 = Greater

variance/smaller
Variance
Residual/Error SSE (c-1) (r-1) MSE = SSE /(c-1) (r-1)
Total SST rc-1
 sum of squares of total of each columns
SSC   correction factor
no. of items in each column
sum of squares of total of each row
SSR   correction factor
no. of items in each row
SST  sum of squares of all the obserations - correction factor
SSE  SST - (SSC  SSR)
T2 
correction factor   
N
 Question

• The following table gives per hectare yield for three varieties
of wheat each grown on fie plots:

Per hectare yield (in tons)

Varieties of wheat
Plot of land A B C
1 5 3 10
2 6 5 13
3 8 2 7
4 1 10 13
5 5 0 17

• Test at 5% level of significance

 Solution

H 0 : 1   2   3
H1 : At least two of the population means are unequal
Calculation of sum of observations of each row
and each column and their grand total

A B C Row Total
5 3 10 18
6 5 13 24
8 2 7 17
1 10 13 24
5 0 17 22
Column 20 60 105 (Grand
Total=25 Total)
 Correction factor

T2
 (25  20  60)2
    735
N  15
SSC (Sum of squares between column)

sum of squares of totals of each column

SSC   correction factor
no. of items in each column
 (25) 2 (20) 2 (60) 2 
SSC       735  190
 5 5 5 
 SSR (sum of squares between rows)

sum of squares of total of each row

SSR   correction factor
no. of items in each row

 (18) 2 (24) 2 (17) 2 (24) 2 (22) 2 

SSR         735  14.66
 3 3 3 3 3 
 SST (Total Sum of Squares)

• SST=sum of squares of all the observations-correction factor

 2 2 2 2 2

SST  (5)  (6)  (8)  (1)  (5)  ...... (17)  735  330
2

SSE  SST  (SSC  SSR)  330  (190  14.66)  125.34

 Solution
S.V Sum of Degree Mean Squares Computed value of F Table value of F
Squares of
freedom
Between SSC=190 2 MSC = 95 F1 =95/15.67 = 6.06 4.46
Columns
Within SSR=14.7 4 MSR = 3.665 F2= 15.67/3.665=4.28 6.04
Samples
Error SSE=125.3 8 MSE =15.67
Total SST=330 14

1) 6.06>4.46. Reject H0. There is significant difference in the variances between

varieties of wheat
2) 4.28<6.04 . Accept H0. There is no significant difference in the variances between
plots of land.
 Question

• The following table gives per hectare yield for three varieties
of wheat each grown on fie plots:

Per hectare yield (in tons)

Varieties of wheat
Plot of land A B C
1 30 26 38
2 24 29 28
3 33 24 35
4 36 31 30
5 27 35 33

• Test at 5% level of significance