Chapter 8:

Analysis of
Variance

Analysis of Variance
-Analysis of Variance or ANOVA is a
comparison test used to determine the
significant difference among normal
population means. The comparison in means
of three (3) or more populations, which
follow normal distributions, can be taken
simultaneously in just one application of
this test. This test is therefore a
generalization of the z and t tests of two
normal population means. This test was
developed by Sir Ronald A. Fisher (18901962).

The following assumptions should

be met in the use of ANOVA:
1. The various groups are assumed
to be with normal populations.
2. The variance of the different
groups is assumed to be equal.
3. The random samples in the groups
should be independent.

Formulas:
1. Total sum of Squares (TSS)

TSS=X - (X)/N
Where TSS= Total sum of squares
X= Value of each entry
N= Total number of items or entry
2. Sum of Squares Between-Columns (SSb)

SSb = 1/No. of Rows (sum of each column) (X)/N

3. Sum of Squares Within-Column (SS w)

SS w = TSS- SSb

4. Mean Sum of Squares Between (MSS b)

MSSb = SSb /
Where

df
b

= no. of columns 1

df
b

5. Mean Sum of Squares Within (MSSw)

MSSw = SSw/dfw
Where
6.

df
w

= (row*column) c

F-Value = MSSb / MSSw

Problem 10:
Let us consider three groups of seven
students, where each group is
subjected to one of the three
strategies or methods of teaching.
Group A was exposed to Explanatory
Approach, Group B for Cooperative
Learning, and Group C for Traditional
Method. The grades of the students
are presented below. Test if there is a
difference in the three methods or
strategies of teaching at 5%level of
significance.

Studen
t

Group
A
Explan
a-tory

Group
Group
B
C
Cooper
Tradia-tive
tional

Xa

Xb

Xc

Group
A

Group
B

Group
C

(Xa)

(Xb)

(Xc)

85

86

100

7225

7396

10000

90

88

89

8100

7744

7921

92

89

88

8464

7921

7744

88

90

87

7744

8100

7569

91

87

83

8281

7569

6889

93

88

85

8649

7744

7225

89

91

80

7921

8281

6400

Steps:
1. Ho: There is no significant difference
among the three methods or
strategies of teaching.
2. = 5%
3. Test statistic to be used: ANOVA
4. Solution: Compute for
1. TSS = X - (X)/N
= 164,887- (1859)/21
= 164,887-164,565.76
= 321.24

Where: X = (Xa)+(Xb)+(Xc)
= 56384+54755+53748
= 164887
X = (Xa)+(Xb)+(Xc)
= 628+619+612
= 1859
2. Sum of Squares Between-Column (SSb)
SSb = 1/ No. of Rows (sum of each column)
- (x)/N
=1/7 (628+619+612) (1859)/21
= 164584.14-164565.76
= 18.38

3. Sum of Squares Within-Column(SSw)

SSw = TSS-SSb
= 321.24-18.38
= 302.86
4. Mean Sum of Squares Between(MSS b)
MSSb = SSb / dfb
= 18.38 / 2
= 9.19
Where: dfb = no. of columns 1
= 3-1
=2

5. Mean Sum of Squares Within (MSS w)

MSSw = SSw/dfw
= 302.86/18
= 16. 825
Where: dfw = (row*column)-3
= (7x3)-3
= 18
6. F-Value = MSSb / MSSw
=9.19/16.83
=0.546
DF = 2 and 18
T.V. = 3.55

After the sum of squares have been computed,

a summary table has to be presented:
Source of
Variation

Sum of
Squares

Degrees of
Freedom

Mean Sum
of Squares

Computed
F

Between
Column
Within Column
Total

18.38
302.86
321.24

3-1=2
(3x7)-3=18
20

9.19
16.83

0.546

7. Accept Ho since the computed value of 0.546 is less than

the tabular value of 3.55
8. Therefore, the three strategies of teaching are not
significantly different from each other at an alpha of 5%.
In as much as the result of the study is not significant, the
researcher may stop at this point for his generalizations.
But if the results of the study showed a significant result,
the data will still be subjected for further testing to
determine which of the pairs will show a significant
difference in means.

Computations of ANOVA using MicroStat

One-way ANOVA
GROUP MEAN N
1 89.7147
2 88.4297
3 87.4297
GRAND MEAN
88.52421
Source Sum of Squares

D.F.

Mean Square F. Ratio

BETWEEN 18.381 2 9.1900.546 0.5884

302.857 18 16.825
TOTAL
321.238 20

Prob.

WITHIN

From MegaStat
One factor
ANOVA
Mean

Std. Dev.

88.52380952

89.7

2.69

Group 1

88.52380952

88.4

1.72

Group 2

88.52380952

87.4

6.35

Group 3

88.5

21

4.01

Total

SS

df

MS

p-value

Treatment

18.38

9.190

0.55

.5884

Error

302.8

18

16.825

ANOVA table
Source

Exercise 22
ANOVA
Name:
Date:
Course & Year:
Score:
Solve completely the following problems:
1. Three brands of reducing pills were tried on a
sample of 10 female adults; the data are reflected
on the table below in terms of weight loss (lb) after
a Responden
month of using
these
Brand
A pills.
Brand B
Brand C
ts
1
2
3
4
5
6
7
8
8
10

4.1
3.1
3.6
4.2
3.8
4.7
4.1
2.8
3.0
4.2

3.1
3.3
3.5
4.9
4.1
3.9
4.0
3.9
4.1
4.0

3.6
3.8
3.0
3.1
3.2
3.3
3.9
4.6
2.9
4.2

continuation.
Test if there is a significant
difference in the average weight
loss (in lb) among the three
groups of respondents using the
three brands of reducing pills at
0.05 level of significance.

Exercise 22
ANOVA
Name:
Date:
Course & Year:
Score:
2. Based from the survey results shown below,
determine if there is a significant difference existing
in the mean achievement of students from the three
non-sectarian schools in Tuguegarao City by using )
0.01
levelNo.
of significance.
Student
School A
School B
School C
1
2
3
4
5
6
7
8
9
10

76
86
88
90
81
75
87
89
92
85

83
82
85
81
96
79
83
93
89
90

82
90
83
86
92
88
75
89
77
90

Exercise 22
ANOVA
Name:
Date:
Course & Year:
Score:
3. The following are the heights in inches of
six male college students of Cagayan
Colleges Tuguegarao from the three regions
of the country. Is there an evidence of
height variation among these groups using
Region 1
Region 2
Region 3
the 0.05 level?
1
2
3
4
5
6

63
58
69
72
60
63

57
63
63
69
61
66

75
72
60
59
61
50

MEASURES OF CORRELATION
-Correlation is a statistical tool to
measure the association of two or more
quantitative variables. It is concerned
with the relationship in the change and
movements of two variables. It is also
defined as the measure of the linear
relationship between two random
variables x and y and is denoted by r. It
measures the extent to which the
points cluster about a straight line.

Three degrees of relationship or correlation between two

variables
1. Perfect correlation (positive and negative)
2. Some degrees of correlation (positive and negative)
3. No correlation
The quantitative interpretation of the degree of linear relationship
existing is shown below.
1.00 Perfect Correlation (negative) correlation
0.91 - 0.99 Very high positive (negative) correlation
0.71 - 0.90 High positive (negative) correlation
0.50 - 0.70 Moderately positive (negative) correlation,
substantial
relationship
0.31 - 0.50 Low positive correlation (negative )

0.01- 0.30 Slight correlation, negligible positive

(negative) correlation
0 No correlation

The five figures in the next page illustrate the

degree of correlation between two variables.
- Figure A is a perfect positive correlation, which
relates two variables whose values are both
increasing.
- Figure B is a perfect negative correlation
describes a situation where one variable
increases, the other variable decreases.
- Figure C and D, some degree of positive or
negative correlation which relates two variables
whose value ranges from 0.1 - 0.99
- Figure E, No correlation, which describes a
situation whose correlation coefficient is 0.

Summary of the different types or degrees of

correlation between two variables

Figure A
Figure B
Perfect Positive Correlation

Perfect Negative Correlation

Figure C Figure DFigure E

Some Degree of Positive Correlation

Some Degree of Negative Correlation

No Correlation

Spearman Rank
-It is used to determine the
degree of relationship of two
variables expressed as ORDINAL
DATA.
Formula:
Rs = 1 - 6D2/ N3 - n

Problem 11: Randomly,

selected jobs are ranked and
stress. Does a significant
relationship exist between
salary and stress using 0.05
level of significance?

Job

Salary
Rank

Stress
Rank

D2

Lawyer

Zoologist

Doctor

College
Dean

Hotel
Manager

Bank
Officer

10

Safety
Inspec

Police
Officer

10

Teacher
Pilot

4
3
1
Solution:
1
1
0
Rs = 1 6 D2 / N3 n

= 1 6(24) / (10)3 10 = 0.145

1
0
D2=24

From MegaStat:
Salary
Rank
Salary
Rank
Stress
Rank

Stress Rank

1.000

.855 1.000
10 Sample size
.632 Critical value .
05(two-tail)
. 765 Critical value .
01(two-tail)

The Pearson Product-Moment

Correlation Coefficient

The Pearson r from Raw

Scores

Formula:

r = N XY - X Y / X2- (X) 2] [NY2 (Y)2]

Where:
N= No. of Class
XY= Sum of the products of X and Y
X= Sum of X
Y= Sum of Y
X2= Sum of the squares of X
Y2= Sum of the squares of Y

Problem 12: Height and weight of 10

basketball players; find the Pearson r.
X
(height
in
inches)

Y(Weigh
t in
kilos)

XY

X2

Y2

65

65

4225

4225

4225

64

64

4096

4096

4096

78

70

5460

6084

4900

72

71

5112

5184

5041

69

65

4485

4761

4225

66

66

4356

4356

4356

70

68

4760

4900

4624

71

69

4899

5041

4761

70

70

4900

4900

4900

67

71

4757

4489

5041

X=692

Y=679

XY=47,0 X2=48,0
50
36

Y2=46,1
69

Solution:
r= XY - X Y / X2- (X) 2] [NY2
(Y)2]
= 10(47050)-(692)(679) /
2]-[10(46169)-(679) 2]
= 470500-469868 /

Continuation
= 632 /
= 632 /
= 632 / 985.35
=0.64
Interpretation: r= 0.64 is moderately positive
correlation. There is substantially degree of
correlation between the height and weight of
10 basketball players.

TESTING THE
SIGNIFICANCE OF r
-The test for significance of r is needed in order to
know, whether the computed r is significant or
not.
Solutions:
1. Ho: There is no significant relationship between
height and weight of the 10 basketball players.
. Ha: There is a significant relationship between
the height and weight of the 10 basketball
players.
2. Level of significance= 5%
df=10-2=8
Tabular Value= 1.859548

3. Test statistic to be used is t for r

4. Compute for t:
t = r / 1 r2 / n-2
t= 0.64 / 1 (0.64)2 / 10 2
t= 0.64 / 0.0738
t= 0.64 / 0.271661554
t= 2.36
5. Decision: The computed value 2.36 is greater the
tabular value 1.8535. Hence, the null hypothesis is
rejected.
6. Interpretation: There is a significant relationship
between the height and weight of the 10 basketball
players

REGRESION ANALYSIS
Bivariate Linear Regression
-Simple and multiple predictions are made with a
technique called Regression Analysis.
Linear Regression Analysis
-We now go beyond the notion of association and
relation to try to examine (possible) casualty
(or prediction). Sometimes, given information
about one characteristics of a phenomenon, we
can have some idea about the nature of
another characteristics.

Continuation..
A statistical technique designed to predict
values dependent variable from knowledge
of the values of the one or more
independent variable.
It uses the principle of ordinary least
squares where line is drawn through a
scatter plot that minimizes the sum of
squared residuals. In other words, a line is
drawn as close as possible to all the cases
in the sample.
When one takes the values of X to
estimate or predict corresponding Y values,
the process is called simple prediction.

Continuation..
Examples:
We associate high caloric intake with body
weight.
If we know the temperature in Celsius, we
can calculate the value in Fahrenheit.
In Social Sciences, we infer the high income
or high education lowers the desired family
size.
We can make these inferences, but we are
not accurate. Therefore, regression is
designed to help us to determine the
probability that our inferences are sound.
Put differently, it helps us to test the degree

GUIDELINES FOR USING LINEAR

REGRESSION:
If there is no significant linear
correlation, do not use the regression
equation to make predictions.
When using the regression equation for
predictions, stay within the scope of the
available sample data.
A regression equation based on old
data is not necessarily valid now.
Dont make predictions about a
population that is different from the
population from which the sample data

The Basic Bivariate Regression

Equation
Non-Stochastic Equation
-It is an error-free equation used to predict
the value of y. It is an equation for perfect
correlations.
Formula: Y = a + bx (exact relationship)
Stochastic Equation
-It is an equation where the estimate yields
an error.
-It is usually common in problems on social
sciences.

Where:
Y = independent or response variable
X = independent or predictor variable (called
explanatory or regressor variable)
a = y-intercept
b = slop of a line
e = residual or error term
e=Y
Where:
= the estimated value of Y using the
rergression equation

Formula:
a = ( Y) ( X2) (X) (XY) / N (X2)
(X)2
b = N (XY) ( X) (Y) / N (X2) (X) 2
a = (679) (48036) (692) (47050) / 10
(48036) (692)2 = 38.666
b = 10 (47050) (692) (679) / 10
(48036) (692)2 = 0.4225

Scatter Plot
-They provide a mean for visual
inspection of data that a list of values
for two variables cannot. They are
essential for understanding the
relationship between variables.

This scatter diagram makes 3 things clear:

1. There seems to be a moderate positive
relationship between x and y.
2. No straight line could be possibly drawn
that would pass through all the point;
most of the points would be above or
below the best fitting line (least square
regression line.)
3. If the relationship were a deterministic
(r=1), the point would lie on the straight
line.

Regression
Analysis

r2

0.411

10

0.641

Std. Error

2.185

Dep. Var

Y(weight in
kilos)

ANOVA table
Source

SS

df

MS

p-value

Regression

26.6995

26.6995

5.59

.0456

Residual

38.2005

Total

64.9000

4.7751

Interpretation: Since probability value is less

Confidence
than alpha, the Ho is rejected. Therefore, there
Interval is
Variables
Coefficients relationship
Std. error
t= between
(df=8)
p-value
95%
lower
a significant
height
and
Intercept weight
38.6658of the
12.3825
3.123
.0142
10.1117
10 basketball
players.

Regression
Output

95% upper
67.2198

Exercise 23
Measures of Correlation
Name
Date
Course and Year
Score
Solve for the coefficient of correlation using the Pearson r formula or
Spearman. Rank the following:
1. Ten students were given tests in Statistics and English. The results are
shown below:
Statistics
English
87

90

67

60

67

76

61

89

67

58

90

91

50

78

78

89

92

90

87

88

Exercise 23
Measures of Correlation
Name
Date
Course and Year
Score
2. The table below shows how the nutrition experts and heads of
household ranked 10 breakfast foods based on their palatability.
Nutrition
Experts

Heads of
household

10

10

Name
Course and Year

Exercise 23
Measures of Correlation
Date
Score

3. The 10 weeks sales of ABC Department Store in Tuguegarao City and its
branch in Santiago City
Sales of ABC Store Sales of ABC
in Tuguegarao City Store in Santiago
City
31

71

42

60

73

11

82

43

91

22

23

35

19

50

28

35

55

18

63

39

Group 8
Santiago, Jarys Christian C.
Santos, Akie
Sarmiento, Lalli Anna
Seduguchi, Kasumi
Valle, Coleen H.
Vallente, Abiathar
Villasper, Arbin
Sumayod, Cressa
Lozada, Elijah

TABLE D. CRITICAL VALUES OF F

TABLE F. SPEARMAN RANK CORRELATION

COEFFICIENT

TABLE E. PEARSON

Table B. Students t-Distribution

df

0.40

0.25

0.10

0.05

0.025

0.01

0.005

0.32492
0

1.00000
0

3.077684

6.313752

12.70620

31.82052

63.6567
4

0.28867
5

0.81649
7

1.885618

2.919986

4.30265

6.96456

9.92484

0.27667
1

0.76489
2

1.637744

2.353363

3.18245

4.54070

5.84091

0.27072
2

0.74069
7

1.533206

2.131847

2.77645

3.74695

4.60409

0.26718
1

0.72668
7

1.475884

2.015048

2.57058

3.36493

4.03214

0.26483
5

0.71755
8

1.439756

1.943180

2.44691

3.14267

3.70743

0.26316
7

0.71114
2

1.414924

1.894579

2.36462

2.99795

3.49948

0.26192
1

0.70638
7

1.396815

1.859548

2.30600

2.89646

3.35539

0.26095
5

0.70272
2

1.383029

1.833113

1.26216

2.82144

3.24984

10

0.26018
5

0.69981
2

1.372184

1.812461

2.22814

2.76377

3.16927

df

0.4

0.25

0.10

0.05

0.025

0.01

0.005

16 0.257885

0.69119
7

1.34060
6

1.75305
0

2.13145

2.60248

2.9467
1

17 0.257347

0.68919
5

1.33337
9

1.73960
7

2.10982

2.56693

2.8982
3

18 0.257123

0.68836
4

1.33039
1

1.73406
4

2.10092

2.55238

2.8788
4

19 0.256923

0.68762
1

1.32772
8

1.72913
3

2.09302

2.53948

2.8609
3

20 0.256743

0.68695
4

1.32534
1

1.72471
8

2.08596

2.52798

2.8453
4

21 0.256580

0.68635
2

1.32318
8

1.72074
3

2.07961

2.51765

2.8313
6

22 0.256432

0.68580
5

1.32123
7

1.71714
4

2.07387

2.50832

2.8187
6

23

0.68530
6

1.31946
0

1.21387
2

2.06866

2.49987

2.8073
4

24 0.256173

0.68485
0

1.31783
6

1.71088
2

2.06390

2.49216

2.7969
4

25 0.256060

0.68443

1.31634

1.70814

2.05954

2.48511

2.7874

0.25697

Df

0.95

0.90

0.70

0.50

0.20

0.10

0.05

0.02

0.01

0.0039
3

0.015
8

0.148 0.455 1.642 2.706 3.841 5.412

6.635

0.103

0.211

0.713 1.386 3.219 4.605 5.991 7.824

9.210

0.352

0.584

1.424 2.366 4.642 6.251 7.815 9.837

11.34
6

0.711

1.064

2.195 3.357 5.989 7.779 9.488 11.66

8

13.27
7

1.145

1.610

3.000 4.351 7.289 9.236 11.07 13.38

0
8

15.08
6

1.635

2.204

3.828 5.348 8.558 10.64 12.59 15.03

5
2
3

16.81
2

2.167

2.833

4.671 6.346 9.803 12.01 14.06 16.62

7
7
2

18.47
5

2.733

3.490

5.527 7.344 11.03 13.36 15.50 18.16

0
2
7
8

20.09
0

3.325

4.168

6.393 8.343 12.24 14.68 16.91 19.67

2
4
9
9

21.66
6

10

3.940

4.865

7.267 9.342 13.44 15.98 18.30 21.16

2
7
7
1

23.20
9

11

4.575 5.578 8.146

10.34
1

14.631 17.27 19.67 22.61 24.72

5
5
8
5

12

5.226 6.304 9.034

11.30

15.812 18.54 21.02 24.05 26.21

9
6
4
7

13

5.892 7.042 9.926

12.34
0

16.985 19.81 22.36 25.47 27.68

2
2
2
8

14

6.571 7.790 10.82

1

13.33
9

18.151 21.06 23.68 26.87 29.14

4
5
3
1

15

7.261 8.544 11.72

7
1

14.33
9

19.311 2.307 24.99 28.25 30.57

6
9
8

16

7.962 9.312 12.62

4

15.33
8

20.465 23.54 26.29 29.63 32.00

2
6
3
0

17

8.672 10.08 13.53

5
1

16.33
8

21.615 24.76 27.58 30.99 33.40

9
7
5
9

18

9.390 10.86 14.44

5
0

17.33
8

22.760 25.98 28.86 32.34 34.80

9
9
6
5

19

10.11 11.65 15.35

7
1
2

18.33
8

23.900 27.20 30.14 33.68 36.19

4
4
7
1

20

10.85 12.44 16.26

1
3
6

19.33
7

25.038 28.41 31.41 35.02 27.56

2
0
0
6

2
1

11.59
1

13.24
0

17.18
2

20.33
7

26.17
1

29.61
5

32.67
1

36.34
3

38.93
2

2
2

12.33
8

14.04
1

18.10
1

21.33
7

27.30
1

30.81
3

33.92
4

37.65
9

40.28
9

2
3

13.09
1

14.84
8

19.02
1

22.33
7

28.42
9

32.00
7

35.17
2

38.96
8

41.63
8

2
4

13.84
8

15.65
9

19.94
3

23.33
7

29.55
3

33.19
6

36.41
5

40.27
0

42.98
0

2
5

14.61
1

16.47
3

20.56
7

24.33
7

30.67
5

34.38
2

37.65
2

41.56
6

44.31
4

2
6

15.37
9

17.29
2

21.79
2

25.33
6

31.79
5

35.56
3

38.88
5

42.85
6

45.64
2

2
7

16.15
1

18.11
4

22.71
9

26.33
6

32.91
2

36.74
1

40.11
3

44.14
0

46.96
3

2
8

16.92
8

18.93
9

23.64
9

27.33
6

34.02
7

37.91
6

41.33
7

45.41
9

48.27
8

2
9

17.70
7

19.75
8

24.37
7

28.33
6

35.13
9

39.08
7

42.55
7

46.69
3

49.58
8

3
0

18.40
3

20.59
9

25.50
8

29.33
6

36.25
0

40.25
6

43.77
3

47.96
2

50.89
2