One-way

ANOVA
Quarter 3 –
Analysis of Variance (ANOVA)
Also known as F-test
• used in comparing the means of three
or more independent groups.
• used if the main SOP is about
significant difference on a variable
when the respondent are grouped
into 3 or more groups.
Analysis of Variance (ANOVA)
Hypothesis
• Hₒ: x₁=x₂=x₃
• Hₐ: At least one group is different
from another
EXAMPLE:
To determine if there is a significant difference on the average
rice yield using 3 types of fertilizer, an experiment was
conducted, and the result of the study is shown below. Test at
0.05 level of significance.
Fermented Rice Hull Commercialized
Plant Juice Fertilizer
25 15 23
27 18 24
28 17 25
30 20 22
26 19 25
1. Formulate the null hypothesis and alternative hypothesis.
2. Follow the procedures for the one-way ANOVA.
3. Compare the computed value with the tabular value then make a decision.
4. Draw a conclusion.
To get the F-computed value, the
following computations should be done:
STEP 1: Compute for the various sum of
squares (SS)
 Sum of Squares “Total”, SST = -
 Sum of Squares “Between”, SSB = -
 Sum of Squares “Within”, SSW = SST -
SSB
To get the F-computed value, the
following computations should be done:
STEP 2: Find the degrees of freedom
(df) for the various sum of squares
 df for total, = N-1, where N is the total
number of samples
 df for between, = k-1, where k is the
number of groups to be compared
 Df for within, = -
To get the F-computed value, the
following computations should be done:
STEP 3: Determine the various mean
squares (MS)
 Mean Squares “Between”, MSB =
 Mean Squares “Within”, MSW =
To get the F-computed value, the
following computations should be done:
STEP 4: Compute for F-ratio, F =

STEP 5: Locate the tabular value using

(2) as the numerator and (12) as the
denominator in the Critical Values of F
table.
To get the F-computed value, the
following computations should be done:
Step 6: Summarize the source of
variations and the computed F-ratio in
a table for the one-way ANOVA
SS df MS
Sources of F
(Sum of (Degrees of (Mean
Variations (F-value)
Squares) Freedom) Squares)
Between
Groups
Within Groups
Total
 ANOVA TABLE
SS df MS
Sources of F
(Sum of (Degrees of (Mean
Variations (F-value)
Squares) Freedom) Squares)
Between K-1
Groups 226.53 MSB =
3–1=2 113.27
Within Groups (N-1) – (K- F=
36.4 1) MSW =
14 – 2 = 12 3.03 F = 37.38
Total N-1
262.93
15 – 1 = 14

Since the F-computed value of 37.38 is greater than the F-tabular value of 3.88 at 0.05
level of significance with =2 and =12, reject the Ho, accept the Ha.
 DMRT
(Duncan’s Multiple Range Test)

Quarter 3 –
Duncan’s Multiple Range Test
•(DMRT)
is a post ANOVA test which is used whenever
ANOVA leads to the rejection of the null
hypothesis.

• Duncan’s test is based on the general notion of

studentized range. The range of any subset p
sample means must exceed a certain value (Rp)
before any of the p means are said to be
different. This value is the least significant range
for the p means and is denoted by Rp.
 =

Where:
rp = can be obtained from the table
MSW = is the MSW in the ANOVA table
n = is the equal sample size
To determine which among the different
groups have a significant different use
DMRT.
Follow the steps:
Step 1: Arrange the means from lowest to
highest.
Step 2: Get MSW and from the ANOVA table.
Step 3: Find rp by getting the intersection of p
and df in the table.
Step 4: Prepare a Duncan’s table and solve for
Rp .
EXAMPLE:
To determine if there is a significant difference on the average
rice yield using 3 types of fertilizer, an experiment was
conducted, and the result of the study is shown below. Test at
0.05 level of significance.
Fermented Rice Hull Commercialized
Plant Juice Fertilizer
25 15 23
27 18 24
28 17 25
30 20 22
26 19 25
EXAMPLE:
To determine if there is a significant difference on the average
rice yield using 3 types of fertilizer, an experiment was
conducted, and the result of the study is shown below. Test at
0.05 level of significance.
Fermented Rice Hull Commercialized
Plant Juice Fertilizer 1. RH, = 17.8
25 15 23 = 23.8
27 18 24
= 27.2
28 17 25
30 20 22
26 19 25
= 27.2 = 17.8 = 23.8
 ANOVA TABLE
SS df MS
Sources of F
(Sum of (Degrees of (Mean
Variations (F-value)
Squares) Freedom) Squares)
Between K-1
Groups 226.53 MSB =
3–1=2 113.27
Within Groups (N-1) – (K- F=
36.4 1) MSW =
14 – 2 = 12 3.03 F = 37.38
Total N-1
262.93
15 – 1 = 14

Since the F-computed value of 37.38 is greater than the F-tabular value of 3.88 at 0.05
level of significance with =2 and =12, reject the Ho, accept the Ha.
 DUNCAN’S TABLE
p 2 3 Number of groups to be compared.

Least Significant Studentized Ranges

rp 3.033 3.178 rp at = 0.05 based on the computed
Rp The intersection of p and in the
table.
Computed
=

To make an analysis of the result:

a. Find the difference of the means of the two groups.
b. If the difference exceeded the value of or the difference is
greater than , there is a significant difference between the
two groups.
=
Where:
rp = can be obtained from the
table
MSW = is the MSW in the ANOVA
table
n = is the equal sample size
= =
=2.36

= =
=2.47
 DUNCAN’S TABLE
p 2 3 Number of groups to be compared.

Least Significant Studentized Ranges

rp 3.033 3.178 rp at = 0.05 based on the computed
Rp 2.36 2.47 The intersection of p and in the
table.
Computed
=

To make an analysis of the result:

a. Find the difference of the means of the two groups.
b. If the difference exceeded the value of or the difference is
greater than , there is a significant difference between the
two groups.
ANALYSIS OF THE RESULTS.

1. Fermented Plant Juice and Commercialized Fertilizer

27.2- 23.8 =3.4 > 2.36 (R2), the difference exceeded R2
There is a significant difference between the mean rice yield of Fermented
Plant Juice and Commercialized Fertilizer. Fermented Plant Juice is a better
fertilizer for rice plants. (Basis: Fermented Plant Juice have a higher mean
than Commercialized Fertilizer.)

2. Fermented Plant Juice and Rice Hull

27.2 – 17.8 = 9.4 > 2.47 (R3), the difference exceeded R3
There is a significant difference between the mean rice yield of Fermented
Plant Juice and Rice Hull. Fermented Plant Juice is a more effective fertilizer
for rice plants. (Basis: Fermented Plant Juice have a higher mean than Rice
Hull.)
ANALYSIS OF THE RESULTS.

3. Commercialized Fertilizer and Rice Hull

23.8-17.8 =6 > 2.36 (R2), the difference exceeded R2
There is a significant difference between the mean rice yield of
Commercialized Fertilizer and Rice Hull. Commercialized Fertilizer is a
better fertilizer for rice plants. (Basis: Commercialized Fertilizer have a
higher mean than Rice Hull)

Therefore, fermented plant juice is the most effective fertilizer for rice
plants.
EXAMPLE:
A researcher wants to test three different diets designed to
maximize weight gain. The weight gains, in kilogram, after a
three-week period are shown below. Test if there is a significant
difference in the weight gain of the pigs using the three
different diets. Use 0.05 level of significance.
Diet A Diet B Diet C
10.5 kg 12.1 kg 6.5 kg
10.9 kg 12 kg 7 kg
10.7 kg 13 kg 6.4 kg
10.3 kg 12.5 kg 7.1 kg
10.35 kg 12.7 kg 6.8 kg
10.3 kg 14 kg 7 kg
 ANOVA TABLE
SS df MS
Sources of F
(Sum of (Degrees of (Mean
Variations (F-value)
Squares) Freedom) Squares)
Between
K-1
Groups 107.27 MSB =
3–1=2
53.64
F=
Within Groups (N-1) – (K-
3.39 1) MSW
F = 233.22
17 – 2 = 15 =0.23
Total N-1
110.66
18 – 1 = 17
Since the F-computed value of 233.22 is greater than the F-tabular value of 3.68 at 0.05
level of significance with 2 and 15 degrees of freedom, reject the Ho, accept the Ha.
2. A researcher is interested in comparing the yield (in sacks of 100 kilos
each) of four rice varieties. The yield per hectare are as follows. Find
out if there is a significant difference in the average yield of the four rice
varieties at 0.05 level of significance.

Variety A Variety B Variety C Variety D

86 75 100 90
89 78 105 80
80 80 98 85
85 77 102 83
87 79 99 85
 ANOVA TABLE
SS df MS
Sources of F
(Sum of (Degrees of (Mean
Variations (F-value)
Squares) Freedom) Squares)
Between K-1
1416.55
Groups 4–1=3 MSB = 472.18
Within Groups (N-1) – (K- F=
144 1)
MSW =9
19 – 3 = 16 F = 52.46
Total N-1
1560.55
20 – 1 = 19

Since the F-computed value of 52.46 is greater than the F-tabular value of 3.24 at 0.05
level of significance with 3 and 16 degrees of freedom, reject the Ho, accept the Ha.
Seatwork
EXAMPLE:
A pharmacy is selling four brands of Vitamin C. The owner is
interested if there is significant difference in the average sales
of the four brands of vitamin C for one week. The following data
are recorded.
BRAND Perform the analysis of variance
A B C D and test the hypothesis at 0.05
7 9 2 4 level of significance that the
average sales of four brands of
3 8 3 5
vitamin C are equal.
5 8 4 7 1. Formulate
hypothesis.
the null hypothesis and alternative

6 7 5 8 2. Follow the procedures for the one-way ANOVA.

3. Compare the computed value with the tabular
9 6 6 3 value then make a decision.
4. Draw a conclusion.
4 9 5 4
3 10 2 5
 ANOVA TABLE
SS df MS
Sources of F
(Sum of (Degrees of (Mean
Variations (F-value)
Squares) Freedom) Squares)
Between
K-1
Groups
F=
Within Groups (N-1) – (K-
1)
Total N-1