Shree Ganesh
An Introduction to
Two-Way ANOVA
PREPARED BY: -
Dr. PORAS PATEL
Call & Whatsapp : 091-7874572005 (M)
Email Id: [email protected]
1
CONTENTS
Introduction
Assumption of two way anova
Related terminology in two way anova
Two way anova calculations-manually
Advantages of two-way anova
Examples covering two way anova – using
MS Excel
2
ANOVA : What is it?
An ANOVA (Analysis of Variance),
sometimes called an F test, is closely
related to the t test.
The major difference:
t - test ANOVA tests
measures the measures the
difference difference
between the between the
means of two groups means of
more than two groups
3
Two-Way ANOVA (Factorial ANOVA)
Extension of one way ANOVA
There are two independent variables (Hence
the name two way)
Two-way ANOVA is an extension of the
paired t test to more than two treatments
4
What is two way independent ANOVA?
Two Independent VariableS (IVs)
Two-way = 2 IVs
Three-way = 3 IVs
Different participants in all conditions
Independent = “Different
Participants”
Several independent variables is
known as a factroial design
5
Two-Way ANOVA
“Two-Way” means groups are defined by 2
independent variables (IVs)
These IVs are typically called factors
With 2-Way ANOVA, there are
two main effects and 1 interaction, so there
are 3 F tests
All, some, or none may be significant
6
there are three types of ANOVA analysis available:
1) Single Factor ANOVA
2) Two-Factor ANOVA Without Replication
3) Two-Factor ANOVA with Replication
Each ANOVA test type is explained below:
7
Single Factor ANOVA
Single Factor ANOVA tests the effect of just
one factor.
Example: the teaching method, on the
measured outputs. The measured outputs
are the mean test scores for the groups that
had the different teaching methods applied
to them. The Null Hypothesis for this one
factor states that varying that factor has no
effect on the outcome.
8
Two-Factor ANOVA Without Replication
Two-Factor ANOVA Without Replication -
Allows testing of the original factor plus one
other factor. For example, in addition to testing
teaching methods, you could also test an
additional factor, such as whether differences in
teaching ability caused additional variation in
the outcome of test average scores. Each
factor has a Null Hypothesis which states that
varying that factor had no effect on the
outcome.
9
Two-Factor ANOVA With Replication
Two-Factor ANOVA With Replication allows for testing
both factors as above. This method also allows us to
test the effect of interaction between the factors upon
the measured outcome. The test is replicated in two
places. This allows for analysis of
whether the interaction between the two factors has an
effect on the measured outcome. The Null Hypothesis
for this interaction test states that varying the
interaction between the two factors has no effect on
the measured outcome. Each of the other two factors
being tested also has its own Null Hypothesis.
10
Logic of One Way ANOVA
One Way ANOVA
SStotal
SSbetween SSwithin
11
Logic of Two Way ANOVA
Two Way ANOVA
SStotal
SSbetween SSwithin
SSfactorA SSfactorB SSinteraction
Main Effect 1 Main Effect 2 Interaction
s 2factorA s 2factorB 2
sinteractio
F F F n
s 2
w s w2 sw2
12
Assumptions for the Two
Factor ANOVA
1. Observations within each sample are
independent
2. Populations are normally or
approximately normally distributed
3. Populations from which the samples
are selected must have equal
variances (homogeneity of variance)
4. The groups must have the same
sample size
13
Factors
The two independent variables in a two-way
ANOVA are called factors
The idea is that there are two variables,
factors, which affect the dependent variable
Each factor will have two or more levels
within it
The degrees of freedom for each factor is
one less than the number of levels
14
Hypotheses
There are three sets of hypothesis with the
two-way ANOVA
The null hypotheses for each of the sets are
given below
1. The population means of the first factor are
equal. This is like the one-way ANOVA for the
row factor
2. The population means of the second factor are
equal. This is like the one-way ANOVA for the
column factor
3. There is no interaction between the two factors
15
Treatment Groups
Treatment Groups are formed by making all
possible combinations of the two factors
All treatment groups must have the same
sample size for a two-way ANOVA
For example,
if the first factor has 3 levels and
the second factor has 2 levels,
Then, 3x2=6 different treatment groups
16
Main Effect
The main effect involves the independent
variables one at a time
The interaction is ignored for this part
Just the rows or just the columns are
used, not mixed
This is the part which is similar to the
one-way analysis of variance
17
Interaction Effect
The interaction effect is the effect that one
factor has on the other factor
The degrees of freedom here is the product
of the two degrees of freedom for each factor
18
Within Variation
The Within variation is the sum of squares
within each treatment group
The within variance = within variation
its degrees of freedom
The within group is also called the error
19
F-Tests
There is an F-test for each of the hypotheses
The F-test is the mean square for each main
effect and the interaction effect divided by the
within variance
F-test = mean square
within variance
The numerator df come from each effect, and the
denominator df is the df for the within variance
in each case
20
Advantages of two-way
ANOVA
1. More efficient to study two factors (A and B)
simultaneously, rather than separately
2. We can investigate interactions between factors
(can investigate complex associations)
3. In a two way anova (A×B design) there are four sources of
variations
1. Variation due to factorA
2. Variation due to factor B
3. Variation due to the interactive effect of A & B
4. Within cell (error) variation
21
BASIC TWO -WAY ANOVA TABLE
Source of SS df MS F Pvalue Fcrit
variation [SS/df]
Main effect
A
Main effect
B
Interactive
effect
Within
Total
22
TWO – WAY ANOVA WITHOUT
REPLICATION WITH
EXAMPLE
23
A Comparison of Dissolution of
Various Tablet Formulations:
Eight laboratories were requested to participate
in an experiment
Objective :
To compare the dissolution rates of two generic
products and a standard drug product
Purpose:
(a) To determine if the products had different
rates of dissolution, and
(b) To estimate the laboratory variability
(differences)
c) Interaction effect between lab & product
24
Tablet Dissolution After 30 Min for Three
Products (% Dissolution)
GENERIC
LABORATORY STANDARD ROW TOTAL
A B
1 89 83 94 266
2 93 75 78 246
3 87 75 89 251
4 80 76 85 241
5 80 77 84 241
6 87 73 84 244
7 82 80 75 237
8 68 77 75 220
25
CLICK ON DATA
26
CLICK ON DATA ANALYSIS
27
CLICK OK AT LAST
SELECT = ANOVA: TWO – FACTOR
WITHOUT REPLICATION
28
CLICK ON INPUT
RANGE SELECTION
29
CLICK ON THIS AT LAST
30
CLICK ON THIS TO SELECT OUTPUT RANGE
31
SELECT THIS B13 CELL TO
DEFINE RESULT AREA
THEN CLICK ON THIS
32
THEN CLICK ON THIS
33
YOUR RESULTS
34
CHECK THIS
CHANGE ALL
RELEVANT DATA
35
36
37
TWO – WAY ANOVA WITH
REPLICATION
38
39
Replicate tablet dissolution data for eight
laboratories testing three products (Percent
distribution)
If we take an example that we had test
content uniformity for Paracetamol and
Diclofenac from USA, INDIA and
AUSTRALIA.
We will take 6 samples from each contry
from 6 different places.
The table can be summarized as follows:
40
WEST CENTER EAST
GROUP
USA INDIA AUSTRALIA
PARACETAMOL 92 98 93
97 99 94
98 97 92
95 98 91
97 96 94
94 99 95
DICLOFENAC 98 95 92
99 99 96
95 97 93
93 98 91
98 98 95
97 99 93
41
ANOVA: TWO-FACTOR WITH REPLICATION
42
THEN CLICK OK
43
44
BENEFIT OF FACTORIAL DESIGN
We can look at how variables interact.
Interactions which…..
Show how the effects of one Independent
Variables (IVs) might depend on the
effects of another
Are often more interesting than main
effects
45
Thank you…!!!
