Factorial Experiments
Factorial Experiments
4. Conclusion – what has been learnt about the original conjecture from
the experiment. Often the experiment will lead to a revised conjecture,
and a new experiment, and so forth.
For example: if there are a levels of factor A and b levels of factor B, each
replicate contains all ab treatment combinations.
When factors are arranged in a factorial design, they are called as crossed.
Factorial Experiments
Basic Definitions
Factor B
Definition of a factor effect: The
Factor A Blow Bhigh change in the mean response when the
Alow 20 30 factor is changed from low to high.
Ahigh 40 52 Main Effect
The main effect of factor A is the difference
between the average response at the high
30 52
+ level of A and the average response at the
(High) low level of A:
40 52 20 30
Factor B
A yA yA 21
2 2
- The main effect of factor B is the difference
(Low) 20 40 between the average response at the high
level of B and the average response at the
- + low level of B:
(Low) (High)
30 52 20 40
Factor A B y B y B 11
2 2
Factorial Design
Basic Definitions
Factor B
Interaction between factors: When the
Factor A Blow Bhigh difference in response between the
Alow 20 40 levels of one factor is not the same at all
Ahigh 50 12 levels of the other factors.
A 12 40 28
Interaction Effect: It is the difference
- in the diagonal averages
(Low) 20 50
28 30
AB 29
- + 2
(Low) (High) OR
Factor A 12 20 40 50
AB 29
2 2
Factorial Experiments
When an interaction is large, the corresponding main effects have very little
practical meaning.
The main effect of factor A in the previous example can be calculated as:
52 12 20 40
A 32 30 2 ( very small )
2 2
Thus, the knowledge of the AB interaction is very useful than the main
effect.
40 52 20 30
A y A y A 21
2 2
30 52 20 40
B yB yB 11
2 2
52 20 30 40
AB 1
2 2
Factorial Experiments
Two-factor with interaction
50 12 20 40
A y A y A 1
2 2
40 12 20 50
B yB yB 9
2 2
12 20 40 50
AB 29
2 2
Factorial Experiments
Regression Model & The Associated Response Surface
y 0 1 x1 2 x2 12 x1 x2
The least squares fit is
yˆ 35.5 10.5 x1 5.5 x2 0.5 x1 x2 35.5 10.5 x1 5.5 x2
Factorial Experiments
Factor B
- A-B - A+B - -
(Low) (Low) A-B - A+B -
- + - +
(Low) (High) (Low) (High)
Factor A Factor A
Effect of Factor A A B A B If a factorial experiment is designed, an additional
treatment combination A+B+ is taken and the
Effect of Factor B A B A B number of observations needed will be reduced to
Since, experimental error is present, it is desirable four.
to take two observations at each treatment
combination and estimate the effects of the factors
using average responses. Thus a total of six
observations are required.
Factorial Experiments
Advantages of Factorial Experimentation
i 1, 2,..., a
yijk i j ( )ij ijk j 1, 2,..., b
k 1, 2,..., n
Other models (means model, regression models) can be useful
Extension of the ANOVA to Factorials
(Fixed Effects Case)
a b n a b
i 1
an . j. ...
( y y ) 2
j 1
a b a b n
n ( yij . yi.. y. j . y... ) ( yijk yij . ) 2
2
i 1 j 1 i 1 j 1 k 1
SST SS A SS B SS AB SS E
df breakdown:
abn 1 a 1 b 1 (a 1)(b 1) ab(n 1)
ANOVA Table – Fixed Effects Case
ANOVA Table – Fixed Effects Case
ANOVA Table – Fixed Effects Case
ANOVA Table – Fixed Effects Case
Example 5.1: The Battery Life Experiment
Residual Analysis – Example 5.1
Residual Analysis – Example 5.1
Factorials with More Than Two Factors
SST SS A SS B SS AB SS AC
SS ABC SS AB K SS E