Design of Experiments

Compiled by
Dr. Deepak Lawrence.K
MED, NIT Calicut
 How to solve problems?

Theory Experience

E=mc2 • Problem with sheer guesswork in

industrial practice

Experimentation

Experiments are series if tests that are performed in all engineering and scientific
disciplines and are important parts of the way we learn about how systems and
processes work
 designed experiment
A designed experiment is an approach to
systematically varying the controllable input
factors in the process and determining the effect
these factors have on the output product
parameters

14-04-2024
3
 Strategy of experimentation
• Extensively used strategy in practice is the one-factor-at-a-
time (OFAT) approach.

• Successively varying each factor over its range with the

other factors held constant

Advantages:
• Simple
• Easy graphical displays

Disadvantages:
• Time consuming
• Extra work
• Shortcut - Less than optimal solution
Strategy of experimentation

➢ Keeping time constant

Strategy of experimentation

➢ Keeping temperature constant

➢ Varying time and temp
Strategy of experimentation to get
optimal solution
 Strategy of experimentation

Usually:
• Response is smooth (no sharp kinks or inflection points)
• Effect of one factor depends on level of one or more of other factors

OFAT approach fails to consider any possible interaction between the

factors
• Interaction is the failure of one factor to produce the same effect on the
response at different levels of another factor.
Analysis of historical data

•Wealth of data already available about the

process of interest (computer databases)
•Certainly much more data available from the
logs than could ever be collected in an
experimental program

•Why experiment?
Correlation ≠ Causation

• A significant correlation between the

response and an independent variable
does not prove causation
Correlated Factors
• Overtime, variation in the factors of
interest, often they vary together
• Impossible to sort out which of the
factors, if any, is having an impact on the
response

Shelf life

Humidity, or
Temperature?
Correlated Factors

Why tasty?
Sugar or Syrup?
No Randomisation = Bias

• No randomization in the variations of the

factors with time, biases to influence the
results
One month old bike
New petrol type
Improved mileage
Effect could be due
to
 Tight Control of factors
• Independent variable when tightly controlled, there will not be enough
variation in that factor to get a measurable effect on the response.

Will the dog bite?

Variation insufficient for effect on
response
Good Design Requirements

• Defined objectives
• Unobscured Effects
- Balanced
• Free of Bias
• Blocking
• Randomization
• Replication
• Variability Estimable
• Better Precision
 Good Techniques of experiments
Blocking
• Keeping noise factors similar
• Balance the effect of noise variables

Replication
•Repetition resetting all
factors
•Improves precision

Randomization
•Carry out experiments in random order
•Eliminate bias due to lurking factors
 Experimental design
• Studying the effects of many factors simultaneously in an
efficient way, is the purpose of experimental design
techniques

• Correct approach to dealing with several factors is to

conduct a factorial experiment

•factors are varied together

•Involves Statistical Design of experiments (DOE)
•Computer software programs are widely used to assist
experimenters in selecting and constructing
experimental designs.
•Gives meaningful conclusions from the data.
•Obtain superior results with fewer experiments
 Experimental design

Factors Response
Process
 Experimental design
Series of tests are conducted by making purposeful
changes to the input variables so that we may
observe some concluded results in the output
variable.
• For determining which variable has the significant
influence on the output.
• For determining optimum combination of process
parameters.

✓ Design of experiments is useful for establishing a

statistical control of the process
✓ Can be used to determine influential process
variables
✓ Improve manufacturing process.
 Experimental design

➢ Steps:
1. Problem statement
2. Selection of response variable
3. Choice of the factors and levels
4. Analyze the measurement system
5. Experimental design
6. Performing the experiment
7. Data analysis
8. Confirmation run and recommendations
 Terms and concepts used

1. Factor:
Independent variable that may affect the response, also called
input variables

2. Level:
Setting or adjustment of a factor

3. Response variable :
Variable that shows the observed result or the output. Also called dependent
variable
4. Background Variable
a variable of which the experimenter is unaware or cannot control, and which
could have an effect on the outcome of an experiment
 Terms and concepts used

4. Experimental Design (also called Experimental Pattern)

Specific setting or a combination of factor levels for an experiment

5. Experimental run:
Single performance of an experiment for a specific set of treatment conditions

6. Effect :
Relationship between a factor and response variable
What are the factors?
• Independent variables
• Can be:
• Numeric factors
• Categoric factors
• Numeric better than
categoric
 • Stage by Stage

Choose experimental design

Rough Final
Screening
optimization optimization
First stage: Screening Design

• To identify the important factors, when >3

B • Detailed experiment on them later
• Effort not wasted
• Factorial Designs
●
• Each factor at two levels
● C • Interactions not considered
• Detailed study only on important factors
● ●
A • Resources not wasted
Second stage: Rough Optimization

• Only significant factors considered

B • Other factors at their easier or cheaper
levels
● ●
• Factorial Designs
● ●
• Each factor at 3 levels (if possible)
●
C • Center points to test curvature
● ● • Interactions identified
● ●
A • Curvature identified
Third stage: Final Optimization
• Central Composite Designs
• Each factor at 5 levels
B • Second degree models can be fitted
● ●
●
● ●

●
●
●
C Do The Experiment!
● ●

● ●
● A • In a random run order
• Watchout and record the results
 Factorial Experiment

A factorial experiment is a powerful technique, that

the experimenter focus on experiments that include
two or more factors
Generally, in a factorial experimental design,
experimental trials (or runs) are performed at all
combinations of factor levels
Thus, if there are two factors A and B with a levels of
factor A and b levels of factor B, each replicate
contains all ab treatment combinations.
When several factors are of interest in an experiment,
a factorial experiment should be used.

14-04-2024
28
 Example : Factorial Experiment example
For example, if a chemical engineer is interested in
investigating the effects of reaction time and
reaction temperature on the yield of a process, and
if two levels of time (1.0 and 1.5 hours) and two
levels of temperature (125°F and 150°F) are
considered important, a factorial experiment would
consist of making experimental runs at each of the
four possible combinations of these levels of
reaction time and reaction temperature.

14-04-2024
29
 Factorial designs

Production process inputs and outputs.

14-04-2024
Factorial designs for the process 30
 Advantages of Factorial Design
They are more efficient than one-factor-at-a-time
experiments.
Furthermore, a factorial design is necessary when
interactions may be present to avoid misleading
conclusions.
Finally, factorial designs allow the effects of a factor
to be estimated at several levels of the other factors,
yielding conclusions that are valid over a range of
experimental conditions

14-04-2024
31
 The 2k Factorial Design
They are factorial design with k factors, each at two levels. Because each
complete replicate of the design has 2k runs, the arrangement is called a
2k factorial design
The simplest type of design 2k is the 22—that is, two factors A and B,
each at two levels
The levels as the “low” or “-” and “high” or “+” levels of the factor.
the design can be represented geometrically as a square with the 22 = 4
runs forming the corners of the square and the four runs in a tabular
format often called the test matrix or the design matrix.

32
14-04-2024
 The 2k Factorial Design
In general, a run is represented by a series of lowercase
letters.
If a letter is present, then the corresponding factor is set at the
high level in that run; if it is absent, the factor is run at its low
level
For example, run a indicates that factor A is at the high level
and factor B is at the low level.
The run with both factors at the low level is represented by (1).
For example, the run in a 24 with A and C at the high level and
B and D at the low level is
denoted by ac
The effects of interest in the 22 design are the main effects A
and B and the two-factor interaction AB. Let the letters (1), a,
b, and ab also represent the totals of all n observations taken
at these design points
14-04-2024
33
 Factorial Experiments
Main effect is the change in response
produced by a change in the level of a
primary factor.
An interaction is present among factors
if a change in the levels of one factor
influences the effect of another factor.
Consider an experiment with two factors
A and B
Interested in
Main effect of A
Main effect of B
Interaction effect of AB
 main effect of A

To estimate the main effect of A, we would average the

observations on the right side of the square when A is at
the high level and subtract from this the average of the
observations on the left side of the square where A is at the
low level

n is number of replications
14-04-2024
35
 main effect of B
the main effect of B is found by averaging the
observations on the top of the square where B is
at the high level and subtracting the average of
the observations on the bottom of the square
where B is at the low level

14-04-2024
36
 AB interaction

the AB interaction is estimated by taking the difference in

the diagonal averages

14-04-2024
37
 Main Effect
The effect of a factor is defined to be the change in response
produced by a change in the level of the factor.
This is frequently called a main effect because it refers to the
primary factors of interest in the experiment.
For example, consider the simple experiment in Figure
(below). This is a two-factor factorial experiment with both
design factors at two levels.
30 52
+
(High)
Factor B

-
(Low) 20 40

- +
(Low) Factor A (High)
14-04-2024
38
 Main Effect Without Interaction
The main effect of factor A in this two-level design can be
thought of as the difference between the average response at
the low level of A and the average response at the high level
of A.

30 52
+
(High)
Factor B

-
(Low) 20 40

- +
(Low) Factor A (High)
14-04-2024
39
 Main Effect Calculation

40 + 52 20 + 30
A= - = 21
2 2 30 52
+
(High)
Increasing factor A from the low

Factor B
level to the high level causes an
average response increase of
21 units. -
(Low) 20 40
Similarly, the main effect of B is +
-
(Low) Factor A (High)

30 + 52 20 + 40
B= - = 11
2 2
14-04-2024
40
 Main Effect With Interaction
In some experiments, we may find that the difference
in response between the levels of one factor is not
the same at all levels of the other factors.
When this occurs, there is an interaction between
the factors. For example, consider the two-factor
factorial experiment shown in Figure below
40 12
+
(High)
Factor B

-
(Low) 20 50

- +
14-04-2024 (Low) Factor A (High)
41
 Main Effect With Interaction
At the low level of factor B, the A effect is
A = 50 - 20 = 30
At the high level of factor B, the A effect is
A = 12 - 40 = -28 40 12
Because the effect of A depends +
(High)
on the level chosen for factor B,

Factor B
we see that there is interaction
between A and B.
The magnitude of the interaction -
(Low) 20 50
effect is the average difference
in these two A effects, - +
(Low) Factor A (High)
 -28 - 30 
AB =   = -29
 2  Clearly, the interaction is large
14-04-2024 in this experiment. 42
 A Factorial experiment without interaction
A Factorial experiment without interaction
60
52
50

40
40
Response

30
30

20
20

10

0
Low (-) High (+)
Series 1 20 30
Series 2 40 52
Factor A

Series 1 Series 2

14-04-2024
43
 A Factorial experiment with interaction
A Factorial experiment with interaction
50

40 40
40

30
Response

20
20

12

10

0
Low (-) High (+)
Series 1 20 40
Series 2 40 12
Factor A

Series 1 Series 2

14-04-2024
44
 22 Design
The quantities in brackets in equations are called contrasts
In these equations, the contrast coefficients are always either + 1 or - 1.
The column headings for the table are the main effects A and B, the AB interaction,
and
I, which represents the total.
The row headings are the runs. Note that the signs in the AB column are the
product of signs from columns A and B.
To generate a contrast from this table, multiply the signs in the appropriate column
of Table by the runs listed in the rows and add

14-04-2024
45
 22 Design
To obtain the sums of squares for A, B, and AB, we use the following
result.
the sums of squares for A, B, and AB are
The analysis of variance is completed by computing the total sum of
squares SST (with 4n - 1degrees of freedom) as usual, and obtaining the
error sum of squares SSE [with 4(n - 1) degrees of freedom] by subtraction

14-04-2024
46
 Factorial Experiments

Statistical Analysis
Completely randomized design with two factors
(A and B) and n replicates.
The model is
y ijk =  +  i +  j + (  ) ij +  ijk
where  = overall mean
i = effect of ith level of factor A
j = effect of jth level of factor B
()ij = effect of the interaction between A
and B.
 = random error component
 Factorial Experiments

Statistical Analysis
Total corrected sum of squares
decomposition, notation:
SST = SSA + SSB + SSAB + SSE
The corresponding degree of freedom
decomposition is
abn – 1 = (a – 1) + (b – 1) + (a – 1)(b – 1) + ab(n – 1)
 Formula for Main effect A
Now

The effect of A at the low level of B is ,  a - (1) 

n
The effect of A at the high level of B is  ab - b 
.
n
1
 ab - b +  a - (1)
Averaging these two
quantities yields the A = 2n
main effect of A:
1
A=  ab + a - b - (1)
2n
14-04-2024
49
 Formula for Main effect B
Now

The effect of B at the low level of A is , b - (1) 

n
The effect of B at the high level of A is  ab - a 
.
n
Averaging these two
quantities yields the
B=
1
2n
 ab - a  + b - (1)
main effect of B:
1
B=  ab + b - a - (1)
2n
14-04-2024
50
 Formula for Interaction effect AB
The interaction effect of AB as the average difference
between the effect of A at the high level of B and the
effect of A at the low level of B, Thus

AB =  ab - b  -  a - (1) 
1
2n
1
AB =  ab + (1) - a - b 
2n

Alternatively, we may define AB as the average difference

between the effect of B at the high level of A and the effect of
B at the low level of A. This will also lead to Equation as
above
14-04-2024
51
 Formula for Sum of Squares
The sum of squares for any contrast can be computed,
which states that the sum of squares for any contrast is
equal to the contrast squared divided by the number of
observations in each total in the contrast times the sum
of the squares of the contrast coefficients.
Consequently, we have
 a + ab - b - (1) 
2

SS A =
n 2k

 b + ab - a - (1) 
2

SS B =
n 2k

 ab + (1) - a - b 
2

SS AB =
14-04-2024 n 2k 52
 Formula for Sum of Squares
The total sum of squares is found in the usual way,
that is,
2 2 n
y...2
SST =    y 2
ijk -
i =1 j =1 k =1 4n
In general, SST has 4n - 1 degrees of freedom. The
error sum of squares, with 4(n - 1) degrees of
freedom, is usually computed by subtraction as

SS E = SST - SS A - SS B - SS AB

14-04-2024
53
 ANOVA

• ANOVA is a powerful statistical tool that can be

used to compare means of two or more
populations.
• It is used to basically compare the test hypothesis
about the differences between two or more means.
• Ho : Null hypothesis – All means are same
• H1 : Any of the means is different
• When populations are normally distributed, we
perform F test of equality of variances to conclude
whether the hypothesis is true or false.
• The calculated F value is compared with F-critical
(Fα) from the table of F distribution (of corresponding
confidence level)
• Analysis of variance (ANOVA) can be extended to handle the two factor factorial
experiment
Error sum of squares, SSE = SST – (SSA + SSB + SSC + SSAB + SSBC + SSAC + SSABC )

𝑀𝑆
F0 =
𝑀𝑆𝐸

The F-test (F value) compares the variance in

each group mean from the overall group
variance.
If the variance within groups is smaller than
the variance between groups, the F-test will
find a higher F-value
Factor main effect or Interaction Effect is Significant if F0 > Fα (F-critical)

✓ Corresponding Fα is obtained from the standard F

table
 The ANOVA Table for a Two-Factor Factorial, Fixed Effects Model

ANOVA is usually performed using software such as Design

expert or Minitab
ANOVA computation and analyses is NOT expected to learn for
exam point of view
14-04-2024
56
 Problem - 1

14-04-2024
57
 Problem - 1
Consider an investigation into the effect of the concentration of the reactant and the
amount of the catalyst on the conversion (yield) in a chemical process. The objective
of the experiment was to determine if adjustments to either of these two factors would
increase the yield. Let the reactant concentration be factor A and let the two levels of
interest be 15 and 25 percent. The catalyst is factor B, with the high level denoting the
use of 2 pounds of the catalyst and the low level denoting the use of only 1 pound.
The experiment is replicated three times, so there are 12 runs. The order in which the
runs are made is random, so this is a completely randomized experiment. The data
obtained are as follows:

Factors Replications
Treatment
A B I II III Total
Combination
-1 -1 A low, B low 28 25 27 80
1 -1 A high, B low 36 32 32 100
-1 1 A low, B high 18 19 23 60
1 1 A high, B high 31 30 29 90

14-04-2024
58
 Solution:
❑ The four treatment combinations in this design are shown
graphically in Figure (next slide).
❑ By convention, we denote the effect of a factor by a
capital Latin letter.
❑ Thus “A” refers to the effect of factor A, “B” refers to the
effect of factor B, and “AB” refers to the AB interaction.
❑ In the 22 design, the low and high levels of A and B are
denoted by “-” and “+” respectively, on the A and B axes.
❑ Thus, on the A axis represents the low level of
concentration (15%), whereas represents the high level
(25%), and on the B axis represents the low level of
catalyst, and denotes the high level.

14-04-2024
59
 Solution:
b = 60 ab = 90
High +
(2 pounds)

catalyst, B
Amount of

Low -
(1 pound) (1) = 80 a = 100
- +
Low High
Reactant
(15%) (25%)
concentration, A

❑ a represents the treatment combination of A at the high level and B at the low
level,
❑ b represents A at the low level and B at the high level, and
❑ ab represents both factors at the high level.
❑ By convention, (1) is used to denote both factors at the low level.
❑ This notation is used throughout the 2k series.
14-04-2024
60
 Solution:

In a two-level factorial design, we may define the

average effect of a factor as the change in response
produced by a change in the level of that factor
averaged over the levels of the other factor.

Also, the symbols (1), a, b, and ab now represent the

total of the response observation at all n replicates
taken at the treatment combination

14-04-2024
61
 Solution: Estimation of the average effects
1
A=  ab + a - b - (1)
2n
1
A= ( 90 + 100 - 60 - 80 ) = 8.33
2(3)
1
B=  ab + b - a - (1)
2n
1
B= ( 90 + 60 - 100 - 80 ) = -5.00
2(3)
1
AB =  ab + (1) - a - b 
2n
1
AB = ( 90 + 80 - 100 - 60 ) = 1.67
14-04-2024
2(3)
62
 Solution:

The effect of A (reactant concentration) is positive;

this suggests that increasing A from the low level
(15%) to the high level (25%) will increase the yield.

The effect of B (catalyst) is negative; this suggests

that increasing the amount of catalyst added to the
process will decrease the yield.

The interaction effect appears to be small relative to

the two main effects.
14-04-2024
63
 Formula for Contrast
In experiments involving 2k designs, it is always important to
examine the magnitude and direction of the factor effects to
determine which variables are likely to be important. The
analysis of variance can generally be used to confirm this
interpretation (t-tests could be used too).
Effect magnitude and direction should always be considered
along with the ANOVA, because the ANOVA alone does not
convey this information.
Consider determining the sums of squares for A, B, and AB.
Note that a contrast is used in estimating A,
Contrast A = ab + a - b - (1)
We usually call this contrast the total effect of A.
14-04-2024
64
 Solution: Estimation of the Sum of Squares
 a + ab - b - (1) 
2

SS A =
n 2k

(100 + 90 - 60 - 80 ) ( 50 )
2 2

SS A = 2
SS A = = 208.33
3(2) 12
 b + ab - a - (1) 
2

SS B =
n 2k
( 60 + 90 - 100 - 80 ) ( -30 )
2 2

SS B = 2
SS B = = 75.00
3(2) 12

 ab + (1) - a - b 
2

SS AB =
n 2k
( 90 + 80 - 100 - 60 ) (10 )
2 2

SS AB = SS AB = = 8.33
3(2) 2 12
14-04-2024
65
 Solution: Estimation of SST and SSE
For the experiment, we obtain
2 2 n
y...2
SST =    yijk
2
-
i =1 j =1 k =1 4n
2 2 3
y...2
SST =    y 2
ijk -
i =1 j =1 k =1 4(3)
SST = 9398.00 - 9075.00 SST = 323.00
SS E = SST - SS A - SS B - SS AB
SS E = 323.00 - 208.33 - 75.00 - 8.33
SS E = 31.34
14-04-2024
66
Analyze the results using ANOVA
• Usually we use statistical
software
• Check all diagnostic graphs

Confirmation run
• Run under the optimum levels
• Result achieved?
• Celebrate!
• If NOT, need to go back 

ANOVA is usually performed using software such as Design

expert or Minitab
ANOVA computation and analyses is NOT expected to learn for
exam point of view
 ANNOVA Table
using SSA, SSB, and SSAB . The complete ANOVA is summarized in
Table below.
On the basis of the p-values (less than 0.05 means statistically
significant), we conclude that the main effects are statistically
significant and that there is no interaction between these factors.
This confirms our initial interpretation of the data based on the
magnitudes of the factor effects

Source of Degree of
Sum of Squares Mean Square F-value p-value
Variation Freedom

A 208.33 1 208.33 53.18 8.45E-05

B 75.00 1 75.00 19.14 0.002363
AB 8.33 1 8.33 2.13 0.182818
Error 31.34 8 3.92
Total 323.00 11 29.36

14-04-2024
68
 Problem 2

14-04-2024
69
 Solution : Factor effect computation

14-04-2024
70
 ANOVA of results

14-04-2024
71
Experimental design example 23 design

Feed rate (A) Depth of cut (B) Tool angle (C)

High (+1) 100 0.1 20
Low (-1) 50 0.05 10

x1, x2, x3 High

Feed rate (A)

Depth of cut (B)

n= 23 Surface finish
y = f(x)
Tool angle (C)
Low
with n = 2
replicates
2k Factorial Design

A 23 factorial design using coded levels Feed Depth of Tool

rate (A) cut (B) angle (C)
Std order A B C High (+1) 100 0.1 20
1 -1 -1 -1 Low (-1) 50 0.05 10
2 +1 -1 -1
3 -1 +1 -1 +1 : High level
4 +1 +1 -1 -1 : Low level
5 -1 -1 +1
6 +1 -1 +1
7 -1 +1 +1
8 +1 +1 +1
 The 2k Design for k = 3 Factors

Three factors A, B and C, each at two levels;

“low” or “−” and “high” or “+”.
This is a 23 factorial design, and it has eight
factor-level combinations
The design can be represented geometrically as;

• A run is represented by a series of lowercase (1)

letters a
• A letter is present, then the corresponding b
factor is set at the high level in that run; if it is ab
absent, the factor is run at its low level c
• The eight runs forming the corners of the ac
cube. bc
abc
 The design allows the estimation of
• Three main effects (A, B, and C)
• Three two-factor interactions (AB, AC, and BC)
and
• One three-factor interaction (ABC).
Thus, the full factorial model could be written
symbolically as:

where μ is an overall mean, ɛ is a random error term, and the uppercase

letters represent the main effects and interactions of the factors
 Table of plus and minus signs for the 23 design

(1)
a
b
ab
c
ac
bc
abc

Effect of factor is the difference between the average response at the high level of A and the average response
at the low level of A.

Effect of A = (Mean of the response variable where A is at the high level – Mean
of the response variable where A is at the low level)
• Effect of C and D:
n = No of replications

• Two-factor interaction:

• Three factor interaction effect ABC :

 ANOVA

• ANOVA is a powerful statistical tool that can be

used to compare means of two or more
populations.
• It is used to basically compare the test
hypothesis about the differences between two or
more means.
• Ho : Null hypothesis – All means are same
• H1 : Any of the means is different
• When populations are normally distributed, we
perform F test of equality of variances to
conclude whether the hypothesis is true or false.
• The calculated F value is compared with F-critical
(Fα) from the table of F distribution (of corresponding
confidence level)
• Analysis of variance (ANOVA) can be extended to handle the two factor factorial
experiment
 The quantities in brackets in equations are
called “contrasts”
Sum of squares ( gives the total variability in the data) is found using equation:

n = No of replications
(𝑫𝒐𝒇
)

Error df = Number of runs in an experiment * (number of replicates – 1)

𝑆𝑆
Mean of MS =
squares; 𝐷𝑜𝑓

Total sum of
squares:
Error sum of squares, SSE = SST – (SSA + SSB + SSC + SSAB + SSBC + SSAC + SSABC )

𝑀𝑆
F0 =
𝑀𝑆𝐸

The F-test (F value) compares the variance in

each group mean from the overall group
variance.
If the variance within groups is smaller than
the variance between groups, the F-test will
find a higher F-value
Factor main effect or Interaction Effect is Significant if F0 > Fα (F-critical)

✓ Corresponding Fα is obtained from the standard F

table
 • The main effect of A:

• The sum of squares for A:

Analysis of Variance for the Surface-Finish Experiment

Analyze the results using ANOVA
• Usually we use statistical
software
• Check all diagnostic graphs

Confirmation run
• Run under the optimum levels
• Result achieved?
• Celebrate!
• If NOT, need to go back 
 Inferences:
• From examining the magnitude of the effects, feed rate (factor A) is clearly
dominant, followed by depth of cut (B) and the AB interaction, although the
interaction effect is relatively small.

• Since small values of the surface finish response are desirable, this would suggest that both A
(feed rate) and B (depth of cut) should be run at the low level

Interaction effects: Extend of non parallelism shows

Main effects:
the extend of interaction between factors
 Randomization

Randomizing allows us to disperse the

experimental error across all the treatments
equally.
 Blocking

• Blocking is the arranging of

experimental units that are
similar to one another in
groups (blocks)

• These variables are chosen

carefully to minimize the
impact of their variability on
the observed outcomes.

Male and female: An experiment is designed to test a new drug for weight
loss on patients. There are two levels of the treatment, drug, and placebo,
administered to male and female patients in a double blind trial.
The sex of the patient is a blocking factor (a source of variability that is not
of primary interest to the experimenter) accounting for treatment variability
between males and females. This reduces sources of variability and thus
leads to greater precision.
14-04-2024
86
factorial design with two replications in two blocks in random order

Feed rate Depth of cut Tool angle

(A) (B) (C)

High (+1) 100 0.1 20

Low (-1) 50 0.05 10

(1)
a
b
ab
c
ac
bc
abc

test matrix in the “standard order”.

 Standard Depth of cut Surface roughness
Run order Block Feed (A) Tool angle (C)
order (B) (Ra)

1 1 7 50 0.1 20 10
2 1 5 50 0.05 20 11
3 1 2 100 0.05 10 10
4 1 1 50 0.05 10 9
5 1 4 100 0.1 10 12
6 1 8 100 0.1 20 16
7 1 3 50 0.1 10 9
8 1 6 100 0.05 20 10

9 2 10 100 0.05 10 12
10 2 13 50 0.05 20 10
11 2 12 100 0.1 10 15
12 2 15 50 0.1 20 8
13 2 9 50 0.05 10 7
14 2 14 100 0.05 20 13
15 2 16 100 0.1 20 14
16 2 11 50 0.1 10 11
 Exercise:
It is required to compare the petrol consumption of two
cars, one by Toyota, the other by Honda. The traffic can
be low or high. The average speed can be 30 or 50 km/hr.
Design a factorial design based experimental plan with
two replications in two blocks in random order.
Text books
 Thank you !

14-04-2024
91