DESIGN OF EXPERIMENTS

Introduction
The subject-matter of the design of
experiment includes:
1. Planning of the experiment,
2. Obtaining relevant information
from it regarding the statistical
hypothesis under study, and
3. Making a statistical analysis of the
data.
Definition
Design of Experiment may be defined
as “the logical construction of the
experiment in which the degree of
uncertainty with which the inference is
drawn may be well defined.”
Terminology in Experimental Designs
Experiments
Treatments
Experimental materials
Experimental Unit
Blocks
Yield
Experimental Error
Replication
Precision
Efficiency of a design
 Experiments
• An experiment is a device or a means of
getting an answer to the problem under
consideration. Experiment can be classified
into two categories as follows:
a) Absolute and
b) Comparative.
 Absolute Experiments
Absolute experiments consists in determining
the absolute value of some characteristics like
(i) obtaining the average intelligence quotient of
a group of people,
(ii) finding the correlations coefficient between
two variables in a bi-variate distribution, etc.
 Comparative experiments
Comparative experiments are designed to
compare the effect of two or more objects
on some population characteristic,
e.g;. Comparison of different manures or
fertilizers, different kinds of varieties of a
crop, different cultivation processes,
different pieces of land in a field
experiment, or different diets or medicines
in a dietary or medical experiment
respectively, etc.
 Treatments
Various object of comparison in a
comparative experiment, are termed
as treatments,
e.g., in field experimentation different
fertilizers or different varieties of crop
or different methods of cultivation, In
medical experiments different doses of
a medicine or diets are the treatments.
For example:
1. suppose an Agronomist wishes to know the effect
of different spacings on the yield of a crop, different
spacings will be treatments. Each spacing will be
called a treatment.
2. If different of fertilizer are tried in an experiment to
test the responses of a crop to the fertilizer doses,
the different doses will be treatments and each
dose will be a treatment.
3. A teacher practices different teaching methods on
different groups in his class to see which yields the
best results.
4. A doctor treats a patient with a skin condition with
different creams to see which is most effective.
 Experimental materials
In crop experiments a piece of land may be
used for treatment trail, feeds under
comparison in a dairy, poultry or fisheries
experiment may be administrated to a group of
dairy animals, a group of poultry birds or fishes
in a pond. A plot of land, a group of dairy
animal, a group of poultry birds etc. on which
treatments are to be applied are called
experimental material.
 Experimental Unit
The smallest division of the experimental
material to which we apply the treatments and
on which we make observations on the variable
under study, is termed as experimental unit,
e.g., in field experiments the plot of “land” is the
experimental unit. In other experiments unit
may be a patient in a hospital, a lump of dough,
a group of pigs in a pen or a batch of seeds.
1. For example i) In laboratory insects may be
kept in groups of five or six. To each group,
different insecticides will be applied to know
the efficacy of the insecticides. In this study
different groups of insects will be the
experimental unit.
2. If treatments are different varieties, then the
objects to which treatments are applied to
make observations will be different plot of
land. The plots will be called experimental
units.
 Blocks
In agricultural experiments most of the times we
divide the whole experimental unit into
relatively homogeneous sub-groups or strata.
These strata, which are more uniform amongst
themselves than the field as a whole, are known
as blocks.
A number of homogeneous plots in a strip
constitute a block in an agricultural experiment
where as the patients of same symptoms having
same age group, same sex etc. may constitute a
block in medical experiments.
 Yield
The measurements of the variable under study
on different experimental units.
(i.e., plots, in field experiments) are termed as
yields.
 Experimental Error
Let us suppose that a large homogeneous field in divided
into different plots and different treatments are applied
to this plots. If the yields from some of the tremens are
more than those of the others, experimenter is faced
with the problem of deciding if the observed differences
are really due to treatment effect or they are due to
chance factors.
In field experimentation, it is a common experience that
the fertility gradient of the soil does not follow any
systematic pattern but behaves in an erratic fashion.
Experience teals us the even if the same treatment is
used on all the plots, the yields would still vary due to the
differences in soil fertility. Such variation from plot to
plot, which is due to random factors beyond human
control, is spoken of as experimental error.
It may be pointed out that the term ‘error’ used
here is not synonymous with ‘mistake’ but is a
technical term which includes all types of
extraneous variation due to:
1. the inherent variability in the experimental
material to which treatments are applied,
2. the lack of uniformity in the methodology of
conducting the experiments or in other words
failure to standardize the experimental
technique and
3. lack of representative ness of the sample to the
population under study.
 Replication

Replication means the execution of an

experiment more than once. In other words, the
repeated application of treatment under
investigation is known as replication.
 Precision
The reciprocal of the variance of the mean is
termed as the precision, or the amount of
information of a design. Thus for an experiment
replicated r times the precision id given by

Where σ2 is the error variance per unit.

 Efficiency of a design
Efficiency of a design in comparison with
another may be defined as the ratio of the
precision of the design to that of the other
design with which the comparison is made.
Example: Let us consider a design D1 having error variance σ12 and replication r1 . The efficiency of
this design D1 in comparison with other design D2 involving r2 replications. and having error
variance σ22 is given by
2 2

E = r1  12 = r1 22
r 2  2 r1 1
i. If E = 1, efficiency of D1 and D2 are equal.
ii. If E > 1, the design D1 is more efficient and
iii. If E < 1, the design D1 is less efficient in comparison D2.

From the above definition it is obvious that efficiency of a design can be increased by increasing the
number of replication and thus decreasing the error variance per unit.
Basic principles of Experimental design:
The purpose of designing an experiment is to
increase the precision of the experiment. In
order to increase the precision, we try to
reduce the experimental error. For reducing the
experimental error, we adopt some techniques.
These techniques form the basic principles
of experimental designs.
According to Professor R A Fishers, the basic
principles of experimental design are-
1. Replication
2. Randomization and
3. Local control.
 Replication
Definition: Repetition of treatment to different
experimental units is known as Replication.
In other words, the repetition of treatments
under investigation is known as replication.
We have no means of knowing about the
variations in the results of a treatment. Only
when we repeat the treatment several times we
can estimate the experimental error.
 Use of Replications
A replication is used
1. to secure more accurate estimate of the experimental
error, a term which represents the differences that would
be observed if the same treatments were applied several
times to the same experimental units;
2. to reduce the experimental error and thereby to increase
precision, which is a measure of the variability of the
experimental error.
𝜎
The standard error of treatment mean is
𝑟
Where 𝜎 is S.D. of treatment in the population and `r‟ is the
number of replications. As `r‟ increases, the standard error of
mean decreases. Also in the analysis of variance the
replication of treatments provides estimate of experimental
error which is essential for the application of F-test.
 Randomization
When all the treatments have equal chances of being
allocated to different experimental units it is known as
randomization or
Random allocation of treatments to different
experimental units known as randomization.
The purpose of randomization is to remove bias and
other sources of extraneous variation
which are not controllable. Another advantage of
randomization (accompanied by replication) is that it
forms the basis of any valid statistical test. Hence the
treatments must be assigned at random to the
experimental units. Randomization is usually done by
using tables of random numbers.
 Local control
It has been observed that all extraneous sources
of variation are not removed by randomization
and replication. This necessitates a refinement
in the experimental technique. For this purpose,
we make use of local control, a term referring to
the grouping of homogeneous experimental
units.
The main purpose of the principle of local
control is to increase the efficiency of an
experimental design by decreasing the
experimental error.
 Types of experimental Design
Completely randomized design (CRD)
Randomized Block Design
Latin Square Design
 Completely randomized design (CRD)
• The CRD is the simplest design. Suppose there are v treatments to be
compared.
• All experimental units are considered the same and no division or
grouping among them exist.
• In CRD, the v treatments are allocated randomly to the whole set of
experimental units, without
• making any effort to group the experimental units in any way for more
homogeneity.
• Design is entirely flexible in the sense that any number of treatments or
replications may be used.
• Number of replications for different treatments need not be equal and
may vary from treatment to
• treatment depending on the knowledge (if any) on the variability of the
observations on individual
• treatments as well as on the accuracy required for the estimate of
individual treatment effect.
 Example
• Suppose there are 4 treatments and 20 experimental
units, then
• - the treatment 1 is replicated, say 3 times and is
given to 3 experimental units,
• - the treatment 2 is replicated, say 5 times and is
given to 5 experimental units,
• - the treatment 3 is replicated, say 6 times and is
given to 6 experimental units
• and
• - finally, the treatment 4 is replicated [20-(6+5+3)=]6
times and is given to the remaining 6
• experimental units
• All the variability among the experimental
units goes into experimented error.
• CRD is used when the experimental material is
homogeneous.
• CRD is often inefficient.
• CRD is more useful when the experiments are
conducted inside the lab.
• CRD is well suited for the small number of
treatments and for the homogeneous
experimental
• material.
 Layout of CRD
Following steps are needed to design a CRD:
1. Divide the entire experimental material or
area into a number of experimental units, say
n.
2. Fix the number of replications for different
treatments in advance (for given total number
of available experimental units).
3. No local control measure is provided as such
except that the error variance can be reduced
by choosing a homogeneous set of
experimental units.
 Layout of CRD with equal no. of replication for each treatment
Replicates
Treatme
nts 1 2 j q Treat
total
1 𝑌11 𝑌12 𝑌1𝑗 𝑌1𝑞 𝑌1 .

2 𝑌21 𝑌22 𝑌2𝑗 𝑌2𝑞 𝑌2 .

i 𝑌𝑖1 𝑌𝑖2 𝑌𝑖𝑗 𝑌𝑖𝑞 𝑌𝑖 .

p 𝑌𝑝1 𝑌𝑝2 𝑌𝑝𝑗 𝑌𝑝𝑞 𝑌𝑝 .

Rep.
𝑌. 1 𝑌. 2 𝑌. 𝑗 𝑌. 𝑞 𝑌. .
total
31
 Linear Model of CRD
The linear model for CRD is
𝑌𝑖𝑗 = 𝜇 + 𝛼𝑖 + 𝜀𝑖𝑗 ; i=1,2,...,p and j=1,2,...,q
where 𝑌𝑖𝑗 is the yield of the ith treatment in the jth array,
𝜇 is the general mean,
𝛼𝑖 is the effect of the ith treatment, and
𝜀𝑖𝑗 is random error due to jth replication of the ith treatment
Assumptions
(i) 𝜇 and 𝛼𝑖 are unknown parameters
(ii) 𝜀𝑖𝑗 ~NID(0, σ2 ) so that 𝐸 𝜀𝑖𝑗 = 0, 𝑉 𝜀𝑖𝑗 = σ2 and
𝐶𝑜𝑣 𝜀𝑖𝑗 𝜀𝑠𝑘 = 0
(iii) Restriction 𝛼𝑖 = 0
32