Scribd
Upload
6 views8 pages

Module

The document discusses the Design of Experiments (DOE), a statistical approach for planning and analyzing experiments to enhance precision and reduce errors. It outlines key concepts such as treatments, experimental units, and basic principles including randomization, replication, and local control. Additionally, it introduces various experimental designs like Completely Random Design (CRD), Randomised Block Design (RBD), and Latin Square Design (LSD), along with their respective ANOVA tables.

Uploaded by

XE 33 SHRIDHARAN VK
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
6 views8 pages

Module

The document discusses the Design of Experiments (DOE), a statistical approach for planning and analyzing experiments to enhance precision and reduce errors. It outlines key concepts such as treatments, experimental units, and basic principles including randomization, replication, and local control. Additionally, it introduces various experimental designs like Completely Random Design (CRD), Randomised Block Design (RBD), and Latin Square Design (LSD), along with their respective ANOVA tables.

Uploaded by

XE 33 SHRIDHARAN VK
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Design of Experiments (DOE)

By experiment, we mean collection of data for a scientific investigation, according to certain


specified sampling procedures. Statistics provides not only the principles and the basis for the
proper planning of the experiments but also the methods for proper interpretation on the
results of the experiment.

Experimental Design is the branch of statistics that deals with the design and analysis of
experiments.

 Design of Experiment: Design of Experiment is powerful data collection and analysis tool
that can be used in a variety of experimental situation, which assigns a proper treatment to
the available experimental units.

 Treatment: It is the factor whose effect is to be studied.

Example: Application of drugs, feeds, conditions, methods, etc.

 Experimental units: The subjects or individuals upon which an experiment is conducted.

Example: Laboratory animals, primates, cattle, buffalo, sheep, goat, poultry, plant, a
plot of land, etc.

Objectives of DOE:

1. To increase the precision of experiment.


2. To reduce the experimental error.

Basic principles of experimental design:

1. Randomization: It is the random process of assigning treatment to the experimental units.

2. Replication: It is the number of times the effect of treatment is measured. In other


words, it is the number of observations made under each treatment.

A C B C A 4 times No. of treatments is 3 (i.e., A, B, C)


C B A B B 4 times Replication = 4
A C B A C 4 times
A C B A D 4 times

3. Local control (Or Blocking): By using local control i.e., block the grouping of experimental
unites is made to make them homogeneous under each group.

Some basic Design of Experiments (DOEs):

1. CRD: Completely Random Design.


2. RBD: Randomised Block Design
3. LSD: Latin Square Design.
1. CRD: Completely Random Design

This is simplest & flexible design. The experimental units & treatments are allotted randomly,
so that every unit gets the same chance of receiving every treatment.

ANOVA Table for CRD (One way ANOVA)

Source of Sum of Degree of Mean Sum of F Calculated


Variation Squares (SS) Freedom (Df) Squares (SS) Value
Between samples No. of Samples
Q’1=MST=
or Class SST=Q1 (Treatments)-1= Q’1/Q’2>1
Q1/(h-1)
(TREATMENT) V1= h-1
Total no. of values OR
Within Sample
RV - No. of Q’2=MSE=
(Error) SSE=Q2
Samples= Q2/(N-h) Q’2/Q’1>1
V2= N - h
Total Q=Q1+Q2 (h-1) + (N-h)=N-1

Steps for calculating Q, Q1 & Q2

1. Calculate T (treatment): T   xij


i j

i.e., xij is the observation of jth member of ith sample


T2
2. Calculate Correction Factor (CF):
N
  xij 
2
3. Calculate
i j

T2
Q    xij 
2
4. Calculate: 
i j N
2
Ti
5. Calculate: n
i
, i.e., calculate T 2 / N for each class and add all
i

Ti 2 T 2
6. Calculate Q1   
i ni N
7. Calculate Q2  Q  Q1
8. Then compare the F calculated value with F table value at DF (V1, V2).
If Fcal < F(V1, V2) for given LOS, the H0 will be accepted.
If Fcal > F(V1, V2) for given LOS, the H1 will be accepted.
2. RBD: Randomised Block Design

In RBD a single observation is made on each of the experimental unit, and then it’s ANOVA for
fixed effect model for two-way classified data with independent observations.

Note: It is classified according to two factors/criteria.

AVOVA table for RBD (Two-way ANOVA Table)


3. LSD: Latin Square Design

In LSD we use three-way ANOVA table:

You might also like