An Intuitive Study of Experimental Design

Experimental design is the formulation of a set of rules and principles according to which an experiment is to be conducted to collect appropriate data whose analysis will lead to valid inferences for the problem under investigation. More precisely, experimental design is a way to carefully plan experiments in advance to get valid and objective based result.

An experiment

An experiment is a well defined act or an investigation conducted to discover the underlying facts about a phenomenon, which are utilized to test some hypotheses of interest to verify the results of previous investigations. More precisely, an experiment is the process of data collection from a non-existent population to get answer to the certain problems under investigation. There are two types of experiment. They are-

• Absolute experiment: An absolute experiment is one in which the absolute value of some characteristic is determined. Sample survey belongs to absolute experiment.
• Comparative experiment: A comparative experiment is one where two or more varieties or treatments are compared to assess the significance of difference among the varieties. Such experimental studies are based on test of hypothesis and estimation of differences among the effect of the different treatments in order to recommend best treatment.

Steps in experimental design

Design of experiment consists of the following steps-

• Choosing a set of treatments for comparisions.
• Selection of experimental units to which chosen treatments will be applied.
• Specification of the number of experimental units for inclusion in the experiment.
• Specification of the method of allocating the treatments to the experimental units.
• Specification of the measurements to be obtained from each experimental unit.
• Specification of the grouping of experimental units to control extraneous sources of variation.

Purposes of experimental design

There are some important purposes of experimental design-

• The main purpose of design of experiment is to collect maximum amount of necessary information for the problem under consideration, at a minimum cost in terms of time and resources.
• Design of experiment is needed to ensure that the requisite assumptions for analysis and interpretation  of data are fulfilled.
• Design of experiment is essential to increase the accuracy of the results of an experiment.

Principles of experimental design

According to R.A. Fisher, there are three basic principles of experimental design. They are-

• Replication: Replication means repetition of basic treatments under investigation. Thus replication is the repetition of same treatment on several experimental units. Experience indicates that even if same treatment is assigned to all experimental units, yields will be differ substantially. So it is essential to replicate the treatments to study the variation in the yields of each variety. 
• Randomization: Randomization is the process of distributing the treatments to the experimental units purely by chance or probability mechanism in such a way that any experimental unit is equally likely to receive any treatment. Then randomization ensures that no treatment is unduly favored or handicapped in the experiment. That means, it eliminates bias from the result of an experiment.
• Local Control: Local control is the procedure of reducing and controlling error variation by arranging the experimental units in blocks. By blocking, variation among the blocks is eliminated from the experimental error. So, it increases the precision of the result.

Requirements of a good experiment

A good experiment should satisfy the following conditions-

• Absence of bias: It is essential to plan an experiment so that unbiased estimates of treatment differences and treatment effect can be obtained from the data of the experiment.
• Measure of experimental error: Since the treatments under comparision apparently produce different results, test of significance in needed to assess the nature of treatment differences.
• Precision: Precisions depend on the experimental error. If the error is less than the precision is high.
• Clearly defined objective: Every experiment should have clearly defined objective on which the design and analysis of data considerably depend.
• Simplicity: Experimental design should be very simple and consistent.
• Range of validity: The conclusion drawn from the experimental data should have a wide range of validity.

Basic experimental designs

There are some basic or commonly used experimental design-

• Completely Randomized Design (CRD): A completely randomized design is a design in which the selected treatments are allocated or distributed to the experimental units completely at random. This design is divided into two category as, balanced completely randomized design and unbalanced completely randomized design.
• Randomized Block Design (RBD): A randomized block design is a design in which the whole set of experimental units are arranged in several blocks which are internally homogeneous and externally heterogeneous. 
• Latin Square Design (LSD): Latin square design is a design in which experimental units are arranged in complete blocks in two different ways, called rows and columns and then the selected treatments are randomly allocated to treatment.

• Time series analysis
• Business forecasting
• Hypothesis testing
• Logistic Regression
• Inference
• Correlation Analysis
• Data analysis using spss