Experimental research design, which can otherwise be called hypothesis-testing research design, were originally made by R.A. Fisher in agricultural research in England. Experimental research design is generally used in experimental studies where hypotheses are tested. Experimental research design is now used in almost all the areas of scientific studies.
The principles of experimental research design which Fisher prescribed are, viz.: (a) the principle of replication, (b) principle of randomization, and (c) the principle of local control.
- The principle of replication suggests that the experiment must be repeated also that the treatment is applied in many experimental units. This increases the statistical accuracy of the experiments.
- The principle of randomization indicating that the variations resulted by extraneous factors can be brought under “chance”.
- The principle of local control indicates that the extraneous factor is allowed to vary deliberately over a wide range so that the variability it causes can be measured and eliminated from the experimental error.
Experimental research designs can either be informal experimental designs or formal experimental designs. Informal experimental designs generally use less sophisticated form of analysis based on differences in magnitudes. Formal experimental designs provide relatively greater control and use statistical procedures for analysis. Informal experimental designs include: (a) Before and after without control design, (b) After only with control design, and (c) before and after with control design. Formal experimental designs include: (I) Completely Randomized Design (CRD), (ii) Randomized Block Design (RBD), (iii) Latin Square Design (LSD), and (iv) Factorial designs.
A single test group or area is selected and the dependent variable is measured before the introduction of the treatment in the “Before and after without control design” The dependent variable is again measured after the treatment. In the “After only with control design” two groups or areas (test area and control area) are selected and the treatment is introduced into the test area only. The dependent variable is then measured in both the areas at the same time.
In the “Before-and after with control design”, two areas are selected in which the dependent variable is measured for an identical time period, before the treatment. Then the treatment is applied into the test area so that the dependent variable can be measured in both for an identical period of time after the treatment.
Completely Randomized Design (CRD) is a simple research design. Subjects are assigned at random for experimental treatment. Principle of replication and principle of randomization of experimental designs are involved in this type of design and random replications design are the two forms of CR design. In the former a sample is selected at random from a defined population. In the latter a number of repetitive treatments are given in each case minimizing the effect of differences.
Randomized block design (RB design) makes use of the principle of local control. Subjects are first divided into groups (blocks). Subjects are relatively homogeneous in regard to selected variables in the RB design. When two extraneous factors are applicable. Latin square design is effective, while items are placed in a square. Out of the many items; say for example, five items are placed in five places each in the square but used only once in each raw/column. The two blocking factors on two sides of the square (vertical and horizontal are words. It is generally used in agriculture.
When a large number of factors affect a particular problem, factorial designs are used. There can be simple factorial designs and complex factorial designs. Generally two factors are involved in the former, and many factors (multi factors) in the latter. The independent variable to be manipulated is called experimental variable, and the extraneous variable to be controlled by homogeneity is called control variable.