Experimental Design

Some of the excerpts below are taken from the compilation of Science Education Institute (SEI), Department of Science and Technology (DOST), Engineering and Science Education Project (ESEP).

Experimental design refers to the complete sequence of steps undertaken to answer the research problem. It involves the logical structure and organization of an experiment. It is a detailed plan of the sampling procedures, data collection and data analysis followed by the investigator during the actual experimentation. It also includes the variables under study.

Functions of the Experimental Design

These include the following:
a. It provides direction during the actual experimentation.
b. It allows a gain of maximum information relevant to the problem at minimum cost.
c. It makes the statistical test of significance valid because it takes into consideration all the assumptions that went into deriving the various statistics.

General Features
The design of an experiment depends on the type of research undertaken and the nature of the conditions under which the study was done. The design of an experiment is dictated by the question it is to answer. There is no common blue print that will serve as a guide in writing an experimental design. Each problem requires its own unique design.

Basic Principles

The following principles are always applied in an experimental design.

• Replication- refers to the duplication of the exposed condition but only differ in one variable.

Ex. replicate 1: 30% extract and 70% water
replicate 2: 50% extract and 50% water
replicate 3: 70% extract and 30% water

All have the same components but differ on one variable which is the concentration.

• Randomization- refers the selecting of sample by chance not by one's preferred perspective. Thereby, bias selection of sample could be avoided and it also neutralizes pre-existing differences between samples that might affect the experimental results
  

Ex. drawing lots like that of a lottery ticket in selecting for the samples/subjects to be tested.

• Local Control- Variables that are kept constant to prevent their influence on the experimental results. If other variables are not controlled then it could lead to difficulty on identifying the exact variable that causes a change on the specified experiment.

Example in testing the strength of Paper-bag material from water-hyacinth fiber. Researcher must control the thickness of the paper when testing for the strength in trial 1, 2, and 3. To really be sure that the measured paper's strength is due to the fiber not because of the difference in thickness.

Other examples of local control are temperature, age, weight, height, sex, etc.