Features of a good research design Research Methodology

Designing a research, particularly in the field of social sciences is very complex as the selection of a method or methods of logic and planning of the design was not always guarantee sound results. As a blue print, the research design may at best be only tentative and useful to the extent of providing the researcher with a series of guide posts to keep him headed is the right direction.

Although every design has its own strengths and weaknesses and simultaneously the possibility of a single perfect research design is difficult, a good research design is often believed to possess characteristic features like flexibility, appropriateness, efficiency, economically sound and so on. A design which minimizes bias and maximizes the reliability of data is construed as a good design.

Similarly the design giving the smallest experimental error is considered to the best design and the design yielding maximal information covering various aspects of a problem is construed as the most efficient design because it is appropriate to the research problem. Hence, consideration of a design as good depends too much upon the objective of the research problem and also the nature of the problem under investigation. 

 A single design can never serve the purpose of all types of research problems because what appears to be suitable in one case may be lacking in one respect or the other in the context of some other research problems. A good research design should always fulfill the following four conditions; objectivity, reliability, validity and generalizability of the findings. 

 (a) Objectivity:

The findings are said to be objective when they pertain to the method of data collection and the scoring of the responses. The objectivity in respect of the procedure may be judged by the degree of agreement between the final scores assigned to various persons by more than one independent observer. The more the agreement among the observers the more objective are the observation, recording and evaluation of the responses. Therefore, a good research design should permit fairly objective measuring instruments in which every observer visualizing a performance comes to the same conclusion.

(b) Reliability:

The question of reliability of knowledge is usually raised when the presence of a problem arouses in the knower a demand, not only for something more than mere conjecture, but for something for which it shall be useful in a given situation and perhaps in other similar situations. Reliable knowledge means any claim that is substantiated as trustworthy for a given purpose. 

 (c) Validity:

Validity implies self-consistency or absence of self-contradiction. It is identified with formal truth or self-consistency. A valid reasoning conforms to the rules of correct reasoning. It is that type of reasoning where conclusions automatically follow from the premises legitimately. 

 (d) Generalisability:

The degree of generalisability is known in terms of the replicability and reproducibility of the findings in spite of different measures and settings respectively.