Understanding Research Design

Research design is a set of advance decisions, a blue print, to help contribute towards the developing the master plan. The selection of research design depends on the research objective, what and how much information is already known about the problem. 

Exploratory design is undertaken to gain background information and to more clearly define the research problem. This design is mainly used in situations to (a) gain background information, (b) to define terms, (c) to clarify problems and hypotheses and (d) to establish research priorities. Attached to exploratory designs involves reviewing existing literature, surveying individuals who have knowledge in the are, use of case situations and focus groups are examples. The benefit of this method, is fast, inexpensive and at times, resolves the research objective. 

Descriptive design is where by the concepts and terms are already known and the concern of the research objective is to describe and measure then phenomena. Descriptive design is primarily concerned to answer questions of who, what, where, when and how and may be conducted at one point in time, known as "cross sectional study". However, if several measurements may be made on the same sample at different points in time, this is known as longitudinal studies.  Often, longitudinal studies are conducted using panels, sample units of people that have agreed to answer questions in periodic intervals, where as continuous panels are longitudinal studies in which sample units are asked the same questions repeatedly. 

A researcher is required to determine the causal relationships between 2 or more variables  Causal relationships provide relationships such as IF X, THEN Y, moreover, a further complex, IF X THEN Y ELSE Z OR A. These can be discovered through special studies known as experiments. Experiments helps determine the effects of variable, known as independent variable. However, if a variable is reliant on another, this is called dependent variable. 

The use of experiments helps to assess both the internal and external validity. If by proceeding with experimental designs, there is a need to ensure the effect observed of a dependent variable is present, in contrast to independent variable and not to others. Should this happen, this is known as extraneous variables.