Because the research process often draws simultaneously from several different types of research, it is useful initially to identify these types. In this chapter, therefore, we discuss the various types of research, and when they are appropriate.

How is Action Research Defined?

Action Research is a three-step spiral process of (1) planning which involves reconnaissance; (2) taking actions; and (3) fact-finding about the results of the action.
Kurt Lewin (1947)

Action Research is the process by which practitioners attempt to study their problems scientifically in order to guide, correct, and evaluate their decisions and actions.
Stephen Corey (1953) 

Action Research in education is study conducted by colleagues in a school setting of the results of their activities to improve instruction. 
Carl Glickman (1992) 

Action Research is a fancy way of saying let’s study what’s happening in our organisation and decide how to make it a better place. 
Emily Calhoun (1994)

As an example, if a company had a problem with absenteeism, then the steps in action research would be as follows:

1. The researcher would gather comprehensive data about both the specific problem (from the company) and the general topic (from a literature studv)

2. The researcher and the stakeholders would agree on some recommendations, and these would be implemented by the company.

3. After a suitable time-period, the researcher would make pre-agreed measurements to determine the effectiveness of the changes.

Some view action research as a philosophy of research rather than a method of research. They reject the attempted separation of the investigator and the problem, and instead embrace research that has the specific goal of social change, where the research is participatory and emancipatory. An example might be a community deciding on the siting of new facilities.

Creative research involves the development of new theories, new procedures and new inventions. For example, a computer scientist might apply new algorithms for managing a computer system, an economist might develop a new model of the world economic system, or an electronic engineer might design a new radio. Creative research is used to some extent in all fields. Unlike experimental research, creative research is much less structured and cannot always be preplanned.

Creative research includes both practical and theoretical research. Practical creative research is about the design of physical things (artefacts) and the development of real-world processes. Theoretical creative research is about the discovery or creation of new models, theorems, algorithms, etc. Practical research mainly proceeds by trial and error.

Historical Research

Studies of the past to find cause-effect patterns are known as historical research. It is often geared towards using past events to examine a current situation and to predict future situations (e.g. stock-market forecasting). The research does not directly study current causes or effects. Data is gathered from primary sources (i.e. records made when the past events took place) and secondary sources (i.e. records made after the events took place).

Expository Research

This is research based purely on existing information, and normally results in 'review'-type reports. By reading widely on a field, and then comparing, contrasting, analysing and synthesising all points of view on a particular subject, a researcher can often develop important new insights. An analysis of the works of a prominent author or a comparison of the tax structures in developed and developing countries are both examples of this type of research.

Experimental Research

The cornerstones of science are experimental and creative research. Experimental research is primarily concerned with cause and effect. Researchers identify the variables of interest, and try to determine if changes in one variable (called the independent variable, or cause) result in changes in another (called the dependent variable, or effect). Experimental research might be used to determine if a certain material is fire-resistant or if a new teaching method achieves better results.

Some variables might be present that are neither the independent nor the dependent variable, but still need to be considered. Say we wanted to test whether a new drug was effective in reducing blood pressure in rats. The independent variable would be the use or non-use of the new drug, the dependent variable the blood pressure. However, other factors may have an effect on blood pressure: for example, the rat's genetic make-up, its weight and its stress level. To get a true answer to the question of the relationship between the independent and the dependent variable, the effect of these other factors must somehow be cancelled out. One common way to do this is to use two groups, an experimental group and a control group, which are treated identically except for the use or nonuse of the drug.

Descriptive Research

Descriptive or 'case-study' research is research in which a specific situation is studied either to see if it gives rise to any general theories, or to see if existing general theories are borne out by the specific situation. An example of this is Mead's anthropological studies of isolated cultures to see whether pervasive social organisations are essential features of humankind (Lou91).

Descriptive research may be used when the object of the research is very complex. For example,, in trying to study the effectiveness of health-care delivery systems, a researcher might undertake an in-depth case study of a selected number of hospitals in a selected number of countries, and then compare them to see if any general trends emerge.

Ex Post Facto Research

While in experimental research the researcher exposes similar groups to different treatments to see the effects of the treatments (so moving from cause to effect), in ex post facto research he/she looks back at the effects and tries to deduce the causes from these effects. Ex post facto means `from after the fact', and this type of research typically occurs when data are available that could not be generated by experimental research. It is important to note that for ex post facto research to be valid, the researcher must eliminate all other possible causes.

The relationship between road development in an area and its current population would be an example. This could, of course, be experimentally tested, but few researchers have the funds to build road systems or the time to see the effects of these over 20 years!

Adapted from Goddard and Melville, 2001