navbar 4Resource papers in action research

Action research: understanding its
philosophy can improve your

Pam Swepson


This is a resource file which supports the regular public program "areol" (action research and evaluation on line) offered twice a year beginning in mid-February and mid-July.  For details email Bob Dick  or

...  in which Pam Swepson identifies and discusses some of the philosophical assumptions which underlie different forms of research, especially action research


Action research: understanding its philosophy can improve your practice

Some comparisons with the philosophy of science


I believe that action research is not widely used because it is not well understood.  There are, I think, two main reasons for this.  Firstly, it is very different from scientific method, with which it is inevitably compared.  Secondly, some practitioners, with a poor understanding of action research, call what they do "action research" when it is simply unevaluated action.  This gives action research a poor reputation.

I am currently exploring this issue in two ways.  I am conducting some collaborative field research with respected scientific and social science practitioners to elicit models of best practice in both of these areas.  Secondly, I am exploring the paradigms of action research and scientific method on a philosophical or conceptual level.  I will them make comparisons between these two sources of data.

This paper is the second strand of my research, exploring the philosophical concepts and underpinnings of action research and science.

I have two reasons for this particular approach.  Firstly, this will explain the differences between action research and science in the "neutral" language of philosophy to overcome some of the problems of explaining action research in terms which have a particular meaning in the scientific paradigm and which do not translate to other paradigms of inquiry.  Secondly, this will demonstrate that a criterion for rigorous research is the link between the philosophy and the methods of a paradigm.  The philosophy of a paradigm sets the parameters for the type and use of methods.  Researchers in any paradigm who are ignorant of the parameters of their paradigm risk sloppy research.


The difference between values and logic in a paradigm

A paradigm of inquiry, like any another intellectual construct, is based on logically unprovable assumptions, i.e.  they are value choices.  The role of logic, then, is to dictate the statements or actions that must follow from those assumptions.  When comparing paradigms, it is important to keep the distinction between the values and the logic of a paradigm in mind.

The assumptions behind paradigms of inquiry are not logically defendable, therefore it is not possible to decide which paradigm is more logical than another.  Neither is it possible to argue that one set of methods is more logical than another.


Fitness for function

The best that we can do is to maintain an appreciation of a number of paradigms of inquiry in terms of their assumptions, and the strengths and weaknesses of their methods.  We will then be in a position to choose the paradigm of inquiry best suited to give the desired results in a given problem situation.

The argument for the best paradigm is largely a pragmatic rather than a logical argument and demonstrates the fitness for function of the chosen paradigm.

This paper is an attempt to increase our appreciation of two paradigms of inquiry and increase our ability to choose between them on the basis of fitness for function.


Which science, which action research?

Part of the problem of understanding science or understanding action research is that there are so many models of what they are.

For the purposes of my comparisons, I have adopted a somewhat Popperian model of post-positivist science.  Popper attempted to overcome the limitations of positivist science which says that science starts with observation with normal unimpaired sense organs and an unprejudiced mind.  If the same observations occur a large number of times and under a wide variety of conditions, then it is legitimate to generalise.  Post-positivists admit that observation is guided by and pre- supposes theory.  They also agree that theories cannot be established as true or probably true in the light of observational evidence because the probability of a finite number of observations, divided by an infinite number of possible observations is still zero.  Therefore post- positivists separate theory generation from theory testing.  Science, they say, progresses by rigorously and ruthlessly testing, and attempting to falsify, speculative theories, by observation and experiment.  It can never be said that a theory is true but that it is the best available.  (Chalmers, 1976.)

It is less easy to adopt a model of action research, but I suggest one where it these "metaprocesses" are in practice.  It pursues action and research outcomes at the same time.  It pursues a cyclic path to allow for a least one stage of critical reflection on the outcomes and the process.  This stage of critical reflection searches for both confirming and disconfirming evidence.  Action research tends, but is not necessarily, participative and dealing with qualitative data.  (After Bob Dick, 1993)


Some assumptions behind action research and scientific method

Each paradigm of inquiry faces some similar problems about inquiring into the world but make different assumptions about the nature of those problems.  These different choices logically lead to different methods.

I will start with the assumption that researchers in all paradigms of inquiry are aware of problems in the real world and are motivated to inquire into the nature of the real world to improve the human condition.


Problem 1.  Who decides if or what is the problem?

Value choice answer of the scientific paradigm to problem 1:

The research community identifies a community "need" that the community is unaware of.

Research driven by the research community's curiosity and need to know.

The logical consequences for scientific method of these assumptions are:

Consult and follow the standards set by the research community.

Value choice answer of action research to Problem 1:

A community identifies that it has a problem, a "want".

The logical consequences for action research method of this assumptions are:

Consult and follow the standards set by the research community.

Consult and follow community standards.  The community is integral in planning, implementing and evaluating.


Problem 2.  What is the aim of the inquiry?

Value choice answer of scientific method to Problem 2:


To improve the human condition by adding to the body of knowledge.  This is underpinned by the further assumption that knowledge is truth.

Logical consequences for scientific method of this assumption:

Gain a thorough understanding of the body on knowledge before starting further research.

Separate research from implementation.

Value choice answer of action research to Problem 2:

To improve the human condition by changing the situation.  This is further under pinned by the assumption that truth is what works.

Logical consequences for action research methods of this assumption:

Continual testing in the problem situation to find what works in practice.

Implementation is an integral part of the research design.


Problem 3.  What is the nature of the problem situation?

Value choice of the scientific paradigm to Problem 3:

The situation is part of the physical world with the assumed properties of being material, determined, and with cause and effect mechanisms.  This includes unconscious, reactive human behaviour.

Logical consequences for scientific method of this assumption:

Treat the world as separate to the researcher, i.e.  it becomes an OBJECT of inquiry and includes humans who may be erroneously called "subjects".

Determine the cause and effect mechanisms by controlling for extraneous variables.

Quantitative measurements best indicate cause and effect.

Value choice of action research to Problem 3.

The situation is part of the human world of ideas and beliefs.  This includes conscious, reflective human behaviour.

Logical consequences for action research methods to this assumption:

The researcher is part of the world.  All humans in the situation are subjects, determiners, co-researchers of the research.

Purposeful human behaviour to change the situation is the aim of the inquiry so overdetermine this outcome with multiple strategies.

Qualitative indicators most valid for data on beliefs and ideas.


Problem 4.  What is the relationship of the researcher to the world?

Value choice answer of the scientific paradigm.

The "real" world is as we perceive it.


We do not have direct access to the "real" world, only our perceptions of it, therefore aim to achieve objectivity by excluding the perceptions of the researcher.

Logical consequences for scientific method of these assumptions.

The researcher reduces their own extraneous perceptions by assuming a disinterested position and using statistics for analysis.

Control for the irrelevant perceptions of human "objects".

Value choice answers of action research to problem 4.

We do not have direct access to the "real" world, only our perceptions of it, therefore aim to achieve objectivity by INCLUDING the perceptions of all researchers.

Logical consequences for action research methods of this assumption.

Include ALL relevant perceptions of ALL relevant co-researchers.

Use techniques to find common and idiosyncratic perceptions.


The epistemologies of scientific methodology and action research.

Epistemology raises the questions of how we know what we know, and what is knowledge.

There are two classical epistemological positions and both are concerned about the relationship of our mind to our experience in generating knowledge.  Empiricism places primary focus on what we experience.  Empirical processes collect data about the world through our five senses and build up a body of knowledge which is seen as probably true rather than absolutely true through a process of induction.  Rationalism places primary focus on our mind.  The human mind is "hard wired" to comprehend certain principles which must exist and are therefore, real, a priori, and absolutely true.  Explanations about experience are deducted from these certain principles.

Scientific method is a particular combination of empirical and rational procedures.

Action research is another combination of empirical and rational procedures.


Scientific method as a combination of rational and empirical processes (Honer, Hunt and Okholm, 1992)

Scientific method starts by making a rational decision about the type of problem that it will pursue.  Scientific method makes the value choice to pursue generalisable knowledge rather than situation specific knowledge, i.e to pursue external validity at the expense, if ncessary, of internal validity.  Therefore, it choses problems where it is possible to extract meaningful relationships between discrete variables.

Scientific method continues with the rational processes of seeking out current knowledge about the relationship between the chosen variables through a literature or other searches.  It is then possible to develop through deductive logic, an a priori theory about the relationship of the crucial variables.  "Theory" is meant in the special sense of having elements which are empirically testable.  It is then possible, to set up a test for the theory.  To overcome the problem of induction (mentioned earlier in regard to postivist science) it is necessary to state the Ho and attempt to falsify it rather than look for evidence to confirm a H1.

Scientific method then proceeds with the empirical processes of observation or experimentation.

Rational processes then follow, given the results of the test.  If the Ho is not falsified, there tends to be a decision not publish but to loop back to the stage of developing a testable theory.  If the Ho is falsified, then a decision is made to accept the H1.  If the H1 fits the current body of knowledge, the results are published.  If the H1 does not fit the current body of knowledge the results may or may not be published.  If not, there tends to be looping back to the stage of assessing the current body of knowledge.


Action research as a combination of rational and empirical processes.

Action research also starts with the rational process of deciding on the type of problem to address.  Action research makes the value choice of pursuing situation specific knowledge rather than generalisable knowledge, i.e.  it will trade off external validity for internal validity, if necessary.  Therefore, it is generally applied to complex, social situations which are a complex set of relationships between indiscrete variables and it is not possible to choose which variables are crucial.

Action research then proceeds with the empirical processes of asking a fuzzy question to get a fuzzy answer to ask a less fuzzy question.  Through this cyclic process of asking general and then progressively more focussed questions, the researchers start to identify some crucial relationships.  It is possible then, through this inductive process, to develop a grounded theory.  "Theory" here has the special meaning of "being a map, not the territory".  The elements of the theory do not need to be testable, but the theory should lead to some action which will improve the situation.

Action research then continues with rational processes by assessing the test of the grounded theory.  It is important, whether the test "worked" or not, to look for and test for evidence to confirm or disconfirm the theory.  Without such a test, the research becomes unevaluated, "one shot" action.

The results of the test of the theory loop back to confirming or altering the theory and loop back to refining knowledge about the original problem situation.


Pam Swepson
Any comments, suggestion or arguments very welcome.
Voice 61 7 378 9670
Fax 61 7 878 9156



Copyright (c) Pam Swepson 1995.  This document may be copied if it is not included in documents sold at a profit, and this and the following notice are included.

This document can be cited as follows:

Swepson, P.  (1995) Action research: understanding its philosophy can improve your practice [On line].  Available at



navbar 4 

Maintained by
Bob Dick; this version 1.02 last revised 991121

A text version is also available at URL