navbar 4Resource papers in action research

Postgraduate programs using action

Bob Dick


This is Chapter 7 of Action learning, action research and process management: theory, practice, praxis, edited by Ortrun Zuber-Skerritt.   Brisbane, Australia: Action Research Unit, Faculty of Education, Griffith University.

It is included here as 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 some key choices faced by higher degree candidates are identified and discussed


Postgraduate students often choose action research, and its cousin action learning, for their theses and dissertations.  In this chapter I consider advantages and disadvantages of these research approaches.  I also consider some of the important choices faced by candidates who consider using action research for their study.  Most of my discussion here concerns action research, since for most postgraduate students this is the component most evident in their thesis.  Some students may also use learning as part of the methodology of their thesis or dissertation.  The most important contribution of action learning, however, is its use to create a support network, as I discuss below.

I am not evangelistic about action research or any of its many different varieties.  To my mind, good research is designed to fit the interests and skills of those involved and it is partly created by and helps to create the research situation and the research questions.  Thus, recognising the need for flexibility mainstream action research, an extension and modification of action research, or something else entirely, may be most appropriate depending upon the researcher and the research task.

Unlike some of my colleagues I hold that most aspects of research design are open to choice.  For example I do not believe that action research must be participative, or qualitative, or published.  It often is and I accept this.  Indeed, many writers such as those in Chapter 3 of this book argue for one or more of these characteristics.  But I do not recognise this as a convincing reason to limit my own research choices.  You will make up your own mind.

If action research is not necessarily participative, qualitative or published, what is it and what are its characteristics? Broadly speaking, I think of it as a family of research methodologies that pursue the dual outcomes of action and research.  In that sense it is true to its label.  This is one of the reasons why it is particularly suitable for postgraduates who wish to improve their own work practice while they pursue a postgraduate qualification.  Like most other proponents I hold that action research profits from the use of a cyclical or spiral process in which the researcher alternates action with critical reflection.  I regard its cyclical/spiral process and its pursuit of both action and research as its defining characteristics.  If these characteristics suit what you plan to do in your research, then action research may be a good choice for you.  I subscribe to the design principle that form follows function: decide what you want to do, then decide how best to do it.


Action research for theses and dissertations

As I said above, I think of action research as a family of methodologies, each of which simultaneously pursues action and research.  The action takes the form of change or improvement or implementation in one's workplace.  The research consists of learning and understanding, often leading to publication of results.

Both processes -- action and research -- and their outcomes are likely to be important to a higher degree candidate.  The research component is obligatory in a thesis or a dissertation.  Some candidates find it useful to emphasise the action component.  Depending on their research question they may face a choice between theory-driven and data-driven research.  Let us consider these choices.


Choice 1: Theory driven or data driven?

I believe that the distinction between qualitative and quantitative research is less important than that between theory-driven and data-driven research.  I think it is a choice of paradigm, in the sense in which Kuhn (1970) and others use that term.

To make this distinction more evident, let me overstate it a little:

Do you want to do research that turns first to a body of extant literature and contributes to knowledge by assuming that literature as a given and extending or refining it, or challenging it?

This is theory-driven research.  Those who prefer to take this path may be better choosing an approach other than action research.  It appears that most current quantitative research and much qualitative research takes this theory-driven form.  But what if you begin from a different source and perspective, what can you do?

Do you wish to deal with the research situation and the people in it as they are, as far as possible putting aside your preconceptions so that you are more open to fully experiencing the research situation?

This is data-driven research.  It is responsive to the situation, and flexible.  The form of action research that most interests me takes this second form.  I argue below that it offers particular advantages to the higher degree candidate who wishes to research her own practice, improve her research situation, and add to her qualifications.

Let me be clear here.  I am not saying that all action research is data-driven.  Sagor (2000), for instance, clearly acknowledges that his approach is more quasi-experimental than anything else.  Nor am I saying that all data-driven research is action research.  Glaser in particular describes a strongly data-driven form of grounded theory in his writing (for example 1992, 1998).  I would also expect heuristic inquiry (Moustakas 1990), among other methodologies, to work well as a strongly data-driven approach.

For a higher degree candidate using a theory-driven or data-driven approach this is a crucial choice.  A data-driven approach carries some risk.  But it also offers advantages because of its dual goals of action and research.  Let us consider the costs and benefits of data-driven action research.  I see potential for risk in three main forms.

  • The candidate may well have prior training or experience in theory-driven research and this may have given her some expectation that can mislead her as she goes about her thesis work.
  • The way in which universities structure higher degrees is often premised on the assumption that good research is theory-driven.  For example, postgraduate research students may be expected to spend their first year digesting the literature and write a summary of the literature before their candidature is confirmed.
  • Some examiners may be surprised by data-driven research because it does not fit their notion of legitimate higher degree research.  Among them may even be some examiners who regard themselves as sympathetic to and supportive of such approaches as action research.

Recognising these risks from the outset, the person choosing a data-driven approach must be conscious of the potential to be mislead by research experience.  They must also have thought carefully about justifying their choice as they will need to argue the case very clearly and cogently.  Those who choose this approach because they imagine it is easier and bypasses a lot of academic reading are mistaken.  These are not risks to be undertaken lightly.

Yet despite these risks, the advantages of data-driven action research may sometimes be very attractive.  I think this is particularly true when the three possible outcomes of data-driven action research are all important to you as a researcher.

  • In the course of doing the research, you are able to study and improve your own practice.  In effect you use the higher degree as valuable personal and professional development.  You may also form a learning set with other postgraduate students for mutual support.  Personal and professional development can then be enhanced further.
  • One of the purposes of action research is to improve the research situation.  If you research your own practice, it is very likely you will also improve your work situation because you better understand your place in it.  The action outcomes you achieve are also likely to benefit your organisation or community.
  • And of course, with successful completion of your action research you gain a higher degree qualification.

I have recently co-supervised a number of action research practitioners who, by the end of their higher degree, had achieved all three of these outcomes.  All of them worked full-time in demanding jobs.  Yet most completed their degree in approximately the same time as do most degree candidates who are enrolled full-time.


The implications of data-driven research

It may appear from what I have said above that data-driven action research is an easy choice for candidates who find the three advantages appealing.  Let me therefore depict in more detail two of the features of this approach that may surprise some prospective degree candidates as well as some supervisors, committee members and examiners.

First, when you use a flexible and responsive approach you probably cannot know where the data will lead you.  Choice of literature is therefore difficult; at the start of a study you may not know what literature will later become relevant.  It therefore makes sense to postpone reading until the relevance of literature can be judged easily.  In turn, this may have consequences for the structure of the eventual thesis.

Second, it is not just the interpretation -- the understanding -- that emerges slowly from the situation.  The same applies for choosing methodology.  You can begin action research by asking initially fuzzy questions using initially fuzzy methods, thereby gaining initially fuzzy answers.  You may then use those initially fuzzy answers to refine your methods as you proceed.  To say it differently, research content and research process both develop as the research proceeds.

Checkland and Holwell (1998) add a further component.  They describe research as a framework F, operationalised as a methodology M, focussed on an area of concern A.  In the course of the research one learns about and therefore modifies F, M and A.  I think they are talking about all research.  And I think they intend their framework F to be an epistemological framework, consisting, for example, of assumptions about what counts as knowledge.  I look forward to the day when all researchers will be expected to explain and justify their epistemology, their methodology and methods, and their conclusions.  I think it is evident that such an approach encourages researchers to be more mindful as they go about their research.  (I take the adjective 'mindful' from Bentz and Shapiro, 1998.)  This brings us to the second choice.  How is the candidate to engage with the research methodology and with the supervisor or committee?


Choice 2: Researcher as technician or performing artist?

I might as easily have taken this as the first choice, for it can apply to all forms of thesis and dissertation research.  You might describe it as an existential choice.

Here I phrase the choice in the following form.

Do you want to be an apprentice who will learn thoroughly a particular approach to research from the supervisor, committee, and literature? That is, will your learning be primarily propositional? At the conclusion of such a research program you can expect to know how to do one form of research.  To overstate the situation, this is research by recipe.

Or do you expect to engage in research with whatever resources and understanding you can bring to bear, learning from your experience? That is, will your learning be primarily through questioning inquiry, with supervisor and committee functioning as mentors rather than as teachers? Such an approach will engage you in examining your assumptions about the nature of knowledge and of methodologies.  This is research as performing art.

You may recognise the influence of Checkland and Holwell (1998) and also of Revans (1983) in the phrasing of this choice.

Action research lends itself to the second of these options.  It alternates action with critical reflection.  That critical reflection can be about the data and the interpretations that the researcher is making from it.  It can also critique and improve the methodology.  Beyond that, it may be used as an opportunity to examine the assumptions about knowledge that inform the research design.

There appear to be enticing parallels between action research and experimental research.  (I was trained as an experimental psychologist.)  It seems to me that good researchers refine their methodology and methods from study to study.  This is a process that resembles the spiral nature of action research.  It applies to a study at a time, whereas in action research each turn of the spiral is an opportunity for learning and change.

If I were embarking on a study with uncertain outcomes I think I would begin with action research.  It would allow me to build enough understanding to decide which methodology and methods best suited my research situation and my research question.  Whatever methodology I chose, I could still use action research as a 'meta-methodology'.  I could thus enhance my learning about whatever methodology I was using.

I expect it is by now apparent that action research offers substantial flexibility and responsiveness to the situation.  It is not for the faint-hearted.  But it does offer a manager or a professional an opportunity to achieve some important practical outcomes.  If you choose action research you can choose to embark immediately on action.  This brings us to a third choice.


Choice 3: 'Action research' or 'action research'?

Action research is action and research.  Either of these terms may be emphasised to produce a distinctive meaning.

Is your main intention to bring about change, with research outcomes as a desired but sacrificable bonus? This will be action-oriented research.

Or do you desire above all to do good research, with change as a hoped-for but not essential outcome? This will be research-oriented research.

This is not to say that action research must be one or the other.  Action and research can fit each other well.  All else being equal, better understanding allows more effective change.  For the most part, trying to change a system will give you a better understanding of it.  As choices, action and research are emphases, not absolutes.

In any event, for thesis purposes you have a more limited choice.  Almost certainly, you need research outcomes.  So your choice may become deciding how much change you are going to pursue.  In turn, that may influence how participatory you are able and willing to be.

You may have little interest in action.  You may nevertheless wish to be flexible in pursuing understanding of the situation on its own terms.  If so, action-research-like processes may serve you well (though you may offend those whose commitment to action research is ideological).  Alternatively, you may find grounded theory a better choice.  It too is a data-driven approach, though this is truer of the version championed by Glaser (1992) than of that by Strauss and Corbin (1997, 1998).

You may also 'mix and match' methods within action research.  There are methods used in grounded theory that may be incorporated in action research, and vice versa.  Yet it is important to note here that action research can be rigorous without surrendering action outcomes.  The sources of rigour may be different to those that characterise other research styles, especially theory-driven research.  But they are no less effective.


Rigour in action research

This is more of an imperative than a choice.  Your thesis will be expected to make a contribution to knowledge.  You will be expected to demonstrate that your claimed contribution is supported by rigorous research, evidence and argument.

The danger is that your supervisors or committee, and your examiners, may assess rigour using criteria appropriate within their own paradigm.  They may value quantification, precise research questions, substantial early literature review, and the like.  It is therefore important that you understand the ways in which action research achieves rigour so that you can justify convincingly what you have done.

Some of the arguments you can use for this purpose are to be found in the qualitative literature generally.  I'm thinking particularly of the well-reasoned monograph by Kirk and Miller (1986).  Also relevant are Patton (1990), and many of the papers scattered through Denzin and Lincoln's (2000) handbook, among others.

For the most part, I have not found the action research literature particularly concerned with issues of rigour.  Some of the better arguments for rigour are to be found in work that some would regard as at the margins of action research.  I'm thinking here of such writers as Argyris (particularly Argyris, Putnam and McLean Smith 1985) and Checkland (particularly the first half of his 1981 book) among others.

The tight spirals of action research, however, deserve special attention here.  Their main function is to provide flexibility and responsiveness for effective change.  They also confer two advantages that are less evident in theory-driven research.

  • Each turn of the spiral provides you with yet another chance to test the interpretations you have so far developed against the data you are collecting.
  • Within each turn you develop plans to test in action.  Each turn of the spiral is a miniature test of the assumptions that guided your plans.

These advantages derive from the nature of each turn of the spiral.  I described this spiral earlier as alternation between action and critical reflection.  The critical reflection can also be regarded as having two components.  First is a review of the results of the previous action.  Second is the planning for the next action.  Action research can be described as a regular cycle of planning, action and review.

It can also be viewed as Kemmis and McTaggart (1988) explain: 'plan, act, observe, reflect'.  You may note its similarity to other cycles, such as that for experiential learning (for example see Kolb 1984) or for quality management (for example see Rothwell, Sullivan and McLean 1995).

Not all varieties of action research emphasise its spiral nature which, to my mind, confers to action research many of its advantages.  You have different versions of action research to choose from.  Here we see how action research offers the flexibility that it advocates.


Choice 4: Which action research methodology?

Methodological choices in action research are many.  Here I attempt to sketch the breadth of choice that researchers face, and some of the varieties that are possible.  They include:

  • many varieties of participative (or 'participatory') action research (PAR), as explained by, for example, Whyte (1991), Greenwood and Levin (1998), Smith and Willms (1997), and Zuber-Skerritt (1996), to name just a few;
  • several varieties of systems methodology, including those of Checkland (Checkland and Scholes 1999) and Flood (1997);
  • teacher research such as that of Mills (2000), Hubbard and Power (1999), or the 'living theory' approach of Whitehead and his colleagues (Whitehead 1993), among others;
  • the substantial body of work on 'action science' developed over many decades by Argyris and his associates: Argyris (1999), Argyris and Schön (1996), or Argyris, Putnam and McLain Smith (1985);
  • approaches whose similarity to action research is more or less explicit; for instance those of Forrester (1999), Heron and Reason (1997), the 'action inquiry' of Torbert (Fisher and Torbert 1995), the 'action evaluation' of Rothman (1997), and some aspects of 'appreciative inquiry' (Elliott 1999);
  • finally, some of the participative evaluation methodologies, such as the utilisation-focussed evaluation of Patton (1997) and the fourth generation evaluation of Guba and Lincoln (1989).

This by no means exhausts the possibilities.

In planning this paper I intended to comment on choosing from this rich array.  But that is a task too large for the parameters of this chapter.  Instead I will suggest the following issues that may be useful to guide decisions on research methodology.

  • Is your chief motivation for using action research your strong commitment to participation and equity? This may indicate that the more deliberately participative approaches are suitable for you.
  • Do you want to develop your understanding of a richly interconnected and complex system? Especially if you can treat it as an information system or a decision-making system, one of the systems methodology may be indicated.
  • Do you work in an academic setting where action research is regarded as a fringe paradigm and is highly suspect? You may then be able to frame your research as 'evaluation', which may be more respectable to others in your academic setting.
  • Are you drawn towards understanding how interpersonal relationships and system dynamics interact to undermine or produce effectiveness? Action science was devised to address this issue.

Each of the methodologies mentioned above offers its own flavour and style for conducting research.  As mentioned earlier there are valid reasons to avoid the literature related to your research theme if your approach is data-driven.  The methodological and philosophical literature, on the other hand, is well worth your early perusal, especially for insights into choosing and gathering your data.

From here we will turn to address a fifth choice concerned with the researcher's level of participation in the research project.  My previous experience suggests this is a contentious point, so I will first state my position clearly.


The virtuous spiral

It seems to me that there are two important dimensions to action research.  The first relates to the action, which is more likely to be achieved if there are high levels of participation by the people involved in the research.  You may also choose strong participation to be consistent with your of ethics or values.  The second dimension be found in the strategies the researcher pursues to maintain rigour.  I turn to these strategies now.

The strategies that the researcher pursues to maintain rigorous research must be effective within a process that is very flexible and very responsive to the research situation.  The strategies are developed for that reason.  They are not a second best approach because the best is not available.  They have virtue in their own right.  The virtue is derived from different sources.

You might decide that extensive participation is not for you.  You may decide it is too difficult for a thesis.  It may introduce complications you wish to avoid.  You might nevertheless decide that you can use action research-like processes for data collection and analysis.  In other words, you might still make use of data-driven action research processes for the research, but be less concerned about the action.  This introduces our fifth choice.


Choice 5: What extent and style of participation?

As mentioned, participation is a contentious issue and in some quarters this has become an ideological choice.  For many people it is therefore not a choice but an imperative.  Yet even if you accept this as imperative there is a spectrum of participation from which one can choose.  This stretches from minimal participation, if any (see Clark 1972), through to 'do it yourself' approaches (cf.  Wadsworth 1984).

At one extreme is much 'teacher research' where an educator researches her own practice and pupils may or may not be involved.  Towards the other extreme are approaches seeking 'emancipation' as a goal and treating political considerations as an issue.  Many of these approaches are to be found in research that takes the ideas of Habermas (1984, 1987) as its foundation.  The Deakin School (Kemmis and McTaggart 1988) which favours this approach has been influential in educational circles in Australia.  Elsewhere there are similar approaches, for instance that of Toulmin and Gustavsen (1996).  Other critical approaches (for example that by Ghaye and Ghaye 1998) draw on yet other foundations.

In any event, participation is far from being an all-or-none choice.  Who will you involve? How will you go about involving them? What will you involve them in? Each of these questions offers further choices.

  • In large communities and organisations, will you involve everyone? Will you also involve those outside the community or organisation who also have an interest? If numbers are very great, will you reduce the depth of participation to make room in your research for everyone?
  • How will you approach people and engage them in the research process? For instance will you broadcast invitations widely? Perhaps you will slowly build relationships as you work your way through the networks.  Or will you set up a steering committee, reference group, or the like, and ask them to involve others?
  • You will probably want to approach people to be informants.  Beyond this, what will you involve them in? Will they interpret the data as well as provide it? Will they help you design the study? Of course, you will write the thesis or dissertation.  But if there are other reports, who will write them?

These are important questions, I believe.  Yet in much of the action research literature they are not given much attention.

Fortunately, other fields address these issues in ways that are relevant to action researchers.  I include here evaluation (see Gregory 2000) and the literatures of organisation development (for instance, Toulmin and Gustavsen 1996, mentioned above), community development (McNair 1998) and rural development (Chamala and Keith 1995).  It would be remiss of me to omit the public participation literature.  Sarkissian and Perlgut (1994) offer an Australian example.

Action learning also fits well with action research.  Current approaches to action learning as described by such writers as Marquardt (1999), Rothwell (1999) or Raelin (2000) bear a close resemblance to group-based participative action research.  An action research group can also function beneficially as an action learning group ('learning set').  Although in earlier times action learning used a different style, current action learning mostly gathers participants from one organisation around a shared project, much as an action research study might do.  Although action learning has been used mostly in organisational settings, it also functions well in a variety of other situations.

Two other important issues concern participation.  First, who decides the nature and extent of participation? You may be committed to involving participants as co-researchers and equals in an emancipatory relationship.  But that may or may not be their desire.  Sometimes you may have to settle for less than you wish, at least initially (cf.  Grundy 1983).

Second, your thesis or dissertation is expected to be an independent and original contribution to knowledge.  You may therefore have misgivings about the heavy involvement of participants.  You may fear it undermines the independence of your contribution.  Zuber-Skerritt and Perry address this issue persuasively and helpfully in Chapter 8 of this book.  As these authors explain, the field study may be done with the style and level of participation that suits you and your participants.  Your individual reflection upon that study then provides your own contribution.

So far in this chapter we have considered the strategic issues; these do much to determine the overall shape of the study.  However a number of other choices facing the researcher are also worthy of note here.  I begin by considering methods to operationalise the methodology.


Choice 6: Which methods for data collection and analysis?

Much of the literature on action research does not explain in detail how it might be done.  But this need be of little concern to researchers considering action research.  You can be vigilant in ensuring that your chosen methods are consistent with both your action and your research aims.  With this protection, many of the methods used for other approaches can be pressed into service here.

The methods used most commonly for data collection are interviewing and focus groups.  These are well covered in the literature, as the following examples demonstrate.

Glesne (1999) gives a readable account of both interviewing and rapport building (the latter a much neglected topic).  In his wide-ranging book, Kvale (1996) mostly treats the interview as conversation, but also addresses its philosophical and scientific context.  Taylor and Bogdan (1998) give considerable attention to interviewing in their general text on qualitative research.

Focus groups have been used most in market research, though their origins were in qualitative research.  The market research literature can be useful for its attention to the practicalities.  For instance, the work by Edmunds (1999) is practical despite its title.  In recent times focus groups have been returned to their origins, and high quality accounts are appearing.  Barbour and Kitzinger (1999)  are a good example.  For a practical account that also considers data analysis see Krueger and Casey (2000), among others.

Focus groups and interviews do not usually involve strong participation.  If your work involves strong participation you may well find other literatures more useful.  Earlier I mentioned the literatures on organisation, community and rural development that could be useful for this approach.

Much of your time is likely to be spent with groups of people and here the literature on group facilitation is instructive.  Heron (1999) is familiar with collaborative research methods and says much that is useful, as does Kaner (1996).  There are also structured group processes such as group feedback analysis (Heller and Brown 1995) that can be valuable in participatory research.

For making sense of the data you gather, Miles and Huberman (1994) provide some useful techniques with detailed and practical descriptions.  The coding methods of grounded theory may also be useful: see Glaser (1998) and Strauss and Corbin (1997), though in my view, the latter study overcodes its data.

As you make these choices about methods, it remains important to continue pressing for appropriate rigour in what you do.  Researchers also need to recognise choice in the criteria they adopt for evaluating research, and in how they treat the very pertinent issue of the generalisability of their research results.


Choice 7: Rigour or relevance?

Fuelled by the development of new paradigms, there has been much debate over the criteria for judging research.  Some believe that qualitative approaches and constructivist philosophies require their own assessment criteria, distinct from those that characterise quantitative research.  Lincoln (for instance 1990) argues this vigorously.  However we cannot accept these views without question.  I have argued that the most important research choice is not qualitative versus quantitative approach, that a theory-driven or data-driven approach is a more influential choice.

This brings us to acknowledge the existential position of action research.  Action research is intended to act on some part of the world, however small or large.  This presumes that part of the world to be acted on is something more than a mental construction.  What we know or think that we know of it may be a mental construction -- and I assume it is.  I assume also that our mental construction stands in some relationship to the world 'out there' that exists as real or actual, rather than simply as a mental construct.  And it is 'out there', a palpably real or actual world, that we wish to influence.  Cherry (1999) examines such issues as they apply to action research.

On the grounds explored above, it seems to me that reliability and validity are still useful concepts.  For example, it is still my intention that as far as possible the information I collect is not idiosyncratic; that it is to some extent reliable.  I attempt also to discover as accurately as I can the experience that my informant describes so it is to some extent valid.  The greater the validity and reliability of the information and insight that I obtain, the greater my confidence when I draw on my understanding to take action.

Generalisability is a key issue.  I agree with Baskerville and Lee (1999) that qualitative researchers have given up too easily on this issue.  These authors recognise several forms of generalisability that are more accommodating than the stricter views upheld by most qualitative researchers.  As a researcher I also see ways to extend the generalisability of case studies, believing as I do that most action research studies can be regarded as case studies.

The experimentalists have some grounds for claiming that their findings have at best the status of general laws.  Suppose they find through their research that variable A consistently influences variable B in a certain way in extensive testing.  The claim can then be made that all else being equal, variable A always influences variable B in that way.  As far as it goes, this is a valid claim.

We could reply that this is less useful than might be imagined.  Generalisability is not imbued with as much virtue as claimed.  In the situations we deal with, it is seldom that only variable A and variable B are operative.  The status of variable B is then indeterminate.

My own approach is to acknowledge that there is some trade-off involved between discovering general truths and being relevant to the local situation.  My choice of 'action research' or 'action research' may tempt me to favour one or the other of these.

However, I can aim for a 'best of both worlds' combination of them since there are grounds I can use to argue for the generalisability of my results.  These grounds can be developed through logical analysis, multiple case studies, or highly diverse samples.  I can also compare my interpretations to those in the relevant literature.

So far, we have talked about choices that influence the shape of the research as it is carried out.  However, except in an indirect manner, these considerations are not the basis on which examiners will assess your research.  No matter how well you carry out the research, you will need to capture it well in your thesis or dissertation.  This is what you will be examined on.


Choice 8: How will I write up my research project?

Here, the choice you face is between a conventional thesis and a more tailored approach.  Based on their training, many supervisors, committee members and examiners have a preconceived notion of thesis structure.  It probably has chapters titled something like Introduction, Methods, Results, Discussion, and Conclusion.  This is the convention, and may be a safer way for postgraduate students to satisfy institutional requirements.

However, if you treat action research as data-driven and responsive, your study may not fit this structure without some trimming.  An alternative is to devise a structure for presentation that allows you to reflect the conduct and style of action research.  To illustrate how you might do this, I refer here to two papers on a web site at Southern Cross University that reflect these two strategies.  One is by Perry (2002) offering suggestions for structuring the paper so that it appears conventional.  I have written the other paper (Dick 1993), intending to capitalise on the particular virtues of action research.  I describe a thesis structure that is organised around the thesis's contribution to knowledge, with less concern for being conventional in form.  From feedback I have reason to believe that postgraduate candidates have found these models instructive, some choosing one approach and some choosing the other.

So far in this chapter I have considered most of the design choices that are important in approaching a thesis or dissertation.  In the remainder of this chapter I turn to the actual conduct of the study.  I discuss some of the traps that await unwary candidates.  I then discuss briefly an action learning set -- a supportive group of colleagues -- as an effective way to help overcome these traps.


Avoiding the traps of procrastination

I hope it is clear from the preceding discussion that action research has distinct qualities that set it apart from most other approaches to research.  If you take into your action research study some assumptions developed within other approaches, these may inhibit you from working most effectively in your chosen paradigm.  In particular, I think there are four traps that can be described as forms of procrastination:

  • postponing data collection until you have read all the relevant literature;
  • postponing analysis until you have collected all the data;
  • postponing action until you have interpreted all the data;
  • postponing writing until all other aspects of the research are complete.

Action research does not require extensive preparatory reading, extensive early data collection or complete analysis.  It lends itself to early action.  It does not even require that you have a research question or 'thematic concern' to begin (Kemmis and McTaggart 1988), though one may well be useful.  It is enough to have a research situation.  After you begin to take action you will soon begin to identify the thematic concerns.

Data-driven research provides an approach that suits activists, managers and other people who wish to act on the world and understand how they do so.  Each turn of the action research spiral can contain data collection, data analysis, action planning, action, and evaluation:

  • data are interpreted as they are collected;
  • interpretations are turned into action plans, which are then immediately acted on;
  • the results of the action can be reviewed immediately;
  • though you will wish to revise later, you can capture on paper your conclusions as you develop them.

It is useful to read some of the methodological literature at the beginning of your research so that you can identify your research paradigm and approach.  You don't wish to find further into your research that your methodology was inadequate.  The content literature can be left until later.  The early imperative is to find a suitable research situation, and achieve early ethical approval.  (This latter topic deserves a paper of its own.)

As I have said, each turn of the spiral tests your assumptions in action.  Each turn gives you another opportunity to test your emerging interpretations by vigorously seeking counter evidence.

There is also a final trap.  If you succeed this far you may then suffer the opposite procrastination: allowing yourself to be distracted from the theory by the action.  As I said earlier, action research can be appealing to those who wish to act on the world.  But it can also tempt them to keep taking action since when an action has been successful, this is incentive to press ahead to the next action.  This a trap since it distracts the researcher from returning to other essential stages in the research process.

Regular, systematic and critical reflection allows an escape from this final trap.  This provides the understanding that accrues to become your contribution to knowledge.  It is also where you develop the understanding to improve your work practice and your work situation.

Continuously documenting, carefully but economically, helps as well.  You can also give more attention to the planning phase of each turn of the spiral.  During planning you can become conscious of your assumptions (about epistemology, methodology, the situation, and action).  And having made them explicit, you are then more likely to notice the evidence that allows you to refine them.

The thesis or dissertation is an individual contribution.  Nevertheless, you don't have to do the study or the written report in isolation.


Creating a support network

As already mentioned, an action research study may draw on action learning methods to engage participants in the research and action.  Action research and action learning have grown together despite their different origins and early history.

In addition, you may use the processes of action learning to organise a support group.  You can meet regularly with other higher degree candidates in an action learning set.  There you can provide each other with mutual challenge and mutual support.  You can thus overcome the loneliness and isolation that afflicts many postgraduates.  You and your colleagues can help each other to be more effective and more satisfied in your thesis work.  You will learn much from each other.

This can be particularly valuable if you are doing your higher degree research by distance education.  Davies and his colleagues (2000) have described an approach where regular but infrequent meetings with supervisors are supplemented with regular and frequent learning set meetings.



Are you planning to do a higher degree by research? If so, action research may be a choice open to you.  In this chapter I have explored a number of choices that you may find useful to guide you through the design:

  • Will your study be theory-driven or data driven: will you take the literature or the research situation as your starting point?
  • Will you engage with the research primarily as technician or as performing artist: will you closely follow a recipe, or design and conduct your study to fit the situation and the theme?
  • Will this be 'action research' or 'action research': how action oriented will your study be?
  • Which of the many available methodologies will you use?
  • Who will you involve as participants in the study, and how, and to what extent?
  • What methods for data collection and analysis will you use to operationalise your methodology?
  • Facing a trade-off between rigour and relevance, what will you do?
  • Will you structure your thesis or dissertation conventionally, or will you design a structure that makes the most of your chosen methodology?

I have also identified some traps that most higher degree candidates face.  I have offered ways to avoid those traps, including forming a group of colleagues to act as an action learning set.

I hope that prospective research degree candidates will find this discussion useful, not just for understanding action research, but for making well and informed choices in their research program whatever approach they find most appropriate for conducting the study they choose.



Argyris, C.  (1999)  On organizational learning, second edition.  Oxford, UK: Blackwell.

Argyris, C., and Schön, D.A.  (1996)  Organisational learning II: theory, method and practice.  Addison-Wesley, Reading.

Argyris, C., Putnam, R., and McLain Smith, S.  (1985)  Action science: concepts, methods and skills for research and intervention.  Jossey-Bass, San Francisco, Ca.

Barbour, R.S., and Kitzinger, J.  (eds)  (1999)  Developing focus group research: politics, theory and practice.  Sage, London.

Baskerville, R., and Lee, A.S.  (1999)  Distinctions among different types of generalizing in information systems research.  In Ngwenyama L.  Introna, M.  Myers and J.I.  DeGross, New information technologies in organizational processes: field studies and theoretical reflections on the future of work.  Kluwer, New York, 49-65.

Bentz, V.M., and Shapiro, J.J.  (1998)  Mindful inquiry in social research.  Sage, Thousand Oaks.

Chamala, S., and Keith, K.  (eds)  (1995)  Participative approaches for landcare: perspectives, policies, programs.  Australian Academic Press, Brisbane.

Checkland, P.  (1981)  Systems thinking, systems practice.  Wiley, Chichester, UK.

Checkland, P., and Holwell, S.  (1998)  Information, systems, and information systems: making sense of the field.  Wiley, Chichester, UK.

Cherry, N.  (1999)  Action research: a pathway to action, knowledge and learning.  RMIT University Press, Melbourne.

Clark, P.  (1972)  Action research and organisational change.  Harper and Row, London.

Davies, A., Dick, B., Hase, S., Sankaran, S., Gloster, M., and Kwok, R.  (2000)  Problem based learning with academic accreditation: a flexible postgraduate program for managers and practitioners using action research at the workplace.  Paper presented at the HERDSA/ASET Flexible Learning Conference, University of Southern Queensland, Monday 3 June 2000.

Denzin, N.K., and Lincoln, Y.S.  (2000)  Handbook of qualitative research, second edition.  Sage, Thousand Oaks, Ca.

Dick, B.  (1993)  You want to do an action research thesis? -- How to conduct and report action research.  Chapel Hill, Qld.: Interchange (mimeo)  Available on line at

Edmunds, H.  (1999)  The focus group: research handbook.  NTC Business Books, Chicago, Ill.

Elliott, C.  (1999)  Locating the energy for change: an introduction to appreciative inquiry.  International Institute for Sustainable Development, Winnipeg, Canada.

Fisher, D., and Torbert, W.R.  (1995)  Personal and organizational transformations: the true challenge of continual quality improvement.  McGraw-Hill, London.

Flood, R.L.  (1997)  Action research and the management and systems sciences.  Systemic Practice and Action Research, 2(1)  Also available on line at

Forester, J.  (1999)  The deliberative practitioner: encouraging participatory planning processes.  MIT Press, Cambridge, Mass.

Ghaye, A.  and K.  (1998)  Teaching and learning through critical reflective practice.  David Fulton, London. 

Glaser, B.G.  (1992)  Basics of grounded theory analysis: emergence vs forcing.  Sociology Press, Mill Valley, Ca.

Glaser, B.G.  (1998)  Doing grounded theory: issues and discussions.  Sociology Press, Mill Valley, Ca.

Glesne, C.  (1999)  Becoming qualitative researchers: an introduction, second edition.  Longman, New York.

Greenwood, D., and Levin, M.  (1998)  Introduction to action research: social research for social change.  Thousand Oaks, Ca.: Sage.

Gregory, A.  (2000)  Problematizing participation: a critical review of approaches to participation in evaluation theory.  Evaluation, 6(2), 179-99.

Grundy, S.  (1983)  Three modes of action research.  Curriculum Perspectives, 2(3), 23-4.

Guba, E.G., and Lincoln, Y.S.  (1989)  Fourth generation evaluation.  Sage, Newbury Park, Ca.

Habermas, J.  (1984)  The theory of communicative action, volume 1: reason and the rationalisation of society.  Beacon Press, Boston.

Habermas, J.  (1987)  The theory of communicative action, volume 2: the critique of functionalist reason.  Polity Press, Oxford.

Heller, F.A., and Brown, A.  (1995)  Group Feedback Analysis applied to longitudinal monitoring of the decision making process.  Human Relations, 48(7), 815-36.

Heron, J.  (1999)  The complete facilitator's handbook.  Kogan Page, London.

Heron, J., and Reason, P.  (1997)  A participatory inquiry paradigm.  Qualitative Inquiry, 3(3), 274-94.

Hubbard, R.S., and Power, B.M.  (1999)  Living the questions: a guide for teacher-researchers.  Stenhouse Publishers, York, Maine.

Kaner, S., with Lind, J., Toldi, C., Fish, S.  and Berger, D.  (1996)  Facilitator's guide to participatory decision-making.  New Society Publishers, Gabriola Island, BC.

Kemmis, S., and McTaggart, R.  (1988)  The action research planner, third edition.  Deakin University, Geelong, Victoria.

Kirk, J., and Miller, M.L.  (1986)  Reliability and validity in qualitative research.  Sage, Beverly Hills, Ca.

Kolb, D.A.  (1984)  Experiential learning: experience as the source of learning and development. Prentice-Hall, Englewood Cliffs, NJ.

Krueger, R.A., and Casey, M.A.  (2000)  Focus groups: a practical guide for applied research, third edition.  Sage, Newbury Park.

Kuhn, T.S.  (1970)  The structure of scientific revolutions, 2nd edition.  University of Chicago Press, Chicago.

Kvale, S.  (1996)  InterViews: an introduction to qualitative research interviewing.  Sage, Thousand Oaks.

Lincoln, Y.S.  (1990)  The making of a constructivist: a remembrance of transformations past.  In Egon S.  Guba (ed.)  The paradigm dialog.  Sage, Newbury Park, Ca., 67-87.

Marquardt, M.J.  (1999)  Action learning in action: transforming problems and people for world-class organizational learning.  Palo Alto, Ca.: Davis-Black.

McNair, R.H.  (ed.)  (1998)  Research strategies for community practice.  New York: The Haworth Press.

Miles, M.B., and Huberman, A.M.  (1994)  Qualitative data analysis: an expanded sourcebook, second edition.  Sage, Thousand Oaks, Ca.

Mills, G.E.  (2000)  Action research: a guide for the teacher researcher.  Merrill, Upper Saddle River, NJ.

Moustakas, C.  (1990)  Heuristic research: design, methodology, and applications.  Sage, Newbury Park.

Patton, M.Q.  (1997)  Utilization-focussed evaluation, third edition.  Sage, Thousand Oaks, Ca.

Patton, M.Q.  (1990)  Qualitative evaluation and research methods, second edition.  Sage, Newbury Park, Ca.

Perry, C.  (2002)  Structure of a five chapter PhD thesis: candidates' notes.  Queensland University of Technology, Brisbane, unpublished paper.  Available on line at

Perry, C., and Zuber-Skerritt, O.  (1992)  Action research in graduate management research programs.  Higher Education, 23(2), 195-208.

Raelin, J.A.  (2000)  Work-based learning: the new frontier of management development.  Prentice-Hall, Upper Saddle, NJ.

Revans, R.W.  (1983)  The ABC of action learning: empowering managers to act and to learn from action, third edition.  Lemos & Crane, London.

Rothman, J.  (1997)  Resolving identity-based conflict in nations, organisations, and communities.  Jossey-Bass, San Francisco, Ca.

Rothwell, W.J.  (1999)  The action learning guidebook: a real-time strategy for problem solving training design and employee development.  Jossey-Bass/Pfeiffer, San Francisco, Ca.

Rothwell, W.J., Sullivan, R., and McLean, G.N.  (1995)  Chapter 2: models for change and steps in action research.  In William J.  Rothwell, Roland Sullivan, and Gary N.  McLean (eds)  Practising organization development: a guide for consultants.  Jossey-Bass/Pfeiffer, 47-72.

Sagor, R.  (2000)  Guiding school improvement with action research.  Alexandria, Va.: Association for Supervision and Curriculum Development.

Sarkissian, W., and Perlgut, D., (eds)  (1994)  The community participation handbook: resources for public involvement in the planning process, second edition.  Institute for Science and Technology Policy, Murdoch University, Perth.

Smith, S.E.  and Willms, D.G., with Johnson, N.A.  (eds)  (1997)  Nurtured by knowledge: learning to do participatory action-research.  Apex, New York.

Strauss, A., and Corbin, J.  (eds)  (1997)  Grounded theory in practice.  Sage, Thousand Oaks, Ca.

Strauss, A., and Corbin, J.  (1998)  Basics of qualitative research: techniques and procedures for developing grounded theory, second edition.  Sage, Newbury Park.

Taylor, S.J., and Bogdan, R.  (1998)  Introduction to qualitative research methods: a guidebook and resource, third edition.  Wiley, New York.

Toulmin, S., and Gustavsen, B.  (1996)  Beyond theory: changing organisations through participation.  Benjamins, Amsterdam.

Wadsworth, Y.  (1984)  Do it yourself social research.  Victorian Council of Social Service and Melbourne Family Care Association, Melbourne.

Whitehead, J.  (1993)  The growth of educational knowledge: creating your own living educational theories.  Hyde, Bournemouth.

Whyte, W.F.  (1991)  Participatory action research.  Sage, Newbury Park.

Zuber-Skerritt, O.  (1996)  New directions in action research.  Falmer, London.



Copyright (c)  Bob Dick 2000.  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:

Dick, Bob.  (2000)  Postgraduate programs using action research [On line].  Available at



navbar 4 

Maintained by
Bob Dick; this version 1.02 last revised 20110508