navbar 4Resource papers in action research

Action research and publication


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 some issues related to the use of action research for publication and theses are examined


In this paper:

This session examines action research done for the purposes of a thesis or for publication.

If you examine the action research which is done, you will find that some of it pursues change as its first priority.  Other action research is done for publication or qualification..  In other words, there is "ACTION research", and there is "action RESEARCH".  It is the second of these which will be our main concern here.

The theme of this document is rigour -- the conduct of research in such a way that you can have confidence in your data, and your interpretations of them.  Rigour is important both to action and research; but it is to research that it relates most directly.

Much of what I say here is relevant both to theses and to publication.  Of the two, the requirements for theses are the most demanding.  The focus, therefore, is on action research for theses.

For thesis purposes, rigour can be crucial.  Most examiners, I think, weight research outcomes much more highly than action outcomes.  Most of them, in my experience, are much more critical of how the research is done.  They may be less interested in how the action was pursued.

It seems to me that three features, in particular, distinguish thesis research (and some research which is primarily for publication) from other action research:

  • As I have said, the research component is usually more heavily weighted.  The methodology, therefore, must often be carefully designed.  Most importantly, it must be carefully explained and justified. 
  • In most instances the research is judged primarily, or entirely, on the written record.  The final write-up is therefore important.  Your ongoing documentation can contribute to that. 
  • A thesis is expected to be a contribution to knowledge which is the individual and original achievement of the person writing it.  You are expected to identify a contribution which is demonstrably your independent achievement.  This can be hard if a study is highly participative, and especially if the participants are involved as co-researchers.

This document looks closely at the process of documenting action research for theses and for publication.  And it examines how you can demonstrate the individual and original contribution required for theses.

In addition, it seems to me that many action research studies are larger than they need to be for thesis or publication.  If you write up everything this may not be a very economical (or interesting) way to report what you have found.  The value of action research can be increased by styles of reporting which are economical while being conceptually and methodologically sound.

You also face important choices about how you use the literature.  This, too, is briefly addressed.


Justifying the methodology

Your study is vulnerable to criticism if you don't adequately argue for your approach.  What you do is often less important that how well you justify it.

For ease of explanation, I'll distinguish between three levels of method:

Paradigm.  By paradigm I mean the overall approach.  For example, I regards the family of experimental and quasi-experimental methods as one paradigm.  Ethnographic approaches I regard as another.  Action research is a third.

Methodology.  As I use the term here, methodology is the particular sub-family used within a paradigm.  For instance, there are varieties of action research: action science and soft systems methodology are two examples.  (In the present terminology, you might say that soft systems methodology is well named.)

Methods.  These are the fine grain processes you use for data collection and the like.  Interviewing is a method, for instance (or several, if you consider different varieties of interview).  Focus groups constitute another.  Document analysis is a third.

For safe thesis writing, it is better to justify your choice of paradigm, methodology and methods.

I suspect it is also safer to choose a methodology which is reasonably acceptable in the academic circles you move in.  What counts as safe may depend on the discipline.  In some disciplines, for instance, qualitative approaches are viewed with suspicion.  In others they may actually be more in favour than quantitative.

Psychology, for example, strongly favours quantitative experimental approaches.  So it is often safer to write up a study as "evaluation" than as "action research".  Evaluation is more respectable, even when this is more a matter of labels than of substance.  (Many academic psychologists would be surprised at how qualitative, and how flexible, much evaluation has become.  But that's another story.)

It is more an issue for theses than for publication.  If you have written a paper, you can choose the publisher or journal to approach.  You may have much less a choice when it comes to thesis examiners. 

It is also easier if your chosen methodology comes with its justification already provided.

For example, Peter Checkland 1 has provided a detailed justification of his soft systems methodology.  To help further, the justification is provided in very rational terms.  Checkland speaks academese well.

In addition, Checkland's approach has been critically but sympathetically reviewed by other well-published academics; 2 and there are published studies using it.  3 This increases the safety.

All of this makes it difficult for an examiner to penalise you for its use, even if she or he has personal doubts about it.  In a sense, it wasn't you who did it.  It was Peter Checkland.  4

It may further enhance your thesis if you can make some small alterations to the methodology to suit the situation.

For instance, you might use soft systems methodology as an evaluation process, as some postgraduates in my programs have done.  (It is more often used for diagnosis and problem solving).  Or you might include some other concepts in addition to systems concepts at the model-building stage.

These modifications may constitute an important contribution to knowledge.  You would need to offer good reason for the changes.  But there is usually some need to tailor your methods to the situation.  It's worth doing: methodological contributions appear to have high status with many examiners, and to be publishable.



The important documentation comes at the end, of course.  That's when you write it up.  And that is what the examiners or reviewers read.  The ease of writing up, however, also depends on the documentation you develop during the study.

How much do you document?  Enough that you can reconstruct, at the end, what you did from the beginning.  I think, though, that some economy in documentation can make lighter what can otherwise be a heavy load.

I've argued that one of the features of action research is that you can combine data collection and interpretation from the first action research cycle.  It seems a pity to go to the effort of documenting all the data.  That doesn't make the most of the specific advantages of action research.  What you really need, it seems to me, are the interpretations. 

Your contribution to knowledge is likely to arise from the changes you introduce.  It makes sense, then, to collect the data which bear most directly on that.

In more detail, then, your documentation might include these items: 5

Before the event:

  • the outcomes you hope to achieve in this next cycle, and why you think they are worth pursuing
  • the contribution you expect those outcomes to make to your long-term goals, and why you expect it
  • the actions you plan to take to achieve those outcomes, and why you think those actions will achieve those outcomes in that situation.

Then, after the event:

  • what actions you carried out, and what outcomes you achieved
  • how and why these differed (if they did) from what you expected
  • what you learned about the client system, your methodology, yourself, and so on.

The final point is the heart of the research component of action research.  It represents your growing understanding.  At the end, it is probably from this learning that your contribution to knowledge will arise.

The "why" questions, those you ask yourself before the action, serve an important function.  They help you identify your expectations and assumptions: about the situation you face, the outcomes you hope for, and the actions you intend to take to achieve them.  Comparing plans to reality then helps you identify which of those assumptions and expectations were incorrect -- and that is something learned.

The final write-up, I think, can be structured around your contribution to knowledge.  You may have to argue for doing it, as it may seem unusual to some examiners.  But when you think about it, that's what the thesis is about. 

For that matter, a conventional empirical thesis is said to be organised around a research question.  Such a thesis documents the question, your approach to answering it, and the conclusions you reached about the answer.  In other words, you could as easily say that a conventional thesis is organised around the contribution to knowledge.

I've suggested a possible structure elsewhere (see the archived file "phd", 6 for example).  Here it is, briefly and with some omissions:

  • Introductory chapters in which you
    • provide the necessary context (briefly!)
    • justify the study
    • justify your choice of action research and your particular methodology
    • in very broad brush style, describe the overall process of the study (this is not a detailed history -- I have read too many excessively long blow-by-blow chronologies to have a very good opinion of them). 
  • Major chapters, each of which deals with a particular conclusions or cluster of conclusions, the evidence supporting them, the relevant literature, and the pains you took to try to disconfirm your conclusions. 
  • A "catch-all" chapter, perhaps, to mention some of the minor contributions to knowledge. 
  • Perhaps a final chapter in which you pick up some of the vulnerabilities in the study, talk of the implications, suggest further research, and do all the other things that typically fill the end pages of a thesis.

I have heard some people argue for writing up an action research thesis so that it resembles a conventional thesis as far as possible.  I think I understand the reasons for this.  But I believe that it doesn't do justice to action research.  I think the special advantages of action research are best expressed within a less conventional thesis structure.


Original contribution

My own thinking about this issue owes much to a paper by Chad Perry and Ortrun Zuber-Skerritt.  7 They take a Deakin cycle (plan -> act -> observe -> reflect) as the basis for the participative element of the study.  Then they suggest treating all of this as the "act" component of an individual reflection cycle.

In other words, each cycle with the client group nests as the action component inside a cycle of personal reflection.

For example...

Suppose you are working with a group of people to plan and implement a particular change.  You meet with them for regular planning.  You disperse to carry out what you have planned.  You meet again to compare notes on what actually happened, and with what results.

In the reflection after action, the group as a whole examines it's methodology, the stakeholders involved, the plans, the goals being pursued, and anything else that is relevant.

After the rest of the team have gone, the researcher reflects on the study.  She examines how the team planned, acted, observed, and reflected.  She examines, too, her own role and her own learning in all of this.  She plans what she will do next.

It is this individual cycle that provides most of the material for the thesis.



I am often asked, "How much should I read before I begin the study?".

Nowadays, I respond by saying that there is some reading that must be done at the outset.  And there is some reading which can only be done later.  And for some reading, you have a choice.

In the beginning, I would usually encourage you to read enough of the methodological literature that you are sure you can later justify your methodology.  You can't afford to arrive at the end and discover that your methodology is fatally flawed.

On occasion you may start without a research question.  You may have only a client group.  At this point, then, you may have no idea what content literature will prove relevant.

More often, though, you have at least a rough research question, or set of questions.  There is almost certainly some content literature which you might usefully read.

For instance, you may be using action research to help set up a community centre.  You would be advised to read the methodological literature on action research, with some emphasis on research in community settings.  I would probably suggest also that you read some general literature on community centres.

An action research project often rapidly develops a life of its own.  It takes you in directions that you didn't predict, and perhaps couldn't have predicted.  For each of these directions, there is likely to be relevant literature.  Some of this can only be read later.

In the example, you might find that existing community networks mostly do not cross any arterial roads which divide the community.  There is probably some literature on this.  You would check your finding against that literature. 

Between these extremes there are choices.  At the start you can deliberately delve into related literatures which seem likely to be relevant.  Or you may instead deliberately limit your reading of the content literature to what is clearly and directly relevant, and avoid the rest.

My own inclination is to read as little as possible at the start.  I can then use the literature to check how generalisable my findings are.  8

In fact, I prefer to do this even with the methodological literature.  I develop a design.  I may even start to use it.  Then I go to the literature to refine and challenge my own ideas and experience.

For thesis purposes, this may be a bit risky.

At the end of it all, you want to be able to offer a contribution to knowledge.  You want some assurance that you can't easily be challenged.  You want to be able to say: "I developed an interpretation, X.  It survived my vigorous attempts to collect disconfirming data and disconfirming literature".



  1. Checkland, P (1981) Systems thinking, systems practice.  Chichester: Wiley.back ]
  2. For example Flood, R.L., and Jackson, M.C.  (1991) Creative problem solving: total systems intervention.  Chichester: Wiley.back ]
  3. For example Davies, L.  and Ledington, P.  (1991) Information in action: soft systems methodology.  Basingstoke, Hampshire: Macmillan.back ]
  4. There is an alternative.  Although it is probably a little riskier, I think it is better research.  Design your study to fit your research questions and your research situation, from first principles, as it were.  Don't worry if your design doesn't look much like any of the established approaches.back ]
  5. I've revised this section of the document since the last revision because of some conversations with Alan Davies, Stephanie Chee, Richard Kwok, Leong Chun Chong, John Man, and Shankar Sankaran.I might mention in passing that I've used this as the basis for learning diaries, and for documenting people's reflection during experiential learning activities.  Here, too, it seems to work well.Part of the advantage of thinking in "action --> outcomes" terms is that you develop contingent principles which you can more easily change than absolute principles such as "Do this" or "Don't do that".back ]
  6. URL ]
  7. Perry, C.  and Zuber-Skerritt, O.  (1992) Action research in graduate management research programs.  Higher Education, 23(2), 195-208.back ]
  8. Glaser, one of the developers of grounded theory, recommends research first and literature later.  See Glaser, B.G.  (1978) Theoretical sensitivity: advances in the methodology of grounded theory.  Mill Valley, Ca: Sociology Press.back ]


Other archived resources

A file, "phd", is archived on the arlist archive.  It describes in a little more detail how an action research thesis might be approached.  The URLs are:

Other reading

There is a large literature on writing in general, and quite a bit on theses in particular.  Much of this addresses thesis writing generally.  You may find this book useful:

Phillips, E.M.  and Pugh, D.S.  (1987) How to get a Ph.  D.: A handbook for students and their supervisors.  Milton Keynes: Open University Press.

A book edited by Ortrun Zuber-Skerritt falls in this category, but has the advantage that she supervises action research and action learning PhDs.

Zuber-Skerritt, O., ed.  (1992) Starting research -- supervision and training.  Brisbane: Tertiary Education Institute, The University of Queensland.

I've written a monograph (still in draft form) specifically on action research theses.  It will appear eventually as a Tertiary Education Institute, University of Queensland, document.  Its present title is:

Dick, B.  You want to do an action research thesis.  Mimeo.

(Its eventual title may be more staid.) You'll also find it on the web at

 For a quite different view you might also want to look at Chad Perry's paper on action research theses.  You'll find it at



This document has examined action research for theses or publication.  It has considered, in particular, issues of documenting the progress of the study, using the literature, and writing the final report.


Copyright (c) Bob Dick 1995-2000.  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, B.  (1999) Action research and publication [On line].  Available at




navbar 4 

Maintained by
Bob Dick; this version 2.04w last revised 20000105

A text version is available at URL