Action research and evaluation on line

Session 2: The change process and
action research



This is Session 2 of areol, action research and evaluation on line, a 14-week public course offered as a public service by Southern Cross University and the Southern Cross Institute of Action Research (SCIAR). which the change process, and the role of action research, are briefly described.The focus of this on-line program is identified as being processes for pursuing change and understanding at the same time


Consider your project, if you have one.  Or think of some other activity which is intended to lead to both improvement and learning.  If you knew nothing about any form of research, how might you approach it? You might begin by thinking of it as a change process, and devise ways of improving the learning.  Or you might think of it as a research process, and look for ways of building more effective change into it


In this session:


Why action research?

As I've said in the earlier material action research is designed for a dual purpose: to yield change (the "action") and understanding (the "research") at the same time.  That is a key strength.  It is a strength which depends upon the flexibility and responsiveness.  In turn, they derive in part from its cyclic process.

You can view action research as a family of processes for combined action and research.  Consider this...

Change often requires flexibility and participation.

Research depends upon high quality data and accurate interpretations.

Often it's useful to be able to meet both these conditions.  Action research is designed to allow you to do so.


Please note that I'm not arguing for action research as "the best" research methodology.  Nor am I arguing for this version of action research over different versions.  Different methods of research or action research achieve different purposes.

For example, sometimes you may wish to identify causal relationships between relatively limited numbers of variables.  Then my recommendation is that you first examine experimental methods.  Or, if you have to take the study into the field, use those quasi-experimental methods which are robust.  (In my view many of them are not.)

Or sometimes you may wish to understand a complex culture on its own terms.  It may be your wish to have as little impact on the culture as possible.  I'd then expect your first choice, more often than not, to be ethnography or related methods.

(I'm not saying that the various research paradigms are limited to these uses.  Ethnographic methods and quasi-experimental approaches, among others, have been used effectively in change situations.)

Action research has a number of uses quite apart from its applicability to change programs.  It's strength, I think, is its emergent nature.  It takes form slowly, informed by the growing understanding of those using it.

This is what allows it to be sufficiently flexible and responsive that it can be used in change programs.  The same quality often allows it to be effective for pilot research, for example.  In some forms, it can deal well with very complex situations.  There, the nature of the situation cannot be identified ahead of time and its emergent nature is valuable.

As my friend and colleague Pam Swepson would say, following John Dewey, fitness of function is the appropriate criterion for choosing a research paradigm.

(By the way, I would rather avoid any debate about the mixing of paradigms.  Personally, I think it's fine if you know what you are doing.  Some disagree.)

What do you want to achieve? Which paradigm will best achieve it? For me, those are the important questions in choosing an approach.

As it is, the boundaries between one research paradigm and another are often unclear.  If researchers were to choose all aspects of their research more carefully, this would be even more so.  There are many dimensions on which a research design can differ.

In my view, the best research involves making informed choices on each of these dimensions.  The aim, I suggest, is not to do "action research".  It is to do effective research.  And effectiveness, I presume, is defined by situation and desired outcomes.

The "best" action research may well borrow processes and approaches from other paradigms, if that is what the situation and the desired outcomes require.  Many action researchers I know are pragmatists, at least when they are in the field.  If it works in practice, they use it.


The process of change

As I've said, other research paradigms can be used in change situations.  Action research, however, is often an appropriate choice.  After all, it is intended to lead to change.

As you might expect, some aspects of a good action research are directed mostly towards achieving change.  Other aspects try to assure the quality of data and interpretations.  For the moment, consider the requirements for change...

There is a large literature (I mention a little of it below).  Much of it is written from the viewpoint of a consultant or facilitator -- a person who helps others achieve the change they desire.  This seems akin to some of what an action researcher has to do.

The overall shape of a change process is often described as a number of distinct stages.  For example:

  • entry and contracting; here you enter the client system and negotiate your role and theirs;
  • diagnosis; here you (and they) decide what requires remedy or change;
  • intervention; here the remedy is put in place;
  • closure; here you withdraw from the system.

(It's common for "evaluation" to be inserted between intervention and closure.  It is a bias of mine that evaluation is better done continuously.  Evaluation -- in the form of critical reflection -- is built into each action research cycle.  I think that that is part of the reason why action research can be effective for its dual purposes.)

I think that this four-stage model is a useful simplification.  Or rather, I think it's useful if you remember it's a simplification. In fact, the stages overlap greatly.  There is often a need to retrace your steps to an earlier stage.

Entry and contracting, for example...  It seems to me that this is needed throughout many change programs.  Each time a new client or informant or participant is identified, there is a new relationship to be formed; there are new roles to be negotiated. For that matter, I have yet to encounter an important relationship (in consulting or elsewhere) that didn't require some renegotiation from time to time.

I think of it like this.  Ultimately, when it comes to facilitating change, all you have to work with is yourself.  You are the only person you can influence directly and predictably.  However, if you change your own behaviour, you will in turn change others.  It may not be in predictable ways; but it will usually occur.

The quality of your relationship with others is important.  It will help them to decide whether to trust you and your processes.  Your processes enable them to develop a better understanding of their situation, and make better use of the understanding they already have.


I have heard an interesting story about Apollo 13.  I don't know how true it is; but I can use it to illustrate the point.

Apollo 13 was the moon expedition that almost ended in disaster.  There was an explosion in a computer panel.  The astronauts lost control of the module: they were faced with circling the earth, as a satellite, forever.  The NASA engineers could not solve the problem.

According to the story, a problem solving consultant was called in.  The consultant knew nothing of the relevant computer technology, or space navigation, or whatever other knowledge was required.  But the consultant's process allowed the engineers to pool their knowledge and their experience.

It is true that the problem was solved.  The astronauts returned safely to earth.

In a sense, as action researchers we are often in the position of the consultant (real or imagined) in the Apollo 13 instance.  We have processes that help people identify problems more effectively, set better goals, develop better plans.  All of these also depend upon better diagnosis (one of the practitioner's labels for research).

And often, the consultant helps the clients to plan action.  The action often happens later.

Planning, in turn, can be done most effectively when people have some prior understanding.  It's a bit like planning a travel route -- and in some respects that is what change is, a journey.  You first have to know where you are, and where you want to go.  It also helps to know something of the terrain between here and there, though you may need to be a bit flexible about this.


So, here is a slightly different version of the change process.

Phase A: pre-planning: identify the stakeholders and set up your personal and working relationships with them

1. build relationships

2. negotiate roles and responsibilities

3. build a climate for change

4. agree on mechanisms for participation

Phase B: planning: help people determine what needs to be done

5. goal setting or visioning (or both)

6. situation analysis or problem-solving

7. action planning (including planning for ongoing evaluation

Phase C: action: implementing the planned changes, checking that they are effective, and modifying them as needed

8. action and on-going monitoring


In other words...  First you build relationships, with clear roles. Then you analyse the situation and decide what action to take. Then you act.

And again, I wouldn't want you to think that it's as simple as that. In practice it isn't a linear sequence.  But it will serve our purpose for the present.  We can develop it further in later sessions.

Now, how can change be pursued so that understanding is also achieved?


Action and research

With the foregoing as background we can begin explore how the two elements of action and research can be combined.


You can usually assume that people are more committed to their own decisions than to other people's.  For effective action, then, the appropriate participants include those who can enable or prevent the desired change.  All else being equal, in most situations these are the people who will be affected by the change, and those who will have to implement it.

(A later session describes a simple procedure which can help to plan your approach to these.  It is called "stakeholder analysis".)

For effective research, you have to have some way of accessing those who have the information necessary to understand the system.

So, change is more likely to occur if certain people are involved.  And you are likely to collect better or more complete data if certain other people are involved.

It tends to be true that the two groups overlap.  Those most affected by the change are often almost the same as those who best understand the system to be changed.  (I don't recommend that you assume this is always true.)

In organisational settings, for example, it is often the people at the workface who are asked to change.  And they are the people who often best understand their work, and their situation.

I would offer a rough generalisation: the more diverse the people you involve, the more likely you are to collect enough of the relevant information.

You may also choose to involve people for other reasons: values, for instance.  As an example, you may place a high value on participation in its own right.  You may therefore choose to involve people who are not needed either for effective change, or effective understanding.

We'll visit this important topic again.


For research outcomes, the most effective methodology is the one which generates appropriate data and interpretations.  For change, you want a methodology which generates commitment to the change.

It is here that the cyclic nature of action research plays an important part.  Consider what happens within each cycle, and also from cycle to cycle...

I've mentioned in earlier sessions two varieties of action research cycle.  The Deakin cycle consists of

plan --> act --> observe --> reflect

The important feature here is alternating action and reflection in an ongoing cycle.  It could be simplified to

action --> reflection

It is theory and practice integrated.  Action and understanding integrated.

In the Deakin cycle, planning can be thought of as reflection before action.  Observation is reflection during action.  Reflection is reflection after action. 


I also mentioned, as a cycle that captures the progression from cycle to cycle

fuzzy questions --> fuzzy methods --> fuzzy answers

This leads in successive cycles to less fuzzy questions, methods and answers.  The later cycles test and refine the conclusions, and the actions, from earlier cycles.  The continuing testing of theory in action adds to understanding.

All else being equal, the better the information and interpretation, the better the decisions.  The better the decisions, the better the decided actions.  But this is true only if the processes generate accurate data, good under-standing, good planning, and commitment to the decisions.


And what about philosophy?

I have been asked what role philosophy plans in all this.  My own answer (many would disagree) is, not very much.  However, there are two relevant files in the archive: "philos" and "naive", mentioned below.



  1. My colleague Alf Lizzio calls these three phases preflection (before action), flection (during action), and reflection (after action).

 [ back ]

Archived resources

For some of the reasons for choosing action research as your paradigm, see "rigour" in the arlist archive.

On change, see "change" in the same archive.  If you sometimes wonder if your research is action research, see "choice".

On philosophy, see "philos" and "naive".

If you are using anonymous ftp the URLs are:

Web versions are available at:


Other reading

There is an extensive literature on the practicalities of change.  For organisational change, the literature on organisational development is relevant.  A good starting point, written from an explicit action research perspective, is:

French, W., and Bell, C.H.  (1995) Organisation development: behavioural science interventions for organisational improvement, fifth edition.  Englewood Cliffs, NJ: Prentice-Hall.

For change in community settings, you might try this book of readings:

Cox, F.M., Erlich, J.L., Rothman, J., and Tropman, J.E., eds.  (1987) Strategies for community organisation: a book of readings, fourth edition.  Itasca: Peacock.

Because of the careful attention it gives to the "people" side of action research, I am impressed by:

Oja, S.N.  and Smulyan, L.  (1989) Collaborative action research: a developmental approach.  London: Falmer Press.



A thought experiment

Identify some past changes (any sort of changes) of which you approved, and some you resisted.  Identify the similarities and differences.  What principles of change would you draw from this?

Bonus questions for the enthusiasts: How well integrated were the action and research components? How involved were you in researching, planning and implementing the change?


An individual activity

During the next several days, act more intentionally.  Try to be aware, before acting, of the outcomes you hope your action will achieve.  Note how often you achieve them.  Note how you behave when you don't achieve them.


For your learning group

You may have already given some attention to getting to know one another in your learning groups.  And you may have begun to exchange information about your projects.  If not, these are useful early priorities.

Recall that your task is to help one another to learn from your experience, with your projects and elsewhere.

Help each other to consider your projects.  Be clear about the desired outcomes.  Begin to plan ways of achieving a balance of action and research through participation and process.


Let's practice action research on areol.  What ideas do you have for improving this session? What didn't you understand? What examples and resources can you provide from your own experience? What else? If you are subscribed to the email version, send your comments to the discussion list.  Otherwise, send them to Bob Dick


In summary...

In this session I've suggested that action research is most suited to those occasions where action and research are both desired.  Action research is action and research.  Change and understanding.

(I've also mentioned that as an emergent methodology it can be used in uncertain or complex settings.)

The action is achieved by applying the principles of change -- most often participative change.

In planning and conducting action research, change and understanding can be achieved together.  Further, each can assist the other. Planning for both can lead to better decisions about participation, and can make good use of the cyclic nature of action research.

Next sessions: participation and ways of achieving it.  See you then -- Bob


Copyright © Bob Dick 2002.  May be copied provided it is not included in material sold at a profit, and this notice is shown.

This document may be cited as follows:

Dick, B.  (2002) The change process and action research.  Session 2 of Areol - action research and evaluation on line.





Maintained by Bob Dick; this version 11.02w; last revised 20020712

A text version of this file is available at