navbar 4Resource papers in action research

Participative processes


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 different ways of involving participants in action research are discussed




The general view of most current writers is that action research is necessarily participative.  1,2,3 Ultimately, that is a question of definition.  It depends upon where you decide to draw the boundary around action research.  Action research is whatever we define it to be.  However, I think there may be advantages at least in considering a different view about participation.  I propose that we regard the issue of participation as offering the researcher, and perhaps the participant, a choice about participation levels.  4

I think it is evident that participation, in any event, is not an all or none matter.  It can vary along a number of dimensions.  On each of those dimensions it can be absent, or substantial, or any point in between.

At any point there may be further choices about how the participation will be achieved.  Sometimes the choice is for partial participation through representation.  Further choices then arise about how the wider body of people will be kept informed.

The choices can depend on the desired outcomes, and the approach to participation which will best achieve them.  Those outcomes are likely to include both action and research, though their relative priority may vary.  As well, there may be outcomes which are decided by values.  For example, I think it is appropriate for a researcher to decide that participation has a value in its own right.  Sometimes this value may be high:  high enough that the issue is not negotiable.

Yet even here there are choices to be made.  Who shall participate?  In what?  To what extent?  For that matter, how much choice do they have in participating?


Dimensions of participation

There are a variety of dimensions of participation.  For present purposes I wish to distinguish seven.  Four of them relate to the content of the situation:

  • providing data;  the participants are informants; 
  • interpreting data;  the participants are interpreters; 
  • planning change;  the participants are planners and decision-makers; 
  • implementation;  the participants are implementers. 

Another two are part of the research process:

  • managing the process of data collection and interpretation;  the participants are facilitators; 
  • designing the overall study;  the participants are researchers or co-researchers. 

The seventh may be about process, or content, or both:

  • being kept informed about the study and its implications;  the participants are recipients only. 

On each of these dimensions, there is a choice:

  • who is to participate? 
  • to what extent are they to participate? 

To develop this further, the first of these questions turns out to have two parts.

First, there is likely to be some body of people who are able to contribute.  For example, some people are equipped to provide data because they have relevant information.

Second, it may not be feasible to involve all potential participants.  You may have to choose some smaller number of these people to involve.  In other words, when full participation is not possible, you may choose representation instead.  A smaller number of participants are chosen to speak in some way for the rest.

It may also be that different people will be involved to different extents.  For example, you might interview a small number of people, on several occasions, in depth.  Others might be interviewed only once.  Yet others may respond briefly to a written survey or the like.  The information from the interviews may be interpreted by a working party or reference group which is broadly similar to the community of stakeholders.


Participants as informants

In most studies, at least some participants are likely to be involved as informants.  This is minimal participation (though participating as recipients is even less).  It may occasionally be possible to obtain data through observation, or from documents.  But in change programs, this is unusual.

To achieve research outcomes, this is often a very important form of participation.  The selection of participants often proceeds in three steps.

The first step is to identify those people who are likely to have information (or an attitude or opinion) about the situation being researched.

The second step is required when it is not possible to involve everyone as informants.  Your task then is to compile a sample which will canvass all views, and be small enough to be managed within the constraints of time and budget. 

In general, your best option is to aim for a "maximum diversity" sample.  5  Sometimes you know enough about the various people that you can achieve a diverse sample almost from the beginning.  You may then get good results by collecting information until new information ceases to emerge.  As with action research generally, the emerging information determines the sample size.

The third step is to decide what information is needed, and how it can best be gained.  Again, action research methods allow you the luxury of being able to change your mind about this in the light of your early experience.

The criterion for selection is:  Who has the information needed to define the situation?

This is not as simple as it sounds.  Obviously, you will usually include members of the system being studied.  Depending on the desired outcomes, you may well want to include outsiders with special knowledge.


Participants as interpreters

A common form of this is where participants first contribute information, and are then asked to comment on its meaning.  In other words, each participant becomes her 6 own interpreter.

In the process for focus groups that I have described elsewhere, a phase of information collection is followed by a phase of interpretation.  7  Many processes for community or organisation development also involve participants in interpreting their own information.

On other occasions, different participants are used as informants and interpreters.  Interpretation may be done, for example, by a small working group which takes overall responsibility for the study.  It is common for a group (sometimes know as a "reference group") to be set up with the express purpose of helping to interpret the information provided.  It may have other duties too.

Other variations are possible.  In some forms of unstructured interviewing, for example, I may ask later informants to comment on the information arising from earlier informants.

The criterion for selection of interpreters is:  who has an adequate knowledge of the situation and the language of the participants?


Participants as planners

There are many techniques used in fields such as community development and organisation development.  Many practitioners who work directly with a clientele have specific procedures for working with participants.  Such techniques and procedures very often involve the same people in providing information, interpreting it, and turning that interpretation into action plans.

For example, consider those forms of team building which work by improving role relationships.  The members of the team first define their roles as they currently exist.  They may then identify the problems and opportunities for improvement.  Finally, they decide how they will change their roles to better manage their interdependencies.

Processes such as search (a technique for developing a shared vision) are more informative, in some respects.  In team building, participants too easily resolve their issues by creating difficulties for those who are not present to speak for themselves.  So techniques such as search may take more care to identify all groups who have some information to give, or some stake in the outcomes.  In other respects, they are similar to team building.  They, too, typically move from information exchange to interpretation to planning.

The point is this...  Techniques such as role negotiation and search (and many others) generate and interpret information.  For the most part they are used primarily to bring about change.  They can also be pressed into effective service as tools for action research.  When used for action research, more care is usually taken to validate the information and interpretation.  They then also become more effective tools for change.

The selection criterion for planners is:  Who is best placed to convert the understanding of the situation into plans for change?  All else being equal, my own usual answer is:  those who are going to be affected by the change. 

Sometimes, as before, you may decide to include outsiders with special knowledge.  You may then also have to decide if they are to be full participants, or advisors to those more directly affected.  Much of the time you can use them effectively by allowing them to offer information, but leaving the planning to the stakeholders.  8  


Participants as implementers

Finally, plans are implemented.  Here the choice is usually relatively easy.  The simple criterion is:  who is in a position to implement the plans?  A more effective criterion is:  who is in a position to implement the plans, and is motivated to do so?

So far, the participants have been involved only in the content of the study.  In addition, you may decide to do what you can to share the research responsibilities -- to make co-researchers of them.


Participants as facilitators

Some common applications are to use local informants as interviewers, or facilitators for some of the group-based methods of information collection and interpretation.  Some processes are robust enough that it is not difficult for members of the client group to facilitate them.  Structured interviews are an example. 

Local informants are often more acceptable to the members of the client group than you are.  They are often able to judge when a process needs fine-tuning to suit the local customs.  Language or dialect may make facilitation difficult for you.  9

The criterion for selecting facilitators is:  Who has the skills and motivation to manage parts of the action research process?

Sometimes you may decide to help participants develop the necessary skills before proceeding.  There are costs in time and effort;  these may be justified by the benefits for the participants.  So it may be part of your intention to help participants to do their own action research in the future.  This may fit in well with their intentions, too.  Action research methods are often a useful addition to the skills they use in the normal course of their work, or whatever it is that they do.


Participants as researchers

It is possible for those involved as participants to be involved also in designing and maintaining the research process. 

As with the preceding category, this may be done so that participants learn to manage their own research.  On other occasions it may come about because a group of people wish to improve their practice, and choose action research as the vehicle.  10  Or again, it may be done for reasons of "emancipation" -- helping participants understand their situation in ways that allow them to take more control of it.  11

The criterion for selecting researchers is similar to the previous category:  Who has the skills and motivation to manage parts of the action research process?  As before, you may often choose to help participants develop the necessary skills before proceeding.


Participants as recipients

Finally, in many instances, you will decide not to involve all of those who might have an interest in what you are doing.  Numbers may be too great.  Often, time will be short and money scarce.  You will then have the task of keeping the wider group of potential participants informed about what is happening.

In fact, the smaller the group of people you involve actively, the more effort it requires to provide some involvement for the rest.  Which brings me to the final point in this document.

Cutting across the above dimensions are two more:  participation (direct involvement);  and representation (involvement mainly through others).  If you want action as well as research it is often an advantage to involve everyone:

  • Participation:  everyone is involved. 
  • Representation:  involvement is offered to a sample of potential participants. 

If you choose the second option, I believe there are two other things to do for best results.  First, encourage those directly involved to act for all of those who have a stake in the issue, not as representatives.  Second, give attention to communication in both directions between your sample, and the others.


Let me sum up...

With some simplification, I have suggested that there are seven dimensions of participation.  Participants may be:  informants;  interpreters;  planners;  implementers;  facilitators;  co-researchers;  or merely recipients of information.  On each of these dimensions, a range of involvement is possible.  This can involve all, or some, of the stakeholders.  For each, you may involve the same, or different, stakeholders.

Sometimes, less than full participation is attempted.  I have argued that it is then important to consider carefully the role of those less directly involved.  They can contribute information, and be informed about what is happening.

As I said at the beginning, for me participation isn't all or none.  It is a choice.  Or, rather, it is a large list of choices.



  1. I have used footnotes rather more extensively than is currently fashionable outside disciplines such as law or philosophy.  There are issues which are too complex to treat well, but I have preferred not to clutter the text too much.back ]
  2. See, for example, Kemmis, S.  and McTaggart, R., eds.  (1988) The action research planner, third edition.  Victoria:  Deakin University.  It is true that, from the beginning, action research was usually participative.  This was because it was directed towards producing social change, and participation was believed to be more effective (as it usually is, in my view).back ]
  3. I incline increasingly to the view that many of the categorisations of research paradigms are somewhat arbitrary.  There are many dimensions along which a research design might differ.  Researchers might design each study by examining each dimension.  They might make their choice on that dimension as the best choice to achieve their research outcomes, taking into account the situation, the stakeholders, and their own skills and experience. 
    I suspect that if this were done, many choices would not correspond to existing forms of research.  Picture each design as a point in multi-dimensional space.  Then it may be that the points would be widely distributed. 
    If you consider actual designs, I suspect a very different image results.  I think that designs would cluster heavily.  And the reason, I suspect, is that people make many of the choices implicitly.  They use the methodology and methods they were taught. 
    In action research, for example, there are distinct schools.  I don't believe these can be explained entirely by the appropriateness of the design choices people make.  I suspect a lot of it is more explained by the history of the person, especially their early tertiary education.back ]
  4. To make my own biases clear...  I value participation highly.  My own autonomy is dear to me;  I try to accord to others the same autonomy.  This means that I prefer not to assume what they might prefer, but rather to consult with them about that. 
    I think that those who know me would agree that I put considerable effort into building high levels of involvement and autonomy in what I do. 
    I agree, too, with many of the advantages claimed for participation.  In particular, I think that people are in general more committed to their own decisions than to the decisions of others.  There are practical reasons, then, to aim at high levels of participation. 
    Having said all this, however, I prefer not to limit my choices by regarding participation as obligatory.  I try to make the choice which is most likely to deliver the desired outcomes in the particular situation?back ]
  5. This is sufficiently different to the procedure for quantitative research that perhaps I should say a little more.  In quantitative research the aim is to develop a description of the sample that is also true for the population from which it is drawn. 
    In theory, this is most often done by random sampling, where people are chosen in such a way that each person has the same chance of being selected.  In practice, at least in quasi-experimental field research (and much survey research) a partial response rate undermines the intention.  (There are ways of dealing with this, but they are not all that often used.)
    To improve representativeness, a "stratified random" sample is often used.  Key dimensions on which people differ are identified.  The sample is chosen randomly within each of those classifications. 
    In qualitative research, on the other hand, you want to make sure that you don't miss any important information.  The proportion of people holding any particular view is usually not so important.  What is more important is that, as far as possible, all views are heard.back ]
  6. Pronouns of feminine gender are intended to refer to both male and female, unless the context dictates otherwise.back ]
  7. Dick, B.  (1997) Structured focus groups.  Electronic file available at URLs

 back ]
  8. A stakeholder is someone who has a stake in the issue being considered, or who can affect the issue.  I often find it useful to distinguish between direct stakeholders (who are directly and immediately affected) and indirect stakeholders (for whom the effect is less direct).back ]
  9. I have done facilitation through an interpreter, and found it very difficult.  First, it takes very much longer to do anything.  Second, it is difficult to act with the same immediacy and relevance when what you say, and the reaction to it, are delayed in time.  Third, it requires skilled interpreters who have a knowledge of the specialist language you use.back ]
  10. The Snyder evaluation process is an example of a process which can be used for this purpose.  One application, for example, is for a team to use it to set up their own systems for performance feedback.  It then becomes much like an action research version of highly-participative quality management.  See the archived file at URLs

 back ]
  11. This is the emphasis of the Deakin approach to action research.  See, also, Fals-Borda, O., and Rahman, M.A.  (1991) Action and knowledge:  breaking the monopoly with participatory action research.  London:  Intermediate Technology Publications.  [ back ]



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.  (1997) Participative processes [On line].  Available at




navbar 4 

Maintained by
Bob Dick;  this version 1.05w last revised 20000103

A text version is available at URL