Maye Taylor

This chapter will at first concern itself with a general view of action research as, simply, 'a way of trying out changes and seeing what happens', and put forward the view that this has particular appeal to researchers motivated by a philosophy of social change such as feminism, anti-racism or socialism. It involves abandoning the idea that there must be a strict separation between science, research and action.


Action research can be seen to have evolved from the work of Kurt Lewin in the 1940s (e.g. Lewin 1946), at which time, drawing upon several traditions of the mid-century, it was lauded as an important innovation in social inquiry, and was taken up in a variety of contexts, notably industry, education and community affairs. In social psychology Lewin probably did the most to promote and popularize the idea of studying things through changing them and seeing the effect, which in essence is the argument that in order to gain insight into a process one must create a change and then observe the variable effects and new dynamics.

Lewin was a strong exponent of action research in its concern with power relations between researcher and researched and the rights of the individual, foreshadowing the new paradigm psychology of the 1980s.

Lewin's 'cycle' of planning, action and 'fact-gathering' was the forerunner of Elliott's (1980) action research spiral. This is illustrated, for example, in one project where Lewin incorporated 'guidance' by sociologists with the actions of socially disadvantaged groups in a democratic process aimed at social change in which the benefits of the research were mutual. Lewin encouraged communities to study the results of their actions and examine the origins of their own biases in an endeavour drastically to change relationships within the communities. This constituted quite a direct challenge to the prevailing practice of using 'proxy' information about 'marginal' populations, because the underlying conceptions of social pathology encouraged researchers to devalue direct accounts from respondents in favour of those mediated by professionals. Current practice increasingly accords 'equality of status1 to those who are researched and with it the right to speak and have their views seen as central to the research enterprise, exemplified in Freire's (1972) concept of 'conscientization', which involves a deepening awareness of their own sociocultural identity and their capacity to transform their lives.

In psychology tensions between the need for experimental rigour and the flexibility demanded by professional standards applied in 'real-world' settings has produced acrimonious debate. According to Frankel (1986), the critics were simply lamenting the 'confusion and noise' foisted upon social science by the 'humanness of human beings'. Like the librarian who dreams of the tidiness of the bookshelves without patrons, the 'neo-positivists' fantasize a spick and span social science where researchers are all identical, unbiased, infallible, measuring instruments. Research would be so much easier if researchers did not have to interact. However, Eisner (1984) argued the need for research 'pluralism' and urged the social science world not only to accept, but also to welcome a proliferation of research paradigms and to take advantage of the new angles they provide for viewing the world. Even though, by the 1970s, action research had been 'rediscovered', there is srill strong opposition to its acceptance as 'real research'. The paucity of action research projects listed in psychology departments indicates this. However, there is healthy activity in some spheres, notably practitioner research (see Sapsford and Abbott 1992). In contemporary practitioner research it is accepted that the research methodology cannot be separated from conceptual analysis; that is, even though research and action are analytically distinguishable, they are inextricably intertwined in practice. In Polanyi's (1962) terms 'knowledge is always gained through action and for action'.

Cohen and Manion (1980) argue that while it is no less scientific and rigorous than applied research, action research interprets scientific method more loosely while focusing on precise knowledge applied to a specific problem in a specific setting: its unique strength is that it is self-evaluative and collaborative, with an ultimate objective being to improve practice. This gives 'practitioners' the kind of knowledge they can apply to their own behaviour in the midst of ongoing events, in such a way that it helps them inquire more effectively with others about their common purposes.

So, in action research the research 'question' arises out of the problems of practitioners and it is an important aspect of this approach that the analysis of the situation is in situ. The immediate aim of the research is to understand these problems, and the researcher, who may or may not be the actual practitioner, formulates speculative and tentative general principles about the problems that have been identified. From these it is possible to generate hypotheses about what action is likely to lead to (desired) improvements. The action can then be tried out and data on its effect collected, and the data can be then used to revise the earlier hypothesis. Two useful definitions of action research are:

the study of a social situation with a view to improving the quality of action within it. [The] total process... review, diagnosis, planning, implementation, monitoring effects - provides the necessary link between self-evaluation and professional development.

Essentially an on-the-spot procedure designed to deal with a concrete problem located in an immediate situation. This means that a step-by-step process is constantly monitored over varying periods of time and by a variety of techniques... diaries, interviews, case studies etc., so that the ensuing feedback may be translated into modifications, adjustments, directional changes, redefinitions as necessary.

(Cohen and Manion 1980: 47)

One last point about ethics in action research comes into sharp focus when we consider the infamous Miligram (1974) 'experiments', which involved deceiving the subjects, and is transformed into a strong set of working principles for action research projects, as will be outlined later in this chapter.

Feminist researchers have further elaborated on the theme of action research, stressing as a matter of central importance that research should be about change. Reinharz (1992) extends the concept and talks about action-in-research, identifying five types of specific action in research, namely action research per se, participatory/collaborative research, prevalence and needs assessment, evaluation research and démystification. She argues that each has validity in its own right.

Action research To qualify under this heading it must be research in which action and evaluation proceed separately but simultaneously. These would be research projects that attempt directly to change people's behaviour. While gathering data in traditional or innovative ways, they intervene and study in a continuous series of feedback loops. A good example of this would be the work of Hanmer and Saunders (1984) into forms of violence against women, where community-based, at-home interviewing with the purpose of feeding the information gained back to the community in order to develop new forms of self-help and mutual aid among women was used. The research involvement led to an attempt to form a support group for survivors of violence and to make referrals to women's crisis and safety services.

Participatory or collaborative research The essential feature here is that the people studied make the decisions about the study format and data analysis. This model of research is designed to create social and individual change by altering the role relations of people involved in the project. This is very clear in feminist participatory research, where the distinction between the researcher(s) and those on whom the research is done disappears. Lather's (1988) work serves as a good example. In her project, low-income women were trained to research their own economic circumstances in order to understand and change them. This links strongly with Lewin's views on power relations, for in participatory research attempts are made to form egalitarian relations, with the researcher abandoning 'control' and adopting an approach of openness, reciprocity and shared risk. Participants thus make decisions rather than function as passive objects, they are 'co-researchers* rather than 'research subjects'. Much work with adult survivors of childhood sexual abuse is progressing along these lines, with the professional therapist acting as resource to self-help groups seeking to understand their own behaviour and change it.

Prevalence and needs assessment Here the research seeks to determine the absolute or relative number of people with a particular experience or need. Conventional research in this area has relied heavily on surveys which 'distanced' results from the situation, a 'let's find out and then act if we find it* approach, often used, according to radical social workers for example, as a way of delaying interventions, in contrast to placing the emphasis on mobilizing people to set up resources and organizations to respond to the needs as they are being identified, measured and re-defined. MacKinnon's work (1979) on sexual harassment first demonstrated the power of this approach, showing how meetings called and women discovering their own experience as they spoke identified common themes which led to the size of the problem becoming evident, and the basis for necessary action being revealed without preconception or suggestion on the part of the researcher.

Evaluation research The purpose here is to evaluate the effectiveness of different types of action in meeting needs or solving problems. It can be used to evaluate individual and organizational behaviour and to evaluate evaluation research itself. An important example of such work is that done on evaluating the effectiveness of different forms of individual behaviour to determine which strategies enabled women to stop a rape in progress. Bart and O'Brien's (1985) intention with this work was to generate data-based advice that could be given to schools, hospitals and courts for them to give to women.

Demystification Central here is the belief that the very act of obtaining knowledge creates the potential for change. There is a paucity of research about certain groups, which accentuates and perpetuates their powerless-ness: researchers on women's employment have found little data on the employment situation of disabled or lesbian women, for example. A crucial point here, for action research, is that because the needs and opinions of these groups are not known their views have less influence on the conditions under which they live. The Boston Women's Health Collective's book, Our Bodies, Ourselves (1984) is a prime example of the empowering impact of action research because it enabled women to understand the working of their own bodies, become more knowledgeable about prevention and treatment of women's health problems and thus take more control themselves. This in turn meant that they were less frequently subjected to the power of male medical experts, which the research project had identified as one of the 'problems' in the first place.

Clearly, action researchers have extra responsibility in addressing the 'does the end justify the means' issue, and the response has to be unequivocal: the action taken has to be in the best possible interests of the people involved; ethically there can be no place for conscious exploitation. Practitioner action researchers must work in such a way that they safeguard the practice aspects of their professional work while maintaining a rigorous and reflexive research stance. Research investigating the sex life of ferns (yes, it does exist!) is quite different from that investigating the best way of carrying out first interviews with rape victims. It is also clear that action research can often involve the practitioner researcher in new sets of relations with colleagues and clients (Kemmis 1982). The ethical guidelines for research have to be stringently applied in action research, which by definition is intervening in people's lives. There has to be respect for the whole life of the person, not just as a research subject.

So far, this chapter has explored some of the definitions, principles and approaches to action research, seeing it to be essentially practical and problem solving. Perhaps the reader can see why it is the hope of administrators, politicians and practitioners and the despair of many academics. Action research means intervention in a world where everything can be happening at once and it is impossible to be sure what arises from what, where there is no ethical way of controlling (or measuring) the 'intervening variables' because those 'intervening variables* are actually people, with their emotional responses, their conceptualizations, their needs, their defence mechanisms etc. It is all a 'scientific researcher's nightmare. There is a Marxist point being made here, that we should not just understand the world but change it. Feminists concur and currently radical social workers loudly agree. Many claim that 'objective value-free' research is a 'cop out', where the research becomes the end in itself and avoids the real issues of what you do about it, what you can do about it and what works.

Some educationalists are anxious, seeing action research as vulnerable to co-option by uncritical policy makers and managers. Teachers are already being trained to view action research in schools as a form of inquiry into the best techniques to produce prespecified curriculum objectives or increases in standardized test scores, which is entirely contrary to the spirit of action research. Elliott (1980} warns that it is only a matter of time before action research will be promoted as the newest strategy to help teachers improve pupil achievement in order to meet national curriculum targets. In other words, the actions action research warrants need to be subjected to a political evaluation in the terms of the context in which they intervene rather than being treated as inherently progressive or worthwhile.

Doing action research

The principle of action research is really quite simple: an existing state of affairs is seen as problematic, it is identified, named and described in an appropriate way, attempts are made to change it and these attempts are monitored, and so on. So, action research is appropriate when 'specific knowledge is required for a specific problem in a specific situation, or when a new approach is to be grafted on to an existing system' (Cohen and Manion 1980).


Action research needs to be planned in the same systematic way as any other type of research. It is useful to draw up a checklist of preliminary questions in order to develop a description of the situation, such as: 'What is happening already? What is the rationale for this? What am I trying to change? What are the possibilities? Who is affected? With whom will I have to negotiate?' A general outline of the stages of action research is:

1 First you need to identify some general idea or problem and clarify just what it is that you are interested in, like ethnography and unlike conventional research, and this general idea can shift as the work progresses. It helps if you are really interested in the area. Sustaining research is often difficult, so a little bit of passion won't go amiss.

2 Spend some time describing the 'facts' of the situation at the outset.

3 Make a preliminary explanation of the 'facts' of the situation by generating hypotheses, by means of brainstorming sessions, for example.

4 Test your hypothesis, i.e. put an action into operation and see if you were right.

This is often referred to as the action research spiral (Elliott 1980) and looks like this:

• Select the general area. Discuss, observe, read and decide on your first action.

• Take your action (monitor the action).

• Examine the information you have collected.

• Evaluate (a) processes, (b) outcomes. ■ Plan next action.

Data collection

You need to gather data in order to be in a position of being able to monitor the action (practice) which is at the centre of the inquiry. Action research is multimethod research (several chapters in this book take a detailed look at some of those methods). The methods selected for gathering the information you need for action will depend on the nature of the information required. It is important to gather information that will tell you more as a practitioner than you already generally know. This should be in such a way that it can inform the thinking of not just you as, perhaps, 'key' researcher, but of all concerned with the intervention programmes, and so offer the opportunity to evaluate the efficacy of intervention programmes in general. This may involve a combination of the following procedures, where appropriate. It is important to stress at this point that if you are carrying out practitioner action research you must select data collection methods that do not distort and intrude on your practice.

1 Collection of documents relating to the situation. These can range from newspaper articles where appropriate, through official documents, policy satements and correspondence.

2 Keeping a detailed diary. It is strongly advisable to keep a research diary whatever the approach being used, but for action research it is crucial. You can record your own ideas, including anything that might not be recorded anywhere else as it seems too general. Your diary will be particularly important when writing up, particularly for the design and methodology, 'what happened when'.

3 Observation notes. It is helpful to keep notes of meetings, lessons etc., which could be in the form of observation schedules. Elliott suggests a 'running commentary' approach to this at first, becoming more specific using checklists of relevant phenomena derived from the earlier analyses (see the research spiral).

4 Questionnaire surveys. Normally these are of other people involved in the project, to gain their impressions, attitudes and experiences. At first you might find an open format helps when you are doing the initial 'exploring'. Later, when you are involved in checking or choosing between interpretations, a closed format might be more helpful.

5 Interviews. The more sustained interaction of interviews allows more subtle nuances of perhaps an unfamiliar perspective to be explored in detail and gradually clarified.

6 Shadowing. Participants in the situation may be followed by an observer over a sustained period, in the ethnographic tradition. Shadow studies can give vital information when the study extends over boundaries into different aspects of practice, as can happen in institutional life.

7 Tape/video recording. This allows for repeated monitoring of the data you have collected.

8 Stil! photographs. These can be very useful for subsequent participant discussion, 'freezing' a situation which is then sharply in focus.

9 Triangulation. This is an essential ingredient whereby you use a range of the above methods to check out information gained, interpretations and your decisions about action.


Two strands are evident: the professional ideal, which means continuing openness to the development of good practice, and the scientific ideal of the continuing growth of understanding through critique and revision. Given this essential nature of action research it is not intended here to rake one completely worked example, as the chapter has already concerned itself with principles and practice. The research spiral always applies, so clearly the crucial issue for undertaking any such project is for you to identify the problem. The research can be big or small, from individual case study work to community-based projects: the choice of area will be constrained only by opportunity and context. Some illustrations should suffice.

Zamorski (1987) did a piece of action research involving a socially isolated Asian boy, apparently suffering from a problem of low self-esteem connected with racial tensions in his school and community. Zamorski:

• carefully observed his behaviour with other children;

• experimented with small group games with varying groups of children, examining patterns of linguistic interaction during this;

• intervened to counteract racist comments by some children;

• negotiated changes in partners for classwork and seating arrangements.

Her work 'documents1 the progress of an individual 'therapy', and analyses the effect of each successive strategy in the light of a developing theory as to the origin and the structure of the problem. The research actions included using groupwork, changing the physical structure of the classroom to change the interaction patterns, actively intervening when any racist comments were made to the boy or in his hearing and giving him attention and support. Changes brought about included transforming the school and teacher's understanding of the nature of the problem: illustrating the nature of the links between attempts to solve the problems, i.e. the isolation of the boy, and attempts to understand the problem, i.e. the patterned effects of institutionalized racism.

Action research, which can start with the identification of a problem, can also start quite dramatically when the problem forces itself to our attention, as the following will show. A sexual assault centre for women found itself with a problem of referrals when a rape victim was referred on, by one of the volunteer counsellors, to a psychiatrist. The counsellor had identified that the victim needed to be seen by a specialist mental health professional, as the counsellor felt that the service the centre was able to offer was not sufficient for her needs. The woman contacted the centre some rime later, highly distressed, and told this story. She had been to see the psychiatrist, who, 'in passing', in the assessment session had humorously(J) said to her, 'Well Mrs X, look on it as community service, if they had not raped you they would have raped your neighbour.' She fled. The centre counsellors, themselves already concerned about their lack of knowledge of the mental health resources and their own limitations, took the incident seriously. I would add a note here that I am quite aware of the shocking nature of this example, that the comment is extremely provocative and that this is a sensitive issue. It is quoted to underline the seriousness: action research concerns quality of life. Some of the stages that might follow are:

• use the incident as a critical incident to start a discussion about the needs of such clients (the starting point for a large action research project);

• organize a day conference for all appropriate workers in the area;

• set up information networks;

• produce a directory of services in the area;

• initiate a programme of training and supervision of voluntary counsellors;

• evaluate the effects of such actions, and so on through the research spiral.

This particular action research project would obviously be a large-scale one and by its very nature is not an exemplar of the small-scale projects that you, the reader, might be envisaging. It is quoted here to underline further the emphasis on the action part of action research. As the action research spiral indicates, all such research starts with the need to select a general area, discuss, observe and decide on your action. At the point of taking such action the project could be 'scaled down1 to manageable proportions to be carried out by one person, or several persons undertaking a collaborative study. The choice is yours, for the range of subjects and problems that could lend themselves to action research is very wide, but the topic has to lie in your own area of expertise and interest. The whole field of counselling psychology, for example, lends itself to this approach, given the increasing turn to the provision of a counselling service as a response to observed 'problems'. The centrality of the 'helping relationship' and the need for research to inform action as well as to provide evaluation of changes brought about by interventions at individual, group and organizational levels clearly places it in this research field. Organizationally a useful example would be a research action project on an in-house counselling service: is it used and does it reduce targeted behaviours?

I will further illustrate the sequence with a representative example.

Identify a problem. An attached youth worker, trained in counselling, informed by equal opportunities issues and new to a centre, might want to make changes in the light of observing the large gender imbalance in users of the centre. There are very few young women at the centre, and when they are there they seem marginalized to few activities.

The questions what, why and how need to be addressed. Reading of previous research on youth work, discussion with colleagues, users of the centre and neighbourhood contacts, and perusal of local newspapers might well elicit information pointing at a history of male domination of the centre.

Formulate a course of action. Start by setting up single-sex discussion groups with the current centre users.

The young women identify verbal sexual harassment, bullying and the exercise of territorial domination, which has 'put off many of their friends. They say they put up with it because at least the club is better than being at home and they cannot afford many of the other places. The young men see no problem, they like things the way they are.

In the light of the fact that the young women said they had enjoyed the discussion groups, continue these as an ongoing support group one evening a week, initially for a ten-week trial period.

Monitor the increase, if any, in women's attendance and record conclusions from the group discussion about the nature of the experiences they have and what they would like to see done about them. Observe activities in the centre to see if there are informal changes in the take-up of the various resources, looking to see if the support group is giving them the confidence to challenge.

Take formal action to ensure that the young women are given equal opportunity time on, say, the pool table and other arenas of activity that are designated as 'open'.

Monitor verbal harassment and intervene. Record in detail all incidents, and set up a group with the young men to look at the issue of verbal harassment in order to get them to stop it themselves. Monitor the effect of the groups.

Perhaps take individual counselling sessions where the only other step would be exclusion. Exclude individuals who refuse to adhere to the rules.

If necessary, allocate separate time periods for use of the resources by specific groups (e.g. girls or boys) to ensure equality of access.

Continue monitoring action as the time allotted to the project allows, and to the point when you are satisfied that the changes to practice have been made and the aims of the research have been met.

Writing up of such action research is necessarily a descriptive process. The comments made in Chapter 3 are applicable here. In addition it is important to report the time sequence very carefully, highlighting the various stages and indicating the problem(s) you identified, what you did, what you saw happen, right through the whole project. It is crucial that at each of the stages where you take action you elaborate on your justification for the action. Full description is important in its own right, as previously stated in Chapter 3, but also because in action research the detail is necessary to allow for close scrutiny of the decisions taken in the light of the evidence collected, and thus to allow for ongoing evaluation.

0 0

Post a comment