ACTION RESEARCH IN LANGUAGE EDUCATION

David Nunan

Abstract
In this paper, I hope to provide a rationale for the use of action research in second and foreign language education. Questions addressed in the paper include:
What is action research in language education?
Is action research ‘real’ research?
What are some of the problems confronted by teachers doing action research?
What are some of the solutions to these problems?
What are the views of the teachers on the action research process?
The paper will be illustrated with data from a longitudinal action research project.

Action research: description and rationale
Until comparatively recently, the focus of concern in much of the writing on second and foreign language education was at the level of method. Methodological prescriptions were generally argued logico-deductively, and prescriptions for practice were generally devoid of data. This tended to reinforce the gap between theory, research and practice, a gap which, according to van Lier, is due in part to the obstacles which prevent teachers from doing research:
‘Those of us who work in teacher education know that one of the most difficult things to balance in a course is the tension between theoretical and practical aspects of the profession. ... theory and practice are not perceived as integral parts of a teacher’s practical professional life. ... This situation is the result of communication gaps caused by an increasingly opaque research technocracy, restrictive practices in educational institutions and bureaucracies (e.g. not validating research time, or not granting sabbaticals to teachers for professional renovation), and overburdening teachers who cannot conceive of ways of theorizing and researching that come out of daily work and facilitate that daily work.’
(van Lier 1992:3)
Despite the difficulties referred to by van Lier, there is some evidence that the picture is beginning to change. The change has been prompted in part by a growing sensitivity on the part of many researchers to the complexities of the teacher’s task. Practitioners, on their part, seem to have grown tired of the swings and roundabouts of pedagogic fashion, and are looking for evidence before embracing the latest trend to appear in the educational market place. This is not to suggest that a revolution has taken place, however.
While position papers, and logico-deductive argumentation have not disappeared from the scene (and I am not suggesting for a moment that they should), they are counterbalanced by empirical approaches to inquiry. I believe that these days, when confronted by pedagogical questions and problems, researchers and teachers are more likely than was the case ten or fifteen years ago, to seek relevant data, either through their own research, or through the research of others. Research activity has increased to the point where those who favour logico-deductive solutions to pedagogic problems are beginning to argue that there is too much research.
(Nunan, 1992a).
An important concept underpinning action research (AR) is that of reflective practice. Wallace (1991), in his excellent book on reflective teaching argues that reflective teaching provides a way of developing professional competence by integrating two sources of knowledge: received knowledge, experiential knowledge, with practice. Wallace's conception is captured in the following figure:
[Insert Nunan Fig. 1: Reflective practice model of professional education/development]
He links this with action research, arguing that:
‘"action research" can be attractive for two reasons:
1. It can have a specific and immediate outcome which can be directly related to practice in the teacher’s own context.
2. The ‘findings’ of such research might be primarily specific, i.e. it is not claimed that they are necessarily of general application, and therefore the methods might be more free-ranging than those of conventional research.
.... "Research" of this kind is simply an extension of the normal reflective practice of many teachers, but it is slightly more rigorous and might conceivably lead to more effective outcomes.’
(Wallace 1991:56-7)
As we can see from the selected extracts presented above, action research is justified on the grounds that it is a valuable professional development tool. It represents what I would call an inside out approach to professional development. It represents a departure from the ‘outside in’ approach (i.e. one in which an outside ‘expert’ brings the ‘good news’ to the practitioner in the form of a one-off workshop or seminar). In contrast, the inside out approach begins with the concerns and interests of practitioners, placing them at the centre of the inquiry process. In addition to being centred in the needs and interest of practitioners, and in actively involving them in the their own professional development, the inside out approach, as realised through action research, is longitudinal in that practitioners are involved in medium to long-term inquiry.
I believe that the benefits to professional development are justification enough for the development of an action research agenda. However, I believe that a further rationale for the development of such an agenda comes from the research process itself, and I shall deal with this in section three of my paper. First, however, I should like to look at the steps involved in the action research process.

Steps in the research process
The action research process is generally initiated by the identification by the practitioner of something which they find puzzling or problematic. This puzzlement or problematicity may, in fact, have emerged from a period of observation and reflection. The second step is the collection of baseline data through a preliminary investigation which is designed to identify what is currently happening in the classroom without trying to change anything. Based on a review of the data yielded by the preliminary investigation, an hypothesis is formed. The next step is the development of some form of intervention or change to existing practice, along with a way of evaluating the effects of this change. The final step is reporting on the outcomes of the interaction, and, if necessary, planning further interventions. Two examples of the action research cycle are presented in Tables 1 and 2.

Table 1: The Action Research Cycle: An ESL Example
1. Problem/puzzle Identification -–> A teacher identifies a problem/puzzle.
‘My students don’t seem interested or
motivated.’
| |
2. Preliminary Investigation -–> What’s going on? Recording and observing
class over several days.
| |
3. Hypothesis -–> Content doesn’t seem to stimulate students.
Exclusive use of display questions.
| |
4. Plan intervention -–> Increase use of referential questions.
Make links between content and learners.
| |
5. Outcome -–> More complex interactions.
More involvement and interest.
More ‘natural’ discourse, e.g. students nominate
topics, Ss disagree with teacher, S-S
interaction
| |
6. Reporting -–> Staff development session

Table 2: The Action Research Cycle: A Foreign Language Example
1. Problem Identification -–> A teacher identifies a problem in her
classroom.
‘My students aren’t using the target language
[German]’
| |
2. Preliminary Investigation -–> What’s going on? Recording and observing
class over several days.
| |
3. Hypothesis –-> Teacher uses too much English.
The important stuff is done in English
| |
4. Plan intervention –-> Teacher increases target language use.
Teacher uses German for classroom
management etc.
| |
5. Outcome –-> Dramatic increase in use of German by
students.
| |
6. Reporting -–> Article in teachers’ newsletter.

Action research and ‘real’ research
In the first part of this paper, I argued that action research can be justified on professional development grounds. However, I believe that AR can also be justified on research grounds. In fact, I believe that there is something essentially patronising in the view that, while AR might be good for professional development, it hardly counts as research. Let us, to use a currently fashionable term, ‘deconstruct’ this view. First of all, what do we mean by ‘research’? What is the function of research?
Elsewhere, I have defined research as ‘a systematic process of inquiry consisting of three elements or components: (1) a question, problem, or hypothesis, (2) data, (3) analysis and interpretation of data’ (Nunan 1992a:3). Action research incorporates these three elements and therefore qualifies as 'real' research. For me the salient distinction between AR and other forms of research is that in AR the research process is initiated and carried out by the practitioner. As far as I am concerned, the opposition is not between action research and ‘real’ research, but between good research and bad research. A further characteristic, perhaps differentiating AR from other forms of practitioner research, is that it incorporates an element of intervention and change.
Fundamental to any discussion of research is a consideration of the researcher’s conception of notions such as ‘truth’, ‘objectivity’, and the status of knowledge. I recently attempted to deal with the tensions of objective and subjective knowledge by suggesting that they represent two alternative ways of looking at the world:
‘Two alternative conceptions of the nature of research provide a point of tension within the book. The first view is that external truths exist ‘out there’ somewhere. According to this view, the function of research is to uncover these truths. The second view is that truth is a negotiable commodity contingent upon the historical context within which phenomena are observed and interpreted. Further ‘[research] standards are subject to change in the light of practice [which] would seem to indicate that the search for a substantive universal, ahistorical methodology is futile.’ (Chalmers 1990:21) ... This second, context-bound attitude to research entails a rather different role for the classroom practitioner than the first. If knowledge is tentative and contingent upon context, rather than absolute, then I believe that practitioners, rather than being consumers of other people’s research, should adopt a research orientation to their own classrooms. There is evidence that the teacher-researcher movement is alive and well and gathering strength. However, if the momentum which has gathered is not to falter, and if the teacher-researcher movement is not to become yet another fad, then significant numbers of teachers, graduate students, and others will need skills in planning, implementing and evaluating research.’
(Nunan, 1992a)
There are those who would argue that my definition of research as a systematic process of inquiry involving formulating a question, collecting relevant data, and analysing and interpreting that data is inadequate, that in order to count as research, the process should also meet the twin strictures of reliability and validity. Key questions for establishing the reliability and validity of research are set out in Tables 2 and 3.

Table 3: Questions for establishing the reliability and validity of a study
______________________________________________________
Type Key Question
______________________________________________________
INTERNAL Would an independent researcher, on reanalysing the
RELIABILITY data, come to the same conclusion?
EXTERNAL Would an independent researcher, on replicating the
RELIABILITY study, come to the same conclusion?
INTERNAL Is the research design such that we can confidently claim
VALIDITY that the outcomes are a result of the experimental treatment?
EXTERNAL Is the research design such that we can generalise beyond
VALIDITY the subjects under investigation to a wider population?
______________________________________________________
Source: D. Nunan. 1992. Research Methods in Language Learning. Cambridge University Press.
While I would argue that any research needs to be reliable, the issue of validity is more problematic. If one is not trying to establish a relationship between variables, but (for example) to describe and interpret phenomena in context, does the imperative to demonstrate that one has safeguarded one’s research from threats to internal validity remain? By the same token, if one is not trying to argue from samples to populations, then it would not be unreasonable to assert that external validy is irrelevant. I would ague that as most AR is not concerned with arguing from samples to populations, external validity is not at issue. (For an excellent discussion of issues to do with reliability and validity in qualitative research, see LeCompte and Goetz, 1982).
It is popularly assumed that the purpose of research is to test theories. For example, ‘That communicative language teaching is more effective than audiolingualism.’ Allwright and Bailey have pointed out that there are problems with this proposition. In the first place, some theories are untestable (for example, Krashen’s attestations on ‘subconscious’ acquisition). Secondly, classrooms are too complex for us to control all the variables in the manner prescribed by experimental research. They propose an alternative purpose for research, namely to try and understand and deal with immediate practical problems facing teachers and learners (Allwright and Bailey 1991). If we accept this alternative purpose, we are drawn immediately into embracing AR, because it makes no sense for an outsider to arbitrate on the practical problems facing teachers and learners. This does not mean that outsiders, such as university based researchers, have no role to play in practitioner-based research. However, the role is one of collaboration and advice rather than direction and control.

Problems and solutions in doing action research
I would now like to reassure those who might feel that I am looking at teacher research through rose coloured glasses. It is certainly not the case that everything is rosy in the AR garden. The principal problems identified by teachers with whom I have worked in a number of different contexts include the following:
• Lack of time
• Lack of expertise
• Lack of ongoing support
• Fear of being revealed as an incompetent teacher
• Fear of producing a public account of their research for a wider (unknown) audience
We have experimented with a number of solutions to the problems. I believe that the chances for an action research agenda to succeed will be maximised under the following conditions:
• There is someone ‘on the ground’ to ‘own’ the project.
One or more individuals with training in research methods are available ‘on tap’ to provide assistance and support to teachers.
• That teachers are given paid release time from face-to-face teaching during the course of their action research.
That collaborative focus teams are established so that teachers involved in similar areas of inquiry can support one another.
• That teachers are given adequate training in methods and techniques for identifying issues, collecting data, analysing and interpreting data, and presenting the outcomes of their research.
In order to facilitate the process, colleagues and I have developed an in-service programme. This programme was initially devised for the LIPT project (Languages Inservice Project for Teachers) in South Australia, and has been further modified and refined in Sydney, where a project has been established bringing together mainstream teachers, ESL teachers, and teachers of LOTE (Languages Other than English – a term which is preferred to Foreign Language Teachers as most non-English languages are widely used within the Australian community, and the term ‘foreign’ is therefore a misnomer). In Table 4, I have provided a summary of the professional development programme as it currently exists.

Table 4: The inservice programme in outline
______________________________________________________
SESSSION 1. An introduction to classroom observation and research
a. A series of reflective activities designed to get teachers thinking about their own teaching style.
b. Reflecting on the teaching of others: teachers examine and critique extracts from a range of classrooms identifying those aspects of the extracts they liked and disliked.
c. Identification of ideological beliefs and attitudes underlying critiques.
Between session task: Teachers record and reflect on their own teaching.
SESSION 2: An introduction to action research
a. Teachers report back on the between session task.
b. Introduction to issues and methods in action research.
c. Introduction to the action research process.
Between session task: Teachers develop a draft action plan.
SESSION 3: Focus groups and action plans
a. Formation of focus groups and appointment of facilitators.
b. Sharing of draft action plans.
c. Refining questions.
Between session task: Baseline observation, focus group meetings, preliminary data collection.
SESSION 4: ANALYSING DATA
a. Participants develop ways of analysing and making sense of their data.
Between session task: Ongoing data collection and analysis, focus group meetings.
SESSION 5: WRITING UP
a. Participants receive input on presenting their research.
b. Development of draft reporting outlines.
Between session task: Production of draft reports.
SESSION 6: REFINING REPORTS
Participants receive feedback on and discussion of their reports.
SESSION 7: EVALUATION
Participants evaluate the LIPT process and provide feedback on how their involvement changed them.

Evaluating action research
From what has already been said, it is clear that action research is difficult, messy, problematic, and, in some cases, inconclusive. It consumes a great deal of time, and often strains the goodwill of the teachers involved, as well as those with whom they work. However, evaluative data from teachers themselves suggest that teachers who have been involved in action research are overwhelmingly favourable. For example, Mickan, who collected data on the reaction of outside teachers to his LIPT project writes:
‘Teachers have welcomed the articles from LIPT. They have found them particularly useful and relevant because they depict the complex circumstances of classroom life in an honest and direct way. They have found them a rich source of ideas and valuable for informing their own practice. The warts and all descriptions (including failures and successes), the research techniques used, the analysis of results and the contextual detail are all elements which readers relate to and understand. As such they possess a validity which derives from the detailed narration of classroom ecology. The experiential reports give other practitioners models and ideas for their own practice. They also suggest topics and procedures for classroom investigations in different contexts.’
(Mickan, 1991).
An evaluation by Lewis (1992) is also favourable. She reports on a study conducted with a group of teachers of French immersion programmes in British Columbia. The focus of her research was the effect on the professional practice of the teachers of engaging in AR. She drew the following conclusions from her research.
1. Through the process of systematically implementing their own choice of action project based on the needs of the students in particular, each teacher learned more about their own theories, or frames for teaching, and modified these frames to a certain extent.
2. The frames for teaching of the participants in this study are related to the bigger questions of second language education and education in general. Practice cannot be understood thoroughly without appreciating how educational theory is expressed within teachers’ frames and neither can theory be useful without recognising that what counts is how theory becomes expressed within practice.
3. [The] ‘teacher as researcher’ or ‘reflection in action’ approach to teacher education can be a very powerful way of facilitating change in curriculum.
In evaluating the last of the LIPT projects, we asked teachers to complete the following statements:
Action research is..............................................................................................................
Action research is carried out in order to....................................................................
We also asked them to respond to the following:
1. What are the most significant things you have learned in carrying out your classroom research?
2. What questions/issues has your classroom research raised for you?
3. What further areas/ideas are you interested in pursuing?
Sample responses to the first of these probes on the most significant outcomes for the participants are set out in Appendix 2. It can be seen that these are overwhelmingly favourable, the participants choosing to focus either on the substantive content outcomes (‘By collecting and analyzing data on my children, I found that they were more highly motivated than I had given them credit for.’ ), learning process outcomes (‘The active involvement of the children in the learning process facilitates learning.’ ‘I discovered that kids know how to learn – the project taught me to listen to them.’), or reflections on the research process itself (‘In working through the action research process, I discovered which methods of data collection are most suited to my research question – next time I will be better prepared as I will be more aware of what I am looking for, and will be better able to match my questions and data.’). The enthusiastic validation of learner-centred approaches to instruction, even though this was not a primary aim of most research, is also worth noting.
Finally, participants were asked to complete a checklist to indicate how their teaching had changed as a result of their involvement in the project. Results are set out below. It can be seen from the survey that, if self-reports are to be believed, the experience was, for most teachers, an overwhelmingly positive one.
How has your teaching changed? Complete the following:
Since I have been doing action research, I find that when I teach I now ...
More About the Less
same__________
1. tend to be directive_______________________________ 1___ 14____ 10___
2. try to use a greater variety of behaviours______________ 16___ 6______0___
3. praise students__________________________________ 15___ 10_____ 0 __
4. criticise students_________________________________ 0___ 11____ 13___
5. am aware of students’ feelings______________________ 18___ 6_____ 0___
6. give directions__________________________________ 4____ 16_____ 5___
7. am conscious of my nonverbal communication 11___ 14_____0_
8. use the target language in class______________________ 19__ 6_____ 0___
9. am conscious of nonverbal cues of students___________ 12____ 12____ 0 _
10. try to incorporate student ideas into my teaching___ 20___ 5____ 0___
11. spend more class time talking myself________________ 1____ 9_____15 __
12. try to incorporate student ideas into my teaching 20__ 5____ 0___
13. try to get my students working in groups ________ 15___ 18____ 0___
14. try to get divergent, open-ended student responses____ 14____ 10____ 0___
15. distinguish between enthusiasm and lack of order_____ 9___ 15____ 0___
16. try to get students to participate__________________ 18___ 7____ 0___

Conclusion
In this paper, I have argued that the adoption of an action research orientation can be justified in professional development terms and research terms. Despite the bureaucratic difficulties and obstacles which are placed in the way of teachers, the elitism of a certain cadre of researchers (some of whom were once classroom teachers themselves!), and the suspicion which is sometimes directed at academics who are trying to promote a closer relationship between theory, research, and practice, there is evidence that things are beginning to change. I can offer no more fitting conclusion to this paper than the following extract from two of the profession’s foremost advocates of the development of harmony between theory, research and practice, who have striven in their own teaching, writing and research, to enhance the status of both practitioner and researcher within language education.
‘Slowly, the profession as a whole is realising that, no matter how much intellectual energy is put into the invention of new methods (or of new approaches to syllabus design, and so on), what really matters is what happens when teachers and learners get together in the classroom. … This shift in emphasis from concentrating on planning decisions … to concentrating on looking at what actually happens in the classroom, has led researchers to have much greater respect for classroom teaching. The more we look, the more we find, and the more we realise how complex the teacher’s job is. And teachers, in their turn, faced at last with researchers who have at least some idea of the enormous complexity of everyday classroom life, are beginning to be more receptive to the whole research enterprise. … Being a good classroom teacher means being alive to what goes on in the classroom, alive to the problems of sorting out what matters, moment by moment, from what does not. And that is what classroom research is all about: gaining a better understanding of what good teachers (and learners) do instinctively as a matter of course, so that ultimately all can benefit.’
(Allwright and Bailey, 1991)

APPENDIX 1
THE LIPT PROJECT: SOME FOCUS GROUP AND RESEARCH TITLES
1. LITERACY AND LOTE 3. LOTE TOPICS
Big books and the German classroom Back to the drawing black board
Journal writing in action Two heads are better than one ... or are they?
‘Reading in German? Not a problem!’ Assessing primary French
Making big books What’s that word again?
Tell me more ... Towards a positive classroom
Developing writing skills Making a game
A literature-based program He can’t hear you nodding Sophie
It’s too hard! Mistakes are for learning
Writing what I want to in German Beginning at the beginning or
Language arts and LOTE immersion of tadpoles
Extending a primary German
program
2. DEVELOPING LEARNING STRATEGIES Dai, Forza Giochiamo! Let’s play!
Cross-age tutoring in German Speak!
‘No stop-start’ learning centres and language Learning through games
learning Let’s play games
Cross-age tutoring in Pitjantjatjara In French, please
Learning together Immersion – or in at the deep end
Learning contracts and LOTE Using more Indonesian in the classroom
Cross-age tutoring in Greek When I speak German I feel great!
Learning centres rule OK? Motivation and language learning
Learning how to learn Heaps more French??
Who’s speaking German?
Using an activities-based approach

APPENDIX 2
1. What are the most significant things you have learned in carrying out your classroom research?
• The active involvement of the children in the learning process facilitates learning.
• Children have different learning preferences and teachers need to allow for this in their instructional practices.
• Children find it difficult to express feelings and opinions on paper.
• It is easy to ‘spoon feed’ children, but this leads to ineffective learning.
• Teaching problems only go away if they are recognized and tackled.
• The most important outcome for me was that I learned how to do action research. To benefit, I therefore have to do it again!
• Working with the children together (e.g. finding their thoughts/feelings and acting on them).
• In working through the action research process, I discovered which methods of data collection are most suited to my research question – next time I will be better prepared as I will be more aware of what I am looking for, and will be better able to match my questions and data.
• The process removed my tunnel vision to teaching.
• It helped me to making links with other teachers of Mandarin , as well as parents and the community.
• The process dramatically enhanced my rapport with students.
• I found that by carefully, step-by-step directing students, I was able to give them tools to manage their own learning.
• By collecting and analyzing data on my children, I found that they were more highly motivated than I had given them credit for.
• The most important outcome for me was that I discovered the children enjoy (and respond well) to being consulted about their learning and being given some say in what they learn.
• There was a negative outcome for me – I’ve learned not to expect children to have completed tasks or to value something just because they’re important to me.
• I found that Year 7 learners still need lots of structure and guidance, even when independent skills are encouraged.
• I was disappointed. I expected too much in my initial project – book flood! Only book trickle is possible in such a short time.
• The most important discovery for me was that my students need more time and opportunities to work in groups as they need to learn to work on their own without teacher directed lessons all the time.
• The need for informed input in this process – one needs to read etc., recent research and thinking in order to maximize value of one’s own research, and move beyond one’s own ‘blinkered’ vision.
• The positive benefit of concentrating on one particular area because the attitude/approach of openness and inquiry carries over into one’s teaching in general.
• I have learned that students with a very limited knowledge of the target language are prepared to try to write more than I expected, and that in future I should try to foster this willingness in my classes.
• Contrary to my expectations, I found that the children were keen to be part of a ‘project’. This led to increased motivation (maybe Hawthorne Effect?)
• I have learned that one needs to undertake classroom research. One needs to intervene – observation alone isn’t a good enough indicator of how much children are learning.
• In my research, I delved into how my lessons were arranged and the effectiveness (or not) of my teaching. I looked closely at my learning strategies. It allowed me to construct a unit that was designed for junior primary students’ needs and interests and my research allowed my to construct strategies accordingly.
• I discovered that kids know how to learn – the project taught me to listen to them.

Acknowledgment
Grateful acknowledgment is made to the British Council for financial support.