- 1 I use the term ‘scientific research’ rather than the term ‘research’ in order to separate it from a (...)
1As humans, we carry out some form of research from a very young age. Even as babies, every new object we taste or feel is a form of experimentation as we try to understand the world in which we live, and our relation to it. We continue to experiment all our lives. We observe, we draw conclusions, we form hypotheses, and then we test the hypotheses to see if they are valid. Our ‘experimentation’ or research will either confirm or negate our original hypotheses. Such curiosity is a natural part of human nature. I shall call this type of research ‘everyday research’. It is something we are all engaged in to some extent. Everyday research is perhaps particularly important in the lives of teachers. The questions that arise from our profession lead automatically to suppositions or hypotheses. Many language teachers are curious about what is going on in their classrooms. Teachers are not only actors in their learning environments, they are also observers. They notice that some things work well, others work badly. They learn from their experience and adapt their teaching accordingly. They ask themselves all sorts of questions concerning the acquisition of language and the efficacy of their lessons or methodology. They are curious to know why the same lesson produces different results with two apparently similar groups of students. They are interested to know what students are actually learning during a lesson. Why is it that some lessons appear more motivating than others? Why do some lessons apparently produce better or more learning than others? But this everyday research is often flawed in terms of the criteria imposed by ‘scientific research’1. As language teachers, usually untrained in scientific methodology, we often form hypotheses based on subjective, qualitative data. We do not construct experiments to test our hypotheses, or if we do, various elements in the research are not carried out scientifically or certain variables are not taken into account (especially in language learning research which is a methodological minefield!).
2Scientific research, on the other hand, follows certain rules of methodology, data collection and publication that take it beyond and above our ‘everyday research’. In ‘scientific’ research, however, the researcher is generally observer and remains outside the experiment. What this article deals with, is yet another form of research that is neither what I have described as ‘everyday research’, nor ‘scientific research’, but is termed ‘action research’. Action research is not everyday research in that it incorporates strict methodology borrowed from scientific research. It cannot, however, be considered the same as scientific research because the researcher him/herself is part of the experimental environment, and is observer-actor-researcher, and it is this element of subjectivity and inclusion in the research environment that separates action research from other research.
A teacher notices that students using the on-line CD-ROM for TOEFL training obtain better TOEFL exam results than those not using this support.
He forms the hypothesis that students using the CD-ROM resource learn better than those using other or no resources.
He pre-tests a group of students, and forms three student groups with the same average ability levels. He provides the CD-ROM resource for one group, a text resource for another group, and no resources for the third group. After a few weeks of training, he re-tests the three groups.
He treats his data statistically to compare the results of the three groups. He finds no statistically significant difference between the CD-ROM and text groups, but both are significantly better than the ‘no resource’ group.
His hypothesis is not supported, and therefore he must find another reason for the success of the CD-ROM students prior to his experimentation.
86. New experimentation
His conclusions lead him to form new hypotheses to explain his observations, and he carries out a questionnaire, the results of which indicate that the students using the CD-ROM resource (prior to experimentation) were also the most highly motivated students who already had above-average TOEFL levels.
He writes up his findings and publishes it, or presents it as a communication at a colloquium. The dissemination of results forms an essential part of any research. It allows other researchers or teachers to test the same hypotheses in their own learning environments which can in turn validate or invalidate the original research that took place. It also allows teachers to take advantage of the research already undertaken, and use the results to alter their own language programmes or methodology.
Now that the teacher has a better idea of why some students perform better than other students at TOEFL tests, he can adapt the resources or teaching methods accordingly, and then once again, restart the cycle as observer. Many of us tend to form conclusions without testing our hypotheses. The teacher in the example could easily have concluded, following observation and hypothesis forming, and prior to experimentation, that CD-ROM use improved performance (based on the real data of his TOEFL results) and thereby launched an expensive project to increase the institute’s intranet resources. Action research imposes a certain rigour and methodology so that not only does experimentation take place, but also that its results are valid and repeatable. Any experimentation within a learning environment is already so full of variables that are difficult to control, that scientific methodology is an absolute must.
11To get straight onto the right track, and also to save a lot of time, our action researcher would do well, prior to elaborating his experimentation, to refer to existing theory and research in the field that interests him. It is well worth going back to original research publications, and not contenting oneself with a book of general theory. A book on the subject matter might well provide an excellent starting point, but a quick look at the bibliography will help locate key research work that were key articles in the formation of related theory. The direct application of theory without referral to the research that took place can be misleading. Learning environments can vary enormously. Research carried out in a lycée in a socially deprived area where pupils are likely to have priorities other than learning a foreign language, will certainly provide very different results from research carried out in a Grande École where one is dealing with a small percentage of highly motivated future engineers for whom languages form part and parcel of their career path. If the lycée research was used to form language learning theory, then the application of that theory in a different environment might not provide the expected results. It is for this reason that the action researcher should look at the original research and take a careful look at the similarities or differences between his and the research’s learning environment. Theory should, of course, be applicable across learning environments, but in that theory is formed from experimentation, and that experimentation always takes place within a real and specific environment, then it is certainly possible to have theory which is environment-specific.
12In the particular case described here, there are two principal theories that underpin our work: negotiation theory and phonetic rehearsal loop theory. Negotiation theory is an extension of communication theory, and suggests that discourse modifications in non-native speaker/non-native speaker (NNS-NNS) interaction aid language acquisition. Negotiation elements vary according to the researcher, but in general include repetition, stressing a word or an element within a word, the provision of synonyms, grammatical transformation, an explanation in L2 or L1, a translation etc. Negotiation in NNS-NNS discourse involves the active manipulation of language between two speakers trying to negotiate the content of their interaction (negotiation of content) and verify that their discourse partner understands what they are saying (negotiation of comprehension). Recent research sets negotiation theory within the Vygotskian framework of ‘proximal zone’ theory (Lantolf 1989; Aljaafreh 1994). For those wanting a fuller description of negotiation theory, a separate bibliography of negotiation research is provided.
13If negotiation theory seems to prove that negotiation assists language acquisition, there remains the question of the mental operations that lead to the assimilation of language and language competence. One such theory is that of Elizabeth Service (1992) who stresses the importance of pronouncing words (within meaningful social or learning contexts) in order to assist the passage of new words, word groups or communication strategies from the short-term memory into the long-term memory. There is no room in this article to go into depth concerning these theories, nor to explain the details of the ramifications involved in applying and testing these theories in the language-learning environment, but I shall describe the general cycle of progress of the application the action research based on these theories at INH.
- 2 INH incorporates two colleges, ENSHAP (École Nationale Supérieure d’Horticulture et d’Aménagement d (...)
14Action started prior to action research at INH2. In the late 1970s, Robert Tuffigo, had the intuition, inspiration and energy to start work on interactive learning. A first ‘test’ language laboratory of 8 posts was used to put students into pairing situations, with the idea that (NNS-NNS) interaction would be beneficial to language acquisition. Technical, administrative and methodological problems had to be overcome, but the experience was considered (subjectively and qualitatively) by both teacher and students to be a positive one. At this point in time there was no particular concern about the methodology of action research. This initial work should therefore be considered as action rather than research in that there was no language team, no experimental methodology and no data collection and analysis. It was, however, an important start to a process that was to influence the whole structure of language teaching and language learning within INH over the years to follow. The original laboratory gave way to a 24-post Tandberg IS10 audio laboratory, in the form of a U to facilitate pair work. Televisions were added to allow the diffusion of video.
15Through dissemination at colloquia and conferences, interest grew in the interactive methodology being used at INH, and many teachers from all over France came to visit to see the innovative work that was being carried out. In 1995 I joined Tuffigo on the language team, and was immediately impressed by two things: the apparent rapidity with which students acquired language competence with only two hours of English per week, and the evident pleasure and enjoyment they took from their language lessons. At the time there were no data from INH to prove or disprove the efficacy of the interactive method and the theories mentioned in the preceding chapter were still unknown to us. As part of my DEA programme with Bordeaux University, I decided to carry out an action research project on this particular and interesting pedagogy (Rees 1997). Although action research is often defined as requiring collegiality or a group of researchers (Narcy 1998), I believe that individual research can also be defined as action research. It is, in my opinion, the research methodology, the research environment and the positioning of the researcher within the research context that defines the type of research, for surely research cannot change its name or character because it is carried out by a different number of researchers?
16My initial research aim was to quantify differences between ‘interactive’ and ‘non-interactive’ groups. We used a total of 240 students from INH and Maîtrise students from the Faculty of Science, Angers University. In a sense, it was not a case of true action research, in that the action - that is to say interactive learning - had already taken place, and the research element was arriving later. We did, however, globally, seem to fulfil the requirements described by Burns:
Action research is a systematic process of investigating practical issues or concerns which arise within a particular social context. (Burns 1999)
La recherche-action est la réalisation pratique de résolutions de problèmes de terrain. (Narcy 1998)
18Initial work was to find a bibliography dealing with NNS-NNS interactive pair work, which led us to work by Pica (1985; 1987a; 1987b; 1988; 1991), Doughty (1986) and Long (1981a; 1981b; 1983a; 1983b; 1983c; 1985; 1990) amongst many others. This served as a theoretical and methodological background from which to design the experimental procedures used.
19Early experiments found that students in interactive groups had a statistically significant improved long-term memory for semantic and visual items in comparison with the students in non-interactive groups. Publication of the findings in ASp (Rees 1998) and dissemination in various colloquia helped spread the encouraging findings. It was now apparent that interactive learning had consequences on long-term recall, and the theory proposed by Service (1992) on the phonological rehearsal loop seemed to provide a theoretical foundation and explanation for the assumption that oral production (output), rather than input, effected the long-term acquisition of new vocabulary.
20The administration at INH became aware of the increasing national interest in the work being carried out at INH, and a 25-post multimedia laboratory (Tandberg 3000) was built in 1998. This laboratory initiated a series of new questions as far as the use of technology was concerned. Experimentation began on the use of internet and intranet, as well as on the setting up of international collaborative projects with other European colleges of horticulture. An account of some of the successes and failures of our early networking years is given in the Cahiers de l’APLIUT (Rees 1999a) and elsewhere (Rees 1999b).
21The use of multimedia labs for language teaching (a second 25-post lab was added in 2001) led to the need for teacher-training and the use of a core of contractual, trained teachers, instead of the unfortunate but common practice of using a large proportion of part-time teachers. Recognition by the county (la région des Pays de la Loire) of the use of technology for language teaching, in which we were cited as the ‘Phare de TICE regional’, led to the financing of a major project within INH (8.5 million francs) and also to radical changes to the perception of language teaching by the administration. The language unit formed part of the project team:
A distinctive feature of action research is that those affected by planned changes have the primary responsibility for deciding on courses of critically informed action which seem likely to lead to improvement, and for evaluating the results of strategies tried out in practice. (Kemmis & McTaggart 1988)
22This brief diary of success is not intended for self-promotion, but to demonstrate the cyclical nature of action research and the empirical repercussions in a variety of institutional changes. But if present success is built partly on the perception of action taken by the language unit, it is with further research that progress will be made. I am currently carrying out research on language laboratory discourse analysis and praxeological change within an interactive learning environment (Rees 2001). Having verified the effect of interactive learning on memory in previous research, current work is focused on whether or not the actual NNS-NNS discourse during inter-group pair work corresponds to what has been written about NNS-NNS negotiation in general. Once again, we return to testing research-based theory in the classroom to check its relevance in a particular learning environment. The discourse analysis carried out can provide data on the quantity of negotiation techniques used (stressing, repetition, provision of synonyms, explanation, grammatical transformation, etc.) and measure the re-utilisation of new key words (and their associated word families) by the learner during a lesson. Results of this research will probably lead, once again, to curriculum changes concerning subject matter and methodology.
23This short overview is of one institution’s cyclical progress within the spiral of action research. It is not intended to provide anything particularly new to the language teacher, but rather to provide a call to teachers to take another look at what they are already engaged in. Teachers can, by applying rigorous methodology, transform observation into research, and exploration into discovery. Many of our non-language teaching colleagues are researchers themselves, and appreciate a scientific approach even though they sometimes consider the field of language learning too complicated to construct ‘pure research’. In that methodologically rigorous experimentation provides coherent data, and that those same data form the foundation for institutional or pedagogical change, the apparent veracity of research, as opposed to simple observation, provides an indisputable pillar upon which to construct new language programmes, new language-learning methods and, above all, a greater awareness by language teachers themselves of the why’s and wherefore’s of what they are teaching.