Action research

I first encountered action research as a secondary school teacher, on a part-time diploma course, during which we were challenged to focus on a series of our lessons in detail. Reflecting on each lesson, we were required to ask, ‘What did the pupils do?’ ‘What did they learn?’ ‘How worthwhile was it?’ ‘What did I do?’ ‘What did I learn?’ and ‘What will I do next?’ I audio-recorded and transcribed lessons on the music of Messiaen, Schoenberg and Harrison Birtwistle, music which the pupils found difficult to understand, and planned each lesson in the light of the answers about the previous lesson. Towards the end of the course we presented summaries of our learning to the whole group; I explained how I had developed my views of teaching, resulting in a less didactic, more practical approach.

Later I discovered that this process was a form of action research, typically conceptualised as research undertaken by practitioners into their own practice, in order to improve it (Elliott, Reference ELLIOTT1991). Practitioners (such as teachers) decide what is worth researching, carry out research and thereby become research-informed. Action research is sometimes contrasted with more traditional methods in which knowledge production is the job of researchers:

Action research is thus associated with the terms ‘teacher research’ (Lytle & Cochran-Smith, Reference LYTLE and COCHRAN-SMITH1992) ‘practitioner research’ (Middlewood et al., Reference MIDDLEWOOD, COLEMAN and LUMBY1999) and ‘self-study’ (Loughran, Reference LOUGHRAN2005). It is used in fields where it is helpful to integrate research with action, and is essentially practical in nature.

Although there are various models of action research, it is possible to generalise about what it is, how, and why it is done. Its process has been encapsulated in various models, which are usually variations of a plan – act – evaluate – reflect cycle, first described by Lewin in the 1940s and since elaborated in many different ways (e.g. Elliott, Reference ELLIOTT1991: 70–71; McNiff, Reference MCNIFF1988: 21–46, Zuber-Skerritt, Reference ZUBER-SKERRITT1996: 99). Educational action research can begin with practitioners asking, ‘how do I improve my practice?’ (Whitehead, Reference WHITEHEAD1989). In seeking answers, they investigate their own practice, plan and carry out interventions to improve it and evaluate the intended and unintended consequences of these interventions, interrogating data in order to ground their evaluations in evidence. They reflect on each stage in order to generate new plans, thus starting the cycle again. As Harris (Reference HARRIS2000) says,

On our diploma course the questions, ‘What did I do?’ ‘What did the pupils do?’ and ‘What did they learn?’ required us to consider our actions and the consequences for the pupils. The question, ‘How worthwhile was it?’ prompted us to evaluate these consequences. The question, ‘What did I learn?’ helped us to reflect and consider changes to our teaching, whilst the final question prompted us to re-start the action research cycle.

Because action research positions practitioners as constructing their own knowledge, it is seen as a democratic process; a ‘grass-roots movement’ (Ormell, Reference ORMELL2000). The aims of action research are frequently political, to do with attaining greater social justice for the participants and the people they serve, and it is sometimes used to challenge unjust systems and practices in organisations, including schools. Action research processes are often collaborative; people working together, in a democratic process, to effect change. Somekh (Reference SOMEKH2006) provides a useful summary of action research by summarising the ‘methodological principles’ that underpin her ‘broad, inclusive definition’ of the process. For her, action research:

1. (1) integrates research and action (1a) in a series of flexible cycles

2. (2) is conducted by a collaborative partnership of participants and researchers

3. (3) involves the development of knowledge and understanding of . . . change and development in a natural (as opposed to contrived) social situation

4. (4) starts from a vision of social transformation and aspirations of greater social justice for all

5. (5) involves a high level of reflexivity and sensitivity to the role of the self

6. (6) involves exploratory engagement with a wide range of existing knowledge

7. (7) engenders powerful learning for participants

8. (8) locates the enquiry in an understanding of broader historical, political and ideological contexts (Somekh, Reference SOMEKH2006: 6–8, numbers not in original)

Educational action research has had a considerable growth in popularity during its 60-year history. However, in a major review of music education research, Rideout & Feldman (Reference RIDEOUT, FELDMAN, Colwell and Richardson2002) reported that, ‘action research has had little impact on research in music education and music student teaching even though its potential contributions were expounded 30 years ago’ (p. 882). In the same review Leglar & Collay (Reference LEGLAR, COLLAY, Colwell and Richardson2002) noted a recent ‘considerable interest’ in action research from music educators, saying,

Previous reviews of action research in music education

Previous reviews have related music education action research to other endeavours and have typically cited a small number of studies to illustrate its characteristics. Regelski (Reference REGELSKI1995) contended that music teaching is conducted largely ‘on the basis of past practice, recipe teaching, and passing fads’ (p. 65). He argued that action research, underpinned by critical theory, can challenge teachers by uncovering their false consciousness and misplaced faith in common-sense teaching, providing ‘a new and disciplined basis for the rational critique of current practices, ideals and rationales of music education’ as well as, ‘facilitat(ing) change through a systematic, scientific process’ (p. 64). Regelski did not consider actual reports of action research but based his argument on philosophical writings, including Lewin (Reference LEWIN1946) and Carr & Kemmis (Reference CARR and KEMMIS1986). Gifford (Reference GIFFORD1997) compared action research with action learning (in which people study their own actions and experience in order to maximise their learning and improve their performance) and conceptualised action research as including action learning but being ‘more deliberate, systematic and rigorous and is made public’ (p. 113). Roulston (Reference ROULSTON2006) situated action research within a qualitative paradigm, stressing its participatory nature and citing Miller's (Reference MILLER1996: Reference MILLER2004) studies as examples. Bresler (Reference BRESLER1996) also situated action research within a family of applied qualitative approaches. She located the roots of such approaches in L'Art de toucher le clavecin (published in 1717), in which ‘Francois Couperin expressed pedagogical assertions based on extensive observations of student behaviour’. Bresler (Reference BRESLER1996) argued that, ‘through action research we learn about the processes of improvement in music instruction’ (p. 5) and described Miller's (Reference MILLER1996) doctoral thesis by way of illustration. A more extended exploration of the subject (Bresler, Reference BRESLER1995/2006) compared the questions, methods and findings that are characteristic of action research with those of ethnography and phenomenology, and characterised action research questions as ‘practical, local, how-to issues, directly concerned with improving a specific classroom practice’. Although she saw action research as operating within a qualitative paradigm, Bresler (Reference BRESLER1995/2006) noted that it can also involve quantitative methods, and described findings as leading to, ‘deep(er) understandings of the operational and experienced curriculum and . . . changes in teaching’ (p. 21). She pointed out that, ‘Because the act of teaching is intense, energy consuming and oriented toward doing as opposed to reflecting, the perspective of an outsider often proves extremely helpful’ (p. 19) and cited Soby (Reference SOBY and Treacher1989) as an illustrative example of action research. Each of these reviews explained the value of action research and related this to music education. However, because the actual reports they refer to were illustrative, they lacked inductive analysis that might explore the characteristics of different studies.

Some studies have investigated action research enquiries by teachers or student teachers. Roulston et al. (Reference ROULSTON, LEGETTE, DELOACH, BUCKHALTER-PITTMAN, CORY and GRENIER2005), investigating a research group of two early-career music teachers and two academics, found the group's work was mutually beneficial, especially in developing the skills or ‘asking critical questions . . . and seeking evidence based answers’ (p. 14). The authors suggested that, to maximise professional development potential, such groups should develop over a substantial period, and adopt a structured framework with specific goals. Both Strand (Reference STRAND2006) and Cain et al. (Reference CAIN, HOLMES, LARRETT and MATTOCK2007) found that being involved in action research encouraged student teachers to read educational literature and engage with theory. The students’ writing went beyond expressing their feelings about teaching and focused more on their personal development as teachers. The present article builds on these studies by analysing 24 action research reports by teachers and academics in music education and asking, ‘What are the distinguishing characteristics of action research in music education?’

Analysis

The analysis in Table 1 reveals a wide range of action research projects, involving single teachers and many teachers, with and without support from academics, in sites which include schools, universities, conservatoires, extra-curricular and community spaces. The focus of the action included curriculum, resources, assessment, behaviour management and teaching approaches, with participants including teachers, parents and students: the very young, school-children, adolescents and adults. Although the most common data collection methods were qualitative, including reflective journals, interviews and participant observations, some studies also employed quantitative methods. Where data analysis was described, it was sometimes inductive, deriving themes from the data (e.g. Davidson, Reference DAVIDSON and Davidson2004) and sometimes deductive, relating data to pre-existing theories (e.g. Miller, Reference MILLER2004). With regard to Somekh's methodological principles, all projects employed some of the principles but none employed all, and some occurred markedly more often than others (see Table 2). The following section explores the extent to which Somekh's principles were observed in these reports.

Table 2

The extent to which Somekh's methodological principles were observed in these reports

Conclusion

This review has demonstrated that reports, of music education action research, focus on a wide variety of subject matter, although there is not yet a substantial body of reports from any particular field within music education. They integrate research and action, are collaborative, grounded in a body of existing knowledge, and lead to powerful learning for the participants, although few are cyclical, deal with aspects of social transformation, or broad historical, political or ideological contexts. There is little focus on reflexivity. Thus most reports describe fairly pragmatic (rather than ideals-driven) attempts to improve practice locally – in the terms used by Carr & Kemmis (Reference CARR and KEMMIS1986) they are more often practical than emancipatory. And, although no single report meets all of Somekh's principles, this might be because at least two of these principles can be seen as mutually incompatible – no report focuses both on the self and on societal contexts and this could be because both foci cannot be equally well served by the same project.