Theorising Social Work Research

Researching the Social Work process Seminar topics

Researching the Social Work process 11th July 2000 Luton

Theory generation and qualitative research: school exclusion and children looked after Dr. Isabelle Brodie Research Fellow, University of Luton


Within social work research, different theoretical perspectives are favoured at different times. This popularity will depend on how well the explanation being offered seems to answer the questions arising from the particular social, economic and political context. In the same way, the context of a research discipline will inform attitudes to processes of theory generation and methodology. As discussions of evidence-based practice have become increasingly central to social work, so a growing amount of debate has been generated regarding the relative merits of different research methodologies. Indeed Trinder (1996) has commented, 'What is immediately apparent is that the future direction of the social work research enterprise is fundamentally contested. The recent flood of methodological contributions contain very different prescriptions for social work research' (p233).

The role of theory and theorising within these debates has, perhaps, not always been clear. Indeed, it may be questioned how far the generation of theory is recognised as a goal of empirical research within social work. Parker (1989) reflecting on the uneasy and complex relationship between research, policy and practice, has observed that 'Paradoxically, perhaps, the failure of much policy-oriented research to be of much help to practitioners may be because it has not been directed towards the development of theory.We may need more research that is specifically aimed at the development of theory; that is, more basic research that has no immediately obvious benefits to either child care policy or practice' (p263).

The criticism that researchers have given insufficient attention to more abstract theorising is a challenging one in the context of social work research, which is widely perceived as being 'applied' in nature. More generally, Parker's comment raises questions about how far theoretical activity is visible within social work research, and in turn how this relates to policy and practice concerns - if indeed it does. As Little (1998) has noted, social work researchers have frequently been 'tardy in writing about how they arrive at their own formulations' (p 51) and this is perhaps especially true in regard to specifying the theory that underpins their work and the nature of their theoretical output.

This paper begins by briefly considering different approaches to theory generation. It will then focus specifically on the role of an interpretive approach to theory generation, by reflecting on this process in relation to a study of the exclusion from school of children looked after by local authorities. Finally, some consideration is given to the question of whether and how research based on qualitative approaches can be linked to practice and professional contexts. Overall it should be noted that the paper is informed by research in the area of child care, and consequently aspects of the discussion may have less resonance in other areas of social work research.

Theory, research and practice

Bullock et al. (1998) state that, at a theoretical level, the links between research and practice are 'fairly direct' in the sense that it would be surprising if a qualified social worker was not aware of certain concepts - such as those relating to attachment theory. These concepts are sufficiently general and flexible they can be related to different levels of policy and practice. This is not entirely unproblematic; as Beresford (1999) has pointed out, social work has often focused on the idea that social workers draw from an eclectic list of theories, which are then applied with varying degrees of skill. At another level, it appears that a process of informal theorising has an important role in social workers' personal evaluation of their work and in the development of professional expertise (Shaw and Shaw, 1997; Fook, 2000). This may involve drawing on a combination of personal values and commitments, the legislative framework, social work practice models and, less commonly, abstract ideas or models. However, this piecemeal and individualised process tends to remain unarticulated, and may indeed appear non-existent. Thus, for example, in research into children's homes, heads of homes - most of whom were qualified and experienced - found it difficult if not impossible to explain what theoretical approaches underpinned their work (Berridge and Brodie, 1998).

The application of theory must be distinguished from methodological processes which enable the generation of theory. The principal methodological positions have been well-rehearsed within this series of seminars. A reasonable distinction can be made between positivist, interpretive/social constructionist and participative/critical research (Everitt et al., 1992), though each of these categories will embrace several slightly different methodological perspectives. Broadly, however, research rooted in a positivist tradition tends to be associated with quantitative methods. Such research proceeds via a theory testing approach. Hypotheses are formulated and subsequently reformulated in the light of data produced through the application of a specific research method or technique. The theory that results will typically specify the relationship between precisely measurable variables. This is something of an over-simplification as inductive reasoning is by no means absent from such processes; nevertheless the picture is one in which theories are gradually, progressively refined through a series of research studies.

Both interpretive/social constructionist and participative/critical research are more likely to adopt qualitative techniques. Interpretive and social constructionist approaches are more ambivalent in regard to the ability of researchers to generate explanations of a social world that is essentially located in the subjective worlds of research participants. The main aim of such research is to interpret and understand these subjective worlds; consequently interpretive research varies across a continuum in terms of the extent to which 'theorising' is balanced with description as the goal of the research (O'Connell Davidson and Layder, 1993). However, other approaches emphasise the role of discovery and exploration in generating theory. The 'grounded theory' approach of Glaser and Strauss (1967) has been particularly influential in regard to the development of middle-range theory from qualitative data. Grounded theorists do not deny that researchers will have some ideas about the nature of the phenomenon being studied, but see the generation of theory as involving an ongoing, dialectical relationship between the emerging data and the development of theoretical categories. These categories are refined and modified as further data emerges.

Participative approaches, on the other hand, challenge the view that research is a neutral, objective process. Trinder (1996; 2000) describes participative research as essentially involving an awareness of the power relationships inherent to the research process. Thus, 'research is not positioned as a neutral fact finding activity. Instead research, researchers and research participants are located in a world where power is unequally distributed between gender, classes, ethnic groups, professionals. Lives and life stories are significantly structured by these unequal power relations, with significant implications for uncovering 'truths'' (p238). In recognising the relationship between research, knowledge and power, participative research will attempt to 'decentre' the 'expert' role of the researcher through a process of dialogue amongst all the participants. Various forms of action research and research undertaken by users have proved influential within this approach.

The distinctiveness of different approaches should not be over-estimated; while there are differences between interpretive and critical/participative approaches, it would at the very least be an oversimplification to imply that, for example, interpretivists deny the political nature of research processes or the significance of gender or ethnicity. Silverman (1989) advocates that 'mediating entities' between different approaches should be welcome, and that the critical test for the adoption of a theoretical framework in the analysis of research findings is whether that framework is appropriate to the research question. It may be, as is certainly the case with the issue of exclusion from school, that a variety of theoretical approaches could be applied. Thus 'we need not be either interpretivists or positivists, micro or macro analysts, or even qualitative or quantitative researchers. Equally, structuralist, neo-Marxist and Foucauldian perspectives may sometimes be fruitfully added to interactionist and ethnomethodological themes' (p11). Though it would be a mistake to suggest that way in which methodologies are evaluated is a neutral activity, the general point that understanding a research problem may be enhanced by elements from different perspectives is an important one. As the following discussion will suggest, the multi-layered nature of the social work context will require a multi-layered approach to theorising.

The education of children looked after

Having briefly considered differing approaches to theory generation, the discussion will now turn to examine the role of theory in a study of the exclusion from school of children looked after in residential accommodation. Over the past couple of years, the education of children looked after in public care has become prominent on the political agenda, even becoming something of a cause celebre. Children looked after were highlighted as an especially high risk group in the Social Exclusion Unit report on Exclusion and Truancy (Social Exclusion Unit, 1998) and new draft guidance specifically addressing the issue of the education of children looked after was given a high profile launch by Secretary of State David Blunkett in June 1999.

At the time this research was initiated, however, the issue had received much less attention (Jackson, 1987). Evidence existed attesting to low levels of educational achievement, the lack of attention given to educational issues within the care system, and difficulties in liaison between social work and education professionals (Fletcher-Campbell and Hall, 1990). Only one large scale survey had been undertaken of the educational placements and attainments of children looked after in foster and residential care (SSI/OFSTED, 1995).

Concerns were also being raised at the level of professional practice. During the fieldwork for an earlier research study of children's homes, residential social workers repeatedly raised the question of education or the lack of it, frequently commenting that the issue of school exclusion in particular represented perhaps the major challenge they encountered in their work. The specific concern was the fact that young people were at home all day, alternative education was not provided and residential social workers (RSWs) were expected to provide supervision. At a practice level, explanations often mirrored those being provided by researchers: namely that education policy, specifically the Local Management of Schools, placed pressure on schools which made them less sympathetic to the needs of children looked after; that children looked after in residential accommodation were more challenging and consequently more likely to experience problems at school and that, generally, children looked after tended to be stigmatised (Berridge and Brodie, 1996; Berridge and Brodie, 1998).

Meanwhile, the issue of school exclusion more generally had become an increasing matter of concern to policy makers and professionals. There was particular alarm at the increase in the number of permanent school exclusions taking place (Department for Education, 1993), and how this could be managed in terms of practice within schools. A profile of excluded pupils was emerging in which those who were socially disadvantaged predominated. However, while children looked after were frequently identified as a group particularly vulnerable to exclusion, very little was known about how the two were related (Department for Education, 1994). At best the problem tended to be explained in terms of the characteristics of the looked after population, and the failure of the care system to act to promote the education of the looked after group.

This was, therefore, a new problem which was generating concern amongst both researchers, policy makers and professionals. Most of the theories being used to explain the issue appeared to lack supporting empirical evidence, and little information was available about the way in which the experience of care related to being out of school. In this respect it is far from unique within social work, where there continue to be a number of areas where even on very basic issues 'there is either no information or such information as does exist is unreliable, limited in coverage or outdated - or all three' (Moss, Owen and Statham, 1998, p263). Within such a context, it would have been not only inappropriate but extremely difficult to specify hypotheses and so to follow a theory testing approach.

The theoretical premise for my own study was that exclusion does not represent a single event or a purely administrative procedure, but a social process defined and negotiated by the individuals involved. The research was therefore essentially interpretive in approach, specifically adopting a symbolic interactionist perspective. Blumer (1969) defines this approach as based on the view that human beings act towards things - whether physical objects or people or social institutions - on the basis of the meanings these things hold for them. These meanings emerge from, and are continually modified through, interaction. Interactional processes are characterised by a reciprocal exchange of meaning, and individuals are therefore actively engaged in constructing and reconstructing their social worlds. Each person will continually engage in redefining his or her position relative to the social group in which he or she is participating. Social relationships are not, therefore, fixed but are in continual flux as definitions of social relations are jointly and reciprocally proposed and established (Joas, 1987).

This approach does not aim to produce an objective 'mirror reflection' of the social world (Miller and Glassner, 1993) but does attempt to provide access to the meanings people attribute to their experiences and social worlds. Consequently symbolic interactionism is typically associated with the use of observational techniques, specifically participant observation. Such methods, it is argued, permit the researcher access to the way in which a particular social world is negotiated by individuals. Observation was not deemed appropriate to study school exclusion and children looked after; instead detailed case studies of the 17 children were undertaken, involving in-depth, semi-structured interviews with young people, carers, teachers and other professionals, together with examination of relevant documents including case files and reports from relevant professionals.

It should be pointed out that this was not the only theoretical framework through which to examine this issue. As Blyth and Milner (1994) point out, exclusion from school cannot be seen only in terms of young people's experiences of education and schooling but 'as part of the complex relationship between individuals, families, the labour market, health and state support and surveillance services' (p293). Research evidence also indicated that school processes are important. These social and institutional structures represent the context within and upon which individuals act, and may be perceived as important explanations for these actions. The fact that this study was concerned with the exclusion from school of children living in children's homes required consideration of a second set of policies and institutional arrangements which may also be significant in understanding the experience of exclusion for this group of pupils. The study did not deny the validity of attempts to understand exclusion in terms of these wider structures. For example, there would be value in examining exclusion within a critical perspective seeking to highlight the power differentials arising from social relationships based on ethnicity and gender.

Definition and redefinition

In focusing upon the subjective understandings of research participants, interpretive approaches have the capacity to challenge taken for granted assumptions about the nature of social phenomenon. Such description and definition is essential in areas of applied research, and must take place prior to designing interventions. At policy level, Levitas (1999), discussing the concept of 'social exclusion', has highlighted the fact that inadequate definition and conceptualisation of social problems at policy level can result in the selection of contradictory indicators of these phenomena and, in turn, confusing measures of the scale of a problem. Challenging the taken for granted is also important within social work given the fact that that social workers are increasingly subject to regulation in the form of audit and the proliferation of bureaucratic procedures. Child protection is the area of social work where this has most obviously occurred. Yet as Parton (1999) has pointed out, social workers will continue to engage with complex and ambiguous situations for which there can be no clear cut bureaucratic prescriptions.

School exclusion emerged as a 'social problem' during the late 1980s, almost concomitant with the introduction of the term into statutory legislation in England and Wales. The statutory definition itself has proved problematic and has undergone several changes. Official guidance and discussion papers also recognised the fact that, alongside the official categories of fixed-term and permanent exclusion, 'informal' or 'unofficial' exclusions took place. These can generally be defined as taking place without reference to official procedures, which at the time the research was undertaken included the requirement that parents/carers should be informed that permanent exclusion had taken place, the reasons for this and details of how an appeal against the exclusion could be made. A typical scenario for an unofficial exclusion would be, for example, a situation where a school encouraged a parent to arrange a transfer to avoid the stigma of a permanent exclusion. However, the range of circumstances through which unofficial exclusion can take place, together with their hidden nature, make the issue a difficult one to research.

Underpinning the study was the theoretical assumption that the nature of the exclusion process - in terms of the constituent parts of that process and the way in which it has progressed - are actively defined by individuals. The exclusion process would not, therefore, necessarily take place according to the procedures outlined in official guidance or school discipline policies. These procedures may mark or symbolise the development of particular definitions of the situation, or may legitimate these definitions, but in themselves do not tell us how or why a child comes to be excluded.

In this research, the difficulties associated with exclusion as a category, through which to understand the fact that many children looked after were out of school, became clear during the research process. Criteria for inclusion in the study, as outlined in access meetings and letters to children's homes, was the young person's permanent exclusion within the last twelve months. However, the generation of the sample proved unexpectedly difficult, and cases were initially very slow to emerge. This was sometimes due to fairly practical matters, notably the care careers of young people which meant that some children changed placement or returned home quickly. Gradually it became evident that significant difficulties surrounded use of the term 'exclusion'. Some of the cases referred to the researcher were soon found to involve young people who were out of school but were not attending for other reasons. In one case the head of home insisted that a young person had not been excluded, yet on further questioning and examination of documentation from the school it seemed to the researcher that exclusion, albeit not in any official form, had indeed taken place. Residential staff appeared uncertain as to the boundaries of 'exclusion' and their definitions frequently did not cohere with those contained in official policies. It was evident, then, that the recent introduction of the term 'exclusion' did not prevent individuals from using the same language to describe quite different phenomena. It also became clear that the distinction between 'official' and 'unofficial' exclusion was in many respects an unhelpful one, and that a more complex understanding of the issue was required. In particular, it was necessary not only to explore the dynamics of the relationship between exclusion and the experience of being looked after, but to understand the meaning of 'exclusion' within the residential care context.

As the research developed, this initial deconstruction of the category of exclusion contributed to the development of a typology of the processes through which children looked after. In doing so it suggested not only that fewer children were officially excluded from school than was generally thought, but that the distinction made between 'formal' and 'informal' or 'official' and unofficial' exclusion also represented an inadequate conceptualisation of the problem. Instead, four processes of 'exclusion' were identified:

Exclusion by non-admission

This group of young people were 'excluded' by virtue of the fact they did not have a school place. All those excluded in this way had serious behavioural problems, but the disruption of their education had also occurred for other reasons, including changes of placement and inadequate planning. Professional interaction in these cases often involved tensions and conflict, and residential carers in particular felt that the stigma attached to looked after children made it more difficult to obtain a school place. The local education 'market' was also significant.

Exclusion on admission

This type of exclusion referred to young people who were officially or unofficially excluded very quickly after their entry to a school, usually within a few days or weeks. The transient nature of their educational and care careers, coupled with a lack of important knowledge about school systems and culture, made successful integration unlikely. Both schools and young people had a low level of commitment to the school placement.

Graduated exclusion

Graduated exclusion occurred via a lengthy build-up to the final decision to exclude. The outcome of the process was highly predictable, and all the main social actors were aware of the probability that exclusion would take place. Despite this, little if any preventative action was taken by professionals, who usually shared a definition of the young people involved as undesirable members of the school community. As the build-up to the exclusion progressed, there was a gradual detachment of the young person from the school setting, making the final decision to exclude something of a non-event.

Planned exclusion

In these cases exclusion was anticipated, but was also 'planned'. As problems escalated, professionals sought to minimise the consequences of this by making alternative plans. Professionals in these cases had more sophisticated networks and field-workers perceived themselves as having a more significant role in advocating for the young person. Significantly, they also perceived the young person to have the ability to respond positively to education, if this was provided in an appropriate context.

These processes therefore reflected individuals' perceptions of how and why the young person was out of school, and the nature of the professional response. This framework also helps explain differences in the experience of 'exclusion' in different types of residential setting - for example, the age of the young people being cared for and the nature of links with field social workers and schools.

Exclusion as a social process

A further dimension of the study concerned the way in which the exclusion was managed by a range of different professionals. The exclusion from school of children looked after is an issue which necessarily straddles the professional domains of social work and education, yet the way in which these different professionals engaged with each other on a day-to-day basis was unknown. Despite increasing calls for social workers to work across organisational boundaries, there continues to be little knowledge as to how this is worked out in practice.

Symbolic interactionism proved a valuable basis for understanding how residential and field social workers related to teachers and other educational professionals and vice versa. If it is taken that individuals communicate on the basis of shared meanings which emerge through interaction, then the question of how agreement is achieved or is not achieved becomes a matter of particular interest. Existing research suggested that a cultural and organisational 'chasm' between social services and education departments was a major barrier to enhancing the educational welfare of children looked after (Fletcher-Campbell and Hall, 1990; SSI/OFSTED, 1995; Berridge et al., 1997), and that liaison was often driven by the efforts of individual professionals rather than being a matter of normal practice.

Exclusion presented a different kind of problem to the professionals involved. For residential carers, exclusion was perceived as a problem in terms of their having to deal with young people staying in a children's home throughout the day. For head teachers, exclusion was often defined as a problem in relation to resource constraints which limit provision for pupils with behavioural difficulties. There was also considerable evidence of disagreements about the respective approaches of social work and education professionals.

Examination of the interaction in specific cases, however, revealed a rather different picture in which consensus was an important element in managing the exclusion process. Where the patterns of interaction were characterised by 'negotiation' and 'strategic collaboration', professionals tended to agree on the likelihood or inevitability of exclusion, but varied in the extent to which they worked with other professionals to achieve specific ends. Where the pattern of interaction was characterised more by 'estrangement', structural issues - such as the educational arrangements within the local authority, the type and location of the children's home - were more important in understanding difficulties in liaison. Overall, social workers and residential carers emerged as responding more proactively to educational issues than might have been anticipated; in the majority of the 17 cases either the social worker or residential carer perceived themselves and were perceived by others to be actively involved either in working positively with other professionals and/or to be working with the young person to resolve educational difficulties.

In terms of the day-to-day management of the exclusion from school of children looked after, therefore, a much more complex picture emerged of the way in which professionals respond to this issue than would be suggested by current discourse regarding inter-agency relationships. In particular, it was possible to identify the points in the process at which different individuals became involved in working with the young person, and how this helped define the nature of the problem and the options available to social services and education professionals. In conceptualising exclusion as a negotiated social process, young people were also recognised as social actors, experiencing a different set of constraints, but who also influenced the development and outcomes of this process. The explanation for the exclusion was not therefore focused on the characteristics of the looked after population, but was recognised as a social and political issue (Jamrozik and Nocella, 1999).

Translating theory into practice

Philpott (1997), discussing the relationship between research and practice in social work, stated that 'the party's over as far as woolly theories are concerned. Research must be a constant touchstone for social work practice' (p25). As the impetus for practice that is more firmly linked to evidence has strengthened, so questions have arisen regarding what constitutes 'good' evidence for practice, and how this evidence can best be translated into different practice contexts. However, it is clear that practitioners access very little research in the form of books or articles, and that the link between empirical research and social work practice is weak (Bullock et al., 1998). Meanwhile, researchers have been criticised for the inaccessibility of their work (Tozer, 1999). This raises questions about how messages from qualitative research can best be communicated to a practice audience. Certainly work in the interpretive tradition is unlikely to produce the 'law-like generalisations' which managerial expertise may seek for its vindication (MacIntyre, cited in Smith, 2000). There is also too much evidence to bear witness to the inappropriate application of theories within social work practice (Berridge and Brodie, 1996). The corollary of this is not, however, that research should take place predominantly in one methodological tradition.

At one level qualitative research will act simply to gather new information and descriptive data which can, at some later date, be used as the empirical basis for theory construction or testing (Layder, 1993). Qualitative techniques will be especially useful in undertaking this task in relation to research questions of a sensitive nature (McKeganey, 1990). However, I would argue that the theoretical scope of such work is rather greater. Detailed description is not only useful in itself, but can challenge current understandings of a problem and in turn generate new explanatory frameworks. In the process, it reveals the complex array of structural constraints which professionals must negotiate in the process of working with individuals, and how such informal negotiation is undertaken. Amid a public discourse where social worker professionals are frequently the target of criticism, such research can also show that the actions of individuals make a positive impact.

The theory generated by qualitative research will often equate more to 'clusters of ideas' rather than advancing hypotheses regarding the relationship between specified variables. The advantage of these theoretical ideas is their flexibility and the fact that they can be developed and related to particular professional contexts. In this sense theory is a tool, which needs to be scrutinised not only for its logical coherence, but also how far it 'makes sense of the inexplicable' (Parker, 1989). The impact of theory will perhaps be greatest where the ideas being advanced have sufficient 'fit' with the issues professionals encounter on an everyday basis, while at the same time providing an alternative framework within which to understand these issues. It is important that research questions reflect the concerns of practitioners, and in this respect research in a more participative tradition, including action research, will have a significant role to play.

The idea of evidence-based practice emphasises that good practice will take account of the best available research evidence regarding both the nature, causes and typical pathways of social problems or about the success of particular methods to deal with these problems (Hill, 1998, p20). It is encouraging that more rigorous questions are being asked about how theories are applied and how they relate to research evidence in specific areas of practice. This may remove some of the wooliness, but does not make the need for theory any less. As standard definitions of evidence-based practice recognise that practitioners must actively engaged in the 'judicious' application of research knowledge (Sackett et al., 1996; Atherton, 1998). It is therefore important that professionals not only understand the methods of empirical research, but also the nature and purpose of theory and how this can relate to practice.

A number of imaginative initiatives are underway to improve the dissemination of research findings to practitioners and to educate professionals in research methods so that they are in a better position to assess critically the relative merits of different research studies (Atherton, 1998; Bullock et al., 1998; Berridge, 1999). The building of more general theory can also usefully take place through the collation of research studies from which broader messages can be drawn (Department of Health, 1995; Department of Health, 1998). The danger is that, in the attempt to improve accessibility to research, complex messages are distilled into 'key messages' from which theoretical content has been removed. Not only does this render theoretical aspects of research even more invisible, but can run the risk of denying the very elements of intensive, qualitative research which make it of value to professional practice.


This paper has reflected on the process of theory generation in the study of the exclusion from school of children looked after by local authorities. This was a modest study and would by no means claim to be the 'last word' on this issue. Not only is this is an area of policy and practice which has changed rapidly in recent years, but any explanation of the problem must be multi-layered, making reference to different aspects of the social world. Theoretically, however, it was fruitful in developing understanding of the complexity of the issue of exclusion within the context of residential child care, and in producing a theoretical model of exclusion. This potentially offers a framework for intervention at different levels of policy and practice, and also has the potential to be extended and developed to other contexts.

More generally, this discussion has considered how qualitative research of this type can generate theory that is relevant within a social work context where, it has been argued, increasing precedence is being given to empirical research in a quantitative tradition (see, for example, Sinclair, 1998). It is important to reassert the need for a range of methodological perspectives to be applied within social work research, and for continued interaction between social work research and other areas of social science. Within this enterprise, qualitative research has an important role to play in contextualising problems, providing detailed description and generating theoretical explanations that are embedded in the subjective experiences of research participants. The dissemination and application of this kind of knowledge will require ongoing discussion between researchers and research participants. Such dialogue is not an adjunct to, but an essential component of, the theorising process.


Atherton, C. (1999) 'Getting a GRIP', Research, Policy and Planning, 17, 1, 1-6.

Berridge, D. (1999) 'Child welfare in England: problems, promises and prospects', International Journal of Social Welfare, 8, 288-296.

Berridge, D. and Brodie, I. (1998) Children's Homes Revisited. London: Jessica Kingsley.

Berridge, D. and Brodie, I. (1996) 'Residential child care in England and Wales: the enquiries and after' in M. Hill and J. Aldgate (eds) Child Welfare Services: Developments in Law, Policy, Practice and Research. London: Jessica Kingsley.

Blumer, H. (1969) Symbolic Interactionism. Englewood Cliffs: Prentice Hall.

Bloor, M. (1997) 'Addressing social problems through qualitative research' in D. Silverman (ed) Qualitative Research: Theory, Method and Practice. London: Sage.

Brodie, I. (1999) Redefining School Exclusion: Children's Homes and the Educational Process. Unpublished PhD Thesis. Luton: University of Luton.

Brodie, I. and Berridge, D. (1996) Exclusion From School: Research Themes and Issues. Luton: University of Luton Press.

Bullock, R., Gooch, D., Little, M. and Mount, K. (1998) Research in Practice: Experiments in Development and Information Design. Aldershot: Ashgate.

Department for Education (1993) Exclusions: A Discussion Paper. London: Department for Education.

Department for Education (1994) Pupils with Problems (Department for Education Circulars 8-13/94). London: Department for Education.

Everitt, A., Hardiker, P., Littlewood, J. and Mullender, A. (1992) Applied Research for Better Practice. Basingstoke: Macmillan/BASW.

Fletcher-Campbell, F. and Hall, C. (1990) Changing Schools? Changing People? The Education of Children in Care. Slough: NFER.

Fook, J. (2000) 'Deconstructing and reconstructing professional expertise' in B. Fawcett, B. Featherstone, J. Fook and A. Rossiter (eds) Research and Practice in Social Work: Postmodern Feminist Perspectives. London: Routledge.

Glaser, B. and Strauss, A. (1967) The Discovery of Grounded Theory. Chicago: Aldine.

Jackson, S. (1987) The Education of Children in Care. University of Bristol: School of Applied Social Studies.

Jamrozik, A. and Nocella, L. (1998) The Sociology of Social Problems. Cambridge: Cambridge University Press.

Joas, H. (1987) 'Symbolic interactionism', in A. Giddens (ed) Social Theory Today. Cambridge: Policy Press.

Layder, D. (1993) New Strategies in Social Research. Cambridge: Polity.

Little, M. (1998) 'Whispers in the library: a response to Liz Trinder's article on the state of social work research', Child and Family Social Work, 3, 49-56.

McKeganey, N. (1990) 'Drug abuse in the community: needle sharing and the risks of HIV infection', in Cunningham-Burley, S. and McKeganey, N. (eds) Readings in Medical Sociology. London: Routledge.

Miller, J. and Glassner, B. (1997) 'The 'inside' and the 'outside': finding reality in interviews' in D. Silverman (ed) Qualitative Research: Theory, Method and Practice. London: Sage.

Moss, P., Owen, C. and Statham, J. (1998) 'Informing ourselves about early childhood services', Children and Society, 12, 263-274.

O'Connell Davidson, J. and Layder, M. (1994) Methods, Sex and Madness. London: Routledge.

Parker, R. (1989) 'Theme and variations' in B. Kahan (ed) Child Care Research, Policy and Practice. London: Hodder and Stoughton.

Philpot, T. (1997) 'Dedicated follower of fashion', Community Care, 31 July - 6 August, 24-25.

Shaw, I. and Shaw, A. (1997) 'Keeping social work honest: evaluating as profession and practice', British Journal of Social Work, 27, 847-869.

Sheldon, B. (1998) 'Evidence-based social services: prospects and problems', Research, Policy and Planning, 16, 2, 16-18.

Sinclair, R. (1998) 'Developing evidenced-based policy and practice social interventions with children and families', International Journal of Social Research Methodology, 1, 2, 169-176.

Smith, D. (2000) 'What works as evidence for practice? The methodological repertoire in an applied discipline', Theorising Social Work Seminar 4.

SSI/OFSTED (1995) The Education of Children Who Are Looked After Away From Home. London: SSI/OFSTED.

Stevens, M. (1999) 'Assessing outcomes in child welfare: a critical review', Research, Policy and Planning, 17, 1, 26-32.

Tozer, C. (1999) '20 questions: the research needs of children and family social workers', Research, Policy and Planning, 17, 1, 17-22.