Social Work Research in the Child Protection Field in Australia

Abstract

Social work makes an important contribution to child protection practice and policy in Australia, but data are limited about the discipline's contribution to research. The aim of this scoping review was to examine the quantity and nature of Australian social work child protection research for the period 2007–14. A comprehensive search of nine bibliographic databases identified 255 peer-reviewed articles. The papers were authored by 287 researchers, most of whom published only one paper during the period. The research was published in seventeen Australian and thirty-five international journals. Non-empirical papers and those using qualitative methodologies were most prevalent, and there was a lack of research depth and quantum on particular child protection topics. The findings indicate that developing more sustained programmes of research, underpinned by a balanced portfolio of methods and approaches, would maximise the potential for research impact on child protection policy and practice.

Child protection, social work, research impact, evidence-based practice

Introduction

While a relatively wealthy country, indicators of social need in Australia reveal disturbing levels of inequality, homelessness, child abuse and inadequate support for people with disability that require ongoing action (Australian Institute of Health and Welfare, 2013). Innovative ways of working in the human services, driven by quality research, are needed to better respond to entrenched poverty and disadvantage. In a quest to improve outcomes for vulnerable children and families, the child protection sector in Australia and internationally has been urged to strengthen evidence-informed practice. Over a decade ago, Macdonald (1998) argued for more uptake of evidence-based practices and more investment to expand the knowledge base underpinning social work practice. Likewise, the need for more research and an evidence-based culture in Australian child protection was advocated (Tomison, 2002). While much international research is relevant to Australian child protection, the context for practice is different to other countries, and arrangements for delivering human services are different. Australia is a federation, with child protection in the remit of state governments. They each have their own laws and approaches, but share a ‘child protection orientation’ that aims to protect children from parental abuse or neglect within a legal and investigative framework (Gilbert et al., 2011). There are policies to encourage supportive family-based responses to concerns about children, but support services are insufficient (Lonne et al., 2009). There are very high rates of racial disparity, with Indigenous children seven times more likely to be in care than non-Indigenous children, but not a high rate of out-of-home care by international standards (Tilbury and Thoburn, 2008). There are also demographic, cultural and historical factors to be taken into account. Therefore, a strong Australian corpus of research is necessary to address local problems. It is necessary for the human services industry, as much as any other sector of the economy, to engage in world-class research that develops new understandings, technologies, and processes to increase productivity and effectiveness.

In order to ascertain the state of Australian child protection research and identify areas for development, several audits have been undertaken. Cashmore and Ainsworth (2004) conducted an audit of Australian out-of-home care research for 1995–2004. Based on a survey, consultations with service and research networks, searches of policy and resource documents produced by federal, state and territory governments and non-government child welfare organisations, and database funding searches, ninety-four research projects were identified. The main findings were that there were no national or large-scale evaluation projects, few replication studies, a prevalence of qualitative methods and small purposive samples, a dominance of research on foster-care, and minimal focus on Indigenous children, children from non-English-speaking cultures and children with disabilities. It was concluded that Australia did not compare well internationally on a number of dimensions (Cashmore and Ainsworth, 2004). Building on the out-of-home care audit, an audit of Australian child protection research was undertaken (Higgins et al., 2005). This charted child protection and family support research for 1995–2004 using similar methods and found that, while all forms of maltreatment were covered, overall there was a lack of research ‘such that it is not possible to claim an adequate evidence-base for sound policy and practice decisions’ (Higgins et al., 2005, p. 28).

An audit for the period 1995–2010 was undertaken to inform the strategic development of the Council of Australian Government's National Framework for Protecting Australia's Children (McDonald et al., 2011). This audit sought information directly from researchers and used literature searches to examine published and unpublished papers on the child protection continuum from prevention to statutory child intervention and out-of-home care. There were 1,359 research projects identified, with around half pertaining to child abuse and neglect and a quarter to out-of-home care (p. vii). This audit provided a broad map of the state of Australian child protection research and found a number of topic, population, methodological and data gaps. It concluded that these gaps hindered policy makers and practitioners in facilitating better outcomes for vulnerable Australian children and families (McDonald et al., 2011).

Governments in the USA and the UK have been more pro-active in building a stronger research and evidence base in child protection than Australian counterparts. The UK has led the way with organisations such as the Social Care Institute for Excellence and the Centre for Excellence and Outcomes in Children and Young People's Services. These initiatives have been important in showing that generating evidence for practice goes beyond ‘gold-standard’ experimental research; it must incorporate the knowledge of service users and the expertise of practitioners, as well as grapple with the messy realities of the service delivery world if evidence is to be implementable on a large scale. The 2008 Economic and Social Research Council initiative in the UK to support the development of a national strategy for social work research aimed to produce a step change in the breadth, depth and quality of social work research (Bywaters, 2008; Shaw and Norton, 2008). Research infrastructure, investment, leadership and vision were identified as necessary to carry forward high-quality, high-impact research (Sharland, 2013). The US federal government funds the National Child Welfare Workforce Institute in order to advance knowledge, workforce development and research skills among child welfare personnel. In addition, several social work research organisations and networks have been established, with the Institute for the Advancement of Social Work Research and the Society for Social Work and Research leading the dissemination of research-based practice knowledge and sponsorship of research. In Australia, neither government nor non-government child protection agencies have much research capacity or funds to commission research, and few have research units. The Australian Research Alliance for Children and Youth and the Australian Institute for Family Studies both undertake research clearinghouse roles, but Australian child and family services lag behind with regard to tripartite practice research enterprise models that incorporate research development, research dissemination and research utilisation (Feldman, 2010).

Shaw (2007) argued that social work research generates and adds to knowledge about social work and human services, highlights the nature of lived experience and the ways that inequality and diversity shape experiences, and promotes social justice and social inclusion. While child protection is an inter-disciplinary field, social work is a key player, especially in front line practice. Practitioners with social work qualifications are preferred by statutory child protection agencies in some Australian states, but most agencies have a mix of social workers and graduates of three-year degree programmes in human services, criminology and psychology (McArthur and Thomson, 2012). Yet other professional groups in law, health and policing have arguably had more influence on child protection policy, especially through high-profile public inquiries that have shaped current policy directions. Social work was noted as a strong contributor to research in the audits (Higgins et al., 2005; McDonald et al., 2011). However, the distinctive contribution of social work to Australian child protection scholarship is not known. This is needed in order to maximise research impact for the benefit of children and families, to advance the standing of the profession and to improve the evidence base for practice.

Methodology

The methodology for the scoping review was guided by the framework developed by Arksey and O'Malley (2005) and other scoping reviews of social work research (Crisp, 2000; Ryan and Sheehan, 2009; Brough et al., 2013). Scoping reviews can be undertaken for a number of purposes, such as examining the nature of research activity, exploring the potential value of a systematic review, summarising and communicating relevant knowledge, and identifying gaps in the evidence base (Arksey and O'Malley, 2005). The purpose of this scoping review was to assess Australian social work research on child protection and specifically to examine the nature, type and purpose of the research undertaken from 2007 to 2014.

Inclusion and exclusion

Peer-reviewed, English-language papers in scholarly journals were retrieved. The focus was on peer-reviewed research as a marker of quality. Searches of grey literature, reports, clearinghouses, books, book chapters, dissertations, general web searches and hand-searching were not undertaken. Journal papers that were excluded were editorials, book reviews, papers with no author stipulated, directors' reports, letters to the editor, news articles, commentaries on research, sector resources, interviews with researchers and papers where the journal could not be accessed electronically.

The research question necessitated definitions of what was included as ‘Australian’, ‘social work’, ‘research’ and ‘child protection’. To qualify as Australian, two criteria were used: undertaken either by Australian authors or by non-Australian authors of research that reported Australian data or content. Social work research was operationalised according to the qualifications of authors: a least one author on a paper had a tertiary social work qualification. The Higher Education Research Data Collection definition of research was used: research involves using systematic methods to create new knowledge, or developing existing knowledge in a new and creative way so as to generate new concepts, methodologies and understandings (Department of Education and Training, 2014). Determining the parameters of child protection research was not simple given that child protection covers many topics on a continuum of preventive and tertiary responses. The breadth of the search terms, as will be outlined later, yielded many results. The topics that were included and excluded are listed in Table 1. Topics that were relevant to child protection, but not directly and explicitly connected, were excluded.

Table 1

Included and excluded topics

Included	Excluded (no substantial link to child protection)
Out-of-home care—e.g. residential care, foster-care, kinship care, adoption, out-of-home care experiences, leaving care, care outcomes, carers	Youth issues—e.g. policy, youth work, substance abuse, school deviance, citizenship, peer relations, teenage risk-taking
Systems and structures—e.g. agency structure, funding arrangements, performance, strategy, regulation, management	Children—e.g. child welfare, child labour, health and well-being, parental separation, homelessness, resilience, bullying, refugee children
Policy and legislation—e.g. legislative change, child death reviews, confidentiality, national framework, links to juvenile justice	Families—family policy, sole parents, parent imprisonment, working hours, family law, family dispute resolution, families with children with a disability; step families, family sociology, family relationship centres
Practice and processes—e.g. interviewing, reunification practices, culturally sensitive practice, child removal, community development	Parenting—experiences of fathers, separation, marital breakdown, interactions with children, parenting stress, intercultural parenting, parental well-being, refugee parenting, parenting assessment, socially disadvantaged, teenage parents
Domestic violence linked to child protection—e.g. children witnessing domestic violence	Couples—couple problems, domestic violence (no connection to child protection), assisted reproductive technology, divorce, same-sex relationships
Workforce—e.g. training, qualifications, roles of different professionals, supervision, burnout, turnover	Aboriginal and Torres Strait Islanders and Indigenous people—education, health, school transition, human rights, gambling, parent–child interactions, offending, housing and social conditions, income management, racial identity, addiction, early intervention programmes (parents)
Interventions and programmes—e.g. therapy for child maltreatment impact, programmes for maltreating parents, child abuse prevention programmes	Interventions and programmes—adolescent therapy for problem behaviour, trauma, family therapy, brief counselling, parent education, interventions for migrant parents, family group therapy, therapeutic programmes for children, substance-dependent parents, family support programmes with no child protection link
Aboriginal and Torres Strait Islander child protection—specific research addressing child protection for Aboriginal people and Torres Strait Islanders	Child protection and out-of-home care in other countries—no Australian link or minimal Australian link
Risk assessment, screening and decision making—e.g. risk assessment and other forms of child protection assessment	Child protection but not relevant—no clinical, counselling, welfare, social programme evaluation or research link, e.g. child abuse in fiction, media, art, nudity, pornography, censorship
Impact of child abuse and neglect—the impact of childhood maltreatment on children and adults	Services—integrated children's services, professionalism of children's services, hard-to-reach service users, alcohol and drugs, community services arrangements
Courts and legal issues—e.g. children's courts, child protection orders, child protection proceedings, legal responses to child protection	Carers—young carers, respite care, paid care work, care constructs, children's unpaid care-giving (unless specifically foster or kinship carers)
Child maltreatment types—e.g. physical abuse, emotional abuse, sexual abuse and neglect	Disability—family support, re-ablement services, community life, discrimination, sibling with intellectual disability
Parents' experiences of child protection processes and services	Early childhood education and childcare
Historical and institutional abuse	Health—non-child protection health issues
Medical issues	Other issues not directly relevant to child protection

Search strategy

We searched nine bibliographic databases: Cinahl Plus, Informit Family and Society Collection, Medline, Social Services Abstracts (Proquest), PsycINFO (Ovid), Sociological Abstracts (Proquest), Scopus, Current Contents Connect and PubMed. An academic research librarian provided recommendations for database choice, search terms and search strategies. Although a consistent search strategy was used for all databases, variation occurred due to differences in functionality and tools available. Combinations of the following subject headings and keywords with Boolean phrases were employed: Austral*, child welfare, child, child abuse, child maltreatment, child protection, family support, foster-care, out-of-home care, care system, human services and social work. This strategy resulted in 5,716 records. Similarly to Serbinski and Shlonsky's (2014) review, this strategy was not efficient because it yielded considerable irrelevant material, but it increased the likelihood that relevant research would be identified.

Screening and selection

In order to manage and further refine material, a second screening process was used. All records were imported into the Thomson Reuters Endnote software program. Duplicates were removed and a journal limit strategy was employed to identify Australian social work research. Words used for the journal limit (field selected: journal/secondary title) were social work, human services, child, children, family, administration, public, policy, social, practice, fostering, adoption, disability, youth, Australasian, Australia, sociology, residential and issues. This resulted in 2,271 records. Each record was then manually screened to see whether it was eligible for the next level of analysis. The title, keywords and abstract of each record were examined and this process resulted in 516 records that were eligible for further analysis. Decisions about inclusion and exclusion were initially classified by one researcher and independently checked by another. The 516 records were then more closely read to ascertain whether there was social work authorship and the final numbers of papers included was 255. Papers excluded at the final stage had no author with social work qualifications.

Analysis

A quantitative analysis was undertaken of the following information obtained from each paper: author name, year of publication, journal of publication, research focus, unit of analysis, method, type of research and funding source. Like Ryan and Sheehan (2009), categorisation was based on authors' own descriptions, when available. The author name and qualification, title, abstract, keywords, research aims and acknowledgements or declarations were examined to obtain the information of interest. When sufficient information could not be obtained from the above sources, the full paper was perused.

Findings

As a scoping review, this paper presents the nature and range of social work research undertaken on Australian child protection between 2007 and March 2014. The search strategy yielded 255 papers. A total of 287 writers (sole or combination) authored the papers. At least one author per paper had a social work qualification. The frequencies and percentages of the authors are presented in Table 2. Most authors (68.0 per cent) only published one paper during the period. A smaller number of authors (26.8 per cent) published between two and five publications. A few authors (fifteen) made a more substantial contribution of six or more publications to Australian child protection research.

Table 2

Number of publications by author

Number of publications	Frequency	Percentage
1 publication by author	195	68.0
2 publications by author	47	16.4
3 to 5 publications by author	30	10.4
6 to 9 publications by author	9	3.1
10 or more publications by author	6	2.1
Total	287	100

The distribution of papers across the time span (Table 3) shows that research was consistently published—an average of thirty-five papers per annum. There was a spike in publication outputs in 2009, but this was not sustained. Only the first quarter of 2014 was included, hence the low count for that year.

Table 3

Year of publication (2007–March 2014)

Year of publication	Frequency	Percentage
2007	37	14.5
2008	28	11.0
2009	55	21.6
2010	33	12.9
2011	30	11.8
2012	27	10.6
2013	37	14.5
2014 (to March)	8	3.1
Total	255	100

Social work child protection research was published in fifty-two journals, covering the fields of child welfare, child protection, law, disability, family therapy, psychiatry, social work, social welfare, social policy, administration, health care, public health, sociology, child care, family, social issues and youth. Most papers were in social work and child welfare journals. Table 4 details the ten highest frequency journals which accounted for three-quarters of the 255 publications. Children Australia was highest (22.4 per cent), followed by Australian Social Work (11.4 per cent) and Communities, Children and Families Australia (10.2 per cent). Most of the papers were published in Australian journals (n = 145, 56.9 per cent of all papers) with the remainder in UK, US and other international journals.

Table 4

Top ten journals of publications

Journal	Frequency	Percentage
Children Australia*	57	22.4
Australian Social Work*	29	11.4
Communities, Children and Families Australia*	26	10.2
Child and Family Social Work	21	8.2
Children and Youth Services Review	14	5.5
Child Abuse Review	10	3.9
British Journal of Social Work	10	3.9
Family Matters*	10	3.9
Journal of Family Studies*	7	2.7
International Journal of Social Welfare	5	1.9
Total	189	74.1

* Australian journals.

The research focus or topics were mainly identified from the abstract of the paper. As shown in Table 5, fifteen topics were identified. Many papers were on out-of-home care (34.1 per cent). Research on child protection systems, policy issues and practice approaches were the next most frequent. Research pertaining specifically to Aboriginal and Torres Strait Islander child protection accounted for 4.7 per cent of the papers and research on medical issues in child protection was the least frequent (two papers, 0.8 per cent).

Table 5

Main topic of research focus

Topic	Frequency	Percentage
Out-of-home care	87	34.1
Systems and structures	42	16.5
Policy and legislation	20	7.8
Practices and processes	19	7.5
Domestic violence links to child protection	15	5.9
Workforce	14	5.5
Intervention programmes	13	5.1
Aboriginal and Torres Strait Islander-specific	12	4.7
Risk assessment	8	3.1
Impact of child abuse and neglect	7	2.7
Courts and legal issues	5	1.9
Child maltreatment types	4	1.6
Parents' experiences of child protection	4	1.6
Historical and institutional abuse	3	1.2
Medical issues	2	0.8
Total	255	100

Research on out-of-home care examined placement journeys within care, young people's perceptions of their needs, educational attainment, transitioning from care, and outcomes for children and care leavers. Other topics were child–carer relationships, kinship care, residential care, foster-care, adoption, parental contact and carer needs. Although diverse topics were examined, some were not substantially researched, with only one or two papers on the issue (e.g. unaccompanied asylum seeking children in out-of-home care). Research on child protection systems and structural arrangements accounted for 16.5 per cent of the papers. This research analysed system trends, regulation, performance, mandatory reporting and management issues. Collaboration between different service sectors (e.g. mental health, domestic violence, criminal justice and child protection) was also examined. This suggests that social work researchers are actively engaged in the analysis and critique of service delivery arrangements in the Australian child protection system. Research on policy issues represented 7.8 per cent of papers. Again, there was a diversity of issues covered (e.g. confidentiality, criminal history checks, poverty) but with slightly more attention given to analysis of child protection inquiries and child death reports. Papers that were categorised as workforce were predominately on capacity building and educating, retaining and supporting staff.

Practice topics included the impact of practice on families, practice frameworks, culturally sensitive work, child removal and community development. Papers that addressed the linkage between domestic violence and child protection mainly considered post-separation parenting such as family violence in family law proceedings, allegations of family violence in parenting disputes, and the interface between child protection and the family court; and the impact of children witnessing domestic violence, including effects, trauma and therapeutic work. The different types and levels of interventions discussed included parenting programmes, group work, family support, child sexual abuse prevention, family group conferencing, child abuse prevention programmes and others. In the sub-theme risk assessment, a variety of risk assessment and decision-making tools were examined. Although specific issues pertaining to Aboriginal and Torres Strait Islander children and families were examined (programmes to address trauma, intervention in communities, over-representation in child protection, community violence, effective practice approaches), compared to other topics, attention could best be described as modest. Most research on the impact of child abuse and neglect was about male survivors of child sexual abuse. Note that the prevalence of particular topics to some extent reflects the research interests of the social work authors who most frequently published.

Papers were analysed according to methodological approach or method (Table 6). Where possible, papers were classified according to statements of methodology made within papers. When papers did not contain explicit identification of methods used, they were categorised using the following definitions: quantitative methods were defined as data collection that measured the quantity of something and data were represented as numbers; and qualitative methods were defined as an approach to data collection that attempted to discover the quality or meaning of something and data were represented as words or pictures (Dudley, 2011). Mixed-methods studies used a combination of the two. A systematic review was coded when recognised guidelines for undertaking the review were used and cited. Non-empirical papers were not data-based (i.e. no primary field data, experimental work or analysis of secondary data was contained in the paper). The non-empirical papers included commentaries, reflections, narrative literature reviews with no stated method, programme descriptions and theoretical or policy critiques of problems, proposals, tools, systems, structures and inquiries. As shown in Table 6, most papers were non-empirical (47.1 per cent). Qualitative methods were most frequently used (32.9 per cent of all papers; representing eighty-four of 135 empirical papers), followed by quantitative and mixed-methods studies which each represented 9.8 per cent of papers (twenty-five in each category). There was one systematic review. The purpose and type of research—whether exploratory, descriptive, evaluative or interpretive—were difficult to assess. Many papers were not clear or did not contain explicit statements or markers clarifying their purpose and the type of research conducted.

Table 6

Methods

Method	Frequency	Percentage
Non-empirical	120	47.1
Qualitative	84	32.9
Mixed method	25	9.8
Quantitative	25	9.8
Systematic review	1	0.4
Total	255	100

For the empirical papers, a variety of samples or units of analysis were employed (Table 7). The highest frequency was a mixed sample (38.8 per cent), which meant that researchers were drawing from more than one type of participant group. Practitioner samples were also frequently used (19.4 per cent) and to a lesser extent case records, administrative data (15.7 per cent), parents (6.7 per cent), carers (6.7 per cent) and children (5.9 per cent).

Table 7

Unit of analysis or participants for empirical studies

Unit of analysis	Frequency	Percentage
Mixed sample (2 or more participant types)	52	38.6
Practitioners/clinicians/professionals/service providers	26	19.3
Record analysis/administrative data/files	21	15.5
Parents	10	7.4
Carers	9	6.6
Children/young people/early adults	8	5.9
Other adults (e.g. care leavers, adult survivors)	5	3.7
Family/family group meetings	2	1.5
Other professionals	1	0.7
Systematic review of literature	1	0.7
Total	135	100

A count was taken if the paper stated the research was funded by the main national competitive grants available to social work researchers, the Australian Research Council, Australian Postgraduate Awards Scheme, and the National Health and Medical Research Council. Shown in Table 8, thirty-four papers acknowledged funding from these sources, representing 13.4 per cent of all papers. A number of papers stated they were funded from sources such as state governments, universities and not-for-profit organisations.

Table 8

Funding source

Funding source	Frequency	Percentage
Australian Research Council (ARC)	29	11.4
Australian Postgraduate Award (APA)	4	1.6
National Health and Medical Research Council (NHMRC)	1	0.4
Other source	13	5.1
Not stated	208	81.6
Total	255	100.0

Discussion

The protection of children from abuse and neglect is a complex social problem. Child protection services reach into the lives of thousands of children and families each year, yet many programmes are not adequately aligned with current knowledge and theory, practice models are contested and not enough is known about the outcomes of different types of intervention. The quality of child protection services is routinely subject to media and public criticism. The growth and development of an Australian child protection research base that includes a strong contribution from social work are vital to understanding current dynamics and improving innovation and responsiveness.

Number of researchers and quantum of research

The finding that the publication output is spread between many researchers has both benefits and limitations. A large number of researchers contributing to child protection discourse potentially increases the likelihood of multi-perspective analysis. It may reflect that many people have an interest in child protection scholarship. Conversely, it may indicate that social work researchers do not have long-term careers in this field, or they are not productive researchers. The low level of publication from doctoral research noted in the USA is likely to also be problematic here (Maynard et al., 2014). Publication is a vital part of research because it opens up findings to peer review and broader audiences, inside and outside academia. Only fifteen authors had published more than six articles during the period (an approximate rate of one per year), whereas 195 authors had published one paper only. Large numbers of authors with one or two publications also suggests that there are few sustained programmes of research that involve the accretion of knowledge over time. Programmes of research—whether pursued by lone researchers or teams—can provide depth and focus, resulting in more substantive outputs. Too many one-off studies is concerning because it suggests a ‘cottage industry’ approach wherein researchers pursue idiosyncratic topics of interest that are not germane to important questions in the field, do not address strategic challenges or do not have international significance. This highlights the necessity to develop strategies that support research communities or teams of scholars, and more research training that builds the scientific community. Building critical mass and a vibrant intellectual environment in a small country is not easy, but more could be done. There is scope to enhance partnerships between governments, child protection agencies and schools of social work to build research teams and proposals, and advocate for investment in the important, longer-term research projects that are needed.

In relation to quantum, there are no benchmarks available for social work outputs in child protection, so we cannot conclude whether the number of papers published over the time period was low or high. It may be of concern that the number of papers from social work child protection researchers did not increase each year over the period, as the number of Australian research outputs (especially journal articles) increased markedly during the period following the first national Excellence in Research Australia (ERA) research assessment exercise in 2010, and social work overall (code 1607) increased from 961 articles in 2010 to 1,264 articles in 2012 (Australian Research Council, 2010, 2012). Note that authors of these papers are not necessarily social workers, and in part this increase can be attributed to more academic staff and better data collection. Nevertheless, we hope that the data presented in this paper will promote discussion and reflection about social work research publication rates.

Quality and impact

Publication of research in both national and international journals shows that Australian research has relevance for domestic and global contexts. Australian child protection is well represented in Australian journals and also published internationally, which suggests that Australian social work researchers are not parochial, or have a purely ‘inward gaze’, they demonstrate a global outlook and their research is relevant to international audiences (Shaw, 2014, p. 524). Australian research had slightly higher uptake in UK compared to US journals. The research was published in a broad range of journals, not only those discipline-specific to social work. This is also a positive result, as it indicates that the research is crossing disciplinary boundaries in its contributions to understanding vulnerable families and formulating equitable social policy.

Shaw and Norton (2008) described two features of research impact: intrinsic signifiers of quality (e.g. a well-considered and argued epistemological and theoretical position) and extrinsic signifiers (incorporating value-for-people, value-for-use and community and peer receptiveness to the research). Although the purpose of this scoping review was not to evaluate the quality of the publications, many papers did not clearly identify the method, purpose and type of research presented. Issues of quality in social work research have been noted by others (Ryan and Sheehan, 2009; Sharland, 2013). Citation counts for publications, while imperfect, also provide an indication of research quality and peer influence, but citation analysis was not undertaken in the present study. Extrinsic indicators of the impact of social work research refers to how it influences, changes or benefits social policy, human services and people's lives. The extent to which Australian social work research impacts upon child protection policy, practice and outcomes is unclear. Research utilisation studies have tended to concentrate on evidence-based practice (using interventions proven to be effective) and, while this is important, child protection is not just a sequence of interventions or programmes that are delivered to children and families. Research also generates knowledge in the form of ideas that can lead practitioners and policy makers to think about problems in new and different ways, and disrupt traditional thinking about issues (Nutley et al., 2007). More nuanced examination of the broader impact of Australian social work research in the child protection field is needed that goes beyond the take-up of evidence-based programmes. Child protection is a contested field, with research being only one potential influence on policy and practice. But it is especially important to pay attention to research in order to moderate the over-critical or ideologically driven responses that can emerge in situations of crises or media scandals (Lonne et al., 2009).

Topics

The topic of out-of-home care attracted the most research attention from social work, consistently with the trend identified in the 2011 Australian audit (McDonald et al., 2011) and with a broader tendency in social care research to focus on people who receive services (clients or service users) rather than those who miss out or drop out. Papers examining child protection systems also received attention compared to other issues. Given that social work aims to contribute to the analysis and critique of social and welfare systems, so as to facilitate more equitable arrangements for individuals and groups, attention to this issue is to be expected. While other disciplines may be addressing the topics that received minimal social work attention, a social work perspective on these strategic areas would be welcome, particularly if social work aims to acquire equivalence with other professional groups and exert impact on the direction of child protection policy.

Methods

The preponderance of qualitative studies has been noted in other social work research audits, not just in child protection (McDonald et al., 2011), but also in health (Brough et al., 2013), disability (Llewellyn, 2014) and social work more broadly (Ryan and Sheehan, 2009). An analysis of research papers (2000–04) in the British Journal of Social Work reported that 33 per cent of the studies were qualitative (McCambridge et al., 2007). While noting the considerable overlaps between the categories of qualitative and quantitative, and above all the need for high-quality research, a balanced portfolio of approaches and methods would generate research capable of answering a broader range of questions. Both qualitative and quantitative methods are essential to the development of an emancipatory social science that is capable of generating useable knowledge about the parameters and consequences of social inequality, and different types of evidence about the effectiveness of professional interventions (Oakley, 2000). Therefore, it is important for research funders and industry partners to appreciate that, in striving for evidence-based practice, we do not overlook a range of methods. Research does not have to measure outcomes or test interventions in order to be useful in practice. There are many types of empirical research essential to child protection practice moving forward, including longitudinal designs, cohort studies, comparative studies, administrative data linkage, policy analyses and prevalence studies.

Data sources for empirical studies

The units of analysis for the empirical research showed that mixed samples were most often used (38.6 per cent), followed by practitioners and service providers (19.3 per cent). Positively, the use of mixed samples suggests researchers are investigating issues from multiple perspectives. However, the more limited inclusion of child and parent perspectives (5.9 per cent and 7.4 per cent, respectively) tends to position them as merely the objects of services or research. Social work research has an important role to play here, in representing the voices of less powerful policy actors.

A high proportion of papers in this review were non-empirical (47.1 per cent). Theoretical debates and critical analyses have a valuable place in scholarship because systematic analysis of accepted practices can lead to new ideas percolating. However, their overabundance as a knowledge source can potentially limit the defensibility of the knowledge base. In striving for practice that is knowledge-guided, policy makers and practitioners also require quality empirical research, including evaluation research. Brough and colleagues (2013) in the health field argued that limited attention to evaluation puts social work in a tenuous position in that the profession can be criticised regarding the effectiveness of its practice, and evaluation may occur but without the influence of social work contribution.

Limitations of the review

This scoping review had a number of limitations to take into consideration. Potentially relevant research may have been missed because of the exclusion of grey literature, books, book chapters and dissertations. But peer review and publication are vital links in the research chain and it is increasingly important that research is accessible on the internet. The identification of relevant papers was dependent on the keywords used by authors being picked up by database search terms. The screening and selection process of importing records into Endnote and employing a journal limit strategy may have also excluded relevant papers published in journals that were not included. An additional limitation was the lack of clarity in many papers on research method, focus and purpose which then required determinations to be made by the coder which may not have reflected the intent of the authors. The qualifications of the authors were not always stated, and web searches were used to ascertain this if not clear, but errors may have been made. Nor was information available about the demographic composition of the social work research workforce (such as gender, ethnicity and age). Despite these limitations, we are confident that this study provides an accurate representation of Australian social work child protection research for the period under review.

Conclusion

In focusing on social work research in this scoping review, we do not diminish the contributions of other disciplines to the child protection field or the inter-disciplinary nature of the social work knowledge base. However, it is vital to develop a picture of the distinctive contribution of social work within the broad range of social science research and to make social work research visible to others. This review has identified a number of features of social work's contribution to Australian child protection—its strengths as well as its weaknesses. To maintain its place as a lead discipline in this field, and its impact on policy and practice, the profession and social work educators need to critically reflect upon and find ways to strengthen its research foundations, including consideration of research training. Social work research in Australian universities is developing, but there is room for improvement. The federal government initiative to assess the quality and impact of research in Australian universities, ERA, is an important driver of research development. The 2012 ERA results for the social work field of research (code 1607) indicated that overall social work was a mid-level performer in research outputs, with ten out of sixteen universities rated at or above world standard (rated 3 or above on a five-point scale) (Australian Research Council, 2012). Eleven universities that offered social work degrees reported no social work research outputs. But the ERA report does not provide a census of social work research, as social work research could have been categorised in other fields such as policy and administration, public health or criminology. In any case, reducing research performance to a single digit does not provide a sufficient basis for assessing quality or impact in the field. More comprehensive data about research in the discipline and the field are needed to interact constructively with the policy and research environment.

Is social work research in the child protection field in Australia distinctive? Undoubtedly there is a strong commitment to debate, critique and challenge of child protection systems and policies. Advocacy for change demonstrates the critical underpinnings of the discipline and the importance of social justice. The use of qualitative research methodologies indicates a research community that values the exploration of lived experience, contexts, complexity and interpretive insights. Studies based on multi-perspectives and mixed analyses which can offer depth and refrain from privileging particular ways of generating knowledge are present, but a smaller part of the picture. The trends highlighted here provide an empirical base for the discipline to critically examine how and in what way it can advance child protection research. There are many types of research essential to policy and practice moving forward. It is vital for social work that its research contribution is contemporary, relevant, robust and credible.