The contribution of qualitative approaches to developmental intervention research

Unique and indispensable

In this special issue, we present four articles that illustrate some of the contributions that qualitative approaches can make to intervention research and thus carry forward social work’s commitment to effective practice. We intend these articles to illustrate possibilities and to stimulate other researchers and funders to take seriously the usefulness of qualitative approaches to the design and development of interventions. From the inception of intervention research until now, researchers have emphasized the importance of qualitative research to the design, development, testing, and reformulation of interventions. From our points of view, the contributions of qualitative approaches are unique and indispensable.

Social work intervention research is significant to the extent that it informs practice in constructive ways. In the call for papers for this special issue, we defined intervention research as:

Research that involves the development and testing of practice models, descriptions of change processes, and the application of models of practice to new populations and contexts. (Gilgun and Sands, 2010: 569)

In the present article, we first share our experience and thoughts about editing the special issue and describe the four articles on intervention research. We have learned a great deal about intervention research and hope the accounts of our experiences stimulate creative thinking in others. We then share the thinking and scholarship of other researchers who have done so much to prompt social work researchers to put intervention research at the center of their research efforts. We join them in stating that the design, development, and on-going reformulation of interventions are central tasks of social work research.

Design and development

Through our experience in editing this special issue and our previous experience with the development of interventions, we have concluded that intervention research involves ongoing design and development (D&D). D&D usually requires teamwork among researchers, service providers, service users, and, often, funders, legislators, and other stakeholders. In some instances, a single researcher builds interventions from a succession of case studies (Reid, 1994), and there is no team.

Design and development is composed of several parts, which, although we present them in linear fashion, may better be visualized as a series of activities that loop back to previous activities, as in a spiral that evolves over time. In short, there are on-going interactions of the various parts of D&D and, therefore, on-going and growing understandings of each segment that, in turn, informs and further contributes to the development of other segments.

The design and development of interventions build upon in-depth understandings of problematic situations for the purposes of changing them. We define problematic situations as interactive contexts in which individuals live their lives over time and where there are injustices, situated vulnerability, and human suffering. The design and development of interventions that show promise of effectiveness require descriptions of problematic situations that are as comprehensive as possible. Such comprehensiveness would include descriptions of how persons view and experience their circumstances in the past, present, and future; the views of service providers familiar with the problematic situations; the multiple influences that affect persons in their various environments; a description of adversities and risks as well as of positives, such as protective factors and resources; and a description of the scope of the problematic situations. Much but not all of the description of problematic situations requires qualitative methods, such as in-depth interviews, observations, and case record reviews.

From these descriptions, the research team develops straightforward goals that are composed of various dimensions of problematic situations and what each of these dimensions would look like if the problematic situation were changed for the better. For example, children who have experienced abuse and neglect often blame themselves for the maltreatment, think they did something to deserve the maltreatment, and come to believe that they are worthless and bad. If the description of this problematic situation were comprehensive, the development team would have information on factors that influence children not to have these beliefs in addition to factors that contribute to these beliefs. The team would state goals in terms of fostering children’s sense of themselves as good and worthy and not as bad and worthless.

The design and development of effective interventions requires familiarity with interventions that others have used, including what worked and what appeared not to work with whom and under what conditions. Thus, teams would review existing interventions that are relevant to the problematic area in which they are interested. Following this, the team would integrate the understandings they developed from the descriptive material with the understandings they developed from their review of interventions. Next, they would craft interventions that are sensitive to the experiences of the persons for whom the intervention is designed.

The next logical steps are piloting the intervention on a small sample, careful evaluation of the pilot, applications of the results of the pilot to any reformulation of the problematic situations and of the interventions that the pilot has evidence for, and then implementation of the intervention on larger samples. Service user and service provider experiences, appraisals of the pilot, and the implementation with larger samples are indispensable for tracking effectiveness/efficacy as well as for identifying issues that require attention.

Qualitative approaches have central roles in these evaluations, as Reid (1994) pointed out almost 20 years ago. Observing and interviewing service users and service providers are typical ways of understanding implementation issues. From these procedures, research teams learn how service users respond to various aspects of interventions and how service providers appraise, experience, and implement the interventions. This is information that research teams feed back into the intervention and into their understandings of problematic situations. Throughout D&D, therefore, qualitative research has a role, including evaluations of randomized controlled trials (RCTs). Like any other phase of design and development, research teams designing and implementing RCTs require information on the variations of responses that persons have to interventions and the circumstances and characteristics of persons and environmental influences that are significant for understanding efficacy.

In some design and development projects, research teams construct assessment and evaluation tools. Such tools are a phase of some intervention research projects. As is becoming increasingly clear, qualitative research makes unique contributions to the conceptual bases of such tools (Gilgun, 2004). This has been the case traditionally for many tools, such as when researchers rely not only upon reviews of the literature, but also upon the clinical expertise of service providers and interviews with persons who are members of the class of people for whom the tools are designed (see the work of Levin and Peled, 2011, for a recent example).

The four articles of this special issue

As this discussion shows, D&D is a complex process. Each of the phases of D&D has multiple dimensions and feedback loops that require on-going evaluation and reformulations. A special issue on the contributions of qualitative approaches to intervention research can cover a limited number of the dimensions of D&D. The first four articles that compose this issue are a sample of the possibilities. We look forward to other D&D research that uses qualitative approaches in these and other ways.

The articles in this special issue have in common the goals of understanding and intervening into problematic situations. These situations are child sexual abuse, care provisions of orphaned children with AIDS in Lesotho, Africa, dating violence among Mexican American youth, and service engagement of formerly homeless adults. All four articles report upon the perspectives of service users and the interactive contexts in which service users live their lives. Two of the four articles also documented the experiences of service providers. In pursuing these goals, the researchers encountered first-hand the complexities of such endeavors.

The clarity of the accounts that the researchers wrote may give the impression that what they accomplished is easy to do. The contrary is more accurate. Understanding persons in their interactive contexts requires that researchers have keen observational skills, sensitivity to research participants, the capacity to manage their own complex emotions that arise in the course of doing qualitative research, and the analytic skills to identify core concepts that underlie the particular processes and interactions in which they are interested. This is not a small set of tasks.

Three of the four articles involved various kinds of design and development of interventions, while the fourth sought to understand the perspectives and interactions of service users and case managers in a housing program that summative evaluation had already found to be more successful than other housing programs. Each of the articles shows how researchers looked for positive elements that could be considered protective factors or service user strengths as well as problematic issues, recommendations that are prominent in Fraser et al. (2009). After the identification and description of these issues, the research teams used them in the design, development, and/or modification of interventions. The fourth, as already stated, had the goal of documenting positive interactive processes in a successful program using qualitative approaches.

As is the case for most qualitative research, the findings reported in the four articles are the most salient in terms of other findings authors could have reported. What is in the articles is based upon a great deal of material for which there is insufficient space in typical academic journals. In addition, some readers may find themselves mystified that some of the authors present their findings in an almost casual or conversational style, in contrast to the highly structured reports that are characteristic of other types of research. Readers who are able to engage with the material with a spirit of openness that characterizes most qualitative research will find the articles informative.

While the emphases on participant perspectives and on accounts of the complexities of participant situations are consistent across studies, there are variations in the four articles as well. Andrew Hill shows how he, as an academic researcher, worked with a treatment team that had decided to modify its practice with children and young people who had experienced children sexual abuse. After years of emphasis on building relationships between therapists and children to the relative exclusion of parents, the team gradually realized the importance of involving supportive family members. Hill joined the team as the service providers were poised to begin re-designing the intervention. Through qualitative methods that involved interviews with parents and evaluations of the change processes involved in the new interventions, Hill developed a theory of change, which involves specifications of chains of causality and/or associations (Maxwell, 2004; Patton, 2011) that take place during interventions. This team approach to intervention research exemplified many of the phases that are characteristic of D&D intervention research.

Hill worked closely with service providers. The re-design and development of the intervention shows that service providers are sources of ideas for practice innovations. This research is also an example of how practice-based evidence, perspectives that arise in the course of engaging in practice, contributes to evidence-based practice. Hill’s research is noteworthy in light of the views of service providers on research. Beddoe (2010), for example, found that many service providers want to do research, including the write up of successful cases for which they had a great deal of enthusiasm. The service providers identified many barriers to such efforts. They often are short of confidence in their capacities to do research, find little encouragement from administrators, and lack the time, resources and training. When service providers team up with academic researchers who have expertise in the design and development of interventions, innovative research and practice-based interventions are possible.

Ellen Block reports on the results of the ‘deep ethnography’ she conducted for 13 months in Lesotho, Africa, for the purposes of understanding issues related to caregiving for children whose parents had died of AIDS and many of whom had AIDS themselves. She used interviewing and participant observation to immerse herself in a culture with which she was not familiar. Block considered her research to be knowledge development, or information that contributes to understanding problematic situations, including the identification of resources and protective processes. Her goal was to understand the context of care so as to develop recommendations for the modification of existing intervention programs. One of many situations that was salient in Block’s understanding of the systems of care was the death of a toddler from dehydration because the grandparent caregivers did not know how to treat such a situation. She knew the family well and had evidence of the quality of care that the grandparents provided. They simply did not know that toddlers could die so quickly from dehydration associated with diarrhea. She also noted the great distance the grandmother had walked to seek help, only to find that the clinic was closed that day. From this and other on-the-ground experiences, Block was able to formulate recommendations for the education of care providers. Block also educated herself about the policies that affected the provision of services, which is essential when researchers want to make policy recommendations themselves. This article shows the importance of immersion in the field and the use of interviews and observations to understand problematic situations.

Through focus groups, Lela Williams, Heidi Adams, and Bianca Altamirano sought to understand Mexican American young people’s perspectives on dating violence for the purpose of modifying existing programs. Williams and co-researchers conducted literature reviews on the issue of teen dating violence and on interventions others have developed and tested. They noted that little information on Mexican American teen dating violence existed. The research team chose focus groups because of their fit with typical developmental trajectories of adolescents, which is their identification with peers. The researchers thought that group interviewing would be more productive if they organized them by levels of acculturation. Their goal was to inform the design and delivery of culturally grounded intervention programs or modification of existing programs for Mexican American youth, and to identify barriers. The students generated many ideas, and they also identified barriers to participation. Such material is foundational to the adaptation of existing programs.

Victoria Stanhope reports on a study that used ethnographic methods, including interviews and observations, to account for, understand, and document the activities and experiences of service users and case managers in a housing program that summative evaluations had already shown to be far more effective than other housing programs. She examined the contexts and processes of implementation of case management services for service users with long-term street homelessness and co-occurring mental illness and substance abuse disorders. Using symbolic interactionism as a conceptual framework, she focused on interactions between service users and providers as a means to provide process knowledge related to the implementation of an established housing intervention. Her research can further the ongoing development and refinement of the model, inform the application of this model in new settings, and contribute to the training of case managers.

These four articles, then, represent the fulfillment of many of our hopes for this special issue. All four used qualitative methods to understand the experiences of service users, and two of the four to understand the experiences of service providers. Three of the four used these experiences as a basis for developing, adapting, or modifying existing interventions. All four used social work perspectives, such as stating where clients are and person-environment interactions, and they also implemented social work values, such as autonomy, dignity and worth, and social justice.

What we learned that we did not expect

In editing the special issue, we found that many researchers who submitted articles did not understand the meaning of the term intervention research. Typically, they equated intervention research with summative evaluations, which focus on outcomes. Few cited references to Thomas, Rothman, Fraser, and others, even though we defined the term as these researchers do and mentioned the work of Thomas and Rothman in the call for papers. We began to wonder whether the idea of intervention research has not had widespread diffusion among social work researchers. We knew the term had little meaning to many of the researchers who submitted articles to the present special issue. In addition, Ian Shaw, editor of Qualitative Social Work and professor at the University of York, York, UK, mentioned to us several times that in England the term intervention research means evaluation research. From early on, researchers who promoted the development of interventions pointed out the rarity of research that does so (e.g. Fraser et al., 2009; Rothman, 1980; 1989; Rothman and Thomas, 1994; Thomas, 1978a, 1978b, 1984). They stated that most if not all social work researchers appear to have graduate training in traditional approaches to social science research, which does not include figuring out how to use the results of such research in the design and development of interventions.

As we thought about our experiences as editors and the observations of others, we began to wonder whether part of the issue is in training. Both of us are seasoned researchers and did our research to contribute to social welfare and therefore to contribute to interventions. Yet, neither of us has participated in the design and development of interventions, despite our long-term enthusiasm for developmental intervention research. Somehow our rhetoric did not connect to reality (cf. Reid, 2002), although editing this special issue is a step in that direction.

We also wondered if the term intervention research itself is an issue. Without knowing the specialized definition of the term, researchers would assume that intervention research means research on interventions, which logically would involve outcome or summative evaluations and perhaps evaluations of processes. The term has no connotations relating to the ideas of design and development, which, from our experience, is the core of what has been called intervention research.

We began to wonder if the connotations of the term intervention research is too narrow in the scope to represent the range of activities in which social workers engage when we attempt to transform various sources of knowledge into interventions and other innovations that contribute to practice excellence. We think that Thomas’s (1978a, 1978b, 1984) term, developmental research is more accurate than intervention research in that it encompasses the idea of development not only of understandings of persons in interactions with various environments and of the environmental influences on individuals, but of the development of other intervention-related technologies, such as assessment and evaluation tools. Such tools also contribute to interventions and to social work practice effectiveness but are not themselves interventions.

In addition, the terms development and developmental capture current thinking about the nature of social interactions as composed of complex systems that continually undergo change (Patton, 2011). Interventions designed to influence complex systems are effective to the extent that they are responsive to the interactive elements that compose the systems. Given that systems continually change, interventions meant to influence them also must continually change, or, in other words, practitioners must continually adapt and re-develop interventions in response to systems changes.

For these reasons, we propose the term developmental intervention research, which has connotations of on-going development, reflective of the complex systems that social workers want to influence. This term may come to have more widespread penetration than the term intervention research.

Scholarship on the design and development of interventions

As a topic of formal discussions complete with explicit procedures and rationales, the design and development of interventions has, at the minimum, a 30-year history in social work. As discussed earlier, Thomas (Thomas, 1978a, 1978b, 1985, 1989) did pioneering scholarship on an approach he called developmental research, which is a way of generating innovative social technologies that can be used in human services. Thomas adopted the word technology from engineering where a research and development (R&D) approach is used to develop new products and procedures or to modify existing ones. The kinds of social technologies that Thomas (1978a) envisioned for social work included physical structures, information systems, assessment methods, intervention methods, and service systems, among others. As discussed earlier, in contrast with traditional behavioral science research in which the goal is to contribute to knowledge-building, the aim of developmental research is to generate new models for practice, service delivery, and policies that can contribute to the attainment of the objectives of social work (Thomas, 1978b, 1985). The focus on generating innovations is why Thomas considered developmental research to be different from conventional social science research.

Thomas (1985, 1989) described four phases of development activity – analysis, design, development, and evaluation. Analysis refers to identifying and examining a problem that could be prevented or remediated through social technology, reviewing information sources, and assessing feasibility (Thomas, 1989). Design, the formation of an intervention, ‘is the planned and systematic application of relevant scientific, technical, and practical information to the creation and assembly of innovations appropriate for interventions in the human services’ (Thomas, 1985: 51). The third phase, development, involves implementing the intervention in a preliminary trial, assessing its adequacy, and making changes that researchers identify as necessary. Changes are then incorporated into future trials, which go through the same processes of testing and refinement. Evaluation, the use of behavioral science research methods to assess the working of an intervention, is built into the successive trials during the development phase. At the point in which the intervention appears to be effective and ready for use, researchers undertake a more comprehensive evaluation that includes applications and outcome studies under ordinary circumstances (Thomas, 1989). Besides the use of measurement tools, experimental designs, and program evaluation, evaluation methods can include empirically oriented practice and the assessment of intervention implementation (Thomas, 1989).

In the opening chapter of their edited book on intervention research, Thomas and Rothman (1994) described three types of intervention research – knowledge development, knowledge utilization, and design and development. Knowledge development refers to social and behavioral science research that aims to generate findings that are applicable to understanding areas that can benefit from intervention, relevant provider conditions, and the social environmental context; knowledge utilization has to do with converting knowledge to concepts and theories that can be applied to specific populations and problems; and design and development is a means of creating and developing innovative means to bring about change in the human service arena. Design and development includes creating an intervention model, modifying it through preliminary trials, and testing and retesting successive versions (Bailey-Dempsey and Reid, 1996). Thomas and Rothman, along with others (e.g. Fraser and Galinsky, 2010; Fraser et al., 2009; Reid and Bailey-Dempsey, 1994; Schilling, 1997), emphasize design and development. As Abell and Wolf (2003) stated, design and development frequently incorporates aspects of knowledge development and knowledge utilization.

Thomas and Rothman (1994) outlined six phases of design and development research – problem analysis and project planning; information gathering and synthesis; design; early development and pilot testing; evaluation and advanced development; and dissemination. Although these authors present the phases sequentially, the process tends to be nonlinear and recursive. (For examples of the application of their model of intervention research, see Rothman (1989), Bailey-Dempsey and Reid (1996), Comer et al. (2004), and Yoshioka (1999)). Fraser and his colleagues (Fraser and Galinsky, 2010; Fraser et al., 2009) modified Thomas and Rothman’s phases to reflect recent developments. Their five steps are: development of problem and program theories; specification of program structures and processes; refinement and confirmation in efficacy tests; effectiveness testing in practice settings; and dissemination of program findings and materials. The Fraser model is also developmental and can be recursive. It has the added features of theory, testing for efficacy (under ideal circumstances) and effectiveness (in natural settings), and the development of treatment manuals and training materials.

Thomas and Rothman (1994), Fraser et al. (2009), and others who describe and conduct intervention research, recognize that qualitative research plays an important role in the design and development of interventions. Schilling (1997) described research processes as moving from qualitative methods at the ‘front end’ to quantitative at the ‘back end,’ leaving open the possibility of returning to qualitative methods at the end. As he explained, qualitative methods at the end of the study can help investigators examine the effect of the research on the agency, staff, and clients and possible compromises to the research. Fraser (2004) suggested that qualitative researchers might ‘devise a change strategy, implement it, and describe the processes leading to outcomes’ (p. 220). He also suggested qualitative studies of successful and unsuccessful cases. Reid and Bailey-Dempsey (1994) used qualitative data (transcripts of case management meetings) and content analysis in a preliminary evaluation of a direct practice innovation.

At a national symposium on doctoral research that took place more than 25 years ago, Briar (1985) stressed that intervention research, particularly developmental research, should be given the highest priority in social work research. He made several points: 1) the results of such research are directly transferable to practice, 2) interventions of central concern to social workers are not likely to be studied by members of other disciplines, 3) developmental research can lead to practice improvement; and 4) developmental research engages social workers in knowledge development.

Summary and discussion

As the present special issue demonstrates, qualitative approaches have unique and indispensable roles to play in the design and development of interventions, an observation that appears to have become self-evident in social work. Examples of how qualitative approaches contribute to interventions, however, remain in short supply. The articles in the present special issue seek to redress this situation. As Reid (2002) said in another context, we as editors sought ‘to connect our rhetoric with reality’ (p. 12).

In the course of editing the special issue, we began to think that the term developmental intervention research may be a more fitting term than intervention research. With the addition of the word developmental, the design and development of interventions has connotations related to the on-going changes that are characteristic of social interactions and other complex systems that are the focus of social work’s efforts. The term developmental also connotes the on-going revisions required to craft effective interventions.

The term intervention research can and apparently does mean any number of things, such as evaluation research and generic research on interventions. In addition, for some researchers, practice research and clinical research may be synonyms for intervention research. The term developmental intervention research does not have the ambiguity of other related terms.

Our efforts at editing this special issue also brought to the foreground questions about social work graduate training in research methods. Social work is an applied discipline whose researchers seek to contribute to the development of interventions that carry forward social work’s commitment to effective practice. At this point in the development of social work research, we have more well-trained researchers than in any other time in history. Now, as others have in the past and more recently, we question whether social work research training has sufficient emphasis on the design and development of interventions. We do not have to give anything up if research training moves to include design and development. The research that social workers already know how to do – such as survey research, the analysis of large data sets, the conduct of experiments, quasi-experiments, evaluation research, and qualitative methods – are relevant to design and development of interventions.

Yet, few doctoral programs appear to guide students to use the procedures that connect their findings to interventions or, if their findings are results of evaluations, to connect their evaluations to understandings of problematic situations in the first place and to the reformulation of interventions. Our experiences as editors of this special issue, our review of scholarship, and our reflections on our own training and careers lead us to conclude that it is time for widespread conversations about the place of developmental intervention research in graduate research training in social work.

The present special issue editorial and the four other articles that follow this one in this special issue represent the tip of what is possible when social workers use information from a variety of sources for the design and development of interventions. As we hope we have shown, qualitative approaches have unique roles in the design and development of interventions.

A note on other features of the present issue

The final main article falls outside the scope of the special issue, although it has potential to contribute to research around intervention. Susan Bahn and Pamela Weatherill describe a relatively simple application of mapping methods to a study of organizational change. Readers will be aware of the extensive growth of spatial and visual methods in the social sciences over the last decade. These have been fueled in part by shifts in technology, but it is easy to forget how central photographic images were to early sociologists and anthropologists such as W I Thomas and Malinowski, and also to later writers such as Howard Becker. Visual methods such as mapping were also at the core of the work of the Chicago School in the 1920s. Bahn and Weatherill illustrate how applications are possible with minimal technological applications. They present visual organizational maps drawn by participants as a means of displaying current and ideal operational models, and as facilitating discussion of responses to change. They link their work to focus group methods and an action learning framework, and assess its advantages.

Finally, Mark Hardy provides an overview, followed by two excellent reviews of books that are relevant to social work research.