Jessica, a residential inhabitant (14 years of age), tells an interviewer about the ‘ideal’ residential care worker: “The ideal care worker is someone who is thoughtful and easy-going, and with whom you can laugh a lot. S/he listens very carefully to the children, takes them serious, helps and supports them, and does a lot of things together with them. The ideal worker never acts harsh” (Meerdink, 1999, p. 27).

This quote comes from a research in which residents have been asked to tell what they think of the care they received, including their living in a group. It turned out that the favorite workers are the ones who participate as equals of the young people, and who ‘follow’ the children in what they plan to do. We talk about things like watching a movie together, shopping, playing games, doing sports, helping children with their school work, organizing activities, cooking together, et cetera. It especially concerns ‘normal’ things and activities, not highly specialized treatment techniques.

A Dutch review concerning seven follow-up studies of former inhabitants of residential group care settings and day treatment centers reports that, on average, a bit less than half of the children (44%) show significant progress; the problems that brought them into care clearly have been diminished or solved (Veerman & Ten Brink, 2001). This is good news and – considering the seriousness of the emotional and behavioral disturbances of the children in all the samples – a major accomplishment. At the same time such a result does imply that in the other cases – a bit more than 50 % – the problems seemingly had not been solved, and in some cases even got worse.

Outcomes like this lead to the question what factors cause or promote a significant decrease of child and family problems. It is the question often articulated as ‘What works?’ (cf. Fonagy et al., 2005). In our opinion helping to answer this question is a most, maybe the most important mission of researchers in the child and youth care field. It is a very complex mission but it is not an impossible one.

An example in a nutshell concerns an empirical study of our Dutch colleagues Van der Ploeg and Scholte (2003). Information on this study can also be found in the Appendix.

In this project outcomes had been investigated of nine ‘promising’ residential programs. A program has been assessed as ‘promising’:

• if key elements of the program (like characteristics of the population to be served and the treatment strategies to be applied) have been described very precisely;

• if the program is based on a clear treatment theory or model; and

• if the program is fully embedded in the organization.

In total 150 children with severe emotional and behavioral problems – mean age 15 years of age – were followed up during one year. Their functioning was monitored using standardized instruments (like CBCL) as well as interviews with residential care workers and the young inhabitants themselves. Treatment goals that were aimed at and the pedagogical approach that was to be applied were also documented.

The study showed that – beside the creation of a social climate of ‘safety’ in the group – the two most frequently recorded treatment goals were:

• strengthening the personality of the child, i.e. improving his/her psychological well-being;

• improvement of social skills of the child or youngster.

Concerning the pedagogical approach chosen the aspects of regulation and structuring the life and behavior of these children had been found most often.

The children who had left the residence after a year were asked to evaluate their stay. More then half of them (58%) were satisfied, near a quarter (23%) was moderately positive (cf. Table l). The residential workers who had been interviewed reported progress in half of the cases. The standardized instruments (like CBCL and NPV-J, a Dutch personality questionnaire) showed the most ‘critical’ results: they register improvement of child behavior in only a quarter of the cases. This shows the meaningfulness of using different instruments in outcome research.

As can be observed in the Appendice the outcomes of the nine programs in this study were quite diverse. An important additional plus of the research design was the opportunity to compare programs at several levels: input, throughput and output (benchmarking). By taking account of client and intervention factors in the design, the impact of these factors could be explored in connection with the outcomes.

Because the study only in broad terms described what kind of care and treatment should have been implemented per child, we do not know what — on a concrete level — has been the interaction between child and worker. And this seems to be equally important as will be shown with the next example.

Some years ago a review study was published, titled The Heart and Soul of Change (Hubble, Duncan, & Miller, 2002). Most interesting in this reader is a contribution of Michael Lambert. Looking for explaining factors in ‘what works’ research, he discerns with his co-writer Ted Asay, four types – named as ‘the big four’ (cf. Asay & Lambert, 2002; also Lambert, 1992), namely:

• extra therapeutic factors, in fact client and problem characteristics;

• relationship factors;

• the factors expectancy and hope of the client;

• the therapeutic factor, i.e. the therapeutic model and techniques.

Based on meta-analysis of outcome-studies on psychotherapeutic interventions Lambert concludes that improvement of the client (personal change) should be attributed for a main part (40%) to the first factor: client characteristics (for instance ego strength, intelligence or motivational state). Second in the ranking are relationship factors (30%), and following this the other two factors, expectancies (15%) and therapeutic techniques (15%) were mentioned. Although the review is not especially made up of child treatment studies, we would like to translate the ‘big four’ to our field.

Doing research on outcomes in residential care and treatment programs implies that attention should be paid to at least the following factors:

• child, family and problem characteristics;

• the treatment approach and techniques;

• the relationship and interaction between child and care worker;

• the expectations of children and their parents concerning the care to be received.

The third and fourth factor – relationship and expectancy – have not been studied much in child and youth care. Still they might be crucial in our understanding of factors explaining ‘what works’. So our recommendation would be to focus on these factors, supplementary to the kind of research done by our Dutch colleagues. It means that more research ‘in depth’ should be done; qualitative studies wherein clients and workers explain their experiences and perceptions. Like Jessica who, in two sentences, said more than can be found in ten books.

Asay, T., & Lambert, M. J. (2002). The empirical case for the common factors in therapy: Quantitative findings. In M. A. Hubble, B. L. Duncan, & S. D. Miller (Eds.) (2002). The Heart and Soul of Change. What Works in Therapy. Washington, DC: American Psychological Association.

Fonagy, P., Target, M., Cottrell, D., Phillips, J., & Kurtz, Z. (2005). What works for whom? A critical review of treatments for children and adolescents. New York/London: The Guilford Press.

Hubble, M. A., Duncan, B. L., & Miller, S. D. (Eds.) (2002). The Heart and Soul of Change. What Works in Therapy. Washington, DC: American Psychological Association.

Lambert, M. J. (1992). Implications of outcome research for psychotherapy integration. In J. C. Norcross, & M. R. Goldstein (Eds.), Handbook of Psychotherapy Integration (pp. 94-129). New York: Basic Books.