When foster children's behavioral problems become unmanageable in community settings, residential treatment centers provide a placement option. Upon entering residential treatment, foster children's social relationships may change more or less drastically. The nature of their social relationships and how they change remain unclear; yet, social relationships likely influence children's behavior both positively and negatively. Using Kahn and Antonucci's (1980) convoy model of social relations as a theoretical and methodological framework, this study examined changes in the structure and function of foster children's social relationships ("social convoys") during their first 3 months in residential treatment and how those relate to behavioral problems.

This study used a prospective longitudinal survey method to assess social convoys and behavioral problems of 9 to 13 year-old foster children in two residential treatment centers. The researcher administered Children's Convoy Mapping Procedure to children and Youth Outcome Questionnaire 30 to child care workers within one month of intake and 3 months later. Data analysis techniques included social network methods, multilevel models, and cluster analysis.

Foster children in this study included a variety of close and important individuals in their social convoys. At baseline, four distinct types of social convoys emerged: balanced-supportive, family-focused more support functions, family-focused fewer support functions, and peer-focused. Over the three months of treatment, children reconstructed their social convoys by adding, keeping, and removing individuals who could or could not provide social support, especially long-term tangible aid of providing care in crisis. Participants with family-focused fewer support functions social convoys, tight-knit networks of family members that provided relatively less support, had the best behavioral outcome.

The results suggest that social convoys of foster children change in many different ways during residential treatment. The environmental attributes of residential treatment, both at the time of transition (e.g. multiple caregivers) and during treatment (e.g. resident turnover), may influence the way children reconstruct their social convoys. This study demonstrated how the convoy mapping method can track short-term changes in foster children's social relationships and has promise for allowing practitioners to regularly assess children's social convoys. Such assessments may provide points of intervention based on the children's perspectives.