Challenges of and returns from large-scale, longitudinal studies with single homeless people

Rigorous, large-scale, longitudinal panel studies of homeless people generate authoritative evidence about the effectiveness of interventions but there have been few. This presentation examines the challenges and returns from such a study in England, FOR-HOME, of 400 single homeless people who were rehoused by six homelessness-sector organizations. They were interviewed just before they moved and after six and 18 months. The staff most closely associated with the rehousing also completed questionnaires. The first challenge was representativeness, for no national data on the population of interest are collated. The second was to recruit respondents, given that often only days separated notice of a housing vacancy and the move. In the early weeks, we learned about some moves only after they had occurred and missed others altogether. Prompt communication between the organizations' staff and the researchers took time to establish and maintain. Each organization nominated a Link Worker to assist with recruitment; nonetheless, it was persistently slow and had to be extended. The third challenge was to minimize attrition. The methods used to maintain contact with the respondents (many being marginalised and vulnerable) will be described, including the tracking system. The strengths of achieving a representative sample, of conducting extended semi-structured interviews, and of the resulting rich, multi-variable dataset will be outlined. The findings have challenged presumptions among policy makers and the provider organizations about the resettlement process and the services provided, but have been confirmed by operational data. The resulting credibility has encouraged government and provider agencies to request FOR-HOME data.