Evaluating needs-based home visiting support: Can administrative data help?

Objective Home visiting is a form of family support which can help families with different problems in different ways. Previous evaluations have utilised either experimental or qualitative designs. However, the needs-based nature of support presents a challenge for evaluation using experimental designs. Main Aim This paper illustrates the unique contribution that administrative data can make to understanding these problems and how it can be used to explore what support works for families in different situations. Methods The analysis of administrative data from one UK home visiting organisation, Home-Start, is presented. Exploratory analysis considers measures describing how well parents are coping with a range of issues and how this changes over the course of support. This highlights problems with evaluation because of the variety of issues Home-Start is supporting parents to cope with and the fact that the duration of support is needs-based. Methodological solutions are proposed for these problems using the administrative data. These include using subgroups to study families with different problems and considering the rate at which improvements occur as an outcome variable. Linear regression models are presented to demonstrate how these methods can identify aspects of support related to improvements in parental self-esteem. Results The methods used are able to demonstrate that the frequency of support and who the support is provided by are related to faster improvements in parental self-esteem. The analysis of sub-groups in the data shows that the frequency of support is important for all parents, but there are differences between parents in different situations, depending on whether volunteers or paid staff provide support. Conclusion The analysis of administrative data is able to make a unique contribution to the evaluation of needs-based home visiting support.