Evaluating children's services: Recent conceptual and methodological developments

This article considers recent innovations and challenges with respect to the evaluation of children's services interventions. It sets out a method for designing and evaluating services that is based on research evidence gathered on individual cases and that permits revision of those services in the light of emerging results. It starts by describing briefly different ways of identifying a discrete group of children with similar needs. It then outlines approaches to ensuring that an intervention for those children is underpinned by theory and research evidence-including the idea of 'logic modelling' before discussing the importance of capturing accurately how well the intervention is implemented. Experimental and non-experimental methods of assessing an intervention's effectiveness are considered, with particular emphasis on techniques for enhancing the causal inference that can be drawn from studies and the importance of matching method and purpose. The article also demonstrates how the principles and methods used in relation to evaluating services for groups can be applied in clinical work with individual cases, and identifies the benefits of this general approach to evaluation over and above the information that it provides on what works.