2001), published as we completed our own study, which found no clear evidence of the impact of a carer training programme on foster children's emotional and behavioural functioning. Our enquiry lends support to their valuable findings; however, our research design differs in many respects and will reveal different insights into this complex field. We return to Minnis and Devine later but first we set out the parameters of our own investigation, which recognized that challenging behaviour is not a single entity and that training in this field cannot be made to fit all needs. In this regard it was important to embark on our research with a clear grasp of the sorts of difficulties typically encountered in the research setting in question. Hence, some initial profiling of current problems was needed early on to examine if the sample contained predefined clinical characteristics or whether challenging behaviour was more a matter of professional judgement by staff and carers. It soon became clear that most children so defined were not the subject of clinical definition and thus it became important to apply a checklist in order to identify the study population. Our checklist assumed a wide and multiple range of challenging behaviours such as absconding, theft, self-harm, violence, attention deficit and hyperactivity disorders
A B S T R AC T This paper sets out key features of a semi-experimental investigation conducted between March 1999 and April 2001 into the impact of training foster carers in techniques to manage challenging behaviour.We identified a sample of children (n = 103) in foster care in four authorities in South Wales who were selected by a checklist as showing behaviours that could be defined as challenging. From this search procedure we identified their foster carers (n = 106) and created a training intervention group for 53 carers looking after 49 children and a non-intervention comparison group of 53 carers looking after 54 children. Both groups of carers were interviewed using a range of measures before and after the delivery of the training in order to identify any changes they perceived in the conduct of children and changes they perceived in their own capacities as carers. The training was delivered in small groups and evaluative feedback was gathered from carers and from training officers in those agencies that participated in the study. The results suggested that training, as designed and delivered in this study, had limited impact on child conduct and carer capacity. The reasons for this are discussed in light of selected findings. We spend some time on describing our methods in order to expose our approach to the critical scrutiny of readers, whose views would be welcome as we make our future research plans in this field.