Abstract
This study set out to develop operational definitions of blame and exoneration, that could be used to code causal attributions about the presenting problem as they arise in naturally occurring conversations in child mental health services. Two studies were undertaken. Study One used the Leeds Attributional Coding System to code causal attributions about the presenting problem made by a sample of ten families and their therapists, and to categorise these attributions as either blaming or exonerating. The different patterns of blame and exoneration made by family members and therapists was investigated and the possible relationship between levels of blame and drop out from therapy was tentatively explored using Multidimensional Scalogram Analysis (MSA). The ten mothers and three children in the study tended to make attributions that blamed the referred child, whilst the two fathers mainly made attributions about the difficulties that exonerated the child. The five therapists tended to make attributions that exonerated the child. MSA indicated that blame of parents by therapists in their interventions, might be a factor worth exploring in relation to premature termination of treatment. Study two set out to explore the validity of the coding system developed, by comparing the results of using the coding system with the results of a Conversation Analysis of a therapeutic encounter. It was found that the coded causal attributions could be taken as representative of the participants' construction of blame and exoneration for the difficulties in that encounter.