The aims of this study were to identify the themes Social Workers regard as important in supporting decisions to remove children from, or return them to, the care of their parents. To further elicit underlying hypotheses that are discernible in interpretation of evidence.A case study, comprising a two-part vignette with a questionnaire, recorded demographic information, child welfare attitudes and risk assessments, using scales derived from standardised instruments, was completed by 202 Social Workers in Northern Ireland. There were two manipulated variables, mother's attitude to removal and child's attitude to reunification two years later. In this paper we use data derived from respondents' qualitative comments explaining their reasoning for in and out of home care decisions.Some 60.9% of respondent's chose the parental care option at part one, with 94% choosing to have the child remain in foster care at part two. The manipulated variables were found to have no significant statistical effect. However, three underlying hypotheses were found to underpin decisions; (a) child rescue, (b), kinship defence and (c) a hedged position on calculation of risk subject to further assessment.Reasoning strategies utilised by social workers to support their decision making suggest that they tend to selectively interpret information either positively or negatively to support pre-existing underlying hypotheses. This finding is in 2 keeping with the literature on 'confirmation bias'. The research further draws attention to the need to incorporate open questions in quantitative studies, to help guard against surface reading of data, which often does not 'speak for itself.'