| This study examined the use of the Child Well-Being Scales (Magura & Moses, 1986) as an aid to structured decision making in a child protection agency in Ontario, Canada. Trained date abstractors reviewed 337 files closed by the agency between September, 1994 and October, 1996 in order to test the hypotheses that the scales, completed by intake workers, would help improve prediction of a number of key decisions and activities taken by subsequent family service workers. The hypotheses were tested by constructing a series of OLS regression models for the continuous dependent variables, and logistic regression models for two dichotomous dependent variables. A number of potentially confounding variables were entered at step one in these analyses, and the CWBS family score was entered at step two in order to examine its unique contribution.;The null hypotheses suggesting the scales would make no contribution to prediction of: office visits, visits to other locations, phone calls, supervisory consultation, services, case, and placement duration, were rejected. The null hypothesis for the model predicting home visits was not rejected, and though the logistic regression null hypotheses were rejected in respect of the scales, the overall goodness of fit of the models was poor, thus obviating the CWBS results.;In conclusion, the scales appear to make a unique contribution to prediction of many case work activities and decisions. Thus they appear to have some utility, both as a clinical tool and as a contributor to the agency management information system. |
