The violence of prediction: The uneasy relationship between social science and social policy

Neither the suggestion that certain categories of people be prevented from childbearing nor the suggestion that involuntary commitment of mental patients be made easier is justified by scientific analysis. Assertion that the high rate of violence in the United States is "caused by" inadequate socialization of biologically susceptible children raised by single parents is unjustified by available data. Similarly, results of recent studies predicting violence among the mentally ill provide only minimal implications for social policy. The problem of false positives continues, in both domains, to be a serious one, and the values, biases, and social constructions of "experts" will continue to influence their interpretation and presentation of available facts.