Pessimism About Prediction of Violence
There has also been some pessimism about “prediction” of violence in the field of criminiology that we will have to face. Gottfredson and Gottfredson (1988) concluded that “there is little evidence supporting our ability to make these predictions with acceptable accuracy” (p. 252). In one study, Piquero (2000) concludes that frequent, violent offenders are indistinguishable from frequent, nonviolent offenders. As with the literature on specialization, we believe that the empirical research on prediction has been fairly conservative for our purposes. While it is not acceptable, in clinical practice, to have high rates of false positives and false negatives (identifying persons as very dangerous when they are not and failing to identify a dangerous person who really is), in the search for risk factors all we really need are statistically significant increases in probabilities. This is a much looser criterion than that sought for clinical prediction of future violence.