The Clinical Prediction of Dangerousness.

The failure to accurately predict violent potential in psychiatric patients is currently an area of much popular and professional interest. Virtually all studies have shown psychological testing to be a poor predictor of violence. The present investigation utilized typed vignettes which described a brief, fictionalized interview as an analogue to an actual contact with a patient. A recent history of violence an actuarial factor, support by a caring person an environmental factor, and psychological testing information were systematically manipulated among the vignettes. Vignettes were mailed to psychiatrists in California who are asked to rate dangerousness and to decide if emergency hospitalization was required for the fictionalized patient. Each psychiatrist received one of 16 possible vignettes. Statistical analyses were utilized to determine how the three manipulated variables actually influenced professional decision making, and these results were compared with what factors the psychiatrists said influenced them. This study produced many practical implications. Psychiatrists in California incorporate a recent history of violence into the decision-making process, but do not give psychosocial support the attention warranted. Psychological testing exerts a powerful but unwarranted, and possible unconscious, influence on the judgmental process. Dissertation