Measurement of aggression in psychiatric patients.

This article analyzes, describes, and characterizes the methods used for measuring aggression in studies on psychiatric patients. The authors reviewed all studies published between 1985 and 1994 in seven major psychiatric journals and found 103 studies on adult aggression and 43 on childhood and adolescent aggression. Almost half (44.7%) of the adult studies and 23.3% of the child and adolescent studies did not use any structured instrument; the remainder used a total of 52 different instruments, indicating that the methodology in this area is not well established. The methods used for measuring aggression were mainly of three types: observational measures; self-report questionnaires; and structured chart reviews. Each of these tools is described herein, with special focus on reliability and validity. Overall the reliability of the various scales is high, but the empirical validity is rather poor. In choosing a suitable method for measuring aggression, researchers should take into account the direction of investigation (e.g. state vs. trait aggression, dynamic or biological aspects of aggression) and the target population. Practical suggestions in this regard are offered.