No abstract

In this paper, we describe the progress of four male, legally-detained personality disordered offenders in a group treatment programme for angry aggression. ''Controlling angry aggression'' is a 15-session, structured, cognitive-behavioural programme that is part of a wider integrated, multidisciplinary treatment programme run within a specialist personality disorder treatment unit. Psychometric tests showed that three of the four patients improved over the course of the treatment programme. Behaviour ratings of staff and patients concurred, allowing confidence in patients' self-report of anger. No change in behaviour was evident over time for any patient, but initial low rates of aggression allowed little room for improvement. Self-monitored anger and aggression scores did vary markedly over time, and since behaviour remains stable despite anger, this indicates that these patients can control their behaviour. One patient did not improve, and reasons for this are examined, concluding with a reminder of the need for rigorous selection of patients for treatment programmes.

Although alcohol-related violence is a common problem, there is wide variability amongst individuals in the extent to which they engage in alcohol-related violence. To assist with identification of individual differences, we developed a comprehensive measure of alcohol-related aggression. Items for the Alcohol-related Aggression Questionnaire (ARAQ) were generated from a priori literature searches and administered to 226 males. Factor analysis reduced the ARAQ from 64 to 28 items, with five subscales showing good internal consistency and concurrent validity. In a second study, 219 offenders and non-offenders were tested on the ARAQ-28, the Buss-Perry Aggression Questionnaire and the Alcohol Use Disorders Identification Test. Confirmatory factor analysis did not support the original factor structure, but an exploratory factor analysis gave a new four-factor structure with broadly similar constructs. The ARAQ-28 showed good internal consistency, test-retest reliability and construct and discriminant validity. The ARAQ-28 coheres as a univariate scale, and its subscales may have clinical utility.

No abstract