Hot spur or tranquil? The adaptation of psychometric anger assessment instruments and their evaluation and application on violent and nonviolent samples in Sweden

Sammanfattning: Anger is a negatively toned emotion and biopsychosocially functional alarm, which automatically activates a colorful arsenal of affective, cognitive, physiological, and behavioral action impulses in the face of experienced irritations and provocations. Anger, on the other hand, is a mixed blessing and a subject of admiration and condemnation since the days of Plato, Aristotle and Thomas Aquinas. This usual and passionate human emotion has also been closely associated with phenomena of human psychopathology: mental and personality disorders, aggression, interpersonal difficulties, and violence. Until the recent past, conceptual confusion and scientific neglect have complicated the research on anger.Following a theoretical review of multiple, classical philosophical and contemporary, scientific approaches to anger, this thesis describes the work towards its principal aim: to develop the first Swedish adaptations of the psychometric anger assessment instruments Novaco Anger Scale (NAS, version 1998), Provocation Inventory (PI), and State-Trait Anger Expression Inventory-2 (STAXI-2). These psychometric anger assessment instruments focus on various theoretical dimensions of the anger concept; NAS assesses the cognitive mediation, physiological arousal, and behavioral action impulses and control efforts that compose anger. PI assesses the intensity of anger reaction as a function of various provocations. Finally, STAXI-2 assesses the various state and trait experience, expression, and control of anger. The adaptations of the instruments were developed through the co-operation of the authors of the three studies included in this thesis, with a native back-translator, and also with the authors of the original instruments. The adaptations, labeled NAS-1998-S, PI-S, and STAXI-2-S, were investigated on the basis of their factor structures, internal reliability, scale correlations, and predictive capacity of group membership of participants, in male samples of violent prisoners (N = 95; Mage = 33.2, SDage = 10.6) and nonviolent university students (N = 100; Mage = 33.2, SDage = 12.5) in Sweden. Furthermore, the level of anger reported by the violent criminal and nonviolent university student samples in Sweden was compared, as were those in turn with published norm values of American standardization samples of the original instruments. According to the results presented in the thesis, NAS-1998-S, PI-S, and STAXI-2-S have appropriate psychometric qualities. More specifically, NAS-1998-S and PI-S (but not STAXI-2-S) demonstrated factor structures that were interpreted as corresponding to the findings of previous studies on the original instruments. The parts that compose NAS-1998-S, PI-S, STAXI-2-S (i.e., subscales, scales, and total scale scores) were in good correspondence with what could be expected on the basis of previous findings on the original instruments. The scales had an appropriate pattern of correlations with concurrent scales of related and distinct theoretical constructs. The adapted assessment instruments demonstrated good capacities to discriminate the prisoners and university students from each other. The results were interpreted in terms of appropriate construct and cross-cultural validity. As expected, the violent prisoners scored higher on the various cognitive, physiological, behavioral, trait, and expressed dimensions of anger, and lower on anger control, compared with nonviolent university students and persons in the American standardization samples. An unexpected, but culturally explainable result was the relatively high levels of anger suppression by both Swedish samples. In particular, the scores of violent prisoners on cognitive inclination and general disposition to anger approached clinically high levels. The high levels of anger in the violent prisoners were interpreted as supportive of previous research indicating the high prevalence of psychopathology in the offender population. Anger being an important antecedent of aggressive behavior and a vital correlate of clinical disorders, the systematic assessment of this emotion has not yet earned a high priority in the research and clinical work with offenders in Sweden. The work described in this thesis is an effort to remedy this condition.