Academic journal article
By Hamberger, L. Kevin; Lohr, Jeffrey M.; Bonge, Dennis; Tolin, David F.
Violence and Victims , Vol. 11, No. 4
A number of studies have described typologies of domestically violent men. Holtzworth-Munroe and Stuart (1994) recently proposed a theoretical model for predicting violence severity and generality from personality "type." The present study, using data from 833 identified abusive men, tested the model. Personality types were determined from cluster analysis of data from the Millon Clinical Multiaxial Inventory, and resulted in a three-cluster solution consistent with the Holtzworth-Munroe and Stuart model. The three main clusters included nonpathological, antisocial, and passive aggressive-dependent groups. Three other, smaller types were also identified. Multivariate and chi-square analyses comparing the main clusters on other variables generally supported the Holtzworth-Munroe and Stuart model. Nonpathological men had the lowest maximum violence and frequency. They restricted their violence primarily to intimate relationships and had the fewest police contacts. Antisocial and passive aggressive-dependent men did not differ in maximum violence, but antisocial men were the most generally violent and had the most police contacts. Passive aggressive-dependent men had the highest frequency of violence. Clinical, theoretical and methodological implications are discussed.
For a number of years, researchers have investigated personality, psychological, and behavioral characteristics of heterosexual men who batter their partners. In general, two major approaches to the systematic, empirical study of domestically violent men have been taken. One approach has compared domestically violent men to groups of nonviolent but maritally distressed men and nonviolent maritally satisfied men. Though not an exhaustive review, these comparison studies generally show that, relative to nonviolent men, domestically violent men show more hostility (e.g, Maiuro, Cahn, Vitaliano, Wagner, & Zegree, 1988), assertiveness deficits (e.g., Maiuro, Cahn, & Vitaliano, 1986) problem-solving deficits (e.g., Barnett & Hamberger, 1992), substance abuse (e.g., Van Hasselt, Morrison, & Bellack, 1985), problems with intimacy (e.g., Allen, Calsyn, Fehrenbach, & Benton, 1989), and higher levels of marital distress (e.g., Rosenbaum & O'Leary, 1981). In addition, other studies have found that abusive men show higher elevations on measures of personality disorder than do nonviolent men (e.g., Beasley & Stoltenberg, 1992; Else, Wonderlich, Beatty, Christie, & Staton, 1993). A major disadvantage of such group comparison approaches is that much of the diversity of characteristics among violent men is lost (Hastings & Hamberger, 1988).
The study of different types of domestically violent men constitutes the second approach to this research area. Such an approach seeks to identify variability among subgroups or types of domestically violent men. In our laboratory, we have attempted to empirically determine different types of psychological functioning and psychopathology among men who batter. For example, Hamberger and Hastings (1986) factor analyzed the basic eight personality scale scores of the Millon Clinical Multiaxial Inventory (MCMI, Millon, 1983) on two samples (n = 105 and 99, respectively) of domestically violent men. Results of both analyses showed three main factors labeled, "schizoidal/borderline," "narcissistic/ antisocial," and "passive-dependent/compulsive." Various combinations of these three factors produced eight subtypes of abusive men, including one group that showed no elevations on any MCMI scale. This nonpathological group made up about 12%-14% of the total sample. In general, men scoring high on the schizoidal/borderline factor showed high levels of depression and anger proneness, abuse in the family of origin and high levels of alcohol abuse. Narcissistic/antisocial abusers exhibited problems with alcohol and drug abuse, but not with depression and anger proneness. Passive-dependent/compulsive abusers showed moderate levels of depression and no clear pattern for anger proneness. …