psy 2
Assessing Gender Role of Partner-Violent Men Using the Minnesota Multiphasic Personality Inventory-2 (MMPI-2): Comparing Abuser Types
David M. Lawson Stephen F. Austin State University
Dan F. Brossart Texas A&M University
Lee W. Shefferman University of Northern Colorado
This study investigated traditional masculine gender role differences between male partner abuser types using the Masculinity/Femininity subsection scales of Minnesota Multiphasic Personality Inventory-2 (MMPI-2) Structural Summary. We examined differences between four groups of partner-violent men (borderline, antisocial, psychotic features, and nonpathological partner violent) and one group of nonpartner-violent men on five MMPI-2 subscales: Masculinity-Femininity, Gender Role-Feminine, Gender Role-Masculine, Ego Inflation, and Low Self-esteem. Results indicated that the borderline group reported the most consistent traditional feminine gender role orientation of all the groups, whereas the antisocial group reported the most consistent traditional masculine gender role orientation of all the groups. The psychotic features group reported characteristics associated with both traditional masculinity and traditional feminine gender role making it distinct among all the groups. The nonpathological intimately violent group and the nonpartner-violent group reported no extreme scores when compared with the other three groups. The borderline and antisocial groups reported significantly more exposure to family of origin violence and use of more severe forms of partner abuse than the other three partner abuse groups. Treatment implications are addressed.
Keywords: masculinity, gender role, partner violence, domestic violence
Intimate partner violence (IPV) is a pattern of abusive behavior (physical and psychological) in a significant relationship and is often used by one partner to gain or maintain power/control over another partner (Office of Violence Against Women, 2009). Fam- ily conflict research indicates approximately equal rates of IPV for men and women (12% each) in national community samples (Stets & Straus, 1990; Straus, 1999). Crime studies, such as the National Violence Against Women in America Survey (NVAW; Tjaden &
Thoennes, 2000), indicate a much higher rate of assaults by men (i.e., 61%), but at a much smaller prevalence rates. Regardless of the type of survey, evidence indicates that women often experience more severe injuries and longer lasting symptoms, such as post- traumatic stress disorder (PTSD), depression, and anxiety (Cas- cardi, Langhinirichsen, & Vivian, 1992; Stets & Straus, 1990). Some scholars account for this phenomena by asserting that men have a greater tendency to use IPV to control and coerce women based on traditional masculine gender role and societal norms (Cascardi et al., 1992; Dobash & Dobash, 1998). This study investigated gender role differences between types of male partner abusers using the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; Butcher, Dahlstrom, Graham, Tellegen, & Kaemmer, 1989).
Because of its negative effects, research has examined numerous predictors of IPV including biological, genetic, stress, and psycho- social factors (Hamberger, 2008; Ehrensaft, Cohen, & Johnson, 2006). Of the psychosocial factors, masculine gender role has received increased attention in both the theoretical and empirical literature (Hamberger, 2005; Moore & Stuart, 2005), with a par- ticular focus on gender role socialization (Harway & O’Neil, 1999; Jakupcak, Lisak, & Roemer, 2002). From this perspective, social- ization plays the primary role in instigating men’s violence toward female partners (Pence & Paymer, 1993). The importance of gender role socialization and IPV is reflected in the large number of treatment programs that emphasize gender role resocialization as the primary goal of treatment (see Babcock, Green, & Robie, 2004).
DAVID M. LAWSON received his PhD in Counseling from University of North Texas. He is a professor in the PhD program in School Psychology in the Department of Human Services at Stephen F. Austin State Univer- sity. His research and practice interests include interpersonal abuse of adults and children, posttraumatic stress disorder with victims of interper- sonal abuse, and treatment of perpetrators and victims of interpersonal abuse. DAN F. BROSSART received his PhD in Counseling Psychology from the University of Missouri. He is an associate professor in the PhD program in Counseling Psychology in the Department of Educational Psychology at Texas A&M University. His research interests focus on intervention re- search, including process and outcomes research, and single-case research. LEE W. SHEFFERMAN received his PhD in Counseling Psychology from Texas A&M University. He is senior staff in the Counseling Center at University of Northern Colorado. His clinical interests include group counseling, men’s issues, ADHD and learning disabilities, and experiential methods. CORRESPONDENCE CONCERNING THIS ARTICLE should be addressed to David M. Lawson, Department of Human Services, Stephen F. Austin State University, Nacogdoches, TX 75962-3019. E-mail: lawsondm@sfasu.edu
Professional Psychology: Research and Practice © 2010 American Psychological Association 2010, Vol. 41, No. 3, 260 –266 0735-7028/10/$12.00 DOI: 10.1037/a0019589
260
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
One widely held view of gender role socialization and IPV posits that a traditional masculine orientation leads to an expecta- tion of male dominance and female subordination (Yllo, 2005). Socialization toward the dominance/subordination dichotomy en- courages men to engage in traditional gender-role behavior pat- terns in an attempt to avoid appearing feminine (Harway & O’Neil, 1999). Any sign of emotional sensitivity, vulnerability, or depen- dency is associated with shame and low self-esteem, which are setting conditions for IPV (Jennings & Murphy, 2000). Research provides support for this view as partner-violent men report lower self-esteem and higher interpersonal dependency than nonviolent men (Holtzworth-Munroe, Stuart, & Hutchinson, 1997; Murphy, Meyer, & O’Leary, 1994).
Feminist theories hold that male IPV is learned and nurtured in a patriarchal society. IPV is an act of dominance and control motivated by a need for power (Dobash & Dobash, 1998). When a man’s control and power are threatened by a female partner, he may react in a number of ways to defend his sense of masculine self-esteem, one of which is IPV (Harway & O’Neil, 1999). Moreover, IPV is reinforced by its effectiveness in modifying a partner’s behavior (Johnson, 1995). Other gender socialization theories emphasize different factors to explain IPV, such as con- striction of vulnerable emotions (Levant, 1996), male shame (Wal- lace & Nosko, 1993), gender role strain (Pleck, 1995), and im- paired male-to-male relationships (Jennings & Murphy, 2000).
Although theory and emerging research appear to support the a link between traditional masculine gender role and IPV, clinicians providing IPV treatment often fail to assess gender role issues in making clinical decisions (Moore & Stuart, 2005). This may occur because of time limitations and/or lack of familiarity with gender- based measures. Given the small treatment effects produced by most IPV treatment programs (Babcock et al., 2004), clinicians may benefit from assessing masculine gender roles as an integral part of the treatment process. The challenge is to identify other more familiar means to accomplish this end.
The MMPI-2 is widely used in the assessment of IPV (e.g., Hale, Duckworth, Zimostrad, & Nicholas, 1988; Flournoy & Wil- son, 1989). Common clinical uses include predicting success and recidivism rates in IPV treatment programs (Sellbom et al., 2008), identifying risk factors for IPV (Erwin, Gershon, Tiburzi, & Lin, 2005), and creating personality profiles to enhance treatment matching (Lawson et al., 2003). In addition to providing informa- tion on psychopathology, the MMPI-2 also contains scales that can provide a general assessment of gender role. Nichols and Greene (1995) developed the MMPI-2 Structural Summary as an empirical and rationally derived means to organize the 129 scales into clinically relevant categories for easier interpretation. They devel- oped six major sections (i.e., test-taking attitudes, factor scales, moods, cognitions, interpersonal relationships, other problem ar- eas) with each section containing several subsections. Masculinity/ Femininity is one subsection within the interpersonal section and includes four scales: Masculinity-Femininity (Mf), Gender Role- Feminine (GF), Gender Role-Masculine (GM), and Ego Inflation (Ma4). The choice of these scales for the Masculinity/Femininity subsection were determined based on content validity and item overlap, scale internal consistencies, and pattern confirmation de- termined by item correlation and factor analysis (Nichols & Green, 1995). All statistical analyses were conducted with a sample of 1,054 psychiatric inpatients and outpatients referred for psycho-
logical assessment. These four scales measure traditional mascu- line and feminine gender role concepts often identified with vio- lent offenders (Beesley & McGuire, 2009). However, the four scales lack the finer-grain gender role detail associated with some current research (see Moore & Stuart, 2005, for review). The scales may be less appropriate for assessing gender role for other populations (e.g., nonoffender populations), or for research en- deavors that examine masculinity constructs such as restrictive emotionality (O’Neil, Helms, Gable, David, & Wrightsman, 1986), attitudes about wife abuse (Dobash & Dobash, 1998) and/or for examining more contemporary constructs of masculinity such as gender role conflict/strain (Pleck, 1995).
IPV Typologies
Research suggests that partner-violent men are a diverse group (Dixon & Browne, 2003). In fact, IPV researchers have identified several typologies of partner-violent men based on various meth- odologies. Different types commit IPV for different reasons and may warrant different treatments (Saunders, 1996). Holtzworth- Munroe and Stuart (1994) proposed three types of partner-violent men (i.e., family only/nonpathological, generally violent/ antisocial, and dysphoric/borderline) based on severity of violence, generality of violence, and psychopathology. This typology has been consistently empirically validated (Dixon & Browne, 2003). Other studies have used one of the following characteristics to distinguish abuser types: (a) psychopathology (Hamberger, Lohr, Bonge, & Tolin, 1996); (b) type of partner violence (e.g., Chase, O’Leary, & Heyman, 2001); or (c) physiological responses and psychopathology (Gottman et al., 1995). Lastly, White and Gon- dolf (2000) employed a profile interpretation of the Millon Clinical Multiaxial Inventory III (Millon, 1983) to determine type catego- ries rather than statistical methods (e.g., cluster analysis). IPV typology research largely has produced consistent results across studies. Nevertheless, these studies have been criticized on a number of issues. Holtzworth-Munroe and Stuart’s (1994) typolo- gies confound violence dimensions and psychopathology in deter- mining abuser types (Hamberger et al., 1996). Hamberger et al. hold that typologies first should be identified based on psychopa- thology and then the resulting types can be compared based on relevant variables such as violence severity. Furthermore, Dixon and Browne (2003) note that Holtzworth-Munroe’s tripartite model was derived from a priori speculation and thus fails to identify other subtypes falling between the boundaries of the types.
To date, little research has examined gender role differences between partner abuser types. Most studies compare partner- violent with nonpartner-violent men with respect to gender role (e.g., Thompson, 1991). However, limited research suggests that more severe types of partner-violent men report stronger beliefs in male dominance than less severe types (Saunders, 1992).
Violence Variables
Research indicates that childhood exposure to parental violence increases the possibility of later aggression toward a partner (Dut- ton, 2007; Widom, 1989). Men with more severe psychopathology report witnessing more frequent and more severe interparental violence in childhood than do men with little or no psychopathol- ogy (Waltz, Babcock, Jacobson, & Gottman, 2000). Furthermore,
261ASSESSING GENDER ROLE OF PARTNER-VIOLENT MEN
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
the more severe the IPV, the more severe the psychopathology (Holtzworth-Munroe, Meehan, Herron, Rehman, & Stuart, 2000). Given this evidence, we would expect IPV typologies to differ with respect to childhood witnessing of parental violence and IPV severity.
Purpose of the Study
The primary purpose of our research was to examine male partner abusers’ gender role using the MMPI-2 Structural Sum- mary (Nichols & Greene, 1995) Masculinity/Femininity scales. In addition, we examined self-esteem and dependency given their association to gender role and IPV (Murphy et al., 1994). With research indicating the importance of distinguishing between abuser types (Dixon & Browne, 2003) we first determined abuser types and then compared the groups along with a nonpartner- violent group on the gender role scales and self-esteem (Ham- berger et al., 1996). Secondarily, we examined differences be- tween abuser types with respect to exposure to childhood parental IPV and severity of IPV used against a partner.
Hypotheses
Hypothesis 1: More severe abuser types (e.g., borderline, antisocial) will report more traditional masculine gender role than less severe abuser types (e.g., nonpathological) and nonpartner-violent men.
Hypothesis 2: More severe abuser types will report using more severe IPV and greater exposure to childhood parental violence than less severe abuser types and men that are nonpartner-violent.
Method
Participants
One hundred fifty-eight men on probation for either IPV or nonpartner violence crimes were invited to participate in the study. Twelve men declined after the initial request. Of the 146 who completed the materials, 25 cases were omitted from analysis because of excessive item omission and/or invalid results based on MMPI-2 validity scales. The final sample included 121 men who were on probation for either violence against their partner (n � 95) or a nonpartner violence offence (n � 26; e.g., illegal drug use or sale, DUI, theft). Of the partner abusive men, 96% had a history of drug/alcohol use or abuse. The nonpartner-violent men had no history of IPV based on self-report and arrest records, whereas the partner-violent men had engaged in at least one incident of partner assault based on arrest record and partner report. The IPV group ranged in age from 17 to 58 years (M � 31.0; SD � 8.5). Thirty (31.6%) were African American, 36 were Hispanic/Latino (37.9%), and 29 (30.5%) were Caucasian. Their education ranged from 5 to 18 years (M � 11.4; SD � 2.1). Twenty-nine were single, 41 were currently married, 9 were separated, and 16 were currently divorced. The nonintimate violent group ranged in age from 18 to 54 (M � 30.3; SD � 10.6). Eleven (42.3%) were African American, 8 (30.7%) were Hispanic/Latino, and 7 (26.9%) were Caucasian. Their education ranged from 7 to 18 years (M �
12.3; SD � 2.8). Fourteen were single, 6 were married, 3 were separated, and 3 were divorced.
Instruments
Typology. We conducted a profile analysis of the MMPI-2’s (Butcher et al., 1989) eight clinical scales (i.e., Hypochondriasis, Depression, Hysteria, Psychopathic Deviant, Paranoia, Psychoas- thenia, Schizophrenia, and Hypomania). This analysis involved identifying the predominant code type (e.g., 2- or 3-point highest score elevations) for each participant.
Masculinity and Self-Esteem scales. Masculinity constructs were examined using the four MMPI-2 scales that emphasize gender role: Masculinity-Femininity (Mf), Gender Role-Feminine (GF), Gender Role-Masculine (GM), and Ego-Inflation (Ma4). Interpretable scores are those with T scores of 65 or higher and T scores less than 45 (Greene, 2000). Men who score higher on the Mf scale are described as dependent, insecure regarding their masculinity, passive, and identify with feminine roles. Lower scoring men strongly identify with traditional masculine role and present as extremely masculine. Men with higher scores on the GF scale have stereotypic feminine characteristics such as sensitivity and interdependency while lower scorers have little interest in stereotypic feminine interests. Higher scores on the GM scale are self-confident, independent, and engage in stereotypic masculine activities. Lower scorers are fearful, socially introverted, and have little interest in masculine activities. Lower scorers on the Ma4 scale have stereotypic feminine characteristics (e.g., ease in speak- ing about emotions, preference for cooperation) and higher scores represent stereotypic masculine characteristics (e.g., impulsive- ness, rebelliousness, and hostility). Higher scorers on the Low Self-Esteem scale (LSE) reflect feeling less capable, less self- confident, and more dependent. Low scorers view themselves as capable and self-confident. The LSE scale measures the same constructs as earlier research (e.g., dependency, self-confidence) using self-report measures (e.g., Murphy et al., 1994). With the exception of the Ma4 scale (.44), coefficient alphas ranged from .60 to .79 (Graham, 2000).
Assessing for IPV and parent violence. Participants were assessed to be either partner violent or nonpartner violent based on arrest records, victim reports, face-to-face interviews, and partner responses on the 19-item Conflict Tactics Scale (CTS; Straus, 1979). The CTS assesses three types of aggression (current study alphas): (a) psychological aggression (.74)(e.g., insulted, swore); (b) mild/moderate physical aggression (.77) (e.g., pushed, slapped, grabbed); and (c) severe physical aggression (.86)(e.g., kicked, choked, beat up). A dichotomous variable was created denoting engagement in either mild/moderate IPV or severe IPV. Evidence supports both the content and construct validity of the CTS (Straus, 1979). Only the mild and severe physical aggression subscales were used as the focus of this research was IPV. Psychological aggression was not included.
We obtained information about exposure to interparental vio- lence with one question: Were you ever exposed to parent-on- parent violence during the time you were growing up at home? Examples of interparental violence were included. Their response was cross checked with the police report, historical information in the men’s files, and interview material. We found no inconsisten- cies between these sources.
262 LAWSON, BROSSART, AND SHEFFERMAN
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Procedures
Participants were administered the CTS and the MMPI-2 in groups of 15 to 30. The men were informed about the study and signed a consent form approved by the university IRB if they were willing to participate. They completed demographic forms and signed consent forms at the time of administration. Participants were interviewed individually to determine the nature of the of- fence, criminal history, family history, work history, and a psy- chosocial history.
MMPI-2 profile analysis. The eight clinical scales were used to identify the predominant code type (i.e., spike, 2- or 3-point highest scores). Profiles were examined for predominant high point code types with T-scores of �65, denoting clinically signif- icant profiles. Next, clinical profiles were separated into similar code type groups based on interpretive materials.
Results
Several analyses were conducted to examine the hypotheses. First, a profile analysis was conducted with the MMPI-2 to deter- mine typology groups based on psychopathology. Next, based on the resulting groups, we compared the groups on demographic variables and gender role variables. Finally, we compared the groups on violence variables.
Profile Analysis of MMPI-2
Prior to the primary analysis, we determined specific typology groups based on the predominant MMPI-2 code type. The group designations (e.g., antisocial) are based on typologies associated with specific behavioral features in the IPV literature (e.g., Holtzworth-Munroe & Stuart, 1994). Group designations should not be viewed as formal diagnostic categories but groups of similar, but not necessarily identical, symptom clusters that have characteristics akin to specific diagnostic categories. Although the groupings are comprised of similar profiles symptoms, there is a degree of heterogeneity within each group.
Code type composition. We identified four groups of inti- mately violent men based on predominant code types. Responses by men in group one (n � 19) indicated chronic distress and serious psychological problems characterized by emotional insta- bility, anxiety, anger, depression, low self-esteem, dependency, impulsivity, irritability, and interpersonal insensitivity. All 19 cases had scale elevation combinations of 2, 4, 6, 7, and 8 code types that are often associated with borderline personality features (Bell-Pringle, Pate, & Brown, 1997; Kelley & King, 1979; Trull, 1991). The majority of cases had one of the following profile configurations: 8-4-2, 8-2-4, 8-4-7, or 8-2-7 (see Table 1 for specific typology code types). Furthermore, the 19 cases were not well defined, because the top two or three scores for the code type definition were often accompanied by two to three additional elevations that were less than 5 points lower and still in the high (65–70) to very high (70 –79) range. Within this group, 21.1% (n � 4) engaged in mild to moderate partner violence, 78.9% (n � 15) engaged in severe partner violence, and 53% (n � 10) reported exposure to parent violence.
The predominant psychological/behavioral characteristic of group 2 (n � 24) was strong antisocial features. The overriding
characteristics of this group were rebellion toward authority, acting out behavior, anger, hostility, and interpersonal problems. Addi- tional characteristics included egocentricity, lack of insight into their own behavior, and low frustration tolerance and poor self- control. Furthermore, scale 4 (Psychopathic Deviant) was either the highest or next highest score. The 4-9/9-4 and spike 4 were the most common code types (see Table 1). Forty-six percent (n � 10) engaged in mild to moderate partner violence, 54% (n � 13) engaged in severe partner violence, and 54% (n � 13) reported exposure to parent violence.
The third group (n � 26) had profiles with a number of char- acteristics similar to group 1, including chronic and serious psy- chological problems, emotional instability, intense anxiety, and interpersonal problems. However, they are distinct from the other groups because of a number of strong psychotic features such as confusion, disorientation, deep mistrust of others, disturbance in thought, and feeling alienated from others. Spike 9, spike 6, 9-6, and 8-9 were the most common (Table 1). Within group 3, 57.7% (n � 15) engaged in mild/moderate partner violence, 42.3% (n � 11) engaged in severe partner violence, and 46.2% (n � 12) reported exposure to parent violence.
Results for the fourth group (n � 26) indicated that all eight clinical scales were within normal limits. Within the group, 61.5% (n � 16) engaged in mild/moderate partner violence, 38.5% (n � 10) engaged in severe partner violence, and 15.4% (n � 4) re- ported exposure to family of origin violence. These men were identified as nonpathological partner abusers.
The fifth group consisted of nonpartner-violent men who were on probation for nonviolent crimes. Of the 26 men, 7 had clinically elevated scores. The clinically elevated code types were as follow: 1-2-9, 1-3-2, 3-9-2, spike 8, spike 9 (n � 2) and 9-8.
Group Comparison for Demographic Variables
A multivariate analysis of variance (MANOVA) was conducted to examine differences between the four profile derived types of intimately violent men and the one nonpartner-violent group on marital status, education, age, and ethnicity. There was a signifi- cant main effect between the five groups on these three variables, multivariate F(16, 345) � 2.3, p � .03 (�2 � .074). Significant differences were found between the groups on age ( p � .05), with the borderline group being significantly older (M � 36.4, SD � 9.2) than the psychotic features group (M � 27.4, SD � 6.2).
Table 1 Minnesota Multiphasic Personality Inventory-2 (MMPI-2) Code Types for Typology Groups of Partner-Violent Men
Borderline features (n � 19)
Antisocial features (n � 24)
Psychotic features (n � 26)
8–4-2 (n � 3) 4–9 (n � 3) 9 (n � 10) 8–4-7 (n � 9) 9–4 (n � 3) 9–6 (n � 5) 8–2-4 (n � 2) 4 (n � 9) 8 8–2-7 4–1 (n � 2) 8–6 2–3-4 (n � 2) 4–2 8–7 2–6 (n � 2) 4–3 8–9 (n � 3)
4–6 (n � 3) 6 (n � 5) 4–7 4–8
263ASSESSING GENDER ROLE OF PARTNER-VIOLENT MEN
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
Group Comparison for Gender Role Variables
A MANOVA was performed to examine differences between the four profile derived partner-violent groups and one nonpartner- violent group on the five gender variables. There was a main effect indicating differences between the groups, multivariate, F(20, 372) � 4.0, p � .001 (�2 � .154). Between-subjects effects were significantly different ( p � .05) for all variables except the Mf scale. Post hoc analyses indicated a number of between group differences (Table 2).
The borderline group reported being more fearful, socially in- troverted, and having less interest in masculine activities (GM) than all the other groups except the psychotic features group. Further, this group reported feeling less capable, less competent, less self-confident, less adequate, more insecure, and more depen- dent (LSE) than all groups.
Similar to the borderline group but less extreme, the psychotic features group reported feeling less capable, less competent, less self-confident, and more dependent (LSE) than all groups except the borderline group. This group also reported more traditional masculine characteristics (e.g., impulsive, rebellious, and hostile; Ma4) than all groups except the antisocial group.
The antisocial group reported a lower level of interest in ste- reotypic feminine activities (e.g., interpersonal affiliation; GF), than the nonpathological domestic violence group. They viewed themselves as more capable and self-confident (LSE) when com- pared with the borderline and psychotic features groups. Scores indicated a narcissistic, over evaluation of their abilities.
As anticipated, the nonpathological IPV group and the nonpartner-violent group reported no extreme scores when com- pared with the other three groups. Overall, results partially support Hypothesis 1, with only the antisocial group reporting a consistent traditional masculine gender role when compared with other groups.
Comparison of Intimate Violent Groups on Violence Variables
The borderline (53%) and antisocial groups (54%) reported the most exposure to family of origin violence, �2(3, N � 95) � 41.98, p � .000, followed by the psychotic features group (46.2%), and nonpathological abuse group (15.4%). Furthermore, men in the
borderline group (78.9%) engaged in more severe IPV than any of the other partner-violent groups, �2(3, N � 95) � 42.60, p � .000, followed by the antisocial, psychotic features, and nonpathological abusers groups (54%, 42.3%, and 38.5 respectively). These results support Hypothesis 2.
Discussion
Results indicated similarities and differences between the vari- ous abuser types and gender role characteristics, thus providing support for using the MMPI-2 to obtain a general profile of gender role for partner-violent men. The borderline group reported the most consistent traditional feminine gender role orientation of all groups. With respect to the borderline group, results support re- search indicating that low self-esteem and dependency are associ- ated with IPV (Flournoy et al., 1989; Hale et al., 1988) and specifically, with borderline partner abusers (Holtzworth-Munroe et al., 2000). However, these results clearly contradict research that attributes IPV to more stereotypic masculine characteristics (Par- rott & Zeichner, 2003). Dutton (2007) holds that IPV perpetrated by men with borderline personality features is often a result of exposure to parental violence, attachment anxiety, dependency, and fear of abandonment. The borderline group in the present study reported similar characteristics such as being fearful, inse- cure, and dependent, along with a majority having been exposed to parental violence. Perhaps these personal characteristics along with traditional feminine gender role contribute to IPV for this group.
In contrast, the antisocial group reported more features associ- ated with a traditional masculine gender role when compared with the borderline, psychotic features, and nonpathological groups. These characteristics are in greatest contrast to those of the bor- derline group but are consistent with theories of masculinity that attribute IPV to hypermasculinity (Parrott & Zeichner, 2003), traditional masculinity (Thompson, 1991), and masculine ideology (Jakupcak et al., 2002).
Research indicates that antisocial and borderline men tend to be the most violent of the IPV types (Holtzworth-Munroe et al. 2000; Waltz et al., 2000). Some suggest that the exposure to family violence in childhood and the severity of psychopathology may account for this fact (Ehrensaft et al., 2006). This is consistent with our results indi-
Table 2 Post Hoc Gender Role Between Group Differences (Profile-Derived Groups)
Variables
Clusters M (SD)
FBorderline Antisocial Psychotic features Nonpathological Nonviolent
Mf 47.5 (6.2) 43.16 (6.5) 43.8 (8.1) 42.1 (6.8) 45.5 (8.4) 0.76 GM 38.5 (7.8) 48.12 (10.4) 44.6 (9.4) 50.5 (7.6) 50.5 (9.8) 6.40a�,b��,c��
GF 45.3 (10.0) 41.4 (7.7) 43.2 (9.1) 50.2 (7.8) 45.1 (10.4) 2.70e�
Ma4 54.4 (11.3) 58.8 (8.8) 64.2 (10.9) 53.3 (7.7) 54.4 (11.7) 3.60 f�,g��,h�
LSE 64.3 (9.3) 44.2 (6.7) 54.4 (9.6) 44.8 (6.5) 45.4 (10.5) 13.8 a��,b��,c��,d�,f�,g��,h��
Note. Mf � Masculinity/Femininity; GM � Gender-Role Masculine; GF � Gender-Role Feminine; Ma4 � Ego-Inflation; LSE � Low Self-Esteem. a Contrast between borderline and antisocial was significant. b Contrast between borderline and nonpathological was significant. c Contrast between borderline and nonpartner violent was significant. d Contrast between psychotic features and antisocial was significant. e Contrast between antisocial and nonpathological was significant. f Contrast between psychotic features and borderline was significant. g Contrast between psychotic features and nonpathological was significant. h Contrast between psychotic features and nonviolent was significant. � p � .05. �� p � .01.
264 LAWSON, BROSSART, AND SHEFFERMAN
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
cating that the borderline and antisocial groups have the highest use of severe violence and the highest rate of exposure to parental violence. A sole, monolithic gender role may be a less prominent influence on IPV when more serious psychopathology and exposure to parental violence are present. However, the fact that the borderline and anti- social groups represent different traditional gender roles may indicate some contribution of gender role to IPV. Although counterintuitive, perhaps either gender role taken to a more extreme traditional level (masculine or feminine) contributes to IPV.
The psychotic features group reported characteristics associated with both traditional masculinity and traditional feminine gender role when compared with the other groups, making it distinct among all the groups. Furthermore, this group reported the third highest amount of exposure to family of origin violence and use of severe violence. This group may approximate what some refer to as a multidimensional view of masculinity (Levant, 1996). It assumes there is no single criterion for masculinity or a single invariant masculine ideology as masculinity is a social construc- tion and varies with age, social class, race/ethnic group, and life stage.
The nonpathological intimately violent and the nonpartner violence groups scores were in the normal range on masculine gender role compared with the other three groups. Their scores support the view that a less traditional gender role is less associated with IPV. Of the five groups, we would expect these two to be the most flexible in gender role. Furthermore, the nonpathological violent group reported the least exposure to parental IPV and used severe violence the least of the violent groups. This is consistent with the evidence linking severity of psychopathology and exposure to parental IPV, to severity of IPV (Chase et al., 2001).
Limitations
This study had several limitations. Arguably, the most signifi- cant limitation was the relatively moderate sample size. This may have reduced the power to detect additional significant differences and relationships. Furthermore, participants may have provided socially desirable responses even though they were informed that names would not be affixed to their response materials nor would they be shared with probation officers. Finally, a limitation noted from the outset was using the MMPI-2 to evaluate masculine gender role. Other instruments would provide more fine grain and contemporary gender role information such as the Male Role Norms Scale (Thompson & Pleck, 1986) or the Gender Role Conflict Scales (O’Neil et al., 1986). However, most clinicians would be less likely to use such instruments in practice, which was a major premise of this article.
Clinical Implications
Treatment programs that employ feminist-based treatment mod- els emphasize resocialization of abusers with respect to their views of women, power, and control. This focus assumes all male partner abusers hold to traditional masculine roles and attitudes, including the use of violence to dominate women (Pence & Paymar, 1993). Thus, much of treatment focuses on challenging and changing traditional masculine gender roles. Although the MMPI-2 scales do not address men’s “right to dominate women,” traditional
masculine gender roles often support such beliefs and ultimately IPV (Loseke & Kurz, 2005). From this perspective, IPV is about challenges to a man’s authority, self-esteem, and status. Our results suggest that treatment should include a focus on both masculine and feminine traditional gender roles and how they contribute to IPV, especially for antisocial and borderline types. Furthermore, a partner-violent male may subscribe to both masculine and femi- nine traditional gender roles, as with the psychotic features group. Thus, therapists should be sensitive to a multidimensional perspec- tive regarding gender role.
In sum, these results support the view that partner-violent men are a diverse group in terms of psychopathology and severity of IPV, but also in gender role. Thus, treatment must be tailored to the person not one stereotypic view of masculine gender role. Not- withstanding its limitations as noted, the MMPI-2 appears to be a useful instrument for obtaining general profile information on traditional gender role for partner-violent men.
References
Babcock, J. C., Green, C. E., & Robie, C. (2004). Does batterers’ treatment work? A meta-analytic review of domestic violence treatment. Clinical Psychology Review, 23, 1023–1053.
Beesley, F., & McGuire, J. (2009). Gender-role identity and hypermascu- linity in violent offending. Psychology, Crime & Law, 15, 251–268.
Bell-Pringle, V. J., Pate, J. L., & Brown, R. C. (1997). Assessment of borderline personality disorder using the MMPI-2 and the Personality Assessment Inventory. Assessment, 4, 131–139.
Butcher, J. N., Dahlstrom, W. G., Graham, J. R., Tellegen, A., & Kaem- mer, B. (1989). Minnesota Multiphasic Personality Inventory-2 (MMPI- 2): Manual for administration and scoring. Minneapolis: University of Minnesota Press.
Cascardi, M., Langhinrichsen, J., & Vivian, D. (1992). Marital aggression: Impact, injury, and health correlates for husbands and wives. Archives of Internal Medicine, 152, 1178 –1184.
Chase, K. A., O’Leary, K. D., & Heyman, R. E. (2001). Categorizing partner-violent men within the reactive-proactive typology model. Jour- nal of Consulting and Clinical Psychology, 69, 567–572.
Dixon, L., & Browne, K. (2003). The heterogeneity of spouse abuse: A review. Aggression and Violent Behavior, 8, 107–130.
Dobash, R., & Dobash, R. (1998). Rethinking violence against women. Thousand Oaks, CA: Sage.
Dutton, D. G. (2007). The abusive personality (2nd ed). New York: Guilford.
Ehrensaft, M. K., Cohen, P., & Johnson, J. G. (2006). Development of personality disorder symptoms and the risk for partner violence. Journal of Abnormal Psychology, 115, 474 – 483.
Erwin, M. J., Gershon, R. R., Tiburzi, M., & Lin, S. (2005). Reports of intimate partner violence made against police officers. Journal of Family Violence, 20, 13–19.
Flournoy, P. S., & Wilson, G. L. (1991). Brief Research Report: Assess- ment of MMPI profiles of male batterers. Violence and Victims, 6(4), 309 –320.
Gottman, J. M., Jacobson, N. S., Rusche, R. H., Short, J. W., Babcock, J., LaTallidade, J. J., et al. (1995). The relationship between heart rate reactivity, emotionally aggressive behavior, and general violence in batterers. Journal of Family Psychology, 9, 227–248.
Graham, J. R. (2000). MMPI-2: Assessing Personality and Psychopathol- ogy (3rd ed.). New York: Oxford University Press.
Greene, R. L. (2000). The MMPI-2: An interpretive manual (2nd ed). Boston: Allyn and Bacon.
Hale, G., Duckworth, J., Zimostrad, S., & Nicholas, D. (1988). Abusive
265ASSESSING GENDER ROLE OF PARTNER-VIOLENT MEN
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .
partners: MMPI profiles of male batterers. Journal of Mental Health Counseling, 10, 214 –224.
Hamberger, L. K. (2005). Men’s and women’s use of intimate partner violence in clinical samples: Toward a gender sensitive analysis. Vio- lence and Victims, 20, 131–151.
Hamberger, L. K. (2008). Twenty-five years of change in working with partner abusers-Part II: Observations from the trenches about changes in understanding of abusers and abusers treatment. Journal of Aggression, Maltreatment & Trauma, 17, 1–22.
Hamberger, L. K., Lohr, J. M., Bonge, D., & Tolin, D. F. (1996). A large sample empirical typology of male spouse abusers and its relationship to dimensions of abuse. Violence and Victims, 11, 277–292.
Harway, M., & O’Neil, J. M. (Eds.). (1999). What causes men’s violence against women? Thousand Oaks, CA: Sage.
Holtzworth-Munroe, A., Meehan, J. C., Herron, K., Rehman, U., & Stuart, G. L. (2000). Testing the Holtzworth-Munroe and Stuart (1994) batterer typology. Journal of Consulting and Clinical Psychology, 68, 1000 – 1019.
Holtzworth-Munroe, A., & Stuart, G. L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psychologi- cal Bulletin, 116, 476 – 497.
Holtzworth-Munroe, A., Stuart, G. L., & Hutchinson, G. (1997). Violent versus nonviolent husbands: Differences in attachment patterns, depen- dency, and jealousy. Journal of Family Psychology, 11, 314 –331.
Jakupcak, M., Lisak, D., & Roemer, L. (2002). The role of masculine ideology and masculine gender role stress in men’s perpetration of relationship violence. Psychology of Men and Masculinity, 3, 97–106.
Jennings, J. L., & Murphy, C. M. (2000). Male-male dimensions of male-female battering: A new look at domestic violence. Psychology of Men and Masculinity, 1, 21–29.
Johnson, M. P. (1995). Patriarchal terrorism and common couple violence: Two forms of violence against women. Journal of Marriage and the Family, 57, 283–294.
Kelley, C., & King, G. (1979). Behavioral correlates of infrequent 2-point MMPI code types at a University Mental Health Clinic. Journal of Clinical Psychology, 35, 576 –585.
Lawson, D. M., Weber, D., Beckner, H. M., Robinson, L., Marsh, N., & Cool, A. (2003). Men who use violence: Intimate violence versus non- intimate violence profiles. Violence and Victims, 18, 259 –277.
Levant, R. F. (1996). The new psychology of men. Professional Psychol- ogy: Research and Practice, 27, 259 –265.
Loseke, D. R., & Kurz, D. (2005). Men’s violence toward women is the serious social problem. In D. R. Loseke, R. J. Gelles, & Mary M. Cavanaugh (Eds.), Current controversies on family violence (2nd ed., pp. 79 –96). Thousand Oaks, CA: Sage.
Millon, T. (1983). Millon Clinical Multiaxial Inventory. Minneapolis, MN: National Computer Systems.
Moore, T. M., & Stuart, G. L. (2005). A review of the literature on masculinity and partner violence. Psychology of Men and Masculinity, 6, 46 – 61.
Murphy, C. M., Meyer, S., & O’Leary, K. D. (1994). Dependency char- acteristics of partner assaultive men. Journal of Abnormal Psychology, 103, 729 –733.
Nichols, D. S., & Greene, R. L. (1995). MMPI-2 structural summary interpretive manual. Odessa, Fl: Psychological Assessment Resources Inc.
O’Neil, J. M., Helms, B., Gable, R., David, L., & Wrightsman, L. (1986).
Gender Role Conflict Scale: College men’s fear of femininity. Sex Roles, 14, 335–350.
Office of Violence Against Women. (2009, March). About domestic vio- lence. Retrieved from http://www.ovw.usdoj.gov/domviolence.htm
Parrott, D. J., & Zeichner, A. (2003). Effects of hypermasculinity on physical aggression against women. Psychology of Men and Masculin- ity, 4, 70 –78.
Pence, E., & Paymar, M. (1993). Education groups for men who batter: The Duluth model. New York: Springer.
Pleck, J. H. (1995). The gender role strain paradigm: An update. In R. F. Levant & W. S. Pollock (Eds.), A new psychology of men (pp. 11–32). New York: Basic Books.
Saunders, D. G. (1992). A typology of men who batter: Three types derived from cluster analysis. American Orthopsychiatry, 62, 264 –275.
Sellbom, M., Ben-Porath, Yossef, S., Baum, L. J., Erez, E., & Gregory, C. (2008). Predictive validity of the MMPI-2 Restructured Clinical (RC) scales in a batterers’ intervention program. Journal of Personality As- sessment, 90, 129 –135.
Straus, M. A. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics Scale (CTS). Journal of Marriage and the Family, 41, 75– 88.
Stets, J. E., & Straus, M. A. (1990). Gender differences in reporting marital violence and its medical and psychological consequences. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families: Risk factors and adaptations to violence in 8,145 families (pp. 151–166). New Brunswick, NJ: Transaction.
Straus, M. A. (1999). The controversy over domestic violence by women: A methodological, theoretical, and sociology of science analysis. In X. B. Arriaga & S. Oskamp (Eds.), Violence in intimate relationships (pp. 17– 44). Thousand Oaks, CA: Sage.
Thompson, E. H. (1991). The maleness of violence in dating relationships: An appraisal of stereotypes. Sex Roles, 24, 261–279.
Thompson, E. H., & Pleck, J. H. (1986). The structure of male role norms. American Behavioral Scientist, 29, 531–543.
Tjaden, P., & Thoennes, N. (2000). Full report of the prevalence, inci- dence, and consequences of intimate partner violence against women: Findings from the National Violence Against Women Survey. Retrieved from http://www.ncjrs.gov/pdffiles1/nij/183781.pdf
Trull, T. (1991). Discriminant validity of the MMPI-Borderline Personality Disorder Scale. Psychological Assessment, 3, 232–238.
Wallace, R., & Nosko, A. (1993). Working with shame in group treatment of male batterers. International Group Psychotherapy, 43, 45– 61.
Waltz, J., Babcock, J. C., Jacobson, N. S., & Gottman, J. M. (2000). Testing a typology of batterers. Journal of Consulting and Clinical Psychology, 68, 658 – 669.
White, R. J., & Gondolf, E. W. (2000). Implications of personality profiles for batterer treatment. Journal of Interpersonal Violence, 15, 467– 488.
Widom, C. S. (1989). Does violence beget violence? Psychological Bul- letin, 106, 3–28.
Yllo, K. (2005). Through a feminist lens: Gender, diversity, and violence. In D. R. Loseke, R. J. Gelles, & Mary M. Cavanaugh (Eds.), Current controversies on family violence (2nd ed., pp. 19 –34). Thousand Oaks, CA: Sage.
Received September 18, 2009 Revision received February 22, 2010
Accepted March 2, 2010 �
266 LAWSON, BROSSART, AND SHEFFERMAN
T hi
s do
cu m
en t i
s co
py ri
gh te
d by
th e
A m
er ic
an P
sy ch
ol og
ic al
A ss
oc ia
tio n
or o
ne o
f i ts
a lli
ed p
ub lis
he rs
. T
hi s
ar tic
le is
in te
nd ed
s ol
el y
fo r t
he p
er so
na l u
se o
f t he
in di
vi du
al u
se r a
nd is
n ot
to b
e di
ss em
in at
ed b
ro ad
ly .