it 311;.

{‘1‘

"um'fvi'a’sd

 

mi-
.1‘

.
v
.‘
5

L‘.':;t-

-:.

-.-

 

 

 

This is to certify that the

dissertation entitled

Antecedents to Psychological Abuse: Adult
Attachment, Gender, and Defense

presented by

Barbara Gormley

has been accepted towards fulfillment
of the requirements for

Ph.D Counseling Psychology

 

' degree in

 

Major professor V V y

Date 7/18/02.

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN REI'URN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

W

(In; new

 

 

EEK! Cum

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIRC/DaleDue.p65—p.15

ANTECEDENTS TO PSYCHOLOGICAL ABUSE: ADULT ATTACHMENT.
GENDER, AND DEFENSE

By

Barbara Gormley

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Counseling, Educational Psychology, and Special Education

2002

ABSTRACT

ANTECEDENTS TO PSYCHOLOGICAL ABUSE: ADULT ATTACHMENT,
GENDER, AND DEFENSE

By

Barbara Gormley

Psychological abuse in heterosexual relationships as perpetrated by college students was
examined. Gender differences and the influences of insecure attachment and defense
mechanisms were investigated. Hierarchical regression was used to test gender,
attachment avoidance, attachment anxiety, narcissism, self-splitting, and other-splitting as
predictors of dominance and emotional abuse. Gender was expected to moderate
attachment-abuse relationships. Defense mechanisms were expected to moderate
avoidance-abuse relationships and to mediate anxiety-abuse relationships. Separate
mediator-moderator models were tested for dominance and emotional abuse. Stress level
was controlled. The interaction between gender and attachment avoidance and narcissism
significantly predicted dominance. Men but not women with higher levels of attachment
avoidance endorsed the use of more dominance strategies. Higher levels of defensive
uses of narcissism predicted higher dominance scores. Significant predictors of emotional
abuse included narcissism and the interaction between narcissism and attachment
avoidance. Higher levels of defensive uses of narcissism predicted more frequent
emotional abuse. Higher levels of attachment avoidance predicted more frequent
emotional abuse only when levels of narcissism were higher than average. Defensive uses
of narcissism led to psychologically abusive behavior in college students, but the

influences of attachment avoidance depended upon moderators such as gender and

narcissism. Posthoc analyses elucidated gender differences in dominance and mediators
between attachment insecurity and emotional abuse. Among women. other-splitting was
the only variable to significantly predict dominance scores. Among men. the significant
predictors were stress. attachment avoidance. and narcissism. Preliminary findings
regarding self-splitting as a mediator of the attachment anxiety-emotional abuse
relationship and other-splitting as a moderator of the relationship between attachment
avoidance and emotional abuse need further study. Interpretations of findings linked male
gender role socialization, difficulties with closeness in romantic relationships, and a

.

narcissistic sense of exploitation and entitlement to psychological abuse and domestic

violence. Implications for practice and research were discussed.

To Dr. Roger Landvoy, for being there and seeing me.

ACKNOWLEDGEMENTS

I am grateful to Dr. Fred Lopez, my advisor and chair. His tireless enthusiasm for
adult attachment research and his challenging presence have contributed much to my
professional development. I appreciate the way my committee members, Dr. Gloria
Smith, Dr. Ken Frank, and Dr. Sheila Maxwell, supported and taught me, helping me
articulate and improve my ideas. I would not have been able to pursue doctoral education
without the Michigan State University Distinguished Fellowship, which provided full
support for four years with a match from the College of Education.

My brother John Gormley provided an incredible amount of concrete support,
including expert computer assistance at all hours of day and night. While I was preparing
my proposal, Dr. Mac Spica listened when I was stuck and Matt Jerram contradicted my
ideas—I thank them both. Stephanie Chervinko was essential; she recruited participants
and collected and entered data. Dr. Donna Neumark and Dr. Erin Emery shared the
journey, and I'm glad we went through this process together. Marilyn Eimon, Dr. Perry
Eimon, and their home in Maine, “Tiderok Ocean’s Blue,” provided unforgettable
inspiration and support for writing. Others believed in my ability to complete this task,

and I relied on their encouragement.

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................... viii
LIST OF FIGURES .................................................................................................... ix
INTRODUCTION ..................................................................................................... 1
Factors Contributing to Psychological Abuse ..................................................... 2
Preventing Psychological Abuse by College Students ........................................ 3
Attachment and Feminist Theories ...................................................................... 5
CHAPTER 1
LITERATURE REVIEW .......................................................................................... 6
Definitions of Domestic Violence and Psychological Abuse .............................. 6
Psychological Abuse ............................................................................................ 7
Psychological Abuse Research ...................................................................... 8
Theoretical Foundations ................................................................................ l O
Feminist Theory ....................................................................................... 12
Attachment Theory .................................................................................. 15
Integration of Theory ............................................................................... 25
Attachment, Gender, and Defenses as Predictors of Abuse ................................. 27
Summary ........................................................................................................ 29
Purpose of the Study ............................................................................................ 30
Hypotheses ........................................................................................................... 31
CHAPTER 2
METHOD .................................................................................................................. 33
Participants ........................................................................................................... 33
Procedures ............................................................................................................ 34
Measures .............................................................................................................. 34
Demographics ................................................................................................ 34
Personal Problems Inventory ......................................................................... 35
Experiences in Close Relationships-Revised ................................................ 36
Splitting Index ............................................................................................... 37
Narcissistic Personality Inventory ................................................................. 38
The Dominance Scale .................................................................................... 40
Multidimensional Measure of Emotional Abuse ........................................... 41
Data Analysis ........................................................................................................ 43
CHAPTER 3
RESULTS .................................................................................................................. 46
Preliminary Analyses ........................................................................................... 46
Hierarchical Regression Analyses ....................................................................... 49
Test of Model Predicting Dominance Scores ................................................ 49

vi

Test of Model Predicting Emotional Abuse Scores ....................................... 51

Posthoc Analyses for Dominance 5-
Posthoc Analyses for Emotional Abuse ............................................................... 54
CHAPTER 4
DISCUSSION ............................................................................................................ 57
Summary .............................................................................................................. 68
CONCLUSIONS ....................................................................................................... 75
REFERENCES .......................................................................................................... 81
APPENDIXES ........................................................................................................... l 30
Appendix A. Informed Consent ........................................................................... 131
Appendix B. Demographics ................................................................................. 133
Appendix C. Personal Problems Inventory .......................................................... 135
Appendix D. Experiences in Close Relationships-Revised ................................. 137
Appendix E. Splitting Index ................................................................................ 140
Appendix F. Narcissistic Personality Inventory .................................................. 142
Appendix G. The Dominance Scale ..................................................................... 146
Appendix H. Multidimensional Measure of Emotional Abuse ............................ 149

vii

LIST OF TABLES

Table 1. Gender Differences in Primary Variables .................................................... 103
Table 2. Gender Differences in Psychological Abuse and Narcissism Subscales ..... 104
Table 3. Correlations of Demographic. Predictor. and Outcome Variables .............. 105
Table 4. Correlations of Demographic, Predictor. and Outcome Variables by

Gender ...................................................................................................... 106
Table 5. Adult Attachment Correlations with Psychological Abuse and

Narcissism Subscales by Gender .............................................................. 107
Table 6. Summary of Hierarchical Regression Analysis for Variables

Predicting Dominance .............................................................................. 108
Table 7. Summary of Hierarchical Regression Analysis for Variables

Predicting Emotional Abuse .................................................................... 110
Table 8. Logarithmic Transformation of Emotional Abuse for Regression .............. 112
Table 9. Separate Models for Dominance by Gender ................................................ 114
Table 10. Test of Mediators in Anxiety-Emotional Abuse Relationship ................... 1 l6

viii

LIST OF FIGURES

Figure 1. Hypotheses for Dominance ........................................................................ l 19
Figure 2. Hypotheses for Emotional Abuse ............................................................... 120
Figure 3. Gender-Avoidance Interaction in Predicting Dominance .......................... 121
Figure 4. Standardized Regression Coefficients for Dominance ............................... 122
Figure 5. Avoidance-Narcissism Interaction in Predicting Emotional Abuse ........... 123

Figure 6. Avoidance-Narcissism Interaction in Predicting Emotional Abuse in
Women ..................................................................................................... 124

Figure 7. Avoidance-Narcissism Interaction in Predicting Emotional Abuse in

Men .......................................................................................................... 125
Figure 8. Standardized Regression Coefficients for Emotional Abuse ..................... 126
Figure 9. Posthoc Model of Dominance for Women ................................................. 127
Figure 10. Posthoc Model of Dominance for Men .................................................... 128
Figure 11. Posthoc Mediational Model for Emotional Abuse ................................... 129

INTRODUCTION

Domestic violence is a prevalent global problem. Universally. men rape. beat. and
kill female partners. Depending on their country. 20% to 50% of women experience such
violence (World Health Organization, 1996. as cited by UNICEF. 2000). Because it
occurs significantly more often to women than men, the United Nations considers
domestic violence a human rights violation, and international laws hold governments
accountable to provide equal protection to women (Thomas & Beasley, 1993). Such
substantial gender differences may be explained in part by gender role socialization. Men
may be socialized to be aggressive toward female partners.

In the United States. an estimated 95% of domestic violence is male-to-female
(USDOJ, 1983, 1993). In the 1998 National Violence Against Women Survey (NVAWS;
Tjaden & Thoennes), 25% of women and 8% of men reported being raped or physically
assaulted by a spouse, partner, or date in their lifetime; men perpetrated approximately
90% of this violence. According to the Federal Bureau of Investigation (FBI, 1998), adult
White females are the most frequent victims of family violence, while their spouses and
cohabitating partners are the most frequent perpetrators. Domestic violence, typically
defined in research as sexual and physical abuse. occurs in the context of romantic
relationships. Intimate attachments are often the most intense relationships, which may
help explain such widespread abuse.

The study of domestic violence has expanded in recent years to include
psychological abuse. Female victims often report that psychological abuse is more

harmful than physical abuse (for review, see Murphy & Cascardi, 1999). In 1994 and

2000. the United States passed The Violence Against Women Acts (L. S. Department of
Justice [USDOJ]. 2000). The mandates from the 1994 Act resulted in the following
findings, among others: (a) the causes of violence against women are poorly understood.
(b) the focus in current research on physical abuse neglects harm caused by emotional
and psychological abuse. and (c) the incidence and dynamics involved in psychological
and emotional abuse need further study (USDOJ, 1998). Discovering characteristics of
perpetrators of psychological abuse may illuminate potential causes and inform
prevention and intervention strategies.
Factors Contributing to Psychological Abuse

Known correlates of psychological abuse include: (a) conflict, (b) alcohol
intoxication, (c) approving of physical aggression, (d) being verbally aggressive outside
the home, (e) younger age and fewer children in married couples, (f) relationship
dissatisfaction, (g) stressful life events (for women not men), (h) witnessing interparental
violence. (i) an effort to control others and/or dissatisfaction with power arrangements,
and (j) lack of argument skill (for review, see Murphy & Cascardi, 1999). All
heterosexual romantic relationships encounter conflict, stress, and dissatisfaction, but
numerous individuals refrain from psychological abuse. More information is needed
about individual differences in perpetration of abuse to evaluate risk and to plan effective
interventions.

One study (Dutton, Starzomski, & Ryan, 1996) showed that early experiences of
trauma such as paternal rejection, child abuse, and absence of maternal warmth contribute
to personality formation that leads to abusive behavior in adulthood. One way to examine

how parent—child relationships and personality correspond to adult relationships is

through attachment theory (Bowlby. 1969/1982). Several studies have found a
relationship between insecure or nonoptimal romantic attachments and abuse (Hoover.
Murphy, 8: Taft, 1999; Roberts & Noller, 1998; also see Lyons-Ruth & Jacobvitz. 1999);
however, links between adult attachment style and psychological abuse need further
study. Getting married or becoming engaged is theorized to be a risk factor for abuse (see
Murphy & Cascardi, 1999), suggesting that activation of attachment strategies in young
adults should be investigated. Insecure adult attachment style may amplify the risk of
psychological abuse.
Preventing Psychological Abuse by College Students

Preventing male violence against women is encouraged by the American
Psychological Association and especially by counseling psychologists with their
emphasis on prevention (Albee, 2000; Hage, 2000). Psychological abuse often coexists
with physical violence (see Burgess & Draper, 1989) and results in similar negative
effects (e. g., O’Leary, 1999). Harmful effects include depression, lowered self-esteem,
reduced sense of autonomy, and fearfulness (Davies, Lyon, & Monti-Catania, 1998:
Sakett & Saunders, 1999). Suicide is another risk related to psychological abuse. Women
report that mental stress from psychological abuse leads to suicidal ideation; 35% to 40%
of battered women attempt suicide in the U. S. (Back, Post, & D’Arcy, 1982), and
battered women are 12 times more likely to attempt suicide than other women worldwide
(United Nations, 1989, as cited by UNICEF, 2000). Domestic violence in general and
psychological abuse in particular heighten suicide and homicide risks.

Homicides by intimate partners have decreased over the past 20 years in all

groups except White, unmarried female victims (Puzone, Saltzman, Kresnow, Thompson,

8: Mercy. 2000). In the United States. young women are the population at greatest risk of
being physically and psychologically abused by romantic partners. Women age 19 to 29
are more likely than women of other ages to experience physical violence perpetrated by
an intimate partner (Bachman & Saltzman, 1995). Over three quarters of college-age
women reported experiencing psychological abuse in the previous six months (Neufeld.
McNamara, & Ertl, 1999). Although these statistics tally victim demographics, most
domestic violence occurs in heterosexual relationships, so young adult males may be at
greatest risk of perpetration.

More research is needed on psychological abuse...in early stages of relationship

formation, and in youthful samples....potentially problematic relationship patterns

often begin in adolescence or early adulthood, and escalating coercive processes

often begin near the onset of intimate relationships. (Murphy & Cascardi, 1999,

pp. 221 -222)

There are many reasons to study psychological abuse in young, heterosexual couples,
including informing interventions that reduce risk. Prevention efforts often target easily
accessible groups of potential victims and perpetrators, such as high school and college
students. Effective assessment strategies and interventions with college student
perpetrators are needed.

College students may be particularly at risk of abusive relationships because of
their developmental status. College students often leave home, form new attachments,
mismanage close relationships as part of a relational learning process, and determine the
quality of relationship they prefer. Over half of college students report psychological

and/or physical abuse in their dating relationships (Sugarman & Hotalin g, 1989). College

students are at great risk: in addition. they often seek help at university counseling centers
for relationship problems and may be uniquely amenable to intervention. Because the
effects of abuse are better known and therefore easier to assess than perpetration risks.
this research seeks to contribute to enhanced assessment of perpetration. In addition.
because counseling centers most often counsel individuals, self-reported perpetration as it
might be presented in therapy is observed and not victim or observer perspectives.
Attachment and Feminist Theories

Romantic relationships and gender are primary contributors to domestic violence
but need to be examined in relationship to psychological abuse. In this study, attachment
theory (Bowlby, 1969/1982, 1988) is used to describe individual differences in styles of
relating in romantic relationships. Because secure attachments are optimal, insecurity in
adult attachment style is analyzed in relationship to psychological abuse. Feminist theory
(J. B. Miller, 1976/1986, 1991) provides a sociological perspective of group differences
related to gender, including gender role socialization. Each theory has been utilized to
study physical abuse, but these studies seldom include psychological abuse. In addition,
these two theories have rarely been integrated, though gender role socialization and
insecure attachment orientations combined may be more explanatory of psychological
abuse (see Mauricio & Gormley, 2001 ). Further, adult attachment orientations may have
different behavioral manifestations depending on affective and cognitive coping
strategies (Collins, 1996), including defense mechanisms; psychodynamic theory is used
to illustrate how certain defense mechanisms may increase the risk of psychologically

abusive behavior.

CHAPTER 1: LITERATURE REVIEW

Controversy surrounds domestic violence. in part because it is a relatively new
field of study that is interdisciplinary and international in scope. There are diverse
viewpoints regarding (a) the definition of domestic violence, (b) the nature of
psychological abuse, and (c) theories that best explain abuse in romantic relationships.

Definitions of Domestic Violence and Psychological Abuse

An internationally accepted definition of violence against women is:

Any act of gender-based violence that results in, or is likely to result in, physical,

sexual or psychological harm or suffering to women, including threats of such

acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in
private life. (United Nations Declaration on the Elimination of Violence Against

Women, 1993)

Psychological abuse is defined as behavior used to intimidate or persecute, including:
“threats of abandonment or abuse, confinement to the home, surveillance, threats to take
away custody of the children, destruction of objects, isolation, verbal aggression and
constant humiliation” (UNICEF, 2000, p. 3). In the United States, domestic violence is
defined as a pattern of emotional, sexual, and economic coercion with periodic physical
assaults or severe threats of bodily injury; a cluster of controlling behaviors that serves to
dominate partners so batterers get their own way (Schechter, 1987). Psychological abuse
is an important aspect of a complex array of behavior clustered under the definition of
domestic violence, but it can also exist separately from other types of abuse. In this

inquiry, psychological abuse is the only construct studied.

Psychological Abuse

Broad definitions of abusiveness are recommended by feminists to maximize the
utility of self-report measures (Smith, 1994); therefore. the broadest definition of
psychological abuse is used. Psychological abuse has been characterized as misuses of
power, verbally abusive behavior, and intentional harmfulness. The power and control
wheel (Domestic Abuse Intervention Project [DAIP], date unknown) has been used for
over twenty years to help victims and perpetrators understand psychological aspects of
domestic violence. Tactics used to control a partner within the context of physical and
sexual violence are: emotional abuse; economic abuse; using the children; threats; using
male privilege; intimidation; isolation; and minimizing, denying, and blaming. Other
reports of psychological tactics include calling partners “crazy,” insults, devaluing their
thoughts and opinions, degrading them, forcing partners to use drugs and alcohol (Davies
et al., 1998), breaking partners' favorite possessions, threatening to kill oneself, and
killing the family pet (Murphy & Cascardi, 1999).

Other authors (see Evans, 1992) report that verbal abuse includes: abusive anger,
name—calling, threatening, undermining, and withholding. Sable (1998) described
emotionally abusive adult attachments as consistently threatening, degrading,
humiliating, and controlling; behaviors run from the extreme of outright hostility and
coercion to more covert activities, such as ignoring and discounting. Others (e. g.,
O’Heam & Davis, 1997) suggest that the perpetrator must intend to lower the status of
the victim and that successful emotional abuse damages a victim’s self—esteem and self-
concept. Dobash and Dobash (1984) suggested that men intend to provoke situations in

which they may punish their wives because it is legitimate to do so within a sociocultural

context of patriarchal domination. Marshall (1996). however. suggested that the
intentions of perpetrators are unimportant. Any actions that result in undermining the
victim are psychologically abusive.

Some authors (see Edleson. 1996) caution that there is a need to differentiate
between being mean and being abusive. Several methods of differentiating these
constructs were found. Extremely stressful events that do not happen to most people are
considered traumatic (Schlenger, Fairbank, Jordan, & Caddell, 1997), so severity of a
partner’s behavior is a consideration. Some authors (see Kelly, 1996) theorize that it is
the chronic nature of abuse, whether physical, psychological, or sexual, that leads to the
greatest violation of one's sense of self, sense of others, and sense of self in relation to
others. Thereby, the repetitive pattern or frequency of psychological attack in a
chronically, verbally abusive relationship is important. Other authors (e.g., Roth,
Lebowitz, & DeRosa, 1997) purport that the eflect on the victim determines whether a
tactic is injurious or not. For example, calling someone “crazy” who has just been
diagnosed with depression and prescribed medication may have a deleterious effect, but a
different tactic would be necessary to have the same effect on someone who was not
being treated for mental health concerns.

Psychological Abuse Research

In a recent special issue of Violence and Victims devoted to psychological abuse,
O’Leary (1999) reviewed research and definitions. He concluded that (a) psychological
abuse has a more severe effect on victims than physical abuse, (b) psychological abuse
predicts physical abuse and lowered self-esteem in victims, and (c) the data suggest the

following definition of psychological abuse:

acts of recurring criticism and/or verbal aggression toward a partner. and/or acts

of isolation and domination of a partner. Generally. such actions cause the partner

to be fearful of the other or lead the partner to have very low self-esteem. (p. 19)
In the same issue. Murphy and Hoover (1999) proposed that psychological abuse
damages the recipient's emotional well being or self-concept.

A number of studies have attempted to clarify the concept of psychological abuse
and measure its effects. Sakett and Saunders (1999) derived four factors: (a) ridiculing of
traits, (b) criticizing of behavior, (c) ignoring, and (d) jealous control. Ridicule was rated
as the most severe tactic and ignoring was the strongest predictor of low self-esteem. A
qualitative study (LaVoie, Robitaille, & Hebert, 2000) discovered categories related to:
(a) death threats, (b) denigration and insults, (c) social control and jealousy, (d)
indifference, (e) threats of separation and reprisals, (f) damaging reputations (of girls),
and (g) harassment after the separation. In addition, one strategy reported by female
victims was illustrated with the following quote: “He knows you too well, you see, so he
knows exactly how to hurt you. It’s very subtle” (p. 16). Another gender difference
besides vulnerable reputations in females was that males were the ones motivated to use
violence to gain power or control. Perhaps men are socialized to dominate women,
women are socialized to submit to men, and the potential for women to be vulnerable to
psychological abuse by men is a negative outcome of gender role socialization.

Another study found six principal components of psychological abuse (Sullivan,
Parisian, & Davidson, 1991, as reviewed by O’Leary, 1999). The six subscales were (a)
criticism and ridicule, (b) social isolation and control, (c) threats and violence, (d)

emotional withdrawal, (e) manipulation, and (f) emotional callousness. Marshall (1999)

found orthogonal relationships between overt and subtle psychological abuse. with each
dimension involving a number of factors. The overt dimension included domination,
indifference, monitoring, and discrediting. Subtle items fell into the categories of
undermining, discounting. and isolating. One of Marshall‘s most interesting findings is
that subtle abuse had more harmful effects on recipients than overt acts. A complete
range of psychological abuse tactics is worthy of inquiry.

Theoretical Foundations

Many theories have been applied to the study of domestic violence, but the theory
with the most success publicizing the problem, helping victims, and improving legal
responses through grassroots efforts is feminist theory (see Enns, 1993). More traditional
psychodynamic views associate battering behavior with personality disorders (e.g.,
Scalia, 1994’), and researchers have found evidence of narcissism, borderline personality,
and other personality disorder traits in batterer populations (Dutton, Saunders,
Starzomski, & Bartholomew, 1994; Rothschild, Dimson, Storaasli, & Clapp, 1997;
Tweed & Dutton, 1998). However, the prevalence of domestic violence is too great to be
fully accounted for by such infrequent pathology.

Cognitive-behavioral theorists (e.g., Ganley, 1989) suggest that batterers learn to
use power and control tactics by watching their fathers or other men while they are
growing up. Researchers have consistently found that men who witnessed domestic
violence as children and were victims of child abuse are more likely to perpetrate
domestic violence (see Heise, 1998). Subsequently, based on the assumption that male
perpetrators need to learn different skills in order to relate more effectively to female

partners many practitioners use social learning theory (Bandura, 1977) to intervene with

10

batterers. However. these interventions are insufficient to stop violence (see Nosko &
Wallace, 1988).

Family systems theorists have proposed that violence in couples results from
problems in communication or conflict resolution to which both partners contribute (see
Straus, 1973). Family systems theory has been criticized for: (a) failing to hold batterers
accountable for their violent behavior, (b) putting women at risk by advocating couples
therapy or avoiding implications of violence (Bograd, 1984, 1992), and (c) failing to
recognize historical inequities and power differentials within families that preclude equal
contributions by men and women (Enns, 1993). Even proponents of family systems
theory have reported difficulty applying it to assessing violence in couples (Aldarondo &
Straus, 1994).

Attachment theory has also been applied to domestic violence research. Adult
attachment concepts overlap with personality disorder traits (Karen, 1994), cognitive-
behavioral schema (Crittenden, 1995), and family systems dynamics (Roberts & Noller,
1998). However, adult attachment classifications are more predictive of relationship
outcomes than personality measures (see Lopez, 1995). Attachment therapists combine
cognitive, relational, emotion—focused, and other types of interventions (Bowlby, 1979;
Stosny, 1995), providing opportunities for more complex conceptualizations than
cognitive theory. Attachment theory may be more effective at conceptualizing couple and
family issues than family systems theories (Lindegger & Barry, 1999). Feminist and
attachment theories combined capture sociological, dyadic, individual behavior, and

intrapsychic levels of influence on psychological abuse.

11

Feminist Theory

Feminist psychological theory is concerned with gender similarities and
differences (Aspy & Sandhu, 1999). Societal structures create and maintain an important
gender difference. that men and masculinity are granted more power and status in
American society than women and femininity (Miller, 1976/ 1986). This power
differential is evidenced two ways: economically and physically. Economically, women
are paid lower wages than men for similar work and are underrepresented in positions of
highest power (see Aspy & Sandhu). Globally, “the occurrence of marital violence seems
to be related significantly to the existence of patriarchal norms and values where men
control the wealth and have, thereby, a power advantage over women” (Burgess &
Draper, 1989, p. 86). A feminist analysis of domestic violence attributes one group’s
domination of another group to social norms that allow this to happen (Walker, 1990).

Reports that women are as physically violent as men abound, most often through
use of the Conflict Tactics Scale (see Straus, 1990). Feminists critique this scale because
it counts behaviors without consideration for differences in physical strength of
perpetrators, the purpose of the behaviors, or the severity of resulting injuries (see
Murphy & Cascardi, 1999). DeKeseredy (1999) found that, although the frequency of
reported incidents overall may be similar to men, (a) women seldom initiate attacks; (b)
women report that many violent behaviors are in self-defense or in order to fight back;
and (c) severe violence is uncommon by women, especially the use of weapons.

“A tendency to ‘blame the victim” or project one’s own unwanted vulnerability
or anxiety onto others also leads to destructive perceptions and behavior towards

marginalized groups” (Jordan, 1997, p. 3). Such defensiveness contributes to perpetuation

12

of abuse and misinterpretations of the dynamics involved. It is difficult to interpret
research results in complicated contexts without defensiveness: this problem can lead to
gender bias, in part because psychological explanations often exclude political or
sociological interpretations (Enns, 1993). Feminist research aspires to emancipate women
by describing nonoppressive family relationships and challenging widely held
professional assumptions that are gender biased (Thompson, 1992). Group effects, such
as gender differences, are considered more explanatory of the prevalence of domestic
violence than individual differences.

Gender Differences Research

Lenton (1995), in a comparison of feminist theory to systems theory (Straus,
1973), reported that the data in the Canadian Violence Against Women Survey supports
the culture of male dominance as central to the etiology of wife abuse but insufficient by
itself. She recommended combining gender with other variables. She also critiqued
research focusing on only one gender as unable to contribute to the literature on gender
differences. Instead, she recommended measuring men and women equally in victim
studies and in perpetrator studies to compare the two gender groups on relevant
characteristics. Comparing male and female self-reports of perpetration has the greatest
chance of elucidating gender differences.

In a review of gender difference research on psychological abuse, Murphy and
Cascardi (1999) found equivalent amounts of psychological abuse by men and women in
ordinary situations but much more male than female psychological abuse in violent
situations. It is interesting that reciprocal verbal abuse may be healthier and less risky

than one-sided abuse. Women tended to report more psychological abuse than men

13

regardless of whether they or their partners were the perpetrators. To compare male and
female self-reports, interpretations of results must consider this gender difference.
According to Murphy and Cascardi, women's definition of emotional abuse may be more
inclusive or they may be more willing or able to identify and disclose such events than
men.

In their research review, Cascardi and Vivian (1995) reported evidence of gender
differences in reports of how violent fights began. Women reported that they challenged
men’s authority or confronted men about undesirable behavior before the men became
violent (also see Dobash & Dobash, 1984). Men reported that violence began because
they lost control. These researchers interviewed married couples. Men and women were
asked about both victimization and perpetration of violence. Regardless of the severity of
the violent incident, women reported that their husbands engaged in five to eight times
more psychological coercion leading up to the physical incident. Men reported equal
amounts of psychological coercion by themselves and their wives. Both partners agreed
that physical violence by men followed psychologically coercive acts by men. In fact,
coercion was the reason men gave most often for the escalation. Women more often
reported that severe physical aggression was self-defense. Aggressive men attributed
their abuse to drinking alcohol more often than aggressive women, and aggressive
women were more likely to apologize than men. Female victims reported numerous
injuries, while male victims reported no injuries. Female perpetrators agreed that their
partners were not injured.

The small number of studies leaves a gap in our understanding of gender

differences in domestic violence. Dobash and Dobash ( 1984) suggested that this

14

systematic lack of attention to gender in studies of violence is based on the erroneous
assumption that all violent incidents can be interpreted similarly to fighting contests
between two males. However, because men are socialized to be aggressive and women
are socialized to inhibit aggression, gender role socialization suggests a greater risk for
perpetration of abuse by men. Whether this applies to psychological abuse is unknown.
Attachment Theory

Violence in romantic relationships may be due, in part, to the closeness of these
relationships. Weis (1989) identified the nature of the relationship between perpetrator
and victim as the most important factor to be studied.
Internal Working Models

By emphasizing how persons in close relationships negotiate and cope with
experiences of insecurity, separation, and loss, Bowlby’s (1969/1982, 1973, 1979, 1980,
1988) attachment theory describes individual reactions to connection and separation.
Bowlby theorized that early attachment relationships result in working models of “self”
and “other” that influence adult relationships. Internal working models of relationships
are cognitive schema providing default reactions to stressful relationship situations.
These mental representations are relatively stable and impact the ability to maintain
satisfying intimate relationships and the ability to be autonomous (Ainsworth, 1989).
Adult Romantic Attachments

Attachment theory has been applied to adult romantic relationships (Hazan &
Shaver, 1987). In particular, conditions that threaten the future of relationships, such as
violence, are likely to activate attachment behavior (Feeney, 1999). Patterns of cognitive

and affective responses have been clustered into attachment styles. Theoretically,

15

attachment styles represent distinguishable expectations of relationships that may be
related to internal working models but are based on observable or reported behavior.
Whereas internal working models are often immutable, attachment orientations or
expectations may be more amenable to intervention (see Creasey & Hesson—Mclnnis.
2001). For this reason, studying attachment orientations yields information most relevant
to therapists.

Adult attachment researchers (Bartholomew, 1990; Bartholomew and Horowitz,
1991) have suggested four categories with different combinations of self and other
models (see Griffin & Bartholomew, 1994). The healthiest category is a secure adult
attachment style. Secure people are comfortable with intimacy and aloneness and able to
traverse effectively and flexibly between the two. Internal models of self and other are
positive. Insecure attachments take three forms: preoccupied, dismissing, and fearful. The
preoccupied and dismissing orientations are not optimal but represent alternative means
of attaching developed in early environments that were less than conducive. Preoccupied
people may compulsively seek approval from others, over rely on intimate connections,
and suffer from developmental disruptions in separation and independence. Self models
are negative and other models are positive. Dismissing people prefer not to depend on
others and may be overly self-reliant and detached. Their other models are negative, but
their self models are positive. Fearful adult attachment is often associated with
disorganized attachment in infants, which represents a hi gh-risk category that may be
unattached. Both self and other models are negative.

Attachment categories have been examined for underlying continuous dimensions

that help explain variation within and across self-report categories (Simpson & Rholes,

16

1998). The two continuous dimensions are “anxiety” and "avoidance." Secure individuals
are low in avoidance and anxiety, while fearful individuals are high in both dimensions.
Dismissing individuals report high avoidance but low anxiety. and preoccupied
individuals report the opposite (Bartholomew. 1990). Anxiety is related to fears of
abandonment or insufficient love, while avoidance is related to fear of intimacy or lack of
emotional expression. There are methodological and conceptual advantages to studying
attachment dimensions instead of categories or types (see Brennan & Shaver, 1995;
Fraley & Waller, 1998; Lopez & Brennan, 2000).

C onflict, Aflect, and Cognition

Adult attachment research addresses three relevant aspects of romantic
relationships: (a) conflict. (b) affect regulation, and (c) cognitive distortions. Attachment
avoidance and attachment anxiety predict different behavioral outcomes in romantic
relationships, especially when cognitive misinterpretations and/or emotional distress are
included in mediational models (Collins, 1996).

Anxious attachment contributed to negative feelings and negative appraisals of
self-confidence in emotion management, while avoidant attachment was related to
escalation and anger (Creasey & Hesson-Mclnnis, 2001). Attachment and level of
conflict interact. In high-conflict relationships, preoccupied (anxious) people had the
most complex expectations, especially in terms of positive expectations (Fishtein,
Pietromonaco, & Barrett, 1999). This was similar to previous findings (Simpson, Rholes,
& Phillips, 1996). Attachment anxiety and discussing major (vs. minor) problems led to

more anger and hostility but also more positive perceptions of partners afterwards.

17

Attachment avoidance was related to low levels of supportiveness and warmth. especially
when discussing major problems.

Affect regulation differs across attachment styles. lnsecurely attached people and
men have shown more difficulty directly expressing emotion (Scarles & Meara. 1999).
This gender difference might contribute to risk of psychological abuse. Affect regulation
has been associated with aggressive behavior (Eisenberg, 2000) and described as a
mediator of behavior in attachment relationships (Lopez & Brennan, 2000). Attachment
theory suggests that people use “others as agents of affect regulation” (Fuendeling, 1998,
p. 307), with avoidance and anxiety predicting different methods of signaling and
responding:

Avoidant individuals tend not to perceive or respond to affective signals, whether

of distress or of love and commitment. Similarly, they tend not [sic] signal their

own distress to partners. ...Ambivalent [anxious] individuals, on the other hand,

over-signal their passion and commitment but appear not to accurately perceive

their partners‘ positive affective signals. (p. 31 1)
Avoidance results in fewer positive and more negative caregiving behaviors, while
anxiety leads to excessive caregiving misaligned with the partner’s needs (Kunce &
Shaver, 1994). Therefore, avoidance may result in cold responses while anxiety may
result in intrusive responses. Gender differences in and attachment styles of affect
regulation may account for some variation in psychological abuse.

Another way to describe attachment differences involves cognitive distortion.
Anxious attachment is associated with attributions of self-criticism or blame, perceptions

of partners as experiencing low intimacy and commitment, and low defensiveness;

18

avoidance is associated with high levels of defensiveness (see Fuendeling. 1998). Though
not all abuse occurs in the context of conflict (for discussion. see Johnson & Sacco.
1995), it is likely that both affect regulation and cognitive distortion are involved.
According to psychodynamic theory, defense mechanisms manage affect by distorting
cognition (see Vaillant, 1992). Attachment theory supports the empirical examination of
defense mechanisms using information processing models of cognitive distortions
(Bowlby, 1979, 1988); however, psychodynamic theory has contributed much to our
understanding of specific types of defensiveness.

There is also evidence that cognitive distortions and defensiveness relate to
violence. Apter et a1. (1989) found that displacement, projection, and denial contributed
to violent behavior. Projection and attachment have also been related. Mikulincer and
Horesh (1999) found that anxious-ambivalent people project their own traits onto
unknown others, perhaps in order to increase feelings of similarity and connectedness.
Avoidant people projected their own unwanted traits onto others, perhaps to effectively
maintain positive senses of self. This study used an elegant experimental design to
capture these interesting effects. Given these findings, attachment avoidance and
defensiveness may contribute to psychological abuse.

By definition, defensive processes are unconscious and therefore difficult to
assess using self-report measures. However, two suitable measures were found, one
related to splitting and the other to narcissism.

Defensive splitting. Splitting involves cognitive and perceptual processes and is
“one of the common denominators in all defense mechanisms” (Grotstein, 1985, p. 16).

The individual deals with emotional conflict or internal or external stressors by

19

compartmentalizing opposite affect states and failing to integrate the positive and

negative qualities of the self or others into cohesive images. Because ambivalent

affects cannot be experienced simultaneously. more balanced views and
expectations of self or others are excluded from emotional awareness. (American

Psychiatric Association, 1994, p. 757)

Kemberg (1975/1985) theorized that splitting originates as “good” and “bad”
internalized objects that interfere with the formation of a stable ego identity. In unhealthy
developmental trajectories, differentiation of self and object, integration of positive and
negative self-images, and integration of positive and negative object images may be
disrupted, affecting autonomous and connective processes.

When splitting is used as a defense, affect-laden information is separated into
good and bad categories rather than integrated. For example, a child is aware of a model
of the parent as good and the child as bad (self-splitting), but excluded from awareness
are negative attributes or perceptions of the parent (see Bowlby, 1973; Bretherton, 1995).
Horowitz (1977) described splitting (or other-splitting) as the “tendency to place good
attributes within the self-representation and to extemalize bad attributes” (p. 550).
“Badness” may be attributed to self or other, depending on whether internalization or
extemalization is the means of defense, resulting in unrealistic appraisals that lead to
maladaptive relationship strategies. A more realistic sense of a relational interaction
would involve recognition of both good and bad characteristics of self and other.

Securely attached persons have demonstrated better integration of constructs
related to self and others than insecure persons (see Lopez, 2001; also Mikulincer, 1995).

Two studies linked splitting to insecure attachment (Lopez, 2001; Lopez. Fuendeling,

20

Thomas. & Sagula. 1997). Dutton (1998) described splitting as a defense mechanism
used by male batterers to manage rage. He studied borderline personality and related
primitive defenses (similar to splitting) and found strong correlations with domestic
violence.

Defensive narcissism. Narcissism is not limited to people with narcissistic
personality disorder but involves both healthy and defensive processes found in most
people. Narcissistic traits involve defensive cognitive distortions. Narcissists are
perfectionistic, lack empathy, experience intense envy, and are unable to be autonomous
or intimate (Fischer, 1989; Klein, 1989). Masterson’s (1988, 1993) theory of narcissism
emphasizes a false self that relates to others through defenses, deceptions, and
“distortions in the normal balance of investment in self and others” (Klein, 1989, p. 34).
Defensiveness narcissism may be a mechanism used to express insecure attachment,
especially avoidance. Love is directed toward the self. and hate or aggression is directed
toward important others (Kemberg, 1975/1985). Because partners may be targets,
narcissism may result in psychological abuse.

Expressions of aggression against important others protect the person from
separation anxiety and involve attacks of intense rage that “contaminate the entire
relationship” (Kemberg, 1975/1985, p. 267). Narcissistic rage is unlikely to occur in
adults because of actual interpersonal events; narcissists are usually out of contact with
reality. Rather, the narcissists’ perceptions that they are not being sufficiently
acknowledged are the cause of the aggression (Fischer, 1989). If perpetrators of

psychological abuse are narcissistic, distorted perception may be a precipitant.

21

Narcissism was related to interpersonal hostility in one study (Rhodewalt ck
Morf, 1995,). A series of studies (Carroll. Hoenigmann-Stovall. & Whitehead. 1996a.
1996b) found that narcissism was likely to elicit interpersonal rejeCIion and reduced
interest in interacting with the narcissistic person. Brennan and Shaver (1998) found that
narcissism related most strongly to fearful and preoccupied attachment. followed by
dismissing attachment. Narcissism was least related to secure attachment.

Dutton (1998) described narcissism as an origin of male batterers” rage.
Numerous studies found narcissism in batterers (Dutton et al., 1996; Gondolf, 1999;
Hamberger & Hastings, 1986, as reviewed by Dutton, 1998; Tweed & Dutton, 1998)
Dutton and Starzomski (1997) compared personality traits commonly found in abusers
with uses of the power and control wheel (DAIP, date unknown). Aspects of power and
control were intercorrelated and personality disturbances (primarily borderline traits)
were related to psychological abuse.

Violent Adult Attachments

Only one study was found that explicitly integrated feminist and attachment
theories in the study of domestic violence, though numerous studies linked insecure
attachment and physical violence. Few attachment studies related to psychological abuse
were found, but several projects looked at related concepts. such as controlling behavior
or hostility.

In a study of male perpetrators that integrated attachment and feminist theories,
Mauricio and Gormley (2001) found that insecure attachment and need for dominance
interacted to predict physical violence. Neither attachment nor dominance alone

significantly related to violence, but insecure men with high needs for dominance

reported committing significantly more acts of violence than secure men with low
dominance needs. This preliminary study underscores the value of integrating feminist
and adult attachment theories to explain domestic violence committed by men. Gender
differences in this interaction remain unknown.

In a review of research using attachment interviews (see Lyons-Ruth & Jacobvitz.
1999), several attachment categories related to physical violence. Categories with
elevated frequency of perpetration or victimization included: (a) disorganized/controlling,
which is comparable to the fearful category and related to unresolved trauma in
childhood; (b) preoccupied, which is overwhelmed by trauma; and (c) “cannot classify,”
those interviews that defy classification. Batterers seemed more likely to have
contradictory states of mind associated with the “cannot classify” category, while abused
women expressed unresolved (fearful) or overwhelmed (preoccupied) mental states.
Another study (Dutton et al., 1994) compared abusive men in all four adult attachment
categories to a control group of men and found that the fearful category was related to
abuse, the preoccupied category was somewhat related, and attachment dimensions of
anxiety and avoidance were both related to abuse. Abuse included physical and
psychological components. Dismissing attachment was not related to abuse in the first
study, but the avoidance dimension was related in the second study. The guardedness
involved in this type of insecure attachment may explain discrepancies in findings.

Mayseless (1991) theorized that courtship violence is a form of anger and protest
over separations in adult attachments that serves the function of regulating intimacy. She
asserted that both avoidance and anxiety lead to violence. Avoidance results in passive-

aggressive hostility and criticism, while anxiety results in aggression, jealousy,

23

possessiveness. and wanting control over the partner. There is some evidence to support
this theory. College students‘ attachments to their parents were related to hostility in their
dating relationships in one study (Morrison, Goodlin-Jones, & Urquiza, 1997). Securely
attached adults reported less attacking and protesting behavior in their intimate
relationships than insecurely attached adults. There is evidence that preoccupied and
fearful adults are controlling and aggressive in their romantic relationships (Bookwala &
Zdaniuk, 1998; see also Kunce & Shaver, 1994). Empirical evidence suggests that
distorted perceptions and cognitive processes are involved in this violent behavior. One
study (Kesner, Julian, & McKenny, 1997) found evidence of perpetrator distortions. Men
with insecure attachments (to parents) were more likely to be violent toward female
partners than secure men. It was hypothesized that these men perceived their partners to
be insensitive and inappropriate like their early caregivers, though the study found that
the female partners were not insensitive to the men’s needs. Abusive behavior might be
better described as a form of protest over perceived separations. Insecure attachments
entail risk factors; efforts to regulate intimacy might have the consequence of destroying
intimacy if expectations of relationships are distorted.

Babcock, Jacobson, Gottman, and Yerington (2000) used attachment interviews to
categorize husbands and found a significant difference between violent and nonviolent
men. Violent men were most often classified in an insecure attachment category while
nonviolent men were most often classified secure. Both preoccupied and dismissing men
were more domineering than secure men, attempting to gain compliance and submission
from their wives through invalidating, patronizing, and lecturing them. Further,

dismissing men were most controlling and distancing (using “stonewalling” tactics) while

24

preoccupied men were least distancing. Dismissing men were most contemptuous. and
preoccupied men were most belligerent. These researchers concluded that dismissing and
preoccupied men differ in their motivations for violence: (a) dismissing men are more
antisocial and more instrumental and deliberate in the use of violence to get what they
want from others: and (b) preoccupied men are more expressive, using violence to reduce
negative affect and because they lack strategies for disengaging from conflict. An
understanding of the effects of attachment style differences on psychological abuse in
both genders is needed.

Murphy and Hoover (1999) found that insecure attachment was associated with
perpetration of psychological abuse by females. In a study of college females and
psychological abuse, O’Heam and Davis (1997) found that secure women both received
and perpetrated the least emotional abuse, while preoccupied/anxious women received
and perpetrated the most. Interview ratings supported these findings but self—report
ratings did not. In contrast to past studies of men (Dutton et al., 1994), the more fearful
women were, the more likely they were to report receiving emotional abuse, the greater
the impact on them, and the less likely they were to have inflicted it. Dismissing women
self-reported high levels of inflicting abuse, and the researchers recommend further
investigation into this unexpected phenomenon. The underlying dimensions of
attachment anxiety and avoidance may contribute to perpetration of psychological abuse
in different ways for men and women.

Integration of Theory
Feminist domestic violence research focuses on why men beat female partners,

while attachment researchers focus on psychopathology leading to violence, why couples

are involved in violent interactions. and why women stay with abusive men (see Burgess
& Draper, 1989: Dutton. 1996). Feminists (e.g.. Browne, 1993) critique studies focusing
on why women stay in abusive relationships as neglecting the causes of male violence.
Examining causes of violent behavior related to perpetrator characteristics is compatible
with both theories.

The overlap between gender and individual differences is especially applicable to
the study of domestic violence. Some individual variables that predict domestic violence,
such as personality disorders, also occur with differing frequencies across gender groups
(Renzetti, 1996). It is unknown if there are gender differences in the use of various
defense mechanisms: however. it might be surmised from gender role socialization that
men are more likely to use externalizing defenses such as narcissism and other-splitting
while women are more likely to use internalizing defenses such as self-splitting. Defense
mechanisms are designed to protect negative or vulnerable attributes, and defenses may
even create the appearance of opposite characteristics. For instance, attachment
avoidance is related to highest levels of anxiety though avoidant people typically deny
discomfort (see Mikulincer & Orbach, 1995). Defensive presentations may contribute to
the enactment of psychological abuse.

Adult attachment theory is used to describe individual differences in (a)
expectations of romantic relationships, (b) means of getting needs met by partners, and
(c) understandings of the self and the other in interactions. This approach is supplemented
by components of feminist theory, especially sociological gender differences and the
influence of gender role socialization. In addition, defense mechanisms as described by

psychodynamic theory help clarify aspects of avoidant attachment related to guardedness.

26

This study examines adult attachment style. gender. and defenses as antecedents to
psychological abuse.
Attachment, Gender, and Defenses as Predictors of Abuse

No studies were found to integrate attachment, gender, and defenses to examine
psychological abuse, nor did any studies consider defense mechanisms as contributors to
psychological abuse. Three papers were found that integrate adult attachment and gender
in efforts to explain domestic violence, with definitions inclusive of psychological abuse.

In a review of research, Stosny (1995) reported that men with attachment deficits
were more vulnerable to negative aspects of gender socialization. As a result, insecure
men were more likely than insecure women to be aggressive, to express all negative
affect as anger, and to have difficulties operating within the context of an intimate
relationship. This author characterized both fears over separation and fears of becoming
closer to someone as threats that activate attachment schemas and incite violence in male
batterers. The inability of insecure men to manage intimate relationship responsibilities
combined with a masculine imperative to be competent leads abusers to blame female
partners in order to defensively extemalize negative affect.

Two recent studies supporting these characterizations were found. In the first,
Roberts and Noller (1998) applied attachment theory and systems theory to both male
and female partners. Attachment was related to domestic violence along the dimension of
anxiety over abandonment but not along the avoidant dimension (discomfort with
closeness). Anxious men and women were more likely to use violence against partners,
and women reported more violence against anxious partners. The only interaction

between two partners’ attachment styles as a significant predictor of violence was for

27

anxious perpetrators with avoidant partners. These researchers concluded that couples
with one anxious and one avoidant partner are at greater risk of violence. This finding
seems to conflict with earlier work that found anxious female and avoidant male partners
to be as stable as secure-secure couples, a finding that was interpreted as gender role
adherence contributing to couple stability (Kirkpatrick & Davis, 1994). The data in this
study (Roberts & Noller) more precisely support the conclusion that anxiety is a risk
factor for perpetration against avoidant partners. Interestingly, this implies that in some
cases, gender role reversal (anxious male perpetrators with avoidant female partners) may
contribute to violence. These researchers further characterized partner withdrawal as
provoking violent behavior, a conceptualization that blames victims and risks gender
bias. Two alternative conceptualizations are better supported by previous research: (a)
those fearing abandonment misperceive their partners as withdrawing, or (b) escalation
prior to the violence involved psychological abuse that created more distance in the
relationship or resulted in the partner pulling away.

In a more recent study (Hoover et al., 1999), emotional abuse was the outcome
measure. Men and women were compared on both self-reports and partner-reports.
Gender and self- versus partner-report had main effects and an interaction effect. Gender
had a main effect on two emotional abuse subscales, Dominance/Intimidation and Hostile
Withdrawal. Women tended to report more of these types of emotional abuse than men
whether reporting about themselves or their partners; therefore, interpretations emphasize
differences in self-report rather than actual differences to reduce potential gender bias.
Self— versus partner-report had a main effect on Denigration, with both men and women

reporting themselves as more deni grating than their partners.

28

The interaction between gender and self- versus partner-report significantly
affected the Dominance/Intimidation and Restrictive Engulfment subscales (Hoover et
al., 1999). There was agreement across self- and partner-reports that men were more
likely to use domination and intimidation. Similarly, women were reported to use
Restrictive Engulfment more often than men, whether women were self-reporting or men
were reporting about female partners. Gender was tested as a moderator of the
relationship between attachment and emotional abuse. Gender and anxious attachment
interacted to predict emotional abuse overall and the subscale of Hostile Withdrawal.
When levels of attachment anxiety were high, women reported being more emotionally
abusive than men.

Summary

Self-reported perpetration of psychological abuse is limited in accuracy and
scope, but clinicians often depend exclusively on self-reports and individual observations
to assess risk. Assessment of associated factors might help, such as insecure adult
attachment style, male gender, and the use of externalizing defense mechanisms. This
study measures self-reports of perpetration, attachment style, and defenses in both
genders to clarify risk of psychological abusiveness.

All of these factors are in keeping with feminist principles that perpetrators be
held accountable for their abusive behavior rather than blaming situations or victims.
How these factors interact with each other to contribute to psychological abuse remains
unclear. Risk factors may vary with type of psychological abuse strategy or may depend
on the interaction between attachment style and gender. Male gender is deemed a risk

factor in general, especially for dominance strategies.

29

Insecure attachment style increases the risk that romantic relationships will be
mismanaged. Protest behavior might occur over perceived abandonment in anxious
attachments or over perceived engulfment in avoidant attachments. Insecurity also
increases the likelihood that negative aspects of gender role socialization will manifest in
behaviors or attitudes. Many men may find aggressive behavior acceptable. Because men
are socialized to extemalize problems, they may be at greater risk of using defense
mechanisms that negatively distort perceptions of intimate partners. Attachment
avoidance puts people at more risk of using defense mechanisms that distance them from
important others. Depending on insecure attachment style, gender, and defenses,
psychological abuse strategies employed may vary. Interaction effects between
attachment, gender, and defense may differ depending on whether dominance or
emotional abuse is the behavioral and attitudinal outcome.

Purpose of the Study

The purpose of this study is to extend research on domestic violence to address
psychological aspects of abuse relevant to individual assessment and intervention.
College students were surveyed to extract information about risk factors at this crucial
stage of developing intimate relationships. Gender differences in self-reported
psychological tactics were investigated rather than focusing exclusively on male
perpetrators and simultaneously gathering female victim or observer reports. Feminist
principles were integrated with attachment theory to interpret results.

Adult attachment research is also extended. The contribution of adult attachment
style to psychological abuse in romantic relationships needs further explication. The

model tested included defense mechanisms that may mediate or moderate the relationship

30

between attachment and psychological abuse. These defensive processes integrate affect
regulation with cognitive distortion. Defense mechanisms were theorized to help capture
contributions of attachment avoidance that have gone undetected because people high in
avoidance are also highly defensive. High levels of avoidance combined with high levels
of externalizing defenses (such as narcissism and other-splitting) were expected to predict
psychological abuse. In addition, defensive processes were expected to explain how
anxious attachment leads to abuse. Anxious attachment may produce severe
mismanagement strategies in romantic relationships that can be described as the use of
primitive defenses such as self-splitting, an internalizing defense expected in more
women than men.
Hypotheses

In the model tested, gender and adult attachment dimensions are key predictor
variables expected to have both direct and indirect (i.e., moderating) effects on
psychological abuse by way of their interaction with each other and with defensive
processes. Splitting and narcissism are also predictor variables expected to mediate the
relationships between adult attachment indices and abuse. Psychological abuse is the
dependent variable. Adult attachment variables include anxiety and avoidance.
Psychological abuse variables include indices of emotional abuse and dominance. In
general, this study addresses a basic question: How do insecure attachment, gender, and
defensiveness interrelate to predict perpetration of psychological abuse in romantic
relationships?

It was expected that men would report higher levels of dominance than women

and that highly anxious adults would be most emotionally abusive. Gender was expected

31

to moderate the relationship between attachment and psychological abuse. Men with high
levels of avoidance and anxiety were expected to be more dominant than secure men.
while insecurely attached women (i.e., avoidant and anxious) were expected to be less
dominant than secure women. Anxious and avoidant women were expected to be more
emotionally abusive than more secure women, while a similar but weaker relationship
was expected in men. Narcissism and splitting were theorized to be positively related to
psychological abuse. Self-splitting was hypothesized to predict more psychological abuse
in anxious people than other-splitting. Other-splitting was expected to be more strongly
related to avoidance than to anxiety. Splitting and narcissism were anticipated to partially
mediate the relationship between attachment anxiety and psychological abuse and to
moderate the relationship between avoidance and abuse. Avoidant people who also report
high levels of defensive distortion would admit more psychological abuse (see Figures 1

and 2).

32

CHAPTER 2: METHOD

Participants

Michigan State University undergraduate students were recruited from education.
criminal justice, and other academic courses as well as from residence halls. Efforts were
made to recruit equal numbers of men and women who had been involved in heterosexual
dating relationships. Although 135 participants completed the survey, eight were
eliminated; of those eliminated, three had never had a dating relationship, one had only
gay relationships, and four neglected to answer a full page of items. In the analyses, 127
participants were considered. Using Cohen’s (1977) formula for power (.80), an alpha
level set at .05, and an effect size of .15, it was calculated that 89 participants were
needed to examine 7 independent variables; therefore, there was sufficient power.

There were 66 women (52%) and 61 men (48%) in the sample. They were
predominantly American citizens (USA; 98%) with a few International students (2%).
Most were White (80.6%), though other racial and ethnic groups were represented
(African 9.7%, Asian and Hispanic 3.2% each, Arabic and Native 1.6% each). The mean
age was 20 with a range from 18 to 41. Most were underclassmen (59.8%) but all were
registered for undergraduate courses. Most were recruited from education courses
(61.1%), though a substantial number were recruited from residence halls (33.3%). Most
were heterosexual (98.4%), but a few were bisexual (1.6%). Most were either in a serious
relationship (41.7%) or casually dating (20.5%), though many were not currently dating

anyone (33.1%). Only a few (2.4%) were living together and reported being committed to

33

their relationship. The mean length of the current relationship was 22 months. although
the range was substantial (from less than one month to 26 years).
Procedures

The study was described as concerned with how people act in dating
relationships. During group administrations in classroom settings or residence hall
meetings, participants completed informed consent forms (see Appendix A) and survey
packets containing measures described below. Adult attachment and defense measures
were counterbalanced with psychological abuse measures and demographic questions
during survey administration to control for possible order effects. The participants
recruited from education courses received some partial course credit, and those recruited
from residence halls received slices of pizza. It took participants approximately 35
minutes to complete the survey packet.

Measures

Participant gender was identified using a demographic questionnaire. To control
for differences in levels of stress, a measure of personal problems was used. There were
measures of adult attachment, splitting, narcissism, and two measures related to
psychological abuse. Multiple measures of psychological abuse, including measures that
contain multiple dimensions, increase the accuracy of self-report data as recommended
by feminists (Smith, 1994).

Demographics

A questionnaire (see Appendix B) was developed to gather data regarding gender,

age, race and ethnicity, sexual orientation, and education level. Number of dating

relationships, current relationship status, and length of time in the most recent

34

relationship (or since it ended) were requested. Students were also tracked as to whether
they were recruited from education. criminal justice, or other courses or from residence
halls.
Personal Problems Inventory (PPI; Cash, Begley, McCown, & Weise. 1975: see
Appendix C)

A version of the PPI modified by Ponce and Atkinson (1989) was used to assess
students’ self-reported levels of stress. Participants rated each of 20 problems often
experienced by college students (e. g., alcohol use, conflicts with parents, trouble
studying) according to its severity in their current lives using a 6-point rating scale (1:
not at all a problem; 6 = very significant problem). Ratings were summed to produce a
total score.

The original 15-item scale yielded four reliable factors: performance anxiety,
interpersonal problems, intrapersonal problems, and substance abuse (Johnson &
Holland, 1986). The added items were designed to address concerns of non-mainstream
students, and the 20-item scale yielded two factors: academic/career and personal/social
concerns (Ponce & Atkinson, 1989). Lopez, Melendez, Sauer, Berger, and Wyssmann
(1998) reported adequate reliability for the scale in their sample of college students (or =
.83); they also found that severity of personal problems was significantly higher among
those students who had previously sought and received counseling. Lopez and Gormley
(in press) reported high reliability in a sample of college freshmen (or = .87); they found
that insecurely attached freshmen reported the most problems, securely attached
freshmen reported the least problems, and those who changed attachment style during

freshman year reported moderate levels of problems, with those changing from insecure

35

to secure reporting a decrease in problems. The PPI was reliable in the current sample (a
= .87: see Wilkinson & The APA Task Force on Statistical Inference, 1999).
Experiences in Close Relations/zips-Revised (ECR-R; Fraley, Wallet. & Brennan. 2000:
see Appendix D)

The 36-item ECR-R provides continuous measures of the adult attachment
dimensions of Avoidance (18 items) and Anxiety (18 items). The original items (ECR)
were derived from an extensive review of adult romantic attachment self-report measures.
According to Brennan, Clark, and Shaver (1998), over 300 items were compiled and
tested, and those selected were most strongly associated with the two components (r =
.95 for both). The components (Anxiety and Avoidance) are almost unrelated (r = .11).
The ECR is highly reliable (Avoidance or = .94; Anxiety or = .91) and has construct
validity at r = .95 when compared to categories on the Relationship Questionnaire (RQ;
Bartholomew & Horowitz, 1991 ), a commonly used single-item measure of attachment
categories. Further construct validity is evident in studies that show that anxiety and
avoidance are related in expected ways to self-other differentiation, self-concealment,
emotional empathy, need for approval (Lopez, 2001), work stress intensity, symptomatic
distress (Schirmer & Lopez, 2001), touch aversion, promiscuity, and postcoital emotions
(Brennan et al., 1998).

Further item analysis was done to develop the ECR-R, the version used in the
present study. According to Fraley et al. (2000), the ECR compared favorably to three
other measures of adult attachment in representing each trait continuum. Some items
were replaced to enhance measurement precision up to 100% without expanding the

number of items. Test-retest correlations for the ECR-R subscales ranged from .93 to .95,

36

far exceeding three altemative multiple-item self-report measures and their subscales (r =
.44 to .82). In addition, stability across time was statistically simulated. and the ECR-R
demonstrated greatest stability of the four measures. Both test-retest reliability and
stability were improved in the ECR-R over the ECR.

Avoidance is characterized as discomfort with dependence, and Anxiety is
characterized as the failure to explore without support. Avoidance items include: “I prefer
not to show a partner how I feel deep down” and “I get uncomfortable when a romantic
partner wants to be very close.” Anxiety items include: “I worry a lot about my
relationships” and “I often worry that my partner doesn’t really love me.” Participants are
asked to rate each item on a scale from 1 to 7 (1 = Disagree Strongly, 7 = Agree
Strongly).

In the current sample, the ECR—R was highly reliable (or = .93), as were the
subscales (anxiety or = .91; avoidance or = .92). The subscales were moderately correlated
(r = .40), which may indicate a difference between the ECR-R and the ECR. Attachment
categories were calculated, and 40% of the sample was secure (n = 50), 24% were fearful
(n = 31), 19% were dismissing (n = 24), and 17% were preoccupied (n = 22). These
frequencies were similar to other college student samples that used the RQ to categorize
adult attachment (Bartholomew & Horowitz, 1991; Brennan & Morris, 1997; Duggan &
Brennan, 1994; Lopez & Gormley, in press; Searles & Meara, 1999).

Splitting Index (SI; Gould, Prentice, & Ainslie, 1996; see Appendix E)

The 24-item SI measures the propensity to use ego splitting as a defense (see

Kemberg, 1975/ 1985). There are three subscales related to Self, Family, and Other that

were derived using factor analysis. Only Self-splitting (8 items) and Other-splitting (8

37

items) subscales will be used for this study. so there are 16 items. Items are rated on a
scale from 1 to 5 (1 = Strongly Disagree. 5 = Strongly Agree). A Self—splitting item is “I
feel different about myself when I am with different people." An Other-splitting item is
“I have doubts about my closest friends.” The subscales are highly reliable (self-splitting
(it = .88 to .89; other-splitting (1 = .84 to .85). The two subscales were significantly
intercorrelated (rs = .47 to .56). Test-retest results were shown to be stable over four
weeks (p < .001, r = .86; self-splitting r = .83, other-splitting r =.75). Gender differences
were not significant overall, but in one sample, women reported high self-splitting scores.
The authors provided evidence of construct validity. Splitting was related to affect
(positive, negative, and depressed), narcissistic and borderline personality features, and
dogmatism, but not to intolerance of ambiguity or authoritarianism. Evidence of
predictive validity was also found: (a) the SI differentiated borderline personality
characteristics and severe psychopathology from less severe psychopathology and control
groups, and (b) the SI differentiated between therapist-identified groups of clients who
were either assessed as using splitting as a defense or not (Armbrust, 1997). In the current
sample, the S1 was reliable (a = .84), as were the subscales (self a = .74; other a = .85).

Narcissistic Personality Inventory (NPI; Raskin & Hall, 1979, 1981; see Appendix F)

The NPI is a forced-choice measure with subscales related to Authority,

Entitlement, Exhibitionism, Exploitativeness, Self-sufficiency, Superiority, and Vanity.
This measure was normed on male and female college students, and is reliable for men
and women of all ages ((1 = .83 overall; for males or = .84 and for females or = .82; R.
Raskin, personal communication, January 8, 2001). Each subscale is a principal

component, and the constructs have validity in relationship to observer reports (Raskin &

38

Terry. 1988). The construct of narcissism and each subscale also have convergent and
discriminant validity in relationship to MMPI scales (Raskin & Novacek. 1989). Example
items include: “I have a strong will to power" and “I can usually talk my way out of
anything.” A factor analysis of the NPI yielded three factors: grandiosity. self-
sufficiency, and vanity. Loadings on the grandiosity factor included Authority.
Entitlement, Exploitativeness, Exhibitionism, and Superiority (Raskin, Novacek, &
Hogan, 1991). Though the most recent version of the complete measure has 40 items (R.
Raskin, personal communication, January 8, 2001), the items related to grandiosity are of
most interest to this study. Upon removing Self-sufficiency and Vanity items (Raskin &
Terry, 1988), 31 items remain.

According to Watson and Biderman (1993), the NPI addresses the complex
relationship between positive and negative aspects of narcissism through separate
subscales. Exhibitionism, Exploitativeness, and Entitlement were related to
maladjustment, but Exhibitionism was also related to higher self-esteem. Only
Entitlement and Exploitativeness were uncorrelated with self-esteem (Raskin et al.,
1991). Entitlement was associated with needs for power and revenge (Raskin & Novacek,
1991). Authority, Superiority, Self-sufficiency, and Vanity were found to be less risky
aspects of narcissism (Raskin & Novacek, 1989). Another study (Emmons, 1987) found
four factors in a revised version of the test: Leadership/Authority (UA), Self-
absorption/Self—admiration (S/S), Superiority/Arrogance (S/A), and
Exploitativeness/Entitlement (E/E). The E/E subscale was found to correlate with
measures of maladaptive or pathological narcissism, and L/A was considered adaptive.

E/E was specifically considered interpersonally maladaptive, while both E/E and S/S

39

were correlated with hostility (Rhodewalt & Morf. 1995). The NPI was reliable overall
(or = .87) and the subscales had adequate reliability (L/A or = .69. 5/5 or = .81. S/A or =
.70, E/E at = .68). The NPI and these subscales have been differentially correlated with
various measures of narcissism (see Emmons). For example, E/E was related to an
aggressive/sadistic interpersonal style as well as a rebellious/distrustful style (Emmons,
1984). To preserve the E/E subscale, it was necessary to return one item to the measure
and to add one item (from Emmons), for a total of 33 items. This measure was reliable in
the current sample (a = .81). The subscales were also reliable (grandiosity a = .81; E/E a
= .65).
Tire Dominance Scale (TDS; Hamby, 1996; see Appendix G)

The 32-item Dominance Scale is a psychological abuse measure of
Attitudes and behaviors involved in exerting power over a partner. Based in feminist
theory, this scale relates to non-egalitarian and overly controlling approaches to
relationships. Whether dominance is a form of psychological aggression or a predictor of
psychological aggression is debated (see O’Leary, 1999). The three subscales are
Authority, Restrictiveness, and Disparagement. Authority describes decision-making
power, as in “Sometimes I have to remind my partner who’s boss.” Restrictiveness means
intrusiveness into the partner’s business, such as “I insist on knowing where my partner is
at all times.” Disparagement involves negative appraisals of the partner, such as the
following item: “My partner doesn’t have enough sense to make important decisions.”

Factor loadings and item analysis resulted in reliable subscales (Authority or =
.80, Restrictiveness or = .73, Disparagement or = .82). The Authority subscale is

significantly and moderately correlated with the Restrictiveness (r = .38) and

40

Disparagement (r = .58) scales (both ps < .0001 .). Evidence for construct validity was
provided. All three subscales were significantly related to the psychological aggression
subscale of the CTS2 (Authority r = .35, Restrictiveness r = .33. Disparagement r = .22).
After multiple regression analysis, the Restrictiveness subscale was the one that remained
a significant predictor of psychological aggression ([3 = .26, p < .01 ). The sample was
college students in dating relationships, and validation efforts continue to expand the
usefulness of this measure. TDS was reliable in the current sample (a = .89). The
subscales were also reliable (authority a = .86; restrictiveness or = .70; disparagement a =
.84).

Multidimensional Measure of Emotional Abuse (MMEA; Murphy & Hoover, 1999;

Murphy, Hoover, & Taft, 1999; see Appendix H)

The MMEA is a measure of psychological abuse that was developed on samples
of college students in dating relationships and that yielded four factors that correlate with
physical aggression: Denigration, Dominance/Intimidation, Hostile Withdrawal, and
Restrictive Engulfment. Samples of both males and females were used in the
development of the study. Items were taken from the Psychological Maltreatment of
Women Inventory (Tolman, 1989, 1999), the Psychological Aggression Scale of the CTS
(Straus, 1979), and other sources. An attachment measure was used in the scale
development study (see Murphy & Hoover).

Participants were asked to rate themselves (self) and partners on emotional abuse.
Only responses regarding “self" were used in this analysis. Items are rated by frequency

of occurrence in a recent six-month period, from “never” to “over twenty times.” The

41

measure is reliable (or ranges = .83 to .89 for each subscale on self ratings). MMEA self-
ratings are significantly related to independent measures of physical abuse.

According to Murphy et al. (1999), the measure was reduced to 28 items on the
basis of confirmatory factor analysis. By relating subscales to interpersonal problems,
additional construct validity was provided. In the two studies reported by the authors. all
four subscales significantly and moderately relate to problems being vindictive,
domineering, and intrusive. None of the subscales relate to being overly nurturing,
exploitable, or nonassertive. In one sample of men, Hostile Withdrawal was the only
subscale associated with social avoidance. Social desirability measures had limited
associations with MMEA responses. Dominance/Intimidation and Denigration were
strongly associated with physical abuse, while the other subscales had moderate
associations.

Deni gration (DN) is the use of direct verbal attacks designed to undermine
partners’ self-esteem. Dominance/Intimidation (D/I) tactics such as threats, property
damage, and verbal abuse are directed at creating fear and submissiveness in partners.
Hostile Withdrawal (HW) is a punishing withholding of emotional contact that makes
partners feel insecure about the relationship. Restrictive Engulfment (RE) is the use of
coercion to isolate partners and express possessiveness in order to maximize partner
availability and dependence. Example items include: “Belittled the other person in front
of other people” and “Became angry enough to frighten the other person.” The measure
of reports regarding “self” was highly reliable in the current sample (a = .95), as were the

corresponding subscales (RE (1 = .85; DN (1 = .90; HW a = .89; BI] a = .88).

42

Data Analysis

Bivariate statistics were used to detect demographic differences in variables of
interest so that background effects could be appropriately controlled. Gender differences
in all primary variables, including psychological abuse subscales, were calculated and
reported. Correlations among all independent and dependent variables were calculated for
the entire sample and for each gender separately. The two measures of psychological
abuse were moderately correlated (rs < .63; ps < .01; see Tables 2 and 3) but considered
to measure different constructs. Therefore, separate models were tested for each outcome
variable. Multivariate statistics included a series of multiple regressions used to test the
proposed mediator-moderator models (Baron & Kenny, 1986). An alpha level of .05 was
set.

In preparation for multiple regression, several assumptions must be met. First, it is
assumed that the relationships being analyzed are linear and that all necessary variables
are included in the model (Rencher, 1995). All independent-dependent variable
relationships were observed visually and found to be linear rather than curvilinear. The
inclusion of a potential confound, the stress variable, aids in ensuring that all necessary
variables are in the model. Second, the assumption of independence (Shavelson, 1996) is
met because participants are reporting their own behavior. If their partners happened to
participate, this did not violate the assumption of independence because reports about
partners were not included in the analysis.

Third, the shapes of the distributions of variables were inspected to see if they
were normal (Shavelson, 1996). Three variables, stress, attachment avoidance, and

emotional abuse, had slight positive skews. Only emotional abuse was determined to be

43

nonnormal after viewing Q-Q plots. Because this is a sample of a normal population.
there are numerous participants with low levels and a few participants with high levels of
emotionally abusive behavior. The tails of the emotional abuse variable did not conform
to the trend (see Rencher, 1995); a split histogram showed that this was more a problem
in the male than the female sample. Although the means were similar, men had a wider
range of responses and dispersion than women (see Table 1). In the entire sample, there
was a thick tail at the low end and a thin tail at the high end.

Fourth, error variances were tested to see if they were equal, and residuals were
observed to check whether error terms were correlated (Shavelson, 1996); that is, the
final regression model for dominance was plotted to observe residuals in relationship to
predicted scores, and all assumptions appeared to be met. However, emotional abuse
residuals showed heteroscedacity and some nonnorrnality at the high ends. Though error
terms were uncorrelated, error variances were unequal. Several outliers at upper levels
were inspected and found to be reports of high levels of emotional abuse. To address this
issue, transformation of the data is recommended (Draper & Smith, 1981; Shavelson).
When the emotional abuse variable was transformed with a logarithmic function,
Levene’s test of homogeneity of variance and a Q-Q plot demonstrated homoscedacity
and normality. The hierarchical regressions were run using the transformed variable, and
the same primary variables were significant (presented in Table 8), indicating that the
initial procedure was arguably robust to these violations of the assumptions.

Fifth, multicollinearity between attachment anxiety, stress, and self-splitting was a
concern. Though they were only moderately intercorrelated (rs < .64; ps < .01; see Tables

2 and 3), regression was used to examine multivariate effects. Although these three

variables were significant in predicting each other. only moderate overlap was found (R:
= .408). Therefore, multicollinearity was ruled out (Lewis-Beck, 1980).

Continuous variables were centered in preparation for creating and testing
interaction terms (Aiken & West, 1991; Holmbeck, 1997). The control variable, stress
level, was added first. To test hypotheses, subsequent variables were entered in the order
planned (W ampold & Freund, 1987). Gender was entered second. Attachment avoidance
and anxiety were entered third as a block. Fourth, gender x avoidance and gender x
anxiety interaction terms were entered. Fifth, narcissism, self-splitting, and other-splitting
were entered as a block. Sixth, a block of interaction terms followed: avoidance x
narcissism, avoidance x self-splitting, and avoidance >< other-splitting. The step with the
last change in significance for relationships tested was retained as the final model. This
process was repeated for each outcome variable. The contributions of variables to the
explanation of psychological abuse and the effects of independent variables on each other

in terms of mediation and moderation were considered (Baron & Kenny, 1986).

45

CHAPTER 3: RESULTS

Preliminary Analyses

Demographic variables were assessed for their effects on the variables of interest.
More women were recruited from education courses and more men were recruited from
residence halls [x2 (1, N = 127) = 49.69, p < .001]. To check the effects of gender and
recruitment on primary variables, 2 x 2 analyses of variance (ANOVAs) were used (with
the alpha level set at .01). Because none of the interactions between gender and
recruitment were significant, the effects of recruitment on variables of interest were
assumed to be consistent across gender groups. In addition, none of the main effects of
gender or recruitment demonstrated significant differences in relationship to primary
variables.

Gender differences in all primary variables were examined using descriptive
statistics (see Table 1). Men reported a wider range of problem levels or stress; a higher
mean and wider range of narcissism; greater dispersion and range in dominance; and a
higher mean, dispersion, and range in emotional abuse. Men’s upper ranges on stress,
narcissism, dominance, and emotional abuse measures were higher than women’s. Most
notably, women’s high scores on emotional abuse did not exceed 73, while men had
scores as high as 155. Men had more above-average reports of dominance than women
(the average score of 64 could be gained without admitting dominance). Almost all
students reported some emotional abusiveness, and a subgroup of men reported frequent

and severe emotional abusiveness.

46

When the effects of gender and recruitment on subscales of psychological abuse
and narcissism variables were examined with 2 x 2 ANOVAs (alpha level = .01). the
interactions between gender and recruitment were nonsignificant. Therefore, the effects
of recruitment on subscale scores were assumed to be consistent across gender groups.
Only gender had a significant main effect on subscale scores. Gender had a main effect
on the exploitative-entitled aspect of narcissism [F (1 , 127) = 10.01 , p < .01] and the
authority aspect of dominance [F(1 , 127) = 6.98, p < .01). Men on average reported more
exploitative-entitled attitudes and more misuse of authority in romantic relationships than
women (see Table 2). Though men had higher means and standard deviations in all
subscales, group differences were significant only in the two subscales mentioned above.

Correlational analyses were used to examine interrelationships of primary
continuous variables (see Table 3). Because stress level was moderately related to both
outcome variables, this variable was controlled in subsequent analyses. As expected,
attachment anxiety and avoidance were interrelated, self— and other—splitting were
interrelated, and the two psychological abuse measures were interrelated. Attachment
anxiety had a small but significant relationship with both psychological abuse variables
(dominance, r = .23, p < .01; emotional abuse, r = .25, p < .01). Attachment avoidance
was moderately and significantly related to dominance (r = .33, p < .01) but not
emotional abuse. Narcissism and splitting were related to both outcome variables. Neither
anxiety nor avoidance was related to narcissism. Attachment anxiety was positively
related to stress, self- and other-splitting, dominance, and emotional abuse, while

attachment avoidance was related to all of these variables except emotional abuse.

47

Gender differences in bivariate relationships were examined by running separate
correlations (see Table 4). There was a stronger relationship between the dependent
variables for men, and stress was significantly related to the dependent variables only in
men. None of the primary variables significantly correlated with emotional abuse among
women, and only attachment anxiety and other-splitting correlated with dominance.
Among men, narcissism was significantly related to both psychological abuse variables,
attachment anxiety and self— and other—splitting were related to emotional abuse, and
attachment avoidance was related to dominance. Narcissism was significantly related
negatively to self—splitting and attachment avoidance in women but not men. Attachment
anxiety and avoidance were significantly related to self- and other-splitting in women; in
men, attachment anxiety was related to both types of splitting but attachment avoidance
was only correlated with self-splitting. Both self- and other-splitting were significantly
related to emotional abuse in men but not women, while both forms of splitting were
related to dominance in women but not men.

To further examine insecure attachment orientations as they related to
psychological abuse and narcissism subscales, separate correlations by gender were
inspected (see Table 5). Only significant relationships are described. For women,
attachment anxiety had a small, positive relationship to disparagement, and attachment
avoidance had a small, positive relationship to misuse of authority and a small negative
relationship to grandiosity. For men, attachment anxiety had a small relationship to uses
of intimidation and restrictive engulfment and moderate relationships to restrictiveness
and hostile withdrawal. Attachment avoidance in men was moderately related to misuse

of authority, disparagement, and exploitative-entitlement. Attachment avoidance was

48

related to misuse of authority in both genders. Attachment anxiety was the only primary
predictor related to restrictiveness, restrictive engulfment, hostile withdrawal. and the use
of intimidation, though only in men. Attachment anxiety in women and attachment
avoidance in men were related to disparagement. Attachment avoidance in women was
negatively related to grandiose aspects of narcissism, while attachment avoidance in men
was related to exploitative-entitled aspects of narcissism. Neither insecure attachment
style related to denigration in either gender.
Hierarchical Regression Analyses

Test of Model Predicting Dominance Scores

Summaries of regression analyses for dominance are shown in Table 6. Only
steps involving a change in significance are shown in the tables. At step 1, stress was
significantly and positively related to dominance. Stress alone explained 14% of the
variance in dominance (R2 = .136, p < .001). Higher stress levels predicted more
endorsement of use of dominance strategies. Also shown at step 1 in Table 6, gender was
significantly predictive of dominance scores. Gender explained an additional 3% of the
variance in dominance after controlling for stress level (AR2 = .029, p < .001). Men
demonstrated higher dominance scores than women.

Attachment avoidance was shown to be a better predictor of dominance than
gender in step 2. Attachment avoidance was significantly predictive of dominance. The
more attachment avoidance was reported, the more dominance was reported. Insecure
attachment orientations explained an additional 4% of the variation in dominance (see
Table 6). However, it was the interaction between gender and avoidance shown in step 3

that remained significantly predictive throughout the analyses. The more avoidant men

49

were. the more likely they were to use dominance strategies (see Figure 3). The same was
not true for women. The gender-attachment interactions explained an additional 7% of
the variance in dominance. It is interesting to note that the gender-anxiety interaction had
a negative effect on dominance, though this variable was not significant in the model.

The defense variables are shown in step 4 (Table 6). Although narcissism did not
have the predicted mediator or moderator relationships with anxiety and avoidance, it did
significantly predict dominance. As more narcissism was reported, dominance scores
increased. Narcissism and splitting variables contributed a unique 8% to explaining
variation in dominance scores. Interactions between avoidance and defense mechanisms
did not have significant effects and did not improve the model, and so they were
removed.

In the final model, three variables had moderate effects in significantly predicting
dominance, including stress ([3 = .35, p < .01), the gender-avoidance interaction ([3 = .35,
p < .01), and narcissism (B = .28, p < .01). Higher stress levels predicted more use of
dominance in romantic relationships; however, after controlling for level of stress, other
primary variables had substantial effects. The effect of one insecure attachment style,
attachment avoidance, depended on gender. Men with higher levels of attachment
avoidance reported higher levels of dominance. Higher levels of the defensive use of
narcissism also predicted higher dominance scores. The final model accounted for 36%
of the variance in dominance and is presented pictorially with coefficients in Figure 4.

With or without controlling for stress level, attachment anxiety and splitting
variables were not significantly predictive of dominance in this multivariate analysis. In

the final model presented in Table 6, attachment anxiety had a small relationship to

50

dominance. and the gender-anxiety interaction and self-splitting had small negative
relationships to dominance: however. these relationships were not significant. Posthoc
analyses separating gender groups further examined the impact of attachment anxiety and
splitting on dominance, and these analyses are described below.

Test of Model Predicting Emotional Abuse Scores

Hierarchical regression results for emotional abuse are shown in Table 7. In step
1, stress moderately and significantly predicted emotional abuse. Higher levels of stress
predicted more frequent and severe reports of emotionally abusive behavior. Stress alone
accounted for 18% of the variation in emotional abuse (R2 =.175, p < .001). Gender, also
shown in step 1 (Table 7), was not a significant predictor of emotional abuse. Gender
explained less than 1% additional emotional abuse variance (AR2 = .008, p < .001).

In step 2, insecure attachment orientations were not significant predictors and
explained no unique variance in emotional abuse (AR2 = .001, p < .001). Gender-
attachment interactions were not significant predictors, though adding them to the model
explained an additional 2% of the variance in emotional abuse (AR2 = .015, p < .001).
Splitting variables were not significant predictors of emotional abuse but narcissism had a
small, positive effect. The more narcissistic people were, the more emotionally abusive
they were. Defense mechanisms contributed a unique 7% to the explanation of emotional
abuse (AR2 = .065, p < .001).

In step 3, defenses were assessed as moderators of the relationship between
attachment avoidance and emotional abuse. Narcissism significantly moderated the
relationship between attachment avoidance and emotional abuse, though the effect was

small. With higher levels of avoidance, those who also reported higher than average

51

levels of narcissism were more emotionally abusive. but this relationship did not hold
true for those with average or below levels of narcissism (see Figure 5). When this
interaction was examined by gender, interesting differences were found. For women
reporting higher than average levels of narcissism, higher levels of attachment avoidance
predicted more frequent emotional abuse (see Figure 6). For women reporting average or
lower levels of narcissism, higher levels of attachment avoidance predicted less frequent
emotional abuse. For men, higher levels of attachment avoidance predicted more frequent
emotional abuse, but those with higher than average levels of narcissism had higher
emotional abuse scores regardless of level of attachment avoidance (see Figure 7).
Adding defensive moderators of attachment avoidance to the model contributed 4% to
explaining variance in emotional abuse (AR2 = .036, p < .001).

Stress (B = .31, p < .01) and narcissism (B = .31, p < .01) had moderate effects and
the avoidance-narcissism interaction ([3 = .19, p < .05) had a small effect in significantly
predicting emotional abuse. The final model explained 30% of variation in emotional
abuse and is presented with coefficients in Figure 8. Because of concerns about whether
the emotional abuse variable met assumptions for regression (see Data Analysis), the
variable was submitted to a logarithmic transformation. When the hierarchical regression
analyses were run with the transformed emotional abuse variable, the same primary
predictors were significant but stress did not remain significant in the final model (see
Table 8). Narcissism and the interaction between narcissism and attachment avoidance
were the most robust predictors of emotional abuse. Higher levels of narcissism predicted
increased frequency in emotional abuse tactics. Higher levels of attachment avoidance

predicted more emotional abuse only when levels of narcissism were high.

52

Posthoc Analyses for Dominance

Posthoc analyses were done in an effort to understand why attachment anxiety
was not related to psychological abuse in planned models. To better understand gender
differences in the anxiety-dominance relationship, separate regression models were run
for each gender (see Table 9). These analyses were exploratory, and because of
limitations in sample size by gender, significance is not reported for individual variables
but only for models overall. At step 1 for women, attachment anxiety predicted
dominance. Approximately 9% of the variation in dominance was explained by stress and
insecure attachment. However, the model overall was not significant at this step [F (1 , 66)
= 1.10 , p = .299], indicating a likelihood that all effects could be zero. At step 2, other-
splitting predicted dominance for women (B = .28, p < .05). Given the nonsignificance of
the model at the first step, other-splitting may be the only variable in this model that
predicts dominance in women. The final model is presented in Figure 9.

At step 1 for men, stress and attachment avoidance moderately predicted
dominance (see Table 9). Increased levels of stress and higher levels of attachment
avoidance predicted higher dominance scores. In addition, attachment anxiety was
negatively related to dominance. Less than average attachment anxiety in men, or
increased attachment security, was related to greater than average use of dominance
strategies. With stress and insecure attachment in the model, 41% of the variation in
dominance was explained (see Table 9).

At step 2, with all relevant variables in the model, narcissism was better at
explaining dominance than attachment security. Stress had a large effect (B = .70) on

dominance scores among men. Attachment avoidance (B = .49) and narcissism (B = .33)

53

had moderate effects on dominance scores. The more stressed. the more avoidant in
attachment orientation. and the more narcissistic men were, the more likely they may be
to self-report perpetration of psychological abuse using dominance strategies. The final
model explained 62% of the variation in dominance (see Figure 10).

Posthoc Analyses for Emorional Abuse

For emotional abuse, the overlap between stressors and attachment anxiety was
considered. It is likely that stress or personal problems manifest as a result of attachment
anxiety rather than vice versa. To test whether stress mediated the relationship between
attachment anxiety and emotional abuse, their order in a hierarchical regression analysis
was reversed (see Table 10). The gender—attachment interaction terms were left out of the
model because they were not previously significant predictors.

In the first step, attachment anxiety had a small, positive effect in significantly
predicting emotional abuse. With gender and insecure attachment terms in the model, 8%
of the variation in emotional abuse was explained (R2 = .079, p < .05). In step 2,
narcissism moderately and significantly predicted emotional abuse, which is similar to
findings in the planned test. Adding defense mechanisms to the model contributed a
unique 10% to the explanation of emotional abuse (see Table 10). Attachment anxiety
was no longer significant at step 2. The relationship between attachment anxiety and
emotional abuse was mediated by self—splitting; self-splitting was marginally significant
(p < .056) at the second step and became significant at step three when related interaction
terms were added.

At step 3, narcissism remained a significant predictor. Self-splitting became a

significant predictor of emotional abuse with a small, positive effect. The avoidance-

54

narcissism interaction was significant as in the planned analysis. In addition. the
avoidance-other-splitting interaction became significant. Other-splitting moderated the
effect of attachment avoidance; higher levels of avoidance predicted more emotional
abuse only among those who also reported greater than average use of other-splitting.
Attachment avoidance, when moderated by high levels of narcissism or other-splitting.
had a small, positive effect in predicting emotional abuse. Moderator relationships
explained an additional 6% of the variation in emotional abuse (AR: = .057, p < .001).

In the final step, the main effect of narcissism and the interaction between
narcissism and avoidance remained significant. Stress was a significant predictor of
emotional abuse as in the planned analysis. Splitting variables were no longer significant.
It is arguable that stressors or problems result from the use of splitting as a defense rather
than vice versa; therefore, the significant effects of self— and other-splitting on emotional
abuse were mediated by stress. This model is presented in Figure 11. When a stress-
anxiety interaction term was added to test for moderator effects, it was not significant (B
= -.1 1, p : .359), so it was removed from the model. High levels of stress were not found
to increase the anxiety-abuse relationship.

The findings were similar when the logarithmic transformation of the emotional
abuse variable was used to check assumptions (see Data Analysis); however, the step that
included attachment anxiety as a significant predictor was nonsignificant overall [F (1 ,
127) = 2.04, p = .11]. There is a high probability that the effect of attachment anxiety on
emotional abuse could have been obtained even if it equaled zero. This reduces
confidence in the mediational model of attachment anxiety presented here, given the

assumptions of regression analyses. When a split—gender regression was performed on

55

emotional abuse. attachment anxiety was not a significant predictor for either group.
Little evidence that high levels of attachment anxiety predicted more frequent

psychological abuse was discovered in posthoc analyses.

56

CHAPTER 4: DISCUSSION

This study examined the psychological abuse aspect of domestic violence in an
attempt to discover causes and risk factors that are assessable by clinicians. College
students were studied to inform early intervention practices and to improve
understandings of how this developmental period contributes to the formation of intimate
relationship problems. Gender differences and attachment insecurity were examined as
influences on abuse. Defense mechanisms were included to describe potentially
destructive responses to stressors and to elucidate how attachment avoidance contributes
to abuse. Gender was hypothesized to predict dominance and to moderate the relationship
between attachment insecurity and psychological abusiveness. Attachment insecurity was
expected to predict psychological abuse. Narcissism and splitting were expected to
predict psychological abuse, to moderate the avoidance-abuse relationship, and to
mediate the anxiety-abuse relationship. Discussions of each predictor, whether related
hypotheses were supported, and how the findings fit with previous literature follow.
Gender influences are portrayed first, followed by attachment avoidance, attachment
anxiety, narcissism, splitting, and stress. Interaction effects and posthoc findings are
elaborated after primary predictor effects have been articulated.

The hypothesis that men would be dominant more frequently than women was
marginally supported, especially in terms of misuse of authority. The main effect of
gender on dominance was not as explanatory as the interaction between gender and
avoidance, however, which is described below. Gender was not expected to predict

emotional abuse, nor did it. It is interesting that the ranges of response were so different,

57

however. with only" men reporting very high levels of emotional abusiveness. Murphy
and Cascardi (1999) found that men and women reported roughly equivalent
psychological abuse unless their romantic relationships were violent, in which case men
were more psychologically abusive. The high levels of emotional abuse reported by some
men in this sample may be an indicator of a violent subpopulation, though violent
behavior was not measured directly in this study.

The hypothesis that attachment avoidance would be related to psychological
abuse was supported in that avoidance was correlated with dominance in men. The goal
of this project, to demonstrate that attachment avoidance is involved in psychological
abuse when moderated by gender and defensiveness, was achieved. Narcissism
moderated the relationship between avoidance and emotional abuse, which is described
below. Gender moderated the avoidance-dominance relationship in partial support of the
hypothesis that gender would moderate insecurity-abuse relationships. Men were more
dominant the more avoidant they were, but there was no relationship between avoidance
and dominance for women. Men with high levels of avoidance were prone to disparage
their partners and to use their authority to manage relationships. Avoidant attachment was
the strongest predictor of dominance for men. The hypothesis that insecure men would be
more dominant than secure men was partially supported in terms of attachment
avoidance. The hypothesis that insecure women would be less dominant than secure
women was not supported. Overall, the use of dominance strategies is best predicted by
considering both gender and adult attachment. Avoidant college-age men are most at risk

of dominating female romantic partners.

58

This finding replicates a previous study that related attachment avoidance and
abuse in men (Dutton et al., 1994). The relationship between avoidance and abuse can be
explained by attachment theory as: (a) high levels of defensiveness based on the
expectation that others will not be responsive to expressed needs (Simpson & Rholes.
1994); and (b) a lack of sensitivity to others characterized by fewer positive and more
negative caregiving behaviors, especially coldness (Kunce & Shaver, 1994). Avoidance
is related to difficulties with intimacy (Bookwala & Zdaniuk, 1998). There is also
evidence that the motivation behind abuse for those high in avoidance is instrumentality
or getting what one wants from another and not affect regulation (Babcock et al., 2000).
The instrumental use of abuse is described by feminists as maintenance of power and
privilege accorded to men in a patriarchal society (see Miller, 1976/1986, 1991), which
may explain why this relationship was found for men but not women.

Theorists combining feminist and attachment constructs (Sheinberg & Penn,
1991) have suggested that insecure men and women have relational difficulties in areas
associated with the opposite gender; therefore, men have difficulty expressing dependent
feelings and women have difficulty expressing independence in intimate relationships.
There is also empirical evidence that psychologically abusive and violent men are
perceived as more traditional in terms of gender roles by female partners (Ellington &
Marshall, 1997). Male gender role socialization promotes independence, attachment
avoidance promotes independence (Pietromonaco & Barrett, 2000), and attachment
insecurity increases vulnerability to gender role socialization, so avoidant men in

particular may fear closeness (Stosny, 1995).

59

Young avoidant men may be unable to express dependency needs in romantic
relationships or use these relationships to fosterjoint coping with stress. instead finding
effective ways to push others away (Simpson & Rholes. 1994). Their efforts serve to
maintain their gender identities as masculine, independent, and in charge but put them at
risk of perpetrating relationship violence. The combination of dominance and insecure
attachment in men has been found to predict physical abusiveness (Mauricio & Gormley,
2001). Male gender role socialization and avoidant attachment serve to limit coping
strategies that preserve and rely on intimate relationships. Avoidant men may sacrifice or
damage romantic relationships in their efforts to preserve themselves. University
counseling centers may find that some of these men seek counseling to understand what
went wrong in their relationships, despite the fact that they seem confident about their
choices, behaviors, and attitudes. However, outreach efforts are recommended because
men and those high in attachment avoidance seldom seek help (see Fuendeling, 1998).

Attachment anxiety was correlated with dominance, especially restrictiveness in
men and disparagement in women. Attachment anxiety was also correlated with
emotional abuse in men, especially restrictive en gulfment, denigration, and hostile
withdrawal. That attachment anxiety would be more related to intrusive or restrictive
aspects of psychological abuse than attachment avoidance was anticipated, but it was not
expected that this would be true only for men. The hypothesis that attachment anxiety
would predict emotional abuse was not supported, however. Posthoc analyses examined
variations from planned models, but the anxiety-abuse relationship was not replicated in
this study. The hypothesis that insecure women would be more emotionally abusive than

secure women, more so than for men, was not supported. Surprisingly, attachment

60

anxiety predicted less dominance by men. though gender did not moderate the
relationship between attachment anxiety and psychological abuse as expected.
Additionally, attachment anxiety was not correlated to narcissism: therefore. narcissism
could not mediate the anxiety-abuse relationship as theorized. Rather, stress was a better
predictor of psychological abuse than attachment anxiety. Though stress and attachment
anxiety were moderately correlated, stress levels predicted psychological abuse while
anxiety did not.

This finding was surprising given previous research suggesting that attachment
anxiety predicts abuse (Dutton et al., 1994; Roberts & Noller, 1998), though this research
emphasized physical rather than psychological abuse. Perhaps attachment anxiety is
related to physical acting out more than subtle forms of mental and emotional abuse. It is
possible that anxiously attached people act out sooner under stress than avoidantly
attached people. Those high in avoidance defend against feeling the anxiety as long as
possible (see Simpson & Rholes, 1994) and, therefore, may be more at risk of
perpetrating psychological forms of abuse.

The differential effects of attachment anxiety and avoidance on physical versus
psychological abuse need further clarification; however, the subscales of the dependent
measures in this study have been differentially linked to physical abuse (Hamby, 1996;
Murphy et al., 1999). The strongest predictors of physical aggression are: (a) the
restrictiveness subscale of the dominance measure, which in this study was related only
to attachment anxiety in men; (b) the dominance/intimidation subscale of the emotional
abuse measure, which was related to attachment anxiety in men, and (c) the denigration

subscale of the emotional abuse measure, which was not significantly related to

61

attachment insecurity in either gender. The other subscales with moderate predictive
value regarding physical abuse are the restrictive engulfment scale, which was related to
attachment anxiety in men. and the hostile withdrawal scale. which was related to
attachment anxiety in men. Given these correlations, attachment anxiety in men is
implicated in types of psychological abuse that lead to physical abuse. Perhaps the
conflict generated by a dependent attachment style and male gender role socialization that
calls for independence contributes to physical abuse; however, in the models tested,
attachment anxiety was not a substantial contributor to psychological abuse, even among
men.

In another study (Hoover et al., 1999), attachment anxiety interacted with gender
to predict emotional abuse and the subscale of hostile withdrawal. In the current sample,
three subscales of the emotional abuse measure, including hostile withdrawal, were
correlated with attachment anxiety in men. However, attachment anxiety was not a
significant predictor in regression models. None of the aforementioned studies that
related attachment anxiety to abuse controlled for stress levels, which may have
confounded their results. Alternatively, the overlap between measures of stress and
attachment anxiety in the present study is a consideration; future studies that utilize
similar measures with less overlap would help clarify these relationships.

The hypothesis that defensive processes would be related to psychological abuse
was partially supported. Narcissism and other-splitting were correlated with both
dominance and emotional abuse, and self-splitting was correlated with emotional abuse.
Among men, all three defenses were correlated with emotional abuse, and narcissism was

correlated with dominance. Among women, the only correlation was between other-

62

splitting and dominance. Narcissism predicted both types of psychological abuse and also
moderated the avoidance—emotional abuse relationship. As higher levels of narcissism
were reported, higher levels of psychological abuse were reported. At higher levels of
narcissism, higher avoidance predicted more frequent emotional abuse. while at lower
levels of narcissism there was no such relationship. Those participants with avoidant
attachment orientations who were also defensively entitled and exploitative were
emotionally abusive to romantic partners. This supported the hypothesis that those high
in avoidance who also reported high levels of defensiveness would admit more
psychological abuse.

There were some interesting gender differences in this relationship, however. As
expected, men were more likely to use an externalizing defense; they were more
narcissistic than women. Men reported higher levels of exploitative-entitled attitudes. For
men, narcissistic defenses predicted dominance. Higher levels of attachment avoidance
predicted more frequent emotional abuse in men, but higher levels of narcissism
increasedthe frequency of emotional abuse across all levels of attachment avoidance. In
women, the relationship between attachment avoidance and emotional abuse depended on
level of narcissism. Highly avoidant women with high levels of narcissism reported more
frequent emotional abuse like men, but when narcissism levels were low, higher levels of
attachment avoidance predicted less frequent emotional abuse. In women, entitled and
exploitative attitudes were correlated with hostile withdrawal. When women used
externalizing defenses, they were psychologically abusive. Women with high levels of
other-splitting, regardless of attachment style or level of narcissism, reported higher

levels of dominance.

63

The narcissistic defense is a risk factor for psychological abusiveness. Narcissism
was previously related to wife abuse (Dutton et al., 1996), and men have been found to be
more narcissistic than women (see Carroll et al., 1996b). A feminist explanation is that
this difference is the result of gender role socialization. There is some support for this
idea in the research literature. The entitlement aspect of narcissism was considered more
negative when displayed by women than men, indicating that gender role socialization
and stereotyping encourages narcissism in men and discourages it in women (Carroll et
al.). To maintain a masculine gender identity, men may develop narcissistic
characteristics despite negative consequences to their romantic relationships, and these
characteristics may be considered socially acceptable in men.

That socialization differences in narcissistic development contribute to
psychological abuse helps explain why men as a group are more abusive than women as a
group, and also why some individual women are abusive. It is interesting that avoidant
and narcissistic men are psychologically abusive, but in women both attachment
avoidance and narcissism must be present to predict psychological abuse. When women
violate gender role stereotypes (i.e., using externalizing defenses such as narcissism and
other-splitting) and act more like men are socialized to behave, women may be at similar
risk to men of becoming psychologically abusive. Whether negative feedback that
women receive for narcissistic traits or the use of externalizing defenses negatively
impacts their sense of self (in comparison to men who are reinforced for similar behavior)
is worthy of further study.

Avoidance and narcissism were separate risk factors for dominance in men, but

both avoidance and narcissism were necessary to predict emotional abuse in the entire

sample and in women. Men get what they want from women through the use of
dominating behavior (Babcock et al.. 2000); men benefit from gender role accordance in
a patriarchal society (Enns, 1993) by being dominant, narcissistic. and overly self-reliant;
and narcissism is considered acceptable in men but not women (e.g.. Carroll et al..
1996b). Gender role socialization seems to contribute to entitled, overly self-reliant
expectations that contribute to male perpetration of psychological abuse. The meaning of
the interaction between avoidance and narcissism may be different for women.
Avoidance in women was negatively correlated with narcissism, and higher levels of
avoidance were related to lower levels of grandiosity. Avoidant women may be less
likely to be narcissistic. However, avoidant women may be less invested in what others
think of them and more likely to violate gender role stereotypes. Women who step out of
the female gender role by using narcissistic or other externalizing defenses, such as other-
splitting, may be at risk of psychologically abusive behavior like their male counterparts.
Narcissism was previously related to insecure attachment categories, though not
the dismissing category (Kelley & Gelso, 2000). Why narcissism predicted psychological
abuse in the present study but was related only to avoidant attachment as a moderator
requires explanation. This discrepancy may have resulted because past measures of
attachment calculated attachment categorically while this study utilized continuous
underlying dimensions. Also, this finding is somewhat similar to previous research
(Rapoza & Malley-Morrison, 1997) that found self-esteem only related to secure
attachment, though in men it was also related to dismissing attachment. Though self-
esteem and narcissism are not the same construct, they are conceptually related as healthy

and unhealthy positive senses of self. Brennan and Bosson (1998) found that the

65

dismissing (avoidant) attachment category differed from the secure attachment category
in two fewer sources of self-esteem: personal growth and positive relations with others.
George and West ( 1999) cautioned that the theoretically positive sense of self in
dismissing people may actually be a defensive self-presentation that hides a negative
sense of self. Further examination of the relationship between defensive and actual
positive self-representations in attachment avoidance is warranted. The positive sense of
self in securely attached people, or genuine self-esteem, may differ from the positive
sense of self in highly avoidant people that may reflect defensive narcissism. A
defensively positive sense of self may have negative interpersonal consequences that
differentiates avoidant from secure attachment. Current findings relating narcissism and
psychological abuse may help explain previous findings that narcissism predicted
interpersonal rejection (Carroll et al., 1996a, 1996b). Psychologically abusive behavior
resulting from narcissism has negative interpersonal consequences.

Trends reported above were significant even after controlling for differences in
stress level. Severity of problems or level of stress was correlated to both types of
psychological abuse. In previous research (see Murphy & Cascardi, 1999), this was true
for women but not men. In the current study, the relationship between stress and
psychological abuse held true only for men. Stress associated with high levels of self-
reported personal problems predicted a large portion of male dominance, which is
important information for college counseling centers. Male college students presenting
with commonly seen problems such as substance abuse should be screened for risk of
perpetrating abuse, including assessment of attachment style and gender role adherence

(see Enns, 1993), as one means of identifying ineffective coping strategies. In addition,

66

all students presenting with common problems should be assessed for attachment style.
the use of externalizing defenses, and perpetration of psychological abuse in romantic
relationships. These risk factors not only interfere with coping. identity. and relationship
skill development, but also they may undermine college students' support networks.

When stress was tested as a mediator, splitting predicted emotional abuse as
hypothesized. The hypotheses that self-splitting would be more strongly related to
attachment anxiety and that other-splitting would be more strongly related to attachment
avoidance in predicting psychological abuse received partial support. Self-splitting
mediated the anxiety-emotional abuse relationship, and other-splitting partially
moderated the avoidance-emotional abuse relationship only without controlling for stress
level. Avoidant people who also use other-splitting as a defense may be more emotionally
abusive than avoidant people who do not use this defense. Stress effectively mediated the
effect of splitting on emotional abuse. Other-splitting helped explain women’s use of
dominance strategies.

Correlations among splitting variables, stress, attachment avoidance, and
attachment anxiety corroborated earlier findings (Lopez, 2001; Lopez et al., 1997) with
two exceptions. In these studies and the present study, significant, positive relationships
were found between attachment insecurity and splitting, more moderate ones with
attachment anxiety and smaller ones with attachment avoidance. Gender differences in
other-splitting were found by Lopez (2001), with men reporting higher scores than
women. In the current study, no gender differences were found; however, attachment
avoidance and other-splitting were not significantly correlated for men, the only

nonsignificant relationship between insecurity and splitting for either gender in both of

67

these studies. In the other study (Lopez et al.. 1997). high stress and insecure attachment
interacted to predict splitting, illuminating the relationship between the dismissing
attachment category and splitting. In the present study, high stress did not moderate the
anxiety-abuse relationship. Stress was a better predictor of emotional abuse than
attachment avoidance or other-splitting, but without stress in the model, there was an
interaction between avoidance and other-splitting. Additional exploration of
interrelationships in attachment avoidance, high stress, and other-splitting as they
correspond to domestic violence is recommended.
Summary

Attachment avoidance is an important contributor to psychological abuse in
dating relationships and explains psychological abuse better than attachment anxiety.
Attachment avoidance implies a more negative response to conflict than attachment
anxiety (Creasey & Hesson-Mclnnis, 2001; Fishtein et al., 1999). Conflicts during the
college years may aggravate developmental issues related to both identity and intimacy;
those high in avoidance may be overly inclined to resolve conflicts by preserving their
identity, losing ground in the area of intimacy skills. The need to gain distance from a
partner in order to cope, a deactivation of the attachment system, is problematic for
romantic relationships. Distancing tactics suggest that the perpetrator of such tactics is
enacting withdrawal to avoid intimacy, while the defensive need to hide this by
criticizing others demonstrates how this can be cognitively distorted as the other person’s
withdrawal. Failure to signal distress exacerbates negative effects of deactivation on the
relationship. Indirect expression of needs may occur through acting out in abusive ways.

It is interesting that insecurity more effectively predicted abuse by men than women. This

68

replicates findings that highly insecure women are least likely to perpetrate abuse
(O’Heam & Davis, 1997). Perhaps in order to abuse others one must feel somewhat
powerful, and men are more likely to feel powerful over women than vice versa given
gender role socialization of men as aggressors and women as submitters.

The utility of including cognitive and affective variables in models that attempt to
explain behavior from attachment styles (Collins, 1996) is demonstrated by this study. It
was necessary to consider gender and defenses as moderators to capture the relationship
between attachment avoidance and psychological abuse. Only attachment avoidance and
male gender together predicted the use of dominance, supporting the need for integration
of feminist and attachment theories in the study of domestic violence. Gender differences
found were subtle and potentially interpretable as due to gender role socialization or
gender “schema.” Extemalizing defenses, often associated with male gender role
socialization, helped predict psychological abuse. In particular, narcissism predicted
psychological abuse, occurred primarily in men, and exacerbated the effects of
attachment avoidance on emotional abuse. The interaction of narcissism and avoidance is
interesting; narcissistic misperception of being insufficiently acknowledged by another is
theorized to cause aggression (Fischer, 1989), while misperceptions regarding another’s
lack of availability are related to attachment avoidance. These two types of distortions
combined lead to erroneous judgments about and misbehavior in relationships;
narcissistic defenses help illustrate the implications of extreme levels of underlying
anxiety in attachment avoidance (Mikulincer & Orbach, 1995). Studies examining
attachment avoidance as a predictor of behavioral outcomes should include defensive

measures in moderator models.

69

Narcissism. male gender. and attachment avoidance are characteristics related to
perpetration of psychological abuse in college students. Narcissism is a defensive effort
at self-preservation that may be activated by an attachment threat such as a desire for
connection or a partner’s request for closeness; however, narcissistic responses to
avoidant expectations are psychologically abusive and damage relationships with
important others. Putting self-preservation before relationship preservation is
stereotypically part of the male gender role that values independence over
interdependence or dependence. A sense of entitlement, a belief that one has the right to
exploit others, and a grandiose sense of self are associated with male privilege.
Ultimately, however, damaging important relationships erodes the sense of self that these
efforts strive to maintain.

Though it is surprising that attachment anxiety was not a significant risk factor for
psychological abuse, this may be due to an inherent desire to preserve relationships and a
higher tolerance for conflict (Fishtein et al., 1999). In fact, men with high levels of
attachment anxiety were less inclined to be dominant than securely attached men. Those
with high levels of attachment anxiety may be more motivated to develop relationship
skills during early adult years, sacrificing their sense of self if needed. Though sacrificing
autonomy will ultimately impact intimacy, these effects may be more perceptible in older
adults. Follow up studies with middle-age heterosexual couples are recommended to
clarify the effects of attachment anxiety on psychological abuse across the lifespan.

This study provided empirical evidence of ideas that feminist practitioners have
suggested about domestic violence for many years. Anger management classes are

seldom recommended for male batterers by feminist counseling organizations because

70

affect dysregulation is not assessed as the problem: rather. distorted beliefs about gender
and relationships are challenged in psychoeducational group settings. The power and
control wheel is used to inform men that acting like the master of the house is a misuse of
male privilege (DAIP, date unknown). Perpetrators are taught altemative means of
getting what they want, such as negotiation and direct communication skills. Feminists
purport that blaming the victim for the abuse is defensive. Abusers often do whatever is
necessary to protect themselves from negative appraisal, even at the expense of important
others, in an effort to ensure that they will be able to get what they want in the future.
This is a narcissistic defense. Men feel entitled to women’s services; if they do not get
their own way, they resort to coercion or force (Walker, 1999). If the police are called,
abusers may convincingly manipulate the situation to their own advantage. In a sense,
such tactics are embedded in male culture as “normal” means of taking care of relational
business. This was supported by the finding that attachment security more than
attachment anxiety predicted emotional abuse by men. As feminists have suggested, men
may abuse women because they feel comfortable doing so. Boys are socialized to manage
relationships aggressively with groups of male peers and may be ill prepared for
conflicting moral perspectives that arise when relating to females who are socialized
differently (see Gilligan, 1982/1993).

There were limitations to this study. The sample was limited to undergraduate
students at a large, Midwestern university. It was not a random sample, and this limits the
generalizability of these results to other samples of college students. Replication with a

random sample is recommended. This was a normal (non-clinical) sample, and it is

71

recommended that similar studies be done with clinical populations. especially those with
relationship problems and histories of domestic violence.

The study focused on self-reports of psychological abusiveness and did not
examine victimization. This study did not solicit input from victims or observers about
perpetrators; therefore, psychological abuse was likely underreported (Smith. 1994).
Severity of the psychological abuse, usually assessed by its impact on the victim, was not
measured in this study. This study did not examine the situations in which events occur,
largely because this information is best gathered through interview methods. Physical
abuse was not analyzed. The findings are limited to how people self-report perpetration
of psychological abuse. Because past studies found gender differences in self-reports of
psychological abuse (see Murphy & Cascardi, 1999), gender was controlled to reduce
potential bias.

The comparison across genders measured self-reported frequencies of behaviors
and did not address contextual differences related to these behaviors (see Murphy &
Cascardi, 1999). Whether women were using psychological tactics to defend themselves
against abusive partners remains unknown. Future studies should use interviews to
discover gender differences in construction of meaning and motivations behind
psychological abuse (see DeKeseredy, 1999). In addition, only biological differences in
gender were measured. Gender role adherence or perceptions of heterosexual
relationships as traditional versus egalitarian (e. g., Ellington & Marshall, 1997) were not
included in this analysis.

The ECR-R (Fraley et al., 2000) is a new version of the attachment measure, and

as such, it has not been utilized in much research. The Adult Attachment Interview

72

(George. Kaplan. & Main. 1996) is often preferred to self-report measures of attachment
to bypass defenses (for discussion, see George & West. 1999): however, self-report
methods are able to capture attachment-related perceptions relevant to romantic
relationships (Crowell, Fraley, & Shaver, 1999; Simpson & Rholes, 1998). This study
was concerned with how people view themselves, and measures of defenses were
included in the model. In addition, how people present themselves contributes to the
literature on assessment, prevention, and treatment. However, future studies of
psychological abuse in romantic relationships using the Adult Attachment Interview to
establish connections between insecurity and psychological abuse are recommended.

The MMEA (Murphy & Hoover, 1999; Murphy et al., 1999) is also a new
measure. Participants did not normally distribute in their responses, and it is unclear
whether this is attributable to the scale or the sample. Continued investigation into the
merits of using this scale to capture an important psychological abuse construct is
recommended.

Affective measures were not included in this model. Responses to powerful affect
were captured in defense variables, stress level was measured, and relevant affect such as
hostility overlaps with the dependent measures. Stress was included as a control variable
and secondarily examined as a potential mediator of observed effects. Why one person
becomes psychologically abusive and another does not, regardless of their level of stress,
was the focus of this study.

Only two defensive processes were studied. Splitting was not as relevant to
models testing avoidance as they have been when attachment anxiety was examined in

relationship to abuse. Psychological abuse research including other externalizing defenses

73

such as projection and projective identification is recommended. especially in the study

of male psychological abuse. Different processes may be more relevant to females.

74

CONCLUSIONS

Theory integration in research aids in the development of a more complete
understanding of complex psychological processes such as psychological abuse.
Although attachment theory is deemed an integrative theory (Lopez, 1995). especially
when explaining psychological problems found in couples (Lindegger & Barry, 1999),
this study demonstrates the utility of integrating attachment theory with feminist theory to
consider sociological constructs such as gender role socialization (also see Mauricio &
Gormley, 2001) and with psychodynamic constructs like defense mechanisms. Such
research can contribute to efforts to improve psychotherapy effectiveness through theory
integration (see Goldfried, Castonguay, & Safran, 1992). Efforts to understand
individuals simultaneously at sociological, relational, and intrapsychic/cognitive levels
are immediately applicable to clinical work. Theoretically integrative approaches may
widen clinical perspectives to encompass cultural differences in combination with
individual complexity, but more research in this area is needed.

Walker (1999) cautioned against requiring domestic violence researchers to be
constricted to one ideology. She also emphasized that stopping physical violence is not
sufficient when batterers simply learn to become better psychological batterers. She
identified an attitude of entitlement in males that has been immutable despite feminism’s
best educational efforts over several decades. This study utilized the concept of
narcissism to examine attitudes of entitlement, which were found more often in men than

women and predicted psychological abuse. The construction of a measure that captures

75

entitled attitudes specific to men‘s approach to relationships with women would help
further study of this difficult problem.

Empirical demonstrations of feminist concepts are needed to influence
mainstream clinical practice. To stop psychological abuse and intervene so that
perpetrators become motivated to change, linking problematic attitudes with known
clinical problems may provide solutions. For instance, both narcissism and attachment
avoidance are difficult problems to address, as is the entitled attitude associated with
male privilege, but effective interventions may be similar and informed by related
theories when integrated with feminist theory. Studies that test the utility of such
interventions in helping psychologically abusive men change are needed. For example,
clinical effectiveness studies that compare anger management interventions, feminist
psychoeducational approaches, psychodynamic interventions found effective with
narcissism, and interventions that enhance negative senses of self and others in order to
improve “felt security” (Sroufe & Waters, 1977) would be helpful.

This study begins to demonstrate the influence of gender role socialization and
insecure attachment style on perpetration of psychological abuse in college students.
Emotional abuse was prevalent in the current sample, and dominance was prevalent in
men. Most students reported some emotional abuse of others, and most men reported
some dominance. Although a comprehensive sense of the causes of mental and emotional
abuse in romantic dating relationships remains elusive, characteristics of perpetrators
were discovered that will aid efforts at prevention, assessment, and intervention.
Prevention measures that assess insecurity in adult attachments, perspectives on gender in

regard to relationships, pathological narcissistic defensiveness, and use of

76

psychologically abusive strategies are recommended. In addition. group intervention
protocols used with male batterers can be expanded to address avoidance of intimacy in
adult attachments and narcissistic entitlement and exploitativeness. with negative
consequences to such approaches clarified and altematives recommended. In addition.
individual interventions may be more fruitful than groups for those with such problems.
Whether the results of this study are useful in prevention or intervention efforts should be
empirically tested. In addition, only two types of abuse were examined, and measures of
other types of psychological abuse strategies are needed.

Not only does psychological abuse cause harm in its own right, but it can lead to
physical violence. The damage done to victims is most likely attributable to perpetrator
misperceptions. Why the research literature under-interprets these misperceptions and
blames victims is perhaps best explained by sociological influences. If these distortions
are part of male gender role socialization, and therefore considered superior ideas in a
patriarchal society and heavily reinforced, most people may fail to consider these ideas
distortions or misperceptions. For example, getting what you want, even at another
person’s expense, may be an acceptable or even valued approach in our society (also see
Lasch, 1979). Rather than intervening with women to improve their adaptation to cultural
narcissism, relational theories emphasized in this study suggest alternative changes that
men can make to help preserve and build healthy family relationships. Focus groups of a
normal population of men could be interviewed to uncover non-relational views that are
part of male culture, to discover motivations in men to preserve romantic relationships,
and to find out how men feel about their gender role socialization as it impacts close

relationships.

77

Risk assessments by clinicians. especially in inpatient psychiatric settings. should
consider domestic violence. Suicide risk assessments are typically more comprehensive
than assault risk assessments. but few suicide assessments question female patients about
whether they are victims of domestic violence or psychological abuse despite statistical
indications of such health risks. This study contributes to assault risk assessment by
providing information about factors and combinations of factors that increase the risk of
perpetrating psychological and physical abuse. Whether improved risk assessment
decreases assaults should be examined.

To test differential effects of various insecure attachment orientations and gender
constructs on aspects of domestic violence, separate models for each attachment style,
each gender, and each abuse strategy may be required. Patterns in how attachment
avoidance and anxiety lead to abuse are quite different, and models that continue to
elucidate this difference are essential. Too much emphasis has been placed on separation
fears as the catalyst to abuse, and not enough attention has been given to fear of intimacy.
Attachment avoidance and defensiveness are important variables to include in predictive
models, as is stress level when examining attachment anxiety. Whether avoidance
contributes to physical abuse in romantic relationships with moderators in predictive
models remains unknown. New measures of defensiveness are needed, especially to
discriminate externalizing defenses and acting out from other types of primitive defenses.

To detect gender differences in psychological abuse, separate models may need to
be tested for men and women, and the inclusion of variables that capture potentially
important motivational differences is recommended. In addition to measuring biological

gender, adherence to gender role is relevant and should be included in future studies of

78

domestic violence. Acknowledging that the male gender is a risk factor for assault is
useful to clinicians, but assisting men with altering attitudes that lead to abuse requires
that we understand the effects of gender role socialization. Examining numerous
psychological abuse strategies is essential and clearly calls for separate predictive
models; thus, larger’samples may be necessary to clarify separate models for men and
women. Further development of psychometrically sound psychological abuse scales that
measure overt and subtle power and control strategies used by both men and women is
needed. In addition, multiple measures of physical abuse that are sensitive to gender
differences should be developed and included in future studies so that interrelationships
between physical and psychological abuse strategies for each gender can be delineated.
Frequency and severity of different types of abuse and their impact on victims should be
assessed separately. Self-reports and observer/partner reports are helpful to improve the
accuracy of measures of abuse.

Advanced multivariate techniques may be needed to compare and contrast the
many variables involved in predicting abusiveness. When multiple sound scales of each
construct are available (including gender schemata, insecure intimate attachment styles,
defensive management of affect and cognition, and psychological and physical abuse)
structural equation modeling can be used to minimize measurement error and map out
interrelationships more carefully. To simultaneously examine self- and partner-reports of
abuse in couples, statistical procedures that do not assume independence, such as
hierarachical linear modeling (Bryk & Raudenbush, 1992), will be of use.

Follow up research should examine narcissism as a defensive moderator of the

relationship between attachment avoidance and other behavioral outcomes, especially

79

physical abuse in romantic relationships. Similar models testing gender. attachment. and
defense mechanisms might be useful in examining child abuse and neglect by parents.
particularly verbal abuse. Extending these findings to see if similar variables predict date
rape and marital rape would be beneficial. Replicating this study with a random or cross-
regional sample of college students as well as older populations and clinical populations
would increase the generalizability of findings related to psychological abuse.
Interventions that increase the “felt security” (Sroufe & Waters, 1977) of battered
women should be examined empirically to determine if attachment theory can inform
efforts to help these women repair damage done to senses of self by destructive adult
relationships. Studies that attempt to assist battered women in their efforts to create safe
lives should consider neuropsychological damage from traumatic brain injuries suffered
at the hands of intimate partners as potential barriers. Standard neuropsychological
screening is recommended for any victim of domestic violence who has suffered
numerous blows to the head, with or without losses of consciousness. Further study of the

neuropsychological consequences of domestic battery is suggested.

80

REFERENCES

Aiken. L. S.. & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions. Newbury Park, CA: Sage.

Ainsworth. M. D. S. (1989). Attachments beyond infancy. American Psychologist. 44.
709-716.

Albee, G. W. (2000). Commentary of prevention and Counseling Psychology. The
Counseling Psychologist, 28, 845-853.

Aldarondo, E., & Straus, M. A. (1994). Screening for physical violence in couple
therapy: Methodological, practical, and ethical considerations. Family Process,
33, 425-439.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental
disorders (4th ed.). Washington, DC: Author.

Apter, A., Plutchik, R., Sevy, S., Korn, M., Brown, S., & van Praag, H. (1989). Defense
mechanisms in risk of suicide and risk of violence. American Journal of
Psychiatry; 146, 1027-1031 .

Arrnbrust, C. A. (1997, March). Measurement of the defensive use of splitting and related
object relations deficits in a clinical population. Dissertation Abstracts
International, 5 7(9-B), 5905. (University Microfilms International No. 1997-
95006-269)

Aspy, C. B., & Sandhu, D. S. (1999). Empowering womenfor equity: A counseling

approach. Alexandria, VA: American Counseling Association.

81

Babcock, J. C.. Jacobson. N. 5.. Gottman. J. M., & Yerington. T. P. (2000). Attachment.
emotional regulation, and the function of marital violence: Differences between
secure, preoccupied, and dismissing violent and nonviolent husbands. Journal of
Family Violence, 15. 391-409.

Bachman, R., & Saltzman, L. E. (1995). National crime survey: Violence against women:
Estimates from the redesigned survey. In U. S. Department of Justice, Office of
Justice Programs, Bureau of Justice Statistics: Special Report (NCJ-154348, On-
line). Retrieved March 9, 2001: http://www.inform.umd.edu/EdRes/Topic/
WomensStudies/Genderlssues/Violence+Womenlnational-crime-victimization-
survey

Back, S. M., Post, R. D., & D‘Arcy, G. (1982). A study of battered women in a
psychiatric setting. Women and Therapy, 1, 13-26.

Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall.

Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in
social psychological research: Conceptual, strategic, and statistical considerations.
Journal of Personality and Social Psychology, 5] , 1 173-1 182.

Bartholomew, K. (1990). Avoidance of intimacy: An attachment perspective. Journal of
Social and Personal Relationships, 7, 147—178.

Bartholomew, K., & Horowitz, L. (1991). Attachment styles among young adults: A test
of a four-cate gory model. Journal of Personality and Social Psychology. 6] , 226-
244.

Bograd, M. (1984). Family systems approaches to wife battering: A feminist critique.

American Journal of Orthopsychiatijv, 54, 558-568.

82

Bograd. M. (1992). Values in conflict: Challenges to famin therapists‘ thinking. Journal
of Marital and Family Therapy, 20, 69-75.

Bookwala, J., & Zdaniuk, B. (1998). Adult attachment styles and aggressive behavior
within dating relationships. Journal of Social and Personal Relationships, 15.
175-190.

Bowlby, J. (1969/ 1982). Attachment and loss. Vol. 1: Attachment (2nd ed.). New York:
Basic Books.

Bowlby, J. (1973). Attachment and loss. Vol. II: Separation: Anxiety and anger. New
York: Basic Books.

Bowlby, J. (1979). The making and breaking of affectional bonds. London: Routledge.

Bowlby, J. (1980). Attachment and loss. Vol. III: Loss: Sadness and depression. New
York: Basic Books.

Bowlby, J. (1988). A secure base: Parent-child attachments and healthy human
development. New York: Basic Books.

Brennan, K. A., & Bosson, J. K. (1998). Attachment style differences in attitudes toward
and reactions to feedback from romantic partners: An exploration of the relational
bases of self-esteem. Personality and Social Psychology Bulletin, 24, 699-714.

Brennan, K. A., Clark, C. L., & Shaver, P. R. (1998). Self-report measurement of adult
attachment: An integrative overview. In J. A. Simpson & W. S. Rholes (Eds),
Attachment theory and close relationships. New York: Guilford Press.

Brennan, K. A., & Morris, K. A. (1997). Attachment styles, self-esteem, and patterns of
seeking feedback from romantic partners. Personality and Social Psychology

Bulletin, 23, 23-31.

83

Brennan. K. A.. 8: Shaver. P. R. (1995). Dimensions of adult attachment. affect
regulation, and romantic relationship functioning. Personality and Social
Psychology Bulletin. 2], 267-283.

Brennan. K. A.. & Shaver. P. R. (1998). Attachment styles and personality disorders:
Their connections to each other and to parental divorce, parental death. and
perceptions of parental caregiving. Journal of Personality, 66, 835-878.

Bretherton, l. (1995). The. origins of attachment theory: John Bowlby and Mary
Ainsworth. In S. Goldberg, R. Muir, & J. Kerr (Eds.), Attachment theory: Social.
developmental, and clinical perspectives (pp. 45—84). Hillsdale, NJ: Analytic
Press.

Browne, A. (1993). Violence against women by male partners: Prevalence, outcomes,
and policy implications. American Psychologist, 48, 1077-1087.

Bryk, A. S., & Raudenbush, S. W. (1992). Hierarchical linear models: Applications and
data analysis methods. Newbury Park, CA: Sage.

Burgess, R. L.. & Draper, P. (1989). The explanation of family violence: The role of
biological, behavioral, and cultural selection. In L. Ohlin & M. Tonry (Eds.),
Family violence (pp. 59-1 16). The University of Chicago Press.

Carroll, 1... Hoenigmann-Stovall, N ., Whitehead, G. 1., III. (1996a). Interpersonal
consequences of narcissism. Psychological Reports, 79, 1267-1272.

Carroll, L., Hoenigmann-Stovall, N., Whitehead. G. I., III. (1996b). The interpersonal
impact of narcissism: A comparative study of entitlement and self-absorption

factors. Journal of Social Behavior and Personality. 11, 601-613.

84

Cascardi. M.. 8; Vivian. D. (1995). Context for specific episodes of marital VlOlCflCCZ
Gender and severity of violence differences. Journal of Family Violence. 10. 265-
293.

Cash. T. F., Begley, P. J., McCown, D. A., & Weise, B. C. (1975). When counselors are
heard but not seen: Initial impact of physical attractiveness. Journal of Counseling
Psychology. 22, 273-279.

Cohen, J. (1977). Statistical power analysis for the behavioral sciences. New York:
Academic Press.

Collins, N. L. (1996). Working models of attachment: Implications for explanation,
emotion, and behavior. Journal of Personality and Social Psychology, 7], 810-
832.

Creasey, G., & Hesson-Mclnnis, M. (2001). Affective responses, cognitive appraisals,
and conflict tactics in late adolescent romantic relationships: Associations with
attachment orientations. Journal of Counseling Psychology, 48, 85-96.

Crittenden, P. M. (1995). Attachment and psychopathology. In S. Goldberg, R. Muir, &
J. Kerr (Eds.), Attachment theory: Social, developmental, and clinical
perspectives (pp. 367-406). Hillsdale, NJ: Analytic Press.

Crowell, J. A., Fraley, R. C., & Shaver, P. R. (1999). Measurement of individual
differences in adolescent and adult attachment. In J. Cassidy & P. R. Shaver
(Eds.), Handbook of attachment: Theory, research. and clinical applications (pp.

434-465). New York: Guilford Press.

85

Davies. J., Lyon, E.. & Monti-Catania. D. (1998). Safety planning with battered women:
Complex lives/difficult choices: Vol. 7. Sage series on violence against women.
Thousand Oaks, CA: Sage.

DeKeseredy, W. S. (1999). Tactics of the antifeminist backlash against Canadian
National Woman Abuse Surveys. Violence Against Women, 5, 1258-1276.

Dobash. R. E., & Dobash, R. P. (1984). The nature and antecedents of violent events.
British Journal of Criminology, 24, 269-288.

Domestic Abuse Intervention Project. (Date unknown). The Power and Control Wheel.
(Available from DAIP, 206 West Fourth Street, Duluth, MN 55806)

Duggan, E. S., & Brennan, K. A. (1994). Social avoidance and its relation to
Bartholomew’s adult attachment typology. Journal of Social and Personal
Relationships, 1 1, 147-153.

Draper, N. R., & Smith, H. (1981). Applied regression analysis. New York: Wiley.

Dutton, D. G. (1996). Patriarchy and wife assault: The ecological fallacy. In L. K.
Hamberger & C. Renzetti (Eds.), Domestic partner abuse (pp. 125-151). New
York: Springer.

Dutton, D. G. (1998). The abusive personality: Violence and control in intimate
relationships. New York: Guilford Press.

Dutton, D. G., Saunders, K., Starzomski, A., & Bartholomew, K. (1994). Intimacy-anger
and insecure attachments as precursors of abuse in intimate relationships. Journal
of Applied Social Psychology, 24, 1367-1386.

Dutton, D. G., & Starzomski, A. J. (1997). Personality predictors of the Minnesota Power

and Control Wheel. Journal of Interpersonal Violence, 12, 70-82.

86

Dutton, D. G., Starzomski, A.. 8: Ryan. L. (1996). Antecedents of abusive personality
and abusive behavior in wife assaulters. Journal ofFamily Violence, II. 1 13-132.

Edleson, J. L. (1996). Controversy and change in batterers' programs. In J. L. Edleson 8:
2. C. Eisikovits (Eds.), Future interventions with battered women and their
families: Vol. 3. Sage series on violence against women (pp. 154-169). Thousand
Oaks, CA: Sage.

Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review of
Psychology, 5], 665-697.

Ellington, J. E., & Marshall, L. L. (1997). Gender role perceptions of women in abusive
relationships. Sex Roles. 36, 349-369.

Emmons, R. A. (1984). Factor analysis and construct validity of the Narcissistic
Personality Inventory. Journal of Personality Assessment, 48. 291-300.

Emmons, R. A. (1987). Narcissism: Theory and measurement. Journal of Personality and
Social Psychology, 52, 1 1-17.

Enns, C. Z. (1993). Twenty years of feminist counseling and therapy: From naming
biases to implementing multifaceted practice. The Counseling Psychologist, 21, 3-
87.

Evans, P. (1992). The verbally abusive relationship: How to recognize it and how to
respond. Holbrook, MA: Bob Adams.

Federal Bureau of Investigation. (1998). Incidents of family violence: An analysis of
1998 NIBRS data. Uniform Crime Reports: Crime in the United States 1998 [On-
line]. Retrieved November 25, 2000:

http://www.fbi.gov/ucr/Cius_98/98crime/98cius29.pdf

87

Feeney, J. A. (1999). Adult romantic attachment and couple relationships. In J. Cassidy
8: P. R. Shaver (Eds). Handbook of attachment: Theory, research, and clinical
applications (pp. 355-377). New York: Guilford Press.

Fischer, R. (1989). Psychotherapy of the Narcissistic Personality Disorder. In J. F.
Masterson & R. Klein (Eds.), Psychotherapy of the disorders of the self: The
Masterson approach (pp. 69-89). New York: Brunner/Mazel.

Fishtein, J., Pietromonaco, P. R., & Barrett, L. F. (1999). The contribution of attachment
style and relationship conflict to the complexity of relationship knowledge. Social
Cognition, I7, 228-244.

Fraley, R. C., & Waller, N. G. (1998). Adult attachment patterns: A test of the
typological model. In J. A. Simpson & W. S. Rholes (Eds.), Attachment theory
and close relationships (pp. 77-114). New York: Guilford Press.

Fraley, R. C., Waller, N. G., & Brennan, K. A. (2000). An item response theory analysis
of self-report measures of adult attachment. Journal of Personality and Social
Psychology, 78, 350-365.

Fuendeling, J. M. (1998). Affect regulation as a stylistic process within adult attachment.
Journal of Social and Personal Relationships, 15, 291 -322.

Ganley, A. (1989). Integrating feminist and social learning analyses of aggression:
Creating multiple models for intervention with men who batter. In P. L. Caesar &
L. K. Hamberger (Eds.), Treating men who batter: Theory, practice, and
programs. Springer series, focus on men, volume 5 (pp. 196-235). New York:

Springer.

88

George. C., Kaplan. N.. 8; Main. M. (1996'). Adult Attachment Inten'icu'. L'npublished
manuscript. Department of Psychology. University of California, Berkeley (3rd
ed.).

George, C., & West, M. (1999). Developmental vs. social personality models of adult
attachment and mental ill health. British Journal of Medical Psychology. 72, 285-
303.

Gilligan, C. (1982/1993). In a dififerent voice: Psychological theory and women '5
development. Cambridge, MA: Harvard University Press.

Goldfried, M. R., Castonguay, L. G., & Safran, J. D. (1992). Core issues and future
directions in psychotherapy integration. In J. C. Norcross & M. R. Goldfried
(Eds.), Handbook of psychotherapy integration (pp. 593-616). New York: Basic
Books.

Gondolf, E. W. (1999). Characteristics of court-mandated batterers in four cities.
Violence Against Women, 5, 1277-1293.

Gould. J. R., Prentice, N. M., & Ainslie, R. C. (1996). The Splitting Index: Construction
of a scale measuring the defense mechanism of splitting. Journal of Personality
Assessment, 66, 414-430.

Griffin, D., & Bartholomew, K. (1994). Models of self and other: Fundamental
dimensions underlying measures of adult attachment. Journal of Personality and
Social Psychology, 67, 430-445.

Grotstein, J. S. (1985). Splitting and projective identification. Northvale, NJ: Jason

Aronson.

89

Hage. S. M. (2000). The role of counseling psychology in preventing male violence
against female intimates. The Counseling Psychologist, 28. 797-828.

Hamby, S. L. (1996). The dominance scale: Preliminary psychometric properties.
Violence and Victims, I], 199-212.

Hazan, C., & Shaver, P. (1987). Romantic love conceptualized as. an attachment process.
Journal of Personality and Social Psychology, 52, 51 1-524.

Heise, L. L. (1998). Violence against women: An integrated, ecological approach.
Violence Against Women, 4, 262-290.

Holmbeck, G. N. (1997). Toward terminological, conceptual, and statistical clarity in the
study of mediators and moderators: Examples from the child-clinical and pediatric
psychology literatures. Journal of Consulting and Clinical Psychology, 65, 599-
610.

Hoover, S. A., Murphy, C. M., & Taft, C. T. (1999, July). Gender diflerences in
reporting, perpetration, and correlates of emotional abuse. Paper presented at the
6th lntemational Family Violence Research Conference, Durham, NH.

Horowitz, M. J. (1977). Cognitive and interactive aspects of splitting. American Journal
of Psychiatry, 134, 549-553.

Johnson, H., & Sacco, V. F. (1995). Researching violence against women: Statistics
Canada’s national survey. Canadian Journal of C riminology, 37, 281-304.

Johnson, M. E., & Holland, A. L. (1986). Measuring clients” expectations: The 15
personal problems inventory. Measurement and Evaluation in Counseling and

Development, I9, 151-156.

90

Jordan. J. V. (Ed). ( 1997). ll’omen 's grott'th in diversity: More writingsfrom the Stone
Center. New York: Guilford Press.

Karen, R. ( 1994). Becoming attached: First relationships and how they shape our
capacity to love. New York: Oxford University Press.

Kelley, F. A., & Gelso, C. J. (2000, August). Relationship of secure and insecure
attachment to healthy and defensive narcissism: Implications for psychotherapy.
In C. J. Gelso (Chair), Exploration of attachment phenomena in the
psychotherapy relationship. Symposium conducted at the annual convention of
the American Psychological Association, Washington, DC.

Kelly, L. (1996). Tensions and possibilities: Enhancing informal responses to domestic
violence. In J. L. Edleson & Z. C. Eisikovits (Eds.), Future interventions with
battered WOIHC’II and their families: Vol. 3. Sage series on violence against women
(pp. 67-86). Thousand Oaks, CA: Sage.

Kemberg. 0. (1975/1985). Borderline conditions and pathological narcissism.
Northvale, NJ: Jason Aronson.

Kesner, J. E., Julian, T., & McKenny, P. C. (1997). Application of attachment theory to
male violence toward female intimates. Journal of Family Violence, 12, 21 1-228.

Kirkpatrick, L. A., & Davis, K. E. (1994). Attachment style, gender, and relationship
stability: A longitudinal analysis. Journal of Personality and Social Psychology,
66, 502-512.

Klein, R. (1989). Introduction to the disorders of the self. In J. F. Masterson & R. Klein
(Eds.), Psychotherapy of the disorders of the self: The Masterson approach (pp.

30—46). New York: Brunner/Mazel.

91

Kunce, L. J., 8; Shaver, P. R. (1994). An attachment-theoretical approach to caregiving in
romantic relationships. In K. Bartholomew & D. Perlman (Eds.), Attachment
processes in adulthood: Volume 5: Advances in Personal Relationships (pp. 205-
237). London: Jessica Kingsley.

Lasch, C. (1979). The culture of narcissism: American life in an age of diminishing
expectations. New York: Norton.

Lavoie, F., Robitaille, L., & Hebert, M. (2000). Teen dating relationships and aggression:
An exploratory study. Violence Against Women, 6, 6-36.

Lenton, R. L. (1995). Power versus feminist theories of wife abuse. Canadian Journal of
Criminology, 37, 305-330.

Lewis-Beck, M. S. (1980). Applied regression: An introduction. (Sage University Paper
Series on Quantitative Applications in the Social Sciences, series no. 07-022).
Newbury Park, CA: Sage.

Lindegger, G., & Barry, T. (1999). Attachment as an integrating concept in couple and
family therapy: Some considerations with special reference to South Africa.

Contemporary Family Therapy, 21, 267-288.

Lopez, F. G. (1995). Contemporary attachment theory: An introduction with implications
for counseling psychology. The Counseling Psychologist, 23, 395-415.

Lopez, F. G. (2001). Adult attachment orientations, self-other boundary regulation, and
splitting tendencies in a college sample. Journal of Counseling Psychology, 48,

440-446.

92

Lopez. F. G.. 8: Brennan. K. A. (2000). Dynamic processes underlying adult attachment
organization: Toward an attachment theoretical perspective on the healthy and
effective self. Journal of Counseling Psychology. 4 7, 283-300.

Lopez. F. G., Fuendeling. J .. Thomas, K., & Sagula. D. (1997). An attachment-theoretical
perspective on the use of splitting defences. Counselling Psychology Quarterly.
10. 461 -472.

Lopez, F. G., & Gormley, B. (in press). Stability and change in adult attachment style.
Journal of Counseling Psychology.

Lopez, F. G., Melendez, M. C., Sauer, E. M., Berger, E., & Wyssmarm, J. (1998).
Internal working models, self-reported problems, and help-seeking attitudes
among college students. Journal of Counseling Psychology, 45, 79-83.

Lyons-Ruth, K., & Jacobvitz, D. (1999). Attachment disorganization: Unresolved loss.
relational violence, and lapses in behavioral and attentional strategies. In J.
Cassidy & P. R. Shaver (Eds.), Handbook of attachment: Theory, research, and
clinical applications (pp. 520-554). New York: Guilford Press.

Marshall, L. L. (1996). Psychological abuse of women: Six distinct clusters. Journal of
Family Violence, 11, 379-409.

Marshall, L. L. (1999). Effects of subtle and overt psychological abuse on low-income
women. Violence and Victims, 14. 69-88.

Masterson, J. F. (1988). The search for the real self: Unmasking the personality disorders

of our age. New York: The Free Press.

93

Masterson. J. F. (1993). The emerging self: A (let'elopmental. self; and object relations
approach to the treatment of the closet narcissistic disorder of the self: New
York: Brunner/Mazel.

Mauricio. A. M., & Gormley, B. (2001). Male perpetration of physical violence against
female partners: The interaction of dominance needs and attachment insecurity.
Journal of Interpersonal Violence. I 6, 1066-1081.

Mayseless, O. (1991). Adult attachment patterns and courtship violence. Family
Relations, 40, 21-28.

Mikulincer, M., & Horesh, N. (1999). Adult attachment style and the perception of
others: The role of projective mechanisms. Journal of Personality and Social
Psychology, 76, 1022- l 034.

Mikulincer, M., & Orbach, I. (1995). Attachment styles and repressive defensiveness:
The accessibility and architecture of affective memories. Journal of Personality
and Social Psychology, 68, 917-925.

Miller, J. B. (1976/1986). Toward a new psychology of women (2nd ed.). Boston: Beacon
Press.

Miller, J. B. (1991). Women and power. In J. V. Jordan, A. G. Kaplan, J. B. Miller, I. P.
Stiver, & J. L. Surrey (Eds.), Women '5 growth in connection: Writings from the
Stone Center (pp. 197-205). New York: Guilford.

Morrison, T. L., Goodlin-Jones, B. L., & Urquiza, A. J. (1997). Attachment and the
representation of intimate relationships in adulthood. The Journal of Psychology,

131,57-71.

94

Murphy. C. M.. 8: Cascardi. M. (1999). Psychological abuse in marriage and dating
relationships. In R. L. Hampton (Ed.), Family violence: Prevention and treatment
(2nd ed.). Thousand Oaks, CA: Sage.

Murphy, C. M., 8: Hoover, S. A. (1999). Measuring emotional abuse in dating
relationships as a multifactorial construct. Violence and Victims, I4, 39-53.

Murphy, C. M., Hoover, S. A., 8t Taft, C. T. (1999, November). The multidimensional
measure of emotional abuse: Factor structure and subscale validity. Paper
presented at the meeting of the Association for the Advancement of Behavior
Therapy, Toronto, Canada.

Neufeld, J., McNamara, J. R., & Ertl, M. (1999). Incidence and prevalence of dating
partner abuse and its relationship to dating practices. Journal of Interpersonal
Violence, I4, 125-137.

Nosko, A., & Wallace, B. (1988). Group work with abusive men: A multidimensional
model. Social Work with Groups, II, 33-52.

O’Heam, R. E., & Davis, K. E. (1997). Women’s experience of giving and receiving
emotional abuse: An attachment perspective. Journal of Interpersonal Violence,
12, 375-391.

O'Leary, K. D. (1999). Psychological abuse: A variable deserving critical attention in
domestic violence. Violence and Victims, 14, 3-23.

Pietromonaco, P. R., & Barrett, L. F. (2000). The internal working models concept: What
do we really know about the self in relation to others? Review of General

Psychology, 4, 155-1 75.

95

Ponce. F. Q.. 8: Atkinson. D. R. (1989). Mexican-American acculturation. counselor
ethnicity. counseling style. and perceived counselor credibility. Journal of
Counseling Psychology, 36, 203—208.

Puzone, C. A., Saltzman, L. E., Kresnow, M., Thompson. M. P.._ & Mercy. J. A. (2000).
National trends in intimate partner homicide: United States, 1976—1995. Violence
Against Women, 6, 409-426.

Rapoza, K. A., & Malley-Morrison, K. (1997, August). An examination of attachment
style, personality triggers, and physical and psychological dating abuse. Poster
session presented at the annual convention of the American Psychological
Association, Chicago, IL.

Raskin, R., & Hall, C. S. (1979). A narcissistic personality inventory. Psychological
Reports, 45, 590.

Raskin, R., & Hall, C. S. (1981). The Narcissistic Personality Inventory: Alternate form
reliability and further evidence of construct validity. Journal of Personality
Assessment, 45, 159-162.

Raskin, R., & Novacek, J. (1989). An MMPI description of the narcissistic personality,
Journal of Personality Assessment, 53, 66-80.

Raskin, R., & Novacek, J. (1991). Narcissism and the use of fantasy. Journal of Clinical
Psychology, 47, 490-499.

Raskin, R., Novacek, J., & Hogan, R. (1991). Narcissism, self-esteem, and defensive self-

enhancement. Journal of Personality, 59, 19-38.

96

Raskin. R.. 8: Terry. H. (1988). A principal components analysis of the Narcissistic
Personality Inventory and further evidence of its construct validity. Journal of
Personality and Social Psychology. 54. 890-902.

Rencher, A. C. (1995). Methods of multivariate analysis. In W. A. Shewart 8: S. S. Wilks
(Establishers), Wiley series in probability and mathematical statistics. New York:
John Wiley & Sons.

Renzetti, C. M. (1996). On dancing with a bear: Reflections on some of the current
debates among domestic violence theorists. In L. K. Hamberger & C. Renzetti (
Eds.), Domestic partner abuse (pp. 213-222). New York: Springer.

Rhodewalt, F., & Morf, C. C. (1995). Self and interpersonal correlates of the Narcissistic
Personality Inventory: A review and new findings. Journal of Research in
Personality, 29, 1-23.

Roberts, N., & Noller, P. (1998). The associations between adult attachment and couple
violence: The role of communication patterns and relationship satisfaction. In J.
A. Simpson & W. S. Rholes (Eds.), Attachment theory and close relationships
(pp. 317-350). New York: Guilford Press.

Roth, 8., Lebowitz, L., & DeRosa, R. R. (1997). Thematic assessment of posttraumatic
stress reactions. In J. P. Wilson & T. M. Keane (Eds.), Assessing psychological
trauma and PTSD (pp. 512-528). New York: Guilford Press.

Rothschild, B., Dimson, C., Storaasli, R., & Clapp, L. (1997). Personality profiles of
veterans entering treatment for domestic violence. Journal of Family Violence,

12, 259-274.

97

Sable, P. (1998). Almost all in the family: Emotionally abusive attachments. Journal of
Emotional Abuse, I , 51-67.

Sakett, L. A., & Saunders, D. G. (1999). The impact ofdifferent forms of psychological
abuse on battered women. Violence and Victims, I 4. 105-1 17.

Scalia, J ., III. (1994). Psychoanalytic insights and the prevention of pseudosuccess in the
cognitive-behavioral treatment of batterers. Journal of Interpersonal Violence, 9,
548-555.

Schechter, S. (1987). Guidelines for mental health practitioners in domestic violence
cases. Washington, DC: National Coalition AgainSI Domestic Violence.

Schirmer, L. L., & Lopez, F. G. (2001). Probing the social support and work strain
relationship among adult workers: Contributions of adult attachment orientations.
Journal of Vocational Behavior, 59, 17-33.

Schlenger, W. E., Fairbank, J. A., Jordan, B. K., & Caddell, J. M. (1997).
Epidemiological methods for assessing trauma and posttraumatic stress disorder.
In J. P. Wilson & T. M. Keane, Assessing psychological trauma and PTSD (pp.
139-159). New York: Guilford Press.

Searles, B., & Meara, N. M. (1999). Affective dimensions of attachment styles:
Exploring self-reported attachment style, gender, and emotional experience
among college students. Journal of Counseling Psychology, 46, 147-158.

Shavelson, R. J. (1996). Statistical reasoning for the behavioral sciences (3rd ed.).
Boston: Allyn and Bacon.

Sheinberg, M., & Penn, P. (1991). Gender dilemmas, gender questions, and the gender

mantra. Journal of Marital and Family Therapy, I 7, 33-44.

98

Simpson, J. A.. 8: Rholes, W. S. (1994). Stress and secure base relationships in
adulthood. In K. Bartholomew & D. Perlman (Eds.), Attachment processes in
adulthood: Vol. 5. Advances in personal relationships. London: Jessica Kingsley.

Simpson, J. A., & Rholes, W. S. (1998). Attachment in adulthood. In J. A. Simpson 8: W.
S. Rholes (Eds.), Attachment theory and close relationships (pp. 3-21 ). New
York: Guilford Press.

Simpson, J. A., Rholes, W. S., & Phillips, D. (1996). Conflict in close relationships: An
attachment perspective. Journal of Personality and Social Psychology, 71, 899-
914.

Smith, M. D. (1994). Enhancing the quality of survey data on violence against women: A
feminist approach. Gender and Society, 8, 109-127.

Sroufe, L. A., & Waters, E. (1977). Attachment as an organizational construct. Child
Development, 49, 1 184-1199.

Stosny, S. (1995). Treating attachment abuse: A compassionate approach. New York:
Springer.

Straus, M. A. (1973). A General Systems Theory approach to a theory of violence
between family members. Social Science Information, 12, 105-125.

Straus, M. A. (1979). Measuring intrafamily conflict and violence: The Conflict Tactics
(CT) scales. Journal of Marriage and the Family, 41, 75-88.

Straus, M. A. (1990). The Conflict Tactics Scales and its critics: An evaluation and new
data on validity and reliability. In M. A. Straus & R. J. Gelles (Eds.), Physical
violence in American families: Risk factors and adaptations to violence in 8, I 45

families (pp. 49-73). New Brunswick, NJ: Transaction Publishers.

99

Sugarman. D. B.. 8; Hotaling. G. T. (1989). Dating violence: Prevalence. context and risk
markers In M. A. Pirog-Good 8; J. E. Stets (Eds). Violence in dating
relationships (pp. 3-32). New York: Praeger.

Thomas. D. Q., & Beasley. M. E. (1993). Domestic violence as a human rights issue.
Human Rights Quarterly. 15, 36-62.

Thompson, L. (1992). Feminist methodology for family studies. Journal of Marriage and
the Family, 54, 3-18.

Tjaden, P., & Thoennes, N. (1998). Prevalence, incidence, and consequences of violence
against women: Findings from the National Violence Against Women Survey.
(NCJ Publication No. 172837, U. S. Department of Justice; On-line). Retrieved
November 24, 2000: http://www.ncjrsorg/txtfiles/172837.txt

Tolman, R. M. (1989). The development of a measure of psychological maltreatment of
women by their male partners. Violence and Victims, 4, 173—189.

Tolman, R. M. (1999). The validation of the psychological maltreatment of women
inventory. Violence and Victims, I 4,_25-37.

Tweed, R. G., & Dutton, D. G. (1998). A comparison of impulsive and instrumental
subgroups of batterers. Violence and Victims, 13, 217-230.

United Nations Declaration on the Elimination of Violence Against Women, General
Assembly Resolution 48/ 104. (20 December 1993).

UNICEF. (2000). Domestic violence against women and girls. Innocenti Digest,_6, 1-23.

US. Department of Justice [USDOJ]. (1983). Report to the nation on crime and justice:

The data. Washington, DC: Author.

100

US. Department of Justice. (1993). Report to the nation on crime andjustice: The data.
Washington, DC: Author.

US. Department of Justice (1998). Title IX. Violence against women education.
prevention, and intervention research [On-line]. Retrieved February 12. 1998:
http://www.now.org/issues/violence/vawa/vawal998.html

US. Department of Justice (2000). Federal Violence Against Women laws and
regulations [On-line]. Retrieved November 25, 2000:
http://www.ojp.usdoj.gov/vawo/

Vaillant, G. E. (1992). Ego mechanisms of defense: A guide for clinicians and
researchers. Washington, DC: American Psychiatric Press.

Walker, L. E. (1999, August). Critical issues for the millennium: Domestic violence.
Paper presented at the annual convention of the American Psychological
Association, Boston, MA.

Walker, L. E. A. (1990). Feminist ethics with victims of violence. In H. Lerman & N.
Porter (Eds.), Feminist ethics in psychotherapy (pp. 214-226). New York:
Springer.

Wampold, B. E., & Freund, R. D. (1987). Use of multiple regression in Counseling
Psychology research: A flexible data-analytic strategy. Journal of Counseling
Psychology, 34, 372-382.

Watson, P. J., & Biderman, M. D. (1993). Narcissistic Personality Inventory factors,
splitting, and self-consciousness. Journal of Personality Assessment, 6], 41-57.

Weis, J. G. (1989). Family violence research methodology and design. In L. Ohlin & M.

Tonry (Eds.), Family violence (pp. 1 17-162). The University of Chicago Press.

101

Wilkinson. L.. 8‘; The APA Task Force on Statistical Inference. (1999). Statistical
methods in psychology journals: Guidelines and explanations. American

Psychologist. 54, 594-604.

102

450E335: :a 30:20E0 .282

 

 

mm. - a fig? 8.8 Q - _ 2.2 mm. a 8:2 .2255
2: - 2 8.2 $8 mm - S 5.x 3% 855::8
am - x New 2.: z. - m can 6.2 waaamézo
mm - N. c5. 3.: 3. - N. E 3.2 32333
S. - a. of. one. 8 - _ 42.. EN. :53622
SE - em... 8. _ 3. En - 8.. 5.. EN 85662
3% - 2: 2: o; Em - or... 3. EA 525%
2: - a CS. 3.8 cm - 8 £5 ~03. $25
09.5% Qm. E 095% Cu. 3

2c M 5 3:32

30 u 5 8:55”—

 

EEEES .CeEtl E 2823ng 33:66

_ 29;

103

..o. v & 453:6 bassocEwi £3883 258$? 35th bEmchmu 8: $9833. ocam a.

 

 

Ea ._ _ mm :1 .3. Eosgeééeéoém
ism. anew. _.m.m $02 xszEEO E23632
8.: £3. :2 f: co_a:_e:=:8§5:ao
a; .82 E “on: 2323:; 2:8:
2.: £3 :2 :3 5:328
a: an; :3 “was: 28:33: 26:78:
”umznxx ECOCCEM
s _ .m 5%.: Ea some. 2550:2535
m: 5.8. a; 6me eaazesam
me: ..ccem :.m 5.8 5:35.: ”BEES
E. 2 a: 2

So u 5 8.32

Go .I. 5 835m

 

9.333% 25.3.:th 35 34:; Naowmezcsai S $92.6.ng .Shtmb

m 635.

104

.5. V 53." 60. V as.“

3.274 35:95. .w

 

:cm. -- oocmEEoQ .5
*3. im. -- wEEEEEzO .0

:8. m... to.» -- msgaméom .m

invm. :3... o..- 3.3.. -- Emimmuaz .v
m... imm. swam. iwm. Ne: -- 85:23:: .m

inmm. imm. it“. :wm. m..- 3.3. .. beta/x .m

in? 3?: :3. 83%. we: ism. *smm. -- $.86 ._
m n c m w m m _

Q2 .I. >2 €33.23» 2:9:th YE ..Etacwi fizigmcsstc 25:33:39

m 255.

105

._o. vain .mo. V1”:

A E u 5 :2: :8 .9593: 2: 328 25:20:00 Ave N 5 5an :8 .mcowfln 2: 032: 20:36:00 «:32

 

 

-- :No. ..R. :8. an. a. :2. :8. 3.22 35:85 .w

:3: -- t. 8. :3. :fi. 3. :on 005558 s
S. :8. -- :3» 2.- E. :S. :S. aegis-3:0 .:
E. E. :3. -- m..- :3. :5. :8. Egan-:8 «-
E. 8. E.- :m.. -- :. 8.- 8.- :_,,._e_.§z .:
5. 0:. ...R. :2. *8- -- :2. :2. 85:62 .m
t. :2. :2. :ow. a..- :wm. -- :8. 8:2: .N
E.- 2. :wm. :8. m... S. :on - sea ._
x s c r. v m N _

aunts-6 >4 €3ch> mic-6:5 3:: ..E:.,..¢§:& ...aES:.&Q:SQ Re .,.~E.::E.>SU

v 035-

106

.598: .895.
22833 E52682 N Z £2833. 3.3m .mcocoEm u m £238.3- ouanEoD u D
.A E n 5 :2: .8 :25ch of 323 85:22.8 Ace n 5 5:53 .8 .2036 2: gong 8.23.250 8:2

 

-- :2. :3. :3. :3. ...2. :2. *2. :8. :2. 8. 22:88:85.9BEE-z.:
:2. -- :2. E. :2. ow. .m. cm. :3. 8. 8.- 572256.28.
8. 8. -- :8. :8. :8. :8. :8. :2. E. :8. .5825_E:85..:_:_oo-m.8
:8. 8. :8. -- :E. :2. *2. :8. :8. 8. :8. 3332:; 2:82-92
2.. E. a. :8. -- :2. :8. :8. :2. 2. 8. 58825-92
2. 8. :8. :8. :8. -- E. :8. :2. 2.. :8. 225.:ow 2:05.888
.82. 8.- 8. 8. 2. 8. -- cm. :8. :2. 8. Egawsamadw
2. 8.- :8. .82. *8. :8. :8. -- :8. 5.- .82. :2§_§:§_-Q.v
:2. 8. 8. E. :8. 8. :3: *8. -- :2. 8. £553.92
8. :8- 2..- E. 8. 8.- 8. 8.- :2. -- :8. 85§S<N
S. s..- 8. 2. 8. cm. *2. 2. E. :8. -- 525;:

 

__ o. o w h c m w m m _

.3sz0 .3 viz-,aézm 3.4.2.8332 NE: 8.5; 3.833248.an 5.2. 23.259226 :r:=<.5:< €5<

w czar—-

107

Table 6

Summary ainerarclzical Regression Analysisfor Variables Predicting Dominance
(N = 12 7)

 

Variable B SE B [3
Step 1
Stress .26 .06 .36***
Gender 3.67 1.78 .17*
Step 2
Stress .22 .07 .31 **
Gender 2.88 1.79 .13
Anxiety -.27 1.05 -.03
Avoidance 2.43 .97 23*
Step 3
Stress .22 .07 .30**
Gender 2.66 1.72 .12
Anxiety 1.01 1.24 .10
Avoidance -.40 1 .25 -.04
Gender*Anxiety -3.06 1.77 -.20
Gender*Avoidance 6.43 1 .88 .40**

 

Note. Table 6 continued on next page.

108

Table 6 (cont‘d)

 

Variable 8 SE B [3
Step 4

Stress .25 .07 .35“
Gender .15 1.76 .01
Anxiety 1.47 1.22 .14
Avoidance .26 1.22 .02
Gender*Anxiety -2.96 1.69 -.20
Gender*Avoidance 5.55 1.83 .35**
Narcissism .52 .15 .28**
Self-Splitting -.26 .20 -.14
Other-Splittin g .13 .16 .07

 

Note. R2 = .165 for Step 1; AR? = .042 for Step 2; AR2 = .071 for Step 3; AR‘? = .083 for
Step 4 (all [)5 < .001).

*p<.05.**p<.01. ***p<.001.

109

 

Table 7

Summary of Hierarchical Regression Analysisfln‘ \f'ariab/es Predicting
Emotional Abuse (N = 127)

 

Variable B SE B [3
Step 1
Stress .67 .13 .41 ***
Gender 4.39 3.90 .09
Step 2
Stress ‘ .59 .18 .37**
Gender .59 4.21 .01
Anxiety -.80 2.91 -.04
Avoidance -.28 2.90 -.01
Gender*Anxiety 3.96 4.03 .12
Gender*Avoidance .34 4.38 .01
Narcissism 1.17 .36 .28**
Self-Splitting .28 .47 .07
Other-Splitting .02 .37 .00

 

Note. Table 7 continued on next page.

110

 

Table 7 (cont'd)

 

Variable B SE B B
Step 3
Stress .51 .18 .31**
Gender -.29 4.26 -.01
Anxiety -1.26 2.94 -.06
Avoidance 1.61 3.06 .07
Gender*Anxiety 4.92 4.10 .15
Gender*Avoidance -3.52 4.91 -.10
Narcissism 1.26 .36 .31 **
Self-Splitting .41 .47 .10
Other-Splitting -.07 .38 -.02
Avoidance*Narcissism .74 .36 .19*
Avoidance*Self-Splitting -.31 .36 —.08
Avoidance*Other-Splitting .59 .38 .15

 

Note. R2 = .183 for Step 1; AR2 = .081 for Step 2; AR2 = .036 for Step 3 (all ps < .001).

*p<.05, **p<.01, ***p<.001.

lll

Table 8

Logarithmic Transformation ofEmotional Abuse for Regression (N = 12 7)

 

Variable B SE B [3
Step 1
Stress .01 .00 .25**
Gender .01 .08 .01
Step 2
Stress .00 .00 .13
Gender —.02 .09 -.03
Anxiety .02 .06 .06
Avoidance -.01 .06 -.03
Gender*Anxiety .03 .08 .05
Gender*Avoidance .03 .09 .05
Narcissism .02 .01 24*
Self-Splitting .02 .01 .19
Other-Splitting .00 .01 -.03

 

Note. Table 8 continued on next page.

112

Table 8 (cont'd)

 

Variable B SE B [3
Step 3
Stress .00 .00 .09
Gender -.05 .09 -.05
Anxiety .02 .06 .04
Avoidance .03 .06 .06
Gender*Anxiety .05 .08 .07
Gender*Avoidance -. 10 . 10 -.16
Narcissism .02 .01 .26**
Self-Splittin g .02 .01 .22
Other-Splitting .00 .01 -.04
Avoidance*Narcissism .02 .01 .22*
Avoidance*Self—Sp1itting .00 .01 —.03
Avoidance*Other-Splitting .01 .01 .12

 

Note. R2 = .063 for Step 1; AR? = .071 for Step 2; AR? = .038 for Step 3 (all ps < .05).

*p<.05,**p<.01.

113

Table 9
Separate Models for Dominance by Gender

Variable B SE B B

 

Women (n = 66)

Step 1
Stress -.02 .08 -.03
Anxiety 2.44 1.09 .33
Avoidance -.27 1.04 -.03

Step 2
Stress -.02 .09 -.05
Anxiety 1.73 1.09 .23
Avoidance 1.06 1.13 .13
Narcissism .21 .19 .14
Self-Splitting -.19 .21 -.13
Other-Splitting .36 .17 .28
Avoidance*Narcissism .34 .17 .27
Avoidance*Self—Splitting -.25 .20 -.20
Avoidance*Other-Splittin g -.08 .18 -.07

 

Note. Table 9 continued on next page. Women: R2 = .093 for Step 1 (ns); AR? = .197 for

Step 2 (p < .05).

114

 

Table 9 (cont'd)

 

Variable B SE B
Men (n = 61)
Step 1
Stress .43 .10 .55
Anxiety -4.00 .71 -.32
Avoidance 5.72 .53 .43
Step 2
Stress .55 .10 .70
Anxiety -2.95 .57 -.24
Avoidance 6.56 .57 .49
Narcissism .73 .20 .33
Self-Splitting -.60 .32 -.26
Other-Splitting -.40 .26 -.17
Avoidance*Narcissism -.24 .22 -.1 1
Avoidance*Self—Splitting .23 .21 .11
Avoidance*Other-Splitting .45 .26 .18

 

Note. Men: R2 = .408 for Step 1; AR2 = .215 for Step 2 (ps < .001).

115

Table 10

Test of Mediators in Anxiet\»°-Enzoti()tta1 Abuse Relations/zip (N =12 7)

 

Variable B SE B [3
Step 1
Gender 5.96 4.24 12
Anxiety 5.51 2.16 24*
Avoidance .71 2.32 .03
Step 2
Gender 2.41 4.34 .05
Anxiety 3.27 2.41 .14
Avoidance -.36 2.28 -.02
Narcissism 1.23 .37 .30**
Self-splitting .88 .46 .21
Other-splitting .33 .38 .08

 

Note. Table 10 continued on next page.

116

Table 10 (cont'd)

 

 

Variable B SE B [3
Step 3
Gender 1.37 4.32 .03
Anxiety 2.80 2.38 .12
Avoidance -.12 2.25 -.01
Narcissism 1.26 .36 .30**
Self-splitting .91 .45 22*
Other—splittin g .16 .38 .04
Avoidance*narcissism .80 .33 .21*
Avoidance*self-splitting -.37 .36 -. 10
Avoidance*other-splitting .82 .39 .21 *

 

Note. Table 10 continued on next page.

117

Table 10 (cont'd)

 

Variable B SE B B
Step 4
Gender -.24 4.22 -.01
Anxiety .78 2.40 .03
Avoidance .05 2.17 .00
Narcissism 1.22 .35 .30**
Self-splittin g .40 .47 . 10
Other-splittin g -.07 .38 -.02
Avoidance*narcissism .65 .32 .17*
Avoidance*self—splitting -.26 .35 -.07
Avoidance*other-splitting .61 .38 .15
Stress .53 .18 .33**

 

Note. R2 = .079 for Step 1 (p < .05); AR2 = .100 for Step 2 (p < .01); AR2 = .057 for Step

3 (p < .001); AR2 = .055 for Step 4 (p < .001).

*p<.05,**p<.01.

118

 

52:55.3: ES 852295 Eocfionzw 5223 35.22.22 2: 55327. 2:03 8:220 20 22,2 :9: 95:25.2
32::_:_:_.-ou:_6_c>s 2:25.235 2: 2225:. Ea 25:25.2 oucaEEosanEiSw 258583 2: 233:. .oucsEEcu
.312; 3.33 329.30 .5253 2:025 5:: EmEEoc 29: on 2.53 :quB 2:03. :5 .58 2:03. 20:. 5:: 255520
22: on 2.33 :2: .3523 32:03.:— .:oEo>> :9: 25:28 20¢. on 2:03 :22 52525 EoEcogs as... 85662

20:235.: .23 58:0 __c:u...._2:_ ES oucaEEon :0 62? SE: m 2%.: 2:03 5950 323.22ch .Ekzmasqrc12E < Ezui

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>._.m_xz<
82.35:
. 9.523.320
...uz<z_zoa .
4 wEEETEEO
2236.52
mmzmza:

 

 

 

 

 

 

 

:maZm—U

 

 

 

119

00:3 :3 003203 30:50.33 :0033: 25:23.2 05 :05w:23 2:03 00:20: :0 m_0>0_ :9: .m_:m:o_3_2
00.50-003.203 325.033 05 033:2: :3 dEm:0_3_2 002349208: 308.383 05 03:02: .0023 3:0:080 3:02:
2:03 0,030.30 :2: 2:085 :35 00:3 3:0:0E0 22: 20:2 2:03 :0803 8:093 32:000.:— .0m:3 35:95 3:02:

o. x3323 306:093 0:: 003203 30:50:33 50: 53» 320.5 2:03 30:00 .00.:QY Etchefim :0\0000§§£ .N 0.:&.Q

 

 

 

 

 

 

 

 

 

 

 

>.—.m:xz<
. a
maze... . HDZ<A=O><
J<ZOFO$§ wEEEEBm
A wEEETBEO
8032032
mmZm—mmn

 

 

 

 

 

 

 

 

 

.....
I.
......
.....
.......
......
:-
u...
......
o...
.....
.....
.....
.....
o.
..1
...
1..
........... ~— m a zm U
.3
:1
.Z
......
.......
.....
.......
........
..........
..........
.....
.....
.......
.........
.....
.......
.....
...-

 

120

110;.

 

1004

Dominance

 

 

 

Gender
40 ‘ male
30 . . female

 

—1
.4

Attachment Avoidance

Figure 3. Gender-Avoidance Interaction in Predicting Dominance. Higher levels of

attachment avoidance predicted higher levels of dominance only in men.

121

 

._O.VQ..:..

.:::0.: :0: 203 95:23.2 00:m:_~:0:-b_::00m:_ 0:08:33 0:: (:0 5:20:08 :3 53:2: 0>_m:0,:0a

.00:0:_E0: 00 .0362: 0:005:30. 0 33 80:00:03: :3 3:20:22 00:0:_Eo:-00:0:_0>0 0:08:33 0:. :03:0:0E

rz30_.::w_m 8:5»: :05. 0023 :0: w:=_0::00 :0d< .00§:.§0Q :0\.0.E0.:0m\\009 «20:20.:me Eflhéhzzfi .x. 0.3.er

 

mUZ<Z=ZOG

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

  

 

 

 

2.. u a: 50.02.:
3.
. > 8.
mUZ<:_O><
A
v .- n m w Q
:E: 0.0.0m .-
A w:_:::m-:0:5 om
A no. NO 8202032
. mmzmmm:
iwm M 2
~55sz

 

 

 

 

5.

122

 

160

 

 

 

140 —
12o -
2
3 100 ~
.0
<
“:6
C
.9
‘5
E
LU
Narcissism
High
0 . . . . Low

 

Attachment Avoidance

Figure 5. Avoidance-Narcissism Interaction in Predicting Emotional Abuse. Higher
levels of attachment avoidance, when accompanied by above average levels of
narcissistic defenses, predicted higher levels of emotional abusiveness. Attachment
avoidance did not predict emotional abuse when levels of narcissism were average or

below.

80

 

 

 

 

60 1
(D
U)
3
.0
<1:
3
C
.9
‘5
E
UJ
Narcissism
High
0 T 1 T Low

 

Attachment Avoidance

Figure 6. Avoidance-Narcissism Interaction in Predicting Emotional Abuse for Women.
With higher than average levels of narcissism, higher levels of attachment avoidance
predicted more frequent emotional abuse by women. With average or below levels of

narcissism, higher levels of attachment avoidance predicted less frequent emotional abuse

by women.

124

160

 

140‘

120*

1001

80%

60~

Emotional Abuse

404

 

Narcissism

o -— High

-20 . 1 . Low

 

 

 

..t
Ni
0)
b.
01
05
\1

Attachment Avoidance

Figure 7. Avoidance-Narcissism Interaction in Predicting Emotional Abuse in Men.
Higher levels of attachment avoidance in men predicted more frequent use of emotional
abuse strategies. When narcissism levels were also higher than average, overall emotional

abuse scores were higher regardless of level of attachment avoidance.

125

.—O. V Q** “mo. V as.
_ ,_ Q “:0 OED :h D XCN CUE 06 N 0 o C 5 —
.Q—Smcomww— : D
—m=0:*0mu 0: . 3 C
5 ...— * O—ummfioe 0>. _ mu _
“v U * Em . m_—C G O—
. :C:— d . . E “C“ UWSD .

P o N ‘A .V ~&
~ % 0‘

O x o O

(\0 I

1* N o

w:

_ 0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

>.—.m=xz<
cm ”mm cor
BO- 5:
00235:
W.
w:_:=am..:0m
A .um .
. < o_ w:_::am-:0:5 m _.
Eva: ‘ saaflp3z. _or

J<ZC_._.O$E A max I O . mmZm—mm:

A cm.

.1. _ m. H

 

............................................................................................................................................................................................

~53sz
5.-

 

 

 

126

.CQE03

:0: 003550: :2052: w5::%-:0::0 ._0>0_ 0023 :0: w5:0.::00 :0c< .5050: .:0\.00:0=.§0Q\.0 3002 0050.0K .0 0.5me

 

m—DZ<Z_EOG

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm.

 

om. H E
2.
All
+
o .- .I. m n .- n O R. M Z mUZ<A=O><

m .. l m w5::am.:0m 8.-
A I w5::am-._0:.o

mm. H 0 5505032
4 mmZm—mmn mm.

3. M Z

>._.m:xz<

 

 

127

.m5zzam-:0::0 3 :03 0:25:05 3320: :3 0:03 :0: 40:05 35: 0:: 5 :30_:_:w_m
:0: 33 003550: :3 .3053 505:053 :0033: 5:50:33: 05.30: 45:05:90 3355 0: :- :05 :0: 003550:

.2062: 550.503: :3 003503 505:053 .223 :0: w5:0.::00 :0::< .33: .:.0\00=0=.§0Q\0 3.00:: 00520:: 0 N 0.03::

 

 

 

 

 

 

 

 

 

 

 

me. H K
0:.
A »
__.um w_.uO __.-uZ HUZ<A=O><
A
EUZ<Z=ZOG A 0N - u m w5::3-:_0m m:
w55_:m-:0::0
2.- n 0 5505032
1 . 3203: N»...
mm M Z

 

 

 

 

 

 

:Nr >.—.m=xz<

 

 

 

128

.5. V RI. .mc. V R...

52000220: 0003;005:303 0:: 0:50:05 :0 0053 3002050 02:05 :00 E: :0::00 .0003

..0 33. 3 :03_:05 33 02000203: :03:0:05 m2: :3 50000220: 0053 3002050003203 505233 05 :03:0:05
3202000050 02000220: 0003 3002050003203 505233 05 :03:0:05 :3 0003 3002050 :0:0_:0:0 5030332

30:3. :0 _0>0_ >0 :03_:05 33 0003 3002050 :3 32500.20": 0003200 52000220: 05 :3 $5530.28 3 :03305

33 02300220: 0003 300205049033 505233 0:: gnaw 20002052 :0\3:0V< 202025.00: 00:20.52 .: 0.00000:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

om. "mm
A 8.
>Pm=xz<
co.
A \./
\\ //
\ / o..- n 0 0025—03
\\\ *—N. H O //
/
\ / t. u z
A \\\ // * *VN.
m .-HO 95:00:02.0 - .
mm2m< . 3.0030- 0m no
320:020 , 00003 .ewaewmz
1mm. *NN. Hm . .
A mmZm—mm—G
.Iom. M Z man—zany
S.

 

 

 

 

 

129

APPENDIXES

130

APPENDIX A

13]

INFORMED CONSENT

Thank you for your expressed interest in this study of how people act in dating
relationships. Please read this form all the way through before signng below. This
project is being conducted by Barbara Gormley, a doctoral candidate in Counseling
Psychology, under the supervision of Dr. Frederick G. Lopez, Professor in the
Department of Counseling, Educational Psychology, and Special Education, 438
Erickson Hall, Michigan State University, E. Lansing, MI 48824.

Students at Michigan State University are being asked to describe their partner
relationships. If you choose to participate in this study, you will be given a packet of self-
report questionnaires to complete. These questionnaires will ask about your feelings,
thoughts, and behaviors in dating relationships or close relationships with a romantic
partner. It will take approximately 40 minutes to complete this survey packet. No risk to
you should occur, but if you find any of the questions upsetting, please feel free to speak
with the person administering the survey.

Your responses are completely anonymous and confidential. Your name will not at any
time be attached to the answers you provide to the questions. Your name will not be on
the survey packet you complete, nor will the number of the packet be attached to your
name. This Informed Consent form will be kept separate from your survey packet. At no
time will your name be released in association with this study unless your instructor
requires notification of your participation for course credit. Your privacy will be
protected to the maximum extent allowable by law.

Your participation in this study is completely voluntary, and you may withdraw at any
time today with no penalty to you. Some instructors are assisting this project by
providing course credit for participation. These instructors will provide extra credit
alternatives for students who do not wish to participate in the survey. Should you choose
to participate, we ask you to answer all of the questions as honestly as you possibly can.
DO NOT put your name on the questionnaires.

If you would like to participate, read the brief statement below and print and sign your
name and enter today’s date on the lines below. If you have any questions about the
study, please ask the person administering the survey or call Dr. Lopez at 517-355-8502
(flopez@msu.edu). Any concerns about your rights as a participant may be directed to
UCRIHS Chair Dr. Kumar, 246 Administration Building, Michigan State University, E.
Lansing, M148824 (ucrihs@msu.edu), 517-355-2180.

 

My signature below indicates that I agree to participate in this study as described above:

 

 

PRINT your name here SIGN your name here Today’s date

132

APPENDIX B

133

DEMOGRAPHICS
Instructions: Please circle the number (or fill in the blank) to describe yourself.

A. Gender: (1)Female (2) Male (3)Transgender

 

B. Age
C. Race/Ethnicity: D. Nationality:
(1) African descent (l) lntemational student
(2) Arabic descent (2) U. S. A.
(3) Asian/Pacific Islands
(4) Caucasian/White
(5) Hispanic

(5) Native/Indian/Alaskan
(6) Multiracial/Biracial

E. Student status:
(1) Freshman (2) Sophomore (3) Junior (4) Senior

(5) Other

 

F. Sexual orientation:

(1) Lesbian/gay man (2) Bisexual (3) Heterosexual (4) Celibate

 

G. How many romantic relationships have you had?
H. Romantic Relationship Status:

(1) I have a legal, religious, or social commitment/marriage to a partner

(2) I live with my partner, but we have not had a ceremony

(3) I am in a serious relationship with one partner, but we do not live together
(4) I am currently casually dating one or more partners

(5) I am not dating anyone, but I have dated in the past

(6) I have never had a close relationship with a romantic partner

I. If currently in a romantic relationship, how long ago did it begin?

 

J. If your most recent relationship ended, how long ago?

 

134

APPENDIX C

l35

PERSONAL PROBLEMS INVENTORY

Directions: Using the scale next to each of the items below. indicate to what extent each
of these concerns is currently a problem for you. CIRCLE the appropriate number.
Not at all 3 Moderate Very significant

Problem Problem Problem
1. General anxiety .................... 1 2 3 4 5 6
2. Alcohol use ........................ 1 2 3 4 5 6
3. Shyness ............................ 1 2 3 4 5 6
4. Sexual functioning ............... 1 2 3 4 5 6
5. Depression ........................ 1 2 3 4 5 6
6. Conflicts with parents ............ 1 2 3 4 5 6
7. Speech anxiety .................... 1 2 3 4 5 6
8. Dating difficulties ................ 1 2 3 4 5 6
9. Career choice ..................... l 2 3 4 5 6
10. Insomnia ......................... 1 2 . 3 4 5 6
1 1. Drug use ......................... 1 2 3 4 5 6
12. Inferiority feelings ............. 1 2 3 4 5 6
13. Test anxiety ..................... 1 2 3 4 5 6
14. Difficulties making friends... 1 2 3 4 5 6
15. Trouble studying ................ 1 2 3 4 5 6
16. Academic performance ....... l 2 3 4 5 6
17. Financial matters ............... l 2 3 4 5 6
18. Alienation—not belonging... 1 2 3 4 5 6
19. Adjustment to college ......... 1 2 3 4 5 6
20. Loneliness/isolation ............ l 2 3 4 5 6

136

APPENDIX D

137

EXPERIENCES IN CLOSE RELATIONSHIPS-REVISED

Instructions:

The following statements concern how you feel in romantic relationships. We are
interested in how you generally experience relationships, not just in what is happening in
a current relationship. Respond to each statement by indicating how much you agree or
disagree with it. Write the number in the space provided, using the following rating scale:

 

Disagree strongly Neutral/mixed Agree strongly
1 2 3 4 5 6 7

 

_ 1. I prefer not to show a partner how I feel deep down.

2. I’m afraid that I will lose my partner’s love.

3. I am very comfortable being close to romantic partners.

4. I worry a lot about my relationships.

5. My partner really understands me and my needs.

6. I worry that romantic partners won’t care about me as much as I care about them.
7. I get uncomfortable when a romantic partner wants to be very close.

8. My partner only seems to notice me when I’m angry.

9. I don’t feel comfortable opening up to romantic partners.

10. 1 often wish that my partner’s feelings for me were as strong as my feelings for
her or him.

_ 1 1. It’s not difficult for me to get close to my partner.

_ 12. I often worry that my partner will not want to stay with me.

__ 13. I am nervous when partners get too close to me.

__ 14. I often worry that my partner doesn’t really love me.

__ 15. I feel comfortable sharing my private thoughts and feelings with my partner.

16. My desire to be very close sometimes scares people away.

17. I find it easy to depend on romantic partners.

138

 

Disagree strongly Neutral/mixed Agree strongly

1

2 3 4 5 6 7

 

18. When my partner is out of sight, I worry that he or she might become interested

19.

in someone else.

I find it relatively easy to get close to my partner.

20. When I show my feelings for romantic partners, I’m afraid they will not feel the

same about me.

. I find it difficult to allow myself to depend on romantic partners.

. I do not often worry about being abandoned.

. I prefer not to be too close to romantic partners.

. I rarely worry about my partner leaving me.

. Itell my partner just about everything.

. I find that my partner(s) don’t want to get as close as I would like.
. I usually discuss my problems and concerns with my partner.

. My romantic partner makes me doubt myself.

. I feel comfortable depending on romantic partners.

. Sometimes romantic partners change their feelings about me for no apparent

reason.

. It’s easy for me to be affectionate with my partner.

. I’m afraid that once a romantic partner gets to know me, he or she won’t like

who I really am.

. It helps to turn to my romantic partner in times of need.

. It makes me mad that Idon’t get the affection and support I need from my

partner.

. I talk things over with my partner.

. I worry that I won’t measure up to other people.

139

APPENDIX E

140

SPLITTING INDEX

Instructions:
Using the scale next to each of the statements below, indicate your level of agreement or
disagreement with each of these statements by CIRCLING the most appropriate number.

 

Strongly Strongly
DISagree Agree
1. I feel differently about myself when I am
with different people ............................. l 2 3 4 5
2. Being able to keep friends is one of my
strong points ......................................... l 2 3 4 5

3. My feelings about myself shift dramatically. l 2 3 4 5

4. The different parts of my personality are

difficult to put together ......................... 1 2 3 4 5
5. I have doubts about my closest friends. 1 2 3 4 5
6. Sometimes I am not sure who I am. 1 2 3 4 5

7. My feelings about myself are very powerful, but
they can change from one moment to the next. 1 2 3 4 5

8. My friendships are almost always satisfying. 1 2 3 4 5

9. My feelings about myself do not change easily. 1 2 3 4 5

10. I have had many long-lasting friendships. 1 2 3 4 5
1 1. I sometimes feel "pulled apart" by my feelings
about myself .......................................... 1 2 3 4 5

12. My feelings toward those close to me

remain constant ............................... 1 2 3 4 5
13. I have always been aware that my close

friends really cared for me ............... 1 2 3 4 5
14. My opinions of my friends rarely change. 1 2 3 4 5
15. I almost always feel good about those close

to me .................................... 1 2 3 4 5
16. Who I am depends on how I am feeling. 1 2 3 4 5

141

APPENDIX F

142

NARCISSISTIC PERSONALITY INVENTORY

Instructions:
In each of the following pairs of attitudes, choose the one that you MOST AGREE with.
Mark your answer by CIRCLING EITHER A or B. Only mark one answer for each pair,

and please do not skip any items.

1.

Ex)

10.

11.

12.

13.

A.
B.

9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°?

9°?

I have a natural talent for influencing people.
I am not good at influencing people.

Modesty doesn’t become me.
I am essentially a modest person.

I would do almost anything on a dare.
I tend to be a fairly cautious person.

When people compliment me, I sometimes get embarrassed.
I know that I am good because everybody keeps telling me so.

The thought of ruling the world frightens the hell out of me.
If I ruled the world, it would be a better place.

I can usually talk my way out of anything.
I try to accept the consequences of my behavior.

I prefer to blend in with the crowd.
I like to be the center of attention.

I will be a success.
I am not too concerned about success.

I am no better or no worse than most people.
I think I am a special person.

I am not sure if I would make a good leader.
I see myself as a good leader.

I am assertive.
I wish I were more assertive.

I like having authority over other people.
I don’t mind following orders.

I find it easy to manipulate people.
Idon’t like it when I find myself manipulating people.

143

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°? 9°?

.w.>

I insist upon getting the respect that is due me.
I usually get the respect that I deserve.

I can read people like a book.
People are sometimes hard to understand.

I just want to be reasonably happy.
I want to amount to something in the eyes of the world.

I try not to be a show off.
I will usually show off if I get the chance.

Sometimes Itell good stories.
Everybody likes to hear my stories.

I expect a great deal from other people.
I like to do things for other people.

I will never be satisfied until I get all that I deserve.
I take my satisfactions as they come.

Compliments embarrass me.
I like to be complimented.

I have a strong will to power.
Power for its own sake doesn’t interest me.

I don’t care about new fads and fashions.
I like to start new fads and fashions.

I really like to be the center of attention.
It makes me uncomfortable to be the center of attention.

Being an authority doesn’t mean that much to me.
People always seem to recognize my authority.

I would prefer to be a leader.
It makes little difference to me whether I am a leader or not.

People sometimes believe what I tell them.
I can make anybody believe anything I want them to.

I am a born leader.
Leadership is a quality that takes a long time to develop.

144

30.

31.

32.

33.

9°? 9°? 9°? 9’?

9°?

I wish someone would someday write my biography.
I don’t like people to pry into my life for any reason.

I get upset when people don’t notice how I look when I go out in public.
I don’t mind blending into the crowd when I go out in public.

I am more capable than other people.
There is a lotthat I can learn from other pe0ple.

I am much like everybody else.
I am an extraordinary person.

I am envious of other people’s good fortune.
When something good happens to other people, I am happy for them.

145

APPENDIX G

146

THE DOMINANCE SCALE

Instructions:

People have many different ways of relating to each other. The following statements are
all different ways of relating to or thinking about your partner. Please read each statement
and CIRCLE the number that corresponds to how much you agree with it.

Strongly Dis- Strongly
DISagree agree Agree Agree
1. My partner often has good ideas ................ 1 2 3 4

2. I try to keep my partner from spending time
with opposite sex friends .......................... 1 2 3 4

3. If my partner and I can't agree, I usually
have the final say ...................................... 1 2 3 4

4. It bothers me when my partner makes plans
without talking to me first ........................ 1 2 3 4

5. My partner doesn't have enough sense to

make important decisions .......................... 1 2 3 4
6. I hate losing arguments with my partner ..... 1 2 3 4
7. My partner should not keep any secrets from me. 1 2 3 4

8. I insist on knowing where my partner is
at all times .................................................... 1 2 3 4

9. When my partner and I watch TV, I hold the
remote control ............................................... 1 2 3 4

10. My partner and I generally have equal say
about decisions ............................................. 1 2 3 4

11. It would bother me if my partner made more

money than I did .......................................... 1 2 3 4
12. I generally consider my partner's interests

as much as mine ............................................. 1 2 3 4
13. I tend to be jealous ....................................... 1 2 3 4

14. Things are easier in my relationship if
I am in charge ............................................... 1 2 3 4

147

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Strongly

DISagree

Sometimes I have to remind my partner of
who's boss ..................................................

I have a right to know everything my
partner does ...............................................

It would make me mad if my partner did
something I had said not to do ..................

Both partners in a relationship should have
equal say about decisions .........................

If my partner and I can't agree, I should
have the final say ......................................

I understand there are some things my
partner may not want to talk about with me.

My partner needs to remember that I am
in charge ......................................................

My partner is a talented person ...................

It's hard for my partner to learn new things.
People usually like my partner ....................

My partner makes a lot of mistakes ...........

My partner can handle most things that happen.
I sometimes think my partner is unattractive.
My partner is basically a good person .........

My partner doesn't know how to act in public.

I often tell my partner how to do something.

I dominate my partner ..................................

I have a right to be involved with anything
my partner does ............................................

148

1

1

1

1

Dis-
agree

to

3

Strongly
Agree Agree

4

APPENDIX H

149

MULTIDIMENSIONAL MEASURE OF EMOTIONAL ABUSE

Instructions:

The following questions ask about the relationship with your partner or ex-partner. Please
report how often each of these things has happened in the last six months (or the last six
months of the relationship if it is over). Please circle a number using the scale below to
indicate how often you have done each of the following things, and a number to indicate
how often your partner has done each of the following things. Indicate how many times
you have done this where it says ‘you’ and how many times your partner has done this
where it says ‘your partner.’ If you or your partner did not do one of these things in the
past six months, but it happened before that, circle ‘7.’

( I ) Once (4 ) 6-10 times (7) Never in the past six months, but before that
(2) Twice (5) 11-20 times (0) This has never happened
(3 ) 3-5 times (6 ) Over 20 times

Once Twice 3-5 6-10 11-20 20+ Not in Never
6 months

1. Asked the other person where they had been or who they were with in a suspicious
manner.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 0
2. Secretly searched through the other person’s belongings.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

3. Tried to stop the other person from seeing certain friends or family members.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 0

4. Complained that the other person spends too much time with friends.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

5. Got angry because the other person went somewhere without telling (you).

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

150

Once Twice 3-5 6-10 11-20 20+ Not in Never
6 months

6. Tried to make the other person feel guilty for not spending enough time together.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

7. Checked up on the other person by asking friends or relatives where they were or who
they were with.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
8. Said or implied that the other person was stupid.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
9. Called the other person worthless.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
10. Called the other person ugly.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
11. Criticized the other person’s appearance.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
12. Called the other person a loser, failure, or similar term.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
13. Belittled the other person in front of other people.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 0

151

Once Twice 3-5 6-10 11-20 20+ Not in Never
6 months

14. Said that someone else would be a better partner (spouse, girlfriend, boyfriend).

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
15. Became so angry that they were unable or unwilling to talk.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O
16. Acted cold or distant when angry.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 0
l7. Refused to have any discussion of a problem.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

18. Changed the subject on purpose when the other person was trying to discuss the
problem.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

19. Refused to acknowledge a problem that the other person felt was important.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7

20. Sulked or refused to talk about an issue.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

21. Intentionally avoided the other person during a conflict or disagreement.

You 1 2 3 4 5 6 7 0
Your partner 1 2 3 4 5 6 7 O

152

Once

22. Became angry enough to frighten the other person.

You
Your partner

23. Put (your) face right in front of the other person’s face to make a point more

forcefully.

You
Your partner

1
1

1
1

Twice

2
2

2

3-5

3
3

3

6-10

4
4

4

24. Threatened to hit the other person.

You
Your partner

25. Threatened to throw something at the other person.

You
Your partner

1

1

2

2
2

3

3
3

4
4

4
4

11-20 20+

5
5

U1

5
5

5
5

6
6

GO

6
6

6
6

Not in
6 months

7
7

\l

\l\)

\l

26. Threw, smashed, hit, or kicked something in front of the other person.

You
Your partner

27. Drove recklessly to frighten the other person.

You
Your partner

28. Stood or hovered over the other person during a conflict or disagreement.

You
Your partner

1
1

1
1

1
1

2
2

2
2

2
2

3
3

3
3

3
3

4

4
4

4
4

5
5

5
5

5
5

153

6
6

6
6 .

6
6

\l

\l\)

Never