Is Domestic Violence Learned? The Contribution of Five Forms of Child Maltreatment to Men's Violence and Adjustment

Abstract

On the basis of a learning-theory approach to the intergenerational transmission of violence, researchers have focused almost exclusively on violent men's childhood experiences of physical abuse and witnessing family violence. Little consideration has been given to the coexistence of other forms of child maltreatment or the role of family dysfunction in contributing to violence. This study shows the relationships between the level of child maltreatment (physical abuse, psychological maltreatment, sexual abuse, neglect, and witnessing family violence), childhood family characteristics, current alcohol abuse, trauma symptomatology, and the level of physical and psychological spouse abuse perpetrated by 36 men with a history of perpetrating domestic violence who had attended counseling. As hypothesized, a high degree of overlap between risk factors was found. Child maltreatment, low family cohesion and adaptability, and alcohol abuse was significantly associated with frequency of physical spouse abuse and trauma symptomatology scores, but not psychological spouse abuse. Rather than physical abuse or witnessing family violence, childhood neglect uniquely predicted the level of physical spouse abuse. Witnessing family violence (but not physical abuse) was found to have a unique association with psychological spouse abuse and trauma symptomatology. These results present a challenge to the understanding of domestic violence obtained from learning theory.

Full text

Journal of Family Violence, Vol. 17, No. 3, September 2002 (C°2002)
Is Domestic Violence Learned? The Contribution
of Five Forms of Child Maltreatment to Men’s
Violence and Adjustment
Emma Bevan1and Daryl J. Higgins1,2
On the basis of a learning-theory approach to the intergenerational transmis-
sion of violence, researchers have focused almost exclusively on violent men’s
childhood experiences of physical abuse and witnessing family violence. Little
consideration has been given to the coexistence of other forms of child mal-
treatment or the role of family dysfunction in contributing to violence. This
study shows the relationships between the level of child maltreatment (physical
abuse, psychological maltreatment, sexual abuse, neglect, and witnessing fam-
ily violence), childhood family characteristics, current alcohol abuse, trauma
symptomatology, and the level of physical and psychological spouse abuse
perpetrated by 36 men with a history of perpetrating domestic violence who
had attended counseling. As hypothesized, a high degree of overlap between
risk factors was found. Child maltreatment, low family cohesion and adapt-
ability, and alcohol abuse was significantly associated with frequency of phys-
ical spouse abuse and trauma symptomatology scores, but not psychological
spouse abuse. Rather than physical abuse or witnessing family violence, child-
hood neglect uniquely predicted the level of physical spouse abuse. Witnessing
family violence (but not physical abuse) was found to have a unique associ-
ation with psychological spouse abuse and trauma symptomatology. These
results present a challenge to the understanding of domestic violence obtained
from learning theory.
KEY WORDS: domestic violence; child maltreatment; family background; learning theory.
1School of Psychology, Deakin University, Waurn Ponds, Victoria, Australia.
2To whom correspondence should be addressed at School of Psychology, Deakin University,
Geelong, VIC 3217, Australia; e-mail: dhiggins@deakin.edu.au.
223
0885-7482/02/0900-0223/0 C
°2002 Plenum Publishing Corporation

224 Bevan and Higgins
INTRODUCTION
Most domestic violence research has received impetus from social learn-
ing theory, and accordingly, researchers have sought evidence for a direct
link between exposure to violence in childhood and later perpetration of
domestic violence, with little consideration given to other intervening vari-
ables. Many factors associated with domestic violence do not occur uniquely,
and therefore have both individual and combined influences on later adult
adjustment, and in particular, the level of spouse abuse perpetrated by men.
Empirical investigation is required to test the unique contribution of these
risk factors to the level of violence and to establish whether violence is, in
fact, learned, or if the process is far more complex than simply violence
begetting violence. The purpose of this study was to explore the degree to
which childhood family functioning, the experience of child maltreatment,
alcohol abuse, parental divorce, and income intercorrelate, and their rela-
tive contributions to predicting the level of violence perpetrated, and current
adjustment in a sample of domestically violent men.
There is a range of limitations and methodological flaws to research
on domestic violence, and the theories on which it is based. Firstly, difficul-
ties are encountered when attempting to define domestic violence, as not
all domestic violence occurs in the home, occurs between married part-
ners, or is physically violent (Melvin et al., 1999). An aspect of domes-
tic violence that is often not sufficiently addressed is psychological spouse
abuse (Shepard & Campbell, 1992). Further problems arise with the freq-
uent use of clinical samples and retrospective surveys, which can, for the
former, lead to a lack of generalizability of results and, for the latter, be
confounded by inherent problems such as reporting bias, lapses in mem-
ory, emotional blocking, and traumatic repression (Urquiza, 1991). In ad-
dition, domestic violence research is often undermined through theoretical
discussions that infer, or in some cases state, causality when research in
this area is primarily correlational (see Tolman & Bennett, 1990). Finally,
most domestic violence research is dominated by theories that have sought
to explain the perpetration of domestic violence by one or a few factors
(e.g., having experienced childhood physical abuse or witnessed domestic
violence).
Theories postulated thus far have focussed on patriarchy, learned be-
havior, psychopathology, personality traits, and biological factors as expla-
nations for incidence of domestic violence (see Anderson, 1997; Hotaling
& Sugarman, 1986; Kalmuss, 1984; McKenry et al., 1995; Schuerger & Reigle,
1988). Yet, domestic violence researchers have not proposed a theory
capable of thoroughly explaining spouse abuse (Hotaling & Sugarman,
1986; McKenry et al., 1995). This study seeks to test a theoretical model

Is Domestic Violence Learned? 225
of the perpetration of domestic violence (learning theory), as well as in-
cluding a wider range of risk factors known to be associated with domestic
violence.
Domestic violence researchers have focussed on the role of childhood
experiences that expose boys to violent role models and the associated
heightened risk of perpetration of domestic violence in adulthood (Mihalic
& Elliott, 1997). The basic premise of theories concerning the intergenera-
tional transmission of abuse is that being a victim of physical abuse, or wit-
nessing the abuse of other family members teaches boys to become violent.
A few interrelated theoretical mechanisms may be at work: identification
with the aggressor, vicarious reinforcement, and positive reinforcement of
aggression.
Identification With the Aggressor
This view posits that those exposed to family aggression later act aggres-
sively toward their own family members if they identify with the aggressor.
Individuals are more likely to act aggressively if they have been exposed to
violence in the family of origin and have also identified with the aggressor
than if there has been exposure to aggression in the family of origin, but no
identification with the aggressor (MacEwen, 1994).
Observational Learning (Vicarious Reinforcement)
The basic premise of this view is that physical aggression between family
members provides a likely model for the learning of aggressive behavior, as
well as for the appropriateness of such behavior within the family (Bandura,
1973; Kalmuss, 1984). Thus, intergenerational transmission of violence stems
primarily from principles of modeling (Doumas et al., 1994).
Positive Reinforcement
This line of reasoning suggests that a violent father may positively re-
inforce early signs of violent behavior not only by exposing individuals to
violence, but by teaching approval for the use of violence (Gelles, 1972). As
a result, children may conclude that physical violence is sometimes a nec-
essary and effective strategy for achieving behavioral change in family and
intimate relationships (Simons et al., 1998).
The literature has focused therefore on the role of childhood experi-
ences of physical abuse and witnessing family violence in leading to men’s

226 Bevan and Higgins
violence within the family. Few studies have focused on whether it is only
physical abuse and witnessing family violence that influence men’s level of
violence, or whether other forms of child maltreatment may play a role. In
one of the few studies that address this question, Andrews and Brown (1988)
found that neither the direct experience of violence as a child nor witnessing
violence was associated with later violence. Instead, they found violence to
be associated with childhood neglect, irrespective of whether they had been
physically abused in childhood.
An ecological approach involves examining both the environmental
and individual factors to which an individual responds, and the developmen-
tal stage at which these factors occurred to explain behavior (see Carlson,
1997; Fraser, 1996). Child maltreatment researchers have recently advocated
use of developmental and ecological approaches (see Rosenberg, 1987),
whereby the extent of all forms of child maltreatment are assessed, and then
their specific and overlapping associations with various types of psychologi-
cal symptoms are evaluated (Briere & Runtz, 1990). Thus, the contribution
of various factors is evaluated in connection with each other resulting in a
more complex, but authentic insight into child maltreatment. Comparably,
an ecological approach to the study of domestic violence would generate
a more unified theory, whereby influencing variables that have previously
been studied singly could be subsumed into a more useful and realistic con-
ceptualization of factors influencing the occurrence and extent of domestic
violence.
Domestic violence researchers have produced substantial evidence for
the existence of strong associations between male perpetration of domestic
violence and subjection to physical abuse and witnessing domestic violence
during childhood (see Caesar, 1988; Cappell & Heiner, 1990; Kalmuss, 1984),
low socioeconomic status in adulthood (see McKenry et al., 1995), and alco-
hol abuse (see Leonard & Blane, 1992). Approximately half of all domesti-
cally violent men attending counseling have alcohol abuse problems (Tolman
& Bennett, 1990). Researchers have also found separate relationships be-
tween three factors: child maltreatment (subjection to physical abuse and
witnessing family violence during childhood) and the adult perpetration of
domestic violence (Rosenbaum & O’Leary, 1981), child maltreatment
(subjection to physical abuse during childhood) and adult substance abuse
(Duncan et al., 1996), and alcohol abuse and the perpetration of domestic
violence in adulthood (Tolman & Bennett, 1990). Although these findings
suggest the possibility of an overlap in the existence of child maltreatment,
alcohol abuse, and the perpetration of spouse abuse, domestic violence re-
searchers have not yet examined the shared relationships between these
variables, and their shared influence on adult adjustment.

Is Domestic Violence Learned? 227
Child maltreatment types have also been found to co-occur. How-
ever, when considering the relationship between child maltreatment and
the later perpetration of domestic violence, researchers adopting a learning
theory perspective have largely focussed on only two types: the experience
of physical abuse and witnessing family violence during childhood (Maker
et al., 1998; Marshall & Rose, 1988). Recent findings have demonstrated
that child maltreatment types rarely occur in isolation (Briere & Runtz,
1988; Ney et al., 1994; see Higgins & McCabe, 2001a). Multitype mal-
treatment (the coexistence of one or more of the following maltreatment
types: sexual abuse, physical abuse, psychological abuse, neglect, and
witnessing family violence) is common among maltreated individuals
(Bernstein et al., 1994; Higgins & McCabe, 1998, 2000b; Kinard, 1994; Mullen
& Fleming, 1998; Tomison, 1995). Higgins and McCabe (2000b) reported
that although all maltreatment types are significantly intercorrelated, the
coexistence of physical abuse and neglect, and psychological maltreatment
and neglect were the combinations most often reported. Thus, to simply fo-
cus upon the influence of physical abuse and witnessing family violence on
the adult perpetration of domestic violence ignores the potential contribu-
tion of other, unacknowledged but associated, maltreatment types on men’s
adjustment, particularly their level of physical and psychological spouse
abuse.
The experience of child maltreatment is associated with poor adjust-
ment in both childhood and adulthood. Increased aggression has been found
to be a long-term correlate of exposure to family violence (Forsstrom-Cohen
& Rosenbaum, 1985), and psychological maltreatment (Hart & Brassard,
1987), and a short-term correlate of physical abuse (Ammerman et al., 1986;
Kinard, 1982), exposure to family violence (Mathias et al., 1995;
Parkinson & Humphreys, 1998), and neglect during childhood (Ammerman
et al., 1986; Reidy, 1977). Similarly, internalizing psychological problems are
commonly found in children and adolescents subjected to physical abuse
(Kinard, 1982; Malinosky-Rummell & Hansen, 1993), as well as in adults
who report childhood sexual abuse (Mullen et al., 1993, 1994), and those who
witnessed family violence during childhood (Henning et al., 1997). Sexual
abuse has also been associated with sexual difficulties, and psychopathol-
ogy in adult life (Mullen et al., 1993, 1994). Researchers have also demon-
strated that experiencing multiple forms of maltreatment (“multitype mal-
treatment”) is associated with greater pathology (Briere & Runtz, 1990;
Higgins & McCabe, 2000a,b, 2001a). Consequently, not only do the mal-
treatment types individually influence adult adjustment levels, but their
coexistence collectively augments the level of dysfunction experienced in
adulthood.

228 Bevan and Higgins
A dysfunctional family background has also been shown to relate to
poor adult adjustment (Higgins & McCabe, 1994). Low childhood family
income and poor interpersonal relationships have been associated with adult
perpetration of domestic violence (McKenry et al., 1995). The experience of
parental divorce during childhood has also been linked to poor functioning
later on in adulthood, such as reduced marital well-being and higher lev-
els of divorce (Amato & Booth, 1991; Garbardi & Rosen, 1992). Similarly,
childhood family background variables, such as low cohesion and adaptabil-
ity, have been found to significantly predict adult trauma symptomatology
(Higgins & McCabe, 2000b) and adult social maladjustment (Harter et al.,
1988).
There has been considerable debate within the child maltreatment liter-
ature concerning the relative contribution of maltreatment and family dys-
function to adjustment. In particular, many researchers have shown that the
effects of sexual abuse are largely accounted for by the dysfunctional fam-
ily environment (e.g., Fromuth, 1986; Higgins & McCabe, 1994). It should
be noted, though, that these researchers often failed to identify the contri-
bution of other specific types of child maltreatment. Higgins and McCabe
(2000b) found that after accounting for childhood family background vari-
ables, the level of five different forms of child maltreatment made a signif-
icant contribution to the prediction of adult trauma symptomatology and
self-depreciation. Thus, coexistence of both a negative family environment
and child maltreatment was associated with greater levels of trauma
symptomatology and self-depreciation than negative family background
alone.
In this study, the relationships between childhood family functioning,
child maltreatment, alcohol abuse, parental divorce, trauma symptomatol-
ogy, and frequency of perpetration of physical and psychological spouse
abuse will be assessed in a sample of men from a counseling agency pro-
viding services for violent families. In line with previous research on family
characteristics and the interrelationships between maltreatment types, it is
hypothesized that there will be a high degree of overlap between childhood
family background, alcohol abuse, parental divorce, income, and child mal-
treatment, especially the five maltreatment types, reported by domestically
violent men. It is predicted that low childhood family adaptability and cohe-
sion scores, high scores on a measure of maltreating behaviors and high
scores on a measure of alcoholism, will be associated with higher levels of
trauma symptomatology and greater frequency for the perpetration of psy-
chological and physical spouse abuse. To test a learning theory explanation
of domestic violence, it is hypothesized that, after accounting for the scores
on the childhood physical abuse and witnessing family violence scales, scores
on the scales for psychological maltreatment, sexual abuse, and neglect will

Is Domestic Violence Learned? 229
not significantly add to the prediction of the frequency of physical and psy-
chological spouse abuse, and levels of trauma symptomatology.
METHOD
Participants
The sample comprised 36 male clients who had attended a counseling
agency for therapy for various periods of time since 1996–99 and who had
agreed at the time to be contacted at a later stage for research purposes. All
of the men in the sample had a history of domestic violence where they had
abused their female spouse either physically or psychologically. The mean
age of the sample was 41.9 years (SD =9.45). The modal rating on income
was 1 (where 1 =between $20,000 and $35,000 Australian per annum).
Participants’ racial/ethnic background was described as “Australian” by 58%
of the sample (n=21), with 28% describing themselves as WesternEuropean
(n=10), 8% considered themselves to be Eastern European (n=3), 1 (3%)
participant described himself as Central American (n=1), and 1 participant
(3%) indicated that he was Indian. Questionnaires were posted to 257 clients;
however, 35 were returned to sender. Consequently, 222 men were success-
fully contacted resulting in a response rate of 16.2%. This is comparable to
other studies that have used similar methodology with a domestically violent
sample (see Selzer, 1971).
Measures
Childhood family functioning was assessed using an adaptation of the
Family Adaptability and Cohesion Evaluation Scales (FACES II; Olson
et al., 1982) specifically designed for retrospective recall (Higgins & McCabe,
2000b). This is a standardized self-report measure that assesses two dimen-
sions of family functioning: family adaptability (emotional bonding between
family members) and family cohesion (the family’s ability to adapt relation-
ship rules, power structure, and role relationships in response to situational
and developmental stress; see Olson et al., 1983). Using linear scoring, respo-
ndents rated the 30 items on a 5-point scale (1 =almost never to5=almost
always). Low family cohesion scores represent a family that was disengaged,
whereas high scores indicate a family that was very connected. Low family
adaptability scores represent a family that was rigid, whereas high scores
indicate a family that was very flexible. FACES II has evidence of good face,
content, and concurrent validity (Olson et al., 1982). The version of FACES II

230 Bevan and Higgins
modified for retrospective recall has a test-retest reliability at 6–8 weeks of
.91 for adaptability and .92 for cohesion (Higgins & McCabe, 2000b).
Alcohol consumption was measured using the Michigan Alcohol Scree-
ning Test (MAST). The MAST is a 25-item paper-and-pencil test developed
by Selzer (1971) to provide a consistent and quantifiable measure for detect-
ing alcoholism. The responses are dichotomous, with participants circling ei-
ther Yes or No to indicate their answer. Three items are considered uniquely
diagnostic of alcoholism and are assigned a value of 5 points. (Although
only weighted with 2 points in the original MAST, Item 19 was assigned a
value of 5 points, as suggested by Selzer, 1971.) Sixteen items that are highly
discriminating are given a value of 2 points, and the remaining items 1 point.
Once the items are summed, Selzer indicated that a score of 5 points or
more is diagnosistic of alcoholism. Internal reliability coefficients for the
MAST range from .83 to .95. The MAST has consistently shown to possess
good face validity, and concurrent validity coefficients range from .79 to .90
(Hedlund & Vieweg, 1984).
The Trauma Symptom Checklist – 40 (TSC-40) was used to assess the
traumatic impact of childhood abuse on later adult functioning (Briere &
Runtz, 1989). Respondents reported the frequency with which they had ex-
perienced each of the 40 items over a 2-month period on a 4-point Likert
scale (0 =never to 3 =very often). In addition to a total scale score, the
TSC-40 yields six subscale scores: Anxiety, Depression, Dissociation, Sexual
Abuse Trauma Index, Sexual Problems, and Sleep Disturbance. Responses
to all items were summed to provide a total score, and specific items were
summed to provide scores for the subscales. Higher scores indicated greater
levels of trauma symptomatology. Internal reliability coefficients for the sub-
scales range from an alpha of .62 to .77, and alpha for the TSC-40 total is
.90 (Elliott & Briere, 1992). The TSC-40 has also demonstrated predictive
and discriminative validity, discriminating well between sexually abused and
nonabused respondents (Elliott & Briere, 1992).
The Abusive Behavior Inventory (ABI; Shepard & Campbell, 1992) was
used to measure the frequency of physical and psychological spousal abuse
perpetrated by the respondents during the 6 months prior to commencing
counseling. This test was devised to include a wide range of abusive behaviors
not involving the use of physical force, which had not yet been adequately
incorporated into other measures of spousal abuse (Shepard & Campbell,
1992). Accordingly, the test includes both physical and psychological abuse
items. The ABI is a 30-item self-report instrument that uses a 5-point Likert
scale (1 =never to5=very frequently) to measure the frequency of abusive
behaviors over a 6-month period. Scores on the 10 physical abuse items
are summed and then divided by 10 to obtain a physical abuse score. The
scores for the remaining 20 psychological abuse items were summed and then

Is Domestic Violence Learned? 231
divided by 20 to obtain a psychological abuse score. For both scales, scores
ranged from 1 (no abuse)to5(very frequent abuse). The ABI is capable
of distinguishing between groups of spouse abusers and nonabusers, using
the reports of men or women (Shepard & Campbell, 1992). The scale has
also demonstrated good reliability with alpha coefficients ranging from .70
to .92, and good criterion-related, construct, and factor validity (Shepard &
Campbell, 1992).
Subjection to child maltreatment was assessed using the Comprehen-
sive Child Maltreatment Scale for Adults (CCMS-A; Higgins & McCabe,
2001b), a self-report instrument used to assess the frequency of behaviors
that could, potentially, be considered maltreatment. Respondents reported
childhood subjection to sexual abuse (9 items for perpetrating mother,
11 items for perpetrating father, and 11 items for perpetrating other, be-
ing either another adult or an adolescent who was at least 5 years older
than the respondent); physical abuse (3 items each for perpetrating mother,
father, or other); psychological maltreatment (3 items each for perpetrat-
ing mother, father, or other); neglect (3 items each for perpetrating mother,
father, or other); and witnessing family violence (2 items). Each item was
responded to using a 5-point Likert scale (1 =never or almost never to5=
very frequently), and sexual abuse items were rated on a 6-point scale (0 =
never,1=once,2=twice,3=3–6 times,4=7–20 times,5=more than
20 times). Items for each maltreatment type were summed to provide a
score for each subscale. Scores on all five subscales were summed to pro-
duce a total score. The sexual abuse scale had more items than the other
scales because of the large number of different sexual behaviors that may
have been directed toward the child. Test-retest reliability for the subscales at
a 6–8-week interval were sexual abuse, .95; physical abuse, .87; psychological
maltreatment, .84; neglect (by mother and father only), .62; and witnessing
family violence, .77. Tests for internal consistency showed that Cronbach’s
alpha for the total CCMS-A was .92. The CCMS-A has also demonstrated
good concurrent criterion-related validity (Higgins & McCabe, 2001b).
Procedure
A counseling agency that facilitated a family violence program was con-
tacted to request their involvement in the study. The agency as well as the
university provided ethical clearance for the study. Former and current male
clients with a history of perpetrating domestic violence who had agreed to
be contacted in the future for the purpose of research were sent a question-
naire, a plain language statement describing the study, and a large reply-paid
envelope.

232 Bevan and Higgins
RESULTS
Descriptive Sample Statistics
Demographic Characteristics
Parental divorce had been experienced by 10 (28%) of the men, oc-
curring at a mean age of 9 years (SD =6.63); however, ages ranged from
1 to 19 years. The remaining 26 (72%) men came from intact families.
The modal rating for number of counseling sessions was 1 (where 1 =
1–5 sessions). A majority of participants (69.4%, n=25) had undergone
counseling prior to their involvement with the family violence program,
and 22% (n=8) had focussed on issues around domestic violence or abuse
during childhood.
Childhood Family Characteristics
Participants completed the FACES II (Olson et al., 1982), using the
version adapted for retrospective recall by Higgins and McCabe (2000b).
The mean cohesion score was 53.69 (SD =14.79) with scores ranging from
29 to 108. Only 4 (11.1%) respondents were raised in “very connected”
childhood family environments, 7 (19.5%) participants were “connected,” 9
(25%) respondents were “separated,” and 16 (44.4%) participants were from
“disengaged” families. The mean adaptability score was 33.81 (SD =10.53)
with scores ranging from 15 to 57. There was 1 (2.8%) participant who was
raised in a “very flexible” childhood family environment, 4 (11.1%) respon-
dents’ families were “flexible,” 8 (22.2%) participants were “structured,” and
23 (63.9%) respondents were from “rigid” families.
Maltreatment During Childhood
Respondents rated the frequency with which they believed themselves
to have been subjected to five types of maltreating behaviors as a child.
The mean score for the Physical Maltreatment scale was 4.72 (SD =3.39)
with scores ranging from 0 to 14. The mean score for the Psychological
Maltreatment scale was 8.08 (SD =5.93) with scores ranging from 0 to 25.
Witnessing Domestic Violence scale had a mean of 2.75 (SD =2.05) with
scores ranging from 0 to 8. The mean score for the Experience of Neglect
scale was 2.33 (SD =3.47) with scores ranging from 0 to 12, and the mean

Is Domestic Violence Learned? 233
score for Child Sexual Abuse scale was 2.33 (SD =6.72) with scores ranging
from 0 to 33.
Alcohol Abuse
The mean score was 8.35 (SD =11.24) with scores ranging from 0 to 49.
Exactly 50% (n=18) of respondents had scores of 5 or over indicating alco-
holism, 11.1% (n=4) had a score of 4, which suggested possible alcoholism,
and 38.9% (n=14) had scores that were not indicative of alcoholism.
Trauma Symptomatology
Respondents rated the frequency with which they had experienced a
range of symptoms (commonly associated with the trauma of childhood
abuse) in the last 2 months. The means for the subscales were dissociation,
4.11 (SD =3.25); anxiety, 5.41 (SD =3.48); depression, 7.33 (SD =4.20);
sexual abuse trauma index, 3.64 (SD =3.15); sleep deprivation, 6.92 (SD =
4.49); and sexual problems, 5.63 (SD =4.69). The mean for the TSC-40 total
score was 31.22 (SD =17.31).
Physical and Psychological Spousal Abuse
Participants reported the frequency with which they had perpetrated
both physical and psychologically abusive behaviors during the 6-month
period before commencing counseling. The mean for physical abuse was
1.36 (SD =.59). The mean across the 10 items for each respondent ranged
from 1.0 to 3.4. There was no reporting of frequent or very frequent physical
spouse abuse by any of the participants. In addition, 41.7% (n=15) reported
that they had not perpetrated any physical abuse toward their spouse during
the 6 months prior to counseling. The mean for psychological abuse was 1.82
(SD =.71). The mean for each respondent across the 20 items ranged from
1.1 to 4.3, indicating that all of the respondents had perpetrated psycholog-
ical abuse toward their spouse at least once.
Intercorrelations Between All Variables
All of the variables correlated significantly with at least one other vari-
able except for alcohol abuse (see Table I). All of the maltreatment scales

Table I. Intercorrelation Matrix of Variables
1. 11.
Physical 2. 3. 4. 5. 6. 7. 8. 9. Witnessing 13.
spouse Psychological Trauma Alcohol Family Family Physical Psychological Sexual 10. family 12. Childhood
abuse spouse abuse symptoms abuse cohesion adaptability abuse maltreatment abuse Neglect violence Parental family
Variable (ABI) (ABI) (TSC-40) (MAST) (FACES-II) (FACES-II) (CCMS-A) (CCMS-A) (CCMS-A) (CCMS-A) (CCMS-A) divorce income
1. 1.00
2. .65∗∗ 1.00
3. .38∗.53∗∗ 1.00
4. −.19 −.08 .07 1.00
5. −.33 −.36∗−.28 .18 1.00
6. .06 −.05 −.04 −.17 .19 1.00
7. .24 .21 .21 .09 −.24 −.42∗1.00
8. .13 .37∗.40∗.23 −.31 −.41∗.64∗∗ 1.00
9. .51∗∗ .22 .18 −.18 −.23 −.25 .64∗∗ .40∗1.00
10. .53∗∗ .44∗∗ .47∗∗ .12 −.15 −.28 .55∗∗ .54∗∗ .59∗∗ 1.00
11. .16 .45∗∗ .54∗∗ .22 −.25 −.38∗.49∗∗ .58∗∗ .16 .54∗∗ 1.00
12. .21 .24 .17 .05 −.18 .33∗−.21 .04 −.12 −.06 .02 1.00
13. −.35∗−.09 −.25 −.04 .17 .00 −.39∗−.21 −.28 −.32 −.11 −.05 1.00
∗p<.05. ∗∗ p<.01.
234

Is Domestic Violence Learned? 235
were significantly intercorrelated (with positive correlations ranging from
.40 to .64, p<.05), except for the association between witnessing family
violence and sexual abuse (r=.16, p>.05).
Physical spouse abuse positively correlated with psychological spouse
abuse (r=.65). Physical spouse abuse also significantly correlated with
neglect (r=.53), sexual abuse (r=.51), and income (r=−.35). Alcohol
abuse, family cohesion, family adaptability, physical abuse, psychological
abuse, witnessing family violence, and parental divorce did not significantly
correlate with physical spouse abuse. Psychological spouse abuse signifi-
cantly correlated with witnessing family violence (r=.45), neglect (r=.44),
psychological abuse (r=.37), and family cohesion (r=−.36). Correlations
between psychological spouse abuse and the following variables were not
significant: alcohol abuse, family adaptability, physical abuse, sexual abuse,
parental divorce, and current income (p>.05).
Trauma symptomatology scores significantly correlated with witness-
ing family violence (r=.54), neglect (r=.47), and psychological abuse
(r=.40). However, trauma symptomatology was not significantly associ-
ated with alcohol abuse, family cohesion, family adaptability, physical abuse,
sexual abuse, parental divorce, or current income (p>.05). All three of the
outcome variables significantly correlated with each other. Trauma symp-
toms were significantly correlated with psychological spouse abuse (r=.53)
and physical spouse abuse (r=.38). These results indicate moderately large
correlations between the level of trauma symptoms and the levels of both
child maltreatment and perpetration of spouse abuse.
Multiple Regression Analyses to Predict Spouse Abuse and Adjustment
Predicting Physical Spouse Abuse From Alcohol Abuse, Childhood Family
Background, and Child Maltreatment
Standard multiple regression analysis was used to assess the contribu-
tion of alcohol abuse, childhood family adaptability and cohesion, physical
abuse, psychological maltreatment, sexual abuse, neglect, and witnessing
domestic violence during childhood to predict physical spouse abuse (see
Table II). Rfor regression was significantly different from 0, F(8, 27) =
3.47, p<.01. The only independent variable that provided unique signif-
icant contribution to the prediction of physical spouse abuse was neglect
(sr2=.09). Altogether, 51% (36% adjusted) of the variability in physical
spouse abuse was predicted from scores on the eight variables. Physical
spouse abuse was predicted by current alcohol abuse, child maltreatment,
and family dysfunction, but uniquely by childhood neglect.

236 Bevan and Higgins
Table II. Standard Multiple Regression of Alcohol Abuse, Childhood Family Background, and
Child Maltreatment on Physical Spouse Abuse
Physical spouse abuse (ABI)
Independent variables R2Adj. R2βsr2
Alcohol abuse (MAST) −.06
Family cohesion (FACES-II) −.30
Family adaptability (FACES-II) .21
Physical abuse (CCMS-A) .08
Psychological maltreatment (CCMS-A) −.22
Sexual abuse (CCMS-A) .30
Neglect (CCMS-A) .53 .09∗
Witnessing family violence (CCMS-A) .01
All combined together .51∗∗a.36∗∗
aunique variance =.09; shared variance =.42.
∗p<.05. ∗∗ p<.01.
Predicting Psychological Spouse Abuse From Alcohol Abuse, Childhood
Family Background, and Child Maltreatment
Standard multiple regression analysis was used to assess the contribu-
tion of alcohol abuse, childhood family adaptability and cohesion, physical
abuse, psychological maltreatment, sexual abuse, neglect, and witnessing do-
mestic violence during childhood to predict psychological spouse abuse. R
for regression was not significantly different from 0, F(8, 27) =2.16, p>.05.
None of the 10 variables significantly predicted psychological spouse abuse.
Predicting Trauma Symptoms From Alcohol Abuse, Childhood Family
Background, and Child Maltreatment
The contribution of alcohol abuse, childhood family adaptability and co-
hesion, physical abuse, psychological maltreatment, sexual abuse, neglect,
and witnessing domestic violence during childhood to predict the level of
trauma symptomatology was evaluated using standard multiple regression
analysis (see Table III). Rfor regression was significantly different from 0,
F(8, 27) =2.56, p<.05. There was no independent variable that provided
unique prediction of trauma symptomatology. Trauma symptoms were pre-
dicted by the shared influence of all the variables.
Predicting Physical Spouse Abuse, Psychological Spouse Abuse, and
Trauma Symptomatology From the Five Maltreatment Scales
To assess the predictive value of psychological maltreatment, sexual
abuse, and neglect beyond that afforded by physical abuse and witnessing

Is Domestic Violence Learned? 237
Table III. Standard Multiple Regression of Alcohol Abuse, Childhood Family Background,
and Child Maltreatment on Trauma Symptomatology
Psychological spouse abuse (ABI)
Independent variables R2Adj. R2βsr2
Alcohol abuse (MAST) −.01
Family cohesion (FACES-II) −.16
Family adaptability (FACES-II) .22
Physical abuse (CCMS-A) −.24
Psychological maltreatment (CCMS-A) .17
Sexual abuse (CCMS-A) .03
Neglect (CCMS-A) .30
Witnessing family violence (CCMS-A) .44
All combined together .43∗.26∗
∗p<.05.
family violence, hierarchical multiple regression was used. Scores on the
physical abuse and witnessing family violence scales were entered on the
first step. Scores for the psychological maltreatment, sexual abuse, and ne-
glect scales were entered on the second step. Physical spouse abuse, psy-
chological spouse abuse, and trauma symptoms were the dependent vari-
ables (see Table IV). After Step 1, with scores for physical spouse abuse
as the dependent variable, adjusted R2=.00 (R2=.06) was not significant,
F(5, 30) =1.06, p>.05. After Step 2, with psychological maltreatment, sex-
ual abuse, and neglect added to the prediction of physical spouse abuse,
adjusted R2=.29, which was significant, F(5, 30) =3.84, p<.05. Neglect
provided unique prediction (sr2=.09). The addition of the three scales re-
sulted in a significant increment in R2(R2change =.33, Fchange =5.42),
p<.01.
Table IV. Hierarchical Multiple Regression of Child Maltreatment on Physical Spouse Abuse,
Psychological Spouse Abuse, and Trauma Symptomatology
ABI-physical ABI-psychological TSC-40
Independent Adj. R2Adj. R2Adj. R2
variables R2Change βR2Change βR2Change β
Step 1 .00 .06 .16∗.20∗.25∗∗ .29∗∗
Physical abuse .22 −.02 −.07
Witnessing family .05 .46 .57
violence
Step 2 .29∗∗ .33∗∗ .17 .09 .27∗.08
Sexual abuse −.20 .18 .17
Neglect .40 .14 .07
Psychological .47 .24 .28
maltreatment
∗p<.05. ∗∗ p<.01.

238 Bevan and Higgins
Using psychological spouse abuse as the dependent variable, after
Step 1, Adjusted R2=.16, F(5, 30) =4.21, p<.05, which was significant.
Witnessing Family Violence provided unique prediction (sr2=.11). After
Step 2, adjusted R2=.17, F(5, 30) =2.47, p>.05, which was not signifi-
cant. The increment in R2as a result of the addition of the three scales
approached significance (R2change =.09, Fchange =1.24), p=.055.
Finally, with trauma symptomatology as the dependent variable, after
entering physical abuse and witnessing family violence during childhood in
Step 1, adjusted R2=.25, F(5, 30) =6.76, p<.01, which was significant.
Witnessing family violence provided unique prediction (sr2=.25). After
Step 2, adjusted R2=.27, F(5, 30) =3.58, p<.05, which was significant.
The addition of the three scales did not result in a significant increment in
R2(R2change =.08, Fchange =1.32), p>.05. Witnessing family violence
was still the only unique predictor (sr2=.09).
Psychological maltreatment, sexual abuse, and neglect significantly
added to prediction of trauma symptomatology and physical spouse abuse,
but not psychological spouse abuse after accounting for the contribution
of physical abuse and witnessing family violence. With all five maltreat-
ment scales included, witnessing family violence uniquely predicted trauma
symptomatology, whereas neglect uniquely predicted physical spouse abuse.
There was no significant unique contributor to psychological spouse abuse
when all five variables were entered.
DISCUSSION
This study explored the relationships between childhood family func-
tioning, child maltreatment, alcohol abuse, and adult adjustment of domes-
tically violent men (particularly trauma symptomatology and frequency of
physical and psychological spouse abuse). The results supported the first
hypothesis: There were many large intercorrelations between the level of
spouse abuse (physical and psychological), child maltreatment, childhood
family adaptability and cohesion, parental divorce, income, and trauma
symptomatology. In particular, a high degree of overlap was found between
the five maltreatment types, confirming earlier findings that children are of-
ten subjected to more than one maltreatment type (Higgins & McCabe, 1998,
2000b). However, alcohol abuse failed to intercorrelate significantly with
any other risk factor. The second hypotheses was only partially supported:
Physical spouse abuse and trauma symptomatology (but not psychological
spouse abuse) were predicted by the combination of predictor variables.
The third hypothesis was also partly supported: After controlling for phys-
ical abuse and witnessing family violence during childhood, sexual abuse,

Is Domestic Violence Learned? 239
neglect, and psychological maltreatment significantly added to prediction of
physical spouse abuse and trauma symptomatology, but not psychological
spouse abuse. Neglect was found to provide unique prediction of the per-
petration of physical spouse abuse, and witnessing family violence uniquely
predicted both psychological spouse abuse and trauma symptomatology.
The frequency of physical and psychological spouse abuse reported
by respondents was typical of a domestically violent sample (Shepard &
Campbell, 1992). Overall a majority of the sample was raised in families
characterized by low adaptability and cohesion, which is comparable to the
clinical sample described by Harter et al. (1988), who also applied FACES-II
to a clinical sample. Exactly half of the men received scores considered di-
agnostic of alcoholism, which supports previous research that has estimated
that approximately half of all men attending counseling for perpetrating do-
mestic violence abuse alcohol (Tolman & Bennett, 1990). However, alcohol
abuse did not significantly correlate with any other variable in the analyses.
This could be explained by the lack of variability in scores because of the
large number of alcoholic participants.
The participants had high trauma symptom scores, particularly for de-
pression, that were comparably greater than those recorded for a normal
female sample (see Elliott & Briere, 1992), and a self-selected community
sample including both sexes (see Higgins & McCabe, 2000b). Maltreatment
scores were comparable to those scores typical of a self-selected community
sample, except for sexual abuse scores, which were low (Higgins & McCabe,
2001a). These two results are peculiar, given that trauma symptomatology
was measured using a scale specifically designed to assess the traumatic
impact of child abuse. This suggests that the trauma symptomatology expe-
rienced by the sample could be influenced by another unassessed variable,
different to the experience of child maltreatment, or that the respondents
minimized the level of maltreatment experienced during childhood.
The results suggest that rather than physical abuse or witnessing family
violence, it may be other forms of child maltreatment that are important
risk factors for perpetration of domestic violence in adulthood. Although
witnessing family violence uniquely contributed to prediction of psycholog-
ical spouse abuse, it was childhood neglect that accounted for the largest
amount of unique variance in physical spouse abuse scores. Consistent with
the results of Andrews and Brown (1988), neglect was the only factor found
to significantly correlate with all three of the outcome variables, and witness-
ing family violence was found to significantly correlate with psychological
spouse abuse and trauma symptomatology.
The failure of subjection to child physical abuse to correlate with the
level of physical or psychological violence is not consistent with the assump-
tions of learning theory and the research that shows a relationship between

240 Bevan and Higgins
the child physical abuse and adult perpetration of domestic violence (Caesar,
1988; Rosenbaum & O’Leary, 1981). Obviously, this lack of significance could
be due to the low sample size; however, this did not prevent neglect and wit-
nessing family violence from significantly correlating with physical spouse
abuse and psychological spouse abuse respectively. Trauma symptomatology
significantly correlated with all of the maltreatment types except for sexual
abuse supporting similar findings demonstrated by Higgins and McCabe
(1998, 2000b). Although researchers have linked the experience of parental
divorce to poor adult functioning (Amato & Booth, 1991; Garbardi & Rosen,
1992), parental divorce did not significantly correlate with any of the out-
comes for this sample of domestically violent men. Men with a lower income
tended to report a greater level of physical abuse experience during child-
hood and a higher frequency of physical spouse abuse. This supports the
connection made between low socioeconomic status and the perpetration of
spouse abuse found by McKenry et al. (1995).
The significant intercorrelations between maltreatment types indicate
that domestically violent men who experienced one maltreatment type dur-
ing childhood were also likely to have been subjected to other types of child
maltreatment. These results are consistent with the findings of Higgins and
McCabe (1998, 2000a,b), who found evidence for the coexistence of mal-
treatment types among self-selected community samples. The only two mal-
treatment types that did not intercorrelate were witnessing family violence
and sexual abuse, possibly because sexual abuse scores were quite low.
Physical spouse abuse and trauma symptomatology were significantly
predicted by the five maltreatment types: childhood family background,
alcohol abuse, parental divorce, and income. This indicates that it was the
coexistence of risk factors that resulted in a shared relationship with the vio-
lent outcome, not the isolated occurrence of individual factors. Neglect was
the only variable found to share a specific relationship with the perpetration
of physical spouse abuse, supporting previous research that has linked the
experience of childhood neglect to aggression problems (Ammerman et al.,
1986; Reidy, 1977). Contrary to the findings of Higgins and McCabe (2000b),
family adaptability and cohesion did not uniquely predict levels of trauma
symptomatology. However, Higgins and McCabe used a self-selected com-
munity sample whose vulnerability to the traumatic influence of a negative
childhood environment may differ to a domestically violent sample. The
failure of the variables to predict, either collectively or individually, psy-
chological spouse abuse suggests psychological abuse could be predicted by
variables other than those tested in this study.
The inadequacy of a social learning model of domestic violence and
the importance of testing for both the shared and individual impact of all
five maltreatment types on the adult adjustment of domestically violent men

Is Domestic Violence Learned? 241
was demonstrated. Hierarchical multiple regression analysis revealed that
physical abuse and witnessing family violence significantly predicted psy-
chological spouse abuse and trauma symptomatology, with witnessing fam-
ily violence individually predicting the two outcomes. Once the influence
of physical abuse and witnessing family violence had been controlled for,
neglect, sexual abuse, and psychological abuse significantly predicted phys-
ical spouse abuse, which suggests that these three maltreatment types pro-
vide a large degree of predictive influence on the frequency of adult male
perpetration of domestic violence. In particular, neglect uniquely predicted
physical spouse abuse. This finding runs contrary to learning theory’s basic
assumptions and implies that adult male perpetration of domestic violence
involves far more complex processes and interactions than simply modelling
and reinforcement.
The major implications of this study are twofold. Firstly, the many large
intercorrelations found between risk factors coupled with the significant
shared prediction of risk factors for the frequency of physical spouse abuse
suggest that individual links made in previous domestic violence research,
between the perpetration of male spouse abuse and its supposed risk factors,
may be inaccurate (e.g., McKenry et al., 1995; Tolman & Bennett, 1990). It is
the coexistence of risk factors that results in a shared relationship with the vi-
olent outcome, not the isolated occurrence of individual factors. The second
implication involves the finding that sexual abuse, neglect, and psychological
maltreatment contributed additionally to the prediction of physical spouse
abuse beyond that afforded by physical abuse and witnessing family vio-
lence during childhood. This suggests that previous learning theory research
that has linked the perpetration of domestic violence to the childhood expe-
rience of physical abuse and witnessing family violence (e.g., Caesar, 1988;
Kalmuss, 1984; Marshall & Rose, 1988; Rosenbaum & O’Leary, 1981) has not
accounted for the powerful influence of coexisting maltreatment types, par-
ticularly neglect, on the perpetration of physical spouse abuse. Although this
finding does not render learning theory’s explanation of domestic violence
false, it does highlight its inadequacy as an appropriate conceptualization
of the phenomenon. An ecological perspective, however, which incorpo-
rates the contribution of various factors and evaluates both their specific
and shared associations with the violent outcome, offers a far more useful
explanation of male perpetration of domestic violence.
Limitations and Directions for Further Research
A key limitation of this study was the small sample size, which may have
compromised the statistical power of the analyses. Therefore, any significant

242 Bevan and Higgins
results should be interpreted with caution. A further limitation was the use
of retrospective self-reports. Additionally, men’s reported frequency of per-
petration of domestic violence was not corroborated with other sources.
This could have affected the results, given the tendency for domestically vi-
olent men to underreport their abuse (Melvin et al., 1999). Similarly, eight
of the men had attended counseling prior to the 6-month period for which
they were asked to report the frequency of violence they perpetrated. Thus,
prior counseling may have reduced the level of violence reported. The study
also used a volunteer, clinical sample, which reduced the generalizability
of results to other domestically violent samples, and included unsubstan-
tiated reports of childhood experiences of maltreating behaviors that may
not have been accurate. The final limitation rests with the inadequacy of
the measures used. The ABI does not record the severity of domestic vio-
lence perpetrated, and similarly, the CCMS-A does not distinguish between
“abused” and “nonabused.”
Accordingly, further research with a larger, nonclinical, sample using
longitudinal or prospective methodology, to verify these results, would be
beneficial. Data were provided here on predicting the level of violence within
a known violent sample; predictors of who perpetrates violence (and who
does not) may be different, and need to be examined. The interrelationships
of maltreatment types and the impact of childhood neglect on the level
of adult perpetration of domestic violence require additional evaluation.
Possibilities for further research are also the incorporation and evaluation
of the influence of coexisting protective and risk factors, which may act to
mediate and moderate the adult adjustment of domestically violent men.
Lastly, more research is required to establish the predictive factors for the
perpetration of psychological spouse abuse.
This study provided evidence for the contention that factors associ-
ated with poor adult adjustment, and the perpetration of domestic vio-
lence, should not just be studied singly, as though they occurred in isolation,
but together, to establish their shared influence on outcomes. Intercorre-
lations between variables were demonstrated, particularly among the five
maltreatment types. Additionally, the collective influence of child maltreat-
ment, childhood family background, and alcohol abuse were found to sig-
nificantly predict physical spouse abuse and trauma symptomatology. This
demonstrated the collective influence of variables on the adult adjustment
of domestically violent men. Furthermore, the importance of examining the
influence of all five maltreatment types on later perpetration of spouse abuse
was shown. Assessing the additional contribution of sexual abuse, psycho-
logical maltreatment, and neglect, after controlling the influence of physical
abuse and witnessing family violence, not only revealed a significant incre-
ment in the prediction of physical spouse abuse, but highlighted the specific

Is Domestic Violence Learned? 243
relationship between neglect and the violent outcome. As well as providing
some initial evidence for the inadequacy of explaining the perpetration of
domestic violence simply in terms of learning theory, it provides support
for an ecological approach to the domain of domestic violence. Assess-
ment of the influence of a wider range of factors on adult outcomes resulted
in a more holistic conceptualization of the childhood histories of domestic
violence perpetrators.
REFERENCES
Amato, P. R., and Booth, A. (1991). Consequences of parental divorce and marital unhappiness
for adult well-being. Soc. Forces 69: 895–914.
Ammerman, R. T., Cassisi, J. E., Hersen, M., and Van Hasselt, V. B. (1986). Consequences of
physical abuse and neglect in children. Clin. Psychol. Rev. 6: 291–310.
Anderson, K. L. (1997). Gender, status, and domestic violence: An integration of feminist and
family violence approaches. J. Marr. Fam. 59: 655–669.
Andrews, B., and Brown, G. W. (1988). Marital violence in the community: A biographical
approach. Br. J. Psychiatry 153: 305–312.
Bandura, A. (1973). Aggression: A Social Learning Analysis, Prentice-Hall, Englewood Cliffs,
NJ.
Bernstein, D. P., Fink, L., Handelsman, L., Foote, J., Lovejoy, M., Wenzel, K., Sapareto, I., and
Ruggiero, J. (1994). Initial reliability and validity of a new retrospective measure of child
abuse and neglect. Am. J. Psychiatry 151: 1132–1136.
Briere, J., and Runtz, M. (1988). Multivariate correlates of childhood psychological and physical
maltreatment among university women. Child Abuse Negl. 12: 331–341.
Briere, J., and Runtz, M. (1989). The Trauma Symptom Checklist (TSC-33): Early data on a
new scale. J. Interpers. Viol. 4: 151–163.
Briere, J., and Runtz, M. (1990). Differential adult symptomatology associated with three types
of child abuse histories. Child Abuse Negl. 14: 357–364.
Caesar, P. L. (1988). Exposure to violence in the families-of-origin among wife-abusers and
maritally nonviolent men. Violence Vict. 3: 49–63.
Cappell, C., and Heiner, B. R. (1990). The intergenerational transmission of family aggression.
J. Fam. Viol. 5: 135–152.
Carlson, B. E. (1997). Mental retardation and domestic violence: An ecological approach to
intervention. Soc. Work 42: 79–89.
Doumas, D., Margolin, G., and John R. S. (1994). The intergenerational transmission of aggres-
sion across three generations. J. Fam. Viol. 9: 157–175.
Duncan, R. D., Saunders, B. E., Kilpatrick, D. G., Hanson, R. F., and Resnick, H. S. (1996).
Childhood physical assault as a risk factor for PTSD, depression, and substance abuse:
Findings from a national survey. Am. J. Orthopsychiatry 66: 437–448.
Elliott, D. M., and Briere, J. (1992). Sexual abuse trauma among profession women: Validating
the Trauma Symptom Checklist – 40 (TSC-40). Child Abuse Negl. 16: 391–398.
Forsstrom-Cohen, B., and Rosenbaum, A. (1985). The effects of parental marital violence on
young adults: An exploratory investigation. J. Marr. Fam. 47: 467–472.
Fraser, M. W. (1996). Aggressive behavior in childhood and early adolescence: An ecological–
developmental perspective on youth violence. Soc. Work 41: 347–361.
Fromuth, M. E. (1986). The relationship of childhood sexual abuse with later psychological and
sexual adjustment in a sample of college women. Child Abuse Negl. 10: 5–15.
Garbardi, L., and Rosen, L. A. (1992). Intimate relationships: College students from divorces
and intact families. J. Divorce Remarr. 18: 25–56.
Gelles, R. (1972). The Violent Home: A Study of Physical Aggression Between Husbands and
Wives, Sage, Newbury Park, CA.

244 Bevan and Higgins
Hart, S. N., and Brassard, M. R. (1987). A major threat to children’s mental health: Psychological
maltreatment. Am. Psychol. 42: 160–165.
Harter, S., Alexander, P. C., and Neimeyer, R. A. (1988). Long-term effects of incestuous child
abuse in college women: Social adjustment, social cognition, and family characteristics.
J. Consult. Clin. Psychol. 56: 5–8.
Hedlund, J. L., and Vieweg, B. W. (1984). The Michigan Alcoholism Screening Test (MAST):
A comprehensive review. J. Oper. Psychiatry 15: 55–64.
Henning, K., Leitenberg, H., Coffey, P., Bennett, T., and Jankowski, M. K. (1997). Long-term
psychological adjustment to witnessing interparental physical conflict during childhood.
Child Abuse Negl. 21: 501–515.
Higgins, D. J., and McCabe, M. P. (1994). The relationship of child sexual abuse and family
violence to adult adjustment: Toward an integrated risk-sequelae model. J. Sex Res. 31:
255–266.
Higgins, D. J., and McCabe, M. P. (1998). Parent perceptions of maltreatment and adjustment
in children. J. Fam. Stud. 4: 53–76.
Higgins, D. J., and McCabe, M. P. (2000a). Multi-type maltreatment and the long-term adjust-
ment of adults. Child Abuse Rev. 9: 6–18.
Higgins, D. J., and McCabe, M. P. (2000b). Relationships between different types of mal-
treatment during childhood and adjustment in adulthood. Child Maltreat. 5: 261–
272.
Higgins, D. J., and McCabe, M. P. (2001a). Multiple forms of child abuse and neglect: Adult
retrospective reports. Aggress. Viol. Behav. Rev. J. 6: 547–578.
Higgins, D. J., and McCabe, M. P. (2001b). The development of the Comprehensive Child
Maltreatment Scale. J. Fam. Stud. 7: 7–28.
Hotaling, G. T., and Sugarman, D. B. (1986). An analysis of risk markers in husband to wife
violence: The current state of knowledge. Violence Vict. 1: 101–124.
Kalmuss, D. (1984). The intergenerational transmission of marital aggression. J. Marr. Fam. 46:
11–19.
Kinard, E. M. (1982). Experiencing child abuse: Effects on emotional adjustment. Am. J.
Orthopsychiatry 52: 82–91.
Kinard, E. M. (1994). Methodological issues and practical problems in conducting research on
maltreated children. Child Abuse Negl. 18: 654–656.
Leonard, K. E., and Blane, H. T. (1992). Alcohol and marital aggression in a national sample
of young men. J. Interpers. Viol. 7: 19–30.
MacEwen, K. E. (1994). Refining the intergenerational transmission hypothesis. J. Interpers.
Viol. 9: 350–365.
Maker, A. H., Kemmelmeier, M., and Peterson, C. (1998). Long-term psychological conse-
quences in women witnessing parental physical conflict and experiencing abuse in child-
hood. J. Interpers. Viol. 13: 574–588.
Malinosky-Rummell, R., and Hansen, D. J. (1993). Long-term consequences of childhood phys-
ical abuse. Psychol. Bull. 114: 68–79.
Marshall, L., and Rose, P. (1988). Family-of-origin violence and courtship abuse. J. Counsel.
Dev. 66: 414–418.
Mathias, J. L., Mertin, P., and Murray, A. (1995). The psychological functioning of children from
backgrounds of domestic violence. Aust. Psychol. 30: 47–54.
McKenry, P. C., Julian, T. W., and Gavazzi, S. M. (1995). Toward a biopsychosocial model of
domestic violence. J. Marr. Fam. 57: 307–320.
Melvin, T., Muller, D., Chapman, A., Shine, R., and Edwards, R. (1999). A study in hope:
The multi-site evaluation and development of a better practice model for family violence
services, Relationships Australia, Melbourne (Report to Victorian Department of Human
Services).
Mullen, P. E., and Fleming, J. (1998). Long-Term Effects of Child Sexual Abuse (Issues in Child
Abuse Prevention, Vol. 9), Australian Institute of Family Studies, Melbourne.
Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., and Herbison, P. S.(1993). Childhood
sexual abuse and mental health in adult life. Br. J. Psychiatry 163: 721–732.

Is Domestic Violence Learned? 245
Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., and Herbison, P. S. (1994). The
effect of child sexual abuse on social, interpersonal and sexual function in adult life. Br. J.
Psychiatry 165: 35–47.
Ney, P. G., Fung, T., and Wickett, A. R. (1994). The worst combinations of child abuse and
neglect. Child Abuse Negl. 18: 705–714.
Olson, D. H., Bell, R. Q., and Portner, J. (1982). FACES II: Family Adaptability and Cohesion
Evaluation Scales, Family Social Science, St. Paul, MN.
Olson, D. H., Russell, C. S., and Sprenkle, D. H. (1983). Circumplex model of marital and family
systems: VI. Theoretical update. Fam. Process 22: 69–83.
Parkinson, P., and Humphreys, C. (1998). Children who witness domestic violence—the impli-
cations for child protection. Child Fam. Law Q. 10: 147–159.
Reidy, T. J. (1977). The aggressive characteristics of abused and neglected children. J. Clin.
Psychol. 33: 1140–1145.
Rosenbaum, A., and O’Leary, K. D. (1981). Children: The unintended victims of marital vio-
lence. Am. J. Orthopsychiatry 51: 692–699.
Rosenberg, M. S. (1987). New directions for research on the psychological maltreatment of
children. Am. Psychol. 42: 166–171.
Schuerger, J. M., and Reigle, N. (1988). Personality and biographic data that characterize men
who abuse their wives. J. Clin. Psychol. 44: 75–81.
Selzer, M. L. (1971). The Michigan Alcoholism Screening Test: The quest for a new diagnostic
instrument. Am. J. Psychiatry 127: 1653–1658.
Shepard, M. F., and Campbell, J. C. (1992). The Abusive Behavior Inventory: A measure of
psychological and physical abuse. J. Interpers. Violence 7: 291–305.
Simons, R. L., Lin, K. H., and Gordon, L. C. (1998). Socialization in the family of origin and
male dating violence: A prospective study. J. Marr. Fam. 60: 467–478.
Tolman, R., and Bennett, L. (1990). A review of quantitative research on men who batter.
J. Interpers. Viol. 5: 87–118.
Tomison, A. M. (1995). Spotlight on Child Neglect (Issues in Child Abuse Prevention, Vol. 4),
Australian Institute of Family Studies, Melbourne.
Urquiza, A. J. (1991). Retrospective methodology in family violence research: Our duty to
report past abuse. J. Interpers. Viol. 6: 119–126.
Mihalic, S. W., and Elliott, D. (1997). A social learning theory model of marital violence. J. Fam.
Viol. 12: 21–47.