MD1

An Overview of the Nature, Causes, and Consequences of Abusive Family Relationships

Toward Differentiating Maltreatment and Violence

Robert E. E m e r y and Lisa Laumann-Billings University o f Virginia

Despite gaps in research, this review argues f o r distin- guishing family maltreatment from family violence, a differentiation between minimal or moderate abuse and serious endangerment, physical injury, or sexual viola- tion. Most acts now defined as violent or abusive are moderate, and stressful life circumstances contribute to their development. Research suggests that abuse may develop through multiple pathways. The consequences o f abuse appear to be general, but more commonalities may be f o u n d by examining more subtle and complex effects. Many interventions have attempted to prevent or treat family violence and its consequences, and several show promise. More systematic research is needed, but several changes in intervention seem warranted now. For instance, (a) mental health professionals should not be required to report maltreatment while a family is engaged in therapy; (b) the social service system needs to redis- cover its roots in supporting families under stress, includ- ing in cases o f moderate maltreatment; and (c) swift and decisive legal intervention is needed in cases o f serious family violence.

F r o m the vantage point o f the present, the public outcry against family violence is shockingly recent. The Society for the Prevention o f Cruelty to Chil- dren was formed only in 1875 (nine years after the forma- tion o f the Society for the Prevention o f Cruelty to Ani- mals), and the arousal o f professional concern about bat- tered children dates only to 1962 with the publication o f K e m p e ' s seminal article (Kempe, Silverman, Steele, Broegemueller, & Silver, 1962; more pervasive profes- sional concern is far more recent.) Spousal assault was made a crime in all states only in recent decades (Harvard Law Review, 1993), and there still are no laws against spousal rape in some states (Dietz, 1996). Professional concern about the children o f battered women is less than two decades old, and sibling and elder abuse are just now beginning to receive clinical and research attention.

We are concerned about the historical inattention to family violence, but we also are troubled by the potential for overreaching in defining family violence, misinter- preting its putative consequences, or intervening broadly

without knowledge o f the risks and benefits o f interven- tion. In fact, a number o f commentators have suggested that some attempts to alleviate the problem of, or the problems created by, family violence m ay actually have iatrogenic effects (Besharov, 1988; Loftus, 1994; Rep- pucci & Haugaard, 1989).

In this article, we attempt to strike a moderate and empirically based course in considering the definition and epidemiology o f family violence and its causes, its consequences, and appropriate intervention. We focus considerable attention on child abuse but extend our dis- cussion to include spousal, sibling, and elder abuse when- ever possible, because the causes and consequences o f family violence are found throughout the entire family system.

As a similar review concluded almost a decade ago, family violence has defied easy description, understand- ing, or amelioration (Emery, 1989). Nevertheless, an overriding theme o f the present overview is the need to make the difficult distinction between levels o f abuse, particularly for purposes o f intervention. We argue for a basic differentiation between family maltreatment, char- acterized b y minimal physical or sexual harm or endan- germent, and family violence, characterized b y serious physical injury, profound psychological trauma, or sexual violation. We recognize that such a categorization re- quires some arbitrary discriminations and that all levels o f abuse can have devastating psychological conse- quences. However, distinguishing between levels o f abuse should move us toward making a much-needed choice between when to offer families supportive interventions and when to use adversary or coercive interventions. Too often psychologists and social service agencies find them-

Robert E. Emery and Lisa Laumann-Billings, Department of Psychol- ogy, University of Virginia.

We would like to acknowledge the support of the William T. Grant Foundation and the University of Virginia Center for Children, Families, and the Law in the preparation of this article. We would also like to thank Mary Jo Coiro, Allyson Crawford, and Frank Fincham for their comments on drafts of this article.

Correspondence concerning this article should be addressed to Robert E. Emery, Department of Psychology, Gilmer Hall, University of Virginia, Charlottesville, VA 22903-2477. Electronic mail may be sent to [email protected].

February 1998 • American Psychologist Copyright 1998 by the American Psychological Association, Inc. 0003-066X/98/$2.00 Vol. 53, No. 2, 121-135

121

Robert E. Emery Photo by Jim Carpenter

selves policing families who might benefit from support and supporting families who need policing.

Definitions and Epidemiology Scientific and legal experts disagree strongly about h o w - - o r how broadly--to define abuse and neglect (e.g., Finkelhor & Dziuba-Leatherman, 1994; Straus & Gelles, 1990; Wald, 1991), and statutory definitions, in- vestigation policies, and data collection procedures differ from state to state (Lung & Daro, 1996). The underlying problem is that our conceptualization of violence or abuse is inherently driven by social judgment and n o t by immutable social standards or by empirical science (Emery, 1989). Moreover, definitions have important and varied implications for different purposes. For instance, researchers attempting to assess the extent and nature of aggressive or violent activity within American families (e.g., Gallup, Moore, & Schussel, 1995; Straus & Gelles, 1990) have very different alms in defining violence or abuse than do child protective workers who must decide when coercive intervention is justified (Besharov, 1988).

These observations stress the need for clear and pre- cise operational definitions of violence in specific re- search or applied contexts, but it also is critical to develop consensual definitions. Offering a useful analogy, Cic- chetti and Barnett (1991) suggested that consensus defi- nitions may produce benefits akin to those stemming from the American Psychiatric Association's (1994) Di- agnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV). In fact, a task force of the National Re- search Council (NRC, 1993) recently recommended es- tablishing consensus definitions across the four areas of child maltreatment (physical abuse, sexual abuse, physi-

cal neglect, and emotional maltreatment). We would go a step further and argue for dividing all forms of family abuse into categories of maltreatment versus violence. As we have noted and discuss in more detail below, one rationale for this distinction is to differentiate between approaches to intervention. A second rationale becomes evident when we carefully examine the epidemiology of family violence. In fact, most acts designated as violent or abusive in leading epidemiological studies involve lit- tle or no physical injury or sexual violation.

Definitions of Violence in Epidemiological Surveys Even given the most conservative definitions, violence within the family is disturbingly common. Young children and women are especially vulnerable to severe or fatal injury at the hands of a family member. Homicide is one of the five leading causes of death for children under 12 years of age (Federal Bureau of Investigation, 1994), with over half of these homicides being perpetrated by a family member. An estimated 1,200 to 1,500 children are killed each year by their parent or parent figure (Lung & Daro, 1996; Sedlak & Broadhurst, 1996); 85% of these children are under the age of five years (Lung & Daro, 1996). Of the 4,967 women murdered in 1995, almost one third were killed by a boyfriend or husband (Federal Bureau of Investigation, 1996). In fact, wife battering is the leading cause of injury to women between the ages of 15 and 44 years, accounting for more injuries than car accidents, muggings, and stranger rapes combined (Federal Bureau of Investigation, 1991).

It is important to note, however, that extreme in- stances of child and spouse abuse make up only a minor- ity of identified cases in epidemiological studies. The Third National Incidence Study of Child Abuse and Ne- glect (NIS-3; Sedlak & Broadhurst, 1996), for instance, estimated that between 20% and 35% of children identi- fied as maltreated suffered a serious injury, defined as "long-term impairment of physical, mental, or emotional capacities or requiring professional treatment aimed at preventing such long-term impairment" (chap. 3, p. 13). Examples of such injuries included loss of consciousness, broken bones, third-degree bums, and schooling loss that required special-education services (Sedlak & Broad- hurst, 1996, chap. 3, p. 13). Other identified, moderate, and inferred cases included educational neglect, where parents failed to ensure children's attendance at school (approximately 25% of cases); moderate physical abuse, where a parent's punishment left the child with an observ- able bruise, pain, or emotional distress for up to 48 hours but did not require professional treatment (approximately 20% of cases); emotional abuse, where a parent habitu- ally belittled, denigrated, or engaged in overtly hostile or rejecting treatment (approximately 10% of cases); in- ferred sexual abuse, where there was some evidence to suggest that the child may have been sexually touched or fondled (approximately 7%); and moderate physical neglect, where a parent was judged to be providing inade- quately for his or her child, even given the conditions of

122 February 1998 • American Psychologist

Usa Laumann- Billings Photo by Jim Carpenter

poverty, inadequate housing, and chaotic neighborhoods in which most families determined to be neglectful reside (approximately 8% of cases; Coulton, Korbin, Su, & Chow, 1995; Garbarino & Kostelny, 1992; Pelton, 1994).

Many estimates of spouse abuse similarly include acts of aggression that cause relatively minor physical harm. Up to one third of abusive married couples are so defined by acts of pushing, throwing objects, or grabbing at each other, whereas about 4% of couples engage in serious acts of violence such as choking, beating up, or kicking a partner (O'Leary et al., 1989; Straus & Gelles, 1990). Finally, some experts have asserted that 80% of children are victims of sibling v i o l e n c e (Finkelhor & Dziuba-Leatherman, 1994); however, such definitions of sibling violence include pushing, grabbing, or hitting sib- lings (Straus & Gelles, 1990). With exceptions, pushing, grabbing, and hitting between siblings, especially younger ones, surely is seen as normal if inappropriate behavior rather than violence by most parents and professionals, including ourselves. The Epidemiology of Child Abuse: How Samples and Methods Affect Estimates We consider epidemiological studies of child abuse in some detail to further demonstrate how research methods can dramatically affect estimates of the epidemiology of family violence. There have been three primary sources for studying rates of child abuse: (a) official statistics, like the National Committee to Prevent Child Abuse and Neglect's (NCPCA) aggregation of state Child Protective Service (CPS) reports across the 50 states (Lung & Daro, 1996)~; (b) information gathered from professionals who come into contact with victims of family violence, like that used in the three National Incidence Studies of Child

Abuse and Neglect (Sedlak, 1981, 1991; Sedlak & Broadhurst, 1996); and (c) surveys of discipline and con- flict resolution methods among population samples of parents and children (e.g., Gallup et al., 1995; Straus & Gelles, 1990). These three types of studies yield widely divergent results, yet each contributes important informa- tion to our understanding of the epidemiology of family maltreatment and violence (see Table 1).

A comparison of the studies illustrates several key problems in determining prevalence rates. The NCPCA's 50-state survey estimates that approximately 3 million children were reported to CPS in 1995, about one third of whom (996,000) were s u b s t a n t i a t e d 2 as victims of maltreatment. Contrary to popular belief, the large major- ity of the cases officially substantiated were of physical neglect (54%), the form of maltreatment most clearly associated with poverty (Pelton, 1992). About one quarter of the substantiated cases were of physical abuse, 11% were primarily sexual abuse, 3% were emotional neglect, and 6% did not clearly fall into any of these categories (Lung & Daro, 1996).

Because not all cases of maltreatment are reported, the National Incidence Studies of Child Abuse and Ne- glect (Sedlak, 1981, 1991; Sedlak & Broadhurst, 1996) have surveyed professionals who potentially come into contact with maltreated children in addition to using state CPS reports. Not surprisingly, this method leads to sub- stantially higher incidence estimates. For example, NIS-3 (Sedlak & Broadhurst, 1996) estimated that be- tween 1.5 and 2.8 million children were maltreated in 1993. The lower figure is obtained when abuse is defined as observable injuries that last at least 48 hours; the higher figure is obtained when abuse is defined as sub- stantial risk of endangerment. These estimates translate into annual incidence rates of 23 to 42 per 1,000 children for the NIS-3 (Sedlak & Broadhurst, 1996) versus 15 per 1,000 children for the NCPCA (Lung & Daro, 1996; see Table 1).

Population surveys yield even higher estimates of the incidence of family violence, presumably partly be- cause much family violence remains behind closed doors. A Gallup poll of a random, representative sample of 1,000 families across the United States found that ap- proximately 5% of children, or 3 million, met the poll's criteria for physical abuse in the previous year (Gallup et al., 1995). Approximately 2% of American children

I The National Center on Child A b u s e and Neglect ( N C C A N ) compiles a similar database, the National A b u s e and Neglect Data System, w h i c h yields very similar e s t i m a t e s ( N C C A N , 1996).

2 T h e t e r m substantiated is often misunderstood. Child a b u s e is substantiated w h e n a child protective services worker d e t e r m i n e s there is sufficient probability o f abuse. T h e procedures and degree o f certainty for substantiating abuse v a r y a c r o s s agencies and states, b u t a determi- nation is to be m a d e based o n the " p r e p o n d e r a n c e o f e v i d e n c e " in m a n y states. T h u s , the threshold and administrative procedures for s u b - stantiating a b u s e are considerably less stringent and formal t h a n in a criminal trial, and they often are a s o u r c e o f controversy ( F i n c h a m , Beach, Moore, & Diener, 1994).

February 1998 • American Psychologist 123

Table 1 Estimated Annual Incidence of Child Maltreatment Across the Three Types of Studies

NCPCA 50-state NIS-3 Gallup pall

Type of abuse No. of children Rate per 1,000 No. of children Rate per 1,000 No. of children Rate per 1,000

Fatalities Physical abuse

Moderate Severe

Sexual abuse Inferred Moderate Severe

Physical neglect Inferred Moderate Severe

Emotional neglect or abuse Educational neglect Specific acts from Gallup poll

Aggressive acts Spanked on bottom,

slapped hand, shook lightly

Hit on boltom with object, slapped face, pinched

Abusive acts Hit with object on other

than bottom Threw or knocked child

down, shook child (under age two)

Hit child with fist, kicked child hard (under age two)

Beat up child, hit as hard as you could

Choked

1,215 .018 1,500 .02 249,000 382,000 5.7

332,000 50,000

110,000 218,000 3.2 115,000 26,000 74,000

540,000 339,000 5.0 67,000 51,000

217,000 29,880 417,000

397,000

To~I 996,000 15.0 1,554,000 23.1

3,000,000 49 2,950,000 44

335,000 5 1,273,000 19

41,540,000 620

38,190,000 570

17,420,000 260

2,500,000 38

470,000 7

270,000 4

200,000 3 0; 7 ever ° 4,273,000 b

Note. NCPCA 50-state = National Committee to Prevent Child Abuse and Neglect, annual 50-state suwey (Lung & Dora, 1996); NIS-3 = Third National Incidence Study of Child Abuse and Neglect (Sedlak & Broadhurst, 1996); Gallup pall JGallup, Moore, & Schussel, 1995). The NIS-3 severe abuse category required immediate medical attention to prevent long-term impairment; the Gallup paws definition rests on the behaviors listed below the category, not on the consequences. The Gallup pall assessed only physical and sexual abuse, not physical or emotional neglect. Empty cells indicate that the data were not obtained for the particular category in the study listed. a NO incidents of choking were reported when parents were asked about the past year, but seven incidents of choking and one incident of burning were reported when parents were asked if they had ever done these acts to their children, b This total includes only an estimate of the number of children who were physically or sexually maltrealed, because the Gallup pall did not assess for physical or emotional neglect.

were determined to be sexually a b u s e d ) These rates o f child physical abuse and sexual abuse are 10 to 16 times higher than comparable figures for officially reported abuse. Questions concerning physical neglect and emo- tional maltreatment were not asked in this survey.

O f considerable importance, the Gallup poll and other population surveys (Straus & Gelles, 1990) include ratings o f very specific acts that further illustrate how the definition o f violence affects estimates o f prevalence

rates (Haugaard & Emery, 1989). For instance, i f spank- ing is considered to be violent (e.g., Finkelhor & Dziuba- Leatherman, 1994), then an estimated two thirds o f chil- dren in the United States are victims o f abuse. When

S i n the Gallup poll (Gallup et al., 1995), physical a b u s e was defined as being hit o n part o f the b o d y other t h a n the b o t t o m with an object. Sexual a b u s e w a s defined as being forced to touch an adult or older child or being forcibly touched in a sexual way.

124 February 1998 • American Psychologist

violence is defined as hitting with an object, as was the case in the Gallup poll, the rate drops to 5%. Naturally, rates become even lower when physical child abuse is defined by increasingly severe actions (see Table 1).

We do not encourage spanking or condone physical punishments with an object. We do suggest, however, that the term violence (or abuse) connotes more serious physical harm or intent, yet mild to moderate acts account for the majority of cases of physical abuse. We further suggest that, although undesirable, a certain degree of aggression in family relationships is normative--at least statistically. In fact, we view much of the task of control- ling family violence, including child abuse, as helping people to learn to control normal angry and perhaps ag- gressive impulses toward family members and providing them with the supports to cope better with the stresses of family life. Finally, most parents currently labeled as abusive are under a great deal of stress, as we discuss in detail shortly, and they are not endangering their chil- dren' s long-term physical health and safety. Such families are more likely to benefit from interventions designed to support them through the challenges of parenting than from interventions that first label them as abusive.

Dish'nguishing levels of abuse. We recognize that there is not a distinct line between what we refer to as maltreatment versus violence and that any specific definitions immediately create controversy. For these rea- sons, we offer no preferred definitions in this article. Instead, we note that a handful of states, including Vir- ginia, Missouri, Florida, Iowa, and Kentucky, are experi- menting with dividing their child protective caseload and directing less serious cases out of the usual investigation protocol (Community Partnerships for Protecting Chil- dren, 1996). In Virginia's experimental program, for in- stance, severe cases undergo traditional CPS investiga- tion, whereas other cases receive a nonadversarial family assessment that ends with referrals to appropriate agen- cies on a voluntary, not a mandatory, basis. In Virginia, serious cases include sexual abuse; child fatality; allega- tions of abuse against a child in the custody of the state or in a child-care facility; children hospitalized because of suspected abuse or neglect; nonorganic failure to thrive; shaken baby syndrome; multiple or old injuries characteristic of battered child syndrome; Munchausen by proxy syndrome; and abuse or neglect resulting in disfigurement, fracture, severe burn or laceration, mutila- tion, maiming, forced ingestion of a dangerous substance, or life-threatening internal injuries. All other cases--the majority of CPS referrals--are offered assessment and referral.

Is family violence increasing? Dramatic in- creases in reports of family violence in recent years raise another important issue about definition and epidemiol- ogy" To what extent does the increasing number of reports of child abuse reflect not an actual increase in abuse but greater awareness of abuse and a lower threshold in the popular perception of what is abusive? In 1976, 669,000 cases of child abuse were reported to social service agen-

cies, but that number grew steadily to 3,111,000 reports in 1995 (Besharov & Laumann, 1996). Much of the in- creased reporting clearly was due to increased recogni- tion of child abuse and not to actual increases in family violence. Increased recognition also may account for the rising number of reports of child sexual abuse in recent years (NRC, 1993). In 1986, approximately 140,000 chil- dren were reported for child sexual abuse; by 1991, this number had more than tripled to 425,000 children, but sexual abuse reports declined steadily in the years subse- quent to 1991 to just over 300,000 in 1995 (Lung & Daro, 1996). Similarly, the increased reporting of elder abuse likely is at least partially due to increased aware- ness. Since elder abuse was first publicly recognized about a decade ago, the number of reports has more than doubled, from 117,000 in 1986 to 241,000 in 1994 (National Center on Elder Abuse, 1997).

The lack of public awareness of family violence clearly was a problem in many respects, but an opposite set of problems has been created by increased awareness and the dramatic increases in reporting in recent years. Social service agencies are so overwhelmed with investi- gating reports of abuse, most of which go unsubstanti- ated, that they are left with few resources to offer sup- portive interventions (U.S. Advisory Board on Child Abuse and Neglect, 1993). In fact, as policing families for abuse has increasingly dominated social service agencies since the 1970s, the philosophy of offering support to families in need has been lost (Besharov & Laumann, 1997). Thus, another reason to differentiate maltreatment from violence is the potential to reduce the resources devoted to adversary investigation and to redirect efforts into supportive family interventions.

As a final note, however, we do want to point out that it is likely that there has been some real increase in child abuse in recent years. Although the number of moderate cases of child abuse remained stable between 1986 and 1993, the number of serious cases in the NIS-3 quadrupled from 142,000 to 565,000 (Sedlak & Broadhurst, 1996). If the increase was solely due to in- creased awareness or "definitional creep" (Besharov, 1996a), one would expect increases across all levels of severity. Researchers who believe that serious abuse is rising attribute the growing problem to increased use of illegal drugs, greater poverty, increased overall violence in the United States, and the disintegration of communi- ties (Garbarino, 1995; Lung & Daro, 1996; Sedlak & Broadhurst, 1996).

Comorbidily of different forms of abuse. As a closing point about epidemiology, it is important to call attention to the comorbidity between different forms of abuse in the family. Large-scale studies typically have not examined this issue, but an overlap in abuse across family relationships, particularly between spouse abuse and child abuse, commonly is found in smaller scale studies. In fact, 40% to 75% of children exposed to mari- tal violence are estimated to be victims of physical child abuse also (Layzer, Goodson, & DeLange, 1986; Straus,

February 1998 • American Psychologist 125

Gelles, & Steinmetz, 1980), and children living with an abused mother have been found to be 12 to 14 times more likely to be sexually abused (McCloskey, Fi- guerdo, & Koss, 1995). Because of this comorbidity and because violence affects other family members who are not victims or perpetrators, it is essential to conceptualize the causes and consequences of violence in terms of the entire family system, as has been discussed elsewhere (Emery, 1989). The need for an even broader ecological perspective is suggested by the fact that violent families tend to live in communities with higher rates of commu- nity violence (Bell & Jenkins, 1993; Garbarino & Kos- telny, 1992).

The Development of Abusive Relationships Many factors contribute to the development of family violence, including individual personality factors, family interaction patterns, poverty and social disorganization, acute stressors, and the cultural context in which the family lives (Belsky, 1993; Bronfenbrenner, 1979; Cic- chetti & Toth, 1995; Vondra, 1990). Consistent with their comorbidity, a number of the same factors increase the risk for child and spouse abuse, which are the types of abuse that are the focus of our overview. We do note where research indicates differences in the etiology of different forms of abuse; however, too little evidence is available on elder or sibling abuse to consider these topics adequately. Ecological Approach to Understanding Family Violence Bronfenbrenner's (1979) ecological model often has been used to integrate research on multiple risks for family violence at four levels of analysis: (a) individual charac- teristics, (b) the immediate social context, (c) the broader ecological context, and (d) the societal and cultural con- text. We briefly review research on each of these broad topics in the following sections.

Individual characteristics. Personality factors such as low self-esteem, poor impulse control, external locus of control, negative affectivity, and heightened re- sponse to stress all increase the likelihood that an individ- ual will perpetrate family violence (Pianta, Egeland, & Erickson, 1989). Alcohol or drug dependence also plays a role both as a background risk factor and as an immediate precipitant of family violence (Kantor & Straus, 1990; Pan, Neidig, & O'Leary, 1994). Although the evidence is less consistent, some research has also suggested that the victims of family violence share some common char- acteristics, including poor physical or mental health, be- havioral deviancies, and difficult temperament or person- ality features (Belsky, 1993). For child victims, age also seems to play a role, as younger children are more prone to be seriously injured as a result of family violence (Lung & Daro, 1996).

Evidence of the association between alcohol and drug use and violence in families is an important illustra-

tion of the role of individual factors. In her review of the literature, Widom (1992) found alcoholism rates ranging from 18% to 38% in child-maltreating parents. Severely abusive husbands also are likely to have alcohol problems (Heyman, O'Leary, & Jouriles, 1995; Pan et al., 1994), and when the husband is abusive and a heavy drinker, alcohol is involved in about half the incidents of spousal violence (Kantor & Straus, 1990). The association be- tween alcohol abuse and family violence is complicated, involving the interplay of a number of other variables, including personality style, the nature of the relationship, the type of alcohol problem, and the degree of family conflict (Murphy & O'Farrell, 1994; Widom, 1992). An important question for further study is whether heavy alcohol use disinhibits aggression, interacting with per- sonality type to predict violence (Murphy & O'Farrell, 1994), or whether aggressive personality types act vio- lently whether or not they are drinking (Pernanen, 1991).

Immediate context. Characteristics of the imme- diate social context, especially of the family system, have important implications for the etiology or perpetuation of violence in that family (Emery, 1989). Studies have examined a host of contributing factors, including family structure and size, acute stressors such as the loss of a job or a death in the family, and characteristic styles of resolving conflicts or parenting.

Research on family stress illustrates this area of investigation. In one of the few prospective studies of child abuse, Egeland and colleagues (Egeland, Breiten- bucher, & Rosenberg, 1981) found that the number of stressful life events distinguished families who were re- ported for child abuse. Most stressed families did not abuse their children, however; thus, it was essential to distinguish stressed families who abused from those who did not. Holding life stress constant, this research team found that abusers were more aggressive, anxious, and defensive, and less succorant, than their nonmaltreating counterparts. Similar findings were found in later work (Egeland, Jacobvitz, & Sroufe, 1988). Thus, it seems clear that coping style, as well as stress, contributes to the development of family violence.

Broader ecological context. Violence in the family also is related to qualities of the community in which the family is embedded, such as poverty, absence of family services, social isolation, and the lack of social cohesion in the community. As Garbarino and Kostelny (1992) noted, family violence is " a social as well as a psychological indicator" (p. 463). The relation between poverty, social isolation, and child maltreatment has now been well established across all categories of abuse (Garbarino, 1976; Garbarino & Kostelny, 1992; Garba- rino & Sherman, 1980; Gelles, 1992; Hampton, 1987; Pan et al., 1994; Pelton, 1981; Polansky, Gaudin, Am- m o n s , & Davis, 1985; Zuravin, 1989). High levels of unemployment, inadequate housing, daily stresses, and community violence also contribute to an increased risk.

Of course, most poor parents do not maltreat their children (Bronfenbrenner, 1979; Garbarino, 1976); thus,

126 February 1998 • American Psychologist

the relation between poverty and abuse is not simple or direct. Garbarino and colleagues have consistently shown that the principal difference between poor families who do and those who do not maltreat their children lies in the degree of social cohesion and mutual caring found in their communities (Coulton et al., 1995; Garbarino & Crouter, 1978; Garbarino & Kostelny, 1992). For exam- ple, neighborhoods with high concentrations of child maltreatment reports tend to suffer from severe social disorganization and lack of community identity. They also have higher rates of juvenile delinquency, drug traf- ticking, and violent crime (Coulton et al., 1995). Other researchers have found that maltreating families are less likely than their nonmaltreating counterparts to socialize with their neighbors, feel that they can ask for help from others, or identify community organizations in which they can participate (Furstenberg, 1993). Some contend that individual personality factors cause the sense of so- cial isolation and that selection factors contribute to more chaotic neighborhoods (Polansky et al., 1985), but others argue that the causal arrow points in the opposite direc- tion (Cicchetti & Lynch, 1993; Garbarino & Kostelny, 1992; Vondra, 1990). The relationship between commu- nity and individual factors clearly is an important topic for further investigation.

Societal or cultural context. A number o f com- mentators contend that family violence is perpetuated by broad cultural beliefs and values, such as the use of physi- cal punishment, extremes in family privacy, and violence in the popular media (Finkelhor & Dziuba-Leatherman, 1994; Garbarino, 1977; Vondra, 1990). As noted earlier, our societal concern with violence in the family is histori- cally recent, and our continued reluctance to intervene forcefully is evident in the reluctance to terminate paren- tal rights even in egregious cases of child abuse (Besh- arov, 1996b). Societal policies may not cause family vio- lence, but many of our practices appear to condone it. Biological Factors We would be remiss not to mention biological contribu- tions to the etiology of family violence both in terms of normative human propensities and individual differences in behavior (A. Burgess & Draper, 1989; R. L. Burgess, Garbarino, & Gilstrap, 1983; Daly & Wilson, 1980, 1988; Malkin & Lamb, 1994). For example, evolutionary psychologists have pointed to potential risk (e.g., jealousy and "mate guarding") and protective (e.g., biological relatedness) factors that offer an innovative framework for conceptualizing violent actions in different family relationships (R. L. Burgess et al., 1983; Daly & Wilson, 1982, 1988, 1996a, 1996b, in press). Indeed, the well- documented link between poverty and child neglect (Pel- ton, 1981, 1989; Straus & Gelles, 1990) can be viewed from the evolutionary perspective. When resources are insufficient to meet basic needs, parents may differen- tially allocate their investment in and resources to their offspring. Research on this process is inconsistent (Bel- sky, 1993), but it is true that stepchildren, unplanned

children, and children in larger families all are at greater risk for child abuse (Daly & Wilson, 1988; Zuravin, 1989).

Research on aversively stimulated aggression offers strong evidence on another normative, biologically based propensity toward aggression. A variety of studies indi- cate that humans and other animals are prone to respond to a variety of unpleasant stimuli, including stressful fam- ily interactions, with aggression (Berkowitz, 1983). From this perspective, frustration and anger in close family relationships is expected, and the task of socialization is learning to recognize emotions and control aggressive impulses. Finally, in a very different area of investigation, we note that the familiality (or intergenerational transmis- sion) of family violence is consistent with biological as well as environmental models of causation, because fami- lies quite obviously share genes as well as environments (DiLalla & Gottesman, 1991).

Mulffple Pathways: Typologies of Abusers As a final point regarding etiology, it is important to highlight that theorists increasingly recognize the impor- tance of multiple pathways, as well as multiple risk fac- tors, in the development of family maltreatment and vio- lence. Early research on the etiology of spouse abuse and child maltreatment attempted to capture the pathological personality of the perpetrator (Melnick & Hurley, 1969; Steele & Pollack, 1968), whereas later reviewers aban- doned the psychopathological model (Belsky, 1993; Em- ery, 1989; Parke & Collmer, 1975). More recent efforts have sought to identify the different typologies of abus- ers, some of whom suffer from psychological disorders and many of whom do not (Francis, Hughes, & Hitz, 1992; Heyman, O'Leary, & Jouriles, 1995; Gottman et al., 1995; Jacobson, Gottman, & Shortt, 1995; Murphy, Meyer, & O'Leary, 1993, 1994; Saunders, 1992).

No consistent typology has yet emerged, but the approach is promising. In their review of the literature, Holtzworth-Munroe and Stuart (1994) identified three primary types of spouse batterers: Type 1 batterers are generally violent-antisocial; Type 2 batterers are family- only; and Type 3 batterers are dysphoric-borderline (emotionally volatile). Type 1 batterers tend to be violent across situations and potential victims. They also are more likely to abuse alcohol, be more belligerent and contemptuous, and have antisocial personality traits. Type 2 makes up the majority of spouse abusers, w h o tend to abuse only in the family, commit less severe acts, be less aggressive in general, suppress angry emotions, and feel remorseful. Type 2 abusers often are dependent and jealous and are unlikely to have personality or other disorders. Type 3 batterers also tend to be violent only within their family. They are more socially isolated and socially incompetent than other batterers, however, and often are depressed, feel inadequate, and are emotionally volatile. Type 3 abusers are more likely to have schizoid or borderline personalities.

February 1998 • American Psychologist 127

Using a similar typology, Gottman and colleagues have found evidence o f a possible physiological marker differentiating two types of spouse abusers (Gottman et al., 1995). One type of batterer (similar to Type 2) showed the expected accelerations in heart rate during conflictual interactions, whereas the second (similar to Type 1) showed marked decelerations in heart rate, suggesting a sort o f focused attention on their violent activity. Al- though these and other typologies are best viewed as tentative, they do imply that all a b u s e r s - - a n d all acts of a b u s e - - a r e not the same. Increasing recognition of alternative pathways to abuse is critically important to future research and intervention and is broadly consistent with our suggestion of distinguishing family mal- treatment and violence.

Consequences for Victims Family violence can cause a wide range of adverse conse- quences for victims--whether they are children, spouses, siblings, or elders--including serious physical injury, im- mediate and delayed psychological distress or disorder, and a variety o f practical upheavals such as placement in foster care or the need to flee the home. The consequences of victimization are a function of at least five broad classes of variables: (a) the nature of the abusive act (e.g., hitting, forced sex), as well as its frequency, intensity, and duration; (b) individual characteristics of the victim (e.g., age, self- blame); (c) the nature of the relationship between the victim and the perpetrator (e.g., spouse, stepparent); (d) the re- sponse of others to the abuse (e.g., social support, legal or psychological intervention); and (e) factors correlated with abuse that may exaccerbate its effects or, in fact, may account for some o f the putative consequences o f abuse (e.g., family chaos). The number of variables involved, as well as methodological limitations (Kendall-Tackett, Wil- liams, & Finkelhor, 1993; Malinosky-Rummell & Hansen, 1993; NRC, 1993), prevent us from reaching unambiguous conclusions about the consequences of family violence. However, we are able to point to some general patterns concerning physical, emotional, and practical outcomes, especially in regard to child and spouse abuse.

Physical Injuries Physical injury is a clear consequence of many instances of family violence, but information on the physical conse- quences of family violence is surprisingly incomplete. As noted earlier, an estimated 1,200 to 1,500 children, most of whom are under the age of five years, die each year as a result of either physical abuse (48%), neglect (37%), or both (15%) by a parent or parent figure (Lung & Daro, 1996). In addition, almost one third of the 4,967 women murdered in 1995 were killed by a boyfriend or husband (Federal Bureau of Investigation, 1996). Official data may somewhat underestimate actual deaths due to family violence, however, because of mis- classification of causes of death (McClain, Sacks, & Frohlke, 1993).

Information on nonlethal injuries is much more sketchy. According to NIS-3 (Sedlak & Broadhurst, 1996), nearly 50,000 children were victims of " s e r i o u s " physical abuse in 1993, that is, they suffered from life- threatening injuries, long-term physical impairment, or required professional treatment to prevent long-term physical impairment. It has also been estimated that over 18,000 children become severely disabled each year as a result of severe child abuse (Baladerian, 1991). We could not locate epidemiological data on more specific physical outcomes, such as the rate of nonorganic failure to thrive (resulting from gross neglect) among infants, specific injuries such as broken bones, or pregnancies or sexually transmitted diseases resulting from sexual abuse. Evidence on the physical consequence of spouse abuse is even more vague. For example, data indicate that at least 5% of women presenting at emergency rooms have been abused by their partner in the past year (McCauley, et al., 1995; Wilt & Olson, 1996), but we do not know whether these women sought help for injuries stemming from abuse, let alone the nature of the physical injuries sustained by battered women in the general popu- lation. Finally, there simply is no good information on the frequency of various physical injuries stemming from sibling or elder abuse.

The threat of injury often is devastating; however, we need more specific information on injury and disease for several reasons. First, physical safety is the foremost concern about family violence; thus, clear evidence of harm is essential for a number o f purposes ranging from basic knowledge to justifying intervention. Second, spe- cific physical injuries provide one relatively unambigu- ous (if limited) definition of family violence; thus, data would allow investigators to clearly chart the scope of the problem and changes over time. Third, clear informa- tion on the physical consequences of family violence should allow researchers to more adequately investigate the psychological consequences of physical abuse, espe- cially given evidence for a d o s e - r e s p o n s e relation, which we review shortly.

Psychological Consequences A number of recent articles have reviewed the large body of research on the psychological adjustment of the vic- tims of various types of family violence, including the child victims of physical abuse and neglect (Cicchetti & Toth, 1995; NRC, 1993), child victims of sexual abuse (Kendall-Tackett et al., 1993), adults physically (Mali- nosky-Rummell & Hansen, 1993) or sexually (Nash, Hulsey, Sexton, Harralson, & Lambert, 1993) abused as children, child witnesses of spousal abuse (Jaffe, Wolfe, & Wilson, 1990; Margolin, in press), and battered women (Margolin, in press). We can draw two general conclusions from this broad and sometimes conflicting literature. First, although some have found child neglect to be more psychologically harmful than abuse (Erick- son & Egeland, 1996) and experiencing violence to be more harmful than witnessing it (Sternberg et al., 1993),

128 February 1998 • American Psychologist

victims of all forms of family violence are at an increased risk for a variety of psychological problems. Second, there may be a modest relation between types of abusive acts and specific psychological outcomes, for example, between physical abuse and aggression among victims (Malinosky-Rummell & Hansen, 1993; Widom, 1989) or between sexual abuse and sexualized behavior in children (Kendall-Tackett et al., 1993). However, the more promi- nent finding is that all types of family violence are linked with diverse psychological problems ranging from ag- gression, to anxiety, to depression. Still, evidence on more specific psychological outcomes may be clouded by the need to consider (a) risk factors correlated with family violence; (b) clusters of symptoms (e.g., disorders like posttraumatic stress disorder [PTSD]) in addition to specific symptoms; (c) subtle psychological conse- quences that are difficult to document empirically, partic- ularly among children; and (d) psychological processes (not just psychological outcomes) set into motion or dis- rupted by the experience of family violence.

Correlated risk factors. Family violence is asso- ciated with a number of factors known to place children and adults at risk for psychological problems, for exam- ple, poverty (Pelton, 1992), troubled family environments (Egeland, in press), genetic liability (DiLalla & Gottes- man, 1991), and so on. These risk factors may account for the apparent relation between family violence and specific psychological outcomes, or their effects may in- teract with the consequences of abuse. Thus, for example, research comparing abused and nonabused children may actually reflect the psychological effects of anxious at- tachments, social isolation, or general family stress rather than the consequences of violence per se (Cicchetti & Toth, 1995; Egeland, in press; NRC, 1993). It is essential to consider this possibility both in research and in clinical interventions, which appropriately may focus on risk fac- tors in addition to, or even instead of, the abuse, particu- larly when abuse is less serious.

Acute a n d posttraumallc stress disorders. As- sessment at the level of psychological disorders, not just psychological symptoms, also should further understand- ing of the psychological consequences of family violence. Recent research on PTSD illustrates the promise of such broader assessments. Several investigators have now doc- umented PTSD among one quarter to one half (or more) of child victims of physical (Famularo, Fenton, Kinsch- erff, Ayoub, & Barnum, 1994; Famularo, Kinscherff, & Fenton, 1992; Kiser, Heston, Millsap, & Pruitt, 1991; Livingston, Lawson, & Jones, 1993) and sexual (Kiser et al., 1991; Livingston et al., 1993; McLeer, Deblinger, Hentry, & Orvaschel, 1992; D. A. Wolfe, Sas, & Wekerle, 1994) abuse. Moreover, the risk for PTSD increases when physical abuse is more severe and long-lasting (Kiser et al., 1991), and when sexual abuse occurs in a close relationship (McLeer et al., 1992) or involves threats, coercion, or guilt on the part of the child victim (D. A. Wolfe et al., 1994). Furthermore, researchers also have found high rates of PTSD among battered women who

are residents at shelters (Presty, 1996), as well as among abused women in couples therapy (Cascardi, O'Leary, Lawrence, & Schlee, 1995).

Further epidemiological research might usefully document more precise risks for PTSD associated with specific forms (or intensities or frequencies) of physical and sexual abuse, as has been done with the traumas in general. For example, rape has been demonstrated to pose a far greater risk for PTSD than other common traumas (Breslau, Davis, Andreski, & Peterson, 1991). Moreover, the PTSD diagnosis suggests new possibilities for psy- chological intervention, especially given recent research on the successful treatment of PTSD (Frueh, Turner, & Beidel, 1995; Rothbaum & Foa, 1996). Of course, the difficulties involved in diagnosing PTSD in children, as well as the need to develop sound measures of acute stress disorder, pose basic challenges to both researchers and clinicians.

SubHe psychological effects on vidlms. The diagnoses of PTSD and acute stress disorder include symptoms such as reexperiencing and dissociation; thus, they offer further impetus for studying some of the more subtle psychological consequences of abuse, particularly among children (Friedrich et al., 1992; V. V. Wolfe, Gen- tile, & Wolfe, 1989). The vehement controversy over "re- covered memories" of abuse provides another rationale for research on more subtle psychological reactions (Koss, Tromp, & Tharan, 1995; Loftus, 1994). A review of the recovered memories controversy is beyond the scope of this article, but we do note that one longitudinal study found that documented sexual abuse among girls ages 10 months to 12 years was not reported by nearly 40% of the victims when they were asked about their history as young adults (Williams, 1994a). Our present concern is not how supportive or unsupportive these data are of the recovered memories controversy (Loftus, Garry, & Feldman, 1994; Williams, 1994b) but the need to better understand the emotional and cognitive pro- cesses that may contribute to the forgetting of (or the dissociation from) abuse in the 17 years between victim- ization and reinterview.

We similarly are intrigued by the psychological mechanisms that may underlie other long-term outcomes, particularly the so-called intergenerational transmission of abuse. Evidence indicates that children who are vic- tims of violence are at an increased risk for becoming violent themselves as adults (Widom, 1989), although it is important to note that only a minority of abused or neglected children (perhaps one third) go on to become violent in their own families (Kaufman & Zigler, 1987; Malinosky-Rummell & Hansen, 1993; Widom, 1989). Some (or much) of the familiality of family violence may be accounted for by genetic rather than environmental risk (DiLalla & Gottesman, 1991), but environmental influences presumably are prominent nevertheless. Our main question about environmental transmission is how a propensity toward violence in the family might be psy- chologically represented across the years from childhood

February 1998 • American Psychologist 129

to adult life. Because fear is children's most immediate response to violence, processes more subtle than direct imitation must account for this continuity (Emery, 1982, 1989). We also urge more attention to such reactions as guilt and self-blame, particularly among victims o f sexual abuse (D. A. Wolfe et al., 1994).

Process of adaptation. Most research focuses on the outcome o f family violence, but it is obvious that the psychological consequences o f abuse are best described as a process that unfolds over time. W e n e e d more adequate models to describe this process. Finkelhor and Browne (1985) offered one example o f what is needed in describing sexual abuse victimization in terms o f the individual's development o f a sense o f traumatiza- tion, betrayal, stigmatization, and powerlessness as time passes following abuse. At the interpersonal level, we have previously outlined how violence can affect the en- tire family system, as when children attempt to intervene in parental disputes both as a means o f protecting their parent(s) and o f regulating their own affect (Emery, 1989). A third approach is described by Cicchetti and Toth (1995), who have thoroughly articulated how abuse may disrupt the normal tasks o f child development such as the formation o f attachments, affect regulation, the self system, and peer relationships. Such observations argue for the development o f both theory and methodol- ogy in research on family violence, perhaps including descriptive qualitative research, because processes are difficult to observe in traditional research designs. Practical Consequences Finally, we must underscore the fact that a b u s e - - a n d interventions with abusive f a m i l i e s - - c a n have tremen- dously adverse practical consequences for children and families. For example, 254,000 children were placed in foster care in 1994, the majority as a result o f abuse or neglect (Tatara, 1994). Foster-care placements often are less than ideal in many ways, but even excellent foster care causes considerable upheaval for children and some- times for parents who are not abusive (as may occur when children are removed from the home because an abused woman refuses to leave her batterer). As another example, the practical consequences o f leaving a marriage, such as uncertainties about residence, lost social support, threats to maintaining custody, and financial hardships, often explain why battered women fail to leave abusive relationships. In focusing on psychological concerns, mental health professionals must not overlook such basic issues. In intervening with abusive families, we must be certain that we can offer better alternatives. When mental health or social service professionals can offer better al- ternatives, we suggest that supportive rather than adver- sarial intervention can be more e f f e c t i v e - - e x c e p t in cases o f serious abuse.

Prevention and Intervention Attempts to prevent or stop family violence, or to treat victims or perpetrators, involve a huge array o f interven-

tions involving mental health, social service, and legal professionals, as well as other professionals who have regular contact with children and families (e.g., teachers, family physicians) and the popular media. Unfortunately, the creation and implementation o f new interventions has far outstripped research on their effectiveness, as previ- ous reviewers have noted (Becket et al., 1995; Melton et al., 1995; Melton & Flood, 1994; NRC, 1993; D . A . Wolfe, Reppucci, & Hart, 1995). Our review o f research and practice nevertheless leads us to advocate for new directions in intervention, all o f which relate to distin- guishing levels o f abuse. First, we urge clarification and revision o f child abuse reporting laws. Second, we en- courage the rediscovery o f the helping model o f social services for less serious cases o f abuse and for families under stress (partially as an effort toward the primary prevention o f abuse). Third, we advocate more forceful legal intervention in cases o f serious family violence.

Child Abuse Reporting B y all accounts, child protective services are driven b y - - and overwhelmed w i t h - - t h e investigation o f child abuse reports (Kamerman & Kahn, 1990; Murphy-Berman, 1994). As we have noted, over three million reports are made every year, but less than one third are substanti- ated (see Footnote 2). Moreover, social service workers are so swamped with reports and investigations that about 40% o f substantiated cases receive no services at all (McCurdy & Daro, 1993). Finally, only a small propor- tion o f substantiated cases, less than 20%, involve any type o f formal court action. Rather, child protective work- ers encourage the great majority o f those involved in substantiated cases to enter treatment voluntarily (Melton et al., 1995).

Mental health professionals thus are in a rather ironic position in regard to reporting child abuse. Most cases involve only suspected or mild abuse, in which families are stressed and parents have indicated at least some interest in receiving help b y virtue o f their contact with the mental health professional. Yet, under existing law, mental health professionals are required to report these cases, knowing that the likely outcomes are failure to substantiate, no provision o f services, no legal action, and eventually, encouraging the family to seek treat- m e n t - - e x a c t l y where they began the long, expensive, and intrusive process. Not surprisingly, mental health pro- fessionals co m m o n l y report that they have little faith in the reporting system and believe that it undermines therapeutic relationships (Levine, 1990; Zellman, 1990). Indeed, one prominent group o f commentators termed the resultant failure to report cases an action o f "civil d i s o b e d i e n c e " (Melton et al., 1995, p. 52).

Given this circuitous chain o f events, a simple but not necessarily minor change would be to alter reporting requirements for mental health professionals to reduce investigations and improve services. Specifically, we would exempt mental health professionals fro m reporting less serious cases o f abuse, whether known or suspected,

130 February 1998 • American Psychologist

when a family is actively engaged in treatment. We do not attempt to define these cases here, but we do note the need to define less serious cases (what we have called maltreatment) clearly in the law, as has been done in some experimental programs. Such a change would be a first step toward the broader goal of refocusing the child protection system on supporting rather than policing fam- ilies under stress, while simultaneously pursuing more vigorous, coercive intervention with cases of serious fam- ily violence. Supportive Interventions A number of supportive interventions have been devel- oped in an attempt to reduce violent behavior within families, including individual and group therapies for both victims and perpetrators, couples therapy for victims of domestic violence, parent-training and family therapy, and home-visiting programs for the prevention of child abuse. In general, the more serious and chronic the nature of the abuse, the less success these programs have in changing behaviors (Barth, 1991; Cohn & Dart, 1987; NRC, 1993; O'Leary, Neidig, & Heyman, 1995). With problems of mild to moderate abuse, however, multilevel programs, which combine behavioral methods, stress management, and relationship skills (parent-child or spousal), lower stress in families and may reduce the likelihood of continued aggression in both child and spouse abuse cases (Lutzker & Rice, 1987, Murphy, 1994; NRC, 1993).

Some interventions show promise, but the need for early intervention and especially prevention is under- scored by the difficulty of changing entrenched family violence and the stressful life circumstances that promote abuse (Garbarino & Kostelny, 1992; Pelton, 1992). Home-visitor programs for new parents living in difficult circumstances are one especially promising form of pre- vention. Home-visitor programs simultaneously assist with material needs (e.g., cribs, child care, transporta- tion), psychological needs (e.g., parenting education and support), and educational needs (e.g., job skills) and may both improve general family well-being and reduce child maltreatment (MacMillan, MacMillan, Offord, Grif- fith, & MacMillan, 1994).

An outstanding example is the Prenatal and Infancy Home Visitors' Program, which targeted low-income, teen, and single-parent mothers who were pregnant with their first child. On the basis of evaluations conducted during the program (Olds, Henderson, Chamberlin, & Tatelbaum, 1986; Olds, Henderson, Tatelbaum, & Chain- berlin, 1986), 1 to 2 years following termination (Olds, Henderson, & Kitzman, 1994; Olds, Henderson, Kitz- man, & Cole, 1995), and l0 to 15 years later (Eckenrode, personal communication, April 1997), home-visited mothers had fewer and less serious CPS reports, had fewer subsequent births, spent fewer months on welfare, had fewer arrests, and were less likely to abuse sub- stances in comparison to control families. In large part because of such benefits, in 1992 the NCPCA launched

a national initiative of home-visitor programs, Healthy Families America, in over 240 communities. Such efforts represent a return to the roots of a social service system originally intended to support families, not to police them (Besharov & Laumann, 1996). Ultimately, supportive in- terventions may be more effective in alleviating mal- treatment and in preventing family violence than the sin- gular use of investigation, which has come to dominate social service agencies. Coercive Interventlon As a final point, we note that distinguishing between levels of abuse also may help to improve intervention in cases of serious and extreme family violence. One benefit would be to free investigators and police to focus on the most serious cases. Shockingly, between 35% and 50% of all fatalities that are due to child abuse or neglect occur in cases that have already been brought to the attention of law enforcement and child protection agen- cies (Lung & Dart, 1996). Another benefit might be to help to clarify when coercive legal intervention is and is not appropriate. For example, debates have errupted over the once promising policy of mandatory arrest for violent partners (even over the objection of victims), because the reduced recidivism found in early research (Sherman & Berk, 1984) has not been replicated in subsequent studies (Berk, 1993; Buzawa & Buzawa, 1993; Schmidt & Sher- man, 1993). As another example, many commentators have questioned the overriding goal of family reunifica- tion following child abuse, especially in cases of serious physical and sexual abuse. In extreme cases like these, the termination of parental rights and early adoption may be the appropriate intervention, especially given the many problems with the overwhelmed foster-care system (Ta- tara, 1994).

In both of these examples, distinguishing between levels of abuse should help to resolve controversy and thereby clarify the appropriate use of coercive legal inter- vention. The idea of termination of parental rights or of arrest is far more threatening when our definitions of abuse include relatively minor acts, as they currently do. Clearly, it is difficult to draw a line between cases that should or should not lead to arrest or termination of parental rights (Azar, Benjet, Fuhrmann, & Cavallero, 1995), but the challenge of distinguishing between levels of abuse should not deter us from the task. What we have called maltreatment may be on a continuum of what we have termed violence, and both acts of abuse may differ from normal family aggression only by a matter of de- gree. As we have argued throughout this article, however, drawing distinctions between levels of abuse seems con- sistent with the state of our knowedge about the preva- lence of abuse, its development, its consequences, and appropriate intervention.

REFERENCES

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

February 1998 • American Psychologist 131

Azar, S.T., Benjet, C, L., Fuhrmann, G. S., & Cavallero, L. (1995). Child maltreatment and the termination of parental rights: Can behav- ioral research help Solomon? Behavior Therapy, 26, 599-623.

Baladerian, N. J. (1991). Abuse causes disabilities. Disability and the Family. Culver City, CA: Spectrum.

Barth, R. P. (1991). An experimental evaluation of in-home child abuse prevention services. Child Abuse & Neglect, 15, 363-375.

Becker, J. V., Alpert, J. L., BigFoot, D. S, Bonner, B. L., Geddie, L. E, Henggeler, S. W., Kaufman, K. L., & Walker, C. E. (1995). Empirical research on child abuse treatment: Report by the Child Abuse and Neglect Treatment Working Group, American Psychological Associa- tion. Journal o f Clinical Child Psychology, 24(Suppl.), 2 3 - 4 7 .

Bell, C. C., & Jenkins, E. J. (1993). Community violence and children in Chicago's southside. In D. Reiss, J.E. Richters, M. Radke- Yarrow, & D. Scharff (Eds.), Children and violence (pp. 46-54). New York: Guilford Press.

Belsky, J. (1993). Etiology of child maltreatment: A developmental- ecological analysis. Psychological Bulletin, 114, 413-434.

Berk, R. A. (1993). What the scientific evidence shows: On the average, we can do no better than arrest. In R.J. Gelles & D. R. Loseke (Eds.), Current controversies in family violence (pp. 323-336). Thousand Oaks, CA: Sage.

Berkowitz, L. (1983). Aversively stimulated aggression: Some parallels and differences in research with animals and humans. American Psy- chologist, 38, 1135-1144.

Besharov, D. J. (1988). The need to narrow grounds for state interven- tion. Protecting children from abuse and neglect: Policy and practice (pp. 47-90). Springfield, IL: Charles C Thomas.

Besharov, D. J. (1996a, October 4). Child abuse: Threat or menace? How common is it really? Slate, 9, 1-5.

Besharov, D. J. (1996b, December 1). When home is hell: We are too reluctant to take children from bad parents. The Washington Post, pp. C1, C 5 - C 6 .

Besharov, D.J., & Laumann (Billings), L. (1996). Child abuse re- porting. In I. Garfinkel, J. Hochschild, & S. McLanahan (Eds.) Social policies for children. Washington, DC: Brookings Institute Press.

Besharov, D., & Laumann (Billings), L. (1997). D o n ' t call it child abuse if it's really poverty. Journal of Children and Poverty, 3, 5 - 36.

Breslau, N., Davis, G. C., Andreski, P., & Peterson, E. (1991). Traumatic events and posttraumatic stress disorder in an urban population of young adults. Archives o f General Psychiatry, 4& 216-222.

Bronfenbrenner, U. (1979). The ecology of human behavior. Cambridge, MA: Harvard University Press.

Burgess, A., & Draper; P. (1989). The explanation of family violence: The role of biological, behavioral, and cultural selection. In L. Oh- lin & M. Tonry (Eds), Family violence (pp. 59-116). Chicago: Uni- versity of Chicago Press.

Burgess, R. L., Garbarino, J., & Gilstrap, B. (1983). Violence to the family. In E. J. Callahan & K. McCluskey (Eds.), Life span develop- mental psychology: Non-normative life events. San Diego, CA: Aca- demic Press.

Buzawa, E. S., & Buzawa, C. G. (1993). The scientific evidence is not conclusive and arrest is no panacea. In R. J. Gelles & D. R. Loseke (Eds.), Current controversies in family violence (pp. 337-356). Thousand Oaks, CA: Sage.

Cascardi, M., O'Leary, K. D., Lawrence, E. E., & Schlee, K. A. (1995). Characteristics of women physically abused by their spouses and who seek treatment regarding marital conflict. Journal o f Consulting and Clinical Psychology, 63, 616-623.

Cicchetti, D., & Barnett, D. (1991). Toward the development of a scien- tific nosology of child maltreatment. In D. Cicchetti & W. Grove (Eds.), Thinking clearly about psychology: Essays in honor o f Paul E. Meehl (pp. 346-377). Minneapolis: University of Minnesota Press.

Cicchetti, D., & Lynch, M. (1993). Toward an ecological/transactional model of community violence and child maltreatment: Consequences for children's development. Psychiatry, 56, 9 6 - 1 1 8 .

Cicchetti, D., & Toth, S. L. (1995). A developmental psychopathology perspective on child abuse and neglect. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 541-565.

Cohn, A. H., & Daro, D. (1987). Is treatment too late: What ten years of evaluative research tell us. Child Abuse & Neglect, 11, 433-442.

Community Partnerships for Protecting Children. (1996, Winter). News- letter, 1.

Coulton, C. J., Korbin, J. E., Su, M., & Chow, J. (1995). Community level factors and child maltreatment rates. Child Development, 66, 1262-1276.

Daly, M., & Wilson, M. (1980). Discriminative parental solicitude: A biological perspective. Journal of Marriage and the Family, 42, 2 7 7 - 288.

Daly, M., & Wilson, M. (1982). Male sexual jealousy. Ethology & Sociobiology, 3, 11-27.

Daly, M., & Wilson, M. (1988). Evolutionary social psychology and family homicide. Science, 242, 519-524.

Daly, M., & Wilson, M. (1996a). Male sexual propietariness and vio- lence against wives. Current Directions in Psychological Science, 5, 2 - 7 .

Daly, M., & Wilson, M. (1996b). Violence against stepchildren. Current Directions in Psychological Science, 5, 7 7 - 8 1 .

Daly, M., & Wilson, M. (in press). Crime and conflict: Homicide in evolutionary psychological perspective. In M. Tonry (Ed.), Crime and justice: An annual review of research. Chicago: University of Chicago Press.

Dietz, E. (1996). Violence against women in the United States. Arizona Journal o f lnternational Comparative Law, 13, 551-595.

DiLalla, L. E, & Gottesman, I. I. (1991). Biological and genetic contrib- utors to violence: Widom's untold tale. Psychological Bulletin, 109, 125-129.

Egeland, B. (in press). Mediators of the effects of child maltreatment on developmental adaptation in adolescence. In D. Cicchetti & S. L. Toth (Eds.), Rochester Symposium on Developmental Psychopathol- ogy, VoL 8: The effects of trauma on the developmental process. Rochester, NY: University of Rochester Press.

Egeland, B., Breitenbucher, M., & Rosenberg, D. (1981). Prospective study of the risk of life stress in the etiology of child abuse. Annual Progress in Child Psychiatry & Child Development, 666-682.

Egeland, B., Jacobvitz, D., & Sroufe, A. (1988). Breaking the cycle of abuse. ChiM Development, 59, 1080-1088.

Emery, R. E. (1982). Interparental conflict and the children of discord and divorce. Psychological Bulletin, 92, 310-330.

Emery, R. E. (1989). Family violence. American Psychologist, 44, 321 - 328.

Erickson, M. E, & Egeland, B. J. (1996). Child neglect. In J. Briere, L. Berliner, J. Bulkley, C. Jenny, & T. Reid (Eds.), The APSAC handbook on child maltreatment (pp. 4 - 2 0 ) . Thousand Oaks, CA: Sage.

Famularo, R., Fenton, T., Kinscherff, R., Ayoub, C., & Barnum, R. (1994). Maternal and child posttranmatic disorder in cases of child maltreatment. Child Abuse & Neglect, 18, 2 7 - 3 6 .

Famularo, R., Kinscherff, R., & Fenton, T. (1992). Psychiatric diagnoses of maltreated children: Preliminary findings. Journal of the American Academy of Child and Adolescent Psychiatry, 31, 863-867.

Federal Bureau of Investigation. ( 1991 ). Uniform Crime Reports. Wash- ington, DC: U.S. Government Printing Office.

Federal Bureau of Investigation. (1994). Uniform Crime Reports. Wash- ington, DC: U.S. Government Printing Office.

Federal Bureau of Investigation. (1996). Uniform Crime Reports for the United States. Washington, DC: U.S. Government Printing Office.

Fincham, F. D., Beach, S. R. H., Moore, T., & Diener, C. (1994). The professional response to child sexual abuse: Whose interests are served? Family Relations, 43, 244-254.

Finkelhor, D., & Browne, A. (1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal o f Orthopsy- chiatry, 55, 530-541.

Finkelhor, D., & Dziuba-Leatherman, J. (1994). Victimization of chil- dren. American Psychologist, 49, 173-183.

1 3 2 F e b r u a r y 1 9 9 8 • A m e r i c a n P s y c h o l o g i s t

Francis, C. R., Hughes, H. M., & Hitz, L. (1992). Physically abusive parents and the 16-PF: A preliminary psychological typology. Child Abuse & Neglect, 16, 673-691.

Friedrich, W. N., Grambsch, P., Damon, L., Hewitt, S. K., Koverola, S., & Wolfe, V. (1992). Child Sexual Behavior Inventory: Normative and clinical comparisons. Psychological Assessment, 4, 303-311.

Frueh, B. C., Turner, S. M., & Beidel, D. C. (1995). Exposure therapy for combat-related PTSD: A critical review. Clinical Psychology Review, 15, 799-815.

Furstenberg, E E (1993). How families manage risk and opportunity in dangerous neighborhoods. In W.J. Wilson (Ed.), Sociology and the public agenda (pp. 231-258). Newbury Park, CA: Sage.

Gallup, G. H., Moore, D. W., Schussel, R. (1995). Disciplining children in America: A Gallup Poll Report. Princeton, NJ: The Gallup Organization.

Garbarino, J. (1976). A preliminary study of some ecological correlates of child abuse. Child Development, 47, 178-185.

Garbarino, J. (1977). The human ecology of child maltreatment: A conceptual model for research. Journal o f Marriage and Family, 39, 721-735.

Garbarino, J. (1995). Growing up in a socially toxic environment: Life for children and families in the 1990s. In G. B. Melton (Ed.), Ne- braska Symposium on Motivation: Vol. 42. The individual, the family, and the social good: Personal fulfillment in times o f change (pp. 1 - 20). Lincoln, NE: University of Nebraska Press.

Garbarino, J., & Crouter, A. (1978). Defining the community context of parent-child relations: The correlates of child maltreatment. Child Development, 49, 604-612.

Garbarino, J., & Kostelny, K. (1992). Child maltreatment as a commu- nity problem. Child Abuse and Neglect, 16, 455-467.

Garbarino, J., & Sherman, D. (1980). High-risk neighborhoods and high-risk families: The human ecology of child maltreatment. Child Development, 51, 188-198.

Gelles, R. J. (1992). Poverty and violence toward children. American Behavioral Scientist, 35, 258-274.

Gottman, J. M., Jacobson, N. S., Rushe, R. H., Shortt, J. W., Babcock, J., La Taillade, J.J., & Waltz, J. (1995). The relationship between heart rate reactivity, emotionally aggressive behavior; and general violence in batterers. Journal o f Family Psychology, 9, 227-248.

Hampton, R. L. (1987). Race, class, and child maltreatment. Journal o f Comparative Family Studies, 18, 113-126.

Harvard Law Review. (1993). Developments in the l a w - - L e g a l re- sponses to domestic violence. Harvard Law Review, 106, 1528- 1550.

Haugaard, J. J., & Emery, R. E. (1989). Methodological issues in child sexual abuse research. Child Abuse & Neglect, 13, 89-100.

Heyman, R., O'Leary, D., & Jouriles, E. (1995). Alcohol and aggressive personality styles: Potentiators of serious physical aggression against wives? Journal o f Family Psychology, 9, 44-57.

Holtzworth-Munroe, A., & Stuart, G.L. (1994). Typologies of male batterers: Three subtypes and the differences among them. Psycho- logical Bulletin, 116, 476-497.

Jacobson, N. S., Gottman, J. M., & Shortt, J. W. (1995). The distinction between q~ype l and "Pype 2 batterers--Further considerations: Reply to Ordnuff et al. (1995), Margolin et al. (1995), and Walker (1995). Journal o f Family Psychology, 9, 272-279.

Jaffe, E G., Wolfe, D. A., & Wilson, S. K. (1990). Children of battered women. Newbury Park, CA: Sage.

Kamerman, S. B., & Kahn, A.J. (1990). The problems facing social services for children, youth, and families. Children and Youth Ser- vices Review, 12, 7-20.

Kantor, G. K., & Straus, M. A. (1990). The "drunken bum" theory of wife beating. In M. A. Straus & R. J. Gelles (Eds.), Physical violence in American families (pp. 203-224). New Brunswick, NJ: Transac- tion Press.

Kaufman, J., & Zigler, E. (1987). Do abused children become abusive parents? American Journal o f Orthopsychiatry, 57, 186-192.

Kempe, C.H., Silverman, E N., Steele, B. E, Broegemueller, W., & Silver, H. K. (1962). The battered-child syndrome. Journal of the American Medical Association, 181, 17-24.

Kendall-Tackett, K. A., Williams, L. M., & Finkelhor, D. (1993). Impact of sexual abuse on children: A review and synthesis of recent empiri- cal studies. Psychological Bulletin, 113, 164-180.

Kiser, L. J., Heston, J., Millsap, P. A., & Pruitt, D. B. (1991). Physical and sexual abuse in childhood: Relationship with posttraumatic stress disorder. Journal o f the American Academy o f ChiM and Adolescent Psychiatry, 30, 776-783.

Koss, M. P., Tromp, S., & Tharan, M. (1995). Traumatic memories: Empirical foundations, forensic, and clinical implications. Clinical Psychology: Science & Practice, 2, 111-132.

Layzer, J. I., Goodson, B. D., & DeLange, C. (1986). Children in shel- ters. Response, 9, 2 - 5 .

Levine, M. (1990). The need for a special relationship doctrine in the child protection context. Brooklyn Law Review, 56, 329-376.

Livingston, R., Lawson, L., & Jones, J. G. (1993). Predictors of self- reported psychopathology in children abused repeatedly by a parent. Journal of the American Academy o f Child and Adolescent Psychia- try, 32, 948-953.

Loftus, E. (1994). The repressed memory controversy. American Psy- chologist, 49, 443-445.

Loftus, E. E, Garry, M., & Foldman, J. (1994). Forgetting sexual trauma: What does it mean when 38% forget? Journal o f Consulting and Clinical Psychology, 62, 1177-1181.

Lung, C. T., & Daro, D. (1996). Current trends in child abuse reporting and fatalities: The results o f the 1995 annual fifty state survey. Chi- cago: National Committee to Prevent Child Abuse.

Lutzker, J., & Rice, J. M. (1987). Project 12-Ways: Measuring outcome of a large in-home service for the treatment and prevention of child abuse and neglect. Child Abuse & Neglect, 8, 519-524.

MacMillan, H. L., MacMillan, J. H., Offord, D. R., Griffith, L., & Mac- Millan, A. (1994). Primary prevention of child abuse and neglect: A critical review. Part I. Journal o f Child Psychology, Psychiatry, and Allied Disciplines, 35, 835-856.

Malinosky-Rummell, R., & Hansen, D.J. (1993). Long-term conse- quences of child abuse and neglect. Psychological Bulletin, 114, 6 8 - 79.

Malkin, C. M., & Lamb, M. E. (1994). Child maltreatment: A test of sociobiological theory. Journal o f Comparative Family Studies, 25, 121-133.

Margolin, G. (in press). The effects of domestic violence on children. In P. K. Trickett & C, Schellenbach (F.xls.), Violence against children in the family and community. Washington, DC: American Psychologi- cal Association.

McCanley, J., Kern, D.E., Kolodner, K., Dill, L., Schroeder, A. E, DeChant, H. K., Ryden, J., Bass, E. B., & Derogatis, L. R. (1995). The "Battering Syndrome": Prevalence and clinical characteristics of domestic violence in primary care internal medicine practices. Annals o f Internal Medicine, 123, 737-746.

McClaln, P., Sacks, J., & Frohlke, R. (1993). Estimates of fatal child abuse and neglect, United States, 1979-1988. Pediatrics, 91, 3 3 8 - 343.

McCloskey, L. A., Figuerdo, A. J., & Koss, M. P. (1995). The effects of systemic family violence on children's mental health. Child Devel- opment, 66, 1239-1261.

McCurdy, K., & Daro, D. (1993). Current trends in child abuse re- porting and fatalities: The results o f the 1992 annual fifty state survey. Chicago: National Center on Child Abuse Prevention Research.

McLeer, S., Deblinger, E., Hentry, D., & Orvaschel, H. (1992). Sexually abused children at high risk for post-tranmatic stress disorder. Jour- nal o f the American Academy of Child and Adolescent Psychiatry, 31, 875-879.

Melnick, B., & Hurley, J. R. (1969). Distinctive personality of child- abusing mothers. Journal o f Consulting and Clinical Psychology, 33, 746-749.

Melton, G. B., & Flood, M. E (1994). Research policy and child mal- treatment: Developing the scientific foundation for effective protec- tion of children. Child Abuse & Neglect, 18, 1-28.

Melton, G. B., Goodman, G. S., Kalichman, S. C., Levine, M., Saywitz, K., & Koocher, G.P. (1995). Empirical research on child mal-

F e b r u a r y 1 9 9 8 • A m e r i c a n P s y c h o l o g i s t 133

treatment and the law. Journal o f Clinical Child Psychology, 24(Suppl.), 4 7 - 7 7 .

Murphy, C. M. (1994). Treating perpetrators of adult domestic violence. Maryland Medical Journal, 43, 877-883.

Murphy, C. M., Meyer, S. L., & O'Leary, K. D. (1993). Family of origin and MCMI-II psychopathology among partner assaultive men. Vio- lence and Victims, 8, 165-176.

Murphy, C. M., Meyer, S. L., & O'Leary, K.D. (1994). Dependency characteristics of partner assaultive men. Journal o f Abnormal Psy- chology, 103, 729-735.

Murphy, C. M., & O'Farrell, T. J. (1994). Factors associated with mari- tal aggression in male alchoholics. Journal o f Family Psychology, 8, 321-335.

Murphy-Berman, V. (1994). A conceptual framework for thinking about risk assessment and case management in child protective services. Child Abuse & Neglect, 18, 193-201.

Nash, M. R., Hulsey, T. L., Sexton, M. C., Harralson, T. L., & Lambert, W. (1993). Long-term sequelae of childhood sexual abuse: Perceived family environment, psychopathology, and dissociation. Journal o f Consulting and Clinical Psychology, 61, 276-283.

National Center on Child Abuse and Neglect. (1996). Child mal- treatment 1994: Reports from the states to the National Center on Child Abuse and Neglect. Washington, DC: U.S. Government Print- ing Office.

National Center on Elder Abuse. (1997). Elder abuse information series #2: Elder abuse in domestic settings [On-line]. Available: http:/l www.interinc.com/NCEA/Settings/Statistics.

National Research Council. (1993). Understanding child abuse and neglect. Washington, DC: National Academy Press.

Olds, D. L., Henderson, C. R., Chamberlin, R., & Tatelbaum, R. (1986). Preventing child abuse and neglect: A randomized trial of nurse home visitation. Pediatrics, 78, 6 5 - 7 8 .

Olds, D. L., Henderson, C. R., & Kitzman, H. (1994). Does prenatal and infancy home visitation have enduring effects on qualities of parental caregiving and child health at 25 to 50 months of life? Pediatrics, 93, 8 9 - 9 8 .

Olds, D. L., Henderson, C. R., Kitzman, H., & Cole, R. (1995). Effects of prenatal and infancy nurse home visitation on surveillance of child maltreatment. Pediatrics, 95, 365-372.

Olds, D. L., Henderson, C. R., Tatelbaum, R., & Chamberlin, R. (1986). Improving the delivery of prenatal and outcomes of pregnancy: A randomized trial of nurse home visitation. Pediatrics, 77, 16-28.

O'Leary, K. D., Barling, J., Arias, I., Rosenbaum, A., Malone, J., & Tyree, A. (1989). Prevalence and stability of physical aggression between spouses: A longitudinal analysis. Journal o f Consulting and Clinical Psychology, 57, 263-268.

O'Leary, K.D., Neidig, P. H., & Heyman, R.E. (1995). Assessment and treatment of partner abuse: A synopsis for the legal profession. Albany Law Review, 58, 1215-1233.

Pan, H. S., Neidig, P. H., & O'Leary, K. D. (1994). Predicting mild and severe husband to wife physical aggression. Journal o f Consulting and Clinical Psychology, 62, 975-981.

Parke, R. D., & Collmer, C. W. (1975). Child abuse: An interdisciplinary analysis. In E. M. Hetherington (Ed.), Review o f child development research, Vol. 5 (pp. 509-590). Chicago: University of Chicago Press.

Pelton, L. H. (1981). The social context o f child abuse and neglect. New York: Human Services Press.

Pelton, L. H. (1989). For reasons o f poverty. New York: Praeger. Pelton, L. H. (1992). The role o f material factors in child abuse and

neglect. Unpublished paper prepared for the U.S. Advisory Board on Child Abuse and Neglect.

Pelton, L. (1994). The role of material factors in child abuse and neglect. In G. B. Melton & E D. Barry (Eds.), Protecting children from abuse and neglect: Foundations f o r a new strategy (pp. 131-181). New York: Guilford Press.

Pernanen, K. (1991). Alcohol in human violence. New York: Guilford Press.

Pianta, R. B., Egeland, B., & Erickson, M. E (1989). The antecedents of maltreatment: Results of the Mother-Child Interaction Research

Project. In D. Cicchetti & V. Carlson (Eds.), Child maltreatment: Theory and research on the causes and consequences o f child abuse and neglect (pp. 203-253). New York: Cambridge University Press.

Polansky, N. A., Gaudin, J. M., Ammons, P. W., & Davis, K. B. (1985). The psychological ecology of the neglectful mother. Child Abuse & Neglect, 9, 265-275.

Presty, S. K. (1996). Psychological sequelae o f battered women residing in rural community shelters. Unpublished doctoral dissertation, Uni- versity of Virginia, Charlottesville.

Reppucci, N. D., & Haugaard, J. J. (1989). Prevention of child sexual abuse: Myth or reality. American Psychologist, 44, 1266-1275.

Rothbaum, B. O., & Foa, E. B. (1996). Cognitive-behavioral therapy for posttraumatic stress disorder. In B. A. van der Kolk, A. C. Mac- Farlane, & L. Weisaeth (Eds.), Traumatic stress: The effects o f over- whelming experience on mind, body, and society (pp. 491-509). New York: Guilford Press.

Saunders, D. G. (1992). A typology of men who batter: Three types derived from cluster analysis. American Journal o f Orthopsychiatry, 62, 264-275.

Schmidt, J. D., & Sherman, L. W. (1993). Does arrest deter domestic violence? American Behavioral Scientist, 36, 601-609.

Sedlak, A. J. (1981). National incidence and prevalence o f child abuse and neglect: 1980. Rockville, MD: Westat, Inc.

Sedlak, A. J. (1991). National incidence and prevalence o f child abuse and neglect: 1988, Revised. Rockvilte, MD: Westat, Inc.

Sedlak, A. J., & Broadhurst, D.D. (1996). Third National Incidence Study on Child Abuse and Neglect. Washington, DC: U.S. Depart- ment of Health and Human Services.

Sherman, L. W., & Berk, R. (1984). The specific deterrent effects of arrest for domestic assault. American Sociological Review, 49, 261 - 272.

Steele, B. E, & Pollack, C. B. (1968). A psychiatric study of parents who abuse their infants and small children. In R. E. Heifer & C. H. Kempe (Eds), The battered child (pp. 89-133). Chicago: University of Chicago Press.

Sternberg, K. J., Lamb, M. E., Greenbaum, C., Cicchetti, D., Dawud, S., Cortes, R. M., Krispin, O., & Lorey, E (1993). Effects of domestic violence on children's behavior problems and depression. Develop- mental Psychology, 29, 4 4 - 5 2 .

Straus, M.A., & Gelles, R.J. (1990). Physical violence in American families. New Brunswick, NJ: Transaction Press.

Straus, M. A., Gelles, R. J., & Steinmetz, S. K. (1980). Behind closed doors: Violence in the American family. Garden City, NY: Anchor.

Tatara, T. (1994). The recent rise in the U.S. child substitute care population: An analysis of national child substitute care flow data. In R.P. Barth, J.D. Berrick, & N. Gilbert (Eds.), Child welfare research review, Vol. 1 (pp. 126-145). New York: Columbia Univer- sity Press.

U.S. Advisory Board on Child Abuse and Neglect. (1993). Neighbors helping neighbors: A new national strategy f o r the protection o f children (Fourth report). Washington, DC: U.S. Government Printing Office.

Vondra, J. I. (1990). The community context of child abuse. Marriage and Family Review, 15, 19-38.

Wald, M. S. (1991 ). Defining psychological maltreatment: The relation- ship between questions and answers. Development and Psychopa- thology, 3, 111-118.

Widom, C. S. (1989). The cycle of violence. Science, 244, 160-166. Widom, C. S. (1992, June). Child abuse and alcohol use. Paper pre-

pared for the Working Group on Alcohol-Related Violence: Fostering Interdisciplinary Perspectives, convened by the National Institute on Alcohol Abuse and Alcoholism, Washington, DC.

Williams, L . M . (1994a). Recall of childhood trauma: A prospective study of women's memories of child sexual abuse. Journal o f Con- sulting and Clinical Psychology, 62, 1167-1176.

Williams, L . M . (1994b). What does it mean to forget child sexual abuse? A reply to Loftus, Garry, and Feldman. Journal o f Consulting and Clinical Psychology, 62, 1182-1186.

1 3 4 F e b r u a r y 1 9 9 8 • A m e r i c a n P s y c h o l o g i s t

Wilt, S., & Olson, S. (1996). Prevalence of domestic violence in the United States. Journal o f American Medical Women's Association, 51, 77-82.

Wolfe, D. A., Reppucci, N. D., & Hart, S. (1995). Child abuse preven- tion: Knowledge and priorities. Journal o f Clinical Child Psychology, 24(Suppl.), 5-23.

Wolfe, D. A., Sas, L., & Wekerle, C. (1994). Factors associated with the development of post-traumatic stress disorder among child victims of sexual abuse. Child Abuse & Neglect, 18, 37-50.

Wolfe, V. V., Gentile, C., & Wolfe, D. A. (1989). The impact of sexual abuse on children: A PTSD formulation. Behavior Therapy, 20, 215- 228.

Zellman, G. (1990). Report decision-making patterns among man- dated child abuse reporters. Child Abuse & Neglect, 14, 3 2 5 - 336.

Zuravin, S. J. (1989). The ecology of child abuse and neglect: Review of the literature and presentation of data. Violence and Victims, 4, 101-120.

F e b r u a r y 1 9 9 8 • A m e r i c a n P s y c h o l o g i s t 135