Etiology of Child Maltreatment
Certain characteristics of child maltreatment complicate research into its etiology. These characteristics include: (1) the extreme socially deviant nature of the behavior, (2) its low prevalence, (3) the presence of multiple factors in the context of child maltreatment, such as poverty and violence, (4) changing political and historical definitions of the behavior, and (5) the troubling and complex nature of the behavior that requires a rethinking of conventional wisdom about human nature and parenting.
Variation in operational definitions and theoretical concepts of child maltreatment is a major problem in reviewing the etiology of child maltreatment. Although this chapter sometimes distinguishes among the etiologies of different kinds of maltreatment, the necessary data to support these distinctions are generally unavailable. The panel believes that, rather than separating research on subpopulations divided by types of maltreatment, it is more useful to review research within a framework that focuses on the range of factors associated with child maltreatment as a general phenomenon. The panel recognizes that some factors are more closely linked with certain forms of child abuse and neglect (such as the relationship between poverty and child neglect). However, as noted in Chapter 2, similarities and differences in the etiologies of physical abuse, physical punishment, sexual abuse, emotional abuse, and neglect have not been well articulated in the scientific literature. In many cases research has not differentiated the etiologies and outcomes associated with multiple forms of maltreatment especially when various forms co-occur in one individual, either within the
same contemporaneous period or during a lifetime. Comparative studies of the origins and correlates of different kinds of abuse are rare.
Most forms of maltreatment are part of a pattern of maladaptive behavior that emerges over time, but research evidence regarding the origins and maintenance of this pattern of behavior is not clear. Investigators disagree about whether child maltreatment is a continuum of behaviors (ranging from mild physical discipline to severe forms of physical or sexual abuse) or a set of unique behavioral problems with distinctive etiologies (Gelles, 1991). Since studies of multiple forms of maltreatment are rare and researchers generally deal with one type of maltreatment in their work, such disagreement may result from the manner in which research projects have been organized.1 More recently, researchers are giving more attention to factors such as the severity and chronicity of abuse and neglect and the co-occurrence of multiple forms of maltreatment. With few exceptions (Wolfe, 1991), most etiological models lack a vocabulary for understanding the temporal organization of child maltreatment or demonstrating potential connections between maladjustments (including attitudes and beliefs) and abusive behaviors of the perpetrators. The existing models also do not resolve uncertainties about the continuum that may or may not exist between physical punishment and physical abuse, or between inadequate care giving and parental neglect. As a result, we currently know very little about the significant causes and pathways that influence risk factors in the etiology of child maltreatment.
Overview Of Etiological Models
Etiological models of child maltreatment are beginning to evolve from isolated cause-and-effect models to more sophisticated approaches that consider multiple pathways and interactive effects among factors that contribute to child maltreatment.2 In the early 1970s, recognizing the limitations of focusing on only parent or only child characteristics, researchers started to emphasize interactions among child, parent, and environmental risk factors. Gil (1970), for example, was one of the first to document the role of poverty and family disadvantage on the rates of child abuse. His work was followed by investigations by Garbarino (1977), who noted that isolation from social support systems was a significant, but not a sufficient, condition of child maltreatment (Wolfe, 1991).
The recognition of the role of ecological or ''situational" factors gradually led to the development of contemporary multicausal interactive models, which emphasize the importance of the sociocultural context of child maltreatment. Current theoretical models include: (a) the ecological models of Belsky and Garbarino, based on the conceptions of Urie Bronfenbrenner (Belsky, 1980, Garbarino, 1977; Lutzker, 1984); (b) the transitional model
of Wolfe, which views child maltreatment as an escalating process and as one end of the continuum of maladaptive parenting (Wolfe, 1991); and (c) the transactional model of Cicchetti (Cicchetti and Carlson, 1989), based on Sameroff and Chandler's (1975) formulations.
Although simple models identified key variables associated with child maltreatmentoften termed "risk factors"they did not establish a firm etiology of child maltreatment or specify causal relationships or sequences between the associated variables. Furthermore, results across these studies are often conflicting, and the predictive power of single variables, such as the individual characteristics of the parent, child, or environment alone, is limited. The emerging social interactional models emphasize the importance of viewing child maltreatment in the context of the family, community, and society rather than emphasizing only individual parental psychopathology or individual stressors (Belsky, 1980, 1992; Cicchetti and Carlson, 1989; Garbarino, 1977; Parke and Collmer, 1975; Wolfe, 1991). The phenomenon of child abuse and neglect has thus been moved away from the conception of an individual disorder or psychological disturbance, toward the conception of a symptom of an extreme disturbance of childrearing, often part of a context of other serious family problems, such as poverty, alcoholism, or antisocial behavior (Burgess, 1979; Pelton, 1989; Starr, 1979; Wolfe, 1991). New empirical findings invoking interaction models suggest that, although studies of abusive and nonabusive parents have not detected important significant differences in terms of personality dimensions, studies of the interactions of abusive and nonabusive family processes have yielded important distinctions, including unrealistic expectations of their children, the tendency to view their own children's behavior as extremely stressful, and their view of themselves as inadequate or incompetent in the parenting role (Wolfe, 1991).
As a result of these shifting paradigms, the panel has observed that terms in the research literature on the origins of child maltreatment are often confused in discussions of cause and effect and risk relationships. The use of terms in child maltreatment studies such as risk factors, intermediate or moderator variables, mitigators, mediators, confounding variables, and so forth lacks the precision found in fields that have more developed sources of statistical and epidemiological data to test theories.3 Furthermore, theoretical terms used in child maltreatment discussions are generally not matched by empirical data, and factors that are hypothesized as significant correlates have often not been tested in rigorous controlled studies. Much of the data base relies on anecdotal material derived from clinical research. As a result, many variables are hypothesized as acting in multiple ways, sometimes as antecedents to child maltreatment, sometimes as consequences, sometimes as factors that are present with or without a modifying effect on the causal relationships that result in child maltreatment.
For example, a strong association has been shown between the role of poverty in generating stressful experiences and the anger that become precipitating factors in child abuse and neglect (Gil, 1970; Pelton, 1978, 1989). Yet the relationship between poverty and child maltreatment is complexmost poor parents clearly are not abusive and poverty alone is not a sufficient or necessary antecedent for child maltreatment. In addition, the effects of parent and family characteristics on the etiology of child maltreatment may vary significantly with social class (Trickett et al., 1991).
Interactive models generally build on a probabilistic risk assessment process, assuming that child maltreatment occurs when multiple risk factors outweigh protective, compensatory, and buffering factors (Cicchetti and Carlson, 1989). Some factors may be relatively enduring and others transient. Some factors may play important roles in instigating maltreatment, while others may help sustain patterns of abuse and neglect. A factor may be protective in some combinations or increase the potential for abuse (potentiating) in others. It is the combination of risk potentiating and protective factors in all levels of the system that determine the likelihood of maltreatment, rather than a single factor serving as a causal influence in isolation from the others (Cicchetti and Carlson, 1989). In reviewing potentiating and contributing factors, researchers often focus on risk factors that appear to be malleable, that is, that can be changed as a result of a treatment or preventive interventions.
This perspective suggests that maltreatment results from complex constellations of correlated variables whose influence may increase or decrease during different developmental and historical periods. The combined effects of multiple variables provide diverse possible pathways to maltreatment. Furthermore, interactive models recognize that risk and protective factors are not static, but change over time as individuals, their life circumstances, and the society in which they live change. The interactive models, although relatively new, show promise and suggest issues that need to be addressed in research on the etiology of child maltreatment. However, the complexity of analysis associated with interactive models and the difficulties of distinguishing causal effects from observational data have inhibited their testing and application.
The panel has selected a developmental/ecological/transactional model of the etiology of child maltreatment as the basis for reviewing the key literature relevant to this chapter (Belsky, 1980; Cicchetti and Lynch, in press; Garbarino, 1977). As summarized in Figure 4.1, this model was selected for its breadth and advantages in organizing the large and often conflicting literature on the etiology of child maltreatment. Although the selected model identifies promising strategies or questions that should be addressed in future research, it is not intended to exclude others in research on child maltreatment.
The selected model views maltreatment within a system of risk and protective factors interacting across four levels: (1) the individual or ontogenic level, (2) the family microsystem, (3) the exosystem, and (4) the social macrosystem. The ontogenic level involves individual characteristics and the changing developmental status of family members. The family microsystem includes the family environment, parenting styles, and interactions among family members. The exosystem consists of the community in which the family lives, the workplace of the parents, school and peer groups of the family members, formal and informal social supports and services available to the family, and other factors such as family income, employment, and job availability. Finally, the social macrosystem consists of the overarching values and beliefs of the culture.
The ecological/developmental framework indicated in Figure 4.1 begins with an analysis of individual factors at the individual level and proceeds through the other ecological levels. This approach follows the conception
of human development as being nested within a set of transacting ecological systems. This conceptual framework is presented by the panel to highlight emerging research priorities in the field of child maltreatment.
Individual Ontogenic Factors
The influence of ontogenic factors such as adult4 and child characteristics in contributing to child maltreatment is often moderated by interactions with other factors. Attempts to identify adult or child characteristics related to maltreatment have produced an inconsistent and contradictory research literature. Consequently, the effects of individual factors need to be studied in conjunction with other factors and not studied in isolation.
Adult Personality Characteristics
Early studies of the etiology of child maltreatment assumed that a distinct psychiatric syndrome or disorder would be found to characterize parents or other caretakers (such as stepparents, grandparents, and foster parents) who maltreat children under their supervision and care. Although a small percentage of parents involved in child maltreatment could be diagnosed with a psychotic disorder, most individuals were identified as troubled or anxious persons who rarely exhibited extreme psychopathology (Steele and Pollock, 1968). A consistent profile of parental psychopathology or a significant level of mental disturbance has not been supported (e.g., Melnick and Hurley, 1969; Polansky et al., 1981, 1992; Spinetta and Rigler, 1972). However, certain types of psychiatric disorders can be important factors in determining outcomes for the maltreated childchildren reported as maltreated are less likely to remain with their biological family if evidence of a parental psychiatric disorder is obtained (Runyan et al., 1981; Widom, 1991).
Early psychiatric studies stimulated a search for parental characteristics and a personality profile of abusing parents (Milner and Chilamkurti, 1991).5 Recent prospective studies (e.g., Pianta et al., 1989) have identified a set of parental personality attributes associated with child maltreatment that have emerged with sufficient frequency to warrant attention. These attributes are low self-esteem, external locus of control, poor impulse control, negative affectivity (including depression and anxiety), and antisocial behavior (including aggression and substance abuse) (for reviews, see Baumrind, 1992; Belsky, 1992; Cicchetti and Lynch, in press). Central in these attributes is a triad of highly correlated personality characteristics, involving depression, anxiety, and antisocial behavior.
Depression, anxiety, and antisocial behavior are associated with disrupted social relations, social isolation, unavailability or lack of utilization of social supports, and an inability to cope with stress (Crittenden, 1985; Wolfe, 1985). Disruptions in social relations are also found in studies of maltreating parents who are described as insular, alienated, unhappy and dissatisfied in relationships with friends, neighbors, spouses, and children. This pervasive discontent and lack of skill in social relations can be exacerbated by additional stressors (Belsky, 1980; Cicchetti and Lynch, in press; Garbarino, 1977). Furthermore, these attributes and attitudes are likely to increase the probability of encountering stressful life experiences and inhibit the development of supportive relationships with a spouse, friends, and family that could help buffer the affected individual from the effects of stress.
Polansky et al. (1981) have proposed that parental characteristics help explain the origins of neglect, particularly chronic neglect. Early studies of neglectful families have suggested that child neglect is only one expression of pervasive and deeply rooted inadequacies in the life of a parent that sometimes appear early in adolescence. This condition has been termed as a character disorder of neglectful parents, usually expressed as the "apathy-futility syndrome" and the "impulse-ridden character" (Polansky et al., 1972, 1981, 1992). Although neglectful parents appear to be less depressed, anxious, angry, and confused than physically abusive parents (Pianta et al., 1989), such parents have been termed childlike or infantile, revealing an absence of self-esteem and an inability to plan important life choices such as marriage, having children, or getting a job. Impulse-ridden behaviors can result from early deprivations in the parent's own life, usually involving the absence of mature adults with whom the child may identify.
The literature on adult personality characteristics associated with child sexual abuse is more extensive than that of other forms of child maltreatment, since the primary etiology of child sexual abuse has been sought in the profile of the adult offender in contrast to other forms of child maltreatment, which often focus on parent-child interactions. Although no specific syndrome or diagnostic category has been associated with child sexual abuse, personality characteristics frequently found in child molesters have contributed to various etiological theories of pedophilia (DSM-III-R, 1987). Some child molesters are reported to be timid, unassertive and awkward; others
exhibit conduct disorder and poor impulse control; others are successful community leaders who have achieved professional respect. Empirical evidence clarifying the role of psychological and psychosexual development and maturity levels of child molesters is needed (Araji and Finkelhor, 1986).
Psychiatric profiles used to classify sex offenders frequently report the presence of an antisocial personality disorder among child molesters, but sex offenders have a heterogeneous range of psychopathology and personality disorders and an accepted system for sexual offender classification and the contribution of perpetrator characteristics has not been established6 (Conte, 1984; Hartman and Burgess, 1989; Lanning, 1992; Prentky, 1990).
Currently, Faller has suggested an incest-assault continuum, noting that, although contributing factors from the cultural, environmental, individual, and family context will vary from case to case, sexual abuse requires "an adult who has sexual desires toward children and the willingness to act upon them" (Faller, 1988:115). Efforts to classify pedophiles and incest offenders have also focused on the style of abuse, drawing on information obtained from offender self-reports, criminal investigative reports, and victim reports (Hartman and Burgess, 1989). A series of research studies have sought to highlight critical factors in the style of sexual abuse, such as the degree of violence (Finkelhor, 1984; Wyatt and Newcomb, 1990), the relationships among sexual stimuli and violence stimuli and their respective arousal components (Hartman and Burgess, 1989); the offender-victim relationship (Panton, 1978; Wyatt and Newcombe, 1990), the victim's and offender's age (Armentrout and Hauer, 1978); and the offender's level of education and mediators to negative outcomes (Knight, 1985; Wyatt and Newcombe, 1990).
Although the large majority of adult offenders in reported child sexual abuse cases are male, the increasing number of reports of female offenders suggests an unexplored pathway in examining the dynamics and origins of child sexual abuse (Finkelhor, 1987). Although women have been reported in a smaller number of cases (Finkelhor, 1987), concerns about detection bias, general research inattention to women, and the significance of maternal-child relations suggest that the role of female sexual offenders has been underestimated in research on child sexual abuse. Clinical studies of child victims of sexual abuse as well as adult offenders (based on retrospective studies) indicate that behavioral and perceptual disorders resulting from childhood sexual victimization may contribute to subsequent assault behavior (Becker, 1988; Hartman and Burgess, 1989).
One promising area of research inquiry has examined outcomessuch as power, control, sadistic pleasure, or displaced angerthat offenders seek to achieve in the sexual victimization of a child (Knight et al., 1985). Finkelhor (1987) has proposed four major theories, often presented as competing explanations, to explain child sexual abuse:
Abusers obtain powerful, developmentally induced emotional gratification from the acts;
Abusers have deviant physiological sexual arousal patterns;
Abusers are blocked by arrested psychosexual development and emotional immaturity in their capacity to meet their sexual needs in more conventional ways; and
Abusers have problems in their capacity for behavioral inhibition (Finkelhor, 1987).
The search for a biological basis for child sexual abuse has also not been successful. Although mental retardation or physiological abnormalities sometimes provoke arousal and disinhibition in sexual abusers, such abnormalities have not been substantiated as a major cause of sexual abuse (Araji and Finkelhor, 1986; Kelly and Lusk, 1992; Langevin, 1983). Hormone levels and chromosomal makeup have been studied extensively, but definitive evidence of these factors accounting for specific sexual interest toward children has not emerged (Goy and McEwen, 1977; Kelly and Lusk, 1992). The disinhibiting contribution of alcohol and alcoholism is the most frequent and well-established biologic agent often associated with sexual abuse, ranging from 19 to 70 percent of reported cases (Aarens et al., 1978; Finkelhor, 1987; Morgan, 1982). Other potent psychoactive agents, such as opiates (including heroin), amphetamines, and cocaine, may be additional pharmacological contributors to abuse.
However, the nature of the relationship of substance use and abuse, different types of substance abuse situations, and the use of violence against children is not well understood. Empirical studies have been biased by a reliance on reported incidents of drunkenness or drug use rather than studying the emerging relationship between such phenomena and child maltreatment as they occur (Pernanen, 1991). Cultural acceptance of the disinhibiting effects of alcohol (MacAndrew and Edgerton, 1969) or drugs has been proposed as one theory that provides an explanation for the breach of social norms and standards involved in child sexual abuse, but theoretical work in this area is in a very early stage of development.
The etiology of emotional abuse and neglect is less developed than that of the other three forms of maltreatment discussed above. However, emotional maltreatment appears to be more prevalent, and some investigators believe that its consequences are more destructive than other forms of child abuse and neglect (Garbarino and Vondra, 1987; Hart and Brassard, 1987; Hart et al., 1987). The relationship between the etiology of emotional maltreatment and other forms of child abuse and neglect is currently not known.
Although we have limited knowledge about processes that link adult and child characteristics and child maltreatment, a considerable research literature on child maltreatment, stress and coping, developmental psychopathology, and normal child development indicates that parental personality characteristics influence child development primarily through the interactive process of parenting. Disrupted parenting can occur in a variety of ways, especially when a parent's personality attributes (such as anger, anxiety, etc.) are compounded by additional stressors such as marital conflict, poverty, unemployment, and having a difficult child (Conger et al., 1984; Hetherington, 1991; Patterson et al., 1992). It is vital, therefore, that scientists examine individual or psychological factors in combination with each other to develop a more comprehensive understanding of their contributions to child maltreatment.
Adult Attitudes, Attributions, and Cognition
Cognitive factors in adults who maltreat children, including negative attitudes and attributions about their children's behavior and inaccurate knowledge and expectations about child development, also play a contributing role in child maltreatment, especially neglect (Holden et al., 1992; Zuravin, 1987). Attitudes held before the birth of the child, such as negative maternal attitude toward an unwanted or unplanned pregnancy, have also been associated with later maltreatment (Altmeier et al., 1982; Brunquell et al., 1981; Egeland and Brunquell, 1979; Murphy et al., 1985; Zuravin, 1987).
Abusive parents may have incomplete or distorted knowledge and understanding of normal child development or their own children's behaviors. The tendency of physical abusers to impart negative attributes to others, including their own children and interpersonal relationships, is associated both with differences in abusive parents' expectations and attributions about children's behavior and with psychophysiological hyperresponsiveness to stimuli. In comparison to nonabusive parents, abusive parents show greater physiological reactivity as well as irritation and annoyance in response to children's positive and negative affective states and behavior (Casanova et al., 1992).
For example, abusive parents, in comparison with nonabusive parents, sometimes perceive their children as more aggressive and intentionally disobedient, annoying, and less intelligent, although other observers fail to detect such differences in the children's behavior (Mash et al., 1983; Reid et al., 1987). In addition, physically abusive mothers perceive their children's negative behavior as a result of stable internal factors such as a
personality trait, but their positive behaviors as a result of unstable external factors. The reverse is true for nonabusive mothers (Larrance and Twentyman, 1983).
Research has been contradictory on abusive parents' knowledge of normal child development as a contributing factor to maltreatment. While some studies have pointed to abusive parents' limited understanding of child development (Disbrow et al., 1977; Spinetta and Rigler, 1972), others have found no significant differences from nonabusive parents (Starr, 1982). Starr (1992) suggests that, even if abusive parents have adequate child development knowledge, they may not apply such knowledge to their childrearing practices.
The absence of studies on how transactions between fathers' and children's characteristics and life circumstances promote or buffer children from the risk of maltreatment is a major gap in the research literature. With the exception of studies of sexual abuse, researchers generally exclude analysis of fathers' attributes or roles within the family or rely on maternal reports of such information, which is a major methodological limitation (Holden et al., 1992). This exclusion results from the difficulties of recruiting fathers into child maltreatment studies.
Intergenerational Transmission of Abusive Parenting
The notion that abused children become abusing parents has received significant attention and has been one of the most pervasive and popular themes in the literature over the past several decades (Cicchetti and Aber, 1980; Kaufman and Zigler, 1987; Kempe and Kempe, 1978; Steele and Pollack, 1968; Widom, 1989).
Two clinicians at the forefront of child maltreatment research in the 1970s observed that "the most constant fact (concerning child abusers) is that parents themselves were nearly always abused or battered or neglected as children" (Fontana, 1973:74, quoted in Belsky, 1992) and that "we see an unbroken line in the repetition of parental abuse from childhood into the adult years" (Steele, 1976:15, quoted in Belsky, 1992).7
The intergenerational hypothesis is controversial because it is supported largely by retrospective analyses. Retrospective studies suggest that the rate of intergenerational transmission is high and that the vast majority of abusing parents were abused as children. For example, Steele and Pollack (1968), in a study evaluating clinical data, found that all 60 abusing parents had been abused during childhood. Retrospective studies indicate a range between 7 percent (Gil, 1970) and 70 percent (Egeland and Jacobvitz, 1984) in the intergenerational transmission of child maltreatment. Kaufman and Zigler's (1987) partial review of the literature estimated a 30 percent rate (plus or minus 5 percent) of intergenerational transmission.
However, retrospective studies of intergenerational transmission are limited by methodological difficulties of definition, design, and reliance on reports of offenders already labeled as maltreaters (Belsky, 1992).8 Inherent limitations of the retrospective approach include the impossibility of determining the proportions of adults who were maltreated as children who have provided adequate care for their children. In addition, the studies do not indicate the extent to which past and current concepts of abuse are in agreement. It is also difficult to assess whether abusive parents may provide distorted reports of their childhood. These factors may lead to an overestimate of the rate of intergenerational transmission.
Other theoretical paradigms (e.g., attachment, social learning) have been generally overlooked in the discussions of the intergenerational hypothesis, with some exceptions (Zeanah and Zeanah, 1989). Such theoretical explanations deserve closer examination in linking types of maltreatment in childhood and the multiple possible pathways to abusive parenting.
The methodological and measurement problems of retrospective studies emphasize the importance of testing the intergenerational hypothesis with longitudinal prospective studies that examine the ongoing caretaking practices of samples of adults who were abused as children (Cicchetti and Aber, 1980; Aber and Cicchetti, 1984). A few prospective studies document a linkage between a reported history of childhood maltreatment and the perpetration of maltreatment but prospective studies generally have significantly smaller rates of transmission associated with the intergenerational hypothesis (Hunter et al., 1978; Egeland et al., 1987). Hunter and Kilstrom's (1979) one-year prospective study of premature infants yielded an intergenerational transmission rate of 18 percent when examined from a prospective vantage point, in contrast to a transmission rate of 90 percent when the same data were examined from a retrospective position.9 Since some studies have shown that ''antisocial behavior patterns are passed from one generation to the next at a rate well beyond chance," particularly when controls for confounding factors such as family size, area of residence, or rates of criminal behavior have been established, the relationship between antisocial behavior and child maltreatment deserves further exploration in intergenerational studies (Huesmann et al., 1984; Wahler and Dumas, 1986:50, quoted in Belsky, 1992).
Several investigators have tentatively identified protective factors that break the cycle of abuse (Egeland, 1988; Egeland et al., 1988; Hunter and Kilstrom, 1979). In retrospective studies, parents with reported histories of childhood maltreatment who do not maltreat their own children are more likely than those who perpetuate the intergenerational cycle to have: (a) better current social support, including a supportive spouse; (b) a positive relationship with a significant adult in childhood or the experience of therapy as an adolescent or an adult; (c) an ability to provide a clear account of their
childhood trauma with anger and responsibility for that abuse directed toward the perpetrator and not themselves.10
In studying why some individuals do or do not perpetuate a cycle of violence, all possible outcomes need to be examined: (1) the maltreated child who maltreats his or her own children; (2) the maltreated child who does not maltreat his or her own children; (3) the nonmaltreated child who maltreats his or her own children; and (4) the nonmaltreated child who does not maltreat his or her own children. Examining the processes and mechanisms of continuity in the first and fourth cells, and of discontinuity in the second and third cells, are important research priorities.
Prospective research on the intergenerational transmission of abuse has not adequately considered the interaction of parental characteristics and children's age and developmental stage (Hunter and Kilstrom, 1979). The one-year duration of the Hunter and Kilstrom prospective study, for example, encompasses only parents who display difficulties in the parenting of infants. Since child maltreatment is not limited to onset in infancy, additional abuse or neglect may be identified at later stages, such as toddlerhood or adolescence (Belsky, 1992).
Research has not demonstrated the transmission of specific types of abuse. For example, mothers who were physically abused as children increase the risk of both physical and sexual abuse for their children (Goodwin et al., 1981). The current studies emphasize the importance of examining factors related to breaking the cycle of abuse in the context of prospective investigations.
Alcohol and Drugs
The often noted association between substance use or abuse and aggression has suggested that the use of alcohol and drugs may be a significant risk factor in abusive families, but the associations among alcoholism, drug use, and child maltreatment are not well understood (for reviews see Hamilton and Collins, 1982; Orme and Rimmer, 1981; Widom, 1992). In particular, the severity and chronicity of intoxication and substance abuse remains poorly documented in studies of child maltreatment (Widom, 1992). Studies of alcoholism among maltreating parents have consistent methodological problems resulting from sample selection and research design.
Hamilton and Collins (1982) concluded that the results were contradictory, with some studies finding a significant relationship and others not (Widom, 1992). A second review published about the same time also found "no empirical data to support an association between alcoholism and child abuse" (Orme and Rimmer, 1981: 273).
More recent studies are based on improved methodologies, but their results are also contradictory. Estimates of the extent of alcoholism among
maltreating parents range from 18 to 38 percent compared with estimates of 6 to 16 percent in the general population (Harford and Parker, 1985; Robins et al., 1984; Widom, 1992). Reports of parental alcohol use by abused psychiatric patients ranged from 30 to 51 percent (Widom, 1992). Only one study has offered strong evidence of a connection between child abuse and alcohol use (Famularo et al., 1986), but authors of this report have noted that limitations in their study design may have affected the incidence of reported alcoholism in their groups (Widom, 1992). One prospective study that compared matched groups of sons of alcoholic and nonalcoholic fathers found no significant differences in the extent of childhood physical abuse (Pollock et al., 1990; Widom, 1992).
Although alcohol often is cited as a principal risk factor in the etiology of child maltreatment, its relationship to child abuse and neglect remains uncertain (Widom, 1992). More needs to be known about the unique and immediate effects of alcohol, its co-occurrence with other problem behaviors such as antisocial personality disorder and substance abuse, the circumstances under which different types of drinking situations lead to or sustain violence against children, and cultural factors that mitigate or exacerbate connections between substance use or abuse and aggression (Abram, 1990; Fagan, 1990; Pernanen, 1991; Robins and Regier, 1991).
The use of drugs during pregnancy, especially cocaine, and their effects of fetal development and the care of infants, has prompted some researchers to examine the relationships among substance use, abuse, and child maltreatment. Such studies are often complicated, however, by the presence of other social and economic variables, such as poverty, that confound the analysis of the contributing role of drugs themselves. At this time the scientific literature on substance abuse and child maltreatment is not well enough developed to allow for inferences by the panel.
Biology and Child Maltreatment
There is no direct evidence that biological factors contribute to child maltreatment, and social and biological scientists tend to agree that the most important influences on aggressive behavior are experiential or environmental. However, efforts to understand biological aspects of aggression, and the role of experience and environment in enhancing or modifying these factors, may contribute to the identification of specific variables and interactive processes that affect the maltreatment of children.
Several studies have examined aspects of child maltreatment and aggression involving both human subjects and animal analoguesmost important, perhaps, rhesus monkeys and other species of monkeys and apes that resemble human behavior in the development of mother-infant and
other social relationships (Suomi, 1978). Biological studies have generally focused on physiological consequences and reactive mechanisms by which childhood victimization can lead to dysfunctional behavior (Lewis, 1992), a topic that is discussed in Chapter 6. Some studies, however, have focused on examples of primate behavior in both field and laboratory environments that demonstrate cases of infant neglect and abuse. Other studies have examined the role of hormones, neurotransmitters, noxious substances (e.g., drugs, alcohol), diet, and abnormalities of brain functions in aggression research in both human and animal models.
Primate studies by Suomi and colleagues suggest that, although some species of monkeys and apes have a capacity to neglect or abuse their young, a valid and useful primate model of human child abuse has not been identified (Suomi, 1978; Suomi and Ripp, 1983). The most extensive data on infant maltreatment by nonhuman primates comes from laboratory settings (Suomi, 1978). The incidence of infant neglect or abuse by biological primate parents in feral environments is relatively rare (Hrdy, 1976).
Stress and social isolation, in particular, show negative effects on primate maternal-child interactions, particularly if stress or isolation occurs during critical periods of development such as the mother's own infancy or birth of her offspring (Suomi, 1978). Rearing experiences may be associated with changes in central nervous system neurotransmitter activity, affecting norepinephrine and serotonin monamine systems (Higley and Suomi, 1989). Primates reared in isolation from their own mothers and peers exhibited inadequate care as well as excessive and inappropriately directed aggression toward their own offspring, especially their first born and male offspring (Suomi, 1978). However, mothers reared in isolation who were given limited exposure (even a minimum of two weeks) to stable social groups during their pregnancy and postnatal period improved their care for subsequent infants, suggesting that socialization can improve the quality of parental care (Suomi, 1978).
The absence of a valid animal model for studies of child maltreatment may be an important barrier to research interventions that would be ethically unacceptable for human subjects. Some causal factors suggested but not proven by the monkey data could be explored in further research, such as the relationship between elevated levels of testosterone in the mother and the increased risk of abuse for male offspring (Suomi, 1978). The contribution of the infant's behavior to its treatment by an inadequate or abusive mother could also be tested experimentally by studies of nonhuman primates through the use of stimulants or depressants. Suomi has noted (1978), however, that the rarity of infant abuse among primates in their natural environments, and the sharply decreasing availability of animal research subjects, suggest that extended experimental studies of infant abuse involving nonhuman primates will not occur unless justified by researchers who
believe that the animal data present important insights into human phenomena.
Beyond the studies of infant maltreatment, the results of aggression research involving both human and animal models are contradictory and inconsistent. No known biological factor in and of itself causes aggression. A changing reciprocity between biological and environmental factors appears to determine whether or not an individual will behave aggressively. As a result, an aggressive temperament may be the reflection of a passing physiological state induced and reinforced by environmental stressors.
Studies of both humans and animals indicate that aggressive behaviors engendered by environmental conditions are often mediated at least in part by physiology. Different parts of the brain continuously interact with each other, and violent or abusive behaviors represent a combination of stimulation and suppression of particular brain areas, past experience and learning, and immediate environmental stimuli or stressors.
Results of studies regarding the relationship between testosterone levels and aggression in humans are equivocal (Ehrenkraz et al., 1974; Kreuz and Rose, 1972; Meyer-Bahlberg, 1974; Monti et al., 1977; Rada et al., 1976). The relationship of endocrine status to behavior is extremely complex and poorly understood, in part because the effects of hormones and their interactions differ from species to species and extrapolations from animal studies are controversial.11 Neurotransmitters, especially norepinephrine, dopamine, and serotonin are involved in both the genesis and suppression of aggressive behaviors.
Clinical data and data from experimentation with animals have shown that aggressive behaviors can be elicited or suppressed, depending on which parts of the brain (the hypothalamus, the amygdala, or the orbital prefrontal cortex) are stimulated or ablated (Bard, 1928; Floody and Pfaff, 1972; MacLean, 1985; Weiger and Bear, 1988). Acetylcholine has been shown to be an important neurotransmitter in the hypothalamus (Bandler, 1970; Bear et al., 1986; Smith et al., 1970).
None of the neuroanatomical or physiological factors considered above, in and of themselves, results in violence or abusiveness. Studies of young offenders (Lewis et al., 1980, 1988, 1989) and follow-up studies of delinquents (Lewis et al., 1989), however, suggest that, when neuropsychiatric vulnerabilities to irritability, oversuspiciousness, impulsivity, and extreme emotionality exist, then parental maltreatment may be a potent catalyst for the child's aggressive behavior.
Children whose ability to function is compromised by virtue of brain dysfunction, brain damage, or psychiatric illness may have difficulty in controlling impulses, distinguishing fantasy from reality, and modulating behavior in response to abusive treatment. Extrapolating from research on animals, maltreatment may modify the physiology of the child, diminish
concentrations in the brain of substances such as serotonin that ordinarily help to modulate feelings, and increase substances such as dopamine and testosterone that enhance hypervigilance and retaliatory aggression in animals.
For these reasons the neurophysiological consequences of maltreatment deserve further study. Furthermore, the biological factors that affect parenting behaviors, especially in primates and other animals closely related to humans, deserve further study.
Evolutionary theory in the form of sociobiology has been applied to studies of child maltreatment (Belsky, 1992; Burgess and Draper, 1989; Daly and Wilson, 1983). Sociobiology holds that parental behavior is influenced by genetic factors and strategies of investment or disinvestment in offspring to enhance the chances of survival of the species (Hrdy, 1976). According to this theory, conditions that amplify biological conflicts of interest between parent and child contribute to child maltreatment; conditions that reduce such conflict prevent its occurrence (Belsky, 1992:47). Although primate studies have documented reproductive and parenting strategies that enhance survival of the species (Hrdy, 1976), the evidence for human behavior is less compelling because the contributions of culture and biology are extremely difficult to disentangle.
Although research on characteristics of maltreating parents has focused on their personality and cognitive features, the risk of maltreatment varies by demographic factors as well. Demographic variables are often of less interest to researchers because evidence of their significance in the etiology of child maltreatment is conflicting, their contribution to child maltreatment can be understood only by analyzing the interaction of individual characteristics and circumstances with other factors, and they cannot be manipulated through treatment or other intervention programs. As a result, demographic variables often acquire importance only in sociological models that require fundamental social reforms (such as reducing the incidence of teen pregnancy) as a basis for intervention.
Reports of specific demographic factors associated with child maltreatment are derived primarily from clinical research, which is subject to reporting and labelling bias. Such reports are inconsistent. For example, some studies associated young maternal age with maltreatment (Benedict et al., 1985; Creighton, 1985; Egeland and Brunquell, 1979; Leventhal, 1981; Leventhal et al., 1993; Whipple and Webster-Stratton, 1991; Zuravin, 1988). Others find no such relationship or suggest that maternal age is confounded by social class (Altmeier et al., 1982; Earp and Ory, 1980; Hunter et al., 1978; Kinard and Klerman, 1980; Leventhal et al., 1993; Murphy et al.,
1985; Oates, 1986). However, when age is operationalized as mother's age at time of abuse there is no relationship, but when age is operationalized as mother's age at time of birth of the abused child, then younger mothers appear to have higher rates of physical abuse (Kinard and Klerman, 1980; Connelley and Straus, 1992). This example illustrates how seemingly small differences in conceptualization and operationalization of etiologic variables can account for discrepancies in the literature.
Abusive and neglectful parents are more likely to be single (Caplan et al., 1984; Holden et al., 1992; Zuravin, 1988), to have a large number of closely spaced children (Belsky, 1992; Holden et al., 1992), and to have a larger family size (Belsky, 1992; Creighton, 1985; Polansky, 1981). They have also been reported to be less educated, but not less intelligent, than nonmaltreating parents (Dubowitz, 1987; Egeland and Brunquell, 1979; Starr, 1982). As discussed below, they are more likely than nonabusive parents to be poor and unemployed (Holden et al., 1992).
Attention to child characteristics associated with maltreatment was stimulated by the recognition that some abused children were abused again in foster homes. Since the prevailing etiological model was a parental pathology model, the abuse of children in foster care directed attention toward characteristics of the child that might provoke anger from both adults at risk of abuse and previously nonabusive adults.
Research on child risk factors associated with maltreatment have included prematurity, temperament, age, and gender. Retrospective research has suggested that factors such as prematurity, low birthweight, and illness or handicapping conditions in the infant or child interfere with attachment and bonding, making the child more vulnerable to maltreatment (Lynch and Roberts, 1977; Oates et al., 1979). Others have found this not to be true and suggest that such findings are more likely due to methodological flaws in study design (Leventhal, 1981).
The identification of child risk factors associated with abuse sometimes generates controversy as to whether child behaviors and characteristics contribute to or are consequences of maltreatment. A focus on the abused child has been thought to be a victim-blaming strategy. Researchers have sought to distinguish between child characteristics that may be causal and those that maintain or perpetuate maltreatment (Ammerman, 1991; Drotar, 1992). In general, little is known about the processes and interactions through which child characteristics and behaviors become risk factors, either as contributing to or maintaining abusive situations.
Child factors are viewed as ones that increase the potential for abuse only if other causal or predictive factors are presentsuch as parental fac-
tors and societal/environmental factorsin a transactional model (Starr, 1992). Both prospective studies (e.g., Egeland, 1988; Hunter et al., 1978) and retrospective studies (e.g., Leventhal et al., 1981) indicate that when researchers control for parental and societal variables (such as social isolation, poverty, substance abuse, or socioeconomic status), child variables such as prematurity and low birthweight do not appear to be major risk factors for child maltreatment, even though such factors have been linked with abuse in some studies involving retrospective designs (e.g., Hunter et al., 1978; Lynch and Roberts, 1977; Oates et al., 1979).
Mothers of low-birthweight infants usually have higher anxiety and depression, negative parenting styles, and less positive interactions with their infants than full-birthweight infants (Beckwith and Cohen, 1984; Brooten et al., 1988; Shosenberg, 1980). Low-birthweight infants are often less soothable, are less responsive, are often perceived as less attractive, and often demand a great deal of care (DiVitto and Goldberg, 1979; Klein and Stern, 1971; Meisels and Plunkett, 1988; Parmelee, 1975). Maternal anxiety and distress, high caretaking demands, and difficulty in soothing low-birthweight infants may account for some abusive treatment (Maccoby and Martin, 1983; Parke and Collmer, 1975). However, since low birthweight is often associated with poverty and low parental education, it may simply represent confounded effects of the cycle of inadequate parenting rather than an independent causal effect.
Infants with difficult temperaments are characterized by high levels of irritability and negative mood, fearfulness, rapid arousal but difficulty in soothing, lack of rhythm in sleep and feeding patterns, and lack of adaptability. In view of these characteristics, it is not surprising that mothers with temperamentally "difficult" children are likely to report more childrearing stress. When mothers of difficult infants experience multiple life stressors and few social supports, both parenting and the bonds of attachment are likely to be disrupted (Goldberg, 1983). What is surprising is that research has not demonstrated that either nurses' or mothers' ratings of infant temperament are associated with maltreatment of infants (Egeland, 1988).
Older children with difficult temperaments, especially boys, are more likely than easy children to be the target of mothers' coercive punitive discipline (Rutter, 1987), especially when the mother is depressed or antisocial, stressed, and has few available supports (Hetherington, 1989, 1991).12
Studies that consider developmental status often focus on the outcomes of abuse at various ages rather than developmental level as a cause of child abuse. However, some parents may be better able to deal with the dependency of an infant than with the combativeness of a toddler or adolescent. Children under age 3, perhaps because of their physical vulnerability, are the most likely to suffer from fatal child abuse (Belsky, 1992). Straus and colleagues (1986) found that toddlers and adolescents were subjected to the
most serious acts of violence and postulated a connection with increased oppositionality in both of these age groups. It is unclear whether there is an onset of abuse at adolescence or whether it represents an ongoing pattern of violence (Farber and Joseph, 1985). More attention needs to be directed to age and gender, and their combination, in increasing or decreasing the vulnerability of children to maltreatment.
Child characteristics may play only a minor role in the initiation of child maltreatment, but they may be important in the maintenance or persistence of abusive relationships or the escalation of at-risk relationships (Ammerman, 1991; Wolfe, 1985). Child characteristics also may be important in reabuse or revictimization. Sexually abused children may develop or learn sexualized behaviors that put them at risk of continued abuse by the original and/or other perpetrators (Frederich, 1988). Similarly, they may engage in sexual intercourse at earlier ages and, if abused by a family member, be poor contraceptors and be more likely to engage in unprotected, high-risk sexual behaviors (Wyatt et al., in press). Toddlers who have been physically abused exhibit aggressive, provocative, and approach-avoidant behaviors with teachers and peers (George and Main, 1979; Main, 1983), characteristics associated with provoking irritability, rejection, aggression, and abuse in others. Indeed, abused children may have learned distorted interaction patterns so well that they evoke similar interactions from both their parents and other social contacts such as teachers (Sroufe, 1983; Dodge et al., 1990).
Factors that trigger child neglect also should be separated from factors that maintain this behavior, especially in the development of infants. In early periods of neglectful behavior, the child may exhibit stressful behaviors in the forms of feeding problems, irritability, or deficits in social responsiveness that place increased demands on the parent's caretaking duties (Powell and Low, 1983; Powell et al., 1987). In some cases, nutritional deprivation, combined with increased maternal detachment, results in nonorganic failure to thrive and sets into motion a "vicious cycle of cumulative psychological risk" (Drotar, 1992:121). Eventually, the parent may begin to perceive the child as quiet, sickly, or not very competent, perceptions that may not be shared by others who observe the child (Ayoub and Milner, 1985; Kotelchuck, 1982).
Child characteristics may be contributing factors rather than independent causes for abuse or neglect. Certain child characteristics, such as low birth weight, prematurity, or non-organic failure to thrive, represent targets for intervention if they are found to be associated with increased maltreatment after the effects of confounding variables have been controlled. However, other factors associated with parental behavior or parent-child interactions appear much more promising sites for intervention at this time.
The Family Microsystem
Building on this review of ontogenic factors within the child, the parent, and the extra-familial offender, we now move to an analysis of significant factors in the family functioning of maltreating families.
Family Functioning in Maltreating Families
Disruptions in all aspects of family relations, not just parent-child interactions, are often present in the families of maltreated children, although it is not clear if such disruptions contribute to or are consequences of child maltreatment. Anger and conflict are pervasive features of maltreating families, although conflict may be more characteristic of abusive families and social isolation may be more associated with neglectful families (Crittenden, 1985). Husbands and wives in maltreating families are less warm and supportive, less satisfied in their conjugal relationships, and more aggressive and violent than those in nonabusive families (Fagan and Browne, 1990; Rosenbaum and O'Leary, 1981; Rosenberg, 1987; Straus, 1980).13 Furthermore, sibling relationships are more conflicted and less supportive in families characterized by high marital conflict or coercive punitive parenting, and tolerance of sibling violence sets the stage for later family violence (Hetherington, 1991; Hetherington and Clingempeel, 1992; Patterson et al., 1992; Straus, 1980). In many cases of maltreatment, there often is not a single maltreated child, but multiple victims (Faller, 1988). Thus, maltreated children may be exposed to considerable violence involving other family members as well as violence directed toward them (Rosenberg, 1987).
In addition, violence and maltreatment are often not confined to the boundaries of the family. Parents with violent, antisocial, or criminal records or those who are aggressive outside the family are more likely to be aggressive in family relations (Patterson, 1982). Similarly, many incestuous offenders do not limit their activities to children within the family. One study indicated that 49 percent of incestuous fathers and stepfathers abuse children outside the family at the same time they are abusing their own children (Abel et al., 1988). The deviant behavior exhibited toward children by maltreating parents is often part of a network of disrupted relationships within the family and in extrafamilial relationships.
Single-parent, particularly female-headed, families are inextricably linked with poverty (Coulton et al.. n.d., 1990a,b), and the contribution of family structure to abuse and neglect is difficult to disentangle from conditions of poverty. Poor, young, single mothers with young children are at the greatest risk of reporting that they use violent behaviors toward their children (Gelles, 1992).
Family structure has been implicated in child sexual abuse in that stepfathers are more likely perpetrators than are biological fathers, and children who have had a stepfather are at greater risk of abuse. The potential for role confusion in reconstituted families and the greater exposure of children to unrelated men as the mother seeks a new relationship have been postulated as contributing to this increased risk (Faller, 1990; Finkelhor et al., 1986; Russel, 1986).
A distinctive feature associated with chronically neglecting families is the chaotic and unpredictable character of the family system.14 One recent examination of the family dynamics and structures of neglectful families concluded that the changeable membership of a neglectful household presents unique challenges for the application of basic principles of family therapy to child neglect (Polansky et al., 1992). Rather than the two-parent family living in a stable location with two or three small children, the neglectful household is often characterized by a shifting constellation of adult and child figures, representing at times desperate efforts by the parent to keep the family together during times of economic and other social crisis. During periods of change, the household of the neglectful mother and her children becomes increasingly fragile, isolated, and detached from adult figures in the neighborhood, church, or other community structures that could offer assistance and support during times of unusual stress or deprivation. The family also may eventually be rejected by relatives and friends, who may show increasing disapproval of the mother's or children's behaviors, especially if substance abuse, delinquency, or other forms of addiction or dysfunctional behaviors are present. The effect on children of repeated fluctuations in the makeup of their household, in addition to child neglect, has not been examined in the research literature, although such changes are suspected to contribute to unrelatedness and detachment (Polansky et al., 1992).
Family relationships that affect the quality of parent-child interactions have also been considered in developing the context for understanding child sexual abuse (Finkelhor, 1984; Hartman and Burgess, 1989). Such factors include an estranged family; one in which the victim is closest to no one individual (Wyatt et al., in press); a mother who is absent, ill, or otherwise not protective of the child; social isolation of the family; lack of supervision of the child; unusual sleeping or rooming conditions; the erosion of social networks; and the lack of social supports for the mother (Finkelhor, 1984). Factors in the child's behavior, education, and relationships have also been considered, including the emotional security or social isolation of the child; knowledge about sexual abuse (which may be affected by school sex education programs); the relationship of trust between the offender and the child; and coercion.
Although a substantial research literature describes the effects of parenting styles on child adjustment, few attempts have been made to link this body of knowledge with child maltreatment research (Baumrind, 1992). Research on parenting styles (Baumrind, 1989, 1991; Dornbusch et al., 1987) has identified two styles that have relevance for studies of maltreatment: (1) a neglecting/disengaged style involving low involvement, nurturance, warmth, control, and monitoring, and (2) an authoritarian style involving punitivness, coercion, restrictiveness, and low warmth and support. The first appears to be related to child neglect and the second to child abuse (Maccoby and Martin, 1983). Although most studies of parenting styles have not been based on clinical samples of parents or of parents identified as maltreating, the outcomes reported for children who experience these types of parenting styles are similar to those reported for maltreated childrenhigh rates of aggression, antisocial behavior, depression and anxiety, and problems with peers, in school, and in intimate relations. These observations suggest that parenting dysfunction is directly related to maltreatment and that examinations of different levels of parental dysfunction would be beneficial, particularly when certain behaviors have adverse effects on the well-being of children. This issue is directly relevant to the next section on the relationship between physical punishment and maltreatment.
Few comparative studies have examined variations in parenting styles and dysfunctional parenting patterns (including abuse and neglect) and their effects on children in different cultural or ecological settings. For example, authoritative parenting stylesinvolving high levels of responsiveness, warmth, control, monitoring, communication, and demands for mature, responsible behaviorare most accepted and successful in promoting competence in white middle-class children. But the authoritative parenting style may not be equally advantageous for other groups (e.g., poor families) and in other situations, especially for families that live in communities characterized by violence.
An important gap in the literature on child maltreatment is the lack of comparative analysis of risk factors for child maltreatment for families in different social, ethnic, and cultural groups. One study has investigated social class differences in the etiology and consequences of child maltreatment (Trickett et al., 1991). Such research could identify different cultural patterns of transactions between risk potentiating and protective factors in the etiology and outcome of child maltreatment. For example, some Latin-American and Asian communities value close adult supervision into late adolescence, a cultural practice that might serve an important protective role while also being viewed as dysfunctional or intrusive by other ethnic groups (Brown et al., 1992). Standards of parental control and discipline
may also be conditioned by cultural factors, especially in environments characterized by danger and violence.
The few existing studies on parenting styles and child maltreatment identify important distinctions between maltreating and nonmaltreating parents, but within maltreating samples the behavior of neglectful and abusive parents indicates that the etiology of physical abuse and neglect shares more similarities than differences (Pianta et al., 1989).
Observational studies have indicated that in comparison to nonabusing parents, physically abusing parents have less pleasant interactions with their children even if not always more negative.15 Abusing parents are less supportive, affectionate, playful, and responsive with their children (Burgess and Conger, 1978; Egeland et al., 1980; Kavanaugh et al., 1988; Reid et al., 1987; Trickett and Susman, 1988; Twentyman and Plotkin, 1982). Even with infants, abusive parents are more controlling, interfering, and covertly, if not overtly, hostile (Crittenden, 1981, 1985).
Moreover, aversive behavior in abusive families is more likely to be reciprocated (Lorber et al., 1984), with escalating acrimonious exchanges of longer duration than those in nonabusive families, described as a coercive cycle by Patterson (1982). Abused children, as a consequence of abusive and inept parenting, may eventually develop characteristics (including brain damage) that make them disagreeable and difficult to manage, resulting in greater involvement in hostile, coercive cycles and an increased risk of abuse.
For chronically neglectful families, child neglect is not a single form of poor hygiene or inadequate nutrition in an otherwise well-ordered family unit, but it exists as a broader aspect of household disorganization, insularity, and lack of cognitive stimulation and emotional nurturance in the household (Polansky et al., 1992). Neglectful parents are unresponsive both with infants (Crittenden, 1981, 1985) and with older children (Burgess and Conger, 1978). They tend not to initiate interactions and not to respond to the initiations of their children; they exhibit little prosocial behavior toward their children; and, under some circumstances, they may actually exhibit more negative behavior than abusive parents (Burgess and Conger, 1978).
Abusive and nonabusive parents usually, but not consistently, exhibit differences in control and disciplinary practices. Abusive parents are more likely to use punishment, threats, coercion, and power and they are less likely to use reasoning and affection in controlling their children (Lorber et al., 1984; Trickett and Sussman, 1988). Moreover, their discipline is less likely to be contingent on the type of behavior exhibited by the child (Crittenden, 1981; Trickett and Kuczynski, 1986). The more frequent use of coercive, physical, punitive disciplinary techniques by abusive parents appears to reflect their negative perceptions of their children's behavior and their be-
lief in the effectiveness of power assertion in controlling it (Milner and Chilamkurti, 1991).
The relationship between physical discipline and child abuse is one that elicits strong beliefs and opinions, but few definitive conclusions. Societal tolerance and use of physical discipline have been suggested as risk factors in child maltreatment (Gil, 1970; Kadushin and Martin, 1981; Straus, 1980; Zigler and Hall, 1989). Acceptance of physical discipline or punishment of children, including infants, is widespread among lay persons (Carson, 1986; Gelles and Straus, 1988; Gil, 1970) and physicians (McCormick, 1992).
Physical discipline has been viewed as one end of a continuum of abusive behavior, but the linkage between physical discipline and abuse has not been established (Gelles, 1991). International and cross-cultural comparisons also do not support an inevitable tie between physical discipline and physical abuse. Sweden, which has laws prohibiting physical punishment, was significantly lower than the United States on self-reports of physical punishment on the Conflict Tactics Scales (Gelles and Edfeldt, 1986). However, there was no significant difference between the United States and Sweden in severe, or abusive, violence against children. Kadushin and Martin (1981) have pointed out that, while most discipline does not become abuse, retrospective reports suggest that most abuse begins with parental intentions to discipline the child. At least two studies also show that severe punishment that falls below the threshold of reported abuse appears to occur at the same rates as child maltreatment (Egeland, 1988; Straus and Yodanis, 1994).
Parke and Collmer (1975) suggest that physical discipline may be most dangerous among parents who disapprove of its use, since it is then used as a method of last resort, when parental anger is highest. They speculate that ''As a result of this cultural shift in attitude [away from the use of physical discipline], the manner in which physical punishment is employed makes the contemporary use of this type of discipline potentially more dangerous than in the past." (p. 27).
Physical discipline, or at least harsh discipline, may be involved in the intergenerational transmission of abusive parenting. Harsh disciplinary practices have been causally related to a maladaptive style of processing social information that then leads to aggression, suggesting that severe discipline fosters more aggressive behavior by the child (Weiss et al., 1992).
The evidence on cultural differences in the use of physical discipline, and its relationship to abuse, is mixed. Self-report data on the Conflict Tactics Scale (CTS) did not exhibit a significant difference between blacks and whites, although parents in an "other category" (including Native American
Indians, Asians, and "other" cultural groups) were most likely to exhibit violent acts toward children (Straus, 1980). The lack of black-white differences supported earlier literature that blacks are not more violent toward children than whites (Billingsley, 1969), and that blacks and whites do not differ in self-reports of spanking (Stark and McElvoy, 1970). Weller et al. (1987) also found that Anglo and Hispanic adolescents did not differ in the proportion of each group reporting the experience of physical punishment.
Stressful Life Events and Child Maltreatment
Several investigators have examined the relationship between stressful life events and parenting outcomes, including quality of mother-infant attachment (Crockenberg, 1981) and child maltreatment (Egeland et al., 1980; Straus, 1980).16 Most research has compared maltreating and nonmaltreating families on a scale that typically consists of a checklist of stressful life events, such as loss or reduction in family income, sickness in family, moves, and death or loss of family member.
The relationships among stressful life events, the characteristics of people affected by them (particularly anger, hostility, and depression), and the role of stress as an etiological factor in child maltreatment are not well understood. Traits of negative affectivity may help generate or exacerbate stressful life events themselves, thereby contributing to the maltreatment process. Stress seems to aggravate the level of conflict among family members. The effects of stressors on parental abilities depend on their overall coping strategies, the availability of support, and other buffering factors. One study has suggested that mothers who were subjected to high rates of stressful events were inconsistent in matching their disciplinary actions to the behavior of their children, in contrast to mothers experiencing low levels of stress (Dumas and Wahler, 1985).
In a prospective study of the antecedents of child maltreatment, Egeland et al. (1980) compared families who had been reported for maltreatment with a subsample of high-risk mothers who provided adequate care. They found a significant difference between stressful life event scores, although the relationship between stress and child maltreatment was far from perfect. Comparing the predictive value of family stressful life events with other variables, a number of interactional and maternal characteristics were better predictors than family stress of membership in the maltreatment and nonmaltreatment groups. Egeland et al. (1980) found that many high-stress mothers were providing adequate care, and that many mothers experiencing low stress were maltreating their children. The high-stress mothers who maltreated their child were often more angry, suspicious of others, and highly anxious compared to high-stress mothers who had low scores on
these variables. Abuse was also more likely to occur among high-stress mothers who had poor quality interaction with their child than among high-stress mothers with good-quality interaction.
The complex relation between stressors and parenting outcomes, including child maltreatment, needs more attention. The exceptions to predicted outcomes, such as why some high-stress parents provide adequate caregiving and why some families experiencing low levels of stress maltreat their children, are fertile areas for future research. The relationships between stress, poor parenting outcomes, family dysfunction, and child maltreatment also need to be clarified.
Family functioning in maltreating families is affected by combinations of factors that are present in many families that do not maltreat their children. Different levels of parental dysfunction need to be studied over time in order to identify conditions that give rise to the emergence of child maltreatment. Although a parent's own history of victimization during childhood is thought to be highly correlated with child maltreatment, this association is based on retrospective studies that are methodologically flawed. The relationship between physical discipline and child maltreatment is also largely unknown, particularly in terms of cultural differences and practices. Finally, stressful life events are thought to play an important role on individuals, abilities to parent, but relations between stressors and parenting outcomes are complex and poorly understood at this time.
Individual and family characteristics do not function in isolation from the larger community that surrounds both maltreating and nonmaltreating families. Although research on exosystem factors has concentrated on neighborhood and community environments, other factors may affect individual or family functioning as well, including the workplace, the media, the school, the church, and peer groups.
One of the most significant research developments in the past decade has been the recognition of the importance of viewing family functioning in the context of various social institutions and external forces that govern family and parent-child behaviors. We now examine the interactions and experiences between families and family members with networks of these extrafamilial social systems, termed the "exosystem."
In 1990, one in five of all American children, approximately 13.4 million, lived in families with total incomes below the poverty level (e.g., $13,254 for a family of four, U.S. Bureau of the Census, 1991). The rates are higher in families with children age six and under than in those whose children are ages 7-16 (23 percent versus 19 percent), indicating that almost one in four preschoolers lives in a poor household at any point in time. Research has shown that living in poverty exacts a toll on children's well-being (Chase-Landale and Brooks-Gunn, in press; Chase-Landale et al., in press; Huston, 1991).
Discussion of the relationship of poverty to child maltreatment has persisted since publication of the early professional papers on child abuse (e.g., Gil, 1970; Gelles, 1983, 1992; Kadushin, 1976; Kempe et al., 1962; Pelton, 1978; Steele and Pollock, 1968). Although child maltreatment is reported across the socioeconomic spectrum, it is disproportionately reported among poor families. Further, child maltreatmentespecially child neglectis not simply concentrated among the poor, but among the poorest of the poor (Giovannoni and Billingsley, 1970; Pelton, 1981; Wolock and Horowitz, 1979, 1984). Whether this association results from greater stress due to poverty-related conditions that precipitate abuse or from greater scrutiny by public agencies that results in overreporting continues to be debated. Nevertheless, the deleterious effects of poverty on children and their families is well documented (Baumrind, 1992).
Self-reports on the Conflict Tactics Scales indicate that lower socioeconomic status is a risk factor for violent behaviors toward children (Straus, 1980; Gelles and Straus, 1988). Although violence toward children occurs in all social strata, violent behaviors toward children, particularly severe violence, is more likely in poor families. Furthermore, although maternal age by itself is not a significant risk factor for child maltreatment, mothers with young children living below the poverty line are at the greatest risk of violent behavior toward children (Connelly and Straus, 1992; Gelles, 1992).
Poverty has been highly correlated with both physical abuse and neglect. Polansky et al. (1981), however, argue that, although poverty often pervades the lives of neglectful families, it is not the major cause of neglect. In contrast, reported cases of child sexual abuse do not seem to occur disproportionately among the poor, but appear to be concentrated in the middle-class (Finkelhor, 1986). However, this finding may be flawed by the selection of samples for study, since poor children may have limited access to the treatment services that often provide the pool of subjects for this kind of research.
Despite general agreement in the literature on a linkage between poverty and maltreatment (both official reports and self-reports), the processes and mechanisms involved in this association require further research. Why are all poor families not equally at risk for maltreatment? Why does maltreatment occur in families that are not poor? How does economic deprivation translate into child abuse and neglect?
The link between unemployment and maltreatment is significant in understanding the relationship between poverty and maltreatment. While Gil (1970) found child maltreatment to be highly linked with poverty, Light's (1973) reanalysis of Gil's data found unemployment to be the most powerful predictor of child abuse and neglect. The relationship between unemployment and maltreatment has been documented in several research studies (e.g., Gabinet, 1983; Gelles and Hargreaves, 1981; Krugman et al., 1986; Whipple and Webster-Stratton, 1991). Steinberg et al. (1981) used an aggregate longitudinal approach, replicated in two distinct metropolitan communities, to demonstrate that increases of child abuse are preceded by periods of high job loss, consistent with the hypothesis that unemployment can cause family stress, subsequently resulting in child abuse.
The relationship between unemployment and violent behaviors toward children is strongest for fathers, and for fathers employed part time, possibly due to higher frustration than the totally unemployed (Straus, 1980). The association between unemployment and maltreatment has also been documented in studies of individuals (Gabinet, 1983; Krugman et al., 1986; Whipple and Webster-Stratton, 1991) and communities (Bycer et al., 1984; Lichtenstein, 1983; Steinberg et al., 1981). Although these studies have demonstrated such an association, the mechanisms (such as stress) by which a job loss or lack of employment may stimulate child maltreatment have not been identified.
Poor neighborhoods differ in their social and physical conditions and in their ability to influence the specific risks posed to children by poverty, unemployment, drugs, and community violence (Coulton et al., 1990a,b; Coulton and Pandey, 1992; Duncan and Aber, in press).17 Garbarino and colleagues found that, although socioeconomic conditions have predictive value for explaining child maltreatment rates, some neighborhoods have higher child abuse rates than would be expected (high-risk neighborhoods) and some have lower rates than would be expected (low-risk neighborhoods) based on socioeconomic conditions alone. Child abuse rates were
found to be higher in poor neighborhoods with fewer social resources than in equally economically deprived neighborhoods where social resources were perceived to be higher. Parents in the high-risk neighborhoods did not use resources in a preventive fashion, but in response to crises; did not use informal resources such as scouting or youth groups; and often fell back on formal public agencies when intervention was necessary. Parents tended not to exchange services and, when such exchanges occurred, abusive parents attempted to exploit others. Residents in poor but low-risk neighborhoods built better environments, made constructive use of resources, and perceived quality of life and goodness of the neighborhood as a place to rear children more positively than those in high risk neighborhoods (Garbarino and Crouter, 1978; Garbarino and Sherman, 1980). While arguments for neighborhood impact on child abuse and child neglect are compelling, Polansky (1981) notes that neglectful families often are socially isolated and their perceptions of themselves, their children, and others do not reflect attitudes prevalent in their neighborhoods.
Although neighborhoods are recognized as important in the ecology of child maltreatment, more insight is needed into the processes by which neighborhood conditions and factors affect family processes in general and child maltreatment in particular. Why do some maltreating families remain socially isolated, yet others can be motivated to take advantage of community or neighborhood resources? What neighborhood resources reaffirm or weaken destructive beliefs about empathy, trust, and self-esteem? What key neighborhood factors have the ability to mitigate the contributions of poverty, unemployment, and young maternal age toward child maltreatment?
Social isolation has been identified as an important etiological risk factor in child maltreatment, but its role as a consequence or cause of maltreatment is uncertain (Polansky et al., 1981, 1992). Abusive parents have been reported to isolate themselves from others, and also to isolate their children from friendship networks (Young, 1964). The presence and use of social networks has generally been regarded as a protective factor against child maltreatment (e.g., Garbarino, 1977). Two important issues require further research attention. First, abusive parents may be isolated by those around them as a consequence of deviant parenting attitudes and behaviors of abusive families that cause neighbors and acquaintances to avoid the family. Second, research attention needs to be directed toward maladaptive qualities of networks. Abusive parents may select networks of individuals who condone their maladaptive parenting styles, and the networks thereby increasing the likelihood of continued abuse rather than acting as a protective factor.
The relationship between child maltreatment and the social networks of organized religion is not well understood. On an individual level, parental
religious preference has not demonstrated a clear relationship with physical (Straus, 1980) or sexual abuse. Little is known about the relationship of religion to either neglect or emotional abuse. Although religious beliefs favoring harsh discipline of children have been suggested to contribute to physical abuse, religious affiliation and participation have been identified as countering the social isolation associated with maltreatment (Garbarino, 1977) and as an important influence in protecting against child maltreatment, particularly among African-Americans (Giovannoni and Billingsley, 1970).
The influence of family ties and organizational affiliations (including employment and education) are poorly understood but increasingly recognized as powerful forces in shaping parenting styles and family functioning. Financial stability, employment, alcohol and drugs, and neighborhoods can create a context that either supports a family during periods of stress or enhances the potential for abuse. Families reported for abuse often have multiple problems; and the abuse may simply be a partor a consequenceof a broader continuum of social dysfunctions.
The last system in our framework of analysis is the set of cultural and social values that pervade and support individual and family life styles and community services in today's society. The macrosystem is the often invisible layer in theoretical models of child maltreatment, yet its influence is increasingly recognized as important in understanding the hidden forces that govern personal and institutional behaviors.
Social and cultural factors can foster or mitigate stress in family life and such factors have achieved new importance in emerging theoretical models of child maltreatment. These values are thought to play a significant role in affecting adult reliance on coercion and violence to control the irritating daily events associated with family stress (Wolfe, 1991).
Cultural and Social Values
The macrosystem encompasses broader cultural and societal values that contribute to our understanding of the etiology of child maltreatment. Two primary topics have attracted research interest in this fourth level of our conceptual model. First, are broader societal and cultural values in American societysuch as the privacy of the family or a cultural preoccupation with violencecontributing to child maltreatment? And second, how does
culture influence the causes of or responses to child maltreatment, especially in a multicultural society such as the United States?
Although the relationship of these factors is not well understood, some cultural and societal values in American society may contribute to child maltreatment. Racism, for example, can lead to an inequitable distribution of resources, education, and employment that undermines many ethnic minority families' abilities to support their children financially and emotionally and to provide parental care. The term societal neglect has been suggested to characterize American tolerance for a situation in which one-fifth of all preschool children live below the poverty line, with a substantially higher rate among ethnic minorities (Baumrind, 1992; Children's Defense Fund, 1991). Societal fascination with violence, including violence toward children, also has been suggested as a risk factor for physical abuse (Gil, 1970; Gelles and Straus, 1988). The lack of coherent family leave and family support policies, such as those in place in many European countries, increases the difficulties faced by many parents (National Research Council, 1990; Ferber and O'Farrell, 1991). In particular, the absence of preventive health care for infants and children increases the risk that abused children will remain undetected or that health problems will contribute to maltreatment. These broader societal and cultural factors should be considered in cross-national and cross-cultural comparisons.
The disproportionate representation of reports of maltreatment of children in ethnic minorities stimulated a search for cultural practices that contribute to abuse and neglect. A major problem with the research literature is that specific cultural practices have not been linked to differential rates or types of maltreatment, but instead post hoc explanations have been sought in cultural patterns. Pathways and mechanisms for a relationship between culture and maltreatment are thus left unclear.
Parents who abuse children are generally behaving outside their culture's acceptable continuum rather than exercising abusive behavior toward children that is culturally sanctioned. In assessing methods of health care, discipline, spiritual beliefs about health, and other aspects of life that are culturally determined, researchers need to obtain information about which practices the populations under study perceive as acceptable due to their culture, their current life circumstances, or the behaviors of the child. Some behaviors that are thought to be cultural from an outsider's perspective are familial practices or are artifacts of circumstances encountered by the families. The continuum of cultural acceptability within and across populations and the reasons for individual violation of these norms have not yet been determined. Greater effort must be directed at differentiating culturally acceptable behaviors from individually deviant behavior, and toward determining whether the interaction of causal factors varies across population groups.
In exploring the relationship between culture/ethnicity and child maltreatment, the cultural variable must be "unpacked" (Tharp, 1991). Since culture is not monolithic, it does not have a uniform impact on all members. The few studies that have considered intracultural variability have identified factors of interest for future research, such as extended family involvement (Dubanoski, 1981), and rural-urban residence (Lauderdale et al., 1980).
Child sexual abuse does not appear to be disproportionately represented in any cultural/ethnic group. In earlier studies of child sexual abuse, ethnic minority families were overrepresented (DeFrancis, 1969; Peters, 1976). However, these studies employed clinical samples from medical or social service settings in which poor and ethnic minority children were likely to be overrepresented. Survey research on the prevalence of child sexual abuse has not supported the existence of racial or ethnic differences in intrafamilial child sexual abuse (Finkelhor, 1984; Russell, 1986; Wyatt, 1985). Furthermore, the two National Incidence Studies have not found a relationship between race and sexual abuse (NCCAN, 1981, 1988).
The causal relationship of cultural factors and child sexual abuse requires further research within and across populations. For example, the practice of parent and child sleeping together may be viewed as a potentiating factor in some populations, according to the preconditions model set for by Finkelhor (1984), but this custom may prevent child sexual abuse in others. Anecdotal evidence suggests that some Native American groups believe that sexual abuse increased when families stopped sleeping together. The dispersal of children to their own rooms provided the conditions of secrecy conducive to abuse that would have been prevented if all adults had been present (Scheper-Hughes, 1987).
Searching for transactions or pathways that cause maltreatment across and within cultural groups has significant potential to improve understanding of the etiology of maltreatment. More punitive or restrictive approaches may be necessary under some circumstances to promote competencies and to protect children from danger in settings such as inner cities (Ogbu, 1981). Moreover, effective parenting styles may differ for different domains of adjustment in different groups. Studies by Steinberg, Dornbusch, and their colleagues have demonstrated that, although authoritative parenting is associated with low levels of problem behavior in all ethnic groups, it is associated with academic competence in white but not black children; in addition, restrictive authoritarian childrearing is associated with low academic competence in white but not Asian children (Dornbusch et al., 1987; Steinberg et al., 1990a,b).
The relationship between pornography and child maltreatment, particularly child sexual abuse, has been examined only sporadically in the research literature. Most of the research has focused on the relationship between pornography and adult sexual behavior. Pornography is rarely a primary or direct motivator of violent sexual acts, although it may indirectly influence child sexual maltreatment (Knudsen, 1988). The 1970 report of a presidential commission declared that pornography does not have any significant, harmful social effects (Mason, 1989). In 1986, however, the attorney general's Commission on Pornography identified several areas of "pronounced harm" due to pornography, including a link between pornography, sexual violence, and child abuse (Mason, 1989), but did not support these findings with empirical evidence.
The fascination of pedophiles with child pornography and erotica is documented by many arrests of pedophiles who possess large amounts of sexually explicit materials involving children (Schetky, 1988; U.S. Senate, Committee on Governmental Affairs, 1986; West et al., 1978). Children who are sexually abused are sometimes further victimized by their involvement in child pornography. Although individual reactions to pornography are highly varied, adolescents may be particularly impressionable viewers of pornography (Marshall, 1989). Pornography has been reported to increase interest in deviant sexual practices in more than 39 percent of offenders (Abel, 1985, cited in Marshall, 1989). Repeated exposure to pornographic materials can alter perceptions toward sexuality and relationships, increase acceptance of physical force, and lessen compassion for victims (Knudsen, 1988). Yet the temporal sequence is unknowna sexual interest in children or repeated victimization experiences may stimulate interest in sexually explicit materials, and an interest in child pornography may follow rather than precede a sexual interest in children (Crewdson, 1988).
Summary Of Etiological Factors
Many factors have been identified as contributing to the occurrence of child maltreatment, but single-factor or unicausal theories of child maltreatment have not been able to identify specific mechanisms that influence the etiology of child maltreatment. Environmenal factors such as poverty or unemployment and individual characteristics such as a prior history of abuse, social isolation, or low self-esteem have been significantly associated with child maltreatment offenders, but the relationships among such factors are not well understood in determining the origins of child maltreatment. The panel believes that the etiology of maltreatment involves a complex interactive process, one that includes constellations of variables that interact along
various dimensions of a child's ecological/transactional system. Potentiating factors that increase risk for maltreatment and protective factors that decrease the likelihood for maltreatment are found at all ecological levels and interact to produce child maltreatment. Although theoretical models that describe the etiological complexity of maltreatment have been developed, they have not been subjected to testing and adequate research. Our recommendations seek to address these limitations.
Recommendation 4-1: Research using multivariate models and etiological theories that integrate ecological, transactional, and developmental factors will improve our understanding of the causes of child maltreatment. Rather than focusing on specific factors (such as depression, unemployment, or history of abuse), the interactions of variables at multiple ecological levels should be examined.
Although considerable research has focused on single-risk factors that contribute to child maltreatment, we know very little about mechanisms or processes by which these factors lead to maltreatment. Research designs and analytic strategies should focus on the multidimensional character of child maltreatment using a framework that accommodates risk and protective factors in an interactive model, examining possible etiological factors in child maltreatment in combination with other contributing agents. Continued reliance on univariate models or isolated risk factors in future research will not be productive. Process-oriented studies are needed to identify mechanisms by which factors interact in promoting or moderating child maltreatment.
Rather than endorsing one approach, the panel recommends that diverse theoretical models and research strategies be developed at this juncture. The capabilities of emerging multisystem interactional models offer significant promise in the study of child maltreatment that deserve to be supported.
In recommending the development of complex models and etiological theories, the panel has identified several key factors that deserve further emphasis at this time:
Research should clarify distinctions between long-term chronic factors and immediate precipitating factors associated with maltreatment. There is a need to support research that would distinguish between factors associated with the initiation of child maltreatment and factors that sustain or escalate it.
The intergenerational transmission of child maltreatment is an issue
that deserves explicit analysis to determine the circumstances under which high- and low-risk circumstances will lead to child abuse and neglect. Prospective studies are needed to identify not only which parents do or do not maltreat children under conditions of high risk (based on their own histories of child maltreatment), but also those that do or do not maltreat under conditions of low risk. Exceptions to expectations under high- and low-risk conditions will improve our understanding of protective and risk potentiating factors. In addition, evaluations of samples of mothers from abusive and nurturing backgrounds should be conducted to determine the conditions that contribute to supportive parenting. Future research should explore the interaction of risk factors based on the concepts of contemporary attachment theory and inner working models as well as models based on cognitive learning skills and other approaches.
Problems experienced by families characterized by multiple problems (such as poverty, unemployment, and violent neighborhoods) need to be better understood in clarifying the role of specific factors, such as alcohol or substance abuse, in the etiology of child maltreatment. We currently know very little about the conditions under which these factors interact with different individual and ecological factors to produce maltreatment. In particular, the role of alcohol as a mitigating factor (including cultural tolerance of drunkenness as an excuse of violence against family members) requires substantial exploration in child maltreatment studies. Relationships between acute intoxication, chronic alcoholism, and various forms of child maltreatment, including physical and sexual abuse, neglect, and emotional maltreatment, require systematic examination and analysis.
The operation of social relationships and social networks is a critical feature in the etiology of child maltreatment. Research that seeks to clarify the conditions under which social networks can serve as risk or protective factors in child maltreatment should be supported. In particular, the roles of fathers and other male figures as offenders and protectors need more clarification and analysis. With the exception of sexual abuse, most research on child maltreatment focuses on mothers as agents of maltreatment. More information is needed on the role that fathers, siblings, grandparents, stepparents, and other household members play as risk or protective factors in child maltreatment.
Recommendation 4-2: Similarities and differences among the etiologies of different forms of child maltreatment should be clarified in order to improve the quality of future prevention and intervention efforts.
Studies that identify the etiologies of different types of maltreatment, and studies that explore diverse patterns of risk and protective factors and
pathways leading to the same and different forms of maltreatment, should be supported. Multiple forms of maltreatment often co-occur in various combinations (physical abuse and neglect, sexual abuse and emotional maltreatment, and so forth), so it is particularly important at this time to emphasize key pathways for child victimization that may be amenable to prevention or other forms of intervention. Research is needed to clarify the most common types of maltreatment with some degree of specificity and then examine these types along a broad continuum of severity. For example, being a victim of physical abuse or emotional maltreatment may be a significant risk factor for child sexual abuse (or vice versa), but the relationships among multiple forms of maltreatment remain unexamined.
The serious and destructive nature of emotional maltreatment has not been fully recognized by researchers and practitioners in child maltreatment studies. Links among verbal abuse, physical punishment, and physical abuse are not well understood. It is not yet known whether physical punishment and emotional maltreatment are part of a progression leading to other forms of maltreatment. We also do not know if the origins of mild, moderate, and severe forms of abuse are similar or different. Continuity or discontinuity in the emergence of emotional maltreatment, physical abuse, and sexual abuse deserves examination.
Recommendation 4-3: Studies of similarities and differences in the etiologies of various forms of maltreatment across various social class, cultural, and ethnic populations should be supported.
The effects of risk-potentiating and protective factors on child maltreatment in diverse social class, cultural, and ethnic populations have not been adequately explored. Researchers have often relied primarily on clinical populations or subjects who have already been identified as offenders as representatives of entire cultural groups. Samples that are more representative of the diversity of contemporary American society are necessary to improve research quality. More needs to be known about ''normal" forms of physical discipline, sexual behavior, and parenting styles within various social class, cultural, ethnic, and residential subgroups, because cultural norms have an impact on child maltreatment. Research is also needed to identify culturally acceptable behaviors that may inhibit the healthy development of the child. Research must address both commonalities and diversity among populations in the pathways and transactions of variables that promote or prevent various forms of maltreatment.
1. A comparison might be made with other fields of study. Research on child maltreatment has often been grouped collectively because of the child status of the victim. Many research sponsors and investigators assume that child physical and sexual abuse, emotional maltreatment, and neglect have more similarities than differences. However, this organizational principle has not been rigorously tested. One could argue, for example, that comparing the etiology of child physical and sexual abuse is similar to comparing the etiology of burglary and rape of adult women. There may in fact be more differences than similarities in the origins of the offending behavior. The four categories of child maltreatment may have quite distinct etiologies and consequences and thus require distinct treatment and prevention efforts.
2. These models were derived from retrospective studies, including early psychiatric models focused on psychopathology, deviance, or psychological or behavioral deficits in abusive parents. Contrasting sociological models soon appeared that emphasized contributions of the social and physical environment. Models based on child attributes also emerged following reports of abuse of children placed in foster care.
3. See, for example, the definitions of terms as presented in A Dictionary of Epidemiology (Last, 1988).
4. Discussions of adult characteristics often focus on factors associated with the parent or caretaker, often the mother. In this framework of analysis the panel uses the term adult, but it should be recognized that the adult usually has a special relationship with the child, either as a parent, caretaker, or trusted authority figure. In cases of sexual abuse, however, the adult may be someone unrelated or unknown to the child.
5. Predictive instruments based on parental personality profiles have been characterized by limited sensitivity and specificity as well as an inadequately validated profile of personality characteristics or perpetrator typology (Wolfe, 1985).
6. Classifications of sex offenders include: preferential versus situational, incest versus pedophilia, and homosexual versus heterosexual. The distinction between preferential (or fixated) molesters and situational (or regressed) molesters (Groth et al., 1982), based on a prison population, is the classification generally regarded as significant (Prentky, 1990; Lanning, 1987), although community-based therapists often find that many offenders have combined features from both categories (Hartman and Burgess, 1989).
Incest is generally defined as sexual contact between people who are biologically related. Distinctions between incest offenders and pedophiliacs are based on certain core assumptions: (1) that incestuous fathers do not act sexually outside the home; (2) that incest is an expression of nonsexual needs; and (3) that other family members contribute psychologically to the development and maintenance of sexual abuse (Conte, 1991; Hartman and Burgess, 1989). However, increasing research evidence challenges the separate classification of incest offenders and pedophiles, suggesting that many incest offenders are sexually aroused by children, that many incestuous fathers are abusing children outside the family, that sexual and nonsexual dimensions are present in all sexuality; and that consistent profiles of incest families do not emerge from empirical studies (Conte, 1985; Conte et al., 1986). For example, one study indicated that 49 percent of a group of incestuous fathers and stepfathers referred for outpatient treatment abuse children outside the family at the same time they are abusing their own children (Abel et al., 1988).
The distinction between homosexual and heterosexual child molesters relies on the premise that male molesters of male victims are homosexual in orientation. Most molesters of boys do not report sexual interest in adult men, however (Conte, 1991).
7. Steele, a prominent and early proponent of the intergenerational hypothesis, did not claim that the cycle was inevitable, even though present in the history of virtually all identified maltreating parents in his clinical samples. Protective factors, particularly a supportive adult in the life of the child, could break the cycle (Steele, 1976).
8. A retrospective approach examining the intergenerational hypothesis is limited by its reliance on autobiographical memory (Bradburn et al., 1987). As adults, children who were abused may have repressed memories of childhood abuse, may be unwilling to talk about such painful memories, or may have accepted their mistreatment as normal, and even acceptable, childrearing behavior.
9. The low figure may have been affected by following the families for only one year.
10. Egeland et al. (1988) interpret the relationship differences between the continuity group and those who broke the cycle of abuse within a framework of contemporary attachment theory. Their analysis stated that the early caretaker-infant attachment relationship is a prototype of an individual's later relationships and influences the capacity for later affectionate bonds. The inner working models of abused individuals, who did not have a supportive adult figure, may be vastly different from the relationship expectations of the exception group, who appear to have experienced a positive, emotional attachment to an adult figure.
11. Some animal studies have indicated, for example, that bright lights can increase serum testosterone levels in pregnant rats as well as in their male fetuses (Rines and vom Saal, 1984). Episodic elevations of testosterone have been associated not only with increased maternal aggressiveness but also with increased postnatal aggressiveness of offspring. Androgens also contribute to hypervigilance and are thought to sensitize the parts of the fetal brain that mediate aggression. Sensitized animals subsequently are able to respond rapidly and aggressively to stimuli that elicit surges of testosterone (Kamel et al., 1975).
12. Although many aspects of the punitiveness measured in these studies, such as criticism, threats, physical punishment, coerciveness, and guilt-inducing behaviors, fall within definitions of maltreatment used by many investigators, the studies of older children dealt with punitiveness and not specifically abuse.
13. Since a supportive spouse can moderate the disruptive effects of other stressors on parenting, such couples lack a salient buffer (Rutter, 1987).
14. For example, the mother may live alone with her children, then live with the father of some of the children (and perhaps conceiving another during this time), then move in with her own mother, then live on her own again (perhaps in a shelter program), then reside with the mother's sister and her own children in a small apartment.
15. Studies using parental reports and observations of parent-child interactions do not consistently demonstrate greater negative or aversive behaviors such as criticizing, slapping, and teasing, but this may result from reactions being observed or a reluctance to report harsh behavior. However, a greater proportion of negative relative to positive parent-child exchanges is found, even in studies that do not report more frequent aversive parental behavior (Burgess and Conger, 1978; Mash et al., 1983).
16. Despite the popularity of this approach, many problems remain in studying the effects of stress, including the failure to consider the significance of a specific stressful event for an individual. Most researchers do not consider the individual's coping strategy, which may alter the impact of stress (Pianta and Egeland, 1990).
17. Neighborhoods may also be a source of reporting bias. Child protection workers in poverty areas with more severe abuse and neglect caseloads tended to judge maltreatment vignettes less seriously, while workers in socioeconomically advantaged areas had a lower threshold in judging the severity of abuse vignettes (Wolock, 1982).
Aarens, M. et al.
1978 Alcohol, Casulaties and Crime. Berkeley, CA: Social Research Group.
1985 Use of Pornography Erotica by Sex Offenders. Paper presented at the United States Attorney General's Commission on Pornography, Houston, TX.
Abel, G.G., J. Becker, J. Cunningham-Rathner, M. Mittelman, and J.L. Rouleau
1988 Multiple paraphiliac diagnosis among sex offenders. Bulletin of the American Academy of Psychiatry and the Law 16(2):153-168.
Aber, J.L., and D. Cicchetti
1984 The socio-emotional development of maltreated children: an empirical and theoretical analysis. Pp. 147-205 in Fitzgerald, B. Lester, and M. Yogman (eds.) Theory and Research in Behavioral Pediatrics. Vol. II. New York: Plenum Press.
1990 The problem of co-occurring disorders among jail detainees: Antisocial disorder, alcoholism, drug abuse, and depression. Law and Human Behavior 14:333-345.
Altemeier, W.A., S. O'Connor, P.M. Vietze, H. Sandler, and K.B. Sherrod
1982 Antecedents of child abuse. Journal of Pediatrics 100 (May):823-829.
1991 The role of the child in physical abuse: A reappraisal. Violence and Victims 6(2):87-100.
Araji, S., and D. Finkelhor
1986 Abusers: A review of the research. Pp. 89-118 in D. Finkelhor, ed., A Sourcebook on Child Sexual Abuse. Beverly Hills, CA: Sage.
Armentrout, J.A., and A.L. Hauer
1978 MMPIs of rapists of adults, rapists of children, and non-rapist sex offenders. Journal of Clinical Psychology 34(2):330-332.
Ayoub, C.C., and S.S. Milner
1985 Failure to thrive parental indicators, types, and outcome. Child Abuse and Neglect 9:491-499.
1970 Cholinergic synapses in the lateral hypothalamus for the control of predatory aggression in the rat. Brain Research 20:409-424.
1928 A diencephalic mechanism for the expression of rage with special reference to the sympathetic nervous system. American Journal of Physiology 84:490-515.
1989 Rearing competent children. Pp. 349-378 in W. Damon (ed.), Child Development Today and Tomorrow. San Francisco: Jossey-Bass.
1991 The influence of parenting style on adolescent competence and substance abuse. Journal of Early Adolescence 11(1):56-94.
1992 Family Factors Applied to Child Maltreatment. Background paper prepared for the Panel on Research on Child Abuse and Neglect. Washington, DC: National Research Council.
Bear, D.M., J.F. Rosenbaum, and R. Norman
1986 Aggression in cat and man precipitated by a cholesterase inhibitor. Psychosomatics 26:535-536.
1988 The effects of child sexual abuse on adolescent sexual offenders. Pp. 193-207 in G.E. Wyatt and G.J. Powell, eds., The Lasting Effects of Child Sexual Abuse. Newbury Park, CA: Sage.
Beckwith, L., and S.E. Cohen
1984 Home, environmental, and cognitive competence in preterm children during the first 5 years. Pp. 235-271 in A. Gottfried, ed., Home Environmental and Early Cognitive Development. New York: Academic Press.
1980 Child maltreatment: An ecological integration. American Psychologist 35:320-335.
1992 The Etiology of Child Maltreatment: An Ecological-Contextual Analysis. Paper prepared for the Panel on Research on Child Abuse and Neglect. Washington, DC: National Research Council.
Benedict, M.I., R.B. White, and D.A. Cornely
1985 Maternal perinatal risk factors and child abuse. Child Abuse and Neglect 9:214-224.
1969 Family functioning in the low-income black community. Social Casework 50(10):563-572.
Bradburn, N.M., L.J. Ripps, and S.K. Shevall
1987 Answering autobiographical questions: The impact of memory and inference on surveys. Science 236(4798):158-61.
Brooten, D., S. Gennaro, L.P. Brown, P. Butts, A.L. Gibbons, S. Bakewell-Sachs, and S.P. Kumar
1988 Anxiety, depression, and hostility in mothers of preterm infants. Nursing Research 37:213-217.
Brown, E., P. Martinez, and M. Radke-Yarrow
1992 Diversity: Research with diverse populations. Society for Research on Child Development Newsletter (Fall).
Brunquell, D., L. Crichton, and B. Egeland
1981 Maternal personality and attitude in disturbances of child rearing. American Journal of Orthopsychiatry 51:680-691.
1979 Child abuse: A social interactional analysis. Pp. 142-172 in B.B. Lahey and A. Kazdin, eds., Advances in Clinical Child Psychology 2. New York: Plenum.
Burgess, R.L., and R.D. Conger
1978 Family interaction in abusive, neglectful, and normal families. Child Development 49:1163-1173.
Burgess, A., and P. Draper
1989 The explanation of family violence: The role of biological, behavioral, and cultural selection. Pp. 59-116 in L. Ohlin and M. Tonry, eds., Family Violence. Chicago: University of Chicago Press.
Bycer, A., L.D. Breed, J.E. Fluke, and T. Costello
1984 Unemployment and Child Abuse and Neglect Reporting. Denver: American Humane Association.
Caplan, P.J., J. Watters, G. White, R. Parry, and R. Bates
1984 Toronto multiagency child abuse research project: The abused and the abuser. Child Abuse and Neglect 8:343-351.
1986 Parents Who Don't Spank: Deviation in the Legitimation of Physical Force. Unpublished doctoral dissertation, University of New Hampshire.
Casanova, G.M., J. Domanic, T.R. McCanne, and J.S. Milner
1992 Physiological responses to non-child-related stressors in mothers at risk for child abuse. Child Abuse and Neglect 16(1):31-44.
Chase-Lansdale, P.L., and J. Brooks-Gunn, eds.
in press Escape from Poverty: What Makes a Difference for Poor Children. New York: Cambridge University Press.
Chase-Lansdale, P.L., J. Brooks-Gunn, and E. Zamsky
in press Young multigenerational families in poverty: Quality of mothering and grandmothering. Child Development.
Children's Defense Fund
1991 The State of America's Children. Washington, DC: Children's Defense Fund.
Cicchetti, D., and J.L. Aber
1980 Abused children - abusive parents: An overstated case? Harvard Educational Review 50(2):244-255.
Cicchetti, D., and V. Carlson, eds.
1989 Child Maltreatment: Theory and Research on the Causes and Consequences of Child Abuse and Neglect. New York: Cambridge University Press.
Cicchetti, D., and M. Lynch
in press Toward an ecological/transactional model of community violence and child maltreatment: Consequences for children's development. Psychiatry.
Conger, R., J. McCarty, R. Yang, B. Lahey, and J. Kropp
1984 Perception of child, childrearing values, and emotional distress as mediating links between environmental stressors and observed maternal behavior. Child Development 54:2234-2247.
Connelley, C.D., and M.A. Straus
1992 Mothers' age and risk for physical abuse. Child Abuse and Neglect 16:703-712.
1984 Research on the Prevention of Sexual Abuse of Children. Paper presented at the Second National Conference for Family Violence Researchers, Durham, NH.
1985 Clinical dimensions of adult sexual abuse of children. Behavioral Sciences and the Law 3(4):341-354.
1991 The nature of sexual offenses against children. Pp. 11-34 in C.R. Hollin and K. Howells, eds., Clinical Approaches to Sex Offenders and their Victims. New York: John Wiley.
Conte, J.R., C. Rosen, and L. Saperstein
1986 An analysis of programs to prevent the sexual victimization of children. Journal of Primary Prevention 6(3):141-155.
Coulton, C., J. Chow, and S. Pandey
1990a An Analysis of Poverty and Related Conditions in Cleveland Area Neighborhoods. Case Western Reserve University, Center for Urban Poverty and Social Change.
Coulton, C., J. Korbin, M. Su, and J. Chow
(n.d.) Community Level Factors and Child Maltreatment Rates. Unpublished manuscript.
Coulton, C.J., and S. Pandey
1992 Geographic concentration of poverty and risk to children in urban neighborhoods. Special Issue: The impact of poverty on children. American Behavioral Scientist 35(3):238-257.
Coulton, C., S. Pandey, and J. Chow
1990b Concentration of poverty and the changing ecology of low-income urban neighborhoods: An analysis of the Cleveland area. Social Work Research and Abstracts 26(4):5-16.
1985 An epidemiological study of abused children and their families in the United Kingdom between 1977 and 1982. Child Abuse and Neglect 9:441-448.
1988 By Silence Betrayed: Sexual Abuse of Children in America. New York: Harper and Row.
1981 Abusing, neglecting problematic, and adequate dyads: Differentiating by patterns of interaction. Merrill-Palmer Quarterly 27:201-208.
1985 Social networks, quality of child rearing, and child development. Child Development 56(October):1299-1313.
1981 Infant irritability, mother responsiveness, and social influences on the security of infant-mother attachment. Child Development 52:857-865.
Daly, M., and M. Wilson
1983 Sex Evolution and Behavior. Adaptations for Reproduction. 2nd edition. Boston: Willard Grant Press.
1969 Protecting the Child Victim of Sex Crimes Committed by Adults. Englewood, CO: American Humane Association.
Disbrow, M.A., H. Doerr, and C. Caulfield
1977 Measuring the components of parents potential for child abuse and neglect. Child Abuse and Neglect 1:279-296.
DiVitto, B., and S. Goldberg
1979 The effects of newborn medical status on later parent-infant interaction. Pp. 311-356 in T.M. Field, S.M. Sostek, S. Goldberg, and H.H. Shuman, eds., Infants Born at Risk: Behavior and Development. New York: Spectrum Publication.
Dodge, K.A., J.E. Bates, and G.S. Pettit
1990 Mechanisms in the cycle of violence. Science 250 (December 21):1678-1683.
Dornbusch, S.M., P.L. Ritter, P.H. Leiderman, D.J. Roberts, and M.J. Fraleigh
1987 The relation of parenting style to adolescent performance. Child Development 58:1244-1257.
1992 Prevention of neglect and nonorganic failure to thrive. Pp. 115-149 in D.J. Willis, E.W. Holden, and M. Rosenberg, eds., Prevention of Child Maltreatment: Developmental and Ecological Perspectives. New York: John Wiley.
1987 Diagnostic and Statistical Manual of Mental Disorders. 3rd ed., rev. Washington, DC: American Psychiatric Association.
1981 Child maltreatment in European and Hawaiian Americans. Child Abuse and Neglect 5(4):457-466.
1987 Child maltreatment in the United States: Etiology, impact, and prevention. In Healthy Children: Investing in the Future. Health Program, Office of Technology Assessment, U.S. Congress, Washington, DC.
Dumas, J., and R.G. Wahler
1985 Indiscriminate mothering as a contextual factor in aggressive-oppositional child behavior: "Damned if you do, damned if you don't." Journal of Abnormal Child Psychology 13:1-17.
Duncan, G., and J.L. Aber
In press Neighborhood conditions and structure. In G. Duncan and J. Brooks-Gunn (eds.) Neighborhoods, Families, and Poverty: These Effects on Children and Youth.
Earp, J., and M.G. Ory
1980 The influence of early parenting on child maltreatment. Child Abuse and Neglect 4:237-245.
1988 Breaking the cycle of abuse: Implications for prediction and intervention. Pp. 87-99 in K.D. Browne, C. Davies, and P. Stratton, eds., Early Prediction and Prevention of Child Abuse. New York: John Wiley.
Egeland, B., and D. Brunquell
1979 An at-risk approach to the study of child abuse: Some preliminary findings. Journal of the American Academy of Psychiatry 18:219-235.
Egeland, B., and D. Jacobvitz
1984 Intergenerational Continuity in Parental Abuse: Causes and Consequences. Paper presented at the Conference on Biosocial Perspectives in Abuse and Neglect, York, ME.
Egeland, B., M. Breitenbucher, and D. Rosenberg
1980 A prospective study of the significance of life stress in the etiology of child abuse. Journal of Clinical and Consulting Psychology 48:195-205.
Egeland, B., D. Jacobvitz, and K. Papatola
1987 Intergenerational continuity of abuse. Pp 255-276 in R.J. Gelles and J. Lancaster, eds., Child Abuse and Neglect: Biosocial Dimensions. New York: Aldine.
Egeland, B., D. Jacobvitz, and L.A. Sroufe
1988 Breaking the cycle of abuse. Child Development 59(4):1080-1088.
Ehrenkranz, J., E. Bliss, and M.H. Sheard
1974 Plasma testosterone: Correlations with aggressive behavior and social dominance in man. Psychosomatic Medicine 36:469-475.
1990 Intoxication and aggression. Pp. 241-320 in M. Tonry and J.Q. Wilson, eds., Drugs and Crime. Chicago: University of Chicago Press.
Fagan, J., and A. Browne
1990 Marital Violence: Physical Aggression Between Men and Women in Intimate Relationships. Background paper prepared for the Panel on the Understanding and Control of Violent Behavior. Washington, DC: National Research Council.
1988 Child Sexual Abuse. New York: Columbia University Press.
1990 Understanding Child Sexual Maltreatment. Newbury Park, CA: Sage.
Famularo, R., K. Stone, R. Barnum, and R. Wharton
1986 Alcoholism and severe child maltreatment. American Journal of Orthopsychiatry 56:481-485.
Farber, E.D., and J.A. Joseph
1985 The maltreated adolescent: Patterns of physical abuse. Child Abuse and Neglect 9(2):201-206.
Feber, M.A., and B. O'Farrell
1991 Work and Family: Policies for a Changing Work Force, Panel on Employer Policies and Working Families. Washington, DC: National Academy Press.
1984 Child Sexual Abuse. New York: Free Press. 1987 The sexual abuse of children: Current research reviewed. Psychiatric Annals 17(4)(April):233-241.
Finkelhor, D., S. Araji, L. Baron, A. Growne, S.D. Peters, and G.E. Wyatt
1986 A Sourcebook on Child Sexual Abuse. Newbury Park, CA: Sage.
Floody. O.R., and D.W. Pfaff
1972 Steroid hormones and aggressive behavior: Approaches to the study of hormonesensitive brain mechanisms for behavior. Aggression 149-185.
1973 The diagnosis of the maltreatment syndrome in children. Pediatrics 51:780-782.
1988 Behavior problems in sexually abused children. Pp. 171-191 in G.E. Wyatt and G.J. Powell, eds., The Lasting Effects of Child Sexual Abuse. Newbury Park, CA: Sage Publications.
1983 Child abuse treatment failures reveal need for redefinition of the problem. Child Abuse and Neglect 7:395-402.
1977 The human ecology of child maltreatment: A conceptual model for research. Journal of Marriage and the Family 39:721-735.
Garbarino, J., and A. Crouter
1978 Defining the community context for parent-child relations: The correlates of child maltreatment. Child Development 49:604-616.
Garbarino, J., and D. Sherman
1980 High-risk neighborhoods and high-risk familes: The human ecology of child maltreatment. Child Development 51:188-198.
Garbarino, J., and J. Vondra.
1987 Psychological maltreatment: Issues and perspectives. Pp. 24-44 in M.R. Brassard, R.Germain, and S.N. Hart, eds., Psychological Maltreatment of Children and Youth. New York: Pergamon Press.
1983 International perspectives on child abuse and neglect. Child Abuse and Neglect 7(4):375-386. 1991 Physical violence, child abuse, and child homicide: A continuum of violence or distinct behaviors. Human Nature 2(1):59-72. 1992 Poverty and violence toward children. American Behavioral Scientist 35(3):258-274.
Gelles, R.J., and A.W. Edfeldt
1986 Violence towards children in the United States and Sweden. Child Abuse and Neglect 10:501-510.
Gelles, R.J., and E.F. Hargreaves
1981 Maternal employment and violence toward children. Journal of Family Issues 2:509-530.
Gelles, R.J., and M.A. Straus
1988 Intimate Violence. New York: Simon and Schuster.
George, C., and M. Main
1979 Social interactions of young abused children: Approach, avoidance, and aggression. Child Development 50:306-318.
1970 Violence Against Children: Physical Child Abuse in the United States. Cambridge, MA: Harvard University Press.
Giovannoni, J.M., and A. Billingsley
1970 Child neglect among the poor: A study of parental adequacy in families of three ethnic groups. Child Welfare 49(April):196-204.
1983 Parent-infant bonding: Another look. Child Development 54:1355-1382.
Goodwin, J., T. McCarthy, and P. Divasto
1981 Prior incest in mothers of abused children. Child Abuse and Neglect 5:87-96.
Goy, R., and B. McEwen
1977 Sexual Differentiation and the Brain. Cambridge, MA: MIT Press.
Groth, N.A., W.F. Hobson, and T.S. Gary
1982 The child molester: Clinical observations. Journal of Social Work and Human Sexuality 1(1-2):129-144.
Hamilton, A., W.B. Stiles, F. Melowsky, and D.G. Beal
1987 A multilevel comparison of child abusers with nonabusers. Journal of Family Violence 2(3):215-225.
Hamilton, C.J., and J.J. Collins
1982 The role of alcohol in wife beating and child abuse: A review of the literature. Pp. 253-287 in J.J. Collins, ed., Drinking and Crime: Perspectives on the Relationships Between Alcohol Consumption and Criminal Behavior. London: Guilford Press.
Harford, T.C., and D.A. Parker
1985 Alcohol dependence and problem drinking in a national sample. Pp. 29-31 in Alcohol, Drugs and Tobacco: An International Perspective for the Future, Volume II. Proceedings of the 34th International Congress on Alcoholism and Drug Dependence, Calgary, Alberta.
Hart, S.N., and M.R. Brassard
1987 A major threat to children's mental health: Psychological maltreatment. American Psychologist 42(2):160-165.
Hart, S.N., R., Germain, and M.R. Brassard.
1987 The challenge: To better understand and combat the psychological maltreatment of children and youth. Pp. 3-234 in M.R. Brassard, R. Germain, and S.N. Hart, eds., Psychological Maltreatment of Children and Youth. New York: Pergamon Press.
Hartman, C.R., and A.W. Burgess
1989 Sexual abuse of children: Causes and consequences. Pp. 95-128 in D. Cicchetti and V. Carlson, eds., Child Maltreatment: Theory and Research on the Causes and Consequences of Child Abuse and Neglect. New York: Cambridge University Press.
1989 Coping with family transitions: Winners, losers and survivors. Child Development 60:1-15. 1991 The role of individual difference in family relations in coping with divorce and remarriage. In P. Cowan and E.M. Hetherington, eds., Advances in Family Research: Vol. 2 Family Transitions. Hillsdale, NJ: Lawrence Erlbaum Press.
Hetherington, E.M., and W.G. Clingempeel
1992 Coping with marital transitions: A family systems perspective. Monographs of the Society for Research in Child Development 57(no. 2-3).
Higley, J.D., and S.J. Suomi
1989 Temperamental reactivity in non-human primates. Pp 153-167 in G.A. Kohnstamm, J.E. Bates, and M.K. Rothbart, eds., Temperament in Childhood. New York: John Wiley.
Holden, E.W., D.J. Willis, and M. Corcoran
1992 Pp. 17-46 (Chapter 2) in D.J. Willis, E.W. Holden, and M. Rosenberg, eds., Prevention of Child Maltreatment: Developmental and Ecological Perspectives. New York: John Wiley.
1976 Care and exploitation of nonhuman primate infants by conspecifics other than the mother. In D. Lehrman, R.A. Hinde, and E. Shaw, eds., Advances in the Study of Behavior Vol 4. New York: Academic Press.
Huesmann, L.R., L.D. Eron, M.M. Lefkowitz, and L.O. Walder
1984 Stability of aggression over time and generations. Journal of Abnormal Psychology 20:1120-1134.
Hunter, R.S., and N. Kilstrom
1979 Breaking the cyle in abusive families. American Journal of Psychiatry 36(October):1320-1322.
Hunter, R.S., N. Kilstrom, E.N. Kraybill, and F. Loda
1978 Antecedents of child abuse and neglect in premature infants: A prospective study in a newborn intensive care unit. Pediatrics 61:629-635.
1991 Children in Poverty: Child Development and Public Policy. Cambridge: Cambridge University Press.
1976 Child welfare services. Children Today (May/June) 5(3):16-23.
Kadushin, A., and J.A. Martin
1981 Child Abuse: An Interactional Event. New York: Columbia University Press.
Kamel, F., E. Mock, W. Wright, and A. Frankel
1975 Alterations in plasma concentrations of testosterone, LH, and prolactin associated with mating in the male rat. Hormones and Behavior 6(3):277-288.
Kaufman, J., and E. Zigler
1987 Do abused children become abusive parents? American Journal of Orthopsychiatry 57:186-192.
Kavanaugh, K., L. Youngblade, J. Reid, and B. Fagot
1988 Interactions between children and abusive control parents. Journal of Clinical Child Psychology 17:137-142.
Kelly, R.J., and R. Lusk
1992 Theories of pedophilia. In W. O'Donohue and J.H. Geer, eds., The Sexual Abuse of Children: Theory and Research, Volume 1. Hillsdale, NJ: Lawrence Erlbaum Press.
Kempe, C.H., F.N. Silverman, B. Steele, W. Droegemueller, and H.R. Silver.
1962 The battered child syndrome. Journal of the American Medical Association 181(1):17-24.
Kempe, R.S., and C.H. Kempe
1978 Child Abuse. Cambridge, MA: Harvard University Press.
Kinard, E.M., and L.V. Klerman
1980 Teenage parenting and child abuse: Are they related? American Journal of Orthopsychiatry 50(3):481-488.
Klein, M., and L. Stern
1971 Low birth weight and the battered child syndrome. American Journal of Diseases of Childhood 122:15-18.
1985 A taxonomic analysis of child molesters. Pp. 2-20 in R.A. Prentky and V.L. Quinsey, eds., Human Sexual Aggression: Current Perspectives. Annals of the New York Academy of Sciences 528. New York: The New York Academy of Sciences.
1988 Child sexual abuse and pornography: Is there a relationship? Journal of Family Violence 3(4):253-267.
1982 Child abuse and neglect: Prediction and misclassification. Pp. 67-104 in R.H. Starr, ed., Child Abuse Prediction: Policy Implications. Cambridge, MA: Ballinger.
Kreuz, L.E., and R.M. Rose
1972 Assessment of aggressive behavior and plasma testosterone in a young criminal population. Psychosomatic Medicine 34:321-332.
Krugman, R.D., M. Lenherr, L. Betz, and G.E. Fryer
1986 The relationship between unemployment and physical abuse of children. Child Abuse and Neglect 10:415-418.
1983 Sexual Strands: Understanding and Treating Sexual Anomalies in Men.
Hillsdale, NJ: Lawrence Erlbaum Press.
1987 Child Molesters: A Behavioral Analysis. December. National Center for Missing and Exploited Children, Arlington, VA.
1992 Investigators' Guide to Allegations of Ritual Abuse. January. National Center for the Analysis of Violent Crime. Quantico, VA.
1986 Theory and treatment in child molestation. Journal of Consulting and Clinical Psychology 54(2):176-182.
Larrance, D.T., and C.T. Twentyman
1983 Maternal attribution and child abuse. Journal of Abnormal Psychology 92:449-457.
1988 A Dictionary of Epidemiology. 2nd ed. New York: Oxford University Press.
Lauderdale, M., A. Valiunas, and R. Anderson
1980 Race, ethnicity, and child maltreatment: An empirical analysis. Child Abuse and Neglect 4:163-169.
1981 Risk factors for child abuse: Methodologic standards in case-control studies. Pediatrics 68(November):684-690.
Leventhal, J.M., S. Horwitz, C. Rude, and D. Stier
1993 Maltreatment of children born to teenage mothers: A comparison between the 1960s and the 1980s. Journal of Pediatrics 122:314-319.
1992 From abuse to violence: Psychophysiological consequences of maltreatment. Journal of the American Academy of Child and Adolescent Psychiatry 31(3):383-391.
Lewis, D.O., J.H. Pincus, B. Bard, and E. Richardson
1988 Neuropsychiatric, psycho-educational, and family characteristics of 14 juveniles condemned to death in the United States. American Journal of Psychiatry 145(5):584-589.
Lewis, D.O., S.S. Shanok, J.H. Pincus, and G.H. Glaser
1980 Violent juvenile delinquents. Annual Progress in Child Psychiatry and Child Development: Psychiatric, Neurological, Psychological Abuse Factors: 591-603.
1989 Toward a theory of the genesis of violence: A follow-up study of delinquents. American Academy of Child and Adolescent Psychiatry 28(3):431-436.
1983 Prediction Based on Census Data and Economic Indicators. Paper presented at the 3rd National Conference on Research, Demonstration, and Evaluation in Social Services, American Public Welfare Association, Washington, DC.
1973 Abused and neglected children in America: A study of alternative policies. Harvard Educational Review 43:556-598.
Lorber, R., D.K. Felton, and J.B. Reid
1984 A social learning approach to the reduction of coercive processes in child abusive families: A molecular analysis. Advances in Behavior Research and Therapy 6:29-45.
1984 Project 12-Ways: Treating child abuse and neglect from an ecobehavioral perspective. Pp. 260-295 in R.F. Dangel and R.A. Polster, eds., Parent Training: Foundations of Research and Practice. New York: Guilford Press.
Lynch, M.A., and J. Roberts
1977 Predicting child abuse: Signs of bonding failure in the maternity hospital. British Medical Journal 1:624-626.
MacAndrew, C., and R.B. Edgerton
1969 Drunken Comportment: A Social Explanation. Chicago, IL: Aldine.
Maccoby, E.E., and J.A. Martin
1983 Socialization in the context of the family: Parent-child interaction. Pp. 1-102 in P.H. Mussen and E.M. Hetherington, eds., Handbook of Child Psychology: Socialization, Personality, and Social Development. New York: John Wiley and Sons.
1985 Brain evolution relating to family, play and the separation call. Archives of General Psychiatry 42:405-417.
1983 Exploration, play, and cognitive functioning related to infant mother attachment. Infant Behavior and Development 6:167-174.
1989 Pornography and sex offenders. Pp. 185-214 in D. Zillmann and J. Bryant, eds., Pornography: Research Advances and Policy Considerations. Hillsdale, NJ: Lawrence Erlbaum Press.
Mash, E.J, C. Johnston, and K. Kovitz
1983 A comparision of the mother-child interactions of physically abused and nonabused children during play and task situations. Journal of Clinical Child Psychology 12:337-346.
1989 The Harm of Pornography. Address to the Religious Alliance Against Pornography. October 26.
1992 Attitudes of primary care physicians toward corporal punishment. Journal of the American Medical Association 267(23)(June 17):3161-3165.
Meisels, S.J., and J.W. Plunkett
1988 Developmental consequences of preterm birth: Are there long-term effects? Pp. 87-128 in P.B. Baltes, D.L. Featherman, and R.M. Lerner, eds., Life-span Development and Behavior. Hillsdale, NJ: Lawrence Erlbaum.
Melnick, B., and J.R. Hurley
1969 Distinctive personality of child-abusing mothers. Journal of Consulting and Clinical Psychology 33(December):746-749.
1974 Aggression, and androgens, and the XYY syndrome. Pp. 433-453 in R.C. Friedman, R.M. Richart, and R. Vande Wiele, eds., Sex Differences in Behavior. New York: John Wiley.
Milner, J.S., and C. Chilamkurti
1991 Physical child abuse perpetrator characteristics: A review of the literature. Journal of Interpersonal Violence 6(3)(September):345-366.
Monti, P.M., W.A. Brown, and M.A. Corriveau
1977 Testosterone and components of aggressive and sexual behavior in man. American Journal Psychiatry 134:692-694.
1982 Alcohol and family violence: A review of the literature. In National Institute of Alcoholism and Alcohol Abuse, Alcohol Abuse, Alcohol Consumption and Related Problems. (Alcohol and Health Monograph 1) Washington, DC: Department of Health and Human Services.
Murphy, S., B. Orkow, and R.M. Nicola
1985 Prenatal prediction of child abuse and neglect: A prospective study. Child Abuse and Neglect 9(2):225-235.
National Center for Child Abuse and Neglect
1981 National Study of the Incidence and Severity of Child Abuse and Neglect. 81-30325. Washington, DC: U.S. Department of Health and Human Services. [NIS-1]
1988 Study Findings: Study of National Incidence and Prevalence of Child Abuse and Neglect. Washington, DC: U.S. Department of Health and Human Services. [NIS-2]
National Research Council
1990 Who Cares for America's Children? Panel on Child Care Policy, Committee on Child Development Research and Public Policy, Commission on Behavioral and Social Sciences and Education. Washington, DC: National Academy Press.
1986 Child Abuse and Neglect: What Happens Eventually? New York: Brunner Mazel.
Oates, R.K., A.A. Davis, M.G. Ryan, and L.F. Stewart
1979 Risk factors associated with child abuse. Child Abuse and Neglect 3:547-553.
1981 Origins of human competence: A cultural-ecological perspective. Child Development 52:413-429.
Orme, T.C., and J. Rimmer
1981 Alcoholism and child abuse. Journal of Studies on Alcohol 42:273-287.
1978 Personality differences appearing between rapists of adults, rapists of children, and non-violent sexual molesters of children. Research Communications in Psychology, Psychiatry, and Behavior 3(4):385-393.
Parke, R.D., and C.W. Collmer
1975 Child abuse: An interdisciplinary analysis. Pp. 509-590 in E.M. Hetherington, ed., Review of Child Development Research 5. Chicago: University of Chicago Press.
Parmalee, A.H., Jr.
1975 Neurophysiological and behavioral organization of premature infants in the first months of life. Biological Psychiatry 10:501-512.
1982 Coercive Family Processes. Eugene, OR: Castalia Publishing Co.
Patterson, G.R., J.B. Reid, and T.J. Dishion
1992 Antisocial Boys. Eugene, OR: Castalia Publishing Co.
1978 Child abuse and neglect: The myth of classlessness. American Journal of Orthopsychiatry 48:608-617.
1981 The Social Context of Child Abuse and Neglect. New York: Human Sciences Press.
1989 For Reasons of Poverty. New York: Praeger.
1991 Alcohol in Human Violence. New York: Guilford Press.
1976 Children who are victims of sexual assault and the psychology of offenders. American Journal of Psychotherapy 30(3):395-421.
Pianta, R.C., and B. Egeland
1990 Life stress and parenting outcomes in a disadvantaged sample: Results of the Mother-Child Interaction project. Special Issue: The stresses of parenting. Journal of Clinical Child Psychology 19(4):329-336.
Pianta, R., B. Egeland, and M.F. Erickson
1989 The antecedents of maltreatment: Results of the Mother-Child Interaction Research Project. Pp. 203-253 in D. Cicchetti and V. Carlson, eds., Child Maltreatment: Theory and Research on the Causes and Consequences of Child Abuse and Neglect. New York: Cambridge University Press.
Polansky, N.A., R. Borgman, and C. DeSaix
1972 Roots of Futility. San Francisco: Jossey-Bass.
Polansky, N.A., M.A. Chalmers, E.Buttenweiser, and D.P. Williams
1981 Damaged Parents: An Anatomy of Child Neglect. Chicago: University of Chicago Press.
Polansky, N.A., J.M. Gaudin, and A.C. Kilpatrick
1992 Family radicals. Children and Youth Services Review 14:19-26.
Pollock, V.E., J. Briere, L. Schneider, J. Knop, S.A. Mednick, and D.W. Goodwin
1990 Childhood antecedents of antisocial behavior: Parental alcoholism and physical abusiveness. American Journal of Psychiatry 147:1290-1293.
Powell, G.F., and J.L. Low
1983 Behavior in non-organic failure to thrive. Journal of Developmental and Behavioral Pediatrics 4:26-33.
Powell, G.F., J.L. Low, and M.A. Spears
1987 Behavior as a diagnostic aid in failure to thrive. Journal of Developmental and Behavioral Pediatrics 8:18-24.
1990 Sexual Violence. Paper prepared for the Panel on the Understanding and Control of Violent Behavior. Washington, DC: National Research Council.
Rada, R.T., D.R. Laws, and R. Kellner
1976 Plasma testosterone levels in the rapist. Psychosomatic Medicine 38:257-268.
Reid, J.B., K. Kavanagh, and D.V. Baldwin
1987 Abusive parents' perceptions of child problem behaviors: An example of parental bias. Journal of Abnormal Child Psychology 15:457-466.
Rines, J.P., and F.S. Vom Saal
1984 Fetal effects on sexual behavior and aggression in young and old female mice treated with estrogen and progesterone. Hormones and Behavior 18(2):117-129.
Robins, L.N., and D.A. Regier, eds.
1991 Psychiatric Disorders in America: The Epidemiological Catchment Area Study. New York: Free Press.
Robins, L.N., J.E. Helzer, M.M. Weissman, H. Orvaschel, E. Gruenberg, J.D. Burke, and D.A. Regier
1984 Lifetime prevalence of specific psychiatric disorders in 3 sites. Archives of General Psychiatry 41(10):949-958.
Rosenbaum, A., and K.D. O'Leary
1981 Children: The unintended victims of marital violence. American Journal of Orthopsychiatry 51:692-699.
1987 New directions for research on the psychological maltreatment of children. American Psychologist 42:166-171.
Runyan, D.K., C.L. Gould, D.C. Trost, and F.A. Loda
1981 Determinants of foster care placement for the maltreated child. American Journal of Public Health 71(July):706-711.
1986 The Secret Trauma: Incest in the Lives of Girls and Women. New York: Basic Books.
1987 Psychosocial resilience and protective mechanisms. American Journal of Orthopsychiatry 57(3):316-331.
Sameroff, A.J., and M. Chandler
1975 Reproductive risk and the continuum of caretaking casualty. Pp. 187-244 in F. Horowitz, ed., Review of Child Development Research 4. Chicago: University of Chicago Press.
Scheper-Hughes, N., ed.
1987 Child Survival: Anthropological Perspectives on the Treatment and Maltreatment of Children. Boston: Kluwer.
1988 Child pornography and prostitution. Pp. 153-165 in D.H. Schetky and A.H. Green, eds., Child Sexual Abuse: A Handbook for Health Care and Legal Professionals. New York: Brunner/Maxel.
1980 Self-help groups for parents of premature infants. Canadian Nurse (July-August):30-33.
Smith, D.E., M.D. King, and B.G. Hoebel
1970 Lateral hypothalamic control of killing: Evidence for a cholinoceptive mechanism. Science 167:900-901.
Spinetta, J.J., and D. Rigler
1972 The child abusing parent: A psychological review. Psychological Bulletin 77:296-304.
1983 Infant-caregiver attachment and patterns of adaptation in preschool: The roots of maladaptation and competence. Pp. 41-83 in M. Perlmutter, ed., Minnesota Symposium in Child Psychology 16. Hillsdale, NJ: Lawrence Erlbaum Press.
Stark, R., and J. McElvoy, III
1970 Middle class violence. Psychology Today 4:52-65.
Starr, R.H., Jr., ed.
1979 Child abuse. American Psychologist 34(10):872-878.
1982 Child Abuse Prediction: Policy Implications. Cambridge, MA: Ballinger.
1992 Physical abuse of children. In V.B. Van Hasselt et al., eds., The Handbook of Family Violence. New York: Plenum Press.
1976 Violence within the family. Pp. 3-24 in C.H. Kempe and A.E. Helfer, eds., Child Abuse and Neglect: The Family and the Community. Cambridge, MA: Ballinger.
Steele, B.F., and C.B. Pollock
1968 A psychiatric study of parents who abuse infants and small children. Pp. 89-133 in R.E. Helfer and C.H. Kempe, eds., The Battered Child. Chicago: University of Chicago Press.
Steinberg, L., R. Catalano, and D. Dooley
1981 Economic antecedents of child abuse and neglect. Child Development 52:975-985.
Steinberg, L., S.M. Dornbusch, and B. Brown
1992a Ethnic differences in adolescent achievement: An ecological perspective. American Psychologist.
Steinberg, L., S.D. Lamborn, S.M. Dornbusch, and N. Darling
1992b Impact of parenting practices on adolescent achievement: Authoritative parenting, school involvement, and encouragement to succeed. Child Development 63:1266-1281.
1980 Stress and physical child abuse. Child Abuse and Neglect 4:75-88.
Straus, M.A., and R.J. Gelles
1986 Societal change in family violence from 1975 to 1985 as revealed by two national surveys. Journal of Marriage and the Family 48(August): 465-479.
Straus, M.A., and S. Lauer
1993 Corporal Punishment and Crime in Ethnic Group Context. Paper presented at the 1992 meeting of the American Society of Criminology, Durham, NH.
Straus, M.A., and C. Yodanis
1994 Paths from Corporal Punishment to Physical Abuse in a Nationally Representative Sample of American Parents. Chapter 6 in Murray A. Straus, ed., Beating the Devil Out of Them: Corporal Punishment by Parents and Its Effects on Children. Boston: Lexington/Macmillan. (forthcoming)
1978 Maternal behavior by socially incompetent monkeys: Neglect and abuse of off-spring. Journal of Pediatric Psychology 3(1):28-34.
Suomi, S.J., and C. Ripp
1983 A history of motherless mother monkey mothering at the University of Wisconsin Primate Laboratory. Pp. 49-78 in M. Reite and N.G. Caine, eds., Child Abuse: The Nonhuman Primate Data. New York: Alan R. Liss, Inc.
1991 Cultural diversity and the treatment of children. Journal of Consulting and Clinical Psychology 59(6):799-812.
Trickett, P.K., J.L. Aber, V. Carlson, and D. Cicchetti
1991 The relationship of socioeconomic status to the etiology and developmental sequelae of physical child abuse. Developmental Psychology 27(1):148-158.
Trickett, P.K., and L. Kuczynski
1986 Children's misbehaviors and parental discipline strategies in abusive and nonabusive families. Developmental Psychology 22:115-123.
Trickett, P.K., and E.J. Susman
1988 Parental perceptions of child-rearing practices in physically abusive and nonabusive families. Developmental Psychology 24:270-276.
Tuteur, W., and J. Glotzer
1966 Further observations on murdering mothers. Journal of Forensic Studies 11:373-383.
Twentyman, C.T., and R.C. Plotkin
1982 Unrealistic expectations of parents who maltreat their children: An educational deficit that pertains to child development. Journal of Clinical Psychology 38(3):497-503.
U.S. Senate Committee on Governmental Affairs, Subcommittee on Investigations
1986 Child Pornography and Pedophilia. Report, October 9. Washington, DC: U.S. Government Printing Office.
Wahler, R.G., and J.E. Dumas
1986 ''A chip off the old block": Some interpersonal characteristics of coercive children across generations. Pp. 49-86 in P. Strain, M. Guralnick, and H. Walkee, eds., Children's Social Behavior: Development, Assessment, and Modification. New York: Academic Press.
Weiger, W.A., and D.M. Bear
1988 An approach to the neurology of aggression. Journal of Psychiatric Research 244:160-166.
Weiss, B., K.A. Dodge, J.E. Bates, and G.S. Petit
1992 Some consequences of early harsh discipline: Child aggression and a maladaptive social information processing style. Child Development 63(6):1321-1335.
Weller, S.C., A.K. Romney, and D.P. Orr
1987 The myth of a subculture of corporal punishment. Human Organization 16(1):39-47.
West, D.J., C. Roy., and F.L. Nichols
1978 Understanding Sexual Attacks. London: Heinemann.
Whipple, E.E., and C. Webster-Stratton
1991 The role of parental stress in physically abusive families. Child Abuse and Neglect 15:279-291.
1989 Does violence beget violence? A critical examination of the literature. Psychological Bulletin 106(1):3-28.
1991 The role of placement experiences in mediating the criminal consequences of early childhood victimization. American Journal of Orthopsychiatry 61(2):195-209.
1992 Child Abuse and Alcohol Use. Prepared for the working group on alcohol-related violence: Fostering interdisciplinary perspectives, convened by the National Institute on Alcohol Abuse and Alcoholism (NIAAA), Washington, DC (May).
1985 Child-abusive parents: An empirical review and analysis. Psychological Bulletin 97(3)(May):462-482.
1991 Preventing Physical and Emotional Abuse of Children. New York: Guilford Press.
1982 Community characteristics and staff judgments in child abuse and neglect cases. Social Work Research and Abstracts 18(2) Summer:9-15.
Wolock, I., and B. Horowitz
1979 Child maltreatment and material deprivation among AFDC-recipient families. Social Service Review 53(June):175-194.
1984 Child maltreatment as a social problem: The neglect of neglect. American Journal of Orthopsychiatry 54(4)(October):530-543.
1985 The sexual abuse of Afro-American and white American women in childhood. Child Abuse and Neglect 9:507-519.
Wyatt, G.E., and M. Newcomb
1990 Internal and external mediators of women's sexual abuse in childhood. Journal of Consulting and Clinical Psychology 58(6):758-767.
Wyatt, G.E., M. Newcombe, M. Riederlee, and C. Notgrass
in press The Effects of Child Sexual Abuse and Psychological Functioning. Newbury Park, CA: Sage Publications.
1964 Wednesday's Children: A Study of Child Neglect and Abuse. New York: McGraw-Hill.
Zeanah, C.H., and P.D. Zeanah
1989 Intergenerational transmission of maltreatment. Psychiatry 52:177-196.
Zigler, E., and N.W. Hall
1989 Physical child abuse in America: Past, present, and future. In D. Cicchetti and V. Carlson, eds., Child Maltreatment Theory and Research on the Causes and Consequences on Child Abuse and Neglect. New York: Cambridge University Press.
1987 The ecology of child maltreatment: Identifying and characterizing high-risk neighborhoods. Child Welfare League of America 66(6)(November-December):497-506.
1988 Fertility patterns: Their relationship to child physical abuse and child neglect. Journal of Marriage and the Family 50:983-993.