ChildAbuse& Nqkcl. Vol. 14, PP. 337-345. Printed in the U.S.A. All righU~~ed.

014%2134/9OS3.OO+.CUl Copyright 0 1990 Pergamon Press plc

1990

SOME FACI’ORS INFLUENCING ABUSERS’ JUSTIFICATION OF THEIR CHILD ABUSE DOROTHEE

DIETRICH,

LEONARD BERKOWITZ, AND JAMES MCGLOIN

ALFRED

KADUSHIN,

University of Wisconsin-Madison

Abstract-The interviews of abusive caretakers originally carried out by Kadushin and Martin ( 198 1) were coded and subjected to statistical analyses to determine what aspects of the caretakers’ situation and of their interaction with the abused child had contributed to their belief that their treatment of the youngster was justified or not. A multiple regression analysis employing scores on the various indices obtained from the sample of 73 interviews indicated that the abusers tended to believe their behavior was justified if they thought the child had been defiant and they themselves had been under considerable environmental stress. On the other hand, they generally regarded their action as less justified if they had lost their temper and had been experiencing emotional distress. The latter finding suggests, in accord with Berkowitz’s analysis of emotional aggression, that some instances of child battering were impulsive reactions to a provocative event. Although major emphasis is given to the meaning of these findings for the theoretical analysis of aggression, implications for protective service practice are also indicated. Key Words-Justification

of abuse, Child’s defiance, Emotional distress, Environmental stress.

INTRODUCTION STUDIES OF CHILD ABUSE are now emphasizing the social interactive nature of the violent incident (e.g., Kadushin & Martin, 198 1; Parke & Collmer, 1975; Straus, Gelles & Steinmetz, 1980; Wolfe, 1985). As one reviewer has noted, most of the research conducted from this increasingly popular perspective “places heavy emphasis on bidirectional influences of behavior among family members, antecedent events that may precipitate abuse, and consequences that may maintain the use of excessive punishment. . .” (Wolfe, 1985, p. 463). In other words, a multiplicity of factors must be considered if child abuse is to be adequately understood, including the degree of stress experienced by the caretakers in their everyday lives and the interrelationships among the family members. In the latter regard, the general agreement is that the adult caretakers’ behavior in many instances is a reaction to the child as well as the situation they are in. But while growing numbers of investigators have adopted this approach, there is no clear consensus as to what, if any, goal the caretakers are pursuing when they batter their victims (cf., Berkowitz, 1983). What are they hoping to accomplish? In trying to identify what these aims might be, many analysts have focused on the power differences between the abusers and the abused. Victimization is typically unidirectional; the persons having the greatest ability to impose their will on others-because of their physical strength, and/or the social role they occupy, and/or societal norms-strike at those with considerably less power (Finkelhor, 1983; Received for publication June 2, 1989; final revision received September 14, 1989; accepted September 18, 1989. Reprint requests to Dr. Leonard Berkowitz, Vilas Research Professor, Department of Psychology, University of Wisconsin-Madison, Brogden Psychology Building, 1202 West Johnson St., Madison, WI 53706. 337

338

D. Dietrich. L. Berkowitz, A. Kadushin, and J. M&loin

Pagelow, 1984). Proceeding from this observation, these theorists tend to view the powerful persons’ aggressive action as largely an attempt at coercion carried out to stop the others’ offending behavior or, more generally, to make them accede to the powerful persons’ wishes (e.g., Patterson, 1982), or even to influence the distribution of the “scarce resources” within the domestic unit (e.g., Ball-Rokeach, 1980). Yet another analysis of domestic violence regards the caretakers’ behavior as being aimed at impression management. As an example, Felson (1978) maintains that the aggression is often an attempt to impress some audience, particularly by restoring a favorable “situational identity” after this identity has been devalued by a perceived humiliation. Applying this conception to child abuse, we would say the caretakers believe their young victim had threatened the image of themselves they wanted to present to a relevant audience (which could be themselves, the child, or other persons), perhaps by some act of defiance, and then hit the child in order to regain their identity, for example as the dominant member of the domestic unit. Berkowitz ( 1983, 1989) has raised a very different possibility. Defining aggression as a deliberate attempt to hurt someone, he differentiates between emotional (or hostile or angry) aggression and instrumental aggression. Instrumental aggression generally develops through learning that aggression can bring extrinsic rewards, and is basically an attempt to do harm in order to accomplish some purpose other than the intended target’s injury. These objectives might include stopping the victim from engaging in disapproved behavior or winning an audience’s approval. Berkowitz thus regards attacks carried out to coerce someone or to restore a favorable situational identity as being largely inst~mental in nature. Emotional aggression, by contrast, presumably arises when people feel bad. The negative affect they experience theoretically activates aggression-related ideas, memories, and expressive motor reactions; and these affect-generated responses produce an impulsive inclination to attack an available target, especially (but not only) the perceived source ofthe felt displeasure. According to this analysis, the emotional aggression’s primary aim is the in~iction of injury. While Berkowitz recognizes that emotionally aroused aggressors may pursue other, nonaggressive objectives on occasion (such as a position of power and dominance or social approval), he contends that angry persons often strike out impulsively with little thought in mind other than to hurt the target. And furthermore, from this perspective, the force of the attack can be influenced by external as well as internal stimulation so that the aggressors may do more harm than they had consciously intended. The present exploratory study of child-abusing adults attempts to investigate the applicability of these different theoretical approaches to child battering by seeking to determine what factors led these aggressors to regard their action as justified or unjustified. Such an inquiry conceivably could reveal what concerns were predominant in their thinking about the punishment they administered and might even suggest what kind of goal they were pursuing when they struck the victim. Suppose, for example, that those adult caretakers who thought their action had made the child comply with their wishes were especially likely to believe their behavior was justified. This finding could indicate that the aggression was mostly aimed at coercing the youngster so that they were pleased when they had accomplished this program. On the other hand, evidence that these adults tended to view their abuse as morally improper if they had lost their temper at the time they hit the child would suggest that (1) the attack had not been under complete control, and (2) this result was therefore consistent with Berkowitz’s conception of emotionally aroused impulsive aggression. Information about the factors influencing the abusive adults’ belief that their conduct was justified could also be useful to the practitioner. It might help reveal what kind of personal outlook serves to rationalize child abuse or even approve this behavior, thereby increasing the likelihood that the abuse will be repeated. Medical and mental health workers seeking to lessen domestic violence might then have a better idea of what types of beliefs and attitudes they should attempt to modify in their interventions with the adults.

Abusers’ justification of child abuse

339

The research data analyzed for this report were derived from extended interviews with parents identified by a county department of public welfare, and confirmed by protective service social workers, as having physically abused their child (inflicting a noticeable injury on the youngster) (cf. Kadushin & Martin, 198 1, for details). The median age of the children was 8 years. The interviews, conducted by two MSWs with considerable clinical experience, asked the adults for a retrospective and introspective recapitulation of the abusive incident for which they had been cited. The adults were to recall the event and report what had happened in considerable detail from the inception to the immediate aftermath of the incident. The interviews were tape-recorded and transcribed. Some time later each of the case descriptions was scored on a variety of indices, and multiple regression analyses were then conducted employing these measures in order to determine what characteristics had contributed to the abusers’ judgment that they had or had not been justified in punishing the child. Despite the two years and more spent in the data collection, the care taken in the interviews, and the time and effort devoted to coding the interview transcripts, the study should be regarded as an exploratory investigation since the results are based on a relatively limited sample.

METHOD Sample Seventy-three interviews with the abusers were completed, recorded, and transcribed for detailed analysis.. The structured interviews included both open- and close-ended questions. Information was obtained on the frequency and severity of the injury, the age and sex of the abused child, the child’s behavior immediately preceding the abuse, the parent’s response to that behavior, and the child’s reaction to that response. Furthermore, information about the parent’s and child’s feeling and thinking throughout and after the abusive incident was obtained. Finally, demographic information as well as the abuser’s history and family background, whether or not the abuser had been abused as a child, was recorded. Of the 73 abusers, 2 1 were men, 3 of whom were the child’s stepfather, with the remainder being the child’s natural father. Of the 18 natural fathers, 17 were married, whereas one was divorced. Two of the 52 female abusers were not the child’s natural mother; one was a single foster mother and the other the stepmother. The marital status characteristics of the 50 natural mothers were more diverse than those of the natural fathers: 19 were single, never married mothers; 14 were divorcees; 1, a widow; 5 mothers were separated from the father at the time; and the remainder (11) were married. The average age of the abuser was 30.6 years. The majority of the abusers (65.8%) were white; while 28.8% were black, one abuser was of Native American heritage, and the remainder were Hispanic. In the sample of the abused children, 42 were girls and 3 1 were boys. The mean age of the victims was 8.3 years. Note that the sex of the child was not taken into account in the later analysis due to the small size of the sample. Recognizing the validity problems in self-report data, information from the abusive parent was checked against the information available in the agency records, which included the social workers’ report of the incident immediately following investigation. The parents’ accounts presented in the research interview were, except in few instances, essentially comparable with the agency record material. Coding The interviews were coded to the following variables which were later used in the multiple regression analysis: l

Justification for abuse. An index of the abuser’s belief that the abuse was justified or not

340

D. Dietrich, L. Berkowitz, A. Kadushin, and J. M&loin

was established based on explicit indications as to ( 1) whether or not the abuser felt remorse about the abusive behavior (such as indications of feeling sorry for having abused the child, wishing he/she had not abused the child, expressing regret for the abusive behavior, etc.); (2) whether or not the abuser had indicated that he/she would probably react in a similar manner if the same situation arose again, presumably because the adult considered his/her action proper; (3) whether or not the abuser in essence blamed the child for the abuse (such as by indicating that the child was “asking for it,” deserved to be spanked, etc.). l History of abuse. Whether or not the abusers had themselves been abused as a child. l Cajun d&am? An index of the child’s level of defiance immediately preceding the abuse based on indications as to (1) whether or not the child’s behavior immediately preceding the abusive incident was accompanied by a display of anger (such as shouting, cursing, stomping feet, kicking, hitting, etc.); (2) whether or not the victim actively defied the abuser immediately before the abusive incident (such as by doing the exact opposite of what being told); (3) whether or not the victim’s preceding behavior was intentional as opposed to accidental (such as intentionally breaking something as opposed to accidentally spilling food, etc.); (4) whether or not the victim’s preceding behavior was directed toward the abuser. * Abuser depressed. Whether or not the abuser indicated feeling depressed or anxious prior to the abusive incident. l Abuser irritable. Whether or not the abuser indicated feeling irritable or angry prior to the abusive incident. l Abuser stress. A situational stress index was composed based on indications as to (1) whether or not the abuser had problems at the work place (such as impending lay-offs, troubles with colleagues or supervisors, feeling overworked or not challenged enough, etc.); (2) whether or not the abuser was encountering financial difficulties; (3) whether or not the abuser indicated having relationship problems (such as having marriage problems, problems with ~rl/~yf~end, etc.); (4) whether or not the abuser indicated experiencing mental health problems other than depression/anxiety or irritability/anger (such as hallucinations, extreme mood swings, etc.); (5) whether or not the abuser indicated any other unusual life events which could be stressful (such as being ill, crowded living arrangements, etc.). l C7hange.sin obedience. Whether or not there was a change in obedience immediately following the abuse (in other words, if the abusive incident seemed to have been inst~mental in changing the victim’s behavior). * Abuser loss oftemper. Whether or not the abuser indicated a loss of temper immediately preceding the abuse (such as reporting that he/she had “lost it,” “flying off the handle,” having been in a “frenzy,” etc.). rater-rater reliability In order to assess the inter-rater reliability of the content analysis, a second rater who was blind to the purpose of the content analysis was supplied with in-depth descriptions of the coding categories and asked to recode ten randomly chosen interview transcripts. Overall, the raters were found to be in 77% agreement. Inter-rater reliability for the different indices is reported below. Factor Analyses in the Index Construction In constructing four of the indices to be used in the multiple regression analyses, the items intended to be included in each index were first exposed to a factor analysis, as is further explained below, to determine if the component items did indeed cluster together in the same factor and what weight should be given to the individu~ items in arriving at the index score.

Abusers’ justification of child abuse

341

The use of the factor scores to determine the item weightings increases the reliability of the index (Nunnally, 1978). These four indices were the dependent variable dealing with the caretakers’ belief that their treatment of the child was justified, and the three of the four predictor variables, two having to do with the stress the adults were under at the time, and the other assessing their perception that the child was defiant.

RESULTS The Research Sample

in Terms offhe Characteristics

~~~t~~c~ionindex.The descriptive statistics for the individual chamcte~stics composing the jus~i~c~~io~index were as follows: For the remorse characteristic, in 58.9% of the cases the abuser indicated feeling remorse, with the remainder indicating no regrets about the abuse; for the blaming-the-~ctim chamcte~stic, 50.7% of the abusers blamed the victim and 49.3% of the abusers did not assign blame to the victim. For the justification characteristic, in 62.5% of the cases the abuser indicated feeling justified for the abusive behavior, with the remainder feeling not justified. When looking at all three characteristics combined, 33.3% of the uses received a score of 0 (indicating that the abuser neither blamed the victim nor indicated that the abuse was justified); 11.1 ok,of the cases received a score of 1, (having only one of the three chamcte~stics); 22.2% of the cases received a xore of 2; and the remaining 33.3% of the cases received a score of 3 (indicating that the abuser felt justified, blamed the victim, and felt no remorse for the abusive ~havior~. For the purpose of data reduction, as was mentioned before, the responses on the three characteristics, blaming-the-~ctim, feeling justified, and the reversely scored feeling remorse were combined via a factor analysis (p~ncip~ ~mponen~ analysis), and one factor emerged which accounted for 73% of the variance. The internai consistency of this index was high, with the inter~hara~e~stic reliability coefficient of .93 (omega, a variant of coefficient alpha); and the averaged inter-rater a~eement for the characteristic comprising this index was -77. Child de~a~ce index.A large number of the adult caretakers interpreted the child’s behavior preceding the abusive incident as intentional (4s cases), as opposed to accidental (27 cases), and in virtually all of the former cases as having been aimed at them personally. Indeed, almost half of the participants (33 cases) believed the youn~ter had been in active defiance, with the remainder (40 cases) not being classified as such. In the majority of the 73 cases (54), the child’s behavior was not accompanied by a display of temper. The factor analysis forming the child de~a~ce index was based on the i~dividu~ items, child’s display of anger, active defiance, intentionality of preceding behavior, and directionaiity of preceding behavior. One factor, child defiance, surfaced, with the highest loading on the active defiance item. This factor accounts for 5~.6~ffof the variance, with a reliability coefhcient (omega} of .83. Abuser History. Out of the 73 cases reviewed, 53.4% of the abusers (39) were found to have been abused themselves as children, whereas 46.6% (34) indicated not having been abused. Inter-rater agreement on this item was .80. ~i~uaiio~alstressom. The percentages of abusers indicating the presence of stressors were as follows: Some difficulties in the work place were reported by 23.3% of the abusers with the remainder not reporting any problems; financial problems were explicitly mentioned by

342

D. Dietrich.

L. Berkowitz,

A. Kadushin,

and J. McGloin

Table 1. Regression Coefficients, F Values, and Probabilities for the Variables Significantly “Predicting” Abusers’ Justification For Their Abuse Predictor

Beta

F

P

Child Defiance Abuser Loss of Temper Emotional Distress Environmental Stress

.45 -.34 -.22 .I9

15.09 13.88

1.001

Some factors influencing abusers' justification of their child abuse.

The interviews of abusive caretakers originally carried out by Kadushin and Martin (1981) were coded and subjected to statistical analyses to determin...
988KB Sizes 0 Downloads 0 Views