2. Intergenerational Transmission and Gender
Although boys and girls rely on both parents as a relevant source of information in constructing their own beliefs and behaviors, a major question has been the degree to which the process of intergenerational transmission operates differently for males and females. Are boys and girls vulnerable to the same degree (i.e., does childhood exposure to marital violence have the same effects on males and females)? Bandura’s (1973) social learning theory states that the ability to influence through modeling depends upon the degree to which the child identifies with the model. This suggests another research question: Do gender differences exist in the modeling of behavior? e.g., are boys more likely to imitate their fathers and girls their mothers?
A community sample of adolescents, some with histories of childhood maltreatment, sheds some light on the first question. The maltreated adolescents, compared with nonmaltreated youths, showed differential patterns of adjustment problems and dating violence. Female adolescents with maltreatment histories reported considerable emotional distress (such as anger, depression, and anxiety), posttraumatic stress–related symptoms, and acts of violent and nonviolent delinquency compared with girls without such histories. Male adolescents with maltreatment histories reported fewer symptoms of emotional turmoil and delinquent behavior but were significantly more likely to be abusive toward their dating partners than boys without a maltreatment history (Wolfe et al. 2001).
Overall, the study of gender effects has produced mixed findings (Stith et al. 2000), with some studies showing that direct and/or indirect exposure to violence in childhood is more salient for females (Forsstrom and Rosenbaum 1985), some studies showing stronger effects among males (Rosenbaum and O’Leary 1981), some showing a same-gender modeling effect (Heyman and Slep 2002), and some showing no sex-specific differentiation (Cappell and Heiner 1990). A review of eight recent studies (Cummings, Pepler, and Moore 1999) examining the impact of interparental violence on children ranging in age from four to sixteen showed that girls exposed to interparental violence displayed higher internalizing scores than did exposed boys in the six studies that reported on internalizing outcomes. Of the five studies that reported externalizing scores, three studies reported higher scores for girls and two for boys. These studies suggest that the social context of the home may be more salient for girls than for boys. Data from the National Youth Survey also support the premise that prior experiences with violence may be more salient for females than for males. In this study, which examined both witnessing parental violence and experiencing child abuse, only witnessing violence had an effect on later violence and only for females. However, this effect was not direct and operated through other variables, such as marital satisfaction (Mihalic and Elliott 1997). A longitudinal study of parenting practices experienced in three distinct developmental periods while growing up also provides evidence of intergenerational transmission for females only (Belsky et al. 2005).
In contrast, Rosenbaum and O’Leary (1981) found that the effects of witnessing parental violence as children on later violent behavior were especially strong for males. Women who were victims of physical marital violence were no more likely than women in two control groups (composed of women who had suffered no physical abuse; one group claimed to have satisfactory marriages and the other group discordant marriages) to have witnessed spouse abuse between their parents. However, abusive husbands were much more likely to have come from families characterized by marital violence than husbands in the two control groups.
Others theorize that modeling of marital aggression is not sex specific, but is role specific. The 1975 National Family Violence Survey examined this perspective, finding that females who had observed fathers hitting mothers were just as likely to be the perpetrators of violence as the victims, and males were as likely to be the victims as well as perpetrators of marital violence. Kalmuss (1984) concluded that the intergenerational transmission of aggression involves both generalizable and specific models. Generalized models increase the likelihood of any form of family aggression in the next generation, and specific models increase the likelihood of particular types of family aggression (e.g., children who observe aggressive acts between their parents are more likely to model aggressive behavior in their own marriages) (see also Seltzer and Kalmuss 1988). A later analysis using the 1975 National Family Violence Survey also found evidence that the existence of spousal violence in the family of origin increased the likelihood that the respondent, whether husband or wife, would be the target of aggression, but no evidence was found for sex-specific acquisition of the perpetrator role (Cappell and Heiner 1990). Findings from the 1975 National Family Violence Survey are in direct contrast to the 1985 National Family Violence Survey, which provided support for a same-gender modeling effect for perpetration of violence toward partners and children. Men’s risk was increased by exposure to father-to-mother violence, and women’s risk was increased by exposure to mother-to-father violence (Heyman and Slep 2002).