Faculty Bibliography

In this study, the authors assessed men's and women's partner and parent physical aggression among 453 representatively sampled families with young children. The prevalences of partner aggression and of severe parent aggression were higher than previously reported. Substantial rates of co-occurrence were found. Risk ratios and regression analyses indicated that connections between (a) husbands' and wives' partner aggression and (b) mothers' and fathers' parent aggression were especially strong. Patterns of co-occurrence pointed to the probable relative importance of family-level, in comparison with individual, predictors of aggression. Patterns of co-occurring violence are described in light of the theoretical literature. Implications for studying family violence in community samples are discussed.

This study examined the association between attitudes about dating aggression and select dating aggressive behaviors (verbal aggression and jealous behavior) in high school students. Our hypothesis, derived from cognitive dissonance theory, was that discrepancies between self-reported attitudes and aggressive behavior at Time 1 (i.e., putative cognitive dissonance) would predict decreases in aggression between Time 1 and Time 2 beyond what would be predicted by change in attitudes over the same period. Results indicated that cognitive dissonance (as indexed by the discrepancy between attitudes and behavior) was generally a significant predictor of behavior change, providing significant improvement in prediction of behavior over attitude change alone. We discuss the implications of these findings for prevention efforts and directions for future research in this area.

This paper describes two studies investigating the interrater agreement of severity scales for family maltreatment used in America's largest child and family maltreatment agency: the U. S. military's Family Advocacy Program (FAP). The USAF-FAP Severity Index is a multidimensional rating system for clinicians' evaluations of the severity of seven forms of family maltreatment: partner physical, emotional, and sexual abuse; child physical, emotional, and sexual abuse; and child neglect. The first study evaluated the reliability of the scale as it is used in the field. The second study compared a generalizable sample of clinicians' ratings to an established "gold standard" of what the ratings should have been. The Severity Index demonstrated fair-to-good levels of reliability, suggesting that with minimal cost, investigating caseworkers can routinely assess, and make fairly reliable ratings of, the severity of seven forms of family maltreatment for each case they investigate.

The stability of and dyadic influences on physical aggression in adolescents' dating relationships have implications for understanding the etiology of intimate partner violence and, in turn, prevention efforts. We studied the stability of aggression and tested a longitudinal dyadic model of psychological and physical aggression in samples of adolescent males and females who remained in relationships for 3 months. Physical aggression against dating partners was remarkably stable. Verbal aggression, jealous behavior, and controlling behavior formed a latent construct psychological aggression. Psychological aggression predicted physical aggression both concurrently and longitudinally. Dyadic relations were evident for both psychological and physical aggression, and these dyadic relations highlight the need for prevention and intervention incorporating dyadic issues with young dating couples.

We examined the relations between preschool boys' behavior problems and mothers' interpretations of children's emotion expressions. A sample of 31 mothers of oppositional boys and 28 control mothers responded to standard stimuli depicting child emotional reactions to maternal control attempts; mothers were instructed to think of the stimuli as either (a) their own child or (b) an unfamiliar child. Mothers of oppositional boys were more likely to generate negative interpretations than were control mothers when thinking of their own children; however, this difference did not generalize to the explicitly unfamiliar child condition. Mothers of oppositional boys demonstrated negative and comparison mothers demonstrated positive interpretive tendencies toward their own children. Findings suggest that child emotion cues may trigger biased maternal cognitions even in the absence of child misbehavior.

The cycle of violence posits that victimized children grow up to victimize others. Three forms of the cycle have never been tested: whether exposure to physical victimization and interparental violence additively or interactively increase risk for adulthood (a) child abuse perpetration; (b) partner abuse perpetration; or (c) partner abuse victimization. These hypotheses were tested in a nationally representative data set (1985 National Family Violence Survey) comprising 6,002 participants. Dually exposed, compared to singly exposed, women had significantly increased risk for adulthood family violence. Frequency of family-of-origin violence predicted adulthood child and partner abuse through both main and interactive effects.

Aggression in dating relationships is associated with attitudes that justify its use. Attitudes about dating aggression are targeted by prevention efforts, contributing to a need to measure these attitudes sensitively, accurately, and multidimensionally. We describe two new measures of attitudes about aggression, each tapping different attitudinal components, and compare their psychometric properties with an existing scale. The 1st assesses attitudes about physical aggression in provocative situations. The 2nd taps attitudes about verbally aggressive, controlling, and jealous tactics against a dating partner. Data from 2,313 high school students were factor analyzed and cross-validated for each new scale. Compared with an existing measure, the scales had comparable levels of reliability and validity and improved response distributions. A 2nd-order factor analysis lends support to a multidimensional view of attitudes about dating aggression.

Research and treatment for partner abuse and child abuse are relatively distinct enterprises, yet when the theoretical and research literatures related to these two forms of family violence are examined together, the likelihood of meaningful associations is strikingly apparent. Partner abuse and child abuse co-occur more often than one might expect by chance, and in the context of overlapping theories and risk factors, this suggests that a more integrated conceptualization might be fruitful. We summarize and provide a framework for describing the large number of similar predictors of partner and child abuse, identify some potentially interesting dyadic differences, comment on how our research methodologies could address an integrated area of family violence, and suggest some directions for future research and treatment.

Divorce prediction studies (e.g., Gottman, Coan, Carrere, & Swanson, 1998) suggest that couples' eventual divorce can be very accurately predicted from a number of different variables. Recent attention to these studies has failed to consider the need to crossvalidate prediction equations and to consider the prevalence of divorce in the population. We analyze archival data to demonstrate that accuracy and predictive value drops precipitously during crossvalidation. We conclude that results of studies without crossvalidation analyses should be interpreted with extreme caution, no matter how impressive the initial results appear to be.

We review the risk and protective factors for child sexual abuse. Overall, characteristics of perpetrators, victims, and families of victims were moderate to strong risk factors for child sexual abuse. However, it is difficult to distinguish between risk factors for extra-familial and intra-familial child sexual victimization because most of the studies combined these two types of child sexual abuse, although the risk factors for these two types of child sexual abuse most likely differ. Research in this area is difficult because etiological and prevention models of victimization would differ substantially from those of perpetration. Given the low yearly prevalence of child sexual victimization, very large samples would be necessary to obtain sufficient power. Thus, most studies have used lifetime prevalence, which may provide much useful information but which add substantial time confounds. Finally, child sexual victimization is probably a misnomer, as the nature, impact, and etiology of sexual victimization most likely differs over the large age span of childhood and gender. Because improved models and prevention programs require improved etiological models (based on knowledge of risk and protective factors), we hope that this review will focus stakeholders on the need for continued research in this area. (C) 2001 Elsevier Science Ltd. All rights reserved.