J Youth Adolescence (2015) 44:745–760 DOI 10.1007/s10964-014-0195-7

EMPIRICAL RESEARCH

Longitudinal Relationships Between Individual and Class Norms Supporting Dating Violence and Perpetration of Dating Violence Katherine A. Taylor • Terri N. Sullivan Albert D. Farrell



Received: 21 July 2014 / Accepted: 24 September 2014 / Published online: 2 October 2014 Ó Springer Science+Business Media New York 2014

Abstract Dating violence is commonly perpetrated in adolescence, making it imperative to understand risk factors in order to inform prevention efforts. Although individual norms supporting dating violence are strongly related to its perpetration, few studies have examined their longitudinal impact. Moreover, the influence of class norms (i.e., norms for students in the same grade, cohort, and school) supporting dating violence on perpetration has rarely been studied. The current study examined longitudinal relationships between individual and class norms supporting dating violence and perpetration of physical and psychological dating violence. Participants were two cohorts of sixth graders from 37 schools who were in dating relationships at Wave 1 and 6 months later at Wave 2 (N = 2,022; 43 % female; 52 % African American, 21 % Latino/a, 20 % White, and 7 % other). The analyses used a multilevel approach, with students represented at Level 1 and classes (n = 74) at Level 2. The models tested direct effects of Wave 1 individual and class norms supporting dating violence on subsequent changes in perpetration of dating violence at Wave 2 and the extent to which gender moderated these relationships. The findings indicated that greater individual norms supporting male dating violence predicted greater change in perpetration of physical and psychological dating violence and greater individual norms supporting female dating violence K. A. Taylor (&)  T. N. Sullivan  A. D. Farrell Department of Psychology, Virginia Commonwealth University, Richmond, VA 23284, USA e-mail: [email protected] T. N. Sullivan e-mail: [email protected] A. D. Farrell e-mail: [email protected]

predicted greater change in perpetration of psychological dating violence. Greater class norms supporting male dating violence predicted greater change in perpetration of physical dating violence; whereas greater class norms supporting female dating violence predicted less change in perpetration of physical dating violence. These findings highlight the need to address norms in early adolescence. Keywords Norms

Adolescence  Dating violence  Perpetration 

Introduction Adolescent dating violence is a significant public health concern. Recent estimates based on ethnically diverse samples of youth indicate that approximately 30 % of youth in a romantic relationship report perpetration of physical dating violence (O’Leary et al. 2008; Simon et al. 2010) and between 40 % (Foshee et al. 2009) and 90 % (O’Leary et al. 2008) are involved in psychological dating violence, such as controlling behavior and insults. Given these high prevalence rates, it is critical to examine risk factors for perpetration of dating violence that can be used to inform prevention efforts. Previous research indicated that norms supportive of dating violence are strongly related to higher frequencies of perpetration of dating violence during adolescence (e.g., Malik et al. 1997; O’Keefe 1997; Reeves and Orpinas 2012; Simon et al. 2010). However, this research has several limitations. First, it is largely cross-sectional, with only one study examining the longitudinal impact of individual norms on perpetration (Foshee et al. 2001). Second, it focused solely on perpetration of physical dating violence and neglected to examine the predictive role of these norms for

123

746

psychological forms of perpetration. Additionally, very little research has focused on norms supporting dating violence at the class level (i.e., norms for students in the same grade and cohort at the same school). According to social learning theory (Bandura 1977), youth who are exposed to norms accepting aggression are more likely to demonstrate aggressive behavior. Related research found support for this premise in an investigation of the impact of class norms supporting physical and verbal aggression on teacher reports of adolescents’ physical and verbal aggressive behavior (Henry et al. 2000). Although some research has addressed the influence of friends’ norms accepting dating violence on perpetration of dating violence (e.g., Foshee et al. 2004), studies have not examined the impact of norms among the broader peer group on changes in perpetration over time. Thus, we aimed to address these gaps in the dating violence literature by examining the longitudinal impact of individual and class norms supporting dating violence on perpetration of physical and psychological dating violence among early adolescents. We also examined the moderating role of gender on relationships between individual norms and perpetration of dating violence. Adolescent Dating Relationships Involvement in romantic relationships is an important part of development for many adolescents. Approximately 25 % of youth report involvement in a romantic relationship around the age of 12, with rates increasing to 70 % by the age of 18 (Connolly and McIsaac 2009). According to Brown’s (1999) development-contextual stage theory of romantic relationship development, adolescent dating experiences tend to follow a developmental sequence of four stages, which coincide with changes in the dynamics of peer relationships. During the initiation phase, adolescents begin to explore romantic interests within the context of same-sex peer groups. As such, this stage is mostly characterized by talking to same-sex peers about infatuations and, subsequently, learning peer group norms for romantic behavior such as those related to approaching and initiating interactions with a potential dating partner (Connolly and McIsaac 2009). During the status phase, adolescents experience increased opportunities to interact with potential romantic partners and begin to form initial romantic relationships, which are largely influenced by the peer group (Connolly and McIsaac 2009). During the final two phases, the affection phase and bonding phase, romantic relationships become increasingly intimate and committed (Brown 1999). Early adolescence, ages 10–14, represents an important timeframe to study romantic involvement because dating is often novel, heavily influenced by peers, and potentially

123

J Youth Adolescence (2015) 44:745–760

difficult to negotiate as adolescents are learning and attempting to ascribe to behavioral norms related to dating that are often driven by peer groups (Connolly and McIsaac 2009). For example, a qualitative study of early adolescent dating problems revealed that adolescents often have difficulties approaching potential partners, and also experience uncomfortable or inappropriate approaches from other adolescents (Sullivan et al. 2010). Additionally, peers often play an integral role in the development and maintenance of early adolescent romantic relationships (Brown 1999). For instance, during the initial phases of romantic relationship development, peers exert their influence by communicating norms and attitudes regarding appropriate partner selection (Brown 1999). Beyond this influence, peers also play a role in relationship maintenance, acting in ways that may support or harm peers’ romantic relationships (Giordano et al. 2006). Adolescent Dating Violence At every stage of development, dating relationships can have a positive influence on adolescents’ socio-emotional competence, as they play a key role in identity development, improve self-esteem, prepare adolescents for future romantic relationships, and provide a key source of social support (Collins 2003). However, they can also place youth at risk for victimization and perpetration in dating contexts. Adolescent dating violence can include physical and psychological violence within a dating relationship. Physical dating violence involves the use of force to inflict fear or injury (e.g., kicking, pushing, and scratching); whereas psychological violence includes non-physical acts that are intended to inflict emotional harm (e.g., monitoring behavior and whereabouts, insults, and emotional putdowns). Unfortunately, dating violence commonly occurs during adolescence. For example, approximately 30 % of girls and 26 % of boys in a large ethnically diverse sample of sixth graders reported involvement in physical dating violence (Simon et al. 2010). Rates are even higher for psychological violence, with recent studies reporting rates as high as 40 % (Foshee et al. 2009) and 90 % (O’Leary et al. 2008). Previous research has revealed mixed findings regarding gender differences in prevalence rates. For the most part, researchers found that dating violence is reciprocal during adolescence in that boys and girls are equally likely to be both victims and perpetrators of moderate levels of psychological and physical dating violence (Foshee et al. 2009). A growing body of research also shows that, during adolescence, girls perpetrate more physical and psychological dating violence than boys (e.g., Wolfe et al. 2003). Yet, studies have also shown that male youth are more likely to perpetrate severe physical dating violence, making

J Youth Adolescence (2015) 44:745–760

female youth more susceptible to serious injury (Foshee et al. 2009). Previous research has also investigated the influence of other socio-demographic factors on perpetration of dating violence. Several studies suggest racial/ethnic differences in rates of perpetration. For example, Foshee et al. (2009) found that minority youth (represented predominately by African American adolescents and smaller percentages of Asian American, American Indian, and bi-racial adolescents) perpetrated higher levels of moderate physical dating violence than White youth. Lower parental education was also associated with increased perpetration of psychological and moderate levels of physical dating violence, such that adolescents whose parents did not graduate high school perpetrated more dating violence than those adolescents whose parents did and adolescents whose parents finished high school but not college perpetrated more dating violence than those who finished college. Similarly, findings indicated that adolescents from single-parent households perpetrated more psychological and severe physical dating violence than adolescents from two-parent households. Other research highlights the role of neighborhood context in adolescent dating violence. For instance, previous research suggests that youth who live in low-income, inner-city neighborhoods may be at increased risk for dating violence due to relatively higher violence rates that may characterize some of these neighborhoods (Spriggs et al. 2009; Wekerle and Wolfe 1999). Class Norms Supporting Dating Violence It is imperative to investigate risk factors for perpetration of dating violence, so as to better inform comprehensive prevention efforts. Several theories suggest that peers are an important influence on dating behavior, especially during early adolescence when dating is often novel and peer relationships take on greater importance (Brown 1999). Drawing from social learning theory (Bandura 1977), youth who are exposed to peer norms and behavior in support of dating violence may be more likely to perpetrate violence toward a dating partner. Previous research has evidenced the impact of friends’ norms and involvement in dating violence on perpetration of dating violence. For example, in a study of risk factors for dating violence among 1,965 eighth- and ninth-grade youth, Foshee et al. (2001) found that in cross-sectional analyses, having friends who were victims or perpetrators of physical dating violence and being accepting of dating violence were positively related to perpetration of physical dating violence for girls. Qualitative research has revealed similar findings. In a focus group study with middle school students, youth described that boys who do not receive peer support for treating girlfriends in a respectful way might

747

perpetrate physical dating violence against their girlfriends to maintain their reputation with their peers (Noonan and Charles 2009). Thus, previous studies provide evidence for the important influence of friends’ attitudes and behaviors toward dating partners on adolescents’ involvement in dating violence. Although friends represent one influential peer group, adolescents are increasingly exposed to a larger social network. Early adolescence represents an important shift in social context, as youth transition from self-contained elementary school classrooms where they spend most of the day with the same group of students to middle school settings where they are exposed to multiple groups of peers as they change classrooms throughout the day (Henry and Chan 2010). As such, youth are exposed to multiple groups of peers ranging from the closer friend group to a broader group of peers in the same grade, cohort, and school. Subsequently, youth in the same grade, cohort, and school constitute one peer group that we refer to as the ‘‘class.’’ Researchers have also argued that due to changes in school organization, the class represents the ecological unit most proximal to youth in middle school (Bernburg and Thorlindsson 2005; Henry and Chan 2010). With the increasing impetus for prevention and intervention efforts to address risk factors for aggression at multiple levels of the school environment, including the individual and school levels (e.g., Olweus Bullying Prevention Program: Olweus and Limber 2007), it is imperative to consider the role of norms supporting dating violence among youth in the same grade, cohort, and school (i.e., class). Researchers (e.g., Henry et al. 2000) examined class norms supporting general aggression and found that they are strongly associated with aggressive behavior. For instance, among a large ethnically diverse sample of elementary-aged children, Henry et al. (2000) found that class norms supporting aggression significantly predicted changes in aggressive behavior 1 year later. Yet, previous research has not examined the impact of class norms supporting dating violence on perpetration of dating violence, specifically. We sought to address this gap in the literature and inform prevention and intervention efforts targeting school-level risk factors by examining the impact of a class-level measure of norms on perpetration of dating violence. Further, based on previous conceptual and empirical efforts (e.g., Bandura 1977; Foshee et al. 2001; Henry et al. 2000; Noonan and Charles 2009), we anticipated that class norms supportive of female and male dating violence would predict increased perpetration of physical and psychological dating violence over time. Individual Norms Supporting Dating Violence Considerably more research focuses on individual norms supporting dating violence and indicates that youth’s

123

748

beliefs about relationships and specifically their attitudes towards dating violence are strongly associated with perpetration of dating violence. For instance, in an earlier study from the same ethnically diverse sample of sixth graders used in the current study, Simon et al. (2010) found that 30 % of youth with a boyfriend or girlfriend reported attitudes and norms supporting male perpetration of dating violence and 60 % of youth reported attitudes and norms supporting female perpetration of dating violence. Youth who reported greater support for dating violence were more likely than youth who reported lower support to engage in perpetration of physical dating violence. Similarly, in an ethnically diverse sample of 719 high school students, Malik et al. (1997) found that the acceptance of dating violence was associated with increased likelihood of perpetrating physical dating violence for both boys and girls. Among an ethnically diverse sample of 939 14–20 yearolds, O’Keefe (1997) found that beliefs about the justifiability of dating violence were positively associated with perpetration of physical dating violence. Specifically, whereas norms supporting male dating violence were more strongly associated with perpetration of physical dating violence for boys, norms supporting female violence were more strongly associated with perpetration of physical dating violence for girls. Lastly, in an ethnically diverse sample of ninth grade adolescents, Reeves and Orpinas (2012) found that there were strong concurrent relationships between norms supporting dating violence and perpetration of physical dating violence for boys, such that greater norms supporting male and female dating violence were associated with higher levels of perpetration. But, similar relationships were not found for girls. To date, we could only find one study (i.e., Foshee et al. 2001) that examined the longitudinal relationship between individual norms and perpetration of dating violence. Findings from this study indicated that norms accepting dating violence significantly predicted perpetration of dating violence (i.e., a composite of physical and sexual violence) one and a half years later for boys, but not girls. Given the lack of longitudinal research and the mixed findings regarding the role of gender in relationships between individual norms supporting dating violence and perpetration, the current study aimed to examine the impact of individual norms on perpetration of physical and psychological dating violence sixth months later as well as the moderating role of gender on these relationships. Therefore, we anticipated that individual norms supporting female and male dating violence would predict increased perpetration of physical and psychological dating violence over time. We also hypothesized that gender would moderate the relationships between individual norms and perpetration. Specifically, we expected that relationships between norms supportive of male dating violence and

123

J Youth Adolescence (2015) 44:745–760

perpetration of physical and psychological dating violence would be stronger for boys and relationships between norms supportive of female dating violence and perpetration of both forms of dating violence would be stronger for girls.

Hypotheses The current study expands on the literature to date by examining longitudinal risk factors at the individual- and class-level for perpetration of physical and psychological dating violence, among an ethnically diverse sample of early adolescents. Although previous research (e.g., Malik et al. 1997; O’Keefe 1997; Reeves and Orpinas 2012), including a prior study conducted using the same data as the current study (i.e., Simon et al. 2010), has examined the relationship between individual norms and perpetration of dating violence, these studies were cross-sectional and focused solely on physical dating violence. In addition, the current study was based on data from the Multisite Violence Prevention Project (MVPP 2004), which included two cohorts of early adolescents from a large number of schools from different communities in different parts of the U.S. This data offered the unique opportunity to examine the impact of class-level variables (i.e., class norms supporting male and female dating violence) on perpetration of dating violence, an issue that previous studies have not been able to adequately address. Based on the research reviewed above, we made the following hypotheses. First, given research showing that the rates of perpetration of dating violence vary by gender, race/ethnicity, family structure, and neighborhood disadvantage, we hypothesized that the levels of physical and psychological perpetration would be higher among girls, African American youth, adolescents from single–parent homes, and youth in more disadvantaged neighborhoods and controlled for these variables in all models. In regards to the longitudinal relationships among individual- and class-level norms supportive of dating violence, gender, and perpetration of physical dating violence we expected that (a) individual norms supportive of male and female dating violence would predict changes in perpetration of physical dating violence 6 months later (i.e., from the fall to spring of sixth grade) (controlling for demographic characteristics and Wave 1 physical and psychological perpetration), (b) the degree to which individual norms were associated with changes in perpetration of physical dating violence would vary as a function of gender, and (c) class norms supporting male and female dating violence would predict changes in class levels of perpetration of physical dating violence 6 months later (controlling for concentrated disadvantage and intervention condition). In

J Youth Adolescence (2015) 44:745–760

749

regards to the longitudinal relationships among individualand class-level norms supportive of dating violence, gender, and perpetration of psychological dating violence we hypothesized that (a) individual norms supporting male and female dating violence would predict changes in perpetration of psychological dating violence from 6 months later (controlling for demographic characteristics and Wave 1 physical and psychological perpetration), (b) the degree to which individual norms were associated with changes in psychological perpetration would vary as a function of gender, and (c) class norms for male and female dating violence would predict changes in class averages of perpetration of psychological dating violence 6 months later (controlling for concentrated disadvantage and intervention condition).

American, 21 % Latino/a, 20 % White, and 7 % other. The current study focused on the first two waves of data which were collected near the beginning and end of the sixth grade school year, pre- and post-intervention. Only youth who reported being involved in a dating relationship within the last 3 months at both waves (2,022 students, 43 % female) were included in the final sample. Over half (55 %) of this sample identified themselves as African American, 17 % as Latino/a, 16 % as White, 9 % as Multiracial, and 3 % as another race/ethnicity. Approximately 43 % reported living in a two-parent household, 28 % lived in a single parent household, 12 % lived with a parent and a stepparent or parent’s significant other, 9 % lived with a single parent and additional family members, 5 % lived with an adult relative, and 3 % reported another family structure.

Method

Procedure

Setting

The Institutional Review Boards at the four participating universities and the CDC approved all study procedures. The MVPP employed a cluster-randomized design, such that 8 to 12 schools were recruited within each geographic site and were randomly assigned to one of four conditions (i.e., universal intervention, targeted intervention, combined universal and targeted intervention, or no-intervention control), with approximately equal numbers of schools in each condition (see Henry et al. 2004 for details). The universal intervention consisted of the GREAT student curriculum and the GREAT Teacher Program (Orpinas et al. 2004). The 20-session student curriculum used a variety of teaching techniques including repetition, small group activities, experiential learning techniques, and didactic modalities to educate students on how to avoid dangerous situations, ignore teasing, ask for help, talk things through, defuse situations, and help peers. The GREAT Teacher Program included a 12 h workshop and ten consultation sessions designed to increase awareness about forms and risk factors of aggression, improve classroom management, and promote strategies to reduce aggression and assist victimized students. The targeted intervention was designed for a subset of sixth graders who were identified based on teacher ratings indicating high levels of aggression and influence over their peers. This intervention consisted of a 15-week family program conducted in small groups of students and their parents or guardians (Smith et al. 2004). None of these interventions specifically targeted dating violence. The first cohort of youth was recruited in the fall of 2001 and the second cohort the following year. Students selected for the first cohort who repeated the sixth grade were not eligible to be recruited for the second cohort. Consent and assent forms were sent home with students. At three of the

The present study used data from the MVPP (2004), a 7-year project funded by the National Center for Injury Prevention and Control (NCIPC) at the CDC. Middle school students participated in an experimental study that assessed the impact of violence prevention programs for sixth grade students at 37 schools across four geographical sites (Durham, NC; Richmond, VA; Chicago, IL; and northeast GA). The participating schools at every site, but Chicago, included students in sixth through eighth grade. In Chicago, the schools served students in kindergarten through eighth grade. A high percentage of students at the participating schools were from low-income families, as 42 to 96 % of students qualified for the federally subsidized lunch program across sites. At the time of the study, the average poverty rate and youth arrest rate in the selected school districts (28 % and 63/100,000, respectively) were higher than the national average (16 % and 43/100,000, respectively; Henry et al. 2004). Participants At each school, two cohorts of sixth graders were recruited from a random sample of approximately 100 students, except at three Chicago schools where all sixth graders were recruited. Students in self-contained special education classrooms were not eligible to participate. Active parental consent and student assent were obtained from 5,811 of the 7,343 eligible students. At each wave, data were collected only from students who remained in their original school, thus, one or more waves of data were available for 5,465 students (97 % of those consented and eligible). Approximately 49 % of participants were male; 52 % were African

123

750

four sites students who returned consent forms received a $5 gift card, whether or not they or their parents agreed to participation. Students completed the survey using computer-assisted self-interviewing with audio (Audio-CASI). Students were administered the survey at school in groups of 10–20; they were seated such that their responses were not visible to others. Students read the questions on the computer while listening to them through headphones, and then entered responses using the keyboard. When permitted by schools, students received a $5 gift card for participating in the survey. Measures For the present study, measures were used to assess demographic characteristics and three constructs including perpetration of physical and psychological dating violence, norms supporting dating violence, and concentrated disadvantage.

J Youth Adolescence (2015) 44:745–760

never to very often). Participants who responded yes to an item asking if they had a boy/girlfriend in the last 3 months were then asked whether or not they had ‘‘done any of the following things to a boyfriend/girlfriend.’’ Youth were asked not to count instances of self-defense. The Physical Perpetration subscale consists of nine items, such as, ‘‘How many times have you punched or hit him/her with something that could hurt?’’ The Psychological Perpetration subscale consists of four items, such as, ‘‘How many times have you said things to hurt his/her feelings on purpose?’’ Youth indicated the frequency of these behaviors, using a four-point response scale, where 0 = Never, 1 = 1–3 times, 2 = 4–9 times, 4 = 10 or more times. Higher scores indicate higher levels of dating violence. At Wave 1 and 2, Cronbach alphas for the Physical Perpetration subscale were .90 and .91, respectively, and for the Psychological Perpetration subscale .72 and .73, respectively. Norms Supporting Dating Violence

Demographic Characteristics Demographic characteristics were assessed by student report. Students were asked to identify their gender. To assess race and ethnicity students were asked two questions. Students were first asked to indicate whether they were Hispanic or Latino and were then asked to choose as many categories that applied to them (e.g., White, Black or African American, American Indian or Alaska Native). A single variable representing race/ethnicity was derived based on their responses to these questions that represented the following race/ethnicities: White, Black or African American, American Indian or Alaska Native, Asian Indian, Other Asian, Hispanic, other race, non-Hispanic, and Multiracial. To assess family structure students were asked who lived in their home and we allowed to choose as many people that applied (e.g., mother, father, stepmother, grandmother). A single variable representing the following typical family structures was derived from students’ responses: two-parent family, multi-generational single parent family, parent with either stepparent or parent’s significant other, single parent with or without other adults, foster family, adult relative with neither parent, or other. Perpetration of Dating Violence Perpetration of dating violence was measured using an adaptation of the Dating Violence and Norms measure (Foshee et al. 1996). Adaptations included minor wording changes for three items, the addition of two items (i.e., ‘‘Stomped out of the room or house during a disagreement’’; ‘‘Punched or hit you/him or her with something that could hurt’’), and a change in the anchors used for the Physical Dating Violence subscale (i.e., changed from

123

Norms supporting dating violence were measured using an adapted version of the Norms for Dating Violence Scale (Foshee et al. 1998). The measure was adapted so that half of the items reflected norms supportive of female dating violence and the other half reflected norms supportive of male dating violence. Students were asked to rate how strongly they agreed or disagreed with ten beliefs regarding violence toward a dating partner. Norms supporting male dating violence were assessed by five items, such as ‘‘It is okay for a boy to hit his girlfriend if she did something to make him mad.’’ Norms supporting female dating violence were assessed using five corresponding items reworded to address norms regarding violence by girls against their boyfriends. Students rated their agreement with each item on a 4-point scale, where 1 = Strongly disagree, 2 = Disagree somewhat, 3 = Agree somewhat, and 4 = Strongly agree. Wave 1 class-level measures of norms supporting dating violence were created by aggregating norms across students in the same grade and cohort within each school. This aggregate was based on data from the whole sample recruited for this study (average n = 5,417), not just youth who were in a dating relationship at Wave 1 and Wave 2. This was possible as all students completed the norms measure whether or not they were in a dating relationship. At Wave 1 Cronbach alphas for the individual norms supporting female and male dating violence subscales were .73 and .74, respectively. Level 2 reliabilities for the class norms supporting female and male dating violence were .59 and .56, respectively. Concentrated Disadvantage A measure of neighborhood concentrated disadvantage was created based on the approach used by Sampson et al.

J Youth Adolescence (2015) 44:745–760

751

(1997). Utilizing U.S. Census data from 2000 (U.S. Census Bureau 2000) factor scores were calculated for census tracts within each school district. Concentrated disadvantage was a weighted composite of poverty rate, percentage of female-headed households, per capita income, percentage receiving public assistance, percentage renting, percentage with less than a high school diploma, and percentage unemployed. Data Analysis The data from MVPP underwent rigorous data cleaning procedures in order to identify data entry errors, outliers,

Fig. 1 Multilevel model depicting regression of Wave 2 perpetration of dating violence on a Wave 1 individual-level norms supporting dating violence and gender and b classlevel norms supporting dating violence. The individual-level model includes a random intercept (black circle) that is modeled at the class level

and suspect response patterns. The primary analyses were conducted using Mplus version 7 (Muthe´n and Muthe´n 2012). The analyses used a multilevel approach whereby student-level variables were represented at Level 1 and class-level variables were represented at Level 2 (Fig. 1). Class-level variables were based on ratings by students in the same grade and cohort within each school, based on the notion that youth in the sixth grade at the same time would experience a more similar school environment than those in the sixth grade during a different school year (Henry et al. 2011). This yielded a total of 74 clusters, with an average of 27 students per cluster. Missing data on outcome variables (i.e., perpetration of psychological dating violence:

Family Structure Race/Ethnicity Male Gender W1 Physical Perpetration W1 Psychological Perpetration W2 Perpetration of Dating Violence

W1 Norms Supporting Male DV W1 Norms Supporting Female DV W1 Norms Supporting Male DV x Male Gender W1 Norms Supporting Female DV x Male Gender

(a)

Intervention Condition

Concentrated Disadvantage W2 Perpetration of Dating Violence W1 Norms Supporting Male DV

W1 Norms Supporting Female DV

(b)

123

752

n = 9; perpetration of physical dating violence: n = 10) due to incomplete responses were handled using fullinformation maximum likelihood (FIML), a method which estimates missing data based on the variables that are present (Muthe´n and Muthe´n 2012). The models were calculated with the MLR estimator, an Mplus option for maximum likelihood estimation with robust standard errors to account for the non-normality and non-independence of study variables. Gender and family structure and race/ethnicity variables were included as Level 1 covariates. Family structure was dummy-coded into two variables: Single-parent and other family structure, with two-parent family structure as the reference group. Race/ethnicity was dummy coded into the following three variables: White, Latino/a, and other race/ethnicity, with African American as the reference group. Intervention condition and concentrated disadvantage were included as Level 2 covariates. Intervention condition was dummy-coded into three variables: universal intervention, targeted intervention, and combined intervention, with the control condition as the reference group. The interaction terms for norms and gender were defined in Mplus by creating a variable representing the product of gender and norms supporting dating violence. All predictor variables, except for the dummycoded variables, at Levels 1 and 2 were grand-mean centered to facilitate interpretation. The models were run separately for each outcome (i.e., perpetration of psychological and physical dating violence). First, we ran a baseline model with random intercepts. Next, we ran models including Level 1 predictors with fixed slopes and random intercepts followed by models including Level 2 predictors (Hox 2010). The models were run using a hierarchical approach, such that variables were entered in sets in the following step-by-step progression: (1) control variables at Level 1 (i.e., family structure, gender, and race/ethnicity), (2) Wave 1 perpetration of psychological and physical dating violence at Level 1, (3) Wave 1 individual norms supporting female and male dating violence at Level 1, (4) interaction terms between gender and norms at Level 1, (5) intervention condition and concentrated disadvantage at Level 2 and (6) Wave 1 class norms supporting male and female dating violence at Level 2. Within each model, parameter estimates and standard errors were examined to determine the significance of individual predictors. Based on recommendations by Snijders and Bosker (1994), the proportion of Level 1 and Level 2 variance explained by each set of variables was determined by calculating the difference in the amount of Level 1 (R21 ) and Level 2 (R22 ) residual variance as a proportion of the total error variance. This results in a statistic that is comparable to the multiple R2.

123

J Youth Adolescence (2015) 44:745–760

Results Descriptive Statistics The percentage of youth who reported perpetrating each type of dating violence one or more times in the past 3 months at Wave 1 and Wave 2 is shown in Table 1. At Waves 1 and 2, in the fall and spring of sixth grade, doing something to make a dating partner jealous was the most common type of psychological perpetration for girls (32 and 35 %, respectively) and boys (21 %). Across both Wave 1 and Wave 2 slapping was the most common physically violent act perpetrated by girls (19 and 24 %, respectively). At Wave 1 stomping out of the room or house during a disagreement was the most common type of physical perpetration by boys (17 %), whereas pushing or shoving was the most common type at Wave 2 (15 %). The means, standard deviations, and correlations among the uncentered Level 1 and 2 variables were calculated in Mplus (see Table 2). At the student level (Level 1), Wave 2 perpetration of physical and psychological dating violence were significantly correlated (r = .81). Wave 1 norms for male and female perpetration of dating violence were significantly correlated with each other (r = .40) and with Wave 2 perpetration of physical (rs = .16 and .13, respectively) and psychological dating violence (rs = .16 and .17, respectively). At the class level (Level 2) Wave 1 class norms supporting male and female dating violence were significantly correlated with each other (r = .80). Multilevel Analyses Relationships Among Norms Supporting Dating Violence, Gender, and Perpetration of Physical Dating Violence The first set of models tested hypotheses regarding the longitudinal relationships among individual- and classlevel norms supporting dating violence, gender, and perpetration of physical dating violence. The results of these models are shown in Table 3. Model 1, the baseline random intercept model, indicated significant variance in perpetration of physical dating violence at Level 1 (r2 = .21, p \ .001) and Level 2 (s = .01, p \ .001). The intraclass correlation (ICC) was .03. The addition of student demographic characteristics at Level 1 (i.e., family structure, gender, and race/ethnicity) accounted for 1 % of the modeled variance in Level 1 scores and 8 % of the variance in Level 2 scores (see Model 2). Boys reported lower levels of perpetration of physical dating violence than girls (b = -.04, p = .042). Youth who were White also reported lower levels of perpetration of physical dating violence than youth who were African American

J Youth Adolescence (2015) 44:745–760

(b = -.10, p \ .001). In Model 3, the addition of Wave 1 perpetration of physical and psychological dating violence accounted for an additional 17 % of the modeled variance in Level 1 scores and 24 % of the variance in Level 2 scores. Wave 1 perpetration of physical (b = .36, p \ .001) and psychological (b = .09, p = .039) dating violence were significant predictors of Wave 2 perpetration of physical dating violence. The addition of Wave 1 individual norms supporting dating violence accounted for an additional 1 % of the modeled variance in Level 1 and Level 2 scores (see Model 4). Wave 1 norms supporting male dating violence were a significant predictor of Wave 2 perpetration of physical dating violence, such that norms accepting male dating violence were associated with changes in perpetration of physical dating violence from Wave 1 to Wave 2 (b = .04, p = .023). Wave 1 norms supporting female dating violence did not significantly predict changes in perpetration of physical dating violence (b = .02, p = .135). In Model 5, the interactions between gender and individual norms supporting dating violence were added. The interaction terms accounted for an additional 1 % of the modeled variance in Level 1 scores and 7 % of the variance in Level 2 scores. The interactions between gender and norms supporting male dating violence (b = .06, p = .134) and gender and norms supporting female dating violence (b = -.05, p = .066) were not associated with changes in perpetration of physical dating violence from Wave 1 to Wave 2, indicating that relationships between individual norms supporting male and female dating violence and physical perpetration did not significantly vary for boys and girls. Model 6 incorporated intervention condition variables and concentrated disadvantage at Level 2, which did not account for any additional modeled variance in Level 1 or Level 2 scores. Neither concentrated disadvantage, nor any of the intervention conditions (i.e., universal: b = .04, p = .177; targeted: b = .02, p = .561; combined: b = .05, p = .197) were significant predictors of changes in physical dating volence perptration. Model 7 included Level 2 class norms supporting dating violence, to test their impact on changes in class averages of perpetration of physical dating violence from Wave 1 to Wave 2. Wave 1 class norms supporting dating violence accounted for 1 % of the modeled variance in Level 1 and 15 % of the variance in Level 2 scores. Both class norms supporting male (b = .53, p \ .001) and female dating violence (b = -.23, p = .011) were significant predictors of changes in average levels of perpetration of physical dating violence across school classes, such that greater class norms supporting male dating violence were associated with greater change in physical perpetration from Wave 1 to Wave 2 and greater class norms supporting female dating violence were associated with less change in physical perpetration.

753 Table 1 Percentage of sixth graders who reported perpetrating specific forms of psychological and physical dating violence one or more times in the past 3 months at Wave 1 and Wave 2, in the fall and spring of sixth grade Perpetration of specific types of dating violence

Wave 1 Total (%)

Wave 2 Girls (%)

Boys (%)

Total (%)

Girls (%)

Boys(%)

Psychological perpetration Damaged something that belonged to him/her

12.4

12.2

12.5

12.7

12.6

12.8

Said things to hurt his/her feelings on purpose

20.1

21.4

19.1

21.1

24.1

18.2

Would not let him/her do things with other people

12.8

12.0

13.5

14.6

13.8

15.2

Did something just to make him/her jealous

25.8

32.3

21.0

26.6

34.5

20.7

Physical perpetration Threatened to hit or throw something at him/her

14.2

15.7

13.1

14.8

16.7

13.4

Scratched him/ her

14.1

18.0

11.2

14.4

20.1

10.2

Slapped him/ her

13.6

18.8

9.7

16.8

23.5

11.7

8.9

6.4

10.1

10.0

7.8

11.6

Slammed him/ her or held you against a wall Kicked him/her

10.8

10.6

11.9

11.4

11.7

11.2

Pushed or shoved him/ her

15.8

17.0

14.8

16.8

18.8

15.3

Stomped out of the room or house during a disagreement

17.0

17.5

16.7

16.0

19.0

13.7

Threw something at him/her that could hurt

9.1

7.6

10.2

9.8

10.0

9.6

Punched or hit him/her with something that could hurt

9.0

7.8

9.9

9.2

9.4

9.0

Percentages are based on students who reported dating at each wave. At Wave 1 N = 2,022 for psychological perpetration and N = 2,022 for physical perpetration. At Wave 2 N = 2,013 for psychological perpetration and N = 2,012 for physical perpetration

123

754 Table 2 Means, standard deviations, and intercorrelations among study variables

J Youth Adolescence (2015) 44:745–760

Variable

1

2

3

4

5

6

7

M

SD

0.24

0.45

0.18

0.43

0.26

0.48

0.19

0.46

1.47

0.65

2.05

0.83

1.42

.11

1.93

.15

Level 1 variables 1. W1 psychological perpetration



2. W1 physical perpetration

.81***



3. W2 psychological perpetration

.40***

.37***

4. W2 physical perpetration

.37***

.42***

.81***

5. W1 norms supporting male DV

.20***

.19***

.16***

.16***

6. W1 norms supporting female DV

.16***

.15***

.17***

.13***

– – – .40***



Level 2 variables

N = 2,022, DV dating violence * p \ .05, ** p \ .01, *** p \ .001

7. W1 norms supporting male DV













8. W1 norms supporting female DV













Relationships Among Norms Supporting Dating Violence, Gender, and Perpetration of Psychological Dating Violence The next set of models tested hypotheses regarding the longitudinal relationships among individual- and classlevel norms supporting dating violence, gender, and perpetration of psychological dating violence. The results of these models are shown in Table 4. Model 1, the baseline random intercept model, showed significant variance in perpetration of psychological dating violence at Level 1 (r2 = .23, p \ .001) and Level 2 (s = .01, p \ .001). The ICC was .02. The addition of student demographic characteristics at Level 1 (i.e., family structure, gender, and race/ethnicity) accounted for 2 % of the modeled variance in Level 1 scores and 16 % of the variance in Level 2 scores (see Model 2). Boys reported lower levels of psychological perpetration as compared to girls (b = -.05, p = .020) and White youth reported lower levels of psychological perpetration relative to African American adolescents (b = -.15, p \ .001). In Model 3, the addition of Wave 1 perpetration of physical and psychological dating violence accounted for 16 % of the modeled variance in Level 1 scores and 24 % of the variance in Level 2 scores. Wave 1 perpetration of physical (b = .17, p = .002) and psychological (b = .28, p \ .001) dating violence were significant predictors of Wave 2 psychological perpetration. The addition of Wave 1 individual norms supporting dating violence accounted for 1 % of the modeled variance in Level 1 and Level 2 scores (see Model 4). Wave 1 norms supporting male dating violence (b = .04, p = .038) and norms supporting female dating violence (b = .04,

123

– .80***

p = .004), were significant predictors of Wave 2 perpetration of psychological dating violence, such that greater norms accepting male and female dating violence were associated with greater change in perpetration of psychological dating violence from Wave 1 to Wave 2. In Model 5, the interactions between gender and individual norms were added, but did not account for any additional modeled variance in Level 1 or Level 2 scores. The interactions between gender and norms supporting male dating violence (b = .03, p = .433) and gender and norms supporting female dating violence (b = -.04, p = .194) did not significantly predict changes in perpetration of psychological dating violence from Wave 1 to Wave 2, indicating that the extent to which individual norms supporting male and female dating violence predicted changes in psychological perpetration did not vary for boys and girls. Model 6 incorporated intervention condition variables and concentrated disadvantage at Level 2, which did not account for any additional modeled variance in Level 1 scores, but accounted for 8 % of additional variance in Level 2 scores. Neither concentrated disadvantage, nor any of the intervention conditions (i.e., universal: b = .05, p = .067; targeted: b = .01, p = .750; combined: b = .05, p = .074) were significant predictors of changes in perpetration of physical dating volence. Model 7 included Level 2 class norms supporting dating violence, to test their impact on changes in class averages of perpetration of psychological dating violence. Wave 1 class norms supporting dating violence accounted for 0 % of the modeled variance in Level 1 and 2 scores. Neither class norms supporting male dating violence (b = .22, p = .158) nor class norms supporting female dating violence (b = -.09,

J Youth Adolescence (2015) 44:745–760

755

Table 3 Unstandardized parameter estimates for multilevel regression of perpetration of physical dating violence on individual- and class-level norms supporting dating violence and gender Variable

Model 2

Model 3

Model 4

Model 5

Model 6

Model 7

Level 1 (student) Intercept Single parent family

.215*** (.019)

.215*** (.019)

.210*** (.019)

.187*** (.025)

-.020 (.023)

.214*** (.020)

.014 (.023)

.017 (.023)

.016 (.023)

.019 (.023)

.018 (.023)

.029 (.030)

-.011 (.023)

-.010 (.023)

-.008 (.024)

-.008 (.024)

-.010 (.024)

Other family structure White

-.101*** (.030)

-.055* (.021)

-.045* (.022)

-.041 (.022)

-.027 (.022)

-.027 (.024)

.000 (.034)

-.008 (.028)

-.010 (.027)

-.004 (.027)

-.015 (.027)

-.024 (.026)

.010 (.030) -.039* (.019)

.007 (.030) -.031* (.017)

.008 (.026) -.036* (.017)

.009 (.026) -.034* (.017)

.011 (.026) -.037* (.017)

.010 (.025) -.038* (.017)

Latino/a Other race/ethnicity Male gender

.217*** (.026)

W1 physical perpetration

.355*** (.053)

.346*** (.053)

.345*** (.053)

.347*** (.053)

.343*** (.053)

W1 psychological perpetration

.093* (.045)

.082 (.046)

.078 (.045)

.080 (.045)

.079 (.046)

W1 norms supporting male DV

.043* (.019)

.006 (.033)

.041* (.019)

.035 (.019)

W1 norms supporting female DV

.023 (.016)

.052** (.016)

.023 (.016)

.024 (.016)

W1 norms supporting male DV 9 male gender

.063 (.042)

W1 norms supporting female DV 9 male gender

-.054 (.029)

Level 2 (class) Concentrated disadvantage

.026 (.014)

Norms about male dating violence

.017 (.014) .533*** (.112)

Norms about female dating violence

-.233* (.091)

Variance estimates Within Level residual variance Between Level residual variance

.206*** (.020)

.172*** (.016)

.170*** (.016)

.170*** (.016)

.170*** (.016)

.171*** (.016)

.005* (.002)

.003 (.002)

.003 (.002)

.002 (.002)

.002 (.001)

.000 (.001)

R21

.009

.178

.188

.192

.192

.197

R22

.076

.315

.321

.394

.394

.539

N = 2,022. DV dating violence Standard errors are in parentheses. All models were fully-saturated. Parameters for Model 1 (intercept only model) were as follows: Intercept = .207***, Within Level Residual variance = .207***, Between Level Residual variance = .006** * p \ .05, ** p \ .01, *** p \ .001

p = .433) were significant predictors of changes in class averages of perpetration of psychological dating violence.

Discussion The present study examined longitudinal relationships between individual and class norms (i.e., norms for students in the same grade and cohort in the same school) supporting dating violence and perpetration of dating violence. Gender was also examined as a moderator of the above relationships. We sought to extend the existing body of research in several ways. First, we examined norms

supporting dating violence as a risk factor for perpetration of physical and psychological dating violence. Second, we assessed norms at two levels—the individual and class levels. Third, this study included a sample of early adolescents. Key findings indicated that greater individual norms supporting male dating violence significantly predicted greater changes in perpetration of both physical and psychological dating violence from Wave 1 to Wave 2. Greater individual norms supporting female dating violence significantly predicted greater change in perpetration of psychological dating violence only. Gender did not moderate these relationships. Class norms supporting male and female dating violence significantly predicted changes

123

756

J Youth Adolescence (2015) 44:745–760

Table 4 Unstandardized parameter estimates for multilevel regression of perpetration of psychological dating violence on individual- and classlevel norms supporting dating violence and gender Variable

Model 2

Model 3

Model 4

Model 5

Model 6

Model 7

Level 1 (student) Intercept

.302*** (.023)

.304*** (.021)

.302*** (.020)

.299*** (.021)

.272*** (.025)

Single parent family

.026 (.026)

.017 (.026)

.020 (.025)

.020 (.025)

.021 (.025)

.142 (.122) .022 (.026)

Other family structure

.028 (.032)

-.024 (.022)

-.023 (.022)

-.022 (.022)

-.021 (.022)

-.021 (.022)

White

-.147*** (.026)

-.094*** (.025)

-.077 (.025)

-.074 (.025)

-.057 (.027)

-.053 (.028)

Latino/a

-.043 (.030)

-.051* (.025)

-.052 (.025)

-.048 (.025)

-.053 (.024)

-.050 (.025)

Other race/ethnicity Male gender

-.019 (.036) -.047* (.020)

-.024 (.030) -.036 (.019)

.022 (.030) -.040* (.018)

-.021 (.030) -.040* (.019)

-.017 (.030) -.042* (.018)

-.015 (.030) -.041* (.019)

W1 physical perpetration

.169** (.054)

.158** (.054)

.158** (.054)

.160** (.054)

.158** (.054)

W1 psychological perpetration

.281*** (.058)

.266*** (.058)

.263*** (.059)

.263*** (.058)

.263*** (.058)

W1 norms supporting male DV

.043* (.021)

.026 (.035)

.040 (.021)

.037 (.021)

W1 norms supporting female DV

.043** (.015)

.063** (.020)

.043** (.015)

.044** (.015)

W1 norms supporting male DV 9 male gender

.032 (.040)

W1 norms supporting female DV 9 male gender

-.038 (.029)

Level 2 (class) Concentrated disadvantage

.020 (.013)

Norms about male dating violence

.222 (.157)

Norms about female dating violence

-.092 (.118)

Variance estimates Within Level residual variance Between Level residual variance

.224*** (.017)

.190*** (.015)

.188*** (.015)

.187*** (.015)

.187*** (.015)

.187*** (.015)

.003* (.001)

.001 (.001)

.001 (.001)

.001 (.001)

.000 (.001)

.000 (.001)

R21

.017

.173

.182

.186

.190

.190

R22

.156

.401

.406

.409

.484

.484

Standard errors are in parentheses. All models were fully-saturated. Parameters for Model 1 (intercept only model) were as follows: Intercept = .258***, Within Level Residual variance = .226***, Between Level Residual variance = .005** N = 2,022. DV dating violence * p \ .05, ** p \ .01, *** p \ .001

in perpetration of physical, but not psychological dating violence. Specifically, greater class norms supporting male dating violence predicted greater change in perpetration of physical dating violence, whereas greater class norms supporting female dating violence predicted less change in physical perpetration. In the following sections, the study findings are discussed along with limitations, practical implications, and directions for future research. One aim of the present study was to examine longitudinal relationships between individual norms supporting dating violence and perpetration of dating violence. The findings indicated that greater norms supporting male dating violence significantly predicted greater changes in

123

perpetration of both physical and psychological dating violence; whereas, greater norms supporting female dating violence predicted greater change in psychological perpetration. Previous studies found concurrent associations between individual norms and perpetration of physical dating violence (e.g., Malik et al. 1997; Simon et al. 2010). Thus, the current study findings extend this literature by establishing longitudinal links between norms accepting dating violence and perpetration of two different forms of dating violence, physical and psychological. A secondary aim of the study was to examine the moderating role of gender on relationships between individual norms and perpetration of dating violence. It was hypothesized that

J Youth Adolescence (2015) 44:745–760

relationships between norms supporting male dating violence and both forms of perpetration would be stronger for boys and relationships between norms supporting female dating violence and both forms of perpetration would be stronger for girls based on previous research showing differential relationships between norms supporting dating violence and perpetration of dating violence by gender (e.g., Foshee et al. 2001; O’Keefe 1997; Reeves and Orpinas 2012). However, these hypotheses were not supported as we did not find significant gender differences in these relationships. Few studies have investigated relationships between norms supporting dating violence and perpetration of dating violence and findings from these studies have been mixed. For instance, some research shows a stronger concurrent association between norms supporting male dating violence and perpetration of physical dating violence for boys and a stronger association between norms accepting female dating violence and perpetration of physical dating violence for girls (e.g., O’Keefe 1997). Other research shows that relationships between norms accepting male and female dating violence and perpetration of physical dating violence are stronger for boys (e.g., Foshee et al. 2001; Reeves and Orpinas 2012). Thus, additional research is needed to better understand the role of gender in these relationships, as there are important gender differences in the dynamics of adolescent dating violence. First, violence within a dating relationship is often reciprocal, in that boys and girls are equally likely to be both victims and perpetrators (Foshee et al. 2009). Studies also show that girls perpetrate higher levels of physical and psychological dating violence (e.g., Wolfe et al. 2003), but that boys are more likely to perpetrate more severe violence, increasing the potential for injury (Foshee et al. 2009). Previous research also highlights important differences in adolescents’ attitudes towards dating violence perpetrated by boys versus girls. For example, previous research with high school students indicated that youth are generally more accepting of violence perpetrated by girls as compared to boys because of the greater potential for physical injury and the legal implications of male dating violence (Reeves and Orpinas 2012). Overall, the findings suggest that individual norms supporting male and female dating violence are a significant risk factor for perpetration within a dating relationship for early adolescents, and may differentially predict perpetration of physical and psychological forms of dating violence. As such, the findings highlight the importance of addressing specific attitudes about male and female dating violence in school-based violence prevention curricula. We also tested longitudinal relationships between class norms supporting dating violence (i.e., norms for students in the same grade and cohort attending the same school) and

757

perpetration of dating violence. To our knowledge, previous studies have not examined the influence of norms at the class level on perpetration of dating violence; thus, the current study extends the current body of research by testing the extent to which this class-level risk factor impacted changes in perpetration of dating violence. Class norms supporting male and female dating violence significantly predicted changes in perpetration of physical dating violence. Consistent with our expectations, greater class norms supporting male dating violence predicted greater change in perpetration of physical dating violence. However, greater class norms supporting female dating violence predicted less change in perpetration of physical dating violence. The correlation between class norms supporting female and male dating violence suggest that classes with high norms supporting male dating violence also have high norms supporting female dating violence. Therefore, the findings regarding the impact of norms supporting female dating violence on physical perpetration may be more reflective of this high correlation and less indicative of a negative relationship. Nevertheless, these findings also suggest that class norms supporting male dating violence are a stronger predictor of change in perpetration of physical dating violence. It may be that greater class norms supporting male dating violence are related to other important factors associated with perpetration of dating violence. For example, in a sample of 1,699 high school boys, witnessing peers perpetrate violence directed toward a female dating partner, lower intentions to intervene in dating violence incidents, and negative bystander behavior (e.g., not seeking assistance from an adult, laughing at or going along with abusive behavior perpetrated by peers) were positively associated with perpetration of dating violence (McCauley et al. 2013). In addition, contrary to hypotheses, Wave 1 class norms did not predict changes in perpetration of psychological dating violence over time. There may be several possible explanations for the lack of findings for perpetration of psychological dating violence. First, our findings suggest that psychological dating violence may be more strongly influenced by individual-level risk factors (e.g., individual norms supporting dating violence) as compared to broader influences at the class level. Patterns of psychological violence may be more likely to form within the context of dyadic dating relationships based on the characteristics of the dating partners and as such be less influenced by class-level norms supporting dating violence (Bartholomew et al. 2015). This is not to say that relationship dynamics do not also influence physical perpetration. In contrast, peer-based influences on perpetration of physical dating violence are well-established (e.g., Arriaga and Foshee 2004; Foshee et al. 2001). Second, the lack of findings regarding relationships between class norms and perpetration of psychological dating violence may have been due to the characteristics of the sample and the transition from elementary to middle school. More specifically, youth in the current study

123

758

were in the sixth grade and the majority were transitioning to a new school with a different and larger group of peers. Due to this transition and the differential and complex expectations for how boys and girls should behave in dating relationships, the adolescents in the current sample may not have been acutely aware of the norms supporting dating violence among students in their grade. Lastly, there was a degree of variability in norms supporting male and female dating violence among students within the same class (i.e., same school, grade, and cohort), which may explain the limited effect of class norms on perpetration of psychological dating violence. Norms about dating violence are likely more consistent among friends or close associates; therefore, future research should investigate the impact of these norms on perpetration of dating violence. Nevertheless, the current study offers a unique contribution to the literature as it examined the influence of class-level norms accepting dating violence on perpetration of physical and psychological dating violence. Limitations Although the present study had several strengths, it is important to note its limitations. Firstly, the current sample only included youth who reported being in a dating relationship in the past 3 months at Wave 1 and Wave 2, in the fall and spring of sixth grade, in order to control for Wave 1 levels of perpetration of dating violence. Thus, the findings cannot be generalized to youth who were in a relationship in the fall of sixth grade, but discontinued that relationship and did not begin another relationship or to youth who were not in a relationship at Wave 1, but began dating by Wave 2. One other limitation was the limited information on the relationship context, specifically, the quality and duration of the relationship, and if the relationship was the same or different at Wave 1 and Wave 2, and if the dating violence was reciprocal. Secondly, we relied solely on adolescents’ reports of perpetration of dating violence and norms for dating violence. Although youth are likely the best reporters of their dating behavior, these reports may be prone to response bias, especially given the sensitive nature of the topic. Thirdly, the classlevel measure of norms supporting dating violence did not include norms among seventh and eighth grade students at the school. As older students in seventh and eighth grades likely have a strong influence on younger students, this would be an important direction for future research. Implications and Future Research The present study’s findings have implications for violence prevention efforts. Prevalence rates of specific forms of perpetration indicate that large numbers of youth reported perpetrating multiple types of dating violence, including

123

J Youth Adolescence (2015) 44:745–760

scratching, insults, and acts to evoke jealousy. Although these rates mostly pertained to less severe forms of perpetration, dating violence does occur in early adolescence and should be addressed early in middle school. This is critical given research showing that involvement in dating violence during adolescence increases the likelihood of involvement in intimate partner violence in adulthood (e.g., Gomez 2011). Given the low frequencies of more severe forms of dating violence in the current study, another direction for future research would be to examine risk factors for involvement in dating problems (Sullivan et al. 2010), as opposed to violent behaviors, in early adolescence, as these may be more prevalent during this developmental period and may lead to more severe forms of dating violence. There may also be specific problem situations that are more relevant to various developmental stages of dating relationships. For example, problem situations related to approaching dating partners and peer influences on dating relationships may be particularly relevant in early adolescence. The current study’s findings indicated that individual norms supporting male and female dating violence and uniquely predicted changes in perpetration of dating violence. Thus, prevention and intervention efforts should address these norms about dating violence in order to prevent perpetration of dating violence. Given findings that class norms more supportive of male dating violence predicted greater change in physical perpetration, prevention and intervention programs should also include a schoollevel component to promote norms against dating violence. Additionally, in light of findings that individual norms supporting male and female dating violence differentially predicted perpetration of physical and psychological dating violence, future research should continue to examine unique risk factors for these two common forms of dating violence. In addition, future research should assess predictors of other forms of violence that may be common in early adolescent dating relationships, such as cyber and relational aggression. For instance, in a recent study of cyber dating aggression Zweig et al. (2013) found that among a sample of 3,745 youth, 26 % of adolescents reported experiencing cyber abuse in a dating relationship. Future research should also continue to investigate the impact of norms on perpetration of multiple forms of dating violence, including norms among friends as well as the broader peer context. Additionally, future research should examine students’ perceptions of the norms supporting dating violence among their same-grade peers and how they impact levels of perpetration. Future research should also explore youth’s attitudes towards specific types of dating violence, including physical and psychological violence, and how these attitudes relate to specific types of perpetration.

J Youth Adolescence (2015) 44:745–760

Conclusion Previous research indicates that over a third of adolescents are involved in some form of perpetration of dating violence, making it critical to identify risk factors for perpetration of dating violence. The current study sought to do this by examining relationships between individual and class norms supporting dating violence, gender, and perpetration of physical and psychological dating violence among an ethnically diverse sample of early adolescents. The findings indicated that greater individual norms supporting male and female dating violence predicted greater change in psychological perpetration; whereas greater individual norms supporting male, but not female dating violence predicted greater change in physical perpetration. Greater class norms supporting male dating violence also predicted greater change in physical perpetration and greater class norms supporting female dating violence predicted less change in physical perpetration. Therefore, personal and class norms supporting male and female dating violence differentially predicted perpetration of physical and psychological dating violence, such that individual and class norms are associated with changes in perpetration of physical dating violence, but that changes in psychological perpetration may be more impacted by individual-as opposed to class-level norms. Overall, our study expands on the current body of literature by establishing longitudinal relationships between individual- and class-level norms supporting dating violence and perpetration of two forms of dating violence (i.e., physical and psychological) for early adolescents. Acknowledgments The authors are grateful to the Multi-Site Violence Prevention Project for permission to use the data for this study. Investigators from each site are as follows (changes in affiliations in parentheses): Centers for Disease Control and Prevention, Atlanta GA: Thomas R. Simon, Robin M. Ikeda, Emilie Smith (Penn State University); Le’Roy E. Reese (Morehouse University); Duke University, Durham NC: David L. Rabiner, Shari Miller (Research Triangle Institute), Donna-Marie Winn (University of North Carolina— Chapel Hill), Kenneth A. Dodge, Steven R. Asher; University of Georgia, Athens GA: Arthur M. Horne, Pamela Orpinas, Roy Martin, William H. Quinn (Clemson University); University of Illinois at Chicago, Chicago IL: Patrick H. Tolan (University of Virginia), Deborah Gorman-Smith (University of Chicago), David B. Henry, Franklin N. Gay (University of Chicago), Michael Schoeny (University of Chicago), Virginia Commonwealth University, Richmond VA: Albert D. Farrell, Aleta L. Meyer (Administration for Children and Families, Washington, DC); Terri N. Sullivan, Kevin W. Allison. This study was funded, in part, by the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention, CDC Cooperative Agreements U81/CCU317633 and 1U01CE0019\56. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Authors’ contributions KT conceived of the study, participated in its design and the interpretation of the data, performed statistical

759 analysis, and drafted the manuscript. TS assisted in the conceptualization of the study, participated in the design and the interpretation of the data and helped to draft the manuscript. AF participated in the design of the study and the interpretation of the data. All authors read and approved the final manuscript.

References Arriaga, X. B., & Foshee, V. A. (2004). Adolescent dating violence: Do adolescents follow in their friends’ or their parents’ footsteps? Journal of Interpersonal Violence, 19, 162–184. doi:10.1177/0886260503260247. Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall. Bartholomew, K., Cobb, R. J., & Dutton, D. G. (2015). Established and emerging perspectives on violence in intimate relationships. In M. Mikulincer, P. R. Shaver, J. A. Simpson, & J. F. Dovidio (Eds.), APA handbooks in psychology. APA handbook of personality and social psychology, Vol. 3. Interpersonal relations (pp. 605–630). doi:10.1037/14344-022 Bernburg, J. G., & Thorlindsson, T. (2005). Violent values, conduct norms, and youth aggression: A multilevel study in Iceland. The Sociological Quarterly, 46, 457–478. doi:10.1111/j.1533-8525. 2005.00022.x. Brown, B. B. (1999). ‘‘You’re going out with who?’’: Peer group influences on adolescent romantic relationships. In W. Furman, B. Brown, & C. Feiring (Eds.), The development of romantic relationships in adolescence (pp. 291–329). New York, NY: Cambridge University Press. Collins, W. A. (2003). More than myth: The developmental significance of romantic relationships during adolescence. Journal of Research on Adolescence, 13, 1–24. doi:10.1111/15327795.1301001. Connolly, J., & McIsaac, C. (2009). Adolescents’ explanations for romantic dissolutions: A developmental perspective. Journal of Adolescence, 32, 1209–1223. doi:10.1016/j.adolescence.2009.01.006. Foshee, V. A., Bauman, K. E., Arriaga, X. B., Helms, R. W., Koch, G. G., & Linder, G. F. (1998). An evaluation of Safe Dates, an adolescent dating violence prevention program. American Journal of Public Health, 88, 45–50. doi:10.2105/AJPH.88.1.45. Foshee, V. A., Benefield, T. S., Ennett, S. T., Bauman, K. E., & Suchindran, C. (2004). Longitudinal predictors of serious physical and sexual dating violence victimization during adolescence. Preventive Medicine, 39, 1007–1016. doi:10. 1016/j.ypmed.2004.04.014. Foshee, V. A., Benefield, T., Suchindran, C., Ennett, S. T., Bauman, K. E., Karriker-Jaffe, K. J., et al. (2009). The development of four types of adolescent dating abuse and selected demographic correlates. Journal of Research on Adolescence, 19, 380–400. doi:10.1111/j.1532-7795.2009.00593.x. Foshee, V. A., Linder, G. F., Bauman, K. E., Langwick, S. A., Arriaga, X. B., Heath, J. L., et al. (1996). The Safe Dates Project: theoretical basis, evaluation design, and selected baseline findings. American Journal of Preventive Medicine, 12, 39–47. Foshee, V. A., Linder, F., MacDougall, J. E., & Bangdiwala, S. (2001). Gender differences in the longitudinal predictors of adolescent dating violence. Preventive Medicine, 32, 128–141. doi:10.1006/pmed.2000.0793. Giordano, P. C., Manning, W. D., & Longmore, M. A. (2006). Adolescent romantic relationships: An emerging portrait of their nature and developmental significance. In A. C. Crouter & A. Booth (Eds.), Romance and sex in adolescence and emerging adulthood: Risks and opportunities (pp. 127–150). Mahwah, NJ: Lawrence Erlbaum.

123

760 Gomez, A. M. (2011). Testing the cycle of violence hypothesis: Child abuse and adolescent dating violence as predictors of intimate partner violence in young adulthood. Youth & Society, 43, 171–192. doi:10.1177/0044118X09358313. Henry, D., & Chan, W. Y. (2010). Cross-sectional and longitudinal effects of sixth grade setting-level norms for nonviolent problem solving on aggression and associated attitudes. Journal of Community Psychology, 38, 1007–1022. doi:10.1002/jcop. 20411. Henry, D. B., Farrell, A. D., & Multisite Violence Prevention Project. (2004). The study designed by a committee: Design of the Multisite Violence Prevention Project. American Journal of Preventive Medicine, 26, 12–19. doi:10.1016/j.amepre.2003.09. 027. Henry, D. B., Farrell, A. D., Schoeny, M. E., Tolan, P. H., & Dymnicki, A. B. (2011). Influence of school-level variables on aggression and associated attitudes of middle school students. Journal of School Psychology, 49, 481–503. doi:10.1016/j.jsp. 2011.04.007. Henry, D., Guerra, N., Huesmann, R., Tolan, P., VanAcker, R., & Eron, L. (2000). Normative influences on aggression in urban elementary school classrooms. American Journal of Community Psychology, 28, 59–81. doi:10.1023/A:1005142429725. Hox, J. (2010). Multilevel analysis: Techniques and applications. New York, NY: Routledge. Malik, S., Sorenson, S. B., & Aneshensel, C. S. (1997). Community and dating violence among adolescents: Perpetration and victimization. Journal of Adolescent Health, 21, 291–302. doi:10.1016/S1054-139X(97)00143-2. McCauley, H. L., Tancredi, D. J., Silverman, J. G., Decker, M. R., Austin, S. B., McCormick, M. C., et al. (2013). Gender-equitable attitudes, bystander behavior, and recent abuse perpetration against heterosexual dating partners of male high school athletes. American Journal of Public Health, 103, 1882–1887. doi:10. 2105/AJPH.2013.301443. Multisite Violence Prevention Project (MVPP). (2004). The Multisite Violence Prevention Project: Background and overview. American Journal of Preventive Medicine, 26, 3–11. Muthe´n, L. K., & Muthe´n, B. O. (2012). Mplus user’s guide. Los Angeles: Muthe´n & Muthe´n. Noonan, R. K., & Charles, D. (2009). Developing teen dating violence prevention strategies: Formative research with middle school youth. Violence against Women, 15, 1087–1105. doi:10. 1177/1077801209340761. O’Keefe, M. (1997). Predictors of dating violence among high school students. Journal of Interpersonal Violence, 12, 546–568. doi:10. 1177/088626097012004005. O’Leary, K. D., Smith Slep, A. M., Avery-Leaf, S., & Cassardi, M. (2008). Gender differences in dating aggression among multiethnic high school students. Journal of Adolescent Health, 42, 473–479. doi:10.1016/j.jadohealth.2007.09.012. Olweus, D., & Limber, S. P. (2007). Olweus Bullying Prevention Program: Teacher guide. Center City, MN: Hazelden. Orpinas, P., Horne, A. M., & Multisite Violence Prevention Program. (2004). A teacher-focused approached to prevent and reduce students’ aggressive behavior: The GREAT Teacher Program. American Journal of Preventive Medicine, 26, 29–38. doi:10. 1016/j.amepre.2003.09.016. Reeves, P. M., & Orpinas, P. (2012). Dating norms and dating violence among ninth graders in Northeast Georgia: Reports from student surveys and focus groups. Journal of Interpersonal Violence, 27, 1677–1698. doi:10.1177/0886260511430386. Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277, 918–924. doi:10.1126/science.277.5328. 918.

123

J Youth Adolescence (2015) 44:745–760 Simon, T. R., Miller, S., Gorman-Smith, D., Orpinas, P., & Sullivan, T. (2010). Physical dating violence norms and behavior among sixth-grade students from four U.S. sites. The Journal of Early Adolescence, 30, 395–409. doi:10.1177/0272431609333301. Smith, E. P., Gorman-Smith, D., Quinn, W., Rabiner, D., Tolan, P., Winn, D.-M., et al. (2004). Community-based multiple family groups to prevent and reduce violent and aggressive behavior: The GREAT families program. American Journal of Preventive Medicine, 26, 39–47. doi:10.1016/j.amepre.2003.09.018. Snijders, T. A. B., & Bosker, R. J. (1994). Modeled variance in twolevel models. Sociological Methods & Research, 22, 342–363. doi:10.1177/0049124194022003004. Spriggs, A. L., Halpern, C. T., & Martin, S. L. (2009). Continuity of adolescent and early adult partner violence victimization: Association with witnessing violent crime in adolescence. Journal of Epidemiology and Community Health, 63, 741–748. doi:10.1136/jech.2008.078592. Sullivan, T. N., Erwin, E. H., Helms, S. W., Masho, S. W., & Farrell, A. D. (2010). Problematic situations associated with dating experiences and relationships among urban African American adolescents: A qualitative study. The Journal of Primary Prevention, 31, 365–378. doi:10.1007/s10935-010-0225-5. U.S. Census Bureau. (2000). Census tracts and block numbering areas. Retrieved from http://www.census.gov/geo/www/cen_tract.html. Wekerle, C., & Wolfe, D. A. (1999). Dating violence in midadolescence: theory, significance, and emerging prevention initiatives. Clinical Psychology Review, 19, 435–456. doi:10. 1016/S0272-7358(98)00091-9. Wolfe, D. A., Wekerle, C., Scott, K., Straatman, A., Grasley, C., & ReitzelJaffe, D. (2003). Dating violence prevention with at-risk youth: A controlled outcome evaluation. Journal of Consulting and Clinical Psychology, 71, 279–291. doi:10.1037/0022-006X.71.2.279. Zweig, J. M., Dank, M., Yahner, J., & Lachman, P. (2013). The rate of cyber dating abuse among teens and how it relates to other forms of teen dating violence. Journal of Youth and Adolescence, 47, 1–15. doi:10.1007/s10964-013-9922-8.

Katherine A. Taylor is a postdoctoral fellow at Virginia Commonwealth University. She received her Ph.D. in Developmental Psychology from Virginia Commonwealth University. Her research interests include risk factors and consequences of youth violence exposure, including peer victimization and dating violence. She is also interested in school-based prevention programs that target youth violence and promote positive development. Dr. Terri N. Sullivan is an Associate Professor at Virginia Commonwealth University. She received her Ph.D. in Clinical Psychology from Virginia Commonwealth University. Her work focuses on the impact of violence exposure (e.g., exposure to community violence and peer victimization) on children’s psychosocial and emotional development. She is especially interested in the development and evaluation of the effectiveness of school-based youth violence prevention programs for children and adolescents, with a focus on youth with disabilities. Dr. Albert D. Farrell is a Professor and Director of the Clark-Hill Institute for Positive Youth Development at Virginia Commonwealth University. His work focuses on the identification of risk and protective factors related to adolescent problem behaviors, especially youth violence. A particular emphasis is on identifying factors that promote the positive development of youth in high risk environments (e.g., those exposed to community violence). He also has a strong interest in research methodology, particularly as it applies to longitudinal research and prevention science.

Copyright of Journal of Youth & Adolescence is the property of Springer Science & Business Media B.V. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.

Longitudinal relationships between individual and class norms supporting dating violence and perpetration of dating violence.

Dating violence is commonly perpetrated in adolescence, making it imperative to understand risk factors in order to inform prevention efforts. Althoug...
298KB Sizes 0 Downloads 4 Views