Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2011 Nov 1.
Published in final edited form as: Soc Sci Res. 2010 Nov;39(6):863–874. doi: 10.1016/j.ssresearch.2010.03.009

The Characteristics of Romantic Relationships Associated with Teen Dating Violence

Peggy C Giordano 1, Danielle A Soto 1, Wendy D Manning 1, Monica A Longmore 1
PMCID: PMC2964890  NIHMSID: NIHMS199586  PMID: 21037934

Abstract

Studies of teen dating violence have focused heavily on family and peer influences, but little research has been conducted on the relationship contexts within which violence occurs. The present study explores specific features of adolescent romantic relationships associated with the perpetration of physical violence. Relying on personal interviews with a sample of 956 adolescents, results indicate that respondents who self-report violence perpetration are significantly more likely than their non-violent counterparts to report higher levels of other problematic relationship dynamics and behaviors such as jealousy, verbal conflict, and cheating. However, we find no significant differences in levels of love, intimate self-disclosure, or perceived partner caring, and violent relationships are, on average, characterized by longer duration, more frequent contact, sexual intimacy and higher scores on the provision and receipt of instrumental support. Finally, violence is associated with the perception of a relatively less favorable power balance, particularly among male respondents. These findings complicate traditional views of the dynamics within violent relationships, add to our understanding of risk factors, and may also shed light on why some adolescents remain in physically abusive relationships.

Keywords: adolescent romantic relationships, gender, teen dating violence

Prior research on teen dating violence (TDV) has documented the scope and seriousness of this public health problem (O'Leary et al., 2008; Zurbriggen, 2009). There is general agreement that violence within the context of intimate relationships is emotionally and physically costly (Silverman et al., 2001), and that such formative experiences during adolescence may be linked to later violence within adult relationships (Halpern et al., 2001; Henton et al., 1983; O'Leary and Slep, 2003). Research has also focused on the demographic patterning of violence within teen relationships (particularly the issue of gender disparities and symmetries (e.g., Halpern et al., 2001; Stets and Pirog-Good, 1987; Whitaker et al., 2007), and precursors (such as witnessing or experiencing violence within the family of origin (see DeMaris, 1990; Foshee et al. 2008; O'Keeffe et al., 1986; Wolfe et al., 2001)). Yet even though TDV necessarily occurs within a relationship context, research on the character and dynamics of violent relationships is limited, with most research emphasizing directly related phenomena such as controlling behaviors and emotional abuse (see e.g., O'Keefe, 1997; O'Leary and Slep, 2003).

Drawing on a symbolic interactionist perspective and data from the Toledo Adolescent Relationships Study (TARS), we investigate links between a range of qualities/dynamics of adolescent romantic relationships and the odds of self-reported intimate violence. We focus on reports of violence perpetration, but supplemental analyses examine links to victimization, and to the experience of ‘any violence’ (perpetration or victimization) within adolescent romantic relationships. The multidimensional portrait we develop focuses on both positive (e.g., love) and negative (e.g., jealousy) relationship features, as well as patterns of interaction and influence (e.g., perceptions of the power balance) that characterize these early teen dating relationships.

Background

Early research on teen dating violence documented that intimate violence was not limited to adult marital relationships, and highlighted the seriousness of this problem (Henton et al., 1983). Yet many studies of dating violence have relied on convenience samples of college students, who are generally outside the age range of the adolescent period, and whose sociodemographic characteristics and levels of academic achievement do not reflect a true cross-section of teens. Indeed, prior research has shown that TDV may be significantly associated with economic disadvantage (Cleveland, Herrera, and Stuewig, 2003; Foshee et al., 2008; Herrenkohl et al., 2006; Makepeace, 1986), and low achievement in school (Halpern et al., 2001). Other research has focused on reports from high school students, but the restriction to in-school administrations is somewhat limiting, as this tends to increase refusals due to lack of parental consent (O'Keeffe et al., 1986, but see O'Leary et al., 2008), and by definition does not capture youths who are frequently absent, suspended from or no longer attending school (Cornelius and Resseguie, 2007). This is especially problematic because prior work suggests an association between TDV and low academic achievement (i.e., the youths most likely not to be attending school) (Halpern et al., 2001). Another study (O'Keeffe et al., 1986) found no association between teen dating violence and demographic characteristics of respondents, but findings were based on students from a single high school in Sacramento.

Analyses based on the National Longitudinal Study of Adolescent Health (Add Health) data set, a national probability sample, although school-based, are a useful addition to the research literature, and have aided in providing prevalence estimates across the population. Halpern et al. (2001) documented that approximately 12% of respondents had experienced violence within their romantic relationships. However, a limitation of the Add Health study is that, during the first two waves of interviews, questions were only asked regarding victimization, and only relatively minor forms of abuse (e.g., push/shove or thrown something) were included in the survey instrument.

In addition to documenting the extent of the problem and demographic correlates, research has provided useful information about precursors and correlates of teen dating violence. Conceptually, most research on TDV is based either on social learning or feminist frameworks. A number of studies have shown that parental violence is associated with increased odds that young people report violence within their own dating relationships. In addition, the experience of child abuse has been linked to a greater likelihood of experiencing violence in romantic relationships (DeMaris, 1990; Foshee et al., 2008; O'Keeffe et al., 1986). These findings are generally congruent with a social learning framework; that is, the idea that these violent repertoires are observed early within the family context and are later enacted within the context of romantic relationships (McCloskey and Lichter, 2003). Yet a limitation of the social learning perspective as traditionally theorized is that it places all of the emphasis on what is transported into dating relationships, rather than including attention to dynamics within the relationships themselves.

Feminist perspectives offer a more complex view in stressing multiple ways in which cultural and peer group socialization practices influence male-female relationships and in turn create the potential for gendered patterns of intimate violence. Some research within this tradition focuses on the influence of micro-level interaction patterns within the peer group. For example, Eder et al. (1995), in a study of middle school students, showed that boys and girls are socialized quite differently within their same-gender peer groups. Boys receive positive reinforcement for a competitive, one-up style of discourse and behavior, and for communications that objectify and denigrate young women. Consistent with this, those who display caring or other softer emotions are negatively sanctioned by male peers. Eder et al. (1995) argued that this style of socialization not only influences the nature of developing male-female relationships, but increases girls' risk for violent victimization (see also Anderson, 2005; Wight, 1994).

Maccoby (1990) also stressed that boys and girls are socialized in two different (peer) worlds, arguing that when the two sexes begin to interact, the transition is more easily accomplished for boys, who tend to transport their dominant interaction style into the new relationship. Although research and theorizing in this tradition thus provides a general framework for understanding gendered inequalities of power and female victimization, this perspective does not fully illuminate specific sources of variation in male behavior within a given sample, or factors that influence girls' participation in violent acts.

The Relationship Context of Teen Dating Violence

Feminist perspectives have focused theoretical attention on the dynamics of power and control in adolescent and adult relationships as influences on intimate partner violence. However, few studies have explored these power dynamics directly, and many other relationship qualities/dynamics distinct from power have not been systematically examined. The absence of research on relationship characteristics is an ironic omission since dating violence, unlike other problem youth outcomes (e.g., dropping out of school, smoking marijuana), is inherently relational. The symbolic interactionist perspective (Mead, 1934) highlights that behaviors and their meanings are necessarily ‘situated,’ that is, emerge within and derive their meaning(s) within specific social contexts. This basic notion highlights that behavior within the romantic realm may be influenced by but is never fully fashioned on the basis of other, more distal social processes (e.g., peer group norms; family of origin experiences). While building on insights from social learning and feminist theories, the symbolic interactionist framework is especially useful because this perspective focuses central attention on the relationship context itself.

Prior research that has explicitly explored aspects of dating context is limited in scope and attention to subjectively experienced aspects of these relationships. For example, the nature of the relationship has been defined as number of dates, which has been linked (positively) to a greater likelihood of violence (Henton et al., 1983; O'Keefe and Treister, 1998). In general, duration of the relationship is associated with higher risk (Burke, Stets, and Pirog-Good, 1988; Gaertner and Foshee, 1999; Magdol et al., 1998; Ray and Gold, 1996), although Luthra and Gidycz (2006) only found an association for male respondents. Stets and Pirog-Good (1987) included a more complex measure of “involvement,” but this index is a simple function of the length of time dating and number of dates. While research has thus shown that length of time, whether defined in terms of duration or number of dates, is associated with heightened risk, this does not provide a comprehensive portrait of the relationship characteristics, qualities, and dynamics associated with dating violence.

Some researchers have attempted to measure intensity of the relationship, but these studies are also limited in scope. Prior research on links between sexual behavior and teen dating violence has often shown an association, but a majority of studies have focused on risky patterns such as early age at first intercourse, number of partners, or inconsistent use of condoms, rather than sexual behavior within a particular relationship (Howard and Wang, 2003; Silverman et al., 2001). However, Kaestle and Halpern, (2005) relying on data from Add Health, found that sexual intercourse significantly increased odds of experiencing partner violence within a focal relationship. Roberts et al. (2006), also relying on the Add Health, report that involvement in a ‘special romantic relationship’ was associated with heightened risk of abuse. However, about 90% of the dating sample answered affirmatively to this item, suggesting that this is not an ideal measure of variations in the character of these dating relationships. Another study (Stets and Straus, 1989) defined commitment in terms of variations in living arrangements, finding an effect of cohabitation on risk of relationship violence (see also Brown and Bulanda, 2008; Sigelman, Berry, and Wiles, 1984; but see Kenney and McLanahan, 2006). Nevertheless cohabitation is not common during the adolescent period, suggesting the need to explore other dimensions.

Where more specific qualities of the relationship have been assessed, there has been an almost exclusive focus on negative dynamics. Stets and Pirog-Good (1987) indirectly assessed the experience of jealousy by examining whether the respondent reported involvement in a serious relationship, but the partner was still dating others (i.e., non-exclusivity was measured, rather than the subjective experience of jealousy). The authors found that non-exclusivity predicted female, but not male, use or experience of violence. Another theme related to power is the notion of coercive control. Both the adult and teen intimate violence literatures have examined controlling behaviors that often accompany physical violence and the relationships between these various forms of abuse (O'Leary and Slep, 2003). The focus on negative dynamics (jealousy, controlling behaviors) is intuitive, fits with prior research and programmatic emphases, and will be a focus of our analyses. However, while it is critical to study these negative dynamics and their relationship to teen dating violence, it is important to recognize the potential for positive features of relationships to coexist with the more troubling dynamics that have been emphasized in prior research and programming. This emphasis on negative dynamics may be related to what Cohen (1955) early on described as the “evil causes evil” fallacy—that is the tendency of researchers to concentrate only on negative causes as influences on negative outcomes.

A limited number of studies have given attention to positive dimensions of relationships, suggesting that a more multifaceted approach is warranted. For example, Arias et al. (1987) found that ‘loving’ the partner was uncorrelated with using/sustaining violence. However the authors noted a gender difference -- lower levels of liking and less positive feelings were associated with higher female but not male reports of using violence. In the same vein, Vivian and O'Leary (1987) used audio tapes to analyze affect within communications of aggressive and non-aggressive couples. They found that the aggressive and non-aggressive couples did not differ in their reports of positive and neutral content and affect, but observed more negative content and affect within the aggressive couples' ten minute interactions. More recently, Gallaty and Zimmer-Gembeck (2008) explored psychological maltreatment using the diary method over a seven day period, and analyzed respondents' perceptions of daily uplifts, hassles, and affect. Those who had experienced maltreatment reported a similar number of overall uplifts connected to their romantic partners as those who had not experienced maltreatment, and also reported similar levels of happiness associated with those uplifts. However, those within the maltreated group did report more hassles related to their romantic relationships, and more stress and depression associated with these hassles. These studies generally support the idea that a more complex set of relationship experiences may be related to dating violence (see also Larson and Richards, 1994).

The current analysis will assess a broader spectrum of subjective relationship dynamics, with the objective of understanding the total package of positive, negative, and conceptually more neutral dynamics that link to violent expression within dating relationships. Developing this comprehensive portrait should be useful, as young people may not identify with or believe that their own relationship experiences, feelings, and situations ‘fit’ with the predominantly negative themes stressed within some studies and program emphases.

Power dynamics within teen relationships also require more systematic research scrutiny, since, as suggested above, issues of power and control have been central to discussions of violence within male-female relationships, but have not often been studied directly. The logic of prior theorizing is that gendered inequalities of power tend to be reproduced at the couple level, although these dynamics have been studied more often among married than younger dating couples (Komter, 1989). Nevertheless, numerous prevention efforts focusing on teens have emphasized the extent to which violence is linked to issues of power and control, e.g., Cornelius and Resseguie, 2007; Dutton and Goodman, 2005), suggesting the need to examine teens' own perspectives on power relations, and how this relates to the experience of dating violence.

Gender and Relationship Dynamics

In analyzing early romantic relationships, like Maccoby (1990) and Eder et al. (1995), Giordano et al. (2006) used early peer interactions as the point of departure theoretically, but highlighted some different dynamics that may follow from these distinct peer interaction styles. Because girls have more experience with intimacy by virtue of their early friendship experiences (e.g., opportunities for self-disclosure, the experience of conflict and its repair), boys may be less prepared and adept at intimate ways of relating within the romantic context. Consistent with this hypothesis, Giordano et al. (2006) found that boys reported greater feelings of communication awkwardness, and less confidence navigating various aspects of romantic relationships. Further, because girls typically have female friends who provide support and numerous chances for intimacy (Call and Mortimer, 2003), the researchers argued that in some respects may be considered more ‘dependent’ on girlfriends, who represent a relatively unique relationship forum within which to explore newly developing feelings and emotions, as well as the experience of sexual intimacy. In line with these considerations, no significant differences by gender were observed in scores on a passionate love scale, and in-depth narratives underscored that boys frequently accorded much meaning and significance to their romantic relationships (Giordano et al., 2006). Boys also scored lower on perceived power in their relationships, while scoring higher on partner influence attempts and actual influence. A number of recent studies, often relying on ethnographic methods, have similarly documented boys' feelings of vulnerability and interest in intimacy (Korobov and Thorne, 2006; Tolman et al., 2004; Way and Chu, 2004), and other recent work has explored the notion that cohort shifts have generally amplified girls' freedom and efficacy within the heterosexual realm (Risman and Schwartz, 2002). Such findings thus present a picture of adolescent dating relationships that complicates themes of male dominance and female dependence that have been developed in the adolescence literature in general, and the dating violence literature in particular. Nevertheless, previous analyses have not examined links to physical violence within the relationship. Thus, it is possible that more traditionally gendered asymmetries of power and dependence will be observed within those relationships that include physical violence, even though these patterns were not observed across the sample as a whole.

In summary, while family of origin influences and peer norms are important influences on teen dating violence, it is also potentially useful to develop a descriptive portrait of the relationship contexts within which these behaviors unfold. This is consistent with a symbolic interactionist perspective that focuses on the situated nature of meanings and behavior, and emphasizes the need to take actors' own perspectives on these relationships into account. A comprehensive assessment requires attention to positive and negative dynamics within these relationships, as well as more general patterns of interaction and influence.

The Current Study

The objective of the current analysis is to document variations in the qualities and dynamics of adolescent relationships, and the degree to which such dynamics are linked to the experience of teen dating violence. Based on the results of prior research, we expect that physically violent relationships will include more troubling features. Thus, verbal conflicts, jealousy, cheating, as well as a perceived lack of identity support will be assessed. However, our multidimensional portrait will also consider more general patterns of interaction and influence (duration of the relationship, time spent together, whether the relationship is sexually intimate, perceived partner influence and power balance within the relationship). In addition, a measure of time spent with friends is included, based on the notion that violent relationships may be associated with greater isolation from friends (Tolman, 1989). We also assess positive features of the relationship, including feelings of love, caring, and level of intimate self-disclosure. Finally, we measure the receipt and provision of instrumental support, recognizing that dating may be associated with tangible as well as ‘intrinsic’ rewards. We evaluate whether these processes are gendered by examining gender by relationship quality interactions, and include results of separate models estimated for girls and boys as well as across the sample as a whole. For example, greater partner power and influence may be associated with girls' but not boys' reports of violence, and girls in violent relationships may report less time spent with their friends.

The analyses to follow focus on respondents' reports of perpetration of violence rather than their experiences of victimization. As prior research on both adolescent and adult populations shows, there are strong links between perpetration and victimization (most violence includes some level of mutual participation, although the idea of sexual symmetry in violent relationships remains controversial (see Makepeace, 1986; Cleveland et al., 2003). We estimated a series of models that explored variations in findings resulting from a focus on victimization, perpetration, or the experience of any violence (whether as victim or perpetrator). We find only minor differences (i.e., similar patterns of significance were observed). Thus, we focus primarily on the adolescent's own use of violence, because our measurement of relationship qualities concentrates on the respondent's perspectives and feelings about this relationship. In short, it is somewhat more intuitive to connect respondents' own feelings to their own actions. However, we have included some measures that tap the partner's feelings and behaviors (e.g., whether the partner is jealous, partner cheating, partner's receipt of support from the respondent), albeit from the respondent's perspective, and these are also included in our assessments.

Methods

Sample

The sample used in the current study is the Toledo Adolescent Relationships Study (TARS). The sample was drawn from the year 2000 enrollment records of all youths registered for the 7th, 9th, and 11th grades in Lucas County, Ohio, a largely urban metropolitan environment that includes the city of Toledo (n = 1,321). The sample universe encompassed records elicited from 62 schools across seven school districts. The stratified, random sample was devised by the National Opinion Research Center and includes over-samples of black and Hispanic adolescents. School attendance was not a requirement for inclusion in the sample, and most interviews were conducted in the respondent's home using preloaded laptops to administer the interview. This method is particularly important when studying relationship violence because the respondent is guaranteed privacy in responses to questions about violence. Unlike other data sources such as Add Health, this survey includes multiple measures of relationship qualities and dynamics, as well as measures of violence perpetration and victimization.

From the total sample of 1,321, we focus the present analysis on the 956 respondents who reported either that they were currently dating or having recently dated during the previous year. The distribution of the dependent variables is presented in Table 1 and independent variables in Table 2. Respondents who were missing responses on the independent variables were assigned the mean value. This is relatively uncommon (in one instance, four of 956 responses were missing, but for a majority of questions, zero responses were missing).

Table 1.

Prevalence of Violence and Types of Violence Reported

Males Females Total
Any Type of Violence
(Mutual, Perpetrator, or Target)		 30.9% 21.8% 26.3%
 Mutually Violent 46.9% 51.9% 49.0%
 Perpetrator Only 6.2% 34.9% 18.3%
 Target Only 46.9% 13.2% 32.7%
N 145 106 251
Open in a new tab

Table 2.

Descriptive Statistics for Relationship Quality Variables, Traditional Violence Predictors, and Sociodemographic Controls (N = 956)a

Mean/Percentage S.D. Range
Relationship Qualities/Dynamics
Problematic Features
 Verbal conflict 1.97 .87 1-5
 Partner jealous of teen 3.02 1.20 1-5
 Teen jealous of partner 2.63 1.14 1-5
 Teen cheated on partner 1.59 .95 1-5
 Partner cheated on teen 1.51 .87 1-5
 Lack of identity support 1.84 .78 1-5
Intrinsic Rewards
 Intimate self-disclosure 3.25 .98 1-5
 Passionate love 3.53 .89 1-5
 Partner cares 4.14 .85 1-5
 Instrumental support from partner 2.31 1.00 1-5
 Instrumental support from teen 2.27 .99 1-5
Patterns of Influence and Interaction
 Relationship duration 4.79 2.07 1-8
 Time spent with partner 2.34 1.72 0-6
 Time spent with friends 3.16 .92 1-4
 Had sex with partner 28.2% - -
 Partner-on-teen influence 2.65 .41 1-5
 Teen power in relationship 2.35 .57 1-3
Traditional Violence Predictors
 Parental monitoring 1.90 .97 1-5
 Parental arguing 2.48 1.04 1-4
 Parent-to-teen violence 1.85 .56 1-5
 Friends' violence 1.60 1.29 1-9
 GPA 2.58 1.00 0-4
 Current relationship 58.5% - -
Sociodemographics
 Male 48.9% - -
 Female 50.2% - -
 Age 15.49 1.70 12-18
 SES 4.12 1.53 1-7
 White 69.0% - -
 Black 19.8% - -
 Hispanic 11.2% - -
 Two biological parents 52.6% - -
 Step-parent 15.3% - -
 Single-parent 20.6% - -
 Other family type 11.6% - -
Open in a new tab
a

Except where noted, figures represent mean scores.

Source: Toledo Adolescent Relationships Study

Measures

Dependent Variable: Relationship Violence

We relied on four items from the Revised Conflict Tactics Scale (Straus and Gelles, 1990), including “thrown something at,” “pushed, shoved, or grabbed,” “slapped in the face or head with an open hand,” and “hit,” focusing on the current/most recent partner. Respondents who experienced any of the violent behaviors were coded as experiencing violence. These items are asked about the respondent as the target (alpha = .91), and subsequently as the perpetrator (alpha = .89) and reference the current or most recent partner.

Relationship Qualities

Problematic Features

Verbal conflict was measured separately from physical violence. This concept was measured by a three-item scale which asks respondents to indicate how often (responses ranging from “Never” to “Very Often”) the following situations occur: “You have disagreements with X,” “You give each other the silent treatment,” and “You yell and shout at each other” (alpha = .83). To measure partner's jealousy, we use a single item asking respondents how strongly they agree with the statement “When I am around other [girls/boys], X gets jealous.” To measure respondent's jealousy we use a mirror-version of the above item. To measure a lack of identity support, we use a 2-item scale (alpha = .63). Respondents are asked how much they agree with the following statements: “X is often disappointed in me” and “X seems to wish I were a different type of person”. We also include a measure indicating how often the teen cheated on their partner; the item asks how often the respondent “[saw/has seen] another person while with X”. We use a similar measure to indicate how often the partner cheated on the respondent.

Rewards of the Relationship

To measure intimate self-disclosure we rely on a revised version of West and Zingle's (1969) self-disclosure scale. This five-item index asks respondents to report about how often (from 1 (never) to 5 (very often)) they communicate with the partner about a range of topics, for example, “something really great that happened, something really bad that happened, your private thoughts and feelings” (alpha = .87). To measure love we use four items drawn from Hatfield and Sprecher's (1986) passionate love scale, including “I would rather be with X than anyone else,” “I am very attracted to X,” “the sight of X turns me on,” and “X always seems to be on my mind” (alpha = .85). Caring is measured with a single item asking the respondent how much s/he agrees with the statement “X cares about me.” To measure instrumental support from the partner, we use a scale specifically designed for the TARS study. This scale includes four items asking how often X “lets you borrow something,” “loan or give you money,” “give you a present,” and “pitch in and help you do things.” Responses consist of a five-category Likert scale and range from “Never” to “Very Often” (alpha = .86). Similarly,instrumental support from the teen respondent is measured using a four-item scale with identical items as the scale described above (alpha = .87).

Patterns of Interaction and Influence

Relationship duration is a single item, in which the respondent indicates how long they and their partner have been together at the time of the interview (the question was not asked of those who had already broken up; instead, the mean was imputed). Responses could range from “less than a week” to “more than a year”. We use a two-item scale indicating how much time the respondent spends with his/her partner. Items include how often in the typical week the respondent goes to their partner's house, or meets after school to go somewhere or “hang out” (alpha = .65). We also use a single measure of how much time the respondent spends with friends. We use a dichotomous variable based on a single item indicating whether or not the respondent has had sex with their partner (1 = yes).

Our measure of the partner's influence on the respondent is a scale consisting of six items (alpha = .71). Examples include: “X often influences what I do,” “I sometimes do things because X is doing them,” and “I sometimes do things because I don't want to lose X's respect.” It should be noted that this measure does not reflect the amount of influence relative to the respondent, but indicates how much influence the respondent perceives from their partner. We modeled our measure of power on Blood and Wolfe's (1960) decision power index revised for use with this younger sample. The scale includes an overall assessment (“If the two of you disagree, who usually gets their way?”) and also includes items that reference specific situations: “what you want to do together,” “how much time you spend together,” and “how far to go sexually.” Responses include “X more than me,” “X and me about the same,” and “me more than X.” Higher scores indicate that the respondent feels that they have more power, relative to the partner (alpha = .77).

Traditional Violence Predictors

Parental Monitoring is measured by a six item scale completed by the parent that includes items such as: “When my child is away from home, s/he is supposed to let me know where s/he is,” “I call to check if my child is where s/he said,” “My child has to be home at a specific time on the weekends” (alpha = .73). Parental conflict is a single item taken from the parental questionnaire asking how often the parent (most often the mother) argues with their current romantic partner. We also employ a scale measuring parent-to-teen violent behavior (alpha = .80). The items are identical to those in the relationship violence scale. Friends' violence is a single item, indicating how often the respondent's friends have attacked someone with the intention of seriously hurting them. Academic performance was measured by asking the question, “What grades did you get in school this year?” The item was recoded to reflect GPAs ranging from 0.0 to 4.0. We also include a simple dummy variable indicating whether the relationship in question is a current relationship or their most recent romantic relationship (1 = current).

Sociodemographic Variables

We use a dummy variable to control for gender, with females as the contrast category. Age is a continuous variable measured in years. To control for socioeconomic status, we use the highest level of education reported by the parent filling out our parental questionnaire. Because the parental sample is overwhelmingly female, we simply refer to this as “mother's education.” Race/ethnicity includes the following categories: Non-Hispanic white (contrast category), Non-Hispanic black, and Hispanic. To measure the family structure of the respondent, we categorize families into: two biological parents (contrast category), step-family, single-parent family, and any “other” family type.

Analytic Strategy

We first present basic prevalence information regarding the experience of violence within the TARS sample, including reports of perpetration and victimization. Subsequently we focus primarily on variations in the odds of self-reported perpetration. We present bivariate models that document observed relationships between each relationship quality or characteristic and odds of perpetration. The next logistic regression models include controls for traditional violence predictors and basic sociodemographic characteristics. These analyses allow us to determine significant associations between relationship qualities and dynamics and perpetration of violence, once these traditional covariates have been introduced. We chose to focus on each relationship dimension in a separate model, because prior research has not considered these multiple facets of teen relationships. Thus, our intent in this initial investigation is to provide a more comprehensive picture of the character of relationships that include violence, rather than to pit these relationship variables against one another to determine relative impact. We also test for gender and relationship quality interactions to determine whether there are distinctively gendered patterns of association between these relationship qualities and self-reported perpetration.

Findings presented below are based on a cross-sectional analysis. In this initial investigation of the links between relationship qualities and violent actions, we wish to ensure that reports of specific dynamics reference the same relationship in which violence either occurs or does not occur. Nevertheless, a limitation of this strategy is that we cannot conclude from results that particular features of these relationships are causes of the outcome of interest, violence, but only whether a particular dynamic/relationship quality is a significant correlate, net of other factors found in prior research to be predictors of teen dating violence. Thus, for example, if an inverse relationship between feelings of love and violence is observed, it could be that violence itself influenced these decreased feelings of love, rather than the situation in which a lack of emotional intimacy actually fostered the violent outcome. Although this is a limitation, these data allow us to draw on a broad range of indicators with the objective of building a more comprehensive portrait of violent and non-violent relationships. A key issue is that adolescent romantic relationships are typically of relatively short duration (Carver, Joyner, and Udry, 2003); thus a longitudinal design would often not capture sufficient subjects for whom reports of violence and reports of relationship qualities reference the same focal relationship. Nevertheless, it is important to highlight that the cross-sectional nature of the study prevents inferences about the direction of causality.

Results

Table 1 presents the prevalence and distribution of dating violence within the TARS sample. A sizeable minority of respondents do report some type of violence, with 26% of the sample reporting victimization, perpetration, or both (31% of males and 22% of females). Our focal dependent variable is perpetration, and 19% of females and 15% of males report violence perpetration. Table 1 also presents the distribution of the various subtypes of violence these respondents reported experiencing. Mutual violence (defined as reporting both victimization and perpetration over the course of the relationship) is the most common type reported, with 49% of respondents in violent relationships reporting this type of violence. Rates of mutual violence are very similar for both males (47%) and females (52%). Among those respondents who report experiencing either perpetration or victimization, gender differences do appear: only 6% of males in violent relationships report only perpetration, compared to 35% of females; similarly, about 13% of females in violent relationships report victimization only, compared to 47% of males. Thus, it appears that while mutually violent relationships appear to be the most common form of violent relationship for all respondents, female respondents more often report being the perpetrator of violence, while males are more likely to report being the sole target of violence. The remaining analyses focus on violence perpetration, and the association between relationship qualities/dynamics and odds of reporting violence perpetration within the focal relationship. Table 2 presents descriptive statistics for the relationship quality variables, traditional violence predictors and sociodemographic controls.

Table 3 presents bivariate (model 1) and multivariate (model 2) results predicting the odds of perpetrating violence. Model 2 includes sociodemographic and other control variables often associated with teen dating violence. We begin with zero-order regressions, examining each relationship factor with no other controls in the model. As shown in model 1, all of the problematic features measured (verbal conflict, jealousy of respondent and partner, less identity support from partner, and more infidelity on the part of the respondent and partner) are associated with significantly higher odds of perpetration. These findings are consistent with the emphases of prior research studies, which have often focused on these troubling aspects of violent relationships. However, the results add to our knowledge in assessing subjectively experienced aspects such as jealousy directly.

Table 3.

Logistic Regressions on Respondent Perpetration (N=956)b

Model 1 Model 2
Predictor b exp(b) s.e. b exp(b) s.e.
Relationship Qualities/Dynamics
Problematic Features
 Verbal conflict 1.07*** 2.91 .10 .95*** 2.60 .11
 Partner jealous of teen .44*** 1.56 .08 .39*** 1.47 .08
 Teen jealous of partner .37*** 1.44 .07 .37*** 1.45 .08
 Teen cheated on partner .33*** 1.38 .08 .16 1.18 .09
 Partner cheated on teen .35*** 1.42 .35 .27** 1.30 .09
 Lack of identity support .46*** 1.58 .10 .42*** 1.52 .11
Intrinsic Rewards
 Intimate self-disclosure .09 1.10 .09 .09 1.09 .10
 Passionate love .13 1.14 .10 .12 1.13 .11
 Partner cares .05 1.05 .10 −.03 .97 .11
 Instrumental support from partner .43*** 1.53 .08 .39*** 1.48 .10
 Instrumental support from teen .25** 1.28 .08 .23* 1.25 .23
Patterns of Influence and Interaction
 Relationship duration .22*** 1.25 .04 .19*** 1.21 .05
 Time spent with partner .20*** 1.22 .05 .19*** 1.21 .06
 Time spent with friends −.05 .95 .09 .02 1.02 .10
 Had sex with partner 1.13*** 3.10 .18 .93*** 2.53 .22
 Partner-on-teen influence .22 1.25 .21 .36 1.43 .36
 Teen power in relationship −.03* .97 .15 −.49** .61 .16
Open in a new tab
b

Model 1 represents bivariate results. Model 2 includes controls for sociodemographic characteristics (gender, age, mother's education, race/ethnicity, family structure), traditional violence predictors (parental monitoring, parental arguing, parent-to-teen violence, friends' violence, gpa), and whether/not the relationship is current.

*

p < .05

**

p < .01

***

p < .001

Source: Toledo Adolescent Relationships Study

Although a comprehensive treatment of relationship dynamics will thus necessarily include attention to the negative features of these relationships, results also reported in model 1 of Table 3 suggest a somewhat more complex picture. The results of our examination of intrinsic rewards of these relationships indicate that levels of partner caring, love, and intimate self-disclosure are similar in relationships with and without perpetration. In addition, respondents reporting greater provision and receipt of instrumental benefits associated with their dating relationships have significantly higher odds of perpetration.

Turning to more basic contours of these relationships (i.e., patterns of interaction and influence), relationships of longer duration are associated with greater odds of relationship violence. Results also indicate that greater time spent with the partner is associated with increased odds of perpetration, consistent with the idea of enmeshment in the relationship. However, it is also interesting to note that time spent with friends is not systematically linked with violence reports. As Table 3 indicates, having had sex with partner has a large effect, with significantly higher odds of violence perpetration among those who describe their relationships as sexually intimate. In contrast, levels of perceived partner influence are not related to violence reported within the relationship. Finally, perceptions of the power balance within the relationship are negatively associated with reports of violence—those who report a less favorable power balance have greater odds of violence.

The next set of models (model 2) in Table 3 include socio-demographic controls (gender, age, mother's education, race/ethnicity, and family structure) and control variables generally associated with violence perpetration (parental monitoring, parental conflict, parent-to-teen violence, friends' violent behavior, GPA), and a dummy indicating whether or not the relationship is current. All of the significant relationships observed in the bivariate model persist in the multivariate model, with the exception of cheating. Cheating is not significantly related to perpetration of violence once controls for friends' violence and GPA have been introduced (results not shown). It should be noted that cheating remains marginally significant (p = .06) even after the addition of these controls. We do find, however, that a majority of the significant effects are diminished in the multivariate models, indicating that these basic controls are important to consider in analyses of perpetration.

Given the importance of gender to discussions of relationship violence, we next included a series of gender by relationship quality interaction terms in the above models (results not shown). Most interactions are not statistically significant, suggesting generally similar patterns of association for males and females. An exception is the interaction of gender and verbal conflict indicating a stronger relationship between verbal conflict and violence perpetration for female respondents. However, an examination of separate male and female models indicates that verbal conflict remains a significant predictor for both males (p < .001) and females (p < .001). Given the centrality of power to many discussions of teen dating violence, it is also of note that in these separate gender models, perceptions of lower power in the relationship are significantly associated with male but not female perpetration. Examining mean levels of perceived power, we find that male perpetrators have the lowest average score on this index. As Figure 1 illustrates, female respondents, regardless of whether or not they report engaging in violence, on average report a more favorable power balance within their relationships. This basic gender pattern is also observed when we examine the scores on the power index of male and female respondents who correspond to the various subtypes described in Table 2 (i.e., whether they reported a target only, mutual or perpetration only pattern, female respondents score higher on this scale).

Figure 1. Violence and Perceptions of Power within Adolescent Romantic Relationshipsc.

Figure 1

Open in a new tab

c Note: Higher scores represent a more favorable perception of power.

As a final step in the analyses, we estimated models identical to those described above, in which the dependent variable was respondent reports of violent victimization within the focal relationship, or the presence of “any violence” (whether in response to the questions about perpetration or victimization) (results not shown). In these analyses, the pattern of association with relationship qualities and dynamics was quite similar. These findings are not particularly surprising given the strong links between victimization and perpetration observed within the TARS data (see Table 2) and other studies; yet they suggest that results reported are not uniquely influenced by our focus here on variations in perpetration.

Discussion

The analyses described above highlight specific dynamics within adolescent dating relationships that are associated with the experience of violence, suggesting the need to focus greater attention on relationship context, as well as on more distal processes such as family of origin experiences and peer normative climates. The findings generally support but add specificity to prior research and prevention efforts that have focused heavily on negative relationship processes. Respondents who self-report violence perpetration are significantly more likely than their non-violent counterparts to report greater verbal conflict, jealousy, cheating, and a lack of identity support. Importantly, these associations were significant even when controls for traditional violence predictors were introduced. These findings not only support the symbolic interactionist theoretical focus on the immediate contexts in which behaviors unfold, but unlike such risk factors as early family exposure or neighborhood poverty, these relationship dynamics may be more malleable/subject to redirection. The findings also accord well with results of Capaldi and Kim's (2007) recent longitudinal study of teen dating violence among high risk youths, which documented continuity in reports of violence within a specific relationship but not across relationships. Such findings provide evidence that even in the presence of known risk factors, the experience of violence within the dating context is not inevitable.

The current study contributes further to our understanding of these relationships in demonstrating that some positive dynamics are also reported by young people who have experienced violence within their relationships. We find no statistically significant differences between non-violent and violent relationships in levels of love, intimate self-disclosure and perceived partner caring, revealing elements of complexity to relationships that include troubling modes of conflict resolution. Indeed, our view is that it has been useful to examine each of these dimensions of intimacy separately rather than to include them in a summary scale (e.g., one measuring ‘positive features’), as our approach serves to highlight several related but conceptually distinct benefits and rewards of these youthful relationships. The direction of findings is intuitive, since the adolescent who defines a particular relationship as containing only negative dynamics would likely find it an easier matter to break up with such a partner. Thus, even though adolescent relationships do not typically involve issues of economic dependence or concerns about child well-being that may influence adult women's stay/leave decisions (Anderson, 2007), they may nevertheless contain elements of intimacy and perceived importance that makes it difficult to withdraw easily from them. The higher scores on the provision and receipt of instrumental support, longer average durations, and frequent contact reported by respondents who experienced violence contribute to this portrait. And the strong association between sexual intercourse and odds of perpetration develops further the notion of ‘enmeshment,’ suggesting an amplifying role for the heightened emotionality that often accompanies sexual intimacy (see especially Kaestle and Halpern, 2005).

These findings offer a framework for elaborating more complex models of relationship factors associated with risk, and for the content of relationship-focused prevention efforts. For example, young people involved in relationships that contain a number of positive elements may not relate fully to programming that emphasizes only negative processes. Such programs will necessarily focus heavily on the negative dynamics that are warning signs of an abusive relationship, but could include greater attention to the positive feelings and emotions that may coexist with them, potentially influencing actions within relationships, and complicating the decision to leave a particular partner.

The results of our analyses also indicated that in general perceptions of a less favorable power balance were associated with increased odds of perpetration, but differences were only significant for male respondents: young men who reported perpetration perceived lower power when compared with their non-violent male counterparts. These results should be viewed against the backdrop of previous findings indicating that male respondents in general reported a less favorable power balance relative to young women who participated in the TARS study (Giordano et al., 2006). In addition, recent analyses of wave four perceptions of power (when respondents averaged 21 years of age) document a similar gender pattern (Giordano et al., 2009), suggesting that these findings are not a simple function of the relatively young age of respondents at the initial wave of interviews we focused on in this analysis.

Prior research and prevention programming has often focused on issues of power and control as central to an understanding of teen dating violence. For example, numerous prevention programs rely on the “power and control wheel” as a tool to describe abusive relationships (Nicarthy, 1986). In addition to the exclusive focus on negative relationship processes as constituting the various spokes of the wheel (e.g., harassment, intimidation, isolation, humiliation), this visual aid depicts power and control as the central dynamic, and highlights the role of male privilege for an understanding of conduct within these intimate relationships. Thus, the findings with respect to perceptions of the power balance are potentially important, and suggest the need for a more nuanced approach to the power dimension in theory building and within the context of prevention efforts. Such results fit generally with Stets' (1991) view of violence as a form of power assertion rather than as a straightforward reflection of male dominance or privilege, and underscore that macro-level explanations of social phenomena may have an imperfect fit with the individual's own understandings of these same phenomena (see also Kaura and Allen, 2004). It is also possible that some forms of violence reflect motives other than a desire to control the partner, as Felson and Outlaw (2007) recently argued in a study of controlling behaviors and marital violence.

A limitation of this analysis is that the interview data were collected within one metropolitan area. More research is also needed that relies on different measures of relationship dynamics, including power dynamics. For example, direct observations of couple interactions or indices that include specific domains of power may reveal different or more complex patterns. Future research on relationship qualities and dynamics and their association with dating violence will also benefit from additional studies that incorporate a temporal dimension (Capaldi and Kim, 2007). Studies focused on a cross-section of adolescents will by definition capture romantic relationships representing a range of stages, and the short average durations of teen relationships poses a further complication. Longitudinal designs that include shorter periods between assessments and the use of timelines or relationship history calendars could be useful strategies for disentangling causal order, and determining the sequence of relationship dynamics that are associated with the greatest levels of risk for the occurrence of violence. For example, the perception that the relationship has some positive features may stem largely from experiences early on in the relationship, suggesting the need for more fine-grained analyses of shifts over time and factors that are associated with an escalation of unhealthy patterns.

Another potentially important finding is that many relationship dimensions were similarly associated with male and female reports of perpetration. For example, female respondents' own feelings of jealousy and perceptions of a lack of identity support were significantly related to their self-reports of violence perpetration. Such findings suggest (but do not directly demonstrate) that some violent actions on the part of young women may be linked to feelings of anger and frustration with the partner or the state of the relationship, rather than emerging only as a reaction to a partner's jealousy, violence or threats of violence (Dobash et al., 1992). Official sources of information and survey data clearly reveal that male perpetration and female victimization represents the more pressing social problem. Nevertheless, the TARS prevalence data and many other surveys do reveal that a significant minority of young women report hitting their partners. Further, recent analyses suggest that violence involving some level of ‘mutuality’ may be most likely to result in physical injury (Whitaker et al., 2007). Thus, additional research on the generic and distinctively gendered processes associated with inflicting as well as sustaining violence is needed for a comprehensive understanding of the relationship dynamics associated with teen dating violence. In particular qualitative approaches and couple level analyses may be useful strategies for elucidating male and female perceptions of the meanings of violent acts (their own, their partner's), perceived effects on relationship dynamics, and more general consequences for emotional and physical well-being.

Acknowledgments

This research was supported by a grant from The Eunice Kennedy Shriver National Institute of Child Health and Human Development (HD036223), and by the Center for Family and Demographic Research, Bowling Green State University, which has core funding from The Eunice Kennedy Shriver National Institute of Child Health and Human Development (R24HD050959-01).

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Anderson KL. Theorizing gender in intimate partner violence research. Sex Roles. 2005;52(11-12):853–865. [Google Scholar]
  2. Anderson KL. Who gets out?: Gender as structure and the dissolution of violent heterosexual relationships. Gender and Society. 2007;21(2):173–201. [Google Scholar]
  3. Arias I, Samios M, O'leary KD. Prevalence and correlates of physical aggression during courtship. Journal of Interpersonal Violence. 1987;2(1):82–90. [Google Scholar]
  4. Blood RO, Wolfe DM. Husbands and wives: The dynamics of married living. Free Press; Glencoe, IL: 1960. [Google Scholar]
  5. Brown SL, Bulanda JR. Relationship violence in young adulthood: A comparison of daters, cohabitors, and marrieds. Social Science Research. 2008;37(1):73–87. [Google Scholar]
  6. Burke PJ, Stets JW, Pirog-Good MA. Gender identity, self-esteem, and physical and sexual abuse in dating relationships. Social Psychology Quarterly. 1988;51:272–285. [Google Scholar]
  7. Call KT, Mortimer JT. Review of arenas of comfort in adolescence: A study of adjustment in context. Journal of Adolescence. 2003;26(1):143–144. [Google Scholar]
  8. Capaldi DM, Kim HK. Typological approaches to violence in couples: A critique and alternative conceptual approach. Clinical Psychology Review. 2007;27(3):253–265. doi: 10.1016/j.cpr.2006.09.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Carver K, Joyner K, Udry JR. Adolescent romantic relations and sexual behavior: Theory, research, and practical implications. Lawrence Erlbaum Associates Publishers; Mahwah, NJ: 2003. National estimates of adolescent romantic relationships; pp. 23–56. [Google Scholar]
  10. Cleveland HH, Herrera VM, Stuewig J. Abusive males and abused females in adolescent relationships: Risk factor similarity and dissimilarity and the role of relationship seriousness. Journal of Family Violence. 2003;18(6):325–339. [Google Scholar]
  11. Cohen AK. Delinquent boys: The culture of the gang. Free Press; New York: 1955. [Google Scholar]
  12. Cornelius TL, Resseguie N. Primary and secondary prevention programs for dating violence: A review of the literature. Aggression and Violent Behavior. 2007;12(3):364–375. [Google Scholar]
  13. DeMaris A. The dynamics of generational transfer in courtship violence: A biracial exploration. Journal of Marriage and the Family. 1990;52(1):219–231. [Google Scholar]
  14. Dobash RP, Dobash RE, Wilson M, Daly M. The myth of sexual symmetry in marital violence. Social Problems. 1992;39:71–91. [Google Scholar]
  15. Dutton MA, Goodman LA. Coercion in intimate partner violence: Toward a new conceptualization. Sex Roles. 2005;52(11-12):743–756. [Google Scholar]
  16. Eder D, Evans C, Parker S. School talk: Gender and adolescent culture. Rutgers University Press; New Brunswick, NJ: 1995. [Google Scholar]
  17. Felson RB, Outlaw MC. The control motive and marital violence. Violence and Victims. 2007;22(4):387–407. doi: 10.1891/088667007781553964. [DOI] [PubMed] [Google Scholar]
  18. Foshee VA, Karriker-Jaffe K, Reyes HLM, Ennett ST, Suchindran C, Bauman KE, et al. What accounts for demographic differences in trajectories of adolescent dating violence? An examination of intrapersonal and contextual mediators. Journal of Adolescent Health. 2008;42(6):596–604. doi: 10.1016/j.jadohealth.2007.11.005. [DOI] [PubMed] [Google Scholar]
  19. Gaertner L, Foshee V. Commitment and the perpetration of relationship violence. Personal Relationships. 1999;6:227–239. [Google Scholar]
  20. Gallaty K, Zimmer-Gembeck M. The daily social and emotional worlds of adolescents who are psychologically maltreated by their romantic partners. Journal of Youth and Adolescence. 2008;37(3):310–323. [Google Scholar]
  21. Giordano PC, Flanigan CM, Manning WD, Longmore MA. Romantic relationships and the transition to adulthood: Developmental shifts in communication, emotion, influence, and utility; Paper presented at the annual meetings of the American Sociological Association; San Francisco. Aug, 2009. [Google Scholar]
  22. Giordano PC, Longmore MA, Manning WD. Gender and the meanings of adolescent romantic relationships: A focus on boys. American Sociological Review. 2006;71(2):260–287. [Google Scholar]
  23. Halpern CT, Oslak SG, Young ML, Martin SL, Kupper LL. Partner violence among adolescents in opposite-sex romantic relationships: Findings from the national longitudinal study of adolescent health. American Journal of Public Health. 2001;91(10):1679–1685. doi: 10.2105/ajph.91.10.1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hatfield E, Sprecher S. Measuring passionate love in intimate relationships. Journal of Adolescence. 1986;9(4):383–410. doi: 10.1016/s0140-1971(86)80043-4. [DOI] [PubMed] [Google Scholar]
  25. Henton J, Cate R, Koval J, Lloyd S, Christopher S. Romance and violence in dating relationships. Journal of Family Issues. 1983;4(3):467–482. [Google Scholar]
  26. Herrenkohl TI, Hill KG, Hawkins JD, Chung I, Nagin DS. Developmental trajectories of family management and risk for violent behavior in adolescence. Journal of Adolescent Health. 2006;39(2):206–213. doi: 10.1016/j.jadohealth.2005.11.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Howard DE, Wang MQ. Risk profiles of adolescent girls who were victims of dating violence. Adolescence. 2003;38(149):1–14. [PubMed] [Google Scholar]
  28. Kaestle CE, Halpern CT. Sexual intercourse precedes partner violence in adolescent romantic relationships. Journal of Adolescent Health. 2005;36:386–392. doi: 10.1016/j.jadohealth.2004.02.030. [DOI] [PubMed] [Google Scholar]
  29. Kaura SA, Allen CM. Dissatisfaction with relationship power and dating violence perpetration by men and women. Journal of Interpersonal Violence. 2004;19(5):576–588. doi: 10.1177/0886260504262966. [DOI] [PubMed] [Google Scholar]
  30. Kenney CT, McLanahan S. Why are cohabiting relationships more violent than marriages? Demography. 2006;43(1):127–140. doi: 10.1353/dem.2006.0007. [DOI] [PubMed] [Google Scholar]
  31. Komter A. Hidden power in marriage. Gender and Society. 1989;3(2):187–216. [Google Scholar]
  32. Korobov N, Thorne A. Intimacy and distancing: Young men's conversations about romantic relationships. Journal of Adolescent Research. 2006;21(1):27–55. [Google Scholar]
  33. Larson R, Richards MH. Divergent realities: The emotional lives of mothers, fathers, and adolescents. Basic Books; New York: 1994. [Google Scholar]
  34. Luthra R, Gidycz CA. Dating violence among college men and women: Evaluation of a theoretical model. Journal of Interpersonal Violence. 2006;21(6):717–731. doi: 10.1177/0886260506287312. [DOI] [PubMed] [Google Scholar]
  35. Maccoby EE. Gender and relationships: A developmental account. American Psychologist. 1990;45(4):513–520. doi: 10.1037//0003-066x.45.4.513. [DOI] [PubMed] [Google Scholar]
  36. Magdol L, Moffitt TE, Caspi A, Silva PA. Developmental antecedents of partner abuse. Journal of Abnormal Psychology. 1998;107(3):375–389. doi: 10.1037//0021-843x.107.3.375. [DOI] [PubMed] [Google Scholar]
  37. Makepeace JM. Gender differences in courtship violence victimization. Family Relations. 1986;35(3):383–388. [Google Scholar]
  38. McCloskey LA, Lichter EL. The contribution of marital violence to adolescent aggression across different relationships. Journal of Interpersonal Violence. 2003;18(4):390–412. [Google Scholar]
  39. Mead GH. Mind, self, and society from the standpoint of a social behaviorist. University of Chicago Press; Chicago: 1934. [Google Scholar]
  40. Nicarthy G. Getting Free: You Can End Abuse and Take Back Your Life. Seal Press; Seattle, WA: 1986. [Google Scholar]
  41. O'Keefe M. Predictors of dating violence among high school students. Journal of Interpersonal Violence. 1997;12(4):546–568. [Google Scholar]
  42. O'Keefe M, Treister L. Victims of dating violence among high school students: Are the predictors different for males and females? Violence Against Women. 1998;4(2):195–223. doi: 10.1177/1077801298004002005. [DOI] [PubMed] [Google Scholar]
  43. O'Keeffe NK, Brockopp K, Chew E. Teen dating violence. Social Work. 1986;31(6):465–468. [Google Scholar]
  44. O'Leary KD, Slep AMS. A dyadic longitudinal model of adolescent dating aggression. Journal of Clinical Child and Adolescent Psychology. 2003;32(3):314–327. doi: 10.1207/S15374424JCCP3203_01. [DOI] [PubMed] [Google Scholar]
  45. O'Leary KD, Slep AMS, Avery-Leaf S, Cascardi M. Gender differences in dating aggression among multiethnic high school students. Journal of Adolescent Health. 2008;42(5):473–479. doi: 10.1016/j.jadohealth.2007.09.012. [DOI] [PubMed] [Google Scholar]
  46. Ray AL, Gold SR. Gender roles, aggression, and alcohol use in dating relationships. Journal of Sex Research. 1996;33:47–55. [Google Scholar]
  47. Risman B, Schwartz P. After the sexual revolution: Gender politics in teen dating. Contexts. 2002;1(1):16–24. [Google Scholar]
  48. Roberts TA, Auinger P, Klein JD. Predictors of partner abuse in a nationally representative sample of adolescents involved in heterosexual dating relationships. Violence and Victims. 2006;21(1):81–89. [PubMed] [Google Scholar]
  49. Sigelman CK, Berry CJ, Wiles KA. Violence in college students' dating relationships. Journal of Applied Social Psychology. 1984;14(6):530–548. [Google Scholar]
  50. Silverman JG, Raj A, Mucci LA, Hathaway JE. Dating violence against adolescent girls and associated substance use, unhealthy weight control, sexual risk behavior, pregnancy, and suicidality. The Journal of the American Medical Association. 2001;286(5):572–579. doi: 10.1001/jama.286.5.572. [DOI] [PubMed] [Google Scholar]
  51. Stets JE. Psychological aggression in dating relationships: The role of interpersonal control. Journal of Family Violence. 1991;6(1):97–114. [Google Scholar]
  52. Stets JE, Pirog-Good M. Violence in dating relationships. Social Psychology Quarterly. 1987;50(3):237–246. [Google Scholar]
  53. Stets JE, Straus MA. The marriage license as a hitting license: A comparison of assaults in dating, cohabiting, and married couples. Journal of Family Violence. 1989;4(2):161–180. [Google Scholar]
  54. Straus MA, Gelles RJ. Physical violence in American families. Transaction Publishers; New Brunswick, NJ: 1990. [Google Scholar]
  55. Tolman DL, Spencer R, Harmon T, Rosen-Reynoso M, Striepe M. Adolescent boys: Exploring diverse cultures of boyhood. New York University Press; New York, NY: 2004. Getting close, staying cool: Early adolescent boys' experiences with romantic relationships; pp. 235–255. [Google Scholar]
  56. Tolman RM. The development of a measure of psychological maltreatment of women by their male partners. Violence and Victims. 1989;4(3):159–177. [PubMed] [Google Scholar]
  57. Vivian D, O'Leary KD. Communication patterns in physically aggressive engaged couples; Third National Family Violence Research Conference; Durham, NC. 1987. [Google Scholar]
  58. Way N, Chu JY. Adolescent boys: Exploring diverse cultures of boyhood. New York University Press; New York, NY: 2004. [Google Scholar]
  59. West L, Zingle HW. A self-disclosure inventory for adolescents. Psychological Reports. 1969;24(2):439–445. doi: 10.2466/pr0.1969.24.2.439. [DOI] [PubMed] [Google Scholar]
  60. Whitaker DJ, Haileyesus T, Swahn M, Saltzman LS. Differences in frequency of violence and reported injury between relationships with reciprocal and nonreciprocal intimate partner violence. American Journal of Public Health. 2007;97(5):941–947. doi: 10.2105/AJPH.2005.079020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Wight D. Boys' thoughts and talk about sex in a working class locality of Glasgow. Sociological Review. 1994;42(4):703–737. [Google Scholar]
  62. Wolfe DA, Scott K, Wekerle C, Pittman A. Child maltreatment: Risk of adjustment problems and dating violence in adolescence. Journal of the American Academy of Child & Adolescent Psychiatry. 2001;40(3):282–289. doi: 10.1097/00004583-200103000-00007. [DOI] [PubMed] [Google Scholar]
  63. Zurbriggen EL. Understanding and preventing adolescent dating violence: The importance of developmental, sociocultural, and gendered perspectives. Psychology of Women Quarterly. 2009;33(1):30–33. [Google Scholar]

RESOURCES