Bullying in Different Contexts

6 Juvenile dating and violence

Within the complex social world that young people inhabit, in addition to traditional forms of bullying, we find other negative forms of relations which share some characteristics with bullying in terms of dominance and submission. This is true for ‘dating violence’, which we discuss in this chapter. We explore individual factors in relation to involvement in dating violence (both aggressors and victims) and the way that the peer network can ‘cover up’ the problem. We address the roles of affective-emotional links and the appearance of interpersonal violence within the network, as well as the inclusion of negative models of couples. We propose that the appearance of dominance–submission–control–violence within the first romantic relationships during adolescence may originate in the behaviours and attitudes of the peer-group. Early data on this topic (Ortega et al., 2008b) suggest that forms of indirect relational violence may appear within dating relationships and are aimed at controlling the behaviour, attitudes and feelings of the partner, and are difficult to detect and difficult to stop. Sexual violence, on the other hand, is worryingly more present during these first dating relationships than we thought.

The importance of the first romantic relationships in adolescence

The biological changes that accompany adolescence play an important role in the changes which occur in the interpersonal relationships of adolescents. With sexual maturity, changes in the levels of hormones and the development of secondary sexual characteristics, adolescents begin to feel sexual attraction for others.

The majority of adolescent couples develop within the larger peer network. Adolescent groups gradually move from being sex-segregated to becoming mixed sex (Maccoby, 1998) and social groups become less centred on play and start to become places where adolescents can express or experiment with these new dimensions of their lives and identity in leisure activities. In this sense, adolescents need to develop ways of managing and expressing their erotic-sexual emotions in accordance with the norms and social conventions of their culture. Dating involves a continuous process of adjustments of desires, attitudes and sexual and emotional behaviours.

Research describes the importance that these early romantic relationships have for adolescents. Approximately 25 per cent of adolescents between 11 and 13 years state that they have had a girlfriend/boyfriend, whilst in later adolescence we find that only one in four have had no experience of dating (Collins, 2003; Furman, 1999; Menesini and Nocentini, 2008; Muñoz-Rivas et al., 2007a). In our recent research (Sánchez et al., 2008) we have found lower levels: only 8 per cent of adolescents between 17 and 20 years had no experience of dating. To have a girlfriend/boyfriend during adolescence appears to impact positively on identity and satisfies the desire and need for intimacy and an emotional and sexual bond with another person (Furman and Shaffer, 2003).

Considering the importance and value of these first romantic relationships, research in this area has not always explored them from a psycho-social or developmental perspective. Many studies on dating have focused on the psycho-social risks for adolescents, particularly on the risk of sexual relationships resulting in teenage pregnancies and early parenthood. It is not surprising therefore that, from this standpoint, it is concluded that the motivations that guide adolescents in search of a partner are sexual, particularly for boys (Hofstede, 1998). However, research conducted in the USA has found that the most important reasons that guide boys to become involved in romantic relationships are emotional and relational (Smiler, 2008), rather than sexual. Furthermore, many studies have confirmed that erotic-sexual relationships tend to strengthen in the long term into the construction of romantic relationships (Carver et al., 2003; Rice, 1990).

In conclusion, research indicates that the romantic relationships formed by adolescents are both an important developmental landmark and an important social context for adolescents. This does not indicate that both dimensions – the individual socio-emotional factors and the dyadic context – are free of difficulties. Adolescents report that one of the main disadvantages of having a boyfriend/girlfriend is the large number of conflicts and discussions that result from the relationship (Overbeek et al., 2007). In addition, research over the past 20 years has found that the adolescent dating relationship can present a particularly violent social context; more so than among adult couples (Jackson et al., 2000). In the following sections we focus on examining this worrying aspect of adolescent relationships further.

Violence in adolescent couples: studies on dating violence

The first consideration to keep in mind when examining violence among adolescent couples is that, historically, research in this area has been strongly influenced by research on domestic violence at both the theoretical and empirical levels (González and Santana, 2001; Magdol et al., 1997; Makepeace, 1981). In particular, this legacy is seen in the instruments used by researchers which come from traditional explorations of gender violence.

The high prevalence rates of aggression and other violent behaviours that are found among adult couples are also found among adolescent couples. However, it may be that researchers have included behaviours in the same category of offensiveness or criminality among adolescents, which may not be similar, because the adolescent couples’ relationships are more short-lived. It is also important to consider the ways in which these behaviours are interpreted and accepted (or not) by those involved. Behaviours which are considered as offensive or harassing by someone who is not attracted to the other may be interpreted by someone who is interested as the initiation of dating or vice versa. The analysis of violence within the adolescent couple may be more complex, as it is important to note the developmental dimension and immaturity of young people in relation to dating. This means that many of the patterns of interaction among young couples are still not clearly defined.

In this chapter we examine violence, more or less persistent, which takes place among adolescent couples (dating violence). Forms of dating violence vary from verbal violence (insults) to psychological violence (threats of different types, including emotional blackmail), to physical violence (pushes, slaps, punches, beatings) and sexual violence (from pressure to have sexual contact to rape). Less severe forms of violence are surprisingly frequent, whilst the most serious forms are less frequent, although both form part of the lives of some adolescent couples.

To date, most research has been conducted in the USA and Canada, and only more recently in European countries. Since 1957, when Kanin conducted the first study with university students (noting that 62 per cent of first-year students said that they had suffered aggression at the hands of their partner in the last year), there have been only a moderate number of studies on this topic. However, there has been an increase in this in recent years.

The prevalence rates in studies conducted in Europe and North America have not been less alarming and controversial, with indices which have varied between 20 and 60 per cent (Archer, 2000; Chase et al., 1998; Fernández-Fuertes and Fuertes Martín, 2005; Hird, 2000; Lewis and Fremouw, 2000; Menesini and Nocentini, 2008; Moffit et al., 2002; Muñoz-Rivas et al., 2007a; Sánchez et al., 2008). More recently, Muñoz-Rivas et al. (2007a) found that around 90 per cent of students aged 16–20 years had, at some point, verbally abused their partner, whilst 40 per cent had been physically violent towards them. The most severe forms of physical violence (trying to choke, beat up, threaten with a knife or other weapon) were perpetrated by 4.6 per cent of males and 2 per cent of females.

Many researchers have analysed the different prevalence rates found by studies in methodological terms. Researchers have suggested that the different definitions of dating violence used, the type of instrument employed and the age and sex of the participants may have an effect on the results obtained. Additionally, the criteria used to estimate the global level of violence (only one or several indices of violence; only considering frequent and persistent violence or all of those who reported the behaviour regardless of persistence), the temporal criteria used as a reference (during the last year or the last few months) and the consideration of the experience of dating violence ever or the presence of dating violence in the current dating relationship (Archer, 2000; Lewis and Fremouw, 2000; Menesini and Nocentini, 2008; O’Keefe, 2005; Shorey et al., 2008) may influence the prevalence rates recorded.

Without doubt, sex differences have been the most studied and the most controversial aspect of research on dating violence. Some studies have found no difference between males and females in relation to aggression (Brendgen et al., 2002; Moffit et al., 2002); others have found higher prevalence rates among females (Archer, 2000; Sánchez et al., 2008); whilst in Italy, the rates of aggression have been higher among males (Menesini and Nocentini, 2008). These contradictory results are maintained when the different types of violence are examined separately. Research in Spain has concluded that males use more serious physical aggression than females, whilst females use more verbal, relational or psychological forms (Muñoz-Rivas et al., 2007a; Sánchez et al., 2008). In terms of less serious physical forms, the results are less conclusive, with some studies reporting that females are more likely than males to use these forms of violence, whereas others report similar levels across genders. Archer’s (2000) meta-analysis clarified this data by illustrating that girls/young women were more likely than boys/young men to physically assault their partners between the ages of 14 and 22 years, whilst between the ages of 23 and 29 years, men were more likely than women to use physical aggression against their partner.

The results are more conclusive in relation to sexual violence. Several studies have shown that adolescent males are more likely to sexually assault their partners (Fernández-Fuertes and Fuertes-Martín, 2005; Menesini and Nocentini, 2008) or their peers (Ortega et al., 2008b). In addition, females report that they suffer more sexual aggression and harassment than males (Bennett and Fineran, 1998; Foshee, 1996; O’Keefe, 2005).

Age is another variable which has been of interest in research on dating violence. However, as has been noted earlier, very few studies have analysed violence from a developmental perspective across adolescence and during the first years of adult life. An important consideration is that most studies which have looked at the effect of age on violence in couples have used samples of young adults, approximately 20 years old or more, and have concentrated on the analysis of violence against the female partner (Capaldi and Kim, 2002; O’Leary, 1999). In this respect, Archer (2000) found that the development of aggression was different among younger and older samples in relation to sex: before the age of 22, females presented higher levels of aggression, whereas after this age, males were more aggressive. Capaldi and Kim (2002) clarified the conclusions of Archer by considering the stability of the couple’s relationship as a mediating variable in the development of violence by young men towards their female partners. They found that there was a drop in violence of males towards their female partners between the ages of 20 and 27 years. When examining those couples whose relationship was relatively stable, the pattern was slightly different. There was still a decline between 20 and 25 years, but violence then increased after the age of 25. However, when there was lower stability in the couple’s relationship, there was an increase in male violence between the ages of 20 and 25 years which then began to decrease after this age. The authors explain this peak in violence at 25 years in part by the higher level of involvement in romantic relationships at this age in comparison to earlier ages (Capaldi et al., 2005).

Most research in this area has been cross-sectional, and only recently have longitudinal studies been carried out. This means that currently little is known about the development of dating violence during early and mid-adolescence. Nocentini (2008) conducted one of the first longitudinal studies of the development of physical violence in adolescent couples between the ages of 16 and 18 years. Nocentini concluded that, controlling for couple stability, there was a decrease in violence between 16 and 18 years, and that girls began to decrease their levels of physical aggression towards partners earlier than boys.

One area of recent interest among researchers is the prevalence of dating violence among gay, lesbian and bisexual adolescents (GLBA). Results are mixed, with some studies showing that dating victimisation among GLBA couples is as frequent as among heterosexual couples (e.g. Freedner et al., 2002); some showing lower (e.g. Halpern et al., 2004) and others higher (e.g. Trish et al., 2003) rates. Regarding specific sexual minorities, studies seem to conclude that bisexual adolescents and young adults are more likely to suffer any type of dating violence in comparison to lesbian and gay adolescents (Freedner et al., 2002; Moore and Waterman, 1999), and that girls in same-sex relationships suffer more psychological violence and physical violence than boys.

One of the main conclusions of these studies is the methodological considerations to bear in mind in order to understand this disparity in the results. For example, one important issue is the sample size and the difficulty in recruiting representative samples of GLBA (Freedner et al., 2002). Another methodological issue is the consideration of sexual behaviour or sexual identity in the definition of GLBA (Savin-Williams, 1995). For example, Trish et al. (2003) found that 40 per cent of adolescents in their study were still questioning their sexual orientation, whilst just 9 per cent self-identified as gay or lesbian. In this respect, Freedner et al. (2002) underline the important effect of the sex of the partner when examining dating violence among GLBA, because many of them may be in relationships with someone of the opposite sex (i.e. have not ‘come out’). In conclusion, GLBA dating violence rates are very worrying, or at least as worrying as those among heterosexual adolescent couples.

Dating violence and the quality of romantic relationships

One of the most conclusive findings on this topic has been that there is a higher level of violence in more stable relationships, although that is not to say that they are couples who are more satisfied with their relationship (Furman and Buhrmester, 1992; Menesini and Nocentini, 2008; Shulman and Scharf, 2000). It is necessary to highlight the importance of relationship satisfaction and the functioning of the couple in trying to understand dating violence. Studies demonstrate that couples characterised by higher levels of violence also present higher levels of conflict, a greater imbalance of power within the couple, more discussions and less satisfaction in general with the relationship (Bookwala et al., 1994; Sánchez et al., 2008).

We have recently examined this relationship between quality and violence. Ortega et al. (2008b) conducted a study with 524 adolescents between the ages of 15 and 20 years, and found that adolescents with a current or previous partner (within the last six months) whose relationship involved aggressive behaviours with their partner felt that there was a lot of conflict and inequality of power within their relationship. However, paradoxically, they also stated that they felt very satisfied and had positive expectations for the future of their relationship, even more so than those who were not involved in dating violence. The fact that these adolescents had relationships with partners which were stable over time (one year or more) appears to suggest that, for some adolescents, the benefits of the emotional-relational dynamic of the relationship appear to be contaminated by violence and conflict, although this does not make them think about breaking up or changing the nature of the relationship. We have called this perception of intimate relationships ‘dirty dating’ (Ortega et al., 2008b). Although we are still examining this phenomenon, we hypothesise that a passive acceptance that being part of a couple includes a certain inequality of power, toleration of aggressive conflicts and direct violence may be at the heart of young people’s understanding of these relationships.

One of the most relevant findings from research is the large number of adolescents who assault their partners and are assaulted by them at the same time. Gray and Foshee (1997) found that 66 per cent of students in their study were involved in mutual violence, results that have been confirmed by later studies in other countries (Capaldi et al., 2004; Hird, 2000; Menesini and Nocentini, 2008; Whitaker et al., 2007). These studies confirm that, in more adolescent couples than previously expected, there exists a violent and reciprocal dynamic which includes large inequalities of power and dominance of one over the other.

Theories and hypotheses for dating violence in adolescence

As Shorey et al. (2008) have underlined, little attention has been paid to the development of theories that could help to understand these behaviours among adolescents. To date, there is no clear theoretical model which can successfully explain the large variety of types of aggression which are found within this context and this age group.

Traditionally, four theoretical models have been proposed regarding dating violence (Shorey et al., 2008; Werkele and Wolfe, 1999): social learning theory, attachment theory, feminist theory and coercion/conflict theory. Recently new research has used integrative approaches to the four models, presenting very interesting results which contribute to the research on this topic.

Attachment theory

Attachment theory explains the unique and powerful nature of the different forms of emotional attachment in terms of the ways in which they impact on the individual’s life. From secure and stable attachments, which are flexible and well developed, to those which are insecure, unstable, rigid or scarcely operating, one can describe a continuum of attachment which relates to general psychological adjustment and the presence/absence of personal and social psychological problems. According to the theory developed by Bowlby (1969), in the first years of life and through the formation of emotional relationships with parents or caregivers, children develop general styles of relating with others (attachment styles). Through these relationships they develop internal mental representations, which are then used as a basis for later relationships. Applied to problems of dating violence, attachment theory suggests that victims and aggressors are those individuals whose attachments with their primary caregivers are based on a lack of trust, on the continuum of dominance–submission and emotional control. Research conducted in this area has confirmed that boys and girls who have insecure attachments come from families in which they have been mistreated; they present a higher risk of becoming involved in dating violence (Werkele and Wolfe, 1998). Nevertheless, other studies have questioned whether there is a direct link between attachment and dating violence, as it has been found that a significant number of adolescents who have secure attachments assault their partners (Schwartz et al., 2006).

Social learning theory

Bandura (1977) postulated that human beings learn by exposure to and imitation of the behaviour of others. Applying this to violence, repeated exposure to violent contexts within life, as a witness or direct recipient, could provide the basis for a later expression of violence in different contexts and situations. Learning violent behaviour within the family setting could explain the expression of violence in other contexts, such as the dating relationship (O’Leary, 1988). A number of studies have supported this model, some reporting that being exposed to family violence during infancy increases the risk of behaving aggressively with a partner (González and Santana, 2001; Linder and Collins, 2005). Others have emphasised the moderating role of other variables, such as the quality of the couple’s relationship or individual personality. So, the regulation and expression of anger may be influenced by the intergenerational transmission of violence. Wolfe and Foshee (2003) demonstrated that the way in which an individual learns to express anger as a result of his or her family context is related to the way in which anger is expressed in adolescence. The authors classified adolescents in relation to the different styles they used to express anger, finding that the destructive style, characterised by aggression towards the person who is the object of the anger, was directly related to aggression within the adolescent dating relationship.

One of the most important contributions of social learning theory is that it can explain the transmission of aggression and violence across contexts. Connolly et al. (2000) and Pepler et al. (2006) analysed the predictive effect of bullying among peers on dating violence from a developmental perspective, taking into account the predictive or moderating effect of the physical and biological changes that accompany adolescence. The authors considered that bullying is a behaviour that is aimed at obtaining power and domination over others, a tendency which aggressive individuals then take to the new forms of relationships which appear during adolescence, i.e. early dating relationships. The authors interviewed 479 students between the ages of 13 and 19 years, using measures of direct involvement in bullying, being the recipient or perpetrator of sexual harassment, involvement in dating violence and changes related to puberty. The authors concluded that, whilst the prevalence of bullying decreased across adolescence, the presence of sexual harassment and dating violence increased. The appearance of these new types of violence also coincided with changes related to puberty, with those who passed puberty early becoming involved in these new types of aggression earlier on as well. In spite of the different trajectories of bullying, sexual harassment and dating violence, the results showed positive correlations between the three forms of aggression among peers, demonstrating, therefore, the important risk factor that involvement in bullying during primary and secondary school poses for the personal and social adjustment of adolescents.

Work by Ortega and colleagues (Ortega and Mora-Merchán, 2008; Ortega et al., 2008a) has also analysed the transmission of violence across contexts, within the peer-group, from the peer-group, to the couple. The authors examined the key role that dominance–submission within the relationship plays in the dynamic of violence and how this is maintained and repeated in different relational contexts. To date, the authors have found important correlations between sexual harassment among peers, and within the couple, and weak correlations for victimisation in these two contexts. The correlation between sexually harassing others and being sexually harassed in different contexts was weak. Nevertheless, highly significant correlations appeared between aggression and victimisation in the couple scenario, which suggests the important effect of the couple as an individual and specific context in the appearance and maintenance of violence.

Feminist theory

Feminist theory (Walker, 1989) suggests that violence within couples is the expression of an inequality of social power which exists between men and women as a consequence of the patriarchal society in the majority of Western and non-Western societies. This imbalance of power places men in a dominant position and women in a more submissive position, placing the latter at higher risk of suffering violence at the hands of their partners. Various factors support the feminist explanation of violence within young couples: the desire for control and the maintenance of power by men over women (Barnett et al., 1997; González and Santana, 2001), as well as rigid sexist attitudes about male and female roles based on misogyny and blaming women for violence. Recent research has found these beliefs to still be present among young people, including university students (Ferrer et al., 2006), and to be particularly representative and defining of the personalities of those who are violent towards their partners (Lichter and McClosey, 2004).

From this perspective, aggression by girls and women towards their male partners is analysed in terms of self-defence rather than being motivated by a desire for control or power. Nevertheless, a vision of women solely being reactive is not fully supported, as girls are equally (Moffit et al., 2002), if not more (Muñoz-Rivas et al., 2007b), likely to start the aggression. Similarly, other studies have found that among the motivations used by girls/women to assault their partner are those which are personal and based on the established dynamics in the intimate relationship more than self-defence (Stuart et al., 2006) or intense emotional states such as momentary anger or rage (Muñoz-Rivas et al., 2007b).

Conflict or coercion theory

Conflict or coercion theory locates the origin of dating and domestic violence in much more proximal factors, in particular the conflict dynamics which are found among some adolescent and adult couples. From this point of view, relationships where coercion and control are used as a way of resolving conflicts would be where we would be more likely to find aggression between the couple (Feld and Straus, 1989). It is therefore a perspective that considers reciprocal conflict and violence as an interactional style used by the couple, which begins in adolescence and becomes more established and perpetuates itself over time. This approach does not focus exclusively on the individual characteristics of each member of the couple, but on the styles of interaction that occur between them.

Gray and Foshee (1997) have confirmed this perspective, noting that these couples are also at higher risk for conflicts, discussions and dissatisfaction with their partner. Werkele and Wolfe (1999) confirmed that these couples have serious difficulties in the use of prosocial strategies and finding solutions. They also have significant problems in emotion regulation (Cummings and Davies, 1996), which in some way reinforces the use of coercive and aggressive strategies in the day-to-day life of the couple. This also increases the likelihood that they will continue to use aggression within their relationship and that it will become a habitual form of interaction (Feld and Strauss, 1989; Werkele and Wolfe, 1999). One of the aspects that has been most studied in this area has been the regulation of anger, which in many studies has been found to be directly related to aggression in adolescent and adult couples (Muñoz-Rivas et al., 2007b; Nocentini, 2008; Riggs and O’Leary, 1989).

Nevertheless, Katz et al. (2000) concluded that, although conflict resolution strategies which involve coercion and control are found within couples where violence is a problem, they are also found within couples where violence does not occur.

The dynamic developmental systems approach

Capaldi et al. (2003) and O’Leary and Slep (2003) have noted the importance of the development of models which prioritise the study of the dyad rather than models which have been centred on the study of the individual. From this, Capaldi et al. (2004) and Capaldi and Kim (2007) have developed a theoretical model which attempts to explain the behaviour of the couple as a dynamic developmental system. In this sense, the behavioural dynamic which is established by the couple is understood as an intrinsically interactive system but, at times, defined by the individual developmental characteristics of both members of the couple and by contextual factors which affect each member of the couple.

The contribution of this model rests in the possibility of considering the different aspects which are involved in the explanation of violence in couples’ relationships. These are aspects which relate to the biology of those involved, their individual characteristics, contextual factors and their social experiences in the different micro-contexts in which they participate, principally family and peers (Capaldi et al., 2005). The way in which these aspects develop over time will also form part of the process under investigation; the characteristics of each member of the couple are a product of their development and include biological aspects as well as socialisation, and therefore evolve and change over time.

According to this model, three main factors are important in understanding the origin and development of violence among adolescent couples:

• The individual characteristics of the couple which may affect their relationship, such as aspects of personality, psychopathological characteristics and those characteristics which were learnt and reinforced within the family and peer contexts.

• The contextual risk factors which could be influential in increasing the likelihood of aggressive episodes towards the partner, such as the use of alcohol or drugs.

• The nature of the relationship of the couple, the patterns of interaction which are established in the couple and the way in which these evolve and develop over time.

From this point of view, they consider the interaction styles and the aggressive dynamics which are established within the relationship as being predictive factors and factors which may reinforce violence within these early dating relationships (Capaldi and Kim, 2007; Capaldi et al., 2004).

Capaldi’s approach, which is developmental, multi-factorial, multi-probabilistic and integrates diverse perspectives, allows for a much richer and comprehensive view, not only of violence during adolescence and adulthood, but also of its development and stability over time. Research by Capaldi et al. (2003) found that some young people who were involved in risky relationships in which physical violence occurred were more likely to continue to be violent (male partners only) if two years later the couple were still together, whilst this probability decreased if the couple had separated and they had started dating other people. Nocentini (2008) found a decrease in physical aggression in adolescent couples from mid- to late adolescence. This decrease was affected by factors related to certain family characteristics and individual experiences, as well as individual factors and factors from within the dyadic relationship such as antisocial behaviour and the perception of victimisation by the partner. In sum, the study concluded that low educational level of mother and the number of previous partners the adolescent had had placed them at higher risk of being aggressive at age 16 compared with the rest of those studied. On the other hand, the use of physical aggression towards a partner in addition to exhibiting other forms of antisocial behaviour and with the perception of being victimised by a partner delayed the tendency for the decrease in physical aggression between mid- and late adolescence. These were the adolescents, according to the study, who were most vulnerable and at the greatest risk.

Risk factors in the appearance and maintenance of dating violence

In order to evaluate the risk factors and to provide intervention or prevention programmes for dating violence, it is important that attention is paid to a large body of research. In this sense, multi-dimensional and probabilistic perspectives seem to be the most appropriate to understand the appearance of dating violence as well as the conditions in which these problems tend to repeat themselves (Cáceres and Cáceres, 2006; Lewis and Fremouw, 2000).

Many of the risk factors identified for dating violence come directly from research carried out from the different perspectives and theoretical models mentioned earlier. Some researchers have focused on family experiences, social contexts (particularly peer contexts), sociodemographic factors, individual difference factors (such as certain personality factors) and the co-occurrence of other risk factors (such as smoking, drinking alcohol or drug taking).

Research into the influence of sociodemographic factors in the prediction of dating violence has not yet produced many conclusive results. Traditionally, the variables studied have been family composition, socioeconomic status (SES), place of residence, ethnicity and gender. Some studies have indicated that boys who are from a low SES background are more at risk of involvement in violence (Makepeace, 1987; O’Keefe, 1998), although for many authors these results are not entirely conclusive (O’Keefe, 2005). In relation to the area of residence, research appears to conclude that violence among adolescent couples occurs independently of the zone of residence. Some studies have found that there is more violence in urban areas (Makepeace, 1987) whilst others have found that there are higher levels of dating violence in rural areas (Reuterman and Burcky, 1989). Several different studies have shown that adolescents from single-parent families are more at risk of being involved in dating violence (Foshee et al., 2008) and that low levels of parental education increase the risk of being involved in dating violence, especially when mediated by sexist attitudes or family violence (Foshee et al., 2008).

With respect to family factors, parents who show neglectful educational styles and whose relationships with their children are characterised by an abuse of power place their children at an increased risk of becoming involved in violent behaviour during adolescence. A study conducted by Straus and Savage (2005), with university students from 17 countries, found that neglect and a lack of material care (e.g. lack of hygiene, lack of help or support with school tasks) by parents during childhood were related to dating violence during adolescence. Boys who had received poor levels of care from their parents during infancy and childhood were more at risk of physically assaulting their partner during adolescence, a probability which increased exponentially if the social context in which they lived was violent.

In the same vein, O’Keefe (1998) found that experiencing physical abuse during infancy significantly predicted victimisation of girls in dating relationships in adolescence. However, this did not predict victimisation or aggression for boys. Kinsfogel and Grych (2004) found that young people who had witnessed high levels of violence in the family home were more at risk of becoming a victim and aggressor of their partner later on. However, the adolescent’s beliefs and attitudes about violence emerged as a mediating factor. The authors suggest that some young people who have witnessed violence may have learnt that this is justifiable within dating relationships. However, this result was only significant for boys and not girls.

The peer context and experiences with peers have also been an important focus of research on dating violence. Various studies have concluded that being violent with peers predicts violent behaviour with a partner (Capaldi et al., 2001), particularly for boys (Chase et al., 1998). Additionally, having violent friends who have positive attitudes towards violence predicts aggressive behaviour towards a partner in adolescence (Arriaga and Foshee, 2004; Connolly et al., 2000). Arriaga and Foshee (2004) found that witnessing dating violence among peers and interparental violence significantly predicted violence within their own relationship, either as an aggressor or victim. When looking at the relative strengths of the predictive relationships, they found that dating violence by peers was a stronger predictor than interparental violence. In general, many studies have confirmed that proximal factors, such as peers, the consumption of substances and the context of the dating relationship, are more important than distal factors, but that these distal factors influence dating violence in an indirect way, often mediated by proximal factors (Foshee et al., 2008; Magdol et al., 1997; O’Keefe, 2005).

Different personality variables and interpersonal abilities have been found to be fundamental in the explanation of dating violence. Low self-esteem is a strong indicator of dating aggression for boys (O’Keefe, 1998), whereas for girls it is more strongly related to victimisation. In this respect, some authors reflected on the important influence of self-esteem in the normalisation of violence in the context of the couple and in the explanation of the complex process of victimisation (Lewis and Fremouw, 2000).

Another important psychological factor is conflict resolution style. As mentioned earlier, from the perspective of coercion and conflict theories, several studies have described how some young people present serious difficulties regulating anger and frustration when faced with conflict (Muñoz-Rivas et al., 2007a; Werkele and Wolfe, 1999) and are aggressive, impulsive and tend to attribute guilt to their partner, which reinforces this aversive relational context and results in an increase in conflicts and discussions. As we have noted, adolescents who are involved in dating violence find themselves in relationships which are violent and aggressive, with high levels of conflict and reciprocal and mutual violence. In this respect, research has demonstrated that one important predictor of aggression in the adolescent couple is aggression and violence by the partner, at least for physical violence (Capaldi et al., 2003; O’Leary and Slep, 2003). Therefore, it is probable that those adolescents who lack skills in conflict resolution and the regulation of stress are more likely to form relationships with other violent individuals who have similar patterns of conflict resolution. This leads to a relational dynamic in which conflicts escalate and may lead to violence.

Other influential factors are the beliefs held by the individual regarding their own ability to resolve conflicts and their feelings of self-efficacy in regulating their own emotions. Nocentini (2008) carried out one of the first studies on the impact of the perception of self-control of aggression towards a partner on actual physical and psychological dating violence. The research concluded that the perception of self-efficacy in the self-regulation of emotions in intimate relationships was related to physical and psychological aggression via rumination and conflict. Adolescents who thought that they could not exercise any control over their anger and rage in the face of conflicts with their partner were more likely to use rumination, increasing the desire to extract revenge or hurt their partner. In consequence, it was more likely that conflicts would not be resolved and would in fact worsen, involving the use of negative strategies for conflict resolution such as physical and psychological aggression.

Finally, it is important to restate the significance of other proximal factors, such as the presence or co-ocurrence of other behaviour problems during adolescence. The use of alcohol and other substances has repeatedly been found to be one of the most important predictive factors for delinquent behaviour, as well as dating violence (O’Keefe, 1998; Silverman et al., 2001). However, some studies have not found this relationship (Loh and Gidycz, 2006) and others report that it is an important predictor of sexual aggression by males (Casey et al., 2009) and sexual aggression by males towards their partner (Koss and Dinero, 1989). It is suggested that this may occur because alcohol and other drugs may diminish control over behaviour and distort the perception of the situation and the other person’s intentions. Risky sexual behaviour, such as having numerous partners at an early age, has also been found to be a predictive factor of dating violence (Makepeace, 1987; Nocentini, 2008; O’Keefe, 2005; Werkele and Wolfe, 1999).

We do not want to finish this section without considering that these variables are not cause–effect variables, but are dimensions which increase or decrease the probability of an individual becoming involved in dating violence. It is important to underline the summative and/or multiplicative interactions of the co-occurrence of these factors in the explanation and prediction of dating violence. In synthesis, we can say that it is a multi-determined phenomenon, affected by diverse proximal and distal factors and it is the analysis of these factors which may help in the design of prevention and intervention programmes.

Implications for intervention and prevention

The development and implementation of intervention and prevention programmes is only relatively recent and is derived directly from the studies of risk factors for dating violence. Interventions have focused on communication strategies and conflict resolution orientated towards the constructive resolution of problems and increasing awareness of dating violence and its consequences.

Most prevention and intervention programmes have been developed from and within the educational context (Avery-Leaf et al., 1997; Lavoie et al., 1995) and have been primary intervention programmes. Knowledge of their effectiveness is still controversial, as a result of methodological problems which have limited the generalisability of the results. On many occasions these programmes have been implemented with a small number of adolescents; in others the intervention period has been very short or the pre–post evaluation has been carried within a limited timeframe and without the analysis of a comparison control group (Werkele and Wolfe, 1999; Whitaker et al., 2006). However, we describe here two programmes which have been implemented and evaluated satisfactorily, whose design has tried to overcome those methodological problems highlighted and which have presented promising results in changing the attitudes and beliefs of young people, as well as in reducing violent behaviour.

The Safe Date Project (Foshee et al., 1998, 2004) is a community intervention programme which involved 14 schools from North Carolina and was conducted with adolescents aged between 14 and 15 years. The programme focused on challenging sexist attitudes, increasing knowledge about dating violence, developing conflict resolution strategies and the establishment of networks of support and help for those involved in dating violence. The students involved in the study were randomly assigned to a control group or the experimental group, and the evaluation of the effectiveness of the intervention was conducted one month post-intervention, one year later and two to three years later. The focus of the evaluation was on the change in sexist attitudes and attitudes towards violence, as well as the reduction of physical, relational or sexual violence in the dating relationships. The results were mixed. On one hand, the one-month post-test assessment revealed a significant decrease in aggression towards partners, as well as higher levels of knowledge and sensitivity and less tolerance towards this type of violence. However, results relating to a decrease in violent behaviour were not found in the follow-up one year later. Further, a follow-up several years later concluded that the decrease in dating violence was maintained up to four years after the end of the programme (Foshee et al., 2004).

The Youth Relationships Project (Wolfe et al., 2003) is a community project whose results have also been promising. This programme focused on those young adolescents who were at particular risk, specifically those who had a history of family abuse. The young people who took part in the programme received group sessions aimed at improving their conflict resolution skills and emotional well-being and communication skills, as well as their knowledge about abusive relationships, and aimed at decreasing their tolerance towards these. In spite of the fact that more than 160 young people took part in the programme, the experimental design continuously assessed the efficacy of the intervention sessions. The results, although not entirely conclusive, reported reductions in the rates of physical aggression among the participants in this programme.

In spite of the fact that the results of the programmes mentioned above are positive and promising, it remains to be seen whether these interventions will be able to reach those who need them most. As we have mentioned, the dating violence programmes developed have been focused on primary prevention, and few have carried out systematic interventions with at-risk populations. It is important to state that, until now, researchers and theorists have paid little attention to the at-risk populations who are in most need of these interventions.