1. Introduction

Adolescent and young couple relationships pose problems that sometimes lead to serious physical and psycho-logical health consequences (^{Alarcón et al., 2018}; ^{Teten et al., 2009}). Intimate partner violence is a worldwide social and public health problem (^{Ferrer-Pérez & Bosch-Fiol, 2019}; ^{García-Díaz et al., 2018}; ^{Hébert et al., 2017}; ^{Heyman et al., 2018}; ^{Rodríguez-Franco et al., 2017}). It is a heterogeneous, universal, and growing phenomenon in society (^{Muñoz & Echeburúa, 2016}; ^{Rodríguez-Biezma, 2007}), in which violent psychological, physical, or sexual behaviors occur (^{McLaughlin et al., 2012}; ^{World Health Organization, 2018}) at any time of the relationship, increasingly at earlier ages (^{Loinaz et al., 2011}) and in both sexes (^{Rodríguez-Biezma, 2007}).

According to the last results of the Spanish National Statistics Institute (^{National Statistics Institute, 2019}), the number of young people reported for gender violence, compared with the previous year, increased by 14.1% at ages 18 to 19 and by 11.9% at ages 25 to 29. Therefore, studies are focusing on the youngest couples because, in this stage, risky behaviors prevail (^{Borrás et al., 2017}) and, in consequence, violent dynamics increase and start to be considered acceptable (^{Kidman & Kohler, 2020}). Dating violence develops differently than adult intimate partner violence. Most of the relationships occur during youth, including the first dating experiences, which have an impact on the development and well-being of individuals and, unfortunately, constitute a clear indicator of intimate partner violence in the adult stage (^{Exner-Cortnes, 2014}; ^{López-Cepero et al., 2014}; ^{Van de Bongardt et al., 2015}), since an early dysfunctional relationship could have negative consequences for the health of individuals (^{Exner-Cortnes, 2014}; ^{Shorey et al., 2012}). Thus, adolescence and youth are considered risk factors for relationships, given the prevalence of violence in these stages, which is even higher than in marital relationships (^{Rubio-Garay et al., 2019}).

It is worth noting that, when studying dating violence, there may be differences between adolescent couples and young couples. In fact, numerous studies focus on the violence that occurs in both types of couples (^{Cortés-Ayala et al., 2015}; ^{de la Villa et al., 2017}; ^{Pazos et al., 2014}; ^{Rubio-Garay et al., 2019}); in young couples only (^{García-Carpintero et al., 2018}; ^{López-Cepero et al., 2015}); and in adolescent couples only (^{Aizpitarte et al., 2017}; Aizpitarte & ^{Rojas-Solís, 2019}; ^{Zamora-Damián et al., 2018}). In this sense, the relation between age and educational level should be taken into account; namely, as a general rule, adolescents are in the final phase of junior high school and in senior high school, and young people pursue higher education (college) courses. In addition, it is clear that the level of maturity is not the same at the beginning of adolescence as that at the end of youth. Some authors suggest a higher rate of violence in adolescents and a reduction as they grow older (^{Jackson et al., 2000}); however, other authors find no difference in this regard (^{Rubio-Garay et al., 2019}). Even so, this problem exhibits many similarities in terms of violence.

Currently, several studies related to violence in young dating relationships state that men and women equally perpetrate violence. (^{Alegría & Rodríguez, 2015}; ^{Archer, 2000}; ^{Arnoso et al., 2017}; ^{Chen & Chan, 2019}; ^{Graña & Cuenca, 2014}; ^{Rojas-Solís et al., 2017}; ^{Straus, 2009}). In Spain, judicial evidence shows that men usually exercise this phenomenon; nevertheless, community samples indicate that both sexes have the same predisposition to perpetrate violence (^{Graña & Cuenca, 2014}). Further-more, bidirectional violence (in which both partners are victim and aggressor in the same relationship) is more frequent (9%) than unidirectional violence (‘’2%) (^{Alegría & Rodríguez, 2017}; ^{Archer, 2000}; ^{Arnoso et al., 2017}; ^{Graña & Cuenca, 2014}; ^{W. L. Johnson et al., 2015}; ^{Melander et al., 2010}; ^{Rojas-Solís et al., 2017}; ^{Rubio-Garay et al., 2019}; ^{Rubio-Garay et al., 2012}; ^{Straus, 2009}; ^{Zamora-Damián et al., 2018}). In relationships, young people state that they have been mostly victims and aggressors at the same time (81.9% boys and 93% girls) rather than only aggressors (3.6% boys and 3.7% girls) (^{Rodríguez, 2015}). Bidirectionality prevails in psycho-logical aggressions by 80% compared to 25% of physical violence (^{Graña & Cuenca, 2014}). Likewise, some findings reveal that women mainly perpetrate mild psychological and physical violence, but men, mostly physical violence (^{Chen & Chan, 2019}; Graña & Cuenca, 2014; M. P. Johnson, 2011; ^{Muñoz-Rivas et al., 2007}; ^{Rubio-Garay et al., 2017}; ^{Rubio-Garay et al., 2012}; ^{Straus, 2008}).

Based on the above, both partners exercise and experience aggressions to similar extents (^{Riesgo González et al., 2019}), although women are more frequently perceived as victims if the seriousness of the attacks is considered. Consequently, the use of inappropriate strategies for conflict resolution (poor communication, confrontations, reproaches, and criticism) is normalized among young people, exacerbating the levels of violence. (^{Lewis & Fremouw, 2001}; ^{Muñoz & Echeburúa, 2016}; ^{Rubio-Garay et al., 2019}; ^{Rubio-Garay et al., 2012}; ^{Trujano et al., 2010}).

M. P. ^{Johnson (2006}, ²⁰⁰⁸, ²⁰¹¹) calls situational violence the type of violence that is visible in an episodic or reactive way during conflict contexts in which the two partners enter a spiral of violence. This is the most common form of intimate partner violence (^{Muñoz & Echeburúa, 2016}); men and women exercise it at equal rates, with no intention of controlling or coercing the partner and rarely causing harm. Unlike intimate terrorism or coercive controlling violence, situational violence consists of situations of tension in which one or both partners resort to violent aggressions (^{Ferrer-Pérez & Bosch-Fiol, 2019}).

Studies available on conflicts in adolescent and young couple relationships should have a real impact on society by contributing to it. Thus, this research is conducted with the purpose of formulating and solving questions whose answers help to find solutions to the problem, both from the theoretical and practical perspectives.

In summary, violence in adolescent and young couple relationships is frequent, since those couples with dysfunctional relationship dynamics resort to violence as a means of communication and conflict management, entering an escalation of violence, leading to stressful situations, and, as a result, causing adverse health effects (^{Exner-Cortnes, 2014}; ^{Muñoz & Echebur·a, 2016}; ^{Rojas-Solφs et al., 2019}). In addition, recent studies point out that these dysfunctional relationship dynamics are normalized, which leads young people to have low perception of abuse, even under high levels of dating violence victimization; this is what the Andalusian Institute for Women calls technical abuse. Therefore, it is important and necessary to address the individuals awareness of being afraid of their partners or feeling mistreated in their dating relationships (^{López-Cepero et al., 2015}; ^{Riesgo González et al., 2019}; Rodríguez- Franco et al., 2012).

Violence is not natural; individuals learn it (^{Echebur·a, 2019}; ^{Ibabe et al., 2020}) intentionally, as well as to harm others. Saying that violence is only exercised by men means perpetuating traditional gender roles and denying the other side of violence, which currently occurs in most dating relationships (^{Riesgo González et al., 2019}; ^{Trujano et al., 2010}). In this way, assuming that men are aggressors and women are victims should be ruled out, since not all cases are male intimate violence situations (^{Zamora-Damián et al., 2018}). Therefore, the general objective of this research is to analyze the prevalence of victimization and violence perpetration in adolescent and young couples, identify the directionality of violence, and thus contribute to the intervention approach to favor healthy and satisfactory relationships.

The following specific exploratory objectives are proposed based on the general objective:

2. Method

3. Results

3.1 Prevalence of dating violence

Firstly, with the purpose of studying the prevalence of victimization and violence perpetration, the analysis of the chi-square statistic (χ ²) shown in Tables 2 and 3 was conducted. These results demonstrate significant differences concerning violence between men and women for victimization related to alienation, that is, stopping talking, ignoring the feelings of the partner, and even disappearing (61-39%; χ ² = 3.410, p < .05), as well as for physical violence (46.1-53.9%; χ ² = 12.074, p < .001) (Table 2). As it can be seen, the other variables were not significant, but both men and women manifest high prevalence of victimization. Furthermore, in our sample, physical violence and alienation are the factors with the highest percentage in men and in women, respectively. Lastly, overall, victimization (total victimization) was not significant, although the highest percentage occurred in women (58.1-41.9%; χ ² = .019; p > .05).

Table 2 Prevalence of violence victimization in women and men 

Factors of violence	Victimization	Women	Men	χ2
		(%)	(%)
Total victimization		58.1%	41.9%	.019
	Alienation	61%	39%	3.410*
	Humiliation	55.8%	44.2%	1.797
	Coercion	55.9%	44.1%	1.982
	Sexual	60.6%	39.4%	.676
	Physical	46.1%	53.9%	12.074***

Note. *p < .05 **p < .01 ***p < .001

Table 3 shows the prevalence of violence perpetration reported by women and men, where women perpetrate (total perpetration) more intimate dating violence (59.9-40.1%; χ ² = 2.439; p > .05). Specifically, coercion (63.2-36.8%; χ ² = 7.754, p < .001), sexual violence (36.6-63.4%; χ ² = 32.863, p < .001), and physical violence (65.9-34.1%; χ ² = 3.581, p < .05) were significant. In-dices of factors of violence perpetration range between 34% and 66%. In addition, it can be stated that women present higher percentages of perpetrated psychological violence; specifically, related to coercion, control, or retention of the partner, followed by alienation and humiliation of the partner. However, men show higher levels of sexual violence perpetration.

Table 3 Prevalence of violence perpretation in women and men 

Factors of violence	Perpretation	Women (%)	Men (%)	χ2
Total perpetration		59.9%	40.1%	2.439
	Alienation	57.2%	42.8%	.460*
	Humiliation	57%	43%	.373
	Coercion	63.2%	36.8%	7.754**
	Sexual	36.6%	63.4%	32.863***
	Physical	65.9%	34.1%	3.581*

Note. *p < .05 **p < .01 ***p < .001

In order to be more specific about the type of violence experienced and exercised by the young people in our sample, we have created two decision trees (Figure 1 and Figure 2) using sex as the predictor variable (independent variable) and, as criterion variables (dependent variables), the factors related to dating violence victimization (Figure 1) and dating violence perpetration (Figure 2).

We studied the association between the sex variable and victimization in the different factors of abuse. To this end, a classification tree was created using the exhaustive CHAID approach, which performs chi-square automatic interaction detection by choosing in each step the predictor variable that displays the strongest interaction with the dependent variable; and the cross-validation option was chosen to evaluate the behavior of the model created to generalize the results to larger samples.

Focusing on the classification tree (Figure 1), it can be affirmed that there is a higher percentage of physical victimization in men (53.9%), highlighting a corrected p-value = .001; and, when this victimization does not exist, women present a higher victimization related to alienation (65.8%) compared to men, with a corrected p-value = .004. The overall specificity for the model has been 59.6%; however, it should be noted that the correct percentage for women reaches 86.7%, while for men it only amounts to 21.7%. Therefore, the risk estimate is 40%.

In the classification tree shown above, the relation-ships between sex and perpetration of the different factors of abuse were analyzed. The exhaustive CHAID procedure was performed again with cross-validation.

Figura 1 Classification tree of victimization according to sex 

For this model, the overall specificity has been a little higher (62.2%), showing again a very remarkable difference between the correct percentage displayed by men (22.4%) and women (90.8%), as well as a lower risk estimate (below 38%).

In this case, the results obtained indicate that the percentage of sexual perpetration of men (63.4%) is higher than that of women (36.6%) (Figure 2). This result has a corrected p-value = .000 that, since it is lower than .05, is considered significant. The following node suggests that when there is no sexual perpetration, coercive perpetration occurs in 71.7% of women with a corrected p-value = .000. Finally, out of that percentage of women, 82.5% also perpetrates physical violence with a corrected p-value = .032.

Victimization and perpetration were also studied from the age perspective; to this end, the sample was divided into two groups: adolescents (15-19 years) and young adults (20-29 years). An analysis of the chi-square statis-tic (χ ²) was conducted and the results are shown in Table 4. Regarding victimization, the results suggest that there are significant differences between violence related to alienation (59.6-49.3%; χ ² = 5.597, p < .05) and violence related to coercion (52.2-43.4%; χ ² = 4.017,p < .05).

As for perpetration, no significant differences were identified between adolescents and young people. Despite the few significant differences found, data demonstrate that adolescents have the highest rates of both victimization and perpetration. However, these data should be approached with caution, given that the group of adolescents is much smaller than that of young adults and, if it were expanded, it would probably go in the same direction.

Subsequently, each sample was analyzed separately aiming at identifying possible differences regarding sex. The results indicate that, in the sample of adolescents, there were differences in physical victimization (12.8-31.8%; %; χ ² = 7.814, p < .05), being boys the ones with the highest rates of victimization, and in sexual perpetration (8.5-22.7%; %; χ ² = 5.190, p < .05), being boys again the ones who perpetrated this type of violence more frequently. With regard to the sample of young adults, the results were virtually the same: physical victimization (12.2-17.1%; χ ² = 3.226, p < .05), sexual perpetration (9.9-19.5%%; χ ² = 12.667, p < .01), and physical perpetration (13.7-8.5%; χ ² = 4.468, p < .05). The only difference from adolescents was observed in physical perpetration, where girls had the highest percentage.

Figure 2 Classification tree of perpetretion according to sex 

3.3 Directionality of dating violence

Table 5 shows the directionality of violence according to sex. The highest percentage was observed in bidirectional violence, that is, 65.20% of participants reported themselves as aggressors and victims at the same time in their couple relationships, while 14.30% affirmed that they were only victims or only aggressors.

Figure 3 Prevalence of perception of abuse according to the sex of the group 

Table 4 Prevalence of violence victimization and perpetration in adolescents and young people 

Factors of violence	Victimization	Adolescents (%)	Young people (%)	χ2
Total victimization		78.3%	72.6%	2.158
	Alienation	59.6%	49.3%	3.597*
	Humiliation	42.2%	44.1%	.001
	Coercion	52.2%	43.4%	4.017*
	Sexual	21.7%	21.4%	.007
	Physical	18%	14.3%	1.421
Factors of violence	Perpetration
Total perpetration		70.2%	67.5%	.435
	Alienation	53.4%	47.2%	1.998
	Humiliation	39.1%	38%	.065
	Coercion	45.3%	40.8%	1.099
	Sexual	12.4%	14%	.273**
	Physical	10.6%	11.5%	.119*

Note. *p < .05 **p < .01 ***p < .001

Table 5 Directionality of dating violence 

	Total (%)	Women (%)	Men (%)
Healthy relationship (No aggressor-No victim)	20.40%	20.10%	20.90%
Total unidirectional violence	14.30%	13.30%	15.80%
Unidirectional violence (Aggressor-No victim)	4.40%	4.90%	3.60%
Unidirectional violence (No aggressor-Victim)	10%	8.40%	12.20%
Bidirectional violence (Aggressor-Victim)	65.20%	66.70%	63.30%

The most perpetrated factors of violence, according to the directionality of violence (unidirectional or bidirectional), are shown in Table 6. Through the analysis of Students t, it can be observed that all factors displayed significant differences. It can also be concluded that bidirectional violence was the most frequent and that perpetration of psychological violence had the highest prevalence, being alienation (t(683) = 2.847; p <.01), humiliation (t(683) = 2.969; p < .01), and coercion (t(683) = 4.670; p < .001) the most common types; followed by the most aggressive factors, that is, sexual violence (t(683) = 4.840; p < .001) and physical violence (t(683)=12.696;p<.001) the latter not being exercised when violence is unidirectional. All effect sizes were be-tween moderate and considerable.

4. Discusion

In recent times, dating violence has become a problem of social relevance (^{Ferrer-Pérez & Bosch-Fiol, 2019}) that affects our social and psychological well-being. For this reason, studying dating relationships at an early age is vital to identify the relational dynamics and communicative styles that begin to take shape and that will be used in subsequent relationships (^{Exner-Cortnes, 2014}; ^{López-Cepero et al., 2014}). In consequence, the main objective of this study was to analyze victimization and violence perpetration in dating relationships and to determine the directionality of violence in our sample, in order to properly address a phenomenon that is emerging in our society and that is showing the other side of social reality.

In the first place, this work studied the prevalence of victimization and perpetration in relation to sex. Specifically, our data suggest that violence is exercised and experienced by both men and women in the same way. Women exercise more psychological and physical violence (and also experience it). Consequently, men present higher levels of physical victimization and exercise more sexual violence (^{Cortés-Ayala et al., 2015}; ^{García-Carpintero et al., 2018}; ^{Kidman & Kohler, 2020}; ^{Pazos et al., 2014}; ^{Rodríguez, 2015}; ^{Rubio-Garay et al., 2017}). In addition, we have observed that, in the absence of physical victimization, psychological victimization occurs, mainly related to alienation, especially in women. Regarding violence perpetration, when there is no sexual violence, coercive and physical violence occurs, being, again, more commonly perpetrated by women. How-ever, the models analyzed do not classify men properly, so it cannot be stated that there are significant differences in relation to sex. This means that both men and women exercise violence, which confirms the findings of studies with community samples that follow the line of research on the bidirectionality of violent behavior.

Our results are, therefore, consistent with research works that have demonstrated bidirectional violence recurrence in dating relationships (^{Alegría & Rodríguez, 2015}, 2017; ^{Arnoso et al., 2017}; ^{Chen & Chan, 2019}; ^{Graña & Cuenca, 2014}; ^{Melander et al., 2010}; Rojas-

^{Solís et al., 2017}; ^{Rubio-Garay et al., 2019}; ^{Rubio-Garay et al., 2017}; ^{Rubio-Garay et al., 2012}; ^{Zamora-Damián et al., 2018}). In this study, the directionality of violence was studied after identifying those participants with victimization, perpetration, or both indicators in the same relationship. Data reveal that bidirectional violence occurs in a much higher percentage than unidirectional violence among the relationships of our sample. It is predominantly manifested in psychological aggressions (alienation, humiliation, and coercion) and, to a lesser extent, in sexual and physical aggressions, which agrees with the study by ^{Chen and Chan (2019)}, who report that, except for sexual and serious physical violence - which are unidirectional- most violent behaviors in couples are bidirectional. Therefore, after all the analyses conducted, we can conclude that psychological violence is the most frequent among the young people in our sample. These findings are in accord with research works conducted by other authors who claim that bidirectional-ity can be generally found in psychological violence and that this directionality decreases when the severity of the behavior increases with physical violence (^{Chen & Chan, 2019}; ^{Graña & Cuenca, 2014}; ^{Rubio-Garay et al., 2012}).

Another objective of the research was to approach differences in both victimization and perpetration be-tween adolescents and young adults. With regard to adolescents, we should keep in mind that this is a critical stage when individuals start assuming responsibilities and establish their first dating relationships. It is known that adolescents’ and young adults dating relationships are different from those of adult couples in terms of level of commitment, responsibility, duration, etc. In addition, violence within the couple also differs: it is usually reciprocal and less severe. Our results suggest that the youngest couples have the highest percentages in both victimization and perpetration, thus supporting the conclusions of most studies conducted with this type of population (^{Aizpitarte et al., 2017}; ^{Aizpitarte & Rojas-Solís, 2019}; ^{Cortés-Ayala et al., 2015}; ^{de la Villa et al., 2017}; ^{Pazos et al., 2014}; ^{Zamora-Damián et al., 2018}).

It is worth highlighting that the amount of young people who do not feel mistreated in their relationships or who are not aware of being in abusive relationships is much higher than that of individuals who do perceive themselves as being mistreated, which explains such high percentages of psychological violence. Recognizing themselves as mistreated is complicated for young populations, especially for men, who use to be socially labeled as violent (^{López-Cepero et al., 2015}). In our study, as in many others, this is called technical abuse, since, despite having important indicators of violence, individuals do not perceive themselves as abused. In summary, this evidences the need to improve and increase the availability of resources to support young populations, especially in the identification of indicators that are not visible and lead to a lack of perception of abuse (^{López-Cepero et al., 2015}; ^{Rodríguez-Franco et al., 2012}).

Finally, it is possible to draw conclusions that con-tribute to the development of prevention and intervention programs, modifying the approach to this social problem (^{Ferrer-Pérez & Bosch-Fiol, 2019}). In order to help young people by means of respect-based education, we must place greater emphasis on psychological violence and on the use of strategies to deal with situational violence. By addressing risk factors, it is possible to solve episodic or reactive situations using positive strategies that prevent violent aggressions. For this reason, it would be useful to make progress with the study of relationship conflict resolution strategies in future studies (^{Rojas-Solís et al., 2019}). This is a great opportunity to guide young people in healthy interpersonal relationship dynamics, since, after being immersed in a first problematic relationship, the capacity to identify inappropriate ones may increase. In this way, it is fundamental to teach them how to respond positively to conflicts and, therefore, stop those destructive patterns of communication to build solid, healthy, lasting, and respectful relationships. To this end, learning and be-coming aware of psychological violence, but also of conflict coping skills and egalitarian attitudes about gender roles, are indispensable, as they could prevent wrong learning that is inevitably consolidated and transferred to subsequent relationships -as suggested by different studies- besides being a predictor of physical violence (^{Exner-Cortnes, 2014}; ^{Juarros-Basterretxea et al., 2019}; ^{Rubio-Garay et al., 2019}; ^{Van de Bongardt et al., 2015}).

As for the main limitations of this research, we can mention, on the one hand, methodological aspects such as the sample size; therefore, it would be interesting to increase the groups in order to generalize, to a greater extent, the results about victimization and violence perpetration. In this sense, expanding the group of adolescents would also be important. On the other hand, in order to conduct a more exhaustive study on bidirectional violence, it would be interesting to determine whether violent responses are reactive.

As a proposal for future research works, the two partners are being assessed to obtain more information that can be contrasted from the two perspectives of the couple, given that the subject of the study has a great impact on society and needs attention from the different professionals who work in this field.

As a general conclusion, relationship conflicts among adolescent and young couples are bidirectional, being psychological violence (alienation, humiliation, and coercion) the most common manifestation. According to numerous studies, sexual violence is still more frequent among boys. Finally, it is worth noting that adolescents seem to be more susceptible to both perpetration and victimization.

Undoubtedly, it is necessary to highlight the clinical and social relevance of this type of research works. In particular, the clinical relevance lies in the importance of detecting cases of unperceived mistreatment in order to provide help and, of course, make the problem visible. The social relevance, for its part, tries to show that violent behaviors should never be normalized. For all these reasons, it is important to take action as soon as possible, offering the youngest couples adequate tools to solve the conflicts that may arise in their relationships and, in this way, prevent the consolidation of violent behaviors and promote healthy relationships. This would increase the efficiency of intervention programs and improve the support plans and strategies addressed to youth.