Attitudes of adolescents toward violence and their related factors

Behavior research revealed that attitudes are important in understanding the roots of violence. People acquire those attitudes by social learning, particularly in adolescence, because in this stage of life is the influence of social environment strongest. This study aimed to assess adolescents’ attitude toward violence and their related factors. The Questionnaire on Attitudes toward Diversity and Violence 1 was applied. This questionnaire addresses the following factors: justification of violence toward peers as a reaction of courage, sexist beliefs and justification of domestic violence, and tolerant beliefs and rejection of violence. An extra factor was incorporated in the research to evaluate the attitude toward the justification of sexual violence. Analysis of the responses to the questionnaire disclosed middle level of justification and rejection toward violence. The stronger attitudes of justification of violence toward peers, domestic violence and sexual violence were related with male sex, alcohol consumption, violence among siblings and higher level of enjoyment of violence in media. The stronger tolerant beliefs and rejection of violence was related with female sex, the experience of sexual violence and the age group of 14-16 years old. Results clearly recommend the need to work with adolescents and their families to eliminate prejudices that justify violence and to promote violence rejection as preventive measures, considering the related factors.


INTRODUCTION
The World Health Organization defines violence as "the intentional use of physical force or power, threatened or actual, against oneself, another person, or against a group or community, that either results in or has a high likelihood of resulting in injury, death, psychological harm, maldevelopment or deprivation" (WHO, 2002). The causes of violence are complex and occur at different levels, and to represent this complexity one often uses the ecological, or social ecological model. This model, which comes from the systemic perspective, suggests that there are four level of factors that causes violence and it is the combination of the factors which lies at the basis of violence (Buvinic et al., 1999). The recognized four levels of factors are respectively: individual factors such as demographic characteristics, psychological, and personality; relational factors like there are the relationships with family, friends, intimate partners and peers; community factors which encompass the living environment and the communities where social actions are executed as school, work, neighborhoods, among others; and societal factors such as social structure, cultural norms and attitudes which support and/or tolerate violence (Krug et al., 2003;WHO, 2002).
In Latin-America researchers found that the following factors relate to violence: personal factors such as gender, age, biological background, education level, employment status, socioeconomic level, alcohol and substance abuse, and early exposure to violence; relational factors related to the living quarter such as overcrowding of the home, history of family violence, home dynamics and rules, and poverty level; community and social factors comprise social inequality, social history of violence (war), effectiveness of social control institutions (judiciary and police, as well as institutions of civil society such as church and community groups), availability of guns and drugs, violence in the media, poverty level of the neighborhood, crime rate in neighborhood and its environment such as the number of houses, street lighting, and cultural norms that promotes or disapproves violence (Buvinic et al., 1999).
Youth violence represents a serious problem affecting individuals, communities, and the larger society (Davidson & Canivez, 2012). Violence among young people occurs between individuals and between groups at school, in neighborhoods, in families and within intimate partner relationships (Pinos & Ávila, 2012(Pinos & Ávila, , 2013Sundaram, 2013;Ali et al., 2016). Violent behavior in adolescents is described by different authors as "part of their search for identity and self-assertion at the adolescent stage of life" (Braconnier & Marcelli, 1996). In the search for identity are the expectations that emerge from cultural norms fundamental. These cultural norms are formed in the same social environments as the individual lives through action, language and representation (Sundaram, 2013). In adolescence shifts the primary influence increasingly from the family, which has been dominant since childhood, to peers (Sheer, 2010).
Cultural norms, such as traditional gender roles, affect the perceptions of violence and shapes its definition (Yick, 2000;Bucheli & Rossi, 2015). That is, cultural norms such as prejudices against women, minorities and others make them more vulnerable to violence by the dominant group (Worchel et al., 2003;Wallach et al., 2010). Prejudices and beliefs regarding gender roles, diversity and minorities frequently are used by people to explain violence. As a result often the victim is blamed for instigating the violence and the perpetrator feels morally justified for the violence. For instance: a study in Kenya found that some sex offenders are immersed in normative cultural expectations about sexuality and gender; within this framework, they endeavor to create a picture that shifts blame to their victims (Muchoki, 2011). This stereotypical belief about gender is transmitted to new generations and makes the adolescents justify violence. In the UK, a qualitative study in middle and late adolescents it was found that some of them believe that it is acceptable for men to use violence and abuse in interpersonal relationships (McCarry, 2010). In a similar research in Cuenca (Ecuador) it was found, that although most adolescents, rejected all sorts of violence and claimed to possess the knowledge to identify situations of violence and their protagonists, in the argumentation they showed contradictory statements in which they justify violence and thus the perpetrators. The justification is based on internalized knowledge and assumptions about traditional beliefs that normalize violence; consequently, in some circumstances is the victim blamed for the violence .
On the other hand, traditional gender roles, stereological and prejudiced beliefs against minorities make the victims accept violence and/or feel guilty for the violence suffered. For instance, women traditionally are assumed to be responsible for taking care of others and for maintaining interpersonal relations, even if this mean enduring violence (Virkki, 2007;Sequeira, 2009). Regarding this, a qualitative study in India found that the traditional gender roles may influence the acceptance of violence and to create a normative threshold of violence acceptance which is dependent on its intensity, justification and frequency (Go et al., 2003). In a research in Ecuador with adolescents, it was found that LGBT adolescents were victims of all kinds of violence because of their sexual orientation. Some of them showed their fear to be rejected and assaulted due to stereotypical beliefs and traditional gender roles that result in homophobia. In fact, the homophobic attitudes were related to a high level of prejudice against LGBT. Moreover, some of LGBT adolescents presented self-homophobia; they assumed that their sexual orientations are not right, a reason why they believe that they deserve the violence .
Overall, beliefs and cultural norms are often the cause that not only perpetrators but also victims consider violence as normal, deserved and justifiable. This is very concerning because the possibility of a violent occurrence is greater when both the victim and perpetrator justify violence or accepted levels of it (Krug et al., 2003). To understand the roots of violence, cultural norms, prejudices and beliefs to promote or justify violence are the factors to consider. As was established above, cultural norms and beliefs that promote or justify violence are factors related with the exercise of violence, and the preconceptions are related to the other violence related factors such as personal, relational, community and social factors.
A way to understand the beliefs and feelings regarding a social object is through attitude assessment ( Barreiro & Marcuello, 2003). Attitudes are a "relatively constant feeling, predisposition, or set of beliefs directed toward an object, person or situation" (Green & Kreuter, 1991;Myers, 1995). Although the strongest attitudes are supported by our own experiences, it is not possible to experiment with all social objects; hence, attitudes usually are formed through social learning (Rodriguez, 2003). Researchers speculate that positive attitudes toward violence are related to its exercise, and the exercise of the violence is related to the positive attitudes toward violence. For instance, considering social learning theory, children and adolescents experiencing and/or observing violence at home and/or in the community may develop unhealthy attitudes towards interpersonal violence as well as engagement in aggressive behavior (Davidson & Canivez, 2012). Studies regarding violence between peers discovered a link between attitudes that justify this behavior with its exercise (Hoge et al., 1996;Dodge & Crick, 1999;Ochoa et al., 2007;Martínez-Ferrer, 2008;Ruiz et al., 2009a;Ruiz et al., 2009b;Estévez et al., 2010). A study in adolescents in Canada found that male perpetrators had more sexist attitudes than other male participants (Lacasse & Mendelson, 2010).
Regarding attitudes toward violence, studies in Spanish adolescents using the CADV test showed that adolescents presented attitudes of justification of violence toward peers as a reaction of courage, sexist beliefs and justification toward domestic violence. Moreover, they had tolerant beliefs and rejection of violence (Díaz-Aguado, 2004;Viniegra, 2007). Additionally, it appeared that the factor levels of justification of violence toward peers as a reaction to courage and sexist beliefs and justification of domestic violence were higher in male than female adolescents, while factors of tolerant beliefs and rejection of violence were higher in females than male adolescents (Viniegra, 2007). The same tendency of results was found when the CADV test was applied to adults who were studying to be teachers (Caballero et al., 2014). Another study in which the CADV test was applied to young Spanish adolescents showed that sexist beliefs and justification of domestic violence are related to hostile and benevolent sexism, where boys display more negative attitudes toward women than girls. Moreover, high scores in the psychological values, wisdom and knowledge, courage, humanity, justice, temperance, and transcendence seems to be related with the least sexist attitudes towards women (Ferragut et al., 2013). These results agree with the conclusions obtained when attitudes toward violence were measured using others instruments.
Concerning violence between peers, studies detected positive attitudes toward violence between peers and violent behaviors when violence is perceived as a form of entertainment, mode to improve self-esteem, way to handling problems and establish social relations, as well as, perceiving violence as fair revenge and punishment when there are feelings of helplessness and distrust toward justice; e.g., in school, due to considering the world as a hostile and unfair place (Hoge et al., 1996;Dodge & Crick, 1999;Ochoa et al., 2007;Martínez-ferrer, 2008;Ruiz et al., 2009a;Ruiz et al., 2009b;Estévez et al., 2010). About sexist beliefs, some studies detected that adolescents still present sexist attitudes (Heras et al., 2006;Lameiras et al., 2008;Rodríguez et al., 2005). A study in young adults from Salvador revealed that men scored higher than women in attitudes of justified sexual violence (Sierra et al., 2007). In general, results demonstrate that males obtain significantly higher scores than females on generally positive attitudes toward violence. In a study of USA adolescents, males obtained higher scores about attitudes which support corporal punishment, crime and war and partner violence (Davidson & Canivez, 2012). Few researchers have looked for the relationship between personal, relational and social violence related factors and attitudes toward violence. In research with adolescents of African descent in the USA, it was found that lower socioeconomic status, older age, and adolescents' perceptions that their parents support fighting were related to endorsing positive attitudes toward violence as revenge (Copeland-Linder et al., 2007). Research of early adolescents from the USA indicated that cross-gender violence (female-to-male), perpetration, harsh parenting, peer deviance, low family income, and neighborhood hazards accounted for significant variation in attitudes toward dating conflict (Windle & Mrug, 2009).
Overall, the link of some factors with violence, such as attitudes that justify it, have been described. This factor, however, is related to others factors; hence, the attitudes that justify violence may be a way to understand the violence and to take preventive measures. Yet, there is very little information concerning the relation between violence related factors and attitudes that justify violence. Also, there is insufficient information regarding attitudes in our context, Ecuador. Given this, the aim of this study was to assess adolescent attitudes of justification or rejection toward violence and to examine factors associated with their attitudes. The hypothesis which lie at the basis of this research is: a) Justification toward violence to peers and partners and sexist beliefs will be associated with boys, late adolescence (17-19 years old), public high school, low level of family functioning, overcrowding living conditions, low education level of mother and father, alcohol consumption within the family, frequency of violence within the family, violence experienced (negligence, physical, emotional or sexual), enjoy watching violence in media, and a low level of confidence in the justice system as related to violence. b) Tolerant beliefs and rejection of violence will be associated with girls, early adolescence (11-13 years old), private high school, high level of family functioning, high education level of mother and father, no violence within the family, no violence experienced, no enjoyment of watching violence in media, and a high level of confidence in the justice system. It is to be expected that the research results will guide and assist in the setting up of educational programs for adolescents that reduces the prejudices related with the exercise and admission of violence, as well as strengthen factors which favor attitudes of rejection towards violence.

Participants
A transversal and correlational study was conducted to assess the attitudes toward violence and their related factors (Hernandez et al., 2007). High schools and the classes within the high schools were chosen by a random cluster sampling process. Data were gathered using 533 questionnaires in 20 high schools in Cuenca (Ecuador). The final sample was constituted by 372 male adolescents and 158 femal; 118 participants in the age range 11-13 years old, 327 between 14-16 years old, and 88 between 17-19 years old; 463 participants attended a public high school, and 70 attended a private high school. The instrument was applied anonymously and an informed consent was signed by the participants and the school authorities.

Instruments and measures
The applied instrument consisted of two sections. The first section aimed at collecting demographic data such as 1) sex, 2) age (11-13 early adolescence, 14-16 middle adolescence, 17-19 late adolescence), and 3) type of school (public or private); the assessed factors associated with violence were: 4) and 5) education level of mother and father (illiterate, school, high school, and college), 6) overcrowding (yes or no), 7) level of family functioning (high, middle, low, null), 8) alcohol consumption within the family (yes or no) and 9) its frequency (daily, weekly, monthly, irregularly, never), 10) frequency of violence within the family (always, sometimes, never), 11) violence between parents, 12) violence among siblings, 13) violence among parents and children, 14) frequency of experienced violence (always, sometimes, never), type of violence experienced: 15) negligence, 16) physical, 17) emotional and 18) sexual violence (yes or no); 19) preference about watching violence in media (high, middle, low, null); and 20) level of confidence in the justice system (high, middle, low, null).
Overcrowding was assessed with two questions: "how many bedrooms are in your home?", and "how many people sleep in your home?" Reporting more than three persons per room was considered overcrowded. The scales of level of family functioning, preference watching violence in media and level of confidence in the justice system were constructed by the research team, and the questions of these scales were chosen by principal component analysis. In consequence, the level of family functioning had seven items with a correlation mean of 0.55 with asseverations, including "if you have a problem, you can obtain the help of someone in your family?", and "can you talk about any theme without fear within family?"; the reliability by Cronbach´s Alpha was 0.66. Preference watching violence in media included two items with a correlation mean of 0.69, e.g., "do you usually like to watch movies or programs with action, beatings and shootings". Since Cronbach´s Alpha requires at least 3 items, this indicator was not calculated. Finally, the level of confidence in the justice system had three items with a correlation mean of 0.54, e.g., "in our country the judges punish those who act with violence and abuse others"; Cronbach´s Alpha was 0.41.
The second section contained the Survey of Attitudes toward Diversity and Violence (CADV), developed by María José Díaz-Aguado. This instrument considers the following factors: justification of violence toward peers as a reaction of courage (F1), sexist beliefs and justification of domestic violence (F2), and tolerant beliefs and rejection of violence (F3). The factor intolerance and justification of violence against minorities and punishment was eliminated since this phenomenon was less appropriate or relevant for our study region. The research team designed an extra factor called sexual violence justification (F4), to assess attitudes of justification toward sexual aggressions. All factors were assessed with a Likert scale from 1 to 5 (1=strongly disagree, 2=disagree, 3=do not agree or disagree, 4=agree and 5=strongly agree). The score of 5 shows full justification toward violence, with exception of tolerant beliefs, and regarding rejection of violence a factor of 5 shows full rejection toward violence.
Justification of violence toward peers as a reaction of courage contained 16 questions, e.g., "if your friend attacks someone who has annoyed him/her, you should support your friend", and "sometimes, it is right to threaten others so they know you have a strong character"; Cronbach's alpha coefficient was 0.82. Sexist beliefs and justification of domestic violence contained 17 questions, regarding male dominance within the family, gender discrimination and violence against children and women, e.g., "for the sake of their children, women must endure the violence of their husbands or partners and they should not report this behavior to the authorities", and "if a woman is punished by her husband, she must have done something to provoke him"; Cronbach's alpha coefficient was 0.70. Tolerant beliefs and rejection of violence had 6 items, e.g., "the problem of violence against women by their husbands or partners affects the whole of our society", and "the indigenous have had a positive influence in our culture"; Cronbach's alpha coefficient was 0.36. An extra factor of sexual violence justification was included via 7 questions, e.g., "somehow, a man is justified in requiring a woman to have sex with him if they had sex in the past", and "it is the woman's fault if her partner forces her to have sex"; Cronbach's alpha coefficient was 0.67.
As bias, the HumSex team 1 decided to change the CADV's Likert scale from 7 to 5 for the reason of having less dispersion in the choices, as well as to avoid the central tendency bias produced because the respondents tend to choose the middle values such as 2 and 4 instead of 1 and 5. This mistake is more recurrent when there are more intermediate values. Moreover, it was demonstrated that no more information is achieved when the Likert scale is further than 5. Furthermore, the factor of tolerant beliefs and rejection of violence had one less item in comparison with the original scale; the item was removed because it did not fit in our context. Some factors had a Cronbach's alpha values minors than 0.5.

Analysis
The SPSS 19 statistical package for social sciences was used for data analysis. The descriptive analysis encompassed the calculation of frequencies for the demographic data and related factors, as well as the mean and standard deviation for each of the four factors. In a second step, the data were analyzed by inferential methods to confirm mean differences per demographics and associated factors. In all distributions, the Kolmogorov-Smirnov and the Shapiro-Wilk tests for normality were applied to verify the parametric assumptions; Levene's test was used to confirm homoscedasticity.
Due to parametric assumptions, the fact that all distributions did not satisfy homoscedasticity and the unbalance between groups, it was decided to apply parametric (t-Student and ANOVA one-way) and non-parametric (Wilcoxon and Kruskal-Wallis) tests. Only parametric results are depicted in the results section because parametric and nonparametric results agreed in most cases; and when there was no coincidence, the parametric test was the right choice considering post hoc analysis. The t-Student test was used for independent samples consistent with the compliance of homoscedasticity and the nonparametric Wilcoxon test one tail was applied for each factor by independent variables: sex, high school type, overcrowding, alcohol consumption within family, violence between parents, violence among siblings, violence between parents and children, experienced physical -emotional -sexual violence, and negligence or carelessness experienced.
ANOVA one-way and the Kruskal-Wallis test were used to determine per factor if statistical significant differences exist between the independent variables, such as: age, father's educational level, mother's educational level, family functioning level, frequency of alcohol consumption within the family, frequency of violence within the family, frequency of violence experienced, level of enjoyment of violence in media, and level of confidence in the justice system. Post hoc Bonferroni one tail analysis was carried out to perform pairwise comparison within each two groups. The corresponding measures of effect size by means of Cohen's d and r for each statistically significant variable was calculated. Finally, for each factor, multiple regression was performed to determine the relationship with all independent variables.

RESULTS
The results showed that 136 (25.5%) of the participants' mothers received some basic education, 223  Table 2 shows for each of the four factors the results of regression analysis between each of the factors and the independent variables.

Factor justification of violence toward peers as reaction of courage (F1)
A mean of 2.49 (SD=0.66) out of 5 was found as general appreciation that violence toward peers is a reaction of courage; it is 30.44% showed that they strongly disagree, 24.25% showed disagreement, 21.26% did not agree or disagree, 14.2% agreed, and 9.85% showed strong agreement with items that justify peer violence. In summary, the majority, 54.69%, of the participants disagree with violence, while 45.31% moderately to strongly agree. According the results of the t-Student test (see Table 1) significant differences of F1 and the independent variables sex, alcohol consumption of a family member, violence among siblings and experienced negligence were detected. The highest mean values were found in male adolescents and in the presence of alcohol consumption within the family, violence among siblings and negligence experienced. The effect of observed size was small in all variables (data no show). Application of the ANOVA one-way test revealed significant differences by age, level of family functioning, frequency of violence within the family, level of enjoyment of violence in media, and level of confidence in the institutions of the justice system. Bonferroni post hoc found differences by age, between 11-13 year olds and 14-16 year olds (p<0.01), where the older age group presented high justification scores in comparison with the younger age group. Among the level of family functioning between high and low with a large effect size (p<0.001; d=-0.78, r=-0.36), high and middle (p<0.05), and middle and low with a medium effect size (p<0.05; d=-0.52, r=-0.25), high levels of family functioning were related to weaker attitudes of justification toward violence. Regarding the level of enjoyment of violence in media between high and never with a large effect size (p<0.001; d=1.32, r=0.55), high and low with a large effect size (p<0.001; d=0,93, r=0.42), high and middle with a medium effect size (p<0.001; d=0.61, r=0.29), and between medium and never with a medium effect size (p<0.001; d=0.76, r=0.36), middle and low (p<0.05) and between low and never (p<0.05), higher enjoyment due to such programs was correlated with a higher level of attitudes. Finally, the level of confidence in the justice system between high and never with a medium effect size (p<0.001; d=-0.76, r=-0.36), middle and never (p<0.01) and low and never (p<0.01) where less confidence in the justice system was correlated with a higher level of attitudes. Furthermore, Bonferroni post hoc found differences by frequency of alcohol consumption, that is, there were differences between irregular consumption and abstinence from consumption (p<0.05), where adolescents who had irregular alcohol consumption in their families had higher levels of attitudes of justification. Regarding the effect size (Cohen's d and r), only values that indicate a medium or high effect are shown. 19% of this factor was explained by independent variables, which proved to be significant. The significant independent variables found were: sex (male), age (younger age), level of family functioning (low), level of enjoyment of violence in media (high) and level of confidence in the justice system (low).

Factor of sexist beliefs and justification of domestic violence (F2)
The average appreciation of the respondents between sexist beliefs and justification of domestic violence was 2.49 (SD=0.53) out of 5. 32.99% of participants showed strong disagreement, 21.48% disagree, while 20.46% did not agree or disagree, 13.54% stated to agree and 11.54% showed strong disagreement with the phases to calculate this factor. In total, 54.46% of the participants disagreed with violence. There were significant differences, according the t-Student test, between means of this factor by sex, high school type, alcohol consumptions within family, violence among siblings and emotional violence experienced. The highest average values of attitudes were found in male adolescents, those who attended a public high school, those that had alcohol consumption within the family, those with violence between siblings in their home and those that did not experience emotional violence. There were significant differences by the ANOVA one-way test of mothers' educational level, frequency of alcohol consumption within family and enjoyment of violence in the media. Bonferroni post hoc found differences in the same factors; specifically, between mothers with high school and university education (p<0.05) where high attitudes belong to participants whose mothers only had a high school education. The frequency of alcohol consumption showed a difference between irregular and abstinence (p<0.05), where irregular alcohol consumption was associated with a higher mean. The level of enjoyment of violence in media between high and never (p<0.001), high and low (p<0.01), middle and never (p<0.001), and low and never (p<0.01) where more enjoyment showed higher attitudes of justification.
Although the ANOVA one-way tests did not find the frequency of violence within family to be significant, there was a difference between frequencies of violence classified as always and never (p<0.05), where a higher frequency of violence was associated with higher attitudes. The observed effect size was small in all variables (data no show). 16% of this factor was explained by independent variables, which proved to be significant. The identified significant independent variables were sex (male), physical violence (not) and negligence experienced (not) and level of enjoyment of violence on media (high).

Factor tolerant beliefs and rejection of violence (F3)
A mean of 3.33 (SD=0.66) out of 5 was obtained analyzing the respondents' response to this factor. 17.07% of participants strongly disagree, 10.51% disagree even though 21.36% did not agree or disagree, 24.48% agree and 26.58% strongly agree with the variables used for the assessment of tolerant beliefs and violence rejection. In total, 51.06% of the participants rejected violence. This factor showed significant differences based on the t-Student test by sex, where female adolescents had higher attitudes of tolerance and rejection of violence and experienced sexual violence. The highest average values were when the respondent experienced sexual violence. The ANOVA one-way test showed the presence of a significant difference by stage of adolescence. Particularly, the Bonferroni post hoc test found significant differences among 11-13 and 14-16 years olds (p<0.01) were the highest average values were the result of the responses of the middle adolescents. The observed size effect was small in all these variables (data no show). 1% of this factor was explained by independent variables, which did not prove to be significant. Nevertheless, sex (female) and level of family functioning (high) were identified as significant individual variables.

Factor justification of sexual violence (F4)
The respondents gave a mean of 2.63 (SD=0.73) out of 5 to the factor of sexual violence justification. 27.20% of participants strongly disagreed, 20.18% disagree with violence even though 26.35 % did not show agreement or disagreement, 15.14% showed agreement and 11.12% showed strong agreement with phases of sexual violence justification. Globally, 46.53% of the participants disagreed with violence. Significant differences with this factor in accordance with the t-Student test were noticed in the following independent variables: sex, alcohol consumption within the family, violence among siblings, and sexual violence experienced. The highest average values were found in male adolescents, when there was alcohol consumption within the family, violence among siblings, and in the absence of experienced sexual violence. The ANOVA one-way test showed significant differences by frequency of violence within the family and level of enjoyment of violence in the media. According Bonferroni post hoc were differences between violence within the family among always and never with a medium effect size (p<0.01; d=0.58, r=0.28) the highest in the group of adolescents who suffered frequently domestic violence. Level of enjoyment of violence in media between high and never with a medium effect size (p<0.001; d=0.6, r=0.29), high and low (p<0.01), high and middle (p<0.05) and middle and never (p<0.05), disclosed that greater enjoyment of media violence was associated with higher attitude levels.
While the ANOVA one-way test did not find differences with fathers' education level, Bonferroni post hoc found differences between fathers that completed high school and fathers with a university degree (p<0.05). Higher attitudes were associated with adolescents whose fathers had only a high school degree, and the level of family functioning between high and low (p<0.05) and middle and low (p<0.05) was associated with higher attitudes. Only Cohen's d and r corresponding to medium or high effect size Table 1. t-Student, mean differences of the four factors and variables 2 1, 3,6,8,11,12,13,15,16,17,18 and ANOVA one-way mean differences of the four factors and variables 3 2, 4, 5, 7, 9, 10, 19, 20. are presented. 9% of this factor was explained by independent variables, which proved to be significant. The significant independent variables were sex (male) and sexual violence experienced (not).

DISCUSSION
The objective of the study was to assess the violence phenomenon in new generations and the adolescent's attitudes of justification or rejection, toward violence and their related factors. The questionnaire based survey found middle level of attitudes that justify violence toward peers as a reaction of courage, sexist beliefs and justification of domestic violence and beliefs that justify sexual violence. Those findings correspond respectably well with the results of other studies regarding violence between peers (e.g., Díaz-Aguado, 2004;Viniegra, 2007;Copeland-Linder et al., 2007;Ruiz et al., 2009;Davidson & Canivez, 2012;Caballero et al., 2014), sexist beliefs (e.g., Díaz-Aguado, 2004;Heras et al., 2006;Lameiras et al., 2008;Rodríguez et al., 2005;Viniegra, 2007;McCarry, 2010), justification of domestic violence (e.g., Díaz-Aguado, 2004;Viniegra, 2007;Windle & Mrug, 2009;Ferragut et al., 2013;Caballero et al., 2014), and justification of sexual violence (e.g., Sierra et al., 2007). Male adolescents had higher level of attitudes that justify violence in comparison with female adolescents as found in other studies (e.g., Viniegra, 2007;Sierra et al., 2007;Davidson & Canivez, 2012). School, neighborhood and family factors strongly affects the attitudes of violence justification, similar as demonstrated in Sherer (2010) and Sundaram (2013). The study also disclosed that adolescents still maintain traditional ideas about gender roles, where boys had the higher scores. Those findings agree very well with other studies with focus on the attitudes of violence of Ecuadorian adolescents (Jaruseviciene et al., 2014). These results are disquieting since attitudes that justify violence, especially toward women and peers, are related to the aggressive behavior against these groups (Díaz-Aguado, 2004;Ferrer et al., 2006;Krug et al., 2003;Worchel et al., 2003;Yick, 2000;Wallach et al., 2010;Bucheli & Rossi, 2015) due to on the one hand blaming the victims for violence (McCarry, 2010;Muchoki, 2011; and on the other hand advocating that the victim in certain conditions accept violence (Go et al., 2003;Virkki, 2007;Sequeira, 2009;. As suggested in the hypothesis, there is a large amount of overlap in related factors for different types of violence justification, suggesting that the different types of violence are part of the whole concept of violence. Personal factors as the tendency of justifying violence experienced or observed, relational factors such as family components (parent's educational level, dysfunction, overcrowding, alcoholism) and community and social factors (high school type, distrust in the judicial system and media influences) are related, as shown in this study, with socio-cultural based attitudes of violent behavior and violence justification. On the other hand, analysis of the questionnaires revealed among the participants' middle level of tolerant beliefs and attitudes of rejection toward violence, with mean values higher than the mean of any other factor of justification. This finding is in line with the results published by Díaz-Aguado (2004) and Viniegra (2007). Furthermore, it has been shown that female adolescents possess a higher level of rejection toward violence than male adolescents, a result that correspond with the study of Viniegra (2007). Given that the related violence factors are correlated with each other, means that the attitude towards violence is a psychological construct promoting a propensity to aggressive behaviors. By contrast, the tolerant beliefs and rejection of violence did not present a link with violence justification, and can therefore be considered as protective violence factors (Díaz-Aguado, 2004;Viniegra, 2007). Another relevant finding is that factors associated with violence in Latin-America, such as sex, age, having suffered or witnessed violence, degree of overcrowding and inadequate family functioning, violence in the media and little confidence in the institutions of police and judicial systems, are all part of an ecological model relating personal, relational, community and social factors with violence (Buvinic et al., 1999;Krug et al., 2003;WHO, 2002). In other words, the violence related factors match with the related factors of the attitudes of violence justification as suggested in the hypothesis of this research. Hence it is imperative to work with attitudes and their related factor to eradicate violence. Nonetheless, some disagreements with theory were found, that is, when there was emotional violence experienced, there were fewer acceptances to sexist beliefs and justification of domestic violence. Participants who suffered sexual violence had fewer approvals toward justification of sexual violence and they had more tolerant beliefs and rejection toward violence. These can mean that in some types of violence, victims had conscience about violence, so they do not try to justify it. A new research in this area with more victims participating to understand these findings might be appropriate.
The results of the present study urge the development of an educational intervention that addresses different environments. Specifically, the intervention must overcome a micro-system focus and develop a macro-systemic dimension. A method to advance the modification of our current situation is to promote attitudes of tolerance toward diversity through changing attitudes about false myths and beliefs that are responsible for the development and acquisition of violence (Pérez, 2002). The WHO (2002) recommends that it is necessary to develop a preventive educational program for adolescents against violence with an emphasis on strategies of troubleshooting and empowering to eliminate traditional gender roles related to violent behavior (Buvinic et al., 1999). It is important to think ecologically and design interventions with adolescents; interventions must involve apart, from school and institutions, parents and family (Mendelson et al., 2010;Mendelson & Letourneau, 2015).
Unfortunately, the traditional school often gives insufficient responses when violence occurs among students, by the tendency to minimize the severity of assaults between equals, by considering this as "normal". To overcome these situations, teachers must be trained in the subject, promoting attitudes of rejection of violence and developing actions in the classroom to promote respect among peers (Gini, 2004). Specifically, educational projects should be proposing as an essential objective to teach condemnation of all forms of violence, promoting assertive communication as well as identifying and overcoming prejudices in society about certain groups and developing attitudes of tolerance towards diversity (Cowie, 2000;Díaz-Aguado, 2005;Deardorff, 2015). In this process, teachers are crucial, because they are transmitters of knowledge, but also of attitudes, especially in the changing society in which we live. Hence, they require much more training to understand new social and cultural realities, and to develop needed skills (Montero, 2000). Moreover, given the importance of perceptions of parents in the adolescents' attitudes, violence prevention and intervention efforts need to engage parents (Montero, 2000;Dumas, 1989;Tolan et al., 1990). All in all, preventative works should approach this problem from an integrative stance, with the intention of overcoming the basic concept of gender discrimination. Moreover, research is necessary with a positive approach to know the related factors that promote tolerant beliefs and rejection of violence to work toward the strengthening of such attitudes (Ferragut et al., 2013).

CONCLUSIONS
The present study disclosed the existence of a concerning percentage of teenagers who justify violence between peers (24.05%), domestic violence (25.08%) and sexual violence (27.58%), while those who hold attitudes against the tolerance and rejection of violence amounted 26.26%. The variables related to justification of violence toward peers as a reaction of courage were male gender, middle adolescence, alcohol consumption within the family, violence among siblings, negligence experienced, low family functioning, violence within family, high level of enjoyment of watching violence in the media, and low level of confidence in the justice system institutions. The variables associated with sexist beliefs and justification of domestic violence were male gender, attendance at public high school, high school as a mother's education level, alcohol consumption within the family, violence between siblings, no emotional violence experienced and high level of enjoyment of watching violence in the media. The variables related with justification of sexual violence were male gender, high school as father's education level, alcohol consumption within the family, violence between siblings, but no sexual violence experienced, and high level of enjoyment of watching violence in the media. The variables related with tolerant beliefs and rejection of violence were female gender, middle adolescence and sexual violence experienced. The findings favor the development of programs that pursue, detect and combat beliefs that justify violence. Such programs should involve teachers and parents; the activities must include the identification and eradication of a set of beliefs that encourage the violent behavior in the various fields as well as the promotion of factors that promote respectfulness and equal treatment.