Abstract
This work analyzes how the assumption of responsibility by aggressors convicted for gender-based violence is related to sexist attitudes, self-esteem and perceived functional social support. Similarly, the predictive capacity of these variables is studied with respect to the aggressors’ minimization of the harm done and a lack of attributing responsibility to themselves. The participants in the research were males condemned to prison sentences for crimes related with gender-based violence in Spain. The instruments applied were the Attribution of Responsibility and Minimization of Harm Scale, the Ambivalent Sexism Inventory (ASI), the Rosenberg Self-Esteem Scale (RSE), the Functional Social Support Questionnaire (FSSQ), and the Social Desirability Scale (SDS). The study concludes that sexist attitudes are related with a greater lack of attribution of responsibility, as well as with a greater tendency to minimize the harm done by the aggression. In addition, the aggressors with low self-esteem use self-defense as a strategy to justify the violence. Similarly, the presence of an adequate social support network for the aggressor increases the attribution of responsibility on the part of those convicted for gender-based violence.
Keywords
Introduction
Over the last few years, gender-based violence has become a subject of maximum social and scientific interest, due mainly to its high prevalence (Montañés & Moyano, 2006). In spite of this fact, the number of victims reporting this type of violence is not real as many of the cases occur within families and the victims do not report it. The use of such terms as domestic violence or violence against wives induce one to think of private, personal acts (Ferrer & Bosch, 2003). So it has been calculated that the reported cases only represent between 10% and 30% of the real number of cases (Patró & Limiñana, 2005). Nevertheless, in recent years, the prevalence of crimes of gender-based violence has increased as a result of a greater awareness and so more female victims report it (Menéndez, Pérez, & Lorence, 2013).
The traditional social mentality, sexist attitudes, and gender stereotypes all influence the social concealment of gender-based violence. There are, however, other factors that characterize this kind of violence and which lead women to tolerate repeated situations of violent behavior: minimizing the harm done and attributing blame to the victim herself (Perela, 2010). This assumption of blame can be generated by the process of abuse itself, causing the woman to prolong or not to run away from the violent situation to which she is subjected (Escudero, Polo, López, & Aguilar, 2005).
Nevertheless, the study of gender-based violence is not only a question of explaining why the victim maintains and perpetuates the violent relationship, it is also necessary to analyze the aggressor’s behavior and the social context. With respect to the above, recent research in Spain points to the ever greater presence of strategies to justify and deny the harm done among those convicted of crimes related with gender-based violence. The result is that these aggressors have cognitive, emotional and behavioral barriers that allow them to absolve themselves from responsibility for their violent acts (Carbajosa, Boira, & Tomás-Aragonés, 2013; Herrera & Expósito, 2009; Lila, Gracia, & Herrero, 2012; Lila, Oliver, Catalá-Miñana, Galiana, & Gracia, 2014; Loinaz, 2014). In this respect, when the aggressors are compared with nonviolent men, the former have a greater tendency toward using external attributes to justify their behavior, blaming their partner, and presenting a greater tendency to minimize the harm done (Dobash & Dobash, 2011; Wallach & Sela, 2008). Thus, those who have been convicted for gender-based violence frequently resist assuming any responsibility for their violent behavior, even blaming everything on the victim (Bowen, 2011; Expósito & Ruiz, 2010; Henning & Holdford, 2006).
Ferrer and Bosch (2016) and Montañés and Moyano (2006) point out that the patriarchal system still rules supreme in society, so the transmission of values takes place through a system of traditional beliefs, one which champions the superiority of men over women. Sexism, or discriminatory attitudes toward women, simply because they are women, is one of the factors that underpin gender-based violence (Heise, 1997). Sexist attitudes are not usually expressed openly nowadays, so the manifestation of such attitudes is often imperceptible, something which favors their assumption and encouragement by women themselves (Darddene, Dumont, & Bollier, 2007). Glick and Fiske’s (1996) concept of sexism considers such attitudes to be an ambivalent phenomenon, one which includes traditional forms of hostility combined with feelings of attraction and pleasure.
In this respect, sexist attitudes can contribute to the genesis of violence against women in the heart of a relationship with affective bonds, even reaching the point where aggression is legitimized as a way of resolving conflict in the couple and favoring the blaming of the victim (Berkel, Vandiser, & Bahner, 2004; Expósito & Herrera, 2009; Flood & Pease, 2009; García-Leiva, Palacios, Torrico, & Navarro, 2007; Palacios et al., 2005).
Self-esteem has also been related to the use of justification strategies and minimizing the harm done with respect to gender-based violence (Lila et al., 2012, 2014). Several studies demonstrate that men who are violent with their partners have low levels of self-esteem and tend to justify the aggression, given that they perceive situations as more threatening, which leads to a greater frequency and severity of the violent behavior (Echeburúa, Amor, & Corral, 2009; Garofalo, Holden, Zeigler-Hill, & Velotti, 2016; Gottman & Jacobson, 2001; López, 2004; Walker & Bright, 2009; Walker, 1984). Nevertheless, in other cases, the aggressors show a high self-esteem as a result of a defense mechanism that implies a profound level of insecurity (Echeburúa & Amor, 2016). Direct aggression seems to be more commonly related to evaluations that are highly favorable to self-esteem among aggressors, and which can be threatened in the dynamic of an intimate relationship (Cale & Lilienfeld, 2006; Wyckoff & Kirkpatrick, 2016).
Another aspect to be taken into account in gender-based violence is the social support network of both the aggressor and the victim. Heise (2011) explained the model of gender-based violence from a social perspective, in which aggression within the couple is related to individual characteristics that may either increase or decrease the risk of violence, and which can combine with elements and dynamics that arise from the family environment, the context closest to the aggressor, as well as from the aggressor’s other, wider social relationships. Thus, Díaz-Aguado (2003) and Gracia and Herrero (2006) stress the importance of establishing adequate social networks that can act as alternatives to violence, given that the absence and/or inadequacy of such social support could suppose a greater risk of violent behavior patterns (Campbell, 1995). Similarly, it has been determined that a negative support network increases the probability that the aggressor will relapse into gender-based violence (Gracia, Herrero, Lila, & Fuente, 2009).
Similarly, perceived social support has a certain relation with the aggressor’s moral socialization, as this process begins within the convicted aggressor’s closest social context (Marzana, Vecina, & Alfieri, 2016). From this point of view, personal responsibility for aggressive behavior can be avoided by using justification strategies in the form of moral disassociation processes (Bussey, Quinn, & Dobson, 2015). Likewise, there would seem to be a relationship between violence against a partner and certain moral principles, so men found guilty of gender-based violence behave selfishly while being convinced that they are acting correctly, as they have an absolutist conception of what is correct as far as the relationship with their partner is concerned (Benavidez, Neria, & Jones, 2016; Vecina, Marzana, & Paruzel-Czachura, 2015).
However, the use of violence against their partner has been linked to a high level of social desirability on the part of the aggressors (Echeburúa & Fernández-Montalvo, 2009; Lila, Conchell, & Catalá-Miñana, 2016), as the distortion and bias of their responses occurs in a context in which the results have important repercussions and may influence the attainment of certain penitentiary benefits. Thus, social desirability and the misrepresentation of the facts in this population entail determining how far the responses given by the aggressors really reflect reality (Boira & Tomás-Aragonés, 2011).
In view of the above, this research will bring greater knowledge of the variables that make up most of the specific gender-based violence programs in Spain (Conchell, Lila, & Catalá-Miñana, 2012; Duarte, Gómez, & Carrillo, 2010; Hernando, 2007; Lila et al., 2016; Lila, Herrero, & Gracia, 2008; Martínez & Pérez, 2009; Pérez, Giménez-Salinas, & de Juan, 2013).
In this context, the aims of this present study were (a) to analyze the attribution of responsibility, sexist attitudes, the perceived social support, and the self-esteem of a sample of Spanish aggressors convicted for gender-based violence; (b) to examine the existing relationship between the attribution of responsibility by the aggressors and the cognitive, social and emotional variables stated above; and (c) to also determine the predictive capacity of sexist attitudes, the perceived social support, and self-esteem in the assumption of responsibility on the aggressor’s part.
On the basis of the theoretical review carried out, it was expected that aggressors condemned for gender-based violence would not assume responsibility for their actions, minimizing the harm done; show sexist attitudes that exacerbate the violence against women; and have an inadequate perception of social support and self-esteem, thus increasing the risk of gender-based violence (Hypothesis 1). In addition, we also expected that the discriminatory, sexist attitudes toward women, the lack of and/or inadequate social support as well as low self-esteem would be related to the use of such strategies as not attributing responsibility to themselves and minimizing the harm done (Hypothesis 2). We also expected that the discriminatory, sexist attitudes, the inadequate perception of social support, and a low self-esteem would be predictors of the aggressors’ attribution of responsibility (Hypothesis 3).
Method
Participants
This research received the participation of 129 men who had been convicted for gender-based violence and were serving sentences in various penitentiaries. The participants were between 22 and 85 years of age (M = 39.98 years; SD = 10.86). Of these participants, 27.9% had been convicted for physical and psychological abuse. The second most common conviction was that of injuries with 11.6%, followed by threats and duress with 9.3%. As for the criminal records of the aggressors, it is worth noting that 27.1% of those convicted were persistent offenders in gender-based violence, while 44.2% had criminal records for other crimes. 41.9% of the aggressors found guilty of gender-based violence who participated in the study admitted to substance abuse, while 17.1% were suffering from a serious mental illness. It should be noted that 58.6% currently participate, or have participated in the past, in a specific treatment program for those found guilty of gender-based violence.
The participants were interned in several penitentiaries and social integration centers (SICs) in Spain: Cáceres prison (21.7%) and Cáceres SIC (1.5%), Badajoz prison (15.6%), Sevilla I prison (11.6%), Sevilla II prison in Morón de la Frontera (12.4%) and Sevilla SIC (4.6%), Alicante prison (7.8%), Villena prison (10.8%), and Málaga prison (9.4%) and Málaga SIC (4.6%).
Instruments
Various instruments were used in the research, all possessing psychometric reliability and validity guarantees:
Attribution of Responsibility and Minimization of Harm Scale (Lila et al., 2008). This scale evaluates the attribution of blame in those convicted of gender-based violence and the degree to which the aggressors minimize the importance of the violence done to the partner. In this research, we have used the refactorized version of the original scale, in which items referring to the minimization of the harm done are included. It uses a Likert-type scale of 20 items, scoring from 1 point (totally disagree) to 5 points (totally agree). The high score is related to a greater lack of attribution of responsibility, as well as to a greater use of strategies of justification and minimization of the harm done. The scale’s reliability reflects adequate indices, since Cronbach’s alphas for all the subscales fluctuate between α = .68 and α = .79. The internal consistency found in this study is α = .77. It provides information concerning four subscales:
- Blaming the Victim (BV): This measures the degree to which the aggressor blames the victim for the situation, alleging the cause is in the victim’s lies or psychological problems. The reliability reflects adequate indices, as Cronbach’s alpha for this subscale is α = .73. In the sample of our study, the internal consistency is α = .74. - Self-Defense (SD): This evaluates the degree to which the aggressor believes that he behaved in the way he did in response to an attack by his partner. Cronbach’s alpha for this subscale is α = .79. The internal consistency found in our study is α = .71. - Self-Attribution of Blame (SB): This measures the degree to which the aggressor attributes the cause of the events to internal factors or to personal problems. Cronbach’s alpha for this subscale is α = .70. The internal consistency in the sample of our study is α = .69. - Minimization (M): This measures the degree to which the aggressor minimizes the importance of the crime he has been convicted for. Cronbach’s alpha for this subscale is α = .68 (Lila et al., 2012). The internal consistency found in the data of this study is α = .70.
Ambivalent Sexism Inventory (ASI; Glick & Fiske, 1996), Spanish adaptation by Expósito, Moya, and Glick (1998). This has a Likert-type scale of 22 items scoring from 0 points (totally disagree) to 5 points (totally agree). The reliability reflects adequate indices α = .86. It provides information concerning two subscales:
- Hostile Sexism (HS): This is a subjectively negative orientation toward women, with the understanding that women possess characteristics that make them inferior to men. The internal consistency found in this study is α = .83. - Benevolent Sexism (BS): This maintains a subjectively positive orientation toward women, who are conceived as deserving respect, affection and protection, as long as they adhere to the traditional gender roles, thus legitimizing gender inequality and keeping women in a subordinate position with the justification of their weakness. The internal consistency found in this study is α = .73.
Rosenberg Self-Esteem Scale (RSE; 1965), Spanish adaptation by Fernández-Montalvo and Echeburúa (1997). This aims to evaluate the feeling of satisfaction that a person has with her/himself. The scale consists of 10 Likert type items scoring from 1 point (totally disagree) to 4 points (totally agree). The reliability of the instrument is adequate α = .85. The internal consistency found in our study is α = .79.
Functional Social Support Questionnaire (FSSQ) by Broadhead, Gehlbach, Degruy, and Kaplan (1988), Spanish adaptation by Bellón, Delgado, Luna, and Lardelli (1996). This measures the perceived social support in intimate relationships and the self-esteem of the aggressor convicted of gender-based violence. The instrument consists of eight questions measured in a Likert-type format with 5 options for each answer, which range from 1 point (much less than I would like) to 5 points (as much as I like). The reliability reflects adequate indices α = .87. The internal consistency found in this study is α = .83.
Social Desirability Scale (SDS; Crowne & Marlowe, 1960). We used the Spanish adaptation by Ferrando and Chico (2000), widely used in research with aggressors. This scale evaluates the tendency to respond in a socially desirable way, which can lead people to show a distorted image of themselves in order to “disguise their real character” or to “make a good impression” (Michaelis & Eysenck, 1971). The scale is made up of 33 items with a true-false answer format. The adapted version has psychometric properties similar to those of the original scale. The instrument’s reliability is adequate α = .78. The internal consistency found in our study is α = .72.
Procedure
We first of all solicited authorization to implement the instruments from the office of the Secretary General of Penitentiary Institutions (Interior Ministry, Government of Spain). Having obtained the necessary permissions, the different penitentiaries and SICs which were to participate in the study sent us the list of those convicted of gender-based violence. We then made contact with each of the prisoners asking for their written consent to participate in the research. The 129 aggressors who participated agreed to complete the evaluation instruments. The principal motivation of the prisoners to participate in the study was to obtain various penitentiary permits and/or benefits.
We first of all gave them the Attribution of Responsibility and Minimization of Harm Scale, followed by the rest of the instruments: the ASI, the FSSQ, the RSE, and the SDS. The instruments were applied individually. The application time for the instruments was 75 min, so as to avoid tiredness in the participants. The evaluators were present throughout the process of data gathering, to resolve doubts and to ensure that they were correctly completed. A total of six evaluators participated in the administration process of the instruments, divided between the different penitentiary centers that participated in the research.
Statistical Analysis
For the statistical treatment of the data, we used the computer program IBM SPSS Statistics Version 23. We first of all carried out a descriptive analysis for a detailed study of the characteristics of the aggressors regarding their sexist attitudes, self-esteem, perceived social support, and the different dimensions of the attribution of responsibility. Then, with respect to the nature of the variables, we carried out a correlation analysis (Spearman) to examine the relation between the attribution of responsibility of the aggressors and the cognitive, emotional, and social variables mentioned previously. We also studied the relation between the social desirability of those convicted of gender-based violence and their responses in the instruments of measurement. Finally, we carried out a linear regression analysis to determine to what extent the sexist attitudes, self-esteem, and perceived social support could predict the use of justification strategies and minimization of harm done by the aggressors.
Results
To address the first hypothesis of the study, Table 1 shows the descriptive results concerning the distribution of the participants in the different instruments.
Distribution of the Sample, Means, and Typical Deviations in Attributing Responsibility, Sexist Attitudes, Self-Esteem, and Perceived Social Support.
Note. ASI = Ambivalent Sexism Inventory; FSSQ = Functional Social Support Questionnaire.
As for the descriptive statistics corresponding to the scale of attribution of responsibility and minimization, it can be seen that the aggressors have a higher average score in the factor of blaming the victim (M = 2.91; SD = 1.323), followed by the general scale of attribution of responsibility (M = 2.31; SD = 0.737), and by the self-defense factor (M = 2.29; SD = 1.393). As for the general scale of attribution of responsibility, the data indicate that 58.9% of those convicted of gender-based violence assume the responsibility for the acts committed. However, in 38% of the aggressors, the assumption of responsibility is moderate, and low in 3.1%.
As for minimizing the harm done during the aggression, the results indicate that 4.7% of the aggressors have a high score in this aspect, thus avoiding responsibility, while 24% have moderate scores. With respect to blaming the victim, the data conclude that 31.1% of those convicted (n = 40) have high scores, blaming their partner for the violent acts. Similarly, it can be seen that the scores here are moderate for 30.2% of the aggressors (n = 39). As for the self-defense factor, the results indicate that 23.3% of the aggressors have high scores (n = 30), considering the aggression to be a reaction to an assault by the victim, while 17.1% show moderate scores (n = 22). Finally, in attributing blame to oneself, it can be seen that 12.5% of those convicted do not assume responsibility for the aggression (n = 16), while 13.2% do so only moderately (n = 17).
As for the dimensions of the inventory on ambivalent sexism, the data indicate that the highest average score is in benevolent sexism (M = 2.71; SD = 1.055), followed by the subscale of hostile sexism (M = 2.57; SD = 1.103). The lowest average score is in ambivalent sexism (M = 2.54; SD = 0.984). The results indicate that 29.5% of the aggressors have a moderate score in ambivalent sexism and 20.2% have high scores, while the results are low in 50.3%. As for hostile sexism, 23.3% of the aggressors have high scores and 30.2% have moderate scores. However, 24,8% of the convicted aggressors have benevolent sexist attitudes, as opposed to 38% who have low scores in benevolent sexism.
The results obtained in the scale of self-esteem (M = 2.37; SD = 0.560) reveal that 55% of the aggressors have an adequate self-esteem, 41.1% have high self-esteem, and 3.9% have low self-esteem. As for perceived social support (M = 3.53; SD = 1.097), the data show that 60.4% of the aggressors perceive support from their family surroundings, while 19.4% do not.
For our second hypothesis, the results of correlations between ambivalent sexism, self-esteem, social support, and attribution of responsibility are presented in Table 2.
Correlation Between Sexist Attitudes, Self-Esteem, and Social Support With Respect to the Attribution of Responsibility and Minimization of the Harm Done.
p < .05. **p < .01. ***p < .001.
As for sexist attitudes, it can be seen that ambivalent sexism shows positive correlations with the lack of attribution of responsibility (p < .001), minimizing the harm done (p < .001), blaming the victim (p = .024), and self-defense (p = .002). The results were similar for hostile sexism, as this also correlated with a lack of attribution of responsibility (p < .001), minimizing the harm done (p < .001), blaming the victim (p = .005), and self-defense (p = .001). In addition, the results indicate that benevolent sexism is related with a greater lack of attribution of responsibility on the aggressor’s part (p = .016), as well as a greater tendency to minimize the harm done (p = .034). Thus, the greater the presence of sexist attitudes, the greater is the lack of attribution of responsibility; as they tend to minimize the harm done, blaming the victim or justifying their aggression as self-defense.
However, there is no evidence that self-esteem correlates in a significant way with the lack of attribution of responsibility on the aggressor’s part, except in the subscale of self-defense, where it can be seen that a greater score in self-esteem means a lower probability that the aggressors will use self-defense as a justification for their aggression against a partner (p = .031).
As for perceived social support, there is no evidence of a significant correlation with the lack of assumption of responsibility on the part of those convicted of gender-based violence, except in the factor blaming oneself. It can be seen that the existence of a perceived social support by the aggressor is related with lower scores in the factor blaming oneself (p = .030), which reflects a greater attribution of responsibility.
As for the correlation analysis between social desirability and attribution of responsibility, ambivalent sexism, self-esteem and social support, the results indicate the following (Table 3).
Correlation Between Social Desirability and Attribution of Responsibility, Sexist Attitudes, Self-Esteem, and Perceived Social Support.
Note. ASI = Ambivalent Sexism Inventory; FSSQ = Functional Social Support Questionnaire.
As can be seen in Table 3, the data show that there is no significant correlation between the attribution of responsibility, sexist attitudes, self-esteem and functional social support with respect to the SDS, so we can conclude that the aggressors’ responses are not distorted so as to adjust to what is socially desirable.
Finally, for our third hypothesis, to determine to what extent sexist attitudes, self-esteem, and functional social support significantly predict the attribution of responsibility, a regression analysis was carried out (Table 4).
Regression Analysis of Sexist Attitudes, Self-Esteem, Perceived Social Support, and Attribution of Responsibility.
Note. AS = ambivalent sexism; HS = hostile sexism; BS = benevolent sexism; SE = self-esteem; FSSQ = Functional Social Support Questionnaire. Bold text represents statistical significance i.e. p < .05.
As can be seen in Table 4, with respect to sexist attitudes, ambivalent sexism significantly predicts the lack of attribution of responsibility on the part of the aggressors (β = .380; p < .001), minimization of the aggression (β = .314; p < .001), blaming the victim (β = .198; p = .025), and justifications using self-defense as the excuse (β = .273; p = .002). Similarly, hostile sexism acts as a predictor of the lack of attribution of responsibility (β = .397; p < .001), minimization of the aggression (β = .368; p < .001), blaming the victim (β = .247; p = .005), and justifications using self-defense as the excuse (β = .275; p = .002). It can also be seen that benevolent sexism significantly predicts the lack of attribution of responsibility (β = .236; p = .007) and minimization of the harm done (β = .188; p = .033).
As for self-esteem, although it does not act as a predictor of the lack of attribution of responsibility on the part of the aggressor, it can be concluded that it contributes significantly to predicting the strategies justifying the harm done based on self-defense (β = −.178; p = .044).
Finally, with respect to perceived social support on the part of the aggressor, it can be seen that it does not significantly predict the lack of attribution of responsibility, although it is a significant predictor of blaming oneself (β = −.221; p = .012).
Discussion and Conclusion
From the results of the study, it can be concluded that most aggressors convicted of gender-based violence assume the responsibility for their violent acts, and that only a small minority use minimization of the harm done and self-defense as strategies to justify the aggression. However, we have found that the aggressors put more blame on the victim, displacing their own responsibility and externalizing the blame, which in turn leads to the women blaming themselves for what happened. This result is consistent with the research of Bosch and Ferrer (2012), Bowen (2011), Dobash and Dobash (2011), and Pérez, Mendieta, and Vilar (2006). Nevertheless, unlike our study, other works of research have demonstrated a lack of attribution of responsibility on the part of the aggressors (Dutton, 2007; Expósito & Ruiz, 2010; Fernández-Montalvo & Echeburúa, 1997; Ferrer & Bosch, 2003; Lila et al., 2012, 2014; Loinaz, 2014; Scott & Straus, 2007; Wallach & Sela, 2008).
As for sexist attitudes, our research has not found forms of ambivalent sexism that aim to legitimize the subordination of women and reinforce the inequality of the sexes. This contrasts with the work of Expósito and Herrera (2009), Flood and Pease (2009), García-Leiva et al. (2007), Glick and Fiske (1996), Heise (1997), and Palacios et al. (2005).
However, concerning benevolent and hostile sexism, we can conclude that the aggressors use both of these forms of sexism indiscriminately. This is similar to the findings of Lameiras and Rodríguez (2003) and Rodríguez, Lameiras, Carrera, and Faílde (2010). Nevertheless, our research does find evidence of a greater number of benevolent (idealizing women as wives, mothers, and romantic objects) than hostile attitudes in the aggressors, something which also presupposes the inferiority of women. However, Chapleau, Oswald, and Russell (2007) found a greater presence of hostile sexism in aggressors.
As for self-esteem, it can be concluded that numerous aggressors have a high self-esteem. This is consistent with the work of Cale and Lilienfeld (2006), Jiménez, Musitu, and Murgui (2008), and O’Moore & Kirkham (2001). A high self-esteem may seem to be the reflection of a distortion of the self-image with narcissistic overtones, which would explain why they behave violently in situations that they perceive as a threat to their dignity (Echeburúa & Amor, 2016). However, unlike in our study, Goldstein and Rosenbaum (1985) and Walker (1984) found low self-esteem in those convicted of gender-based violence. Thus, we can conclude that self-esteem is a primordial component in the analysis of violent behavior, as both low and high self-esteem can be a potential risk factor for aggression (Conchell et al., 2012; Jones, 2008; McMurran & Howard, 2009; Shaver & Mikulincer, 2011; Walker, 1984).
As for perceived social support, we find an adequate functional social support in the aggressors, in both social and family spheres. This contrasts with the work of Campbell (1995), Díaz-Aguado (2003), and Gracia and Herrero (2006). However, our results can be explained by what was pointed out by Gracia et al. (2009). What needs to be taken into account is the fact that functional social support does not refer to any real support the convicted person receives from those around him, but to the aggressor’s own perception of the same. In addition, this perceived social support on the part of the aggressor may be negative, as it can increase the probability of such aggressions being repeated.
As for the relationship between the aggressor attributing responsibility and the presence of sexist attitudes in those same aggressors convicted of gender-based violence, it can be concluded that the greater the ambivalent sexism, the greater the lack of attribution of responsibility will be (the aggressors tend to justify the violence, minimize the harm done, and externalize their responsibility, blaming their partner). These results are in the same line as the studies carried out by Berkel et al. (2004) and Nayak, Byrne, Martín, and Abraham (2003), who demonstrate that sexist attitudes concerning traditional roles and the subordination of women are related with the tendency to blame the victim, legitimize the violence, and sustain myths concerning gender-based violence.
Similarly, our research shows that both benevolent and hostile sexist attitudes are related with the lack of attribution of responsibility on the part of the aggressor and also a greater tendency to minimize the harm done. This can also be seen in the work of Markowitz (2001), Nayak et al. (2003), and Yoshioka, DiNoia, and Ullah (2001).
As for the predictive capacity of sexist attitudes toward the attribution of responsibility, it can be concluded that ambivalent and hostile sexism allow us to predict the lack of attribution of responsibility, the tendency to minimize the harm done, blaming the victim and justifications for the violence based on self-defense. It can also be seen that benevolent sexism acts as a predictor for the lack of attribution of responsibility and the use of minimization of the harm done. These results coincide with the research of Expósito and Herrera (2009), Expósito and Moya (2005), García-Leiva et al. (2007), and Palacios et al. (2005), which show that sexism contributes to the initiation and maintenance of aggressions, acting as a predictor of gender-based violence.
However, the study of self-esteem allows us to state that aggressors with low self-esteem make greater use of strategies that deny responsibility based on the self-defense justification. The low self-esteem of the aggressors is related with a greater tendency to perceive situations as threatening. The efforts aimed at protecting his own image activate mechanisms of self-protection in order to release him from responsibility (Gottman & Jacobson, 2001; Lila et al., 2012, 2014).
Related with social support, our study concludes that the greater the perceived social support, the aggressor’s attribution of responsibility for his acts is greater. Family, social, and institutional support is primordial for the prevention of gender-based violence (Montañés & Moyano, 2006). It is worth pointing out the relevance of an adequate social support network for aggressors convicted of gender-based violence, as both its absence and inadequateness increase the probability of recidivism (Campbell, 1995; Gracia et al., 2009).
As for the contribution of our study, we should point out that the results provide evidence of the existing relationship between sexist attitudes, self-esteem, and perceived social support on the part of the aggressor convicted of gender-based violence with respect to the strategies of justification and denial of the harm done. It is vital that we reflect on the cognitive, emotional, and social strategies that the aggressors maintain in their intimate relationships, as these contribute to the lack of attribution of responsibility with respect to violent acts (López, 2004). For all the above reasons, we consider it necessary to set up prevention programs whose main focus is the modification of stereotypical attitudes concerning violence against women. Bosch and Ferrer (2012) point out that the stereotypical attitudes maintained socially serve to minimize, justify the aggression, and place the blame on the victim. Interventions based on the risk factors that increase the probability of abuse occurring within couples will bring down the figures concerning gender-based violence (Lila et al., 2016).
The conclusions have an effect on the importance of interventions with aggressors found guilty of gender-based violence, with the principle aim of reducing the impact caused by risk factors through a change in attitudes toward women and the suppression of such defense mechanisms as negation and minimization of the harm done (Catalá-Miñana, Lila, Conchell, Romero-Martínez, & Moya-Albiol, 2013; Lila et al., 2016; Pérez et al., 2013).
Thus, the assumption of responsibility is a key element for the participation of aggressors and their involvement in intervention programs. Similarly, it reduces the risk of the aggressors abandoning the program (Daly & Pelowski, 2000), given that minimization of the harm done to the partner can lead to a rejection of the intervention or a premature abandonment of the program (Echeburúa & Fernández-Montalvo, 2009). Therefore, as the convicted aggressors participating in the study assume responsibility for their use of violence, we can conclude that there is a certain disposition toward a change in attitude (Expósito & Ruiz, 2010). The decrease in the sexist attitudes that feed violence is a signal of a lower risk of further aggressive behavior against the partner (Expósito & Herrera, 2009; Hernando, 2007).
We have seen that the support perceived by the convicted aggressor is related with the assumption of responsibility, so adequate social support is determinant in the cessation of violent behavior toward the partner (Gracia, 2009; Lila, Oliver, Galiana, & Gracia, 2013; Lila, Oliver, Lorenzo, & Catalá-Miñana, 2013). The development of intervention programs requires activities that involve the aggressor’s social network.
Gender-based violence appears as a result of a profound insecurity, as the evaluation of self-esteem is more susceptible to high scores among aggressors (Echeburúa & Amor, 2016). Nevertheless, since there is no agreement on considering the relation between self-esteem and a higher risk of violence in a couple, the programs require greater adaptation to the individual characteristics of the aggressors (Carbajosa & Boira, 2013). To do so, it is necessary to include specific emotional regulation techniques; as well as strategies to change personal variables related with self-esteem (Garofalo et al., 2016; Lila et al., 2016). We would stress here the need for research to analyze the factors that cause a lack of assumption of responsibility among male abusers, as this could lead to a greater number of effective prevention and intervention plans with this particular population.
Finally, it should be pointed out that such variables as the mental health and addiction(s) recognized in the medical records of some aggressors who participated in the research were not studied because they had been receiving psychotherapy treatment since starting their prison sentence and that, at the time of the study, they were not consuming toxic substances. Nevertheless, considering that these variables may be contributing factors involved in the aggression committed by some of those condemned for gender-based violence, it would be of interest in future research.
Limitations
The results of our research should be interpreted with caution. Among the limitations, we should point out the difficulty of accessing this particular population, and this has meant we were unable to have a wider sample so as to guarantee its representative nature in the generalization of the results. In addition, the use of self-reporting and the absence of qualitative items may have contributed to lower scores in the studied variables, and this should be taken into account so as to obtain more significant results in future work.
However, the data obtained through a transversal methodology do not allow us to look at the evolution over time of the scores and thus establish causal relationships between the variables analyzed. It would, therefore, be suitable as a future line of work to carry out a longitudinal study. We trust that this current research work may serve to encourage more study in greater depth of this type of violence.
Footnotes
Authors’ Note
This study is part of a national research project titled “The Aggressor in Gender-Based Violence: Typologies, Strategies and Treatment” (Ref. DER2010-16003 presented in the convocation for projects in R+D (BOE, December 31, 2009) and carried out from the University of Extremadura.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
