Controlling Behaviors in Couple Relationships in the Digital Age: Acceptability of Gender Violence, Sexism, and Myths about Romantic Love

Introduction

Gender-based violence is defined as violence that men exercise against women in order to maintain control and domination over them. The Declaration on the Elimination of Violence Against Women defines this type of violence as “any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women, including threats of such acts, coercion, or arbitrary deprivation of liberty, whether occurring in public or private life” (Res. A. G. 48/104; UN, 1994, p. 2). In Spain, intimate partner violence (IPV) constitutes a public health problem. It is the most common violence suffered by women (Martín-Fernández, Gracia, & Lila, 2018), whose impact reaches all sectors of society (López-Ossorio et al., 2018).

This social phenomenon not only manifests itself in adulthood, but it also has an increasingly greater impact on youth and adolescence (Borrajo & Gámez-Guadix, 2015). Terms such as “courtship violence” or “dating violence” have been coined to refer to the abuse at the stage of courtship or first date. Dating violence is the most used concept globally and concerns “physical aggression, psychological and emotional, verbal or implied abuse and which takes place both in public and private” (Ely, Dulmus, & Wodarski, 2002).

A review by Leen et al. (2013) examined the prevalence of abuse in this stage and found that psychological mistreatment (from 22 to 77%) was the most frequent form, followed by physical (from 2 to 44%) and sexual (from 1 to15%) violence. Psychological abuse is the most-used form amongst young couples. O’Leary and Slep (2003) asserted that it is assessed according to three indicators: verbal aggression (i.e., shouting), behaviors of control and dominance (i.e., controlling the partner’s relationships with friends), and behaviors of jealousy (i.e., checking where the partner has been; Muñoz-Rivas, Graña, & González, 2011).

Gender-based violence embodies multiple forms that evolve according to society. As a result of the incorporation of information and communication technology (ICT) in relationships, violence has not been eradicated but it is occurring in a different way (Flores & Browne, 2017). In this sense, violence through ICT is a recent problem that expresses new forms of traditional violence, but it is still being caused for the same patriarchal cultural reasons.

Information and Communication Technology (ICT)

It is important to pay attention to new forms of socialization that emerge in today’s society and affect people’s lives and their personal and social development. According to the National Statistics Institute (Instituto Nacional de Estadística - (INE), 2016), in Spain there were 28 million internet users, of which 82.9% used it every day; the mobile phone was the most widely used device (93.3%) by young people. The ability to communicate, share personal experiences, find solutions or support, and access any person from any place or at any time have caused ICT to become the primary source of interaction amongst young people (Megías & Rodríguez, 2014) who have been identified as the digital generation.

Controlling Behaviors through New Technologies

Researchers have documented both positive and negative impacts that arise from young people’s use of new technologies (Best, Manktelow, & Taylor, 2014). For example, online interaction via mobile phones can provide opportunities to strengthen relationships with friends and partners (Subrahmanyam & Greenfield, 2008); however, these same situations can become opportunities to threaten, harass, and attack other users (Draucker & Martsolf, 2010; Gómez-Franco & Sendín, 2014). Although ICT has fostered instruments that facilitate interpersonal communication, such devices also have become a means of control and violence against an abuser’s partner. According to Donoso, Rubio, and Vilà (2017), control behaviors are the most frequent form of online violence; examples of such behavior include constantly checking up on the whereabouts of one’s partner and confirming who she is with, checking her mobile phone, forcing the partner to stop chatting with someone, forcing her to delete photos or her social network friends or asking for the password to access her personal accounts or social networks.

The situation is serious. One of the first studies about gender violence and social perception found that abusive online control through a mobile phone is the most exercised form by young people; however, it is not perceived as gender violence (Díaz-Aguado, 2013). Recent research has compared both online and face-to-face contexts and determined that 90% of young people believe there is more gender violence in the online context (Donoso, Hurtado, & Vilà, 2018). On the other hand, young people state that they have observed violent behaviors in their online network, but few of them declare to have been a victim or aggressor (Donoso et al., 2018). In this way, it seems that there is a widespread perception about the magnitude of violence that is exercised through ICT instruments, but few people identify themselves as victims or perpetrators. According to Donoso et al. (2018), gender seems to be the best indicator of the type of response a person has to dating violence: adolescent boys adopt more passive behaviors when they observe gender violence in the online context, whereas girls provide the victims more helpful behaviors.

Relationships are an important source of well-being and happiness (Centro de Investigaciones Sociológicas, 2010). However, when the relationship is conflicting and violent, it could become one of the main causes of suffering (Garrido-Macías, Valor-Segura, & Expósito, 2017; Valor-Segura, Expósito, Moya, & Kluwer, 2014). The inability to properly confront this situation could affect different psychosocial areas such as self-esteem, loneliness, social support, and life satisfaction (Gómez-Franco & Sendín, 2014). Some of the strategies used by women to confront cyber dating violence are to delete published content in their social networks, change publications to avoid anger, diminish activity in social networks, or disable their accounts (Vitak, Chadha, Steiner, & Ashktorab, 2017); that is to say, behaviors that isolate and limit women’s lives in the technological realm are similar to those that occur in conjunction with traditional violence (Expósito, 2011).

In this sense, sexist attitudes and myths about romantic love acquire special importance as they are perpetuated even more intensely through social networks.

Sexist Attitudes and Myths about Romantic Love

Sexist attitudes and myths about romantic love are situated at the base of these new forms of relationship between young people. Sexism is defined as the beliefs and attitudes held in traditional gender stereotypes regarding the roles that are considered appropriate for men and women and the relationships that must be kept between both members of the couple (Moya, 2003). According to the Ambivalent Sexism Theory (Glick & Fiske, 1996), traditional sexism is divided in two different components: hostile sexism and benevolent sexism. The hostile component reflects a negative view of women and is manifested towards those who do not assume traditional roles and, thereby representing a threat to the superiority and domination of the male (Glick & Fiske, 2001). Benevolent sexism carries a positive connotation because it considers that women need affection and protection and positively value those who assume traditional roles (Glick & Fiske, 2001).

Different authors agree that ambivalent sexism is positively related to justification of violent attitudes against the partner in the traditional context (Herrera, Expósito, & Moya, 2012; Herrero, Rodríguez, & Torres, 2017; Valor-Segura, Expósito, & Moya, 2011) and victim-blaming (Gracia, García, & Lila, 2014; Martín-Fernández, Gracia, & Lila, 2018; Vidal-Fernández & Megías, 2014). Furthermore, an association has been found between acceptability of intimate partner violence against women (IPVAW) and perpetration of it (Copp, Giordano, Longmore, & Manning, 2016; Gracia, Rodríguez, & Lila, 2015), so high grades of acceptability of IPVAW increase the likelihood that men exercise violence in the couple and that this violence will be justified and normalized by victims (Martín-Fernández, Gracia, Marco et al., 2018; Waltermaurer, 2012). Additionally, high levels of acceptability of IPVAW have been associated with high levels of ambivalent sexism (Martín-Fernández, Gracia, Marco et al., 2018). In a technological context, several researchers agree that ICT devices facilitate the consolidation of gender stereotypes and a symbolic violence that legitimates models of domination based on patriarchal culture and distinction by sex (Donoso, Baños, Hurtado, & Soto, 2016; Flores & Browne, 2017). In particular, Ellsberg et al. (2015) pointed out that the sexist ideology regularly manifest itself as possessiveness and as controlling behaviors when ICT (WhatsApp, Instagram, Facebook, etc.) is used. However, ICT can also be used as a tool for combating sexism and for educating on equality. For example, Navarro-Pérez, Carbonell, and Oliver (2019) recently determined that an intervention with a mobile app for reducing sexism, which Navarro-Pérez, Oliver, Morillo, and Carbonell (2018) designed, was effective, decreasing the level of sexism in adolescents between 6 and 12%.

On the other hand, myths about romantic love refer to the set of unreal and distorted beliefs about the supposed nature of love (i.e., soul mates, exclusivity, faithfulness, jealousy, etc.; Ferrer, Bosch, & Navarro, 2010; Yela, 2003). These myths are socially accepted and contribute to the maintenance of gender stereotypes and the asymmetric power between men and women (Bosch & Ferrer, 2012; Nardi-Rodríguez, Pastor-Mira, López-Roig, & Ferrer-Pérez, 2018; Rodríguez-Castro, Lameiras-Fernández, Carrera-Fernández, & Vallejo-Medina, 2013). Young people are especially vulnerable to the influence of myths about romantic love; they have a distorted impression about what love is and how members of the couple relate to each other (Ferrer et al., 2010; Sharpe & Taylor, 1999). Borrajo, Gámez-Guadix, and Calvete (2015) found that beliefs in myths about romantic love were related to controlling behaviors in the couple fostered by new technology. Young people justify and accept these abusive behaviors (i.e., constantly checking where and with whom one’s girlfriend may be or sharing passwords in their social networks) because they consider them to be expressions of love or worry in their relationships (Nardi-Rodríguez et al., 2018; Redondo, Ramis, Girbis, & Schubert, 2011). On the other hand, García and Gimeno (2017) observed in a sample of undergraduate students that women are more vulnerable than men to the influence of myths about romantic love. Specifically, they pointed out that women show an idealization of love, an unconditional commitment to the relationship, including a high sense of protection and care of the other above satisfaction of their own needs and interests. As teenagers and young adults prefer the use of technology in order to communicate and traditional forms of contact are less frequent (Colás, González, & De Pablos, 2013), it is necessary to analyse the use of ICT instruments and the variables that affect the process of minimization, normalization, and perpetuation of dating violence.

The Current Research

The present research aims to analyse and understand young people’s social perception about controlling behaviors in the couple that is fostered by new technologies. Two studies were carried out, the first with women and the second with men. The studies share the objectives to, on the one hand, understand the frequency with which young people experience and perceive control in relationships and, on the other hand, analyse the influence of some variables that affect social perception of that control, the adopted role on the scene (protagonist vs. observer) and the means of control that is used (face-to-face vs. WhatsApp) by considering the effect of ideological variables such as acceptability of IPVAW, ambivalent sexism, and myths about romantic love.

Study 1

Hypothesis

• H1. Young women identify more easily with violence against the partner when they adopt the role of observer (vs. protagonist), so it is expected that they:
• H1a. Express a lower justification of violent behavior
• H1b. Perceive a greater severity of the situation
• H1c. Perceive a greater risk of suffering dating violence
• H2. Young women identify controlling behaviors amongst the couple to a lesser extent when it takes place through WhatsApp (vs. face-to-face). Specifically, it is expected that they:
• H2a. Express a greater justification of violent behavior
• H2b. Perceive a lower severity of the situation
• H2c. Perceive a lower risk of suffering dating violence
• H3. Ideological variables (ambivalent sexism, acceptability of IPVAW, and myths about romantic love) affect young women’s social perceptions about dating violence, so it is expected that participants with high scores for these ideological variables express the following:
• H3a. A greater justification of violent behavior
• H3b. A lower perception of severity
• H3c. A lower risk of suffering dating violence

Method

Participants. The sample consisted of 224 female undergraduate students at the University of Granada, Spain. The age of the participants ranged from 18 to 34 years (M = 20, SD = 2.2). A total of 88.8% of participants had Spanish nationality, 10.7% were immigrants and 0.4% did not indicate their nationalities. Regarding their sexual orientations, 90.6% of participants were heterosexual, 2.7 % were homosexual, 5.4% were bisexual and 1.3% did not indicate their sexual orientations. Concerning their civil statuses, the majority of the participants were single (73.2%), 0.4% were married, 0.4% were divorced, and 25.9% were dating.

Design and procedure. A between-subjects 2 (adopted role on the scene: protagonist vs. observer) x 2 (means of control used: face-to-face vs. WhatsApp) factorial design was employed through the scenario manipulation technique.

The sample was obtained through incidental sampling in different classrooms within several faculties at the University of Granada, Spain. First, we contacted the course teacher of each class by email and asked for his/her permission to conduct the study during his or her class period. Next, a researcher was trained to give participants appropriate instructions and to carry out the experiment. All participants were assured that their information and responses would remain anonymous and confidential. The students were informed that their participation in this research was voluntary, and that they could quit the study at any time. Therefore, the task would only be completed in a session in their habitual classrooms, with the course teacher always present. Having got informed consent, the participants were randomly assigned to one of the experimental conditions and were given approximately 15 minutes to complete a questionnaire. Once all students had completed the questionnaire, they were informed of the study’s objectives.

Instruments. A questionnaire containing all of the variables to be measured was designed. The first step was to present a scenario about dating violence, corresponding to experimental manipulation (see Appendix A). To design the fictitious scenarios, we used the previous research by Navarro-Pérez, et al. (2018) as our basis. In this way, we recreated situations of daily life, adapting them to WhatsApp context, with the aim of giving it more realism.

Thereafter, the following instruments were administered:

Manipulation check. Two items with a dichotomous response format (yes vs. no) were used to verify the effectiveness of the experimental manipulation: a) Adopted Role on the Scene independent variable (IV): “Is it a scenario that occurred between the members of a young couple, Juan and María?” (MC₁), “Is it a hypothetical situation between you and your partner?” (MC₂); and b) Means of Control independent variable (IV): “The communication between the man and woman was through a mobile platform (WhatsApp)?” (MC₃), “The communication between the man and the woman has been personal (face-to-face)?” (MC₄).

Thereafter, the perceived severity and justification of violent behavior was evaluated through two items, which were an adaptation of those that other authors used in their studies, such as Garrido-Macías et al. (2017), Milesi, Süssenbach, Bohner, and Megías (2019), Valor-Segura et al. (2011), and Vidal-Fernández and Megías (2014).

Perceived severity. It was evaluated through the item: “How severe do you consider the described episode?”. A 7-point Likert-type response format that ranged from 1 (nothing severe) to 7 (very severe) was used.

Justification of violent behavior. This variable was measured through one of the following items, according to the experimental condition: “How justified do you consider Juan’s behavior to be” (observer condition) or “How justified do you consider your partner’s behavior to be” (protagonist condition). The response format was a Likert type that ranged from 1 (completely unjustified) to 7 (completely justified).

The frequency at which young people experience and perceive controlling behaviors in their relationships and amongst others young couples was evaluated through two items, which were designed based on the Gender Cyber Violence Questionnaire (Donoso, 2014).

Experiences of controlling behaviors in participants’ own relationships. These experiences were asked through the following item: “How often have you experienced similar or equal situations in your relationships?”. This statement was evaluated by using a 7-point Likert scale (1 = never, 7 = always).

Frequency of controlling behaviors in young couples. This variable was evaluated through the item “How often do you think these situations occur amongst young couples?” The response format was a Likert type that ranged from 1 (never) to 7 (always).

Subjective risk perceived of dating violence. This element was evaluated by using an adaptation of the self-anchoring scaling designed by Kilpatrick and Cantril (1960), which consisted of showing the participants a pictorial non-verbal scale, such as the 10-point ladder scale, preceded by the following instructions: “Suppose the next ladder represents various levels of risk of violence in a couple. The highest part of the ladder represents a maximum risk of violence within the relationship, whereas the lowest part represents a minimum risk”. Next, we asked them to mark the box that best represented their perceptions of the risk of suffering dating violence.

The Acceptability of Intimate Partner Violence against Women Scale (A-IPVAW; Martín-Fernández, Gracia, Marco et al., 2018). This scale consisted of 20 items that were scored on a 4-point Likert-type scale (1 = nothing acceptable, 4 = very acceptable). High scores were indicative of greater acceptance of intimate partner violence against women. The A-IPVAW consisted of three dimensions: physical violence (i.e., “I think it is acceptable for a man to hit his partner if she has been unfaithful”), verbal violence or coercion (i.e., “I think it is acceptable for a man to shout at his partner if she is constantly nagging/arguing”), and emotional violence (i.e., “I think it is acceptable for a man to threaten to injure his partner or others if she leaves him”). The alpha coefficient for the total scale was .60. In this study, only global scores were analyzed because the primary objective was to obtain an overview of acceptability of the IPVAW rather than a detailed analysis of each specific dimension.

The Ambivalent Sexism Inventory (ASI; Expósito, Moya, & Glick, 1998). This questionnaire consisted of 22 items rated with a 6-point response format ranging from 0 = totally disagree to 5 = totally agree. High scores revealed more sexist attitudes. Half of the items were related to hostile sexism (HS; i.e., “Women get easily offended,” “Women always exaggerate the problems they have at work”), and the other half were related to benevolent sexism (BS; i.e., “Women are bestowed with a purity that few men possess”). The alpha coefficient of the hostile sexism subscale was .87; that of the benevolent sexism subscale was .82.

Myths Scale toward Love (Bosch et al., 2007; adapted in an adolescent sample by Rodríguez-Castro et al., 2013). The scale consisted of seven items. The response format was a 5-point Likert type that ranged from 1 = totally disagree to 5 = totally agree. The alpha coefficient for the scale was .69.

Demographic information. The students’ gender, age, sexual orientation, and marital status were measured at the end of the questionnaire.

Data Analysis. Data analysis was carried out using the SPSS program, version 23. Firstly, a chi-square test was applied to assess the manipulations’ efficacy. After that, in order to assess the frequency at which young women experience and perceive controlling behaviors in their relationships and amongst other young couples, descriptive analyses of frequencies were performed. Next, several hierarchical regression analyses were performed with the objective of verifying our predictions about the influence of the adopted role on the scene, the means of control used, and the ideological variables (acceptability of IPVAW, sexism, and myths about romantic love) on the social perception of dating violence.

Results

Manipulation check. The analysis revealed that all of the experimental manipulations were effective. Regarding the Adopted Role on the Scene IV, in MC₁, 100% of the participants indicated that the episode happened between Juan and María when they belonged to the observer condition, and 57% of women indicated that the situation did not happen between Juan and María when they belonged to the protagonist condition, so differences were statistically significant, χ²(1, 223) = 78.66, p <.001. According to using the rules of thumb for low, moderate, and large effect sizes (Cohen, 1988, p. 532), the analysis showed a Cramer’s V coefficient = .59, so a large effect size was obtained. In MC₂, 60% of young women said that the situation was a hypothetical situation about their relationships when they were allocated to the protagonist condition, whereas 90.27% of participants did not consider this to be a hypothetical situation about their relationships when they were allocated to the observer condition, χ²(1, 223) = 62.30, p < .001. Cramer’s V coefficient was .53, revealing a large effect size.

Regarding Means of Control IV, in MC₃, 99.1% of participants indicated that the communication took place via a mobile phone when they belonged to the WhatsApp condition, whereas 96.4% of participants who were allocated to the face-to-face condition indicated that communication did not take place via a mobile phone, χ²(1, 224) = 204.59, p <.001. The analysis showed a Cramer’s V coefficient = .96, so a large effect size was obtained. In MC₄, the results revealed that 99.1% of participants considered that communication occurred in person when they were allocated to the face-to-face condition, whereas 99.1% indicated that communication did not take place in person when they belonged to the WhatsApp condition, χ²(1, 223) = 215.07, p < .001. A large effect size was obtained (Cramer’s V = .98).

Frequency of controlling behaviors in relationships. To assess the frequency at which young women experience and perceive controlling behaviors in their relationships and amongst other young couples, descriptive analyses of frequencies were performed. The results showed that 84.8% of young women considered controlling behaviors to frequently occur amongst young couples; however, 82.9% declared that they had never or hardly ever suffered from these behaviors in their relationships.

Influence of adopted role on the scene, means of control used, and ideological variables on social perception of dating violence. To analyse the influence of some variables that affect social perception of that control, specifically adopted role on the scene (Hypothesis 1), means of control that was used (Hypothesis 2), and ideological variables, such as the acceptability of IPVAW, ambivalent sexism, and myths about romantic love (Hypothesis 3), several hierarchical regression analyses were performed with the following dependent variables: 1) perceived severity of the situation, 2) justification of violent behavior, and 3) perceived subjective risk of dating violence.

The same procedure was applied for each analysis. Step 1 assessed adopted role on the scene (protagonist = 0, observer = 1) and means of control (face to face = 0, WhatsApp = 1) and the participants’ centered scores in Hostile Sexism, Benevolent Sexism, Acceptability of Intimate Partner Violence Against Women (A-IPVAW), and Myths about Romantic Love. Step 2 involved second-order interactions between experimental manipulations and ideological variables. The results obtained are shown in Table 1.

According to Hypothesis 1, in the first step of regression analyses we found a significant effect of the Adopted Role on the Scene IV on perceived subjective risk of dating violence, β = -.225, p ≤ .001, 95% CI [.098, .351]. Thus, participants who were allocated to observer condition perceived a greater subjective risk of dating violence in comparison with participants who were allocated to protagonist condition, which supported H1c.

Regarding Means of Control IV, regression analyses did not show any simple effect on dependent variables (p > .05), rejecting Hypothesis 2 (see Table 1). However, the second step revealed the effect of a statistically significant interaction between means of control that was used and the A-IPVAW on the measure of perceived severity of the situation, β = .209, p = .013, 95% CI [.073, .615]. Specifically, in face-to-face condition low levels of A-IPVAW predicted a greater perception of severity in comparison with high levels. However, in WhatsApp condition acceptability of IPVAW did not predict perceived severity (see Figure 1).

In addition, regression analyses showed the effect of another statistically significant interaction between IV Means of Control and benevolent sexism on the ‘perceived severity of the situation’ measure, β = .261, p = .016, 95% CI [.077, .725]. In the same way, in the face-to-face condition, low levels of benevolent sexism predicted a greater perception of severity in comparison with high levels. In contrast, in WhatsApp condition, benevolent sexism did not predict perceived severity (see Figure 2).

Finally, regression analyses indicated that ideological variables predict significantly social perception of controlling behaviors in the relationship, which supported Hypothesis 3. In particular, high levels of hostile sexism predicted a greater justification of violent behavior and a lower perceived risk of dating violence. On the other hand, high scores for A-IPVAW predicted a greater justification of violent behavior, a lower perceived severity, as well as a lower perceived risk of suffering dating violence. In addition, high beliefs in myths about romantic love predicted a greater justification and a lesser perceived risk of dating violence (see Table 1).

Discussion

On the one hand, the present study’s aim was to understand the frequency with which young people experience and perceive controlling behaviors in relationships. The results showed that young women considered controlling behaviors to frequently occur amongst young couples (84.8%); however, they declared that they had never or hardly ever suffered from these behaviors in their relationship (82.9%). These findings are consistent with the results by Donoso et al. (2016), who found that young people perceived controlling behaviors amongst other couples of their ages, yet few of them stated that they experienced control struggles in their relationships.

On the other hand, Study 1 aimed to explore young people’s social perception about controlling behaviors in relationships. Based on Hypothesis 1, we predicted that young women who adopted the role of observer (observer condition) in the described episode would more easily identify violence against the partner, in comparison with women who read the hypothetical episode about her relationship (protagonist condition). The main results indicated that the Adopted Role on the Scene IV predicted perceived subjective risk of dating violence in the expected direction, which supported H1c. Women who were allocated to the observer condition perceived a greater subjective risk of dating violence, in comparison with women who were allocated to the protagonist condition. However, the adopted role on the scene was not found to predict perceived severity of the situation and justification of violent behavior, rejecting H1 (a and b). Again, this result is consistent with the findings by Donoso et al. (2018), given that young women identify a greater risk of suffering from dating violence when they adopt the role of observer, that is, when controlling behaviors occur in couples besides their own relationships, as we observed in descriptive analyses performed previously. In this way, it seems that a widespread perception exists of the magnitude of controlling behaviors that are exercised within relationships, but few women identify themselves as victims. Women tend to use a series of personal and social mechanisms to face this experience, such as denial. Denying abuse constitutes a psychological defense mechanism; it does not mean lying or hiding what happens (Expósito, 2011).

According to Hypothesis 2, it was expected that the Means of Control IV that was used predicted perception of controlling behaviors in relationships. Specifically, it was believed that young women belonging to the WhatsApp condition would identify controlling behaviors in a couple to a lesser extent than young women belonging to the face-to-face condition. However, means of control that was used did not predict any dependent variables (perceived severity of the situation, justification of violent behavior, and perceived risk of suffering dating violence), rejecting H2. These results could be due to the fact that controlling behaviors were reproduced through new technologies; thus, they were perceived with the same normality as in a traditional context. In addition, the results pointed out the effect of two significant interactions on perceived severity: on the one hand, interaction between means of control and acceptability of IPVAW (see Figure 1); and on the other hand, an interaction between means of control and benevolent sexism (see Figure 2). In the face-to-face condition, lower levels (vs. high levels) of acceptability of IPVAW and benevolent sexism predicted a greater perception of severity; however, in the WhatsApp condition, these ideological variables did not predict perceived severity. The fact that acceptability of IPVAW and benevolent sexism did not affect perceived severity of the situation and justification of abusive behavior when the episode occurs in WhatsApp suggests that women are accepting and normalizing controlling behaviors online. These behaviors can be normalized due to the high frequency with which they occur in a technological context (Flores & Browne, 2017; Nardi-Rodríguez et al., 2018), with young people considering what is common to be normal. In addition, according to Estébanez and Vázquez (2013), women could consider these behaviors to be signs of worry and love instead of new manifestations of dating violence. In addition, it is important to consider the subjective nature of communication through WhatsApp, where messages between transmitter and receiver are subject to a high degree of interpretation. For example, a woman could think that her partner is joking or that he is not truly angry. However, the ideological variables predicted perception of severity and justification of abusive behavior when the episode occurred face-to-face. Perhaps, this could be due to the fact that control behaviors occur less frequently in a traditional context and, therefore, are more socially rejected. Young people are less accustomed to observe these behaviors face-to-face, so women with low levels of acceptability of IPVAW and benevolent sexism perceive a greater severity of the situation and justify it to a lesser extent in comparison with women with a high acceptability of IPVAW.

Finally, Study 1 proved the influence of ideological variables on social perception of controlling behaviors exercised against one’s partner, which substantiated Hypothesis 3. According to initial predictions, a high degree of hostile sexism predicted a greater justification of violent behavior and a lower perceived risk of dating violence, replicating the findings by Herrera et al. (2012) and Herrero et al. (2017). On the other hand, in agreement with the results by Martín-Fernández, Gracia, & Lila (2018) and Waltermaurer (2012), high scores for A-IPVAW predicted a greater justification of violent behavior, a lower perceived severity, as well as a lower perceived risk of suffering dating violence. In addition, consistent with the findings by Redondo et al. (2011), high beliefs in myths about romantic love predicted a greater justification of violence and a lower perceived risk of dating violence. These findings contribute to previous research studies, as they demonstrate that sexist attitudes and myths about romantic love are situated at the base of these new forms of relationships between young people, legitimating models of domination based on patriarchal culture and distinction by gender (Donoso, et al. 2016; Flores & Browne, 2017).

Once young women’s perceptions of dating violence had been explored, in a second study we aimed to examine social perceptions of young men, replicating the basic findings of Study 1. At the same time, we added two new variables, identification of controlling behaviors and perceived threat due to the loss of power within the relationship, as we considered these variables to be essential when examining men’s social perceptions of dating violence. To explain the phenomenon of dating violence, some researchers suggest that perception of change or loss of control or power within a relationship could motivate the aggression that a man exercises towards his partner (Dutton, 1988). As result of this perceived loss of power, some men react negatively and with strong resistance, trying to maintain or regain power through the use of violence (Babcock, Waltz, Jacobson, & Gottman, 1993; Dutton, 1988; Herrera et al., 2012).

Study 2

Hypothesis

• H1. Young men identify more easily with dating violence when they adopt the role of observer (vs. protagonist), so it can be expected that they:
• H1a. Express a lower justification of violent behavior
• H1b. Perceive a greater severity of the situation
• H1c. Perceive a greater threat due to loss of power
• H1d. Identify controlling behavior more easily
• H2. Young men identify dating violence to a lesser extent when it takes place through WhatsApp (vs. face-to-face). Specifically, it can be expected that they:
• H2a. Express more justification of violent behavior
• H2b. Perceive a lower severity of the situation
• H2c. Perceive a lower threat due to loss of power
• H2d. Identify controlling behavior to a lesser extent
• H3. Ideological variables (ambivalent sexism, acceptability of IPVAW, and myths about romantic love) affect young men’s social perceptions of dating violence, so it is expected that men with high scores for these ideological variables express:
• H3a. A greater justification of violent behavior
• H3b. A lower perception of severity
• H3c. A greater perception of a threat due to loss of power
• H3d. A lower perception of controlling behavior

Method

Participants. The sample consisted of 120 male undergraduate students at the University of Granada, Spain. The age of the participants ranged from 18 to 34 years (M = 20.41, SD = 2.62). In the sample, a total of 95% of participants had Spanish nationality, 4.2% were immigrants, and 0.8% did not indicate their nationalities. Regarding their sexual orientations, the majority of the participants were heterosexual (89.2%), 6.7% were homosexual, 2.5% were bisexual, and, 1.7% indicated other sexual orientations. Concerning their civil statuses, 83.3 % were single, 15.8% were dating, and 0.8% did not indicate their civil statuses.

Design and procedure. The second study closely replicated the procedures and design of Study 1, adapting the scenarios of dating violence to young men (see Appendix B).

Instruments. The participants responded to the following scales:

Manipulation check. This element was tested through items designed in Study 1.

Perceived severity. The following item was used to evaluate this variable: “How severe do you consider the described episode?” (1 = nothing severe, 7 = very severe).

Justification of violent behavior. This element was measured using one of the following items, according to the experimental condition: “How justified do you consider Juan’s behavior to be” (observer condition) or “How justified do you consider your behavior to be” (protagonist condition), (1 = completely unjustified, 7 = completely justified).

Experiences of controlling Behaviors in their own relationships. These experiences were evaluated through the question: “How often have you suffered similar or equal situations in your relationships?” (1 = never, 7 = always).

Frequency of controlling Behaviors in young couples. The following item was used to evaluate this variable: “How often do you consider that these situations occur in young couples?” (1 = never) to 7 (always).

Perceived threat due to loss of power within the relationship. This variable was evaluated through one of the following items, according to the experimental condition: “To what extent do you think that Juan feels that his power within the relationship is threatened?” (observer condition) or “To what extent do you think that your power within the relationship is threatened?” (protagonist condition). It was scored on a 7-point Likert type scale (1 = nothing, 4 = a lot).

Perceived controlling behavior. Depending on the experimental condition, one of the following questions was used: “Is Juan controlling his partner?” (observer condition) or “Are you controlling your partner?” (protagonist condition). The response format was a 7-point Likert type scale (1 = nothing, 4 = a lot).

Acceptability of Intimate Partner Violence against Women Scale (A-IPVAW; Martín-Fernández et al., 2018). The alpha coefficient for the total scale was .63.

The Ambivalent Sexism Inventory (ASI; Expósito et al., 1998). The alpha coefficient of the hostile sexism subscale was .94; that of the benevolent sexism subscale was .84.

Myths Scale Towards Love (Bosch et al., 2007; adapted to an adolescent sample by Rodríguez-Castro et al., 2013). For this scale, the outcome was α =.63.

Demographic information. Students’ gender, age, sexual orientation, and marital status were measured at the end of the questionnaire.

Data Analysis. Data analysis was carried out using the SPSS program, version 23. Firstly, a chi-square test was used to assess manipulations’ efficacy. Next, in order to examine the frequency with which young men experience and perceive controlling behaviors in their relationships and amongst others young couples, descriptive analyses of frequencies were performed. Thereafter, several hierarchical regression analyses were performed with the objective of verifying the initial predictions about the effect of adopted role on the scene and means of control used on the social perception of controlling behaviors in the couple, considering the ideological variables.

Results

Manipulation check. The analysis revealed that all of the experimental manipulations were effective. Regarding the Adopted Role on the Scene IV, in MC₁, the results showed that 100% of the participants who were allocated to the observer condition indicated that the episode happened between Juan and María, whereas 48.39% of participants who were allocated to the protagonist condition indicated that the situation did not happen between Juan and María, so differences were statistically significant, χ²(1, 120) = 37.42, p < .001. The analysis showed a Cramer’s V coefficient = .56, so a large effect size was obtained. In Mc₂, it was found that 46.77% of men said that the situation was a hypothetical situation about their relationships when they belonged to the protagonist condition, whereas 93.1% of participants did not consider this to be a hypothetical situation when they belonged to the observer condition, χ²(1, 120) = 23.90, p < .001. A moderate effect size was obtained (Cramer’s V = .45).

Regarding the manipulation of the Means of Control IV, in MC₃, it was observed that 98.36% of participants who were allocated to the WhatsApp condition indicated that communication occurred via mobile phone, whereas 98.31% of men who were allowed to the face-to-face condition indicated that communication did not take place via a mobile phone, so the manipulation check was effective, χ²(1, 120) = 112.13, p < .001. The analysis pointed out a Cramer’s V coefficient = .97, so a large effect size was obtained. In MC₄, the results revealed that 96.61% of participants who belonged to the face-to-face condition considered that communication occurred in person, whereas 96.72% of participants who belonged to the WhatsApp condition indicated that communication did not take place in person, χ² (1, 120) = 104.53, p < .001. A large effect size was obtained (Cramer’s V = .93).

Frequency of controlling behaviors in relationships. To examine the frequency with which young men experience and perceive controlling behaviors in their relationships and amongst other young couples, descriptive analyses of frequencies were performed. The results showed that 92.5% of men declared that they had never or hardly ever exercised controlling behaviors in their relationships; however, 79.5% considered that these behaviors frequently happen within young couples.

Influence of adopted role on the scene, means of control used, and ideological variables on social perception of dating violence. For the purpose of analysing the influence of adopted role on the scene (Hypothesis 1) and means of control that were used (Hypothesis 2) on social perception of controlling behaviors in the couple, considering ideological variables, several hierarchical regression analyses were performed. The steps of Study 1 were closely replicated and the following dependent variables were entered: 1) perceived severity of the situation, 2) justification of violent behavior, 3) perceived threat due to loss of power within the relationship, and 4) perceived controlling behavior. The results obtained are shown in Table 2.

According to Hypothesis 1, the analyses pointed out that the Adopted Role on the Scene IV predicted the perception of threat due to loss of power within the relationship, β = .307, p ≤ .001, 95% CI [.131, .169], and identification of controlling behavior, β = .179, p = .036, 95% CI [.012, .347], which supported H1 (c and d). In this way, participants who were allocated to the observer condition predicted a greater perception of threat due to loss of power and a greater identification of controlling behavior, in comparison with participants who were allocated to the protagonist condition (see Table 2).

In the second step, regression analyses revealed the effect of a significant interaction between adopted role on the scene and acceptability of IPVAW on the measure of perceived severity, β = .272, p = .026, 95% CI [.033, .501], so in the protagonist condition, low levels of A-IPVAW predicted a greater perception of severity of the situation in comparison with high levels. However, in the observer condition, acceptability of IPVAW did not predict perceived severity (see Figure 3).

In addition, regression analyses showed the effect of another statistically significant interaction between the Adopted Role on the Scene IV and benevolent sexism on the measure of the justification of the controlling behavior, β = -.233, p = .04, 95% CI [-.445, -.010]. In the protagonist condition, low levels of benevolent sexism predicted a lower justification of controlling behavior in comparison to high levels; however, in the observer condition, benevolent sexism did not predict justification of controlling behavior in the couple (see Figure 4).

Regarding the Means of Control IV (face-to-face vs. WhatsApp) regression analyses did not show a significant simple effect on dependent variables (p > .05) (see Table 2), rejecting Hypothesis 2. However, the second step revealed the effect of a significant interaction between means of control that was used and benevolent sexism on the measure of perceived threat due to loss of power within the relationship, β = -.237, p = .049, 95% CI [-.477, -.001], so in the face-to-face condition, high levels of benevolent sexism predicted a greater perception of threat in comparison with low levels. In contrast, in the WhatsApp condition, benevolent sexism did not predict perceived threat due to loss of power (see Figure 5).

Lastly, consistent with Hypothesis 3, which sustained that ideological variables would affect young men’s social perceptions of dating violence, regression analyses showed a main effect of sexism, myths about romantic love, and acceptability of IPVAW on the dependent variables. Specifically, it was found that high levels of hostile sexism predicted a greater justification of violent behavior and a lower perceived severity of the situation. Additionally, high scores for benevolent sexism predicted a greater justification of aggressors’ behavior. Furthermore, high A-IPVAW predicted a greater justification of violent behavior and a lower perceived severity of the situation, as well as a lower identification of controlling behavior in the couple. Finally, high scores for myths about romantic love predicted high justification and low perception of severity.

Discussion

Study 2 replicated results found in Study 1 regarding the frequency with which young people experience and perceive control in relationships. Data showed that 92.5% of men declared that they had never or hardly ever exercised controlling behaviors in their relationships. However, 79.5% considered that this type of behaviors frequently happen within young couples. Consistent with findings of Study 1 and providing empirical support for previous research, this study pointed out that men frequently identify control in other young couples, but few of them recognize to exercise controlling behaviors against their partners (Donoso et al., 2016; Donoso et al., 2018).

As in Study 1, we predicted that men who adopted the role of observer on the scene (observer condition) would more easily identify violence against the partner compared with men who adopted the role of protagonist (protagonist condition). Main results showed that the Adopted Role on the Scene IV predicted perceived threat due to loss of power and identification of controlling behavior in the expected direction, which supported H1c. and H1d. Men who were allocated to the observer condition perceived a greater threat and a greater identification of controlling behavior, in comparison with men who were allocated to the protagonist condition. However, adopted role on the scene was not found to predict perceived severity and justification of violent behavior, rejecting H1 (a and b). Again, these results are consistent with results found in Study 1, so when controlling behaviors that occur in couples outside of their own relationships, men more easily identify these abusive behaviors and recognize to a greater extent that the perpetrator felt that his power within the relationship was threatened. It seems that male perpetrators tend not to identify violent behaviors exercised against their partners or the threat experienced within the relationship as an adaptive mechanism for reducing their psychological discomfort (Expósito, 2011).

On the other hand, the results pointed out the effect of a statistically significant interaction between adopted role on the scene and acceptability of IPVAW on perceived severity of the situation (see Figure 3). In the protagonist condition, lower scores for A-IPVAW predicted a greater perception of severity in comparison with higher scores; however, in the observer condition, A-IPVAW did not predict perceived severity. When men adopted the role of observer on the scene, the social norm was active and the situation was perceived as severe, as participants submitted above-average scores for this. However, when men adopted the role of protagonist, adaptive mechanisms were activated, so participants with high levels of A-IPVAW perceived the situation as less severe in comparison with participants with low levels of A-IPVAW, who rejected violence to a greater extent. In addition, the effect of a significant interaction between adopted role on the scene and benevolent sexism on justification of controlling behavior was found (see Figure 4). In the protagonist condition, low levels (vs. high levels) of benevolent sexism predicted less justification of controlling behavior; however, sexism did not predict this in the observer condition. When men adopted the role of observer on the scene, they graded above average, so they tended to justify controlling behaviors. In contrast, men who adopted the role of protagonist activated adaptive mechanisms, so when they scored low in benevolent sexism, they rejected the situation of violence more and justified the aggressor’s behavior less. Meanwhile, men with high levels justified the controlling behaviors to a greater extent.

On the other hand, as previous data pointed out, men identified to a greater extent controlling behavior and perceived threat due to loss of power within the relationship in other peer couples. However, at the same time, there appeared to be no effect of ideology on social perception of dating violence of men who adopted the role of observer. According to the data, they perceived the severity of the situation, but at the same time they justified it. This could indicate that they took on passive attitudes toward dating violence in cases where they were not directly involved, which happened in the situation of Juan and Maria. These results are consistent with the findings by Donoso et al. (2018) in a study with adolescents, where boys adopted more passive behaviors when they observed gender violence, in comparison with girls, who provided more helping behaviors when dealing with victims.

According to Hypothesis 2, it was expected that the Means of Control IV that was used (face-to-face vs. WhatsApp) predicted social perception of controlling behaviors in the relationship. Specifically, it was believed men who were allocated to the WhatsApp condition would identify control in the couple to a lesser extent than men who were allocated to the face-to-face condition. In opposition to our predictions, means of control that was used did not predict any dependent variables (perceived severity of the situation, justification of violent behavior, perception of controlling behavior, and threat due to loss of power), rejecting H2. These findings were consistent with results obtained in Study 1 and indicated that young men normalize and accept technologies as new ways of exercising controlling behaviors in the couple (Wright, 2017). Therefore, this type of behaviors is identified similarly in both contexts: technological (WhatsApp) and traditional (face-to-face).

Nevertheless, the results pointed out the effect of a significant interaction between means of control and benevolent sexism on the measure of perceived threat due to loss of power (see Figure 5). In the face-to-face condition, high levels (vs. low levels) of benevolent sexism predicted a greater perception of threat; however, in the WhatsApp condition, benevolent sexism did not predict a perceived threat. These results suggest that men experience threat due to loss of power when an episode of control occurs through WhatsApp, as they submitted above-average scores for this. However, the fact that benevolent sexism did not affect perception of threat seems to indicate that men have normalized these types of technological situations in their relationships, probably because they happen very frequently (Flores & Browne, 2017; Nardi-Rodríguez et al., 2018). Conversely, they are less accustomed to experience these situations face-to-face with their partners, so men with high score for benevolent sexism experience a greater threat due the loss of power, in comparison with those men with low benevolent sexism.

Finally, as in Study 1, data proved the influence of sexism, myths about romantic love, and acceptability of IPVAW on social perception of controlling behaviors against the partner, which supported Hypothesis 3 (a, b and c). Specifically, according to initial predictions, it was found that high levels of hostile sexism predicted a greater justification of violent behavior and a lower perceived severity of the situation, replicating the findings by Herrera et al. (2012) and Herrero et al. (2017). Additionally, in agreement with the results by Valor-Segura, et al. (2011), high scores for benevolent sexism predicted a greater justification of aggressors’ behavior. Consistent with findings by Martín-Fernández, Gracia, & Lila (2018) and Waltermaurer (2012), a high acceptability of IPVAW predicted a greater justification of violent behavior, as well as a lower perceived severity and a lower identification of controlling behavior in the couple. Finally, in agreement with the results by Redondo et al. (2011), high scores for myths about romantic love predicted a high justification of violence and low perception of severity. However, ideological variables did not predict perceived threat due to the loss of power within the relationship, rejecting H3d. Even so, these results revealed that high beliefs in ideological variables constitute an important risk factor of dating violence.

General Discussion

The present research aimed to explore young people’s social perception about controlling behaviors in relationships, analysing the influence of adopted role on the scene and means of control, as well as the effect of ideological variables (i.e., ambivalent sexism, acceptability of IPVAW, and myths of romantic love).

In an exploratory way, both studies pointed out that both women and men frequently perceived controlling behaviors in other young couples; however, few of them recognize suffering (women) or exercising (men) control in their relationships.

Regarding the Adopted Role on the Scene IV, on the one hand, women (Study 1) perceived a greater risk of dating violence when they adopted the role of observer (vs. protagonist) in the described episode; on the other hand, men (Study 2) identified controlling behavior and threat due to the loss of power within the relationship to a greater extent when they adopt the role of observer (vs. protagonist) on the scene of dating violence. Instead, an effect of the Means of Control IV on the perception of dating violence was not found according to the described episode in either study. Even so, these findings prove that controlling behaviors are reproduced through new technologies. Thus, due to the high frequency with which controlling behaviors occur in a technological context (Flores & Browne, 2017; Nardi-Rodríguez et al., 2018), these behaviors are normalized, with young people accepting what is common as normal. Finally, this research provides empirical support to previous studies as it demonstrates the influence of sexism (Gracia et al., 2014; Herrera et al., 2012; Herrero, et al., 2017; Martín-Fernández, Gracia, & Lila, 2018; Valor-Segura et al., 2011; Vidal-Fernández & Megías, 2014), myths about romantic love (Borrajo et al., 2015, Bosch & Ferrer, 2012; Nardi-Rodríguez et al., 2018; Rodríguez-Castro et al., 2013), and the acceptability of IPVAW (Martín-Fernández, Gracia, Marco et al., 2018; Waltermaurer, 2012) on both women’s and men’s perception about dating violence.

It should be noted that this study has several limitations that can most certainly be rectified in the future. An important issue is that it did not ask participants if they had previously received education or academic training on gender-based violence, which could affect their perception of dating violence in the described episode, so future studies should monitor this variable. Furthermore, the methodology of scenes represents another important limitation. Due to the impossibility of recreating real-life situations about dating violence, the degree of spontaneity, precision, and real experience that the hypothetical situations achieve should be treated with caution. In addition, as all participants are university students from Spain, future studies should try to work with heterogeneous samples that would allow for the possible influence of cultural values, age, and socio-demographics to be analyzed. Finally, when we assessed manipulations’ efficacy, we observed that questions that were used for checking the manipulation of the Role on the Scene IV (MC₁ and MC₂) generated confusion among participants. Specifically, we observed that some participants who belonged to the protagonist condition failed in MC₁ and MC₂ in both studies.. We believe it could be due to the fact that these participants did not identify themselves with the protagonist of the episode, either aggressor or victim. Consequently, although we indicated to them that it was a hypothetical scenario about their relationships, they failed in the experimental manipulation. Again, this could be a way to deny that these situations of violence are manifested in their relationships. Therefore, future studies should consider this limitation and evaluate these denial mechanisms that are used by the participants in situations of dating violence. However, despite these limitations, these studies contribute to previous literature, providing new information about the role of young people as observers of dating violence.

The present research evidences that when participants adopt the role of protagonist, women perceive the risk of suffering from dating violence to a greater extent, whereas men experience a lesser threat due to the loss of power within the relationship. Additionally, this research underscores the importance of ambivalent sexism and acceptability of IPVAW women as predictive variables in social perception of dating violence, specifically in perceived severity and justification of violent behavior. Moreover, it shows that myths about romantic love constitute an important risk factor of dating violence, as these myths predict a low perceived risk of dating violence amongst women and a high justification of controlling behavior amongst men. Likewise, given the fact that both studies demonstrated the influence of ideological variables on social perception of dating violence online, this research contributes to previous literature by demonstrating the importance of using ICT as a tool for combating sexism and educating on equality, just as Navarro-Pérez et al. (2019) recently tested in their research. In short, this research reveals the need to develop intervention programs that are based on risk perception of dating violence, addressing the problem from a broad gender perspective, which includes the importance of observers as key figures in the confrontation of violence against women.