Servicios personalizados
texto en 
Inglés (pdf)
Articulo en XML
Referencias del artículo
Traducción automática
Enviar articulo por email
Citado por SciELO 
Citado por Google 
Similares en
SciELO 
Similares en Google 
Permalink
Introduction
Intimate Partner Violence Against Women (IPVAW) is the most frequent form of violence suffered by women (Devries et al., 2013; World Health Organization, 2016), with severe consequences for their physico-mental health (Campbell, 2002; Ellsberg, Jansen, Heise, Watts, and Garcia-Moreno, 2008; Vilariño, Amado, Vázquez, & Arce, 2018). The average prevalence of lifetime intimate partner violence is 22% in the European Union and, in Spain, where this research was conducted, it is 13%, which is among the lowest in the European Union (European Union Agency for Fundamental Rights, 2014; Gracia, Martín-Fernández, Lila, Merlo, & Ivert, 2019; Martín-Fernández, Gracia, & Lila, 2019, 2020). In 2018, Spanish courts received 166,961 complaints of IPVAW, of which 70.5% were convictions (158,590 women were victims of Intimate Partner Violence). Regarding the complaints filed by women victims of Intimate Partner Violence, 49,904 (31.5%) were made by immigrant women (Observatorio contra la Violencia Doméstica y de Género y Consejo General del Poder Judicial, 2019). These data highlight the possibility of some differential factors of Intimate Partner Violence appearing in the immigrant population (Cummings, González-Guarda & Sandoval, 2013; Vargas, Lila, Catalá-Miñana & Gracia, 2017).
Batterer Intervention Programmes (BIPs) form part of the institutional response to combat IPVAW (Gondolf, 2012). Ever since these programmes were set up, several studies have been conducted to evaluate their effectiveness, but no consensus has been reached about the intervention effect on changing batterers’ behaviour (Bowen, 2011; Gondolf, 2012). Several meta-analysis and systematic reviews have attempted to solve this problem by compiling the results of available research in this field (e.g., Arango et al., 2014; Arce, Arias, Novo & Fariña, 2020; Babcock, Green, & Robie, 2004; Eckhardt et al., 2013; Feder & Wilson, 2005). The main findings show small and/or moderate size effects, and problems hindering their effectiveness, such as the high dropout rates and lack of adherence to treatment (Arango et al., 2014; Eckhardt et al., 2013).
One of the proposed recommendations to improve the outcomes of these programmes is to adjust them to participants’ individual characteristics (Juarros-Basterrectxea, Herrero, Fernández-Suárez, Pérez, & Rodríguez Díaz, 2018; Weber, Taylor, Cantos, Amado, & O’Leary, 2019). Taking into account the diversity among intervention groups, such as their cultural origins, motivational level, addictions and other mental health problems, will diminish resistance to treatment, increase motivation for change and adapt the intervention to participants’ needs (Babcock et al., 2016; Catalá-Miñana, Walker, Bowen, y Lila, 2014; Eckhardt et al., 2013; Yáñez-Peñúñuri, Gómez, & Anacona, 2019).
The cultural adaptation of BIPs is considered one of the possible strategies to improve intervention effectiveness in men from both ethnic minorities and immigrants. Nevertheless, current scientific evidence is controversial. On the one hand, several studies have found that cultural adaptation lowers high dropout and recidivism rates, and increases these participants’ attendance to programmes (Gondolf, 2004; Taft, Murphy, Elliott & Keaser, 2011; Waller, 2016; Welland & Ribner, 2010). In line with this, and by focusing on the Latin American immigrant population, some authors suggest they may not feel comfortable and be unable to connect to these conventional programmes (i.e., cognitive-behavioural with a feminist orientation) because programmes do not pay attention to the migration process, and the cultural shock that some topics covered in the intervention might trigger (e.g., sexist attitudes, patriarchy, gender roles) and different males’ ‘masculinity’ ideal (Aldarondo, Kaufman & Jasinski, 2002; Falconier et al., 2013; Hancock and Siu, 2009).
On the other hand, some researchers state that there is insufficient evidence for the effectiveness of culturally-oriented programmes and suggest that conventional programmes would be equally effective for immigrants and ethnic minority groups (Murphy & Ting, 2010; Parra-Cardona et al., 2013). According to these authors, training in skills that contribute to reduce violence and attitudes supporting it may be useful for the majority of males regardless of their cultural origins (Buttell & Carney, 2005). From this perspective, the study conducted by Rothman et al. (2007) found no differences in the dropout rates between the Caucasian and Latin American immigrants who participated in a conventional programme and a culturally-oriented one, respectively. Rather than cultural specificity, this research line suggests that it would be sufficient to address participants’ individual characteristics and certain cultural sensitivity during interventions (Rothman et al. (2007). Therefore including some strategies, such as Motivational Interviewing, stages of change and adherence techniques, could lower dropout rates and resistance to treatment in these groups by, thus, increasing the probability of interventions being successful (Alexander, Morris, Tracy & Frye, 2010; Crane and Eckhardt, 2013; Lila, Gracia & Catalá-Miñana, 2018).
In addition to inconsistent evidence for the effectiveness of culturally-oriented BIPs in both Spain and elsewhere, very few studies have analysed the role of cultural belonging in BIPs despite the marked presence of men from various ethno-cultural origins, especially from the Latin American culture (Echauri, Fernández-Montalvo, Martínez & Azkarate, 2013; Fernández-Montalvo, Echauri, Martínez, Azcárate & López-Goñi, 2015; Vargas, Lila & Catalá-Miñana, 2015; Vargas, Lila, Catalá-Miñana & Gracia, 2017; Welland & Ribner, 2010). This work aimed to determine whether there were differences in the effectiveness of a standard BIP in Spain (i.e., cognitive-behavioural with motivational strategies and with no specific cultural adjustment) by comparing the treatment outcomes of Spanish and Latin American immigrant batterers.
Velonis, Cheff, Finn, Davloor and O’Campo (2016) categorised treatment outcomes as final (i.e., the intervention’s expected outcome) and proximal (i.e., variables that intervene during treatment and affect the final result). Following their classification, the final outcome herein considered was recidivism because its reduction is one of the main objectives to be accomplished in BIPs (Lila et al., 2018). As far as proximal outcomes are concerned, variables previously found to be related to the intervention’s effectiveness and in reducing recidivism were included. Hence the proximal outcomes herein considered were intention to change (i.e., attitude towards intervention and motivation for change, resistance to change and stage of change; Carbajosa, Catalá-Miñana, Lila, Gracia & Boira, 2017; Lee, Uken & Sebold, 2007), responsibility attribution (Lila, Gracia & Herrero, 2012; Lila, Oliver, Catalá-Miñana, Galiana, & Gracia, 2014), adherence to treatment (i.e., dropout and treatment attendance; Carbajosa, Catalá-Miñana, Lila & Gracia, 2017; Gordon & Moriarty, 2003) and treatment compliance (i.e., treatment involvement and treatment satisfaction; Gondolf, 2012; Hancock & Siu, 2009; Santirso, Martín-Fernández, Lila, Gracia, & Terreros, 2018).
Methods
Sample
The sample consisted of 425 male batterers convicted of Intimate Partner Violence in the province of Valencia (Spain) according to their condition of attending a BIP. The inclusion criteria required for participation in the study were: being an adult male older than 18 years; having been convicted of an IPVAW crime; having Spanish or Latin American nationality; having participated in at least the assessment phase and first interview in the intervention programme. The inclusion criteria of those candidates considered eligible to participate in the programme were: not presenting severe personality or mental disorders; not having severe alcohol and other drug dependence types: not displaying disturbing behaviours that could disrupt the functioning of the intervention. Among the 425 men who participated in the assessment phase and the first interview, 343 completed the programme (completers) and 82 dropped out (dropouts). The participants’ ages ranged from 18 to 78 years old, with a mean age of 39.58 years (SD = 12.08). Regarding their country of origin, 324 males were Spanish (76.2%) and 101 (23.8%) were Latin American. The most represented countries of origin in the Latin American group were Ecuador (n = 45; 44.6%), followed by Bolivia (n = 19; 18.8%) and Colombia (n = 17; 16.8%). The socio-demographic characteristics of all the groups are presented in Table 1.
Measures
Final outcomes
Recidivism. The official recidivism rate was taken from the VioGén database, an Intimate Partner Violence monitoring system of the Spanish Home Office (López-Ossorio, González-Álvarez and Andrés-Pueyo, 2016; López-Ossorio et al., 2018). This system includes data from any new IPVAW incident or breach of court-ordered conditions (i.e., restraining order) committed by the participants after the initial assessment phase and the first interview. The follow-up period for recidivism was 10 years, which took into account the starting date of the first group of participants included in this sample. Variables were coded as 0 (if the participants did not reoffend) and 1 (if they reoffended).
Risk of recidivism. Spousal Assault Risk Assessment Guide (SARA; Kropp and Hart, 2000; Spanish version by Andrés-Pueyo and López, 2005). This test is a 20-item protocol employed by professionals to assess the risk factors of Intimate Partner Violence (IPV) re-offending.
Proximal outcomes
Intention to change. It was assessed by professionals based on two ad hoc measures intended to evaluate participants. The Attitudes toward Intervention and Motivation for Change Scale is composed of 10 items divided into two factors: (1) Attitude towards the intervention (7 items), which assesses participants’ disposition towards the intervention and its professionals. Alpha’s Cronbach coefficient for Time 1 and Time 2 was .89; (2) Motivation for change (3 items), which assesses participants’ disposition to change their behaviour and attitudes associated with increased responsibility attribution and raised awareness of their offense and its negative consequences. The internal consistency of this second factor was .79 and .89 for Time 1 and Time 2, respectively. Items were rated on a 5-point Likert-type scale (1 = Negative/low presence of the attitude or behaviour; 5 = Positive/high presence). A higher score implies a better attitude towards the intervention. Alpha’s Cronbach coefficient was .90 for the first factor and .80 for the second one. An item was also designed to assess each participant’s stage of change on a Likert-type scale which includes five stages of change: 1= Precontemplation; 2= Contemplation; 3= Preparation; 4= Action; 5= Maintenance (see Carbajosa et al., 2017 for a similar approach).
Responsibility attribution. The Intimate Partner Violence Responsibility Attribution Scale (IPVRAS; Lila, Oliver, Catalá-Miñana, Galiana and Gracia, 2014). The IPVRAS is a 12-item scale designed to assess where offenders situate the responsibility of the act for which they had been convicted. It is a 5-point Likert-type scale on which participants rate each item according to their degree of agreement (1 = totally disagree; 5 = totally agree). It comprises three dimensions with four items each, which correspond to three possible attributional causalities: (a) Responsibility attributed to the legal system (e.g., ‘I’m here because of an injustice’); (b) Responsibility attributed to the victim (e.g., ‘my partner’s behaviour and the way she treats me are the main reasons why I am in this situation’; (c) Responsibility attributed to the offender’s personal context (e.g., ‘ Alcohol or other drugs use is the reason why I am in this situation’). Cronbach alpha ranged from .60 to .66 for Time 1 and from .61 and .70 for Time 2.
Adherence to treatment. Two measures were used to assess this variable, dropout (0 = dropout, 1 = completion) and treatment attendance, in which the proportion of programme sessions that each participant attended was calculated in relation to the intervention’s total sessions. It ranged from 0 to 1, where 1 = attendance at all the sessions. Attendance average was .8 with a standard deviation of .25.
Treatment compliance. It includes treatment involvement and satisfaction with intervention. Treatment involvement was evaluated considering homework activities that were later used during the group session. It was calculated by dividing the activities performed by each participant by the total activities required during the group session. It ranged from 0 to 1, with 1 equalling the delivery of all activities, whose average was .63 (SD=.3). For satisfaction with intervention, the Satisfaction with Intervention Scale, designed ad hoc, is a 21-item scale that assesses participants’ satisfaction at the end of treatment. It comprises two dimensions: 7 items assessing Programme Satisfaction (e.g., ‘I have learned new things in this programme’) and 14 items for Satisfaction with the intervention group (e.g., I would like to meet with the group again some other time). The first dimension assesses the changes and knowledge that participants consider they have acquired at the end of treatment. The second dimension estimates users’ satisfaction with the way they were treated, and with the relationship established with professionals and partners during the programme. Items were answered on a 5-point Likert-type scale (1= Totally disagree; 5= Totally agree). A higher score implies more satisfaction. Apha’s Cronbach was .90 and .75 for the first factor and the second factor, respectively.
Procedure
Most of the self-report questionnaires were answered by the participants at two different time points: before treatment (T1: baseline) and at the end of treatment (T2). The Satisfaction with Intervention Scale was completed only after treatment (T2). Confidentiality and anonymity were guaranteed. Written informed consent was requested. The sample’s official recidivism record was taken from the Spanish Home Office’s database (VioGén system). Information was also collected from the professionals’ assessment of three variables: (1) attitudes towards intervention and motivation for change; (2) risk of recidivism; (3) stages of change. Professionals’ assessment was obtained before and after group treatments. This research was conducted by taking into account the protection of personal data and following the current guidelines of the Ethics Committee of the University of Valencia.
Data Analysis
The scores obtained by both the Spanish and Latin American groups were compared to assess the treatment outcomes at the intervention’s two different time points. Regarding the final outcomes, a Chi-square test was performed to assess official recidivism. To examine risk of recidivism, repeated-measures ANOVAs were conducted by taking the intervention time point (T1 and T2) as the within-subject factor, and the group’s country of origin (Spanish and Latin Americans) as the between-subject factor.
Regarding the proximal outcomes, repeated-measures ANOVAs were performed to analyse the variables intention to change (i.e., attitudes towards intervention, motivation for change, and stages of change) and responsibility attribution by taking the intervention’s time points as the within-subject factor and the group’s country of origin as the between-subject factor. The measures included in the variable adherence to treatment were examined by a Chi-square test for dropout, and a one-way ANOVA for treatment attendance. Finally, in order to assess treatment compliance, a one-way ANOVA was conducted for each dimension: treatment involvement, satisfaction with the intervention group, and programme satisfaction.
Results
Final outcomes
No significant differences appeared between the Spanish and Latin American batterers for official recidivism: χ²(1) = .84, p > .05; Cramér’s V = .045. The percentage of official recidivism in the total sample was 7.8% (n=33). However, significant differences were found in risk of recidivism when comparing pre- and post-treatments (within-subject factor), F(1, 329) = 84.87; p < .05; (2 = .205. Thus, the risk of recidivism in both groups reduced after treatment was completed. Furthermore, the interaction effect between the main effect (risk of recidivism) and the between-subjects factor (Spanish or Latin American immigrants) was not significant, (F(1, 329) = 0.27; p > .05; (2 = .001), which means that changes in this variable are independent of the batterers group to which individuals belong (see Table 2).
Proximal outcomes
For intention to change, differences in the pre- and post-treatment scores were statistically different for attitudes towards intervention, motivation for change and stage of change. At the end of treatment, the statistical analysis showed a significant change in all these variables, which meant that the individuals who completed the intervention showed a better attitude towards intervention, more motivation for change and moved forward in stages of change. No significant differences were observed in the interaction between the within-subjects factor effect (Spanish and Latin American immigrants) and the intention to change dimensions. Regarding responsibility attribution, the statistical analysis showed significant differences for the main effect in the dimensions responsibility attribution to the legal system and to the victim. After treatment, those individuals who completed the intervention showed diminished responsibility attribution to both the victim and the legal system, although no significant differences were found between the intervention time points when contemplating the responsibility attribution to the personal context variable. Interactions were not significant, which meant that there were no differences between groups according to their country of origin (see Table 3).
As far as the adherence to treatment dimensions are concerned, a Chi-square test revealed that there were no differences in dropout between groups for their country of origin, χ²(1) = 3.538; p > .05; V = .091). For treatment attendance, a one-way ANOVA (see Table 3) did not show any differences between the Spanish and Latin American participants after treatment: F = 3.36; p > .05; (( = .088.
Finally, in treatment compliance terms, significant differences appeared in the treatment involvement between the Spanish and the Latin American immigrant groups: F = 5.76; p < .05; (2 = .014. Spanish batterers delivered more homework activities during the intervention than the Latin American immigrants. For satisfaction with intervention, the univariate analysis showed no significant differences in satisfaction with the intervention group between the Latin American immigrants and Spanish batterers: F = 0.86; p > .05; (( = .002. Nevertheless, significant differences were found for programme satisfaction: F = 12.60; p < .001; (( = .040. The Latin American immigrants showed more programme satisfaction than the Spanish batterers. Finally, the data generally revealed small to moderate size effects (see Table 3).
Discussion
Several recommendations have been made by the scientific community to improve BIP outcomes (Babcock et al., 2016; Bowen, 2011; Gondolf, 2012). Specifically, some researchers have indicated the need to culturally adapt these programme for ethnic-minority men and immigrants (Buttell & Carney, 2005; Gondolf, 2004; Murphy & Ting, 2010). In order to move forward in this field as regards the benefits of these interventions for BIP participants belonging to different countries, this study aimed to compare the final and proximal outcomes achieved by a group of Spanish batterers and a group of Latin American immigrant batterers after a non-culturally adapted intervention.
For the achieved final outcomes, there were no significant differences in official recidivism between the Spanish participants and the Latin American immigrants who completed the intervention. Thus, both groups presented similar recidivism rates, with a relatively low percentage of recidivism for the total sample (7.9%). This result is consistent with previous research conducted in Spain with immigrants, in which similar results were found for therapeutic success, improvement, and recidivism rates in Spanish and immigrant batterers after completing treatment (Echauri et al., 2013). No significant differences were found between the Spanish batterers and the Latin American immigrant batterers for risk of recidivism. At the end of treatment, a lower risk of recidivism was obtained for both groups (the sum of risk factors and the global risk assessment of both), which is consistent with previous research showing a reduction in the risk of recidivism for BIP completers (Echauri et al., 2013; Lila, Martín-Fernández, Gracia, López-Ossorio & González, 2019).
As far as the proximal outcomes are concerned, significant differences were found between the beginning and the end of the intervention, irrespectively of the group to which participants belonged, and for the following dimensions of intention to change: attitudes towards intervention, motivation for change, and stage of change. Attitudes towards intervention, which assesses participants’ disposition towards the programme and professionals, positively varied at the end of treatment. Participants (Spanish and Latin American immigrant batterers) showed lower aggressiveness levels, fewer complaints, and higher levels of positive interaction, communication, cooperation and respect. Similarly, positive progress was made in motivation for change after the intervention: the Latin American immigrants and Spanish batterers were more willing to change their behaviour and attitudes towards greater responsibility attribution and reported being more aware of their offence and its consequences. In relation to stage of change, both groups progressed in stages while the intervention was underway, with no significant differences between them. It is worth mentioning that previous research has associated progress made in the variables related to intention to change with a greater responsibility attribution towards their violent behaviour which, thus, influences the expected final outcome of such interventions; reduction of recidivism (Scott, King, McGinn & Hosseini, 2011; Velonis, Cheff, Finn, Davloor & O’Campo, 2016).
For responsibility attribution types, responsibility attribution to the legal system and to the victim showed a reduction in both groups. Lower levels of responsibility attribution to external factors, such as the legal system and victims, have been associated with a reduction in resistance mechanisms and more motivation for change (Gracia, 2014; Gracia, Rodríguez & Lila, 2015; Levesque, Velicer, Castle & Greene, 2008; Martín-Fernández, Gracia & Lila, 2018; Martín-Fernández, Gracia et al., 2018). These previous findings are consistent with the results obtained in this study for indicators of change. Nevertheless, the differences in attribution to the personal context were not significant after the intervention. This could be because the responsibility attribution to specific circumstances, which influences batterer’s behaviour (e.g., substance abuse, stressful life events), would demand having to accept individual circumstances to make more efforts in their personal area, which is not specifically addressed by the intervention programme (Capaldi, Knoble, Shortt & Kim, 2012; Catalá-Miñana, Lila & Oliver, 2013; Catalá-Miñana, Lila, Oliver, Vivó, Galiana & Gracia, 2017; Lila, Gracia & Murgui, 2013).
For adherence to treatment, no significant differences were found in either dropout or treatment attendance in the Spanish and Latin American immigrant batterers. This finding is inconsistent with a study conducted by Echeburúa, Sarasua, Zubizarreta, Amor and Corral (2010) in Spain, in which being an immigrant was a predictor of dropout or resistance to treatment. However, the major differences between both these studies could explain these controversial results. Our sample included court-mandated batterers who did not voluntarily attend the programme, and included Latin-American immigrants with an immigrant percentage of 23.8% of the whole sample (n=101). In the study by Echeburúa et al. (2010), all the participants attended the programme voluntarily, and the percentage of immigrants whose countries of origin were not specified was much lower (8%; n=36). This is probably because the intervention was mandatory for the participants in our study, as well as the immigrants’ Latin American provenance who, therefore, shared the same language and other similarities with the host country, which could explain why the differences between immigrants and natives were not significant in treatment adherence terms.
Finally, regarding treatment compliance (i.e., treatment involvement and satisfaction with intervention), no significant differences appeared between the Spanish and Latin American immigrant batterers in their satisfaction with the intervention group; that is, the received treatment and the established relationship with the programme’s professionals and peers. The variables showing significant differences between the Latin American immigrants and Spanish Batterers were treatment involvement and programme satisfaction. The Latin American immigrants participated less than the Spanish participants (i.e., delivered fewer homework activities for reflection) and showed more programme satisfaction by expressing more satisfaction with the accomplished changes and knowledge gained by the end of treatment. These results suggest that the Latin American immigrant batterers better accepted treatment than the Spanish batterers for a standard intervention. This finding contrasts with the results obtained in the study by Hancock and Siu (2009) in the USA, in which Latin American immigrant participants showed more acceptance to a culturally adapted programme than a standardised one. More programme satisfaction is not consistent with the lower participation levels for the Latin American immigrant batterers. Further research is necessary to clarify this apparent inconsistency between both results (more satisfaction vs. a lower participation rate in the Latin American immigrants).
This study has certain limitations, of which the main one is lack of equivalence in sample size. Although this study, versus similar ones, recruited a significant number of Latin American immigrants, some analyses required having to adjust sample size to test and meet assumptions, which led to a small sample size for both groups. A second sample-related limitation was that these findings cannot be generalised to all men convicted of intimate partner violence (e.g., batterers serving a prison sentence). Likewise, the immigrant batterers sample size was composed only of Latin Americans, which means that these results can only be generalised to this immigrant population. However, other immigrant populations with a high migration rate in Spain were not herein analysed, such as Romanian and Moroccan populations who, like the Latin American immigrants, present high participation rates in BIPs. In line with this, it is worth mentioning that the Latin American immigrant population, despite its similarities, is not a homogeneous group. Differences between them can appear. In study design terms, the herein proposed comparison of T1 to T2 does not guarantee that any changes taking place during the intervention are due to treatment and not to unpredictable external factors. Hence for future research purposes, we suggest using an experimental design that allows, for instance, comparisons between outcomes in standard BIPs and culturally-adapted programmes for Latin American immigrants. In this population type, two systematic sources of error in the responses attributable to the subject, and not to the construct, are suspected. On the one hand, dissimulation, consisting of false assumptions of positive characteristics to obtain a benefit, e.g. adherence to treatment, intention to change, responsibility attribution and treatment compliance (Arce, Fariña, Seijo & Novo, 2015). On the other hand, simulating damage that is also related to obtaining prison benefits (American Psychiatric Association, 2013; Arce, Fariña, & Vilariño, 2015). Both sources of bias pollute the results in the proximal outcomes to some extent. Finally, the effect size in most analyses ranged from low to moderate, which is consistent with systematic reviews and meta-analyses in this field, where the effectiveness of intervention programmes has been assessed (Arce et al., 2020; Babcock et al., 2004; Feder & Wilson, 2005; Santirso, Gilchrist, Lila, & Gracia, 2020).
Despite these limitations, this study offers important practical implications for BIPs with Latin American immigrant groups. Although possible differences appear between the batterers’ groups in culture of origin terms, the treatment outcomes were similar in most assessed variables. Our study results suggest that by having strategies that promote motivation for change and taking into account individual characteristics, which was the case in the evaluated programme, probably suffice to improve the effectiveness of BIPs for different participant groups regardless of their origin.
