**1. Introduction**

Intimate Partner Violence (IPV) is one of the most frequent types of domestic violence. IPV is characterized by various forms of abusive (physical, psychological, sexual), dominance and control behaviors perpetrated by one partner over the other in the context of an intimate relationship (Burgess et al. 2010). IPV exists in all socioeconomic, religious, and cultural contexts/groups, regardless of sexual orientation, age, or sex, although an unequal gender distribution is recognized (Burgess 2017); gender assumptions are culturally legitimated and influence individual and social discourses on risk as well as informing the policies and practices of risk assessment (Walklate 2018).

Research on IPV has developed a recognition of the factors and processes that contribute to the key decision by victims of abusive relationships of whether to stay or leave their partner (Anderson and Saunders 2003; Barrios et al. 2020). This issue is particularly important in models and intervention strategies aimed at the recovery of the victim of violence (Peterson 2020) and also contributes to the general health and well-being of individuals after experiencing violence (Hamby et al. 2017).

The decision to leave an abusive relationship is very difficult for victims to make, because they must consider their own safety as well as the safety of their children (Jones and Vetere 2017; Peterson 2020; Stephens 1999). The leaving process is complex and contingent upon numerous contexts, such as individual factors (e.g., clinical condition of the victim), interpersonal factors (e.g., support of their family, the needs and safety of their children, the severity of IPV), and community or sociocultural factors (e.g., the existence of victim support services, cultural beliefs about gender roles, coercive control, religious tradition) (Barrios et al. 2020; Domenech del Rio and Valle 2019; Estrellado and Loh 2019; Jones et al. 2017; Sears 2018; Wilhelm and Tonet 2007). The process of leaving the abusive relationship is influenced by a series of micro and macro variables that affect each individual. The process is more abrupt for some victims than for others. The decision may arise in different stages depending on the weighting that different aspects contribute to the decision to leave (Enander and Holmberg 2008; Khaw and Hardesty 2013). Some victims face grea<sup>t</sup> psychological difficulties during the decision-making process to leave and/or after leaving the abusive relationship, and their condition may worsen over time (Anderson and Saunders 2003). Therefore, it is important to know the factors that can affect the decision to leave, as well as the predictors of victim well-being, so that when there is a request for help, the planning for an intervention can consider various factors that can help the victim face adversity, particularly their resources and strengths. We note that some studies in this area recognize the importance of such factors, but do not present them within a model or go on to discuss how they can inform interventions with victims of IPV. This work intends to fill that gap.

This article presents an empirical qualitative, descriptive, and cross-sectional study, which observed the experience of victims of IPV who are mothers. Based on the narratives of victims, the research identifies the main factors underlying victims' decisions to leave abusive relationships. We find that these factors are also fundamental for consideration in the victim support process. After a brief literature review on the issue of leaving an abusive relationship, we focus on the elements of The Resilience Portfolio Model (Grych et al. 2015) to reflect on the process of psychosocial support to victims of IPV. Based on victims' experiences, and through the use of a qualitative and participatory (grounded) research approach, we move toward theory. Although not an exhaustive consideration, we attempt to underline the importance of psychosocial intervention in this area.

### **2. Reasons for Leaving the Abusive Intimate Relationship**

There are many theoretical contributions to the understanding of the process of abandoning an abusive intimate relationship (Anderson and Saunders 2003; Barrios et al. 2020): from Freudian explanation, based on conceptions of relational masochism for not leaving the abusive relationship (Snell et al. 1964; Young and Gerson 1991) and traumatic bonding theory (Dutton and Painter 1993); through critical approaches from feminism or social psychology that emphasize the role of gender in the understanding of violence against women (Andrade 2018; Butler 2004; Deutsch 2007; Wuest and Merrit-Gray 2001); up to the current debate on violence as an attack on human rights (McLeod et al. 2020; Otto 2019). Given that no single theory has been able to predict which reasons weigh most heavily on the decision to leave an abusive relationship, some empirical studies, mainly quantitative, have tried to isolate variables that have some explanatory power (Domenech del Rio and Valle 2019; Rhodes and McKenzie 1998). Research on leaving abusive relationships identifies various personal, interpersonal, and societal factors that are classified in various ways in different studies (Anderson and Saunders 2003; Baly 2010; Gelles 1976; Jones and Vetere 2017; Khaw and Hardesty 2013).

For Gelles (1976) three important factors must be observed: (i) the severity of violence (the less severe, the greater the tendency to stay in the relationship); (ii) history of childhood abuse (such a victim is more likely to remain in the relationship); (iii) material resources of the victim (the fewer resources, the more likely they are to stay). Rhodes and McKenzie (1998), reviewing three decades of studies, concluded that although some of these factors may be important in the decision to leave, the results are contradictory or unclear and that other variables can be considered, for example, individual factors

(e.g., personality characteristics, self-esteem, psychological effects of the abuse). However, many of the studies reviewed by these authors were mainly concerned with recognizing what was wrong with the victims, rather than focusing on the behaviors and attitudes that can favor the leaving process. We corroborate the opinion of Rhodes and McKenzie (1998) when they conclude that there was a tendency to look for a theory that fits all victims and not to try to understand their inner worlds and patterns of behavior in their relationships. Likewise, we support the position that interventions should be tailored to fit each case, attending to the individual experience of each victim of IPV, as advocated by other studies (Khaw and Hardesty 2013).

Anderson and Saunders (2003) mention two broad categories of predictors for leaving an abusive relationship: material resources (e.g., employment and income) and social–psychological factors (e.g., history of childhood abuse). Khaw and Hardesty (2013) theorize that sociocultural factors intersect and interact with individual factors (e.g., informal social support), and the interplay of these elements creates unique circumstances affecting the victim's access to formal support resources (e.g., the police) and the strategies they choose to engage in leaving. We prefer to refer to the existence of intrinsic and extrinsic factors to the victim if we consider the conceptual classification of locus of control (internal vs. external) (Rotter 1966) that can determine decision-making.

Baly (2010) underlines the importance of analyzing the victim's narratives. Through a qualitative study with women who had abandoned abusive relationships, she concluded that meaning-making about the abusive situation was influenced by social and cultural discourses in different ways: some constructions reduced the victim's ability to deal with the situation, while other discourses encouraged a self-construction of personal strength to leave the abusive situation.

We need to highlight that the decision to leave is the result of a process that may include several phases, during which the victim learns how to face the adverse situation of IPV. The process is multidimensional and variable: some victims first prepare a cognitive and emotional leaving (with primary recognition, resistance, and risk management) before the physical leaving (already aware of their own needs) (Anderson and Saunders 2003), while others may experience a faster cut of the abusive relationship. Preparation is more than just a step within the larger process of leaving that must be actively planned (Bermea et al. 2020). Closer to the decision to leave, some factors can be decisive (e.g., the existence of family or friends support) and even after leaving the relationship, other factors continue to contribute (e.g., education of children), which may reinforce the decision to leave or may compromise it, resulting in a reversal of the decision. For example, motherhood can be a source of empowerment, which helps victims care for their children and survive abusive relationships, or a source of pressure to leave abusive situations (Jones and Vetere 2017; Sani et al. 2020; Zink et al. 2003). The dissonance between wanting the best for their children and fearing the involvement of child protection services hinders decision-making by victims of IPV (Sani and Carvalho 2018; Zink et al. 2003) and thus should be considered in planning for supportive interventions.

After leaving an abusive relationship the victim needs practical assistance (Anderson and Saunders 2003). For example, how will the victim ensure their psychological well-being in the post-separation period? This is a question that professionals who are in the line of support to the victim necessarily ask, even before the abusive intimate relationship ends. This reinforces the idea that leaving an abusive relationship is a process that begins before a decision is made and does not end after the decision to leave. Therefore, various intrinsic (e.g., meaning-making, self-regulatory) or extrinsic forces (e.g., social support, material resources) on the victim need to be identified and worked on to protect against negative psychological outcomes. Sometimes the same factors that are identified as predictors of leaving are variables that require additional work to enhance an effective support intervention for victims of IPV (e.g., the existence of young children, recognition of informal social support).

Finally, based on the results of our study and considering the implications of these for the intervention process to support victims of violence, we will defend the adoption of a positive approach centered on the search for the individual's strengths, which can also be found in a model that we will present below.

### **3. Intervention with IPV Victims: Three Components for Resilience**

Research on victims has enabled important developments to be achieved on interventions to violence and crime (Hamby et al. 2017; Sani and Caridade 2016, 2018). In this context, a paradigm shift has been advocated that allows a better understanding of violence, adopts an integrated approach to the phenomenon, takes into account the risk factors (static and dynamic), and also considers the mechanisms that underlie violence and victimization (Ahmadabadi et al. 2018). Accordingly, intervention processes must focus on the identification and promotion of the strengths or skills evidenced by individual victims (Hamby et al. 2017).

In the following, we briefly present The Resilience Portfolio Model (Grych et al. 2015), a conceptual approach that aims to identify the poly-strengths or protective factors of the person that can contribute to their health and well-being after exposure to stressful situations. According to the model, there are three major strengths associated with well-being: self-regulation (ability to sustain motivation and overcome obstacles while striving toward a goal); interpersonal forces (social support); and meaning-making (capacity to find meaning in di fficulty). The gathering of forces from these three functional domains can promote the necessary resilience for the prevention of violence and the development of well-being, as demonstrated in some studies (e.g., Banyard et al. 2016; Taylor et al. 2016).

In this model (Grych et al. 2015), sometimes the factors that promote a person's resilience are simply the reverse of the risk factors (e.g., the non-existence vs. existence of a recognized support system); however, an e ffort is needed to search for the multi-forces that may be more implicit. Additionally, a more positive view of the goal of the intervention process is that it would be defined not by the absence of pathology, but by reaching levels of well-being and health. In the domain of intervention with victims of violence and specifically the decision of staying in or leaving abusive relationships, this model has particular importance when assuming decision-making as a process in which the identification and mobilization of protective factors contribute to the victim's well-being (subjective evaluations of satisfaction).

Following the tradition of research supported by qualitative approaches that underline the importance of understanding individuals' experiential reality, this study aims to contribute to the knowledge of factors associated with leaving an abusive relationship, and on this basis we reflect on interventions in support of the victim of IPV.
