The relationship between length of stay in shelters, post-shelter destination site, and type of violence, emotional state, expectations and experiences from the shelter.

There is a worldwide steady increase in the prevalence of violence against women.  This is a widespread social phenomenon with severe implications on health and emotional wellbeing. Violence against women refers to women who are involved in relationships that are characterized by terror and violence (Krishnan, Hilbert, Mcneil, & Newman, 2004).

The decision to leave a violent spouse represents a more complex decision which is affected by the environmental, social and psychological context (Rhatigan, Street, & Axsom, 2006). Women who leave their violent spouse are subjected to greater violence than those who do not leave, with an escalation in violence that often includes death threats or murder (Bell, Goodman & Dutton, 2009). Nevertheless, the common perception is that it is essential for women in such situations to leave their violent partner, despite the risk involved (Grigging, Ragin, Sage, Madry, Bingham & Primm, 2002).

Residing in a designated shelter represents the desire to leave the violent spouse and change the situation. However, many women leave the shelter and return to their violent spouses (Griffing, Sage, Ragin, Madry, Bingham, & Primm, 2002). Despite this, residing in a shelter affords women the opportunity to engage in a guided, supported thinking process where they can debate the question of whether or not to stay in the relationship, and to consider the factors that may contribute to their decision to leave their spouse (Krishnan, Hilbert, Mcneil & Newman, 2004).

The present study examined the variables associated with the decision to end a violent relationship and leave the violent spouse. In addition, the study examined three issues pertaining to women’s decision to leave the shelter: (1) the destination of women who leave the shelter. It should be noted that a woman's stated destination upon leaving the shelter does not necessarily reflect her final destination though it may predict the final destination (Lerner & Kennedy, 2000). (2) The duration of stay in the shelter. Previous research suggests that longer stays in a shelter are associated with more serious intentions to escape the cycle of violence (Baker, 1997; Rhatigan, et al., 2006). (3) The contribution of characteristics of the violent relationship, the woman’s emotional state, her expectations from the shelter, and her experiences in the shelter to predicting duration of stay in the shelter and destination upon departure.

This study utilized the Psychosocial Readiness Model (Cluss, Chang, Hawker, Scholle, Dado, Buranosky et al., 2006). The premise of this model is that following the strengthening of the woman's internal characteristics - her awareness of the violence inflicted by her partner, her expectations from the shelter, and her emotional state - and the strengthening of external characteristics – namely, her experiences in the shelter, the woman will be motivated to change her situation, the duration of her stay in the shelter will be longer and she will be inclined to end the violent relationship.

Variables that were recorded upon arrival at the shelter will be used as covariates in the statistical analyses examining gaps between variables that were recorded both upon arrival and departure from the shelter.

The study sample included 104 women, victims of violence, who resided in 12 shelters in Israel. The data utilized in the current study is part of a larger study conducted of Professor Dekel, Professor Itzhaki and Dr. Ben Porat from the School of Social Work at Bar Ilan University. The women completed a structured questionnaire upon arrival and departure from the shelter. The questionnaire examined the following variables: demographic characteristics, characteristics of the violent relationship, psychological wellbeing - including a PTSD questionnaire based on DSM-IV (Solomon & Horesh, 2007) and the depression sub-scale of the Brief Symptom Inventory (BSI). The expectations and experiences of women in the shelter were measured with a questionnaire from a study conducted in shelters in the U.S. (Lyon, Lane & Menard, 2008). Duration of stay in the shelter and destination upon departure were recorded by the staff in the shelters. It should be noted that the aforementioned variables were recorded both upon arrival and departure from the shelter except the characteristics of the violent relationship which were only taken at the time of arrival, and experiences in the shelter which were only taken at the time of departure.

The current study, as well as data from the Ministry of Social Welfare, indicates that over 20% of women in violent relationships utilize transitional points (relatives and friends) upon leaving the shelter (Ministry of Welfare, 2011). These findings may suggest that transitional points are the suitable option for many women (Ministry of Welfare, 2011). Therefore, this study will focus on women who choose to use transitional points (relatives and friends) and women who return to independent living.

The study hypotheses were partly confirmed. The findings regarding the destinations upon leaving the shelter (independent living, transitional points and return to spouse) show that over 20% of women leave to transitional points. To our surprise and contrary to our hypothesis, the findings indicate that women who use transition points sense less danger than women who continue to independent living (Compton et al., 1989). In line with our hypothesis, greater improvement in the emotional state was observed in women who resided in the shelter for longer durations and went on to independent living or transition points. Moreover, women who returned to their violent spouse showed a tendency towards higher levels of PTSD. No association was found between depression and final destination after leaving the shelter (Dutton & Painter; 1981; Johnson, Perez, & Zlotnick, 2008). Additionally, there was a significant association between place of birth and destination.

Findings regarding duration of stay in the shelter indicate that longer stays in the shelter are associated with a higher probability of independent living upon departure from the shelter (as opposed to transitional points). Additionally, longer stays are associated with an improvement in the women's descriptions of their situations in all the variables (emotional state, experiences and expectations from the shelter). These findings support our hypothesis regarding the association between duration of stay and likelihood of leaving the violent spouse, specifically when the woman reports an improvement in the aforementioned variables (Compton et al., 1989; Lyon et al., 2008; Rhatigan et al., 2006). The findings also suggest that greater sense of danger is associated with longer durations of stay in the shelter. It is well established that a sense of life-threatening danger is the main reason for turning to shelters, and it seems that many women experience the shelter as a safe haven (Angless et al., 1998; Dutton, Meyer & Wagner, 2009; Gondolf & Shestakov, 1997; Levinson, 1989). A tendency towards a positive correlation (bordering significance) was found between women's experiences in the shelter and duration of stay. Hence it seems that positive experiences in the shelter affect women's decisions regarding the duration of stay in the shelter. This finding confirms the perception that shelters provide supportive means of paramount importance (Ben-Porat & Itzhaky, 2008; Francesca, 2003; Hague & Mullender, 2006; Wallis & Poulton, 2001). Moreover,  several associations between demographic variables and duration of stay in the shelter were found: women who were abused during childhood and women who worked during their stay in the shelter stayed more days in the shelter; immigrants from the former Soviet Union and Ethiopia also tended to stay in the shelter for longer periods (compared to native Israelis).

Regarding duration of stay in the shelter and departure destinations, there is a continuum of readiness to change: one end symbolizes the status-quo in violent relationships (women with short durations of stay in the shelter who return to their violent spouses), and the other end symbolizes readiness to take action for change (women with longer durations of stay in the shelter who continue to independent living). The subgroup of women who went to live with family and friends did not resemble either of the other two groups and appear to represent an independent group which is at the heart of the controversy regarding the definition of violence cycles, as evident in the literature (Compton, Michael, Krasavage-Hopkins, Schneideman & Bickman, 1989).

The findings of this study have several contributions: first, they contribute to broadening the little knowledge that exists about women who leave the shelter to reside with family and friends. Second, the findings identified predictors of destinations upon leaving the shelter and duration of stay in the shelter.

In light of the contribution of prolonged stays in shelters to the ending of violent relationships, it is important to conceive ways to effectively encourage longer stays in the shelter so that women may fully utilize the proposed interventions to their own gain and increase their ability to plan living alternatives upon leaving the shelter.