RETURNING MATERIALS: MSU Place in book drop to LIBRARIES remove this checkout from -—L—_ your record. FINES win — be charged if book is returned after the date stamped below. THE PRELIMINARY INVESTIGATION OF SOCIAL SUPPORT NETWORKS OF WOMEN WHO HAVE BEEN IN ABUSIVE RELATIONSHIPS, AND THE EVALUATION OF A SUPPORT GROUP INTERVENTION AS A SOURCE OF SOCIAL SUPPORT BY Chris Schattmaier A THESIS Submitted to Michigan State University. in partial fulfillment of the requirements for the degree of MASTER OF ARTS Psychology Department June 1987 ABSTRACT THE PRELIMINARY INVESTIGATION OF SOCIAL SUPPORT NETWORKS OF WOMEN WHO HAVE BEEN IN ABUSIVE RELATIONSHIPS, AND THE EVALUATION OF A SUPPORT GROUP INTERVENTION AS A SOURCE OF SOCIAL SUPPORT BY Chris Schattmaier This study examined the social support networks of women who had been in abusive relationships and included a support group as an intervention to increase the availability of social support. The participants were thirteen women who attended support group meetings and received interviews; some of whom also participated in follow-up and drop-out interviews. The findings indicated that the respondents tended to have small networks with relatively few contacts, which may indicate isolation. However, 70% described at least one friend who was available to provide support "most of the time." The support group suffered from low attendance and was utilized more by women who had returned to their abusive partners, although it was developed for those who had ended the relationships. It was proposed that a needs assessment would best indicate directions for future Chris Schattmaier interventions. Suggestions were made to improve the research instruments. ACKNOWLEDGEMENTS My committee deserves the first recognition. Sincere thanks is offered to my chairperson, Robin Redner, for her support of this project from the beginning and for her guidance in helping me to organize the ideas in this paper. Carol Mowbray taught me through her challenging questions much about integrating psychological theory with "real world" research. I also appreciate the wisdom and knowledge of Ellen Strommen, who contributed additional ideas for consideration. Special recognition. belongs to Deb Bybee, who so generously shared her time and invaluable insights, which helped me to keep perspective of the purpose of this project. In addition, I'm eternally grateful to Mary Sullivan, my co-facilitator and research partner, whose optimism and motivation were the lifeblood of this pro- ject. Deepest appreciation is also acknowledged for the women who assisted with the planning of the support group; those who attended the meetings; cooperated with the sometimes lengthy interviews; and shared their wisdom and experiences. ii My greatest debt, however, is owed to my friends and family, my social support network, for their patience, support, and encouragement that were expressed in ways too numerous and personal to mention here. iii TABLE OF CONTENTS Page List of Tables vii Introduction 1 Social Support 10 Components of Social Support Networks 13 Network Size 14 Amount of Contact 17 Nature of Contact 20 Adequacy of Support 22 Network Composition 25 Density 29 Complexity of Relationships 30 Reciprocity 31 Geographic Proximity 33 Stability 35 Closeness 37 Interventions Aimed at Increasing Social Support 41 Volunteer Linking Strategies 42 Natural Helper Strategies 47 Peer Support Group Strategies 51 Summary 64 Description of the Study 66 Research Questions 67 Description of Social Support Networks 67 Impact of the Support Group Intervention on the Participants' access to Supportive Resources 69 Impact of the Support Group Intervention on the Participants' Social Support Networks 70 Methods 72 Support Group History 72 Sponsoring Organization 72 Researcher Involvement 73 Group Development 74 Research Participants 77 Recruitment 79 Initial Contact 81 Restrictions 81 Procedure 82 iv Results Setting. . . . . . . . . . . . . Facilitators . . . . . . . . . . . Format of Meetings . . . . . . . . . Administration of Measures . . . . . . Initial Interview. . . . . . . . Follow-up Interview . . . . . . . Drop-out Survey . . . . . . Group Participation Form . . . . . Group Content Form . . Interviewer and Behavioral Observer Training Initial Training . . . . . . . Weekly Supervision Meetings . . . . Termination of the Group . . . . . . . Measures . . . . . . . . . . . . . Interviews. . . . . . . . . Initial Interview. . . . . . . Development/Construction. . . . Categories . . . . . . . . Follow-up Interview . . . . . . . Categories . . . . . . . . Drop-out Survey . . . . . . . . Items . . . . . . . Group Participation Form . . . . . Development/Construction. . . . Description of Categories . . . Group Content Form . . . . . . . Examination of Social Support Networks of . ° . Support Group Participants . . . . Average Network Characteristics of Support Group Participants . . . . . . . . . Network Size . . . . . . . . . Amount of Contact. . . . . . . . Nature of Contact. . . . . . . . Adequacy of Support . . . . . . . Network Composition . . . . . . . Density . . . . . . . . . Geographic Proximity. . . . . . . Stability . . . . . . . . . . Closeness . . . . . . . Relationships that Emerged for Network Variables of Support Group Participants . . Evaluation of the Support Group Intervention . . Supportive Resources Provided by the Support Group. . . . . . . . . . . Assistance . . . . . . . . . Emotional Support . . . . . . . Commonality, , . . . . . . . . Intimacy, , . . . . . . Influence/Role Modeling, . . . . . V 82 83 85 87 87 88 88 89 90 90 90 92 93 93 94 94 94 96 97 99 99 100 101 102 102 105 108 108 108 110 110 110 111 112 113 113 113 114 115 120 120 120 123 125 126 127 Sociability. . . . . . . . . . 128 The Impact of the Support Group Inter- vention on the Participants' Social Support Networks. . . . . . . . . . 129 Feedback on the Support Group from Drop-out Surveys . . . . . . . . . . . . . . 131 Summary..............135 Discussion - - . - . - - - - - - - - - - 138 Discussion of Findings . . . . . . . - - 139 Network Variables . . . . . . . . . 140 Alternate Interventions . . . . . . . . . 151 Suggestions for Revisions of Measures . . . . 156 Initial Interview and Follow-up Interview . 156 Drop-out Survey . . . . . . . . . . 162 Group Participation Form and Group Content Form. 0 O O O O O O O O C O O O 164 Suggestions for Evaluation of the Intervention . 166 Summary..............167 Appendices Appendix A . . . . . . . . . . . . . 174 Appendix B . . . . . . . . . . . . . 194 Appendix C . . . . . . . . . . . . . 211 Appendix D . . . . . . . . . . . . . 214 Appendix E . . . . . . . . . . . . . 216 References . 219 vi LIST OF TABLES Responses of Participants on Network Variables and Demographic Character- istics. . . . . . . . . . Phi Coefficients for Network Variables and Demographic Characteristics . Point Biserial Correlations for Network Variables and Demographic Character- istics . . . . . . . . . . Spearman's Rank Order Correlations for Network Variables and Demographic Characteristics . . . . . . . Pearson's Product Moment Correlations for Network Variables and Demo- graphic Characteristics . . . . Topics Discussed at Support Group Meet- ings and Number of Meetings at which Topic was Discussed . . . . . Changes in Responses of Participants on Network Variables and Demographic Characteristics from the Initial Interview to Follow-up Interviews . Support Received by Women from Support Group Members who were Members of their Social Support Networks . . vii 109 116 117 118 119 122 130 132 T" INTRODUCTION Since the grassroots efforts by women established the first safe houses in the 1970's, abuse of women has continued to gain attention in the United States. This increase in awareness has led to a variety of services and legislation aimed at assisting women who are trying to leave their abusive partners. Although much has been written describing the experiences of women who have been in relationships with abusive men, few sys- tematic evaluations have been reported. Reasons fbr this absence may include the urgency of providing safety to meet the ever present need and the difficul- ties of conducting research in this area. It should be noted that most of what has been reported about women who have been in abusive relationships comes from those who have sought assistance from shelters, counselors and other formal sources of assistance. However, there are also likely to be a great number of women who have escaped from abusive relationships solely through their own efforts or with the assistance of informal sources of support. Additionally, most of what has been written focuses on the abusive relationship and the establishment of shelter programs, while little has been presented on the lives of women once they have left the relationship and the shelter. The present study attempted to examine women who were in the pro- cess of adjusting to living independent of their abu- sive partner and the shelter. The prevalence of the crime of abuse of women has not been accurately determined, however, all estimates indicate that it is disturbingly commonplace. Aside from a study by Straus, Gelles and Steinmetz (1980), no systematic attempt has been undertaken to collect accu- rate figures of the incidence of this crime. Indirect indications of the extent of woman abuse include: reports that 70% of assault victims who went to Boston City Hospital's emergency room were known to be victims of women battering; 12% of all murders are between hus- bands and wives, with wives and husbands committing the murders. equally’ as often, however, wives are seven times more likely to kill in self-defensefL 37% of female divorce applicants in Cleveland included physi- cal abuse as one of their complaints (Martin, 1976); shelters for' battered. women. and. their' children all across England and the United States report that they are filled to capacity as soon as they open their doors (Walker, 1978); over the course of two years, shelters in Michigan served 21,900 clients (abused women and their children), while over 6,900 women and children had to be turned away for lack of space (Carty, 1983); a systematic study of violence in the home conserva- tively estimates that almost two million women are vic- tims of domestic violence each year (Straus, Gelles & Steinmetz, 1980); and it is further estimated that 50 to 60 percent of marriages contain at least one violent incident (Straus, 1977-78). Abuse of women refers to attempts or actions used by a man to dominate or control a woman with whom he is, or has been, in an intimate relationship. This includes such behavior as physical attacks, threats of physical attacks. and. psychological attacks, such. as repeated derogatory remarks, deliberate destruction of the woman's property, confining her to a limited area, or imposing 24—hour surveillance upon her. Although psychological abuse is often neglected in the literature, it appears to be very common and almost always accompanies physical abuse (Finkelhor, 1983; Pagelow, 1984; Roy, 1977; Schecter, 1982; Walker, 1979). In Walker's (1978) pilot study, she found that psychological humiliation and isolation were the worst forms of abuse that women reported experiencing whether or not they had been physically abused. Furthermore, even when they had never been physically abused, all reported that they feared for their lives. The present study focused on women who chose to end. the relationship ‘with. their' abusive partner’ as opposed to those who were working towards ending the abuse while continuing the relationship. The informa- tion presented below explains this focus. Walker (1978) reported that the most effective way for a woman to end her victimization is by ending the relationship. This is necessary to break the mutual dependency and possibly, in time, reconciliation may be achieved with- out this destructive form of dependency. She claimed also that the abuser has no motivation to change once the woman returns. In fact, most reports indicate that without intervention the abuse is likely to increase in frequency and severity if the relationship continues (Schacter, 1982; Walker, 1984). The present study also focused on women who are attempting to establish a life apart from the abuser because they are often in critical need of support in this transition period. They are likely to experience pressure from a number of sources: financial strain, loneliness, child care responsibilities, and requests for reconciliation from the abusive partner. The woman may find considerable support for her decision if she resides at a shelter. However, upon her departure, she may find much less encouragement. A feminist perspective views violence against women as the result of and reflection of our male supremist society (Fleming, 1979; Schecter, 1982; Stark & Flitcrait, 1983). As long as the structure of soci- ety' continues to systematically‘ oppress and. devalue women economically, sexually and socially, .violence against women will remain prevalent. Primary preven- tion through education, along with the enactment and strict enforcement of laws that protect the rights of women are essential; however, society is equally responsible for providing assistance to those who suf- fer from abusive relationships as long as this problem exists. Although abuse of women cuts across race, class and geographic lines, certain characteristics set abu- sive relationships apart from other intimate relation- ships. First, the abusive relationship is often described as mutually dependent (Walker, 1978). Both partners are often strong believers in traditional sex roles. Often the woman believes that she cannot sur- vive without a man. Two plausible theories of how abused women are kept in the grip of emotional depen- dency have been proposed by Lenore Walker (1977-78; 1978). Walker (1978) observed an abusive cycle which began with a tension building phase, where the woman generally attempts to please the man in the hope of avoiding an attack. Phase I culminates in an explosive battering incident, marking phase II, where the man is frequently described as totally out of control.2 The final phase is when emotional dependency is established and reinforced. During this stage, which is commonly called the honeymoon period, the abuser is repentant, loving and often declares his need for the woman. He begs her forgiveness and promises that it will never happen again. The woman wants to believe this because during this phase everything is how she imagines it should be. She hopes that this time she can prevent the cycle from repeating itself by trying harder to meet her partner's demands. The second theory presented by Walker (1977-78) may operate in conjunction with the cycle of abuse to reinforce the woman's dependency. Walker applies Seligman's learned helplessness model to women in abu- sive relationships. Here, societal institutions act as an accomplice in her victimization. The theory of learned helplessness holds that when an individual's attempts to escape from a threatening situation are repeatedly blocked, s/he will stop trying to escape. After s/he has reached this point, the victim will not attempt escape even in the presence of a clear opportu- nity because s/he believes it is hopeless. Research with animals has shown that the effects of this learn- ing process, although difficult to overcome, do not appear to be irreversible. For example, research involving dogs demonstrated that they needed to be dragged repeatedly to safety before they began to attempt escape on their own. Walker suggests that sex role socialization may teach women to be helpless. They learn that their actions, especially in terms of trying to prevent the abuse, are generally ineffective. In time it becomes very difficult to change this belief, and hence, they often "need to be shown the way out repeatedly before change is possible" (p. 529). Additionally, learned helplessness is heavily rein- forced by many social institutions which refuse to pro- vide assistance to the abused woman. For example, police officers, doctors, psychotherapists and clergy are all reported to have asked women what they did to provoke the abuse. Others have pressured the woman to stay "for the sake of the children." Social service agencies have not provided the emergency funds needed by the woman to establish her own residence and to feed herself and her children, and law enforcement agencies have failed to provide adequate protection (Fleming, 1979; Walker, 1978; Michigan Women's Commission, 1977). Another extremely common factor that contributes to an abused woman's dependency (and real and perceived help— lessness) is the isolation that she often experiences. Isolation can be imposed by a number of sources. Most commonly, the abusive partner will try to keep the woman isolated in an attempt to meet his need to con- trol her, or to try to prevent her from leaving him (Goodstein & Page, 1981; Hanks & Rosenbaum, 1977; Hilberman, 1980; Walker, 1979). Such isolation efforts may include minimizing or preventing the woman from having contact with friends and relatives, and/or leav- ing her at home without access to transportation, money or a telephone. Fleming (1979) reports that abusive men may actively limit their partner from engaging in rewarding work, securing assistance with child care, or participating in educational or recreational activi- ties. It has been noted (Fleming, 1979) that women may also contribute to their own isolation. Generally, women are embarrassed or ashamed of remaining in such a relationship. They may also remain at home to avoid "provoking" their partners (Gelles, 1976). WOmen may also refrain from initiating contact with others because of their responses. As Fleming (1979) reports, "the abused women usually finds little sympathy or support from family members, friends, neighbors, co- workers, or the person in the street. Cultural norms legitimizing wife beating abound" (p. 79). If such factors as severe dependency on the abu- sive partner, learned helplessness and isolation do indeed exist, it seems that women who attempt to end abusive relationships will require a great deal of sup- port to overcome these obstacles. For women who feel lonely and isolated, Fleming (1979) recommends that, you will need to make an extra effort to keep in touch with those who have supported you and helped you to get reestablished. You may feel extremely vulnerable to returning to the marriage, especially if your husband has found you and is pressuring you to return. Continued contact with your support system (women's groups, friends, advocates, etc.) is very important. (p. 63) Support may also be necessary in assisting the women as she grieves the end of the relationship. This may include a significant sense of loss and frightening independence (Fleming, 1979). A more detailed discussion of social support and how it may be mobilized for women who have left abusive relationships is presented next. 10 Social Support The term "social support" in the psychological literature can refer to a wide range of constructs. Although many may feel that they have an intuitive understanding of what is meant by "social support," definitions in the research literature vary consider- ably. Some authors have described social support as a global, unidimensional concept, while others have described it in multidimensional terms. The global definitions tend to be either vague descriptions, such as that of Kaplan, Cassel and Gore, "the relative pres- ence or absence of psychosocial support resources from significant others" (p. 146, cited in Thoits, 1982); or single operationalizations (e.g. marital status, or number of contacts with friends) (Holahan & Moos, 1982; Lin, Dean & Ensel, 1981) . On the other hand, when social support is treated as a multifaceted concept, researchers have generally chosen a subset from the many dimensions that have been identified as relevant aspects of social support (e.g., Duncan-Jones, 1981; Norbeck, Lindsey & Carrieri, 1981). Although some have not made a clear distinction, social support can be separated into two components: the resources that are provided and those who provide these resources. In the present discussion social support will refer to the supportive resources that an individual perceives as 11 available to her or him, while social support networks are defined as the people who provide these resources. The supportive resources that one's social support net- work provides include (a) "commonality" (the sharing of similar interest, values, and/or experiences) (Candy, Troll & Levy, 1981: Davidson a Packard, 1981); (b) "assistance" (the provision of help when needed, including direct help or helpful information) (Candy et al., 1981; Davidson & Packard, 1981; Wright, 1969); (c) "intimacy" (mutual self-disclosure, sharing of feel- ings, indications that one is loved) (Candy et al., 1981; Davidson & Packard, 1981): (d) "sociability" (fun to be with; feel at ease with; interesting) (Candy et al., 1981; Wright, 1969): (e) "emotional support" (feeling encouraged, understood, accepted, and/or cared for) (Davidson & Packard, 1981; Wright, 1969); and (f) "influence or role modeling" (Having influence on each other, viewing another as possessing good judgement or character) (Candy et al., 1981: Davidson & Packard, 1981). In evaluating a definition of social support one may wish to consider the claims of some authors concerning the issue of the validity of global and multidimensional approaches. Although the data from a number of studies have indicated that social support is comprised of a number of distinct dimensions, others 12 have found significant results when employing a more global definition. All of the following studies found that social support is correlated with positive experiences, despite the variations in their respective definitions and measurements of social support. Eaton (1978) found that household members provide support in times of crises that help prevent mental dysfunction; Sarason, Levine, Basham and Sarason (1983) discovered a positive relationship between social support and both extrover- sion and level of happiness; Turner (1981) reported a positive association between the experience of social support and psychological well-being; and Linn and McGranahan (1980) stated that having more contact with close friends can reduce the negative impact of per- sonal disruptions on an individuals' well being. On the other side of the issue, Cauce, Felner and Primav- era (1982) argue very strongly in favor of a multidi- mensional approach. They feel that important differ— ences can go undetected when using unidimensional indices. Their findings indicated that the perceived degree of helpfulness is viewed differentially for for- mal and informal sources of support. They stated, that their findings argue persuasively in favor of the position advanced by [others] that vague and 13 global conceptualizations of support systems, both structurally and functionally, may be, at best, not very information and, at worst, lead to poorly designed and damaging inter- ventions" (p. 426). The findings of other studies, that the effects of social support vary according to the situations and type of support provided, lends further weight to the claim that social support is a multidimensional concept and needs to be examined as such (Sarason et al., 1983; Wandersman, Wandersman & Kahn, 1980). Components of Social Support Networks Theorists who have studied the networks of persons who provide social support to individuals have identi- fied a number of components that help to distinguish the characteristics of these social support networks. An examination of these components can help in assess- ing the amount and quality of support that is available to an individual. The most common network variables used in the literature are presented below. The pre- sent study was of an exploratory nature. It included most of the network components, but due to its limited scope, indepth network analysis was beyond its capac- ity. Rather, a general indication of the extent to which the participants had access to supportive 14 resources and preliminary information concerning their networks was gathered. W Network size is one of the most frequently included dimensions in studies that examine social sup- port. Most researchers ask for the number of people, either overall or in particular categories, while a few others concentrate on the number of different cate- gories of support available without considering the number of persons in each category (e.g., Caldwell & Bloom, 1982; Duncan-Jones, 1981; Henderson, Duncan- Jones, Byrne & Scott, 1980). Some studies ask exclu— sively about friends (e.g., Linn & McGranahan, 1980), while others include relatives and acquaintances (e.g., Mitchell, 1982; Silberfeld, 1978). Another major dis- tinction in network size definitions is whether the respondent is to include everyone s/he knows (e.g., Henderson et al., 1980), or just those felt to be sup- portive (e.g., Mitchell, 1982). In the present study, network size was calculated as the total number of people given in response to the questions, "How' many people do you consider close friends?" ("A close friend is someone you can talk to about your feelings and to whom you can turn for help." (Bybee, 1980, p. 143) and "How many of your relatives do you consider to be supportive?" It was felt that 15 excluding relatives in the calculation of network size might severely limit information on providers of social support. A number of studies have shown that relatives can be an important source of support (e.g., Holahan & Moos, 1982; Silberfeld, 1978). Additionally, this definition includes only "close" or "supportive" rela- tionships because it was believed that descriptions of these relationships would provide the most information on the availability of supportive resources. A number of other studies have limited their focus to relation- ships that are perceived as supportive (e.g., Linn & McGranahan, 1980; Makowsky, 1983; Mitchell, 1982). Most research findings, which are based on correlational data, indicate that in terms of network size, "bigger is better." For example, in a study involving previously hospitalized psychiatric clients, significant relationships were reported between the number of "intimates" and problem solving abilities (Mitchell, 1982). Additionally, the number of support- ers for a group of women who had sought assistance from battered women's shelters was negatively related to the level of ‘violence: they' had. experienced. (Mitchell & Hodson, 1983). However, the reliability of self-report indications of network size has been questioned. In the developmental stages of their social support instrument, Henderson. et a1. (1980) considered. this 16 issue. After administering their measure to a sample of the general population in Canberra, the capitol of Australia, a modified version was administered to peo- ple who knew the respondents well. This modified ver- sion requested information on the original respondents' social relationships. Henderson and his colleagues report that the correlations that they achieved between the respondents' and the informants' scores were sig- nificant beyond the one per cent level. They report that these correlations are in keeping with the level one would expect when one person rates another on such attributes. Unfortunately, network size has little comparative value across studies because most studies used unique calculations. It would be useful if researchers pre- sented the numbers (or means) of people in the networks of their sample. This would allow other researchers to make comparisons with. their own samples. In most instances, the only information provided was the corre- lations of "high" and "low" network size with other variables. Examining network size for women who have been in abusive relationships may be especially relevant. Though it has not often been systematically documented, it is frequently reported that social isolation often accompanies woman battering. Certainly, the previously 17 mentioned findings of Mitchell and Hodson (1983) (that the level of violence experienced by a woman was negae tively related to her number of supporters) point to one advantage of larger networks. If the woman chooses to end the relationship, she may be losing one of her greatest sources of support. Even though the behaviors of the abusive partner do not usually appear to be sup- portive, if the woman had been isolated, he may have been her sole or major source of contact. It seems likely that a woman's transition, as well as her abil- ity to remain away from the abusive relationship, would be aided if her social support network provided ade- quate support. A larger network would be more likely than a smaller network to provide some of the resources lost as a result of ending an abusive relationship. AEQQQE_Q£_§QELQQ§ The amount of contact between an individual and her social support network members is also defined in many different ways. These definitions vary on a few dimensions: the type of contact that is counted; the way that contacts are measured; and the period of time that respondents are asked to report on. Some measures count only in-person interactions, while others include telephone conversations and written correspondence as well. The amount of contact can be calculated in terms of number of contacts, and/or by the amount of time 18 spent in contact. Most frequently, individuals are requested to indicate the amount of contact that they have had with members of their social support networks within the last month. They may also be asked to describe weekly or daily contact in addition to or instead of monthly contact. And finally, the amount of contact may be calculated by contact with everyone that the respondent has interacted with, only friends and relatives, or only close friends. In the present study, the amount of contact was equal to the number of times in the previous month that the respondent had seen in person, talked with (on the telephone), and/or written letters to her three closest friends‘and three most supportive relatives. It was felt that requesting information on contact in the pre- vious month would be a long enough period of time to provide a representative sample of the participant's amount of contact, while a short enough period of time to prevent retrospective distortion. (Mitchell & Hodson, 1983, and Holahan & Moos, 1982, have also employed this time frame.) The time period immediately following the escape from an abusive relationship is likely to have been a period of major change when a woman may need a great amount of support. Therefore, collecting information on support received in the pre- vious month may provide a good indication of a woman's 19 network in action. Additionally, it was felt that a month's length of time was too long to expect accurate estimates on the duration of each contact, therefore, the definition was limited to number of contacts. When reviewing the description of the amount of contact variable in the social support literature, it is once again found that the inconsistency in the oper- ational definitions of the amount of contact impedes meaningful comparison. Furthermore, there is some evi- dence to indicate that the way that amount of contact is measured can influence research findings. Silber- feld (1978), for example, discovered significant dif- ferences between the amount of contact for female fam— ily practice patients and psychiatric out-patients by considering the amount of time in contact that would not have otherwise been revealed if just the number of contacts were measured. While Mitchell and Hodson (1983) found that the severity of violence that bat- tered women experienced was related to the presence or absence of her abusive partner during her interactions with others. Specifically, they found that women who had less contacts with friends and family unaccompanied by their partner experienced a significantly higher level of violence. Again, frequent contact with supportive persons may' be invaluable. in helping 'women adjust. to life 20 without their abusive partner. If the woman remains isolated, his efforts to win her back during the honey- moon period and her remembering the good times may influence her to trade her loneliness for whatever degree of support he seems to be offering (Fleming, 1979). ature 0 Contact The nature of contact that individuals engage in with members of their social support networks can be classified by the medium of the contact (i.e., in-per- son, over the telephone, or through written correspon- dence). Additionally, the content of the interaction (i.e., the activities and conversations of the partici- pants) may be of interest. Information about the con- tent of interactions may indicate the supportive resources that were provided. For example, sharing a ride to work may provide assistance, while visiting a museum may involve commonality and sociability. In a few studies, the authors examined contacts that included in-person, telephone and letter writing inter- actions, however, a greater number were only concerned with in-person contacts (Eaton, 1978; Holahan & Moos, 1982; Mitchell, 1982; Mitchell & Hodson, 1983; Silberfeld, 1978). Still others sought the frequency of contact without specifying the medium (Lin, et al., 1981; Linn & McGranahan, 1980). 21 An assessment of the supportive resources that were provided by network members was frequently included in studies. Some of these resources included the: degree of agreement with. or support for one's actions (Norbeck et al., 1981); assistance, and emo- tional support (Mitchell, 1982); commonality and inti- macy (Duncan-Jones, 1981); and sociability (Mitchell & Hodson, 1983). In the present study, interview items examined the nature of contact with network members that included face-to-face visits, telephone conversations, and let- ter writing. For each of these categories the content of the interaction was also specified through responses to such questions as, "What do you usually do with them?" It was considered important to include more than just face-to-face interactions because much infor- mation may have been lost by excluding other mediums, especially if one or more key network members did not live near the woman. The distinction between the media of contact was included to provide more information on an individuals' supportive interactions. It was felt that. the content. of the interactions could. provide information not available from requesting only the num- ber of contacts. In addition, the information given by the respon- dents on the content of their interactions could be 22 very useful. For example, a woman may report that in almost all of her contacts that she discussed her feel- ings about leaving her abusive partner, and that she frequently received support for her decision. Other such information may be that most women tend to seek material assistance from relatives and emotional sup- port equally from friends and relatives. The kinds of support that women who have left abusive relationships request most frequently can also be useful for pro- gramming considerations. de c o u or An indication of whether individuals are satisfied with the social support received from their networks is central to many authors' definitions of social support. Caldwell and Bloom (1982), Canoe et al., (1982), Duncan-Jones, (1981), Mitchell, (1982), Sarason et al., (1983), and Wandersman et al. (1980) have all included an indication of perceived adequacy of social support in their studies. Others also included items that may be considered indirect indications of satisfaction with available support" For instance, Young, Giles and Plantz (1982) asked how much a situation had improved as a result of discussing it with someone else, and Mitchell and Hodson (1983) looked at whether women who had been in abusive relationships received empathic or 23 avoidance responses when the issue of battering was raised. Items to elicit a direct indication of the respon- dents' perceptions of the adequacy of support received from network members were not included in this study. Indirect indications of the respondents' level of sat- isfaction may be found in the responses to the follow- ing items. Four items asked how the respondent felt after her visits and phone conversations with friends and relatives; one item asked how she felt about her friends' opinions about what she should do about her relationship with her abusive partner; another item requested this information concerning relatives; and one item for each friend and relative asked how supportive she felt that they would be regardless of her decision concerning her abusive partner. It was felt that by including such items, information would be reported on participants' satisfaction with their interactions with network members and more specifi- cally, what they did and did not like (which may be easier to remember and describe in the context of a particular situation or issue). Others have found that situation specific indications of adequacy can reveal important findings (Mitchell & Hodson, 1983). As an indication of how satisfied people are in general with the supportive resources that they 24 receive, Young et a1. (1982), found that 43 per cent felt their situation had improved "a lot" because of the behavior of the most helpful person that they talked to. Sarason et a1. (1983) reported that for female undergraduate students, satisfaction with avail- able support was negatively related to anxiety, depres— sion and neuroticism. Cauce et a1. (1982) found that satisfaction with support may depend on the relation- ships or other factors. More specifically, they found that, among a group of low income youths, black adoles- cents rated family members as more helpful than did Hispanic or white students, and that females rated informal support (friends) as more helpful than did their male peers. The findings of two other studies seem to indicate that just discussing a problem with network members may not provide enough support for the individual. First, Mitchell and Hodson's (1983) corre- lational data revealed that women who received more avoidance responses when attempting to discuss their abusive situation tended to have suffered more frequent and severe beatings. Similarly, family members or psy- chiatric clients perceived their friends to be signifi- cantly less helpful in dealing with problems that were related to the client, as opposed to other stress pro- ducing situations (Mitchell, 1982). 25 This last finding might be similar to the com- plaint of many women that family and friends who have not had an abusive experience were not able to under- stand their situation. In such cases, women may feel that they are not receiving adequate support. Another important factor in determining satisfaction with sup- port (for women who have left their abusive partner) is the degree to which network members are willing to pro- vide the resources that will allow follow through on this decision. It may be essential to have this sup- port because, as mentioned previously, women making this transition to independence are likely to face many pressures, and historically, societal institutions have been less than helpful (Walker, 1978). Network Composition The relationship of network members to the indi— vidual is reflected by network composition. The most frequently included categories can be classified as family/relatives (e.g., spouse/partner, immediate fam- ily, extended family), friends (sometimes a distinction is made between degree of closeness), community sup- porters (e.g. counselors, religious leaders, health care ;providers), and. other’ acquaintances (e.g., co- workers, neighbors). Out of a sample of twelve studies 26 that listed categories, ten included family or rela- tives; nine included friends; seven included acquain- tances; and five included community supporters. In the present study, respondents were asked to list their three closest friends and where they met them, and their three most supportive relatives and their relationship; and they were asked to rank their three top sources of support from the following list: social worker; abusive partner; parents; children; other relative (and relationship); counselor or thera- pist; minister, priest or rabbi; female friend who has experienced an abusive relationship; female friend who has not been in an abusive relationship; male friend (other than abusive partner); and other (and relation- ship) (adapted from Bybee, 1980). In addition, it was asked whether any of the woman's close friends or rela- tives had been in a relationship with an abusive man. These items incorporated all of the dimensions that other researchers included regarding close, informal sources of support (e.g., Norbeck et al., 1981; Silberfeld, 1978). The present study was mainly con- cerned with informal supporters, but an indication of how they perceived these supporters in relation to more formal sources of support was provided by ranking the above list. 27 A few network characteristics related to network composition are given below. In a sample of university staff people, Norbeck, Lindsey and Carrieri (1983) reported that 97% included family members or relatives in their network, 94% included friends, 61% listed a spouse or partner, 50% listed work or school associ- ates, 17% included neighbors, and only 10% listed com- munity supporters. The respondents stated that most of their contact were with friends, followed by contact with family members or relatives. Silberfeld (1978) found that his family practice patients had a greater proportion of close relationships with relatives (other than spouse or child), usually with their parents. However, the psychiatric outpatients in the sample had a greater proportion of close relationships with friends. The method employed by Young et a1. (1982) to measure social support was to present a number of prob- lems in living and then ask the respondent the type of person talked to about the problem. A close distribu- tion resulted, with 21% talking to their spouse, 21% to acquaintances, 19% to friends, 18% to community support providers, 16% to relatives, and 4% discussed the prob- 1em with others. In another study that relates to net- work composition, Straus (1980) found that for individ- uals who were experiencing high levels of stress, higher incidences of domestic violence correlated with 28 networks primarily composed of relatives. One final research finding, which may be applicable to women who have left abusive relationships is that of Mitchell and Hodson (1983). In their study of social support net- works of clients of battered women shelters, they observed a significant negative relationship between the number of friends met through the abusive partner and the amount of avoidance responses of friends (and avoidance responses were related to level of violence and number of times battered). After a woman has left an abusive relationship, it may be very important for her to talk about what she has experienced. It is not uncommon for women to sup- press their anger in an attempt to keep peace and avoid an attack from their partner, and once a woman has left the relationship it may benefit her to be able to dis— cuss her feelings. Additionally, it has been reported that a batterer may deny that the abuse has occurred and others may ignore any evidence of abuse, leaving the woman to question her sanity (Fleming, 1979). In such instances it may be essential for the woman to have her feelings and experiences validated. If the majority of a woman's contacts are with people she knows through her abusive partner, those more likely to avoid discussing the abusive situation (Mitchell & Hodson, 1982), she may be less likely to feel supported 29 in her decision to leave the relationship. Conversely, women who have been in relationships with abusive men might find that the most understanding people are women who have had similar experiences. Density Density refers to the extent to which the ties in an individuals' network are interrelated. In other words, how many network members know other network mem- bers. It is likely that networks that include rela- tives will have some degree of density, and that den- sity will increase as the proportion of relatives in the network increases. The instrument developed for this study, included one item which asked which, if any, of the respondents' three closest friends knew each other, and if so, who they knew. It was assumed that the participant's rela- tives would know each other. No indication was elicited concerning the ties between friends and rela- tives. It is common for those who measure density to do so by asking which network members know each other (e.g., Milardo, 1982). The effects of the degree of density of an indi- vidual's network are clouded by conflicting findings. For instance, Craven. and. Wellman (1973) found that greater’ density’ was related to an increase in the availability of support and affective intangible 30 resources from close friends and relatives. However, others report that high density networks do not always provide positive influences. The research of Hirsch, and Bott (cited in Norbeck et al., 1981) suggested that for individuals who have not been diagnosed as schizophrenic, a network that is lower in density may be of more assistance in situations that involve per- sonal growth or changing social roles. There has been speculation that high density networks may be unsup- portive of change, possibly involving collective pres- sure (Hirsch, 1980). In relating these findings to women who have recently left abusive relationships, it is clear that they are likely to be making changes in their lives, some of which may call for personal growth and changes in social roles. Therefore, a network that is lower in density may be more beneficial. On the other hand, a network that is higher in density may help a women to feel a sense of family from relatives or a close circle of friends that she may otherwise lose by leaving the abusive relationship. Cpmplexity of Relationships Another"useful source. of information. about ‘the structure of social support networks can be provided by examining the number of roles performed by each member of a person's network. For example, an individual may 31 interact with a network member as a co-worker, friend, and/or neighbor. Indicators of the complexity of relationships are not commonly included in social support instruments. However, most of the studies that included the dimen- sion of complexity indicated that multiplex ties are more supportive than uniplex ties. Multiplex ties have been described as stronger, less superficial, more reliable, and significantly related to higher self- esteem, sociability and assistance (Hirsch, 1981). They appear to be most useful for those experiencing life transitions and they are better suited to provid- ing support in areas outside of family life (Hirsch, 1981). The present study did not include a means of col- lecting information on the complexity of network ties. Since complexity has not been included as often as most other network variables there is limited information on effective approaches of quantifying this variable. In the present study, an indication of complexity may be provided by examining responses to the items which ask about the content of interactions. Reciprocity Reciprocity is a network variable that indicates the amount of giving and receiving of supportive 32 resources between network members. A number of differ- ent approaches have been initiated to reveal the degree of reciprocity in a given network. Lin et al. (1981), for example, attempted to reveal the degree of reciprocity by posing the following questions to respondents concerning their relationship to a confi- dant. "how often have you talked with this person when you had a problem?" (p. 77) along with, "How often has this person talked over his/her problem with you?" (p. 77). Similarly, Duncan-Jones (1981) was interested in the number of friends in an individual's network that s/he could visit at any time, without an invitation, as well as the number that could visit her/him without an (invitation. Beyond this, Duncan-Jones also gathered data on the number of people who were in need of the respondent's care and who depended on her/him for help, guidance or advice. Young et al. (1982) additionally assessed overall indications of help-giving and help- seeking behaviors. They asked how frequently the respondents were sought out by others; the relationship of the help-seekers; the kinds of problem brought; and what behaviors they used to help. This same informa- tion was requested in reference to the respondents' own help-seeking behaviors. Responses to some items found in the social sup- port. measure used in the jpresent study' may reveal 33 information on the degree of reciprocity in an individ- ual's network, however, 'the present study was not directly concerned with reciprocity. It was believed that the participants may have experienced a number of changes in their lives in an attempt to end the abusive relationship, therefore, greater emphasis was placed on the supportive resources that were received by the par- ticipants, rather than those that they provided to oth- ers. Although there is a limited amount of evidence concerning reciprocity, there are indications that is is an integral part of healthy networks. For example, ' one study revealed that medical inpatients gave and received equal amounts of support, while those diag- nosed as schizophrenic received more support relative to the amount that they provided to others (Hirsch, 1981). A further indication of the importance of giv- ing as well as receiving support comes from the "helper-therapy" principle (Riessman, 1965). This the- ory will be described in greater detail below, but basically states that in helping another, the helper also benefits. eo ' o ' 't Geographic proximity can be defined as the physi— cal distance between individuals and their network mem- bers. This variable is important because it can give 34 an indication of the availability of some supportive resources. For studies that only included face-to-face con- tacts, geographic proximity may have an effect on amount of contact. Although few studies included direct measures of geographic proximity, some cate- gories that were used to describe network composition include indications of geographic proximity. One may assume, for instance, that relatives living in the same household and neighbors are geographically close, and that co-workers are not likely to be separated by great distances. Geographic proximity was assessed in the present study by asking respondents in what city (or how far away) their three closest friends live, and the city of residence of their three most supportive relatives. Most other studies did not include an indication of geographic proximity, and those that did included only indirect indications. It was felt that a definition that would provide an estimate of the actual distance between an individual and her network members would be more useful for the purposes of the present study. It is common for women who stay at battered women's shelters and decide against returning to their abusive partners to have little choice but to rely on public assistance. This may mean that they may not be 35 able to afford a car or telephone. With these restric- tions it may be most beneficial for them to have at least a few network members nearby. They are likely to need assistance with such things as rides to the gro- cery store and laundromat, child care while they meet with lawyers, etc. It would be more difficult for net— work members to provide these forms of support if they were separated from the individual by great distances. At the same time, anyone who is experiencing the termi- nation of a relationship is also likely to be in need of’ emotional support, .Access to supportive others either in person or via the telephone is likely to aid in the transition. tab' 't Stability refers to the degree to which changes occur in social support network components. Some of the most frequently examined components include network composition, amount of contact, and network size. The length of relationships in a network can also provide an indication of stability. Stability can be assessed by examining the networks of individuals at separate points in time, usually through the readministration of social support measures. Stability may be an especially useful variable for those who are experiencing life transitions because it may reflect the impact that life events have on the 36 personal relationships of an individual. Additionally, it provides a procedure for researching the effects of interventions aimed at altering network components. The design of the present study included a follow- up measure that was to be administered periodically in an attempt to assess change over time. To allow direct comparison, this follow-up measure included all of the social support items found in ‘the initial measure. Readministration of the same items was the procedure used in other studies when examining change over time (e.g., Caldwell & Bloom, 1982; Norbeck et al., 1986). As another indicaticn of stability, participants were also asked how long they had known each of their close friends. This procedure was also employed by Lin et a1. (1981), Norbeck; et al., (1981), and Silberfeld (1978). Because of the life transitions that women who choose to end abusive relationships are likely to face, documenting the stability of their networks may reveal patterns which indicate how this change has effected their availability of support and, in turn, their adjustment. One may expect the length of relationships to decrease as those who provide few of the supportive resources needed for successful adaptation to the life changes are replaced with more supportive persons. It may also be of interest to examine the course of 37 friendships ‘that. develop ‘with other' women that the respondent met while at a shelter to indicate the forms of support that those who share this experience can provide for each other. Closepess This variable may supply qualitative information about relationships. It may give an indication of the degree to which individuals rely on different network members for various support functions. For example, are 'there certain functions that close friends are sought out for, rather than other friends or acquain- tances, and vice versa? Also, one may be interested in the proportion of close friends in an individual's net- work. A In assessing the degree of closeness in a rela- tionship, some studies have provided the respondents with a definition of closeness (Caldwell & Bloom, 1982; Linn & McGranahan, 1980; Mitchell, 1982; Wandersman et al., 1980), while others have left the respondents to use their own conceptualizations (Eaton, 1978; Silberfeld, 1978). Other studies have used labels or descriptions that may indicate closeness, such as the "importance" of the network member (Lin et al., 1981: Mitchell, 1982; Norbeck et al., 1981) being able to visit a friend at any time without an invitation 38 (Duncan-Jones, 1981); listing of people who are consid- ered to be best friends (Makowsky, 1983); and naming someone with whom the respondent can share her/his most private feelings (Henderson et al., 1980). In the present study, an initial definition of closeness was provided ("A close friend is someone you can talk to about your feelings and to whom you can turn for help."; Bybee, 1980, p. 143). Further, a rat- ing of closeness was assessed by having the respondent indicate which of the following statements best described her relationship with each of her three clos- est friends: (1) There is nothing that I cannot talk to her/him about or that s/he would not do for me if s/he could. (2) I can talk to her/him about most of my feelings and I can rely on her/his help most of the time. (3) Sometimes I can talk to her/him about my feelings and I can sometimes rely on her/his help. (4) There are only a few things that I can talk to her/him about or rely on her/him to help me with. Other stud- ies in the social support literature did not include a rating of closeness, but only asked how many people the respondent felt close to. The rating scheme in the present study provided more information on the rela- tionship than a dichotomous distinction between "close" or "not close", regardless of the definition. 39 Although many studies included closeness, in some form, few reported any findings with respect to it. An exception was Silberfeld (1978) who, in comparing fam- ily practice and psychiatric out—patient females, found that the family practice patients had a higher propor- tion of their close relationships with relatives than did the psychiatric patients, who had a higher propor- tion of their close relationships with friends. Also, both groups had the same proportion of close relation- ships, but the family practice patients spent more time with their close network members. Considering the claims that isolation and depen- dency are characteristic of abusive relationships, one may predict that relatively few close relationships would be found in the social support networks of women who have been in abusive relationships. An examination of’ the frequency' and length. of close relationships should provide valuable information relating to this and other issues. Findings concerning social support appear cloudy and even contradictory, primarily as a result of the diversity of variables assessed and the ways in which they were measured. The use of a reliable, valid and standardized measure would help to clear up some of the confusion. Certainly a psychometric evaluation of all instruments should be conducted. However, there are a 40 few cautions to consider regarding the standardization and validity of measures. First, some of the findings that were presented above indicate that while a univer- sal measure of social support may facilitate compar- isons across studies on some dimensions, they may be of limited usefulness. In examining the idiosyncrasies of most social support measures, it appears that the con- structors carefully considered the unique char- acteristics of their population when developing the measures. This has served to document valuable infor- mation. For example, Cauce et a1. (1982), in their study of high school students included items on per- ceived supportiveness of teachers and guidance coun— selors. The findings of Hirsch (1981) and others sug- gest that instruments that are used with those diag- nosed as schizophrenic may want to include detailed items concerning interactions with family members, since they are frequently key members of their net- works. And those who are conducting research with women who have been in abusive marriages will certainly not want to make the assumption that "marriage equals support" as Eaton (1978) has. An issue concerning assessment of the validity of a measure of social support that researchers will need to consider is the belief of Pagelow (cited in Fleming, 1979) that many personality and other psychological 41 instruments which are considered to be valid and reli- able have been constructed with an underlying sexist bias. Using such instruments in an effort to establish concurrent validity could result in inaccurate conclu- sions and possibly damaging generalizations. (For example, a measure that had been standardized on an all-male sample may label women who realistically fear bodily harm at the hands of their spouses as displaying "paranoid" tendencies.) Interventions Aimed at Increasing Social Support Social support has been shown to assist people in times of stress or major life transitions. Therefore, it was believed that women leaving abusive relation- ships many benefit from interventions that increase access to social support. The literature was reviewed to locate interventions that were most relevant for this group of women. The interventions that were located generally follow one of three approaches. First, there are interventions which provide volunteers who are matched with individuals or families, often to teach skills and give information and referrals, as well as to offer more affective forms of support (e.g., emotional support). Second, a number of strategies have been used in an attempt to strengthen already 42 existing networks. Finally, peer support group strate- gies have been applied to provide those in stressful situations with a supportive place to discuss their concerns and meet others who may become part of their network of supporters. For all of the studies to be presented, the authors gave no indications of why they chose their particular model and whether or not they considered any other strategies before selecting their approach. Some suggestions were offered by Gottlieb (1981) to consider when designing such programs. Gottlieb suggests than when selecting a strategy, one should consider that individuals who are going through certain life transi- tions may benefit most from networks that have certain characteristics (such as the information presented above that women who were making role changes did best with low density networks), which may best be achieved by trying to mobilize existing network members or by introducing new supports. Additionally, the quality of the existing network and personality factors of the individual may indicate the most appropriate strategy. Vpluppeer Linking Stgategies Froland, Pancoast, Champan and Kimboko (1981), in their study of a sample of human service agencies and their utilization of helping networks, identified an 43 approach which they called volunteer linking strate- gies. These programs typically involved a one-to-one matching of a volunteer with a client. The volunteers often had backgrounds and experiences similar to the target individual. In addition to the volunteers pro- viding support to the client, they sometimes expanded the client's network by introducing them to their own network members. Another common use of volunteers has been to pro- vide assistance to new parents. One such intervention was the Lay Health Visitor Program described by Gray and Kaplan (1980). The lay health visitors modeled appropriate parenting behavior, supplied information on child rearing, and provided referrals when necessary. One of the major functions of the visitors was to pro- vide emotional support to the mothers. This involved round-the-clock availability, offering suggestions and reassurance, and providing friendship to isolated moth- ers. Results of this study are not yet available. In another study, Gray, Cutler, Dean and Kempe (1979) identified a population believed to be at-risk for child abuse and neglect. Families were selected from a hospital maternity ward and then assigned to either a high- or low-risk group based on interviews, questionnaires and observation. Those assessed as high-risk were randomly assigned to either a "High-Risk 44 Intervene" or a "High-Risk Nonintervene" group. The study also included a low-risk control group. The intervention included bi-monthly examinations of the infant; weekly home visits and referrals by public health nurses; and the assignment of lay health visi- tors to function as liaisons between the family and medical staff and provide emotional support to the fam- ily. The authors found no significant differences for any of their outcome measures (number of official reports of child abuse, number of indications of abnor- mal parenting, scores on child development measures, and others). They did find a significant difference between groups on the severity of injuries. However, they stated that less serious injuries were reported sooner in the intervention group which may have pre- vented more serious injuries from occurring. A third intervention involving home visits by volunteers to provide support for parents of new borns was conducted by Field, Widmayer, Stringer and Ignatoff (1980). In this program teenage, lower-class black mothers with pre-term infants were provided bi-weekly home visits. The visits were made by a "trained inter- ventionist" and a black, female work/study student. The half-hour visits included teaching developmental stimulation exercises for the mothers to practice with their infants. These exercises were selected to aid 45 the infants' development and strengthen the relation- ship between mothers and their infants. This study reported a number of significant findings which included more "desirable" child-rearing attitudes, less difficult infant temperaments (via mothers' ratings), more optimal face-to-face interactions and lower blood pressures for the infants in the treatment group than for the control infants. However, one interesting finding was that the mothers in the intervention group had significantly higher blood pressure. A possible explanation for this, presented by the authors, was that "by 'imposing' on these mothers middle-class mod- els of more stimulating interactions, we may have indi- rectly contributed to 'middle-class anxiety' and higher blood pressures in these mothers" (p. 435). In examining the social support resources provided by these studies, it can be seen that most of them clearly specified the assistance resources that they provided (e.g., referrals, health care), but none of them were specific in their descriptions of how "emotional support" was ‘providedm .And. more impor- tantly, none included objective or subjective indica- tions of their participants perceptions of the level of support provided by the interventionists. Addition- ally, the Field et a1. study (1980) was the only study 46 to report any significant impact from their interven- tion. However, even they did not indicate the degree to which the teenage mothers felt supported by the vis- itors or members of their networks (and, in fact, the mothers' increased blood pressure may indicate a decrease in well-being due to the intervention). Although based on only a few studies, there may be some limits of the use of volunteer linking strategies to increase the availability of social support to tar- get populations. One of the major shortcomings of this approach may be that, in all except the Froland et a1. (1981) study, support was only provided on a temporary basis. Referrals were often provided, but nothing was done to strengthen existing informal networks, which people often turn to first and prefer to receive assis- tance from (Young et al., 1982). In addition to the volunteer interventions generally being of a short-term nature, they often involved volunteers who have lived quite different life styles than the participants. The self-help movement maintains that those who have had similar problems or experiences can assist each other in. ways that. professionals are ‘unable to duplicate (Politser & Pattison, 1980). The agencies that Froland et a1. (1981) examined generally attempted to match their volunteers on background and experiences, but no data were presented on the effectiveness of the 47 attempt. Field et al., (1980) took the issue of cul— tural differences into account and included black, female work/study students with their trained interven- tionists (whose backgrounds were not described). How- ever, one cannot assess the extent to which this prac- tice decreased any cultural gap that may have existed between volunteers and participants. Natpzal Helper Stpategies In light of these cautions, it may be beneficial to consider interventions that attempted to strengthen existing networks, rather than instituting temporary, artificial programs. Increasing the availability of social support from informal sources has been most fre— quently accomplished by enlisting the aid of natural helpers; that is, people in a community who were cen- tral support providers to their networks. Addition- ally, informal support has been mobilized by creating or structuring events or situations that bring commu- nity members or neighbors together in an environment suitable for informal networking. The development of natural helpers as resources of support can follow any number of strategies. Several studies have created systematic interventions to improve the skills of natural helpers. These include two studies reviewed by Gottlieb (1981). The first project, by Wiesenfeld and Weiss, trained hairdressers 48 in listening skills to use with their clients. The Community Helpers Project by D'Augelli, Vallance, Danish, Young and Gerdes trained interested local citi- zens in listening, life development and crisis intervention skills, which they were to use with their own network members and to teach to other local resi- dents. Hirsch (1981) also presented preliminary pro- jects involving "network workshops," which taught net- work analysis to help people understand various network components such as density and multidimensional rela- tionships. It was thought that an understanding of these concepts and their effects would allow individu- als to restructure their networks to better meet their support needs. A structured attempt to strengthen the support that individuals received from existing networks was used by some of the agencies described by Froland and associates (1981). Agency personnel maintained contact with helpful or potentially helpful members of clients' networks. Most often, contact was initiated in response to a crisis, and then the agency staff contin- ued to encourage and support the helpers through phone calls and visits. A similar technique was discussed in Pancoast (1980). In this strategy, an agency staff person identified key individuals in a community or neighborhood who were presently connected to many 49 networks. The agency staff person, after establishing mutual trust, acted as a consultant to these natural helpers as they reached out to those in need of sup- port. The consultant worked within the previously established role that the natural helper held in the networks of those they were trying to reach. The major difference between this approach and that described by Froland et al. (1981), is that the consultants did not have direct contact with the targeted individuals. While all of the above mentioned interventions worked with existing network members, others have attempted to link or create informal supports. Two examples are given below of how neighborhood helpers were enlisted to help provide support to isolated peo- ple in need. In one community, the agency representa- tive located people in neighborhoods who looked out for their elderly neighbors and indirectly sought their help for other elderly neighbors by mentioning a need, which was usually followed by an offer to help (Froland et al., 1981). Another intervention was presented, where isolated target populations, such as the mentally iLl or developmentally disabled, were integrated into the community by encouraging them to participate in community activities or by creating activities for them and the larger community. In once such program, the clients cooked a weekly dinner at a local cafe' and 50 were thereby introduced to the regular customers (Froland et al., 1981) . Neighborhood work projects were organized for this same purpose of facilitating the development of informal support in SwediSh neigh- borhoods (Tiejten, 1980). Community projects included film programs for children, a neighborhood newspaper, a childcare co-op, and other projects. In comparing these natural helper interventions with. the 'volunteer’ linking' interventions, one finds that there are advantages and limitations of each. The major advantage of natural helper strategies is that they build on relationships that are likely to be stronger and more permanent than the volunteer linking strategies tend to be. These natural relationships have a greater potential to be reciprocal, multidimen- sional, and to include shared experiences and values, than relationships between the participant and a volun- teer. Unfortunately, this claim remains unsubstanti- ated since not one of these studies presented any data on the impact of the intervention on the helpees' eval- uation of the support made available to them. Wiesen- feld and Weiss (cited in Gottlieb, 1981) did look at the degree to which the hairdressers in their study learned the skills that they presented, however, they failed to assess the impact on their clients when they used these skills. Again, there is no evidence that 51 any of these interventions met their goal of increasing the availability or quality of support since none of them gathered this information from the target popula- tion. Another shortcoming of all of the interventions that attempted to strengthen existing network ties (except the ones described by Froland et al., 1981) is that they did not assess the network position of the natural helpers. For example, the hairdressers may have done an excellent job of implementing the listen- ing skills, but they may have held minor positions in their clients' networks, and therefore had little impact on clients' perceptions of the availability of social support (Gottlieb, 1981). A further potential hazard of increasing reliance on natural helpers is that they may try to help in areas where they lack the necessary skills, or they may reinforce maladaptive behaviors (Gottlieb, 1981). Most of the studies have taken steps that reduce the chances of this happening either through some form of training, supervision or consultation with the natural helpers, but again, no data is available on the impact of any of the helping projects to evaluate this. Peer Support Group Strategies A third form of intervention to increase the availability of social support are self-help and sup- port groups. Proponents of support groups claim that 52 there are specific and general positive effects on group members. Although support and self-help groups differ in a number of ways (e.g., size, leadership style), information from the self-help group literature is relevant to support groups. Although some of the reported benefits of support groups can also be acquired through other interventions that attempt to increase social support, support groups appear to pro- vide some unique advantages. One way in which peer support groups have been shown to be effective is in reducing symptoms associated with psychological dis- tress. Caplan (cited in Gottlieb, 1981) proposed this theory twenty years ago: when people experience the emotional and cog- nitive uncertainties accompanying life crises and transitions, they have a need to share and compare their own reactions and beliefs with others', preferably with persons cur- rently or recently experiencing similar events. When they have opportunities to do so, there is a moderation in the amount of stress they experience. (p. 220) This idea has been frequently expressed (Burnell, Dworsky & Harrington, 1972; Powell, 1975; Schwartz, 1975) and explored (Gartner & Riessman, 1982; Gottlieb, 53 1981; Politser & Pattison, 1980). For example, self- help programs for widows have reported shorter adjust- ment periods and better outcomes on depression, anxi- ety, self-esteem, coping and well-being for partici- pants than non-participating widows (Gartner & Riessman, 1982). Similarly, participants in mutual- support education groups for new mothers experienced less emotional distress six months after the birth of their child than did mothers in a control condition (Gartner & Riessman, 1982). And in a study of women with terminal breast cancer, those who participated in support groups experienced less tension, depression and phobias than those randomly assigned to a control group (Spiegel, Bloom & Yalom, 1981). A single case study, reported by Meyers-Abell and Jansen (1980) , on the effects of an assertiveness therapy group for residents of a shelter for battered women and their children, found a lower depression score and a higher assertive- ness score for one participant at one-year follow-up. On the other hand, all studies have not been able to demonstrate an improvement in psychological variables. McGuire and Gottlieb (1979) did not observe significant differences in levels of stress or well-being in their study of support groups for new parents. Nor did Spiegel et a1. (1981) find significant differences on 54 levels of self-esteem or denial in their study of women with terminal breast cancer. While these findings about changes in psychologi- cal variables‘are notable, a number of more behavioral outcomes lend impressive evidence for the potential efficacy of support group for interventions. An exper- iment involving a support group for after-care mental health clients reported the following findings. The randomly assigned intervention group had twice as many people functioning without any contact with the mental health. system, while ‘the control group 'members had twice as many re-hospitalizations, with an average stay that was three times longer than for those support group members who required re-hospitalization (Gartner & Riessman, 1982) . Other findings include those of Minde, Shosenberg, Marton, Thompson, Ripley and Burns (1980), who reported that mothers of premature infants in an experimental support group visited their infants more frequently while they were on the post—partum ward, and interacted with them more in the nursery and at home. Another study involving four to six year follow-up observed that mothers who had participated in a support group had healthier infants; fewer marital conflicts, sexual problems and physical illness; and more subsequent healthy children than control mothers (Gartner & Riessman, 1982). 55 Presentation and sharing of information is one of the most common forms of support that support generat- ing interventions claim to provide (Burnell et al., 1972; Gartner & Riessman, 1982; Gottlieb, 1981; McGuire & Gottlieb, 1979; Minde et al., 1980; Peterman, 1981; Powell, 1975; Schwartz, 1975; Spiegel et al., 1981). Some authors claim that access to such information may help to increase the participants' coping skills (Gartner 8: Riessman, 1982; Peterman, 1981; Schwartz, 1975; Spiegel et al., 1981); and may increase their problem solving abilities (Gottlieb, 1981; McGuire & Gottlieb, 1979; Powell, 1975; Tietjen, 1980). The information can be written or oral instruction by group facilitators, visits to the group by people from the community who are knowledgeable in particular areas, or experiential knowledge shared among group members. A. possible example of the effect that information sharing can have on group members was suggested by Minde et a1. (1980). In ‘their study' of parents ‘with premature infants, they suggested that the increase in noninstru- mental touching by the support group mothers may have been attributable to having their uncertainty about touching their infants reduced after attending as few as two support group meetings. In the self—help literature, the breakdown of the extended family in American society is hypothesized as 56 being responsible for inadequate provision of social support to a great number of people (Gartner & Riessman, 1982; Katz, 1965). Some of the studies that employed support group interventions were based on mem- bers of their population having inadequate networks that the group might help to improve. This inadequacy was seen to arise from either a shortage of supportive resources or the inability of the network to meet cur- rent needs (McGuire & Gottlieb, 1979; Pearson, 1983; Powell, 1975; Wandersman et al., 1980). Inability may also stem from a lack of acceptance of the changes that the individual is working toward (Powell, 1975; Wander- sman et al., 1980). Not only might support groups pro- vide legitimization for an individual's experiences and assistance, such as information and referrals, but they could also provide access to others who may become integrated into their networks. This has been reported to occur in some cases (Gottlieb, 1981; Schwartz, 1975) . Such examples include the support groups for women with breast cancer, where the groups continued for a year past the data collection period and when they finally ended, the surviving members continued to be part of an informal supportive network (Spiegel et al., 1981). In addition, participants in the support groups for parents of premature infants, conducted by Minde et a1. (1980) not only remained in contact after 57 the disbanding of the groups, but also formed a "Perinatal Society" and began establishing similar sup- port groups on their own. Such dramatic results are not always the case, however, as the participants in McGuire and Gottlieb's (1979) study of new parents indicate. Although nine out of the twenty-two participants had at least one outside contact with other support group members, few indicated that they had had continuing contact. Inter- estingly, they did discover that being involved in the support groups prompted more contact in the partici- pants' existing networks concerning child rearing issues than was found for those in the control con- dition. The use of a group format can allow for the occur- rence of other benefits that are difficult to replicate in the volunteer linking or natural helper strategies. Attending meetings with others who share a common situ- ation can provide the sense of being part of a commu- nity (Gartner & Riessman, 1982; Schwartz, 1975). This may be especially important in the early stages of a life transition when the current availability of sup- portive resources may need to be supplemented due to an increase in need (Gottlieb, 1981) . Becoming involved with a new group of people with whom one shares signif- icant experiences may replace or help to balance the 58 information that the individual receives from her existing network. As stated above, Caplan (cited in Gottlieb, 1981) has theorized that people who are face ing life transitions can experience a reduction in stress by being able to share and compare their experi- ences with others who have been in comparable situa- tions. McGuire and Gottlieb (1979) consider social comparison to be the central component of social sup- port; and their position was confirmed by twenty out of twenty-two of their participants. It is possible that self-comparison to a reference group derives its benefit from what Borkman (1976) calls experiential knowledge. She defines experiential knowledge as "truth learned from personal experience rather than truth acquired by discursive reasoning, observation, or reflection on information provided by others" (p. 446). Although Borkman stated that experi- ential knowledge and professional knowledge are not mutually exclusive, the strong reliance of self-help groups on experiential knowledge grew out of the Commu- nity Mental Health movement, which argued that profes- sionals were not necessarily the most effective helpers in all situations (Dumont, 1974). According to Bork- man, the experiential knowledge process involves the presentation of the experiences of a number of people, 59 from whidh common and unique characteristics surface. The range of attempted solutions reflects that each situation is different and therefore the individual must integrate the knowledge gained from others' expe- riences in order to create a viable strategy for her/himself. The belief that others who have shared the same concern can provide unique assistance has been echoed by many (Dumont, 1974; Gartner & Riessman, 1982; Gottlieb, 1981; Pearson, 1983; Powell, 1975). Another potential benefit of peer support group formats is the self-growth that can be acquired by assisting others. This phenomenon is sometimes called the "helper-therapy" principle (Riessman, 1965). This theory is based on the premise that when attempting to help another, the helpers are at least as likely to benefit from their efforts as those they attempt to aid. Riessman provided a number of possible explana- tions to explain this phenomenon: (a) When people per- form an action perceived to be worthwhile (i.e., help- ing another), an increase in self-esteem may follow; (b) Research has indicated that people are likely to increase their own belief in something as they attempt to (convince others; (c) "Peer' therapists" may feel important or of a higher status because of their helper role; (d) They may feel that they must be "ok" since they are able to help others with the same problem; and 60 (e) By focusing on another, they may be distracted from their own problem. As an indication of how this prin- ciple may be effectively applied, Riessman (1965) pro- vided an example of a program where sixth grade stu- dents who were experiencing reading difficulties were selected to tutor fourth graders who also needed to improve their reading skills. The result was that the reading abilities of both the fourth and sixth graders improved. Others have recognized the possibility of "helper- therapy" effects within group settings. Spiegel et a1. (1981) suggested that as group members are able to help one another they can reduce their feelings of helpless- ness and uselessness, while increasing their sense of worth and value. Schwartz (1975) has also stated that a support group is a place where people can provide support to others and feel more useful despite the neg- ative feelings that can accompany life transitions. Despite all of these apparent advantages of peer support groups, there are indications that they are not equally beneficial for all people in all situations. Some authors have proposed that those who refer others to various groups should consider the structure of the group when making recommendations. For example, although the structure of Alcoholics Anonymous appeals to many (as evidenced by its large membership) it also 61 has a high drop-out rate (Politser & Pattison, 1980). Durman (1976) suggested that one way to evaluate self- help groups is to examine their attendance records -- people will often reveal their opinion of a group by "voting with their feet" (p. 441). Durman has found that those who are most likely to attend and benefit from self-help groups are also people who are more likely to use traditional services (i.e., verbal, mid- dle class persons). Gottlieb (1981) reported some evi- dence that non-self disclosing people are more likely to suffer psychologically and medically when encouraged to self disclose, than non-self disclosers not ferced to self disclose. Keeping the previously mentioned cautions in mind, support groups have been described as a viable strategy to assist women who are or have been in abusive rela- tionships (Fleming; 1979; jMartin, 1976; Rounsaville, Lifton & Bieber, 1979; Walker, 1978). However, research findings to support this claim are scarce. Very' little research. has been. conducted on support groups for women who have been in abusive relation- ships. One descriptive account, by Rounsaville et a1. (1979), reported a problem with small group membership, but claimed that the group was helpful for those who did attend. Members of this group were quite eager to assist one another and had frequent contact outside of 62 the :meetings. One other’ study, which involved an assertiveness therapy group for shelter residents, described the group as offering a large amount of emo- tional support (Meyers-Abell & Jansen, 1980). Although this study included a control group and random assign- ment, difficulty in collecting post-treatment and follow-up data prevented the authors from drawing any conclusions on the group's effectiveness. Previous isolation and the failure of societal institutions to provide assistance are common ratio- nales for the selection of support group interventions. As Fleming (1979) wrote, support . . . groups seem especially valuable for battered women because they are even more isolated than other women. They rarely meet other battered women, and if they do, the odds are that they do not talk about their abuse. Shame, quilt, fear of reprisal, and the feelings that they alone have this prob- lem contribute to the isolation. Those who are not isolated tend to be surrounded by people who not only have no understanding of their plight or goals but who actually inter- fere with attempts to ameliorate or to get away from abusive situations. Our experience indicated that the group was the one stable 63 force in the lives of these battered women that was consistently operating in their behalf. The usual "magic" of the group pro- cess -- losing the feeling of being alone, giving support, ideas and encouragement to each other, and receiving understanding from women who have had similar experiences and who really understand -- is even more "magical" for the previously isolated bat- tered woman. (pp. 101-102) A support group can also provide its members with the opportunity to overcome some of the victimization that they have received from society. Martin (1978) contends that there is great value in "women supporting women as 21pm" because it serves to break down the barriers imposed by the professional "treatment" model which, in identifying the woman as having the problem, further victimizes and dehumanizes her. Peer support groups can provide a place for women to sort out their own responsibilities from those imposed by society (Martin, cited in Fleming, 1979). Additionally, they can recognize the sexist issues that have shaped their lives and that have worked to keep them in abusive relationships (Fleming, 1979). Support groups can also provide an opportunity to learn from the successful changes of others (Walker, 1978) and to share knowledge 64 and experiences of how to obtain resources that they have a right to receive from society. Also of primary importance, support groups can enable the group partic- ipants to develop mutual support systems (Martin, cited in Fleming, 1979). Summary In the last ten years, woman battering has gained legitimacy as a serious social problem. This problem was originally brought to public attention through the work of Erin Pizzy and her associates in England in the 19705 (Martin, 1976). Soon afterwards, the women's movement in the United States also became aware of the widespread threat of violence to women from their hus- bands and lovers, and began organizing shelters for battered women and their children. In the U.S., interventions for battered women have focused almost exclusively on assisting women to escape from violence and offering them a safe (though tempo- rary) haven. However, the present research and inter- vention efforts focused on another point where assis- tance may be equally important; these efforts concen- trated on women who were ready to leave the shelters and begin their attempts at rebuilding their lives apart from their abusive partners. There is a wide body of research that suggests that having a circle of supporters can assist people in 65 overcoming crises and major life transitions. There- fore, it. was. hypothesized. that 'the availability' of social support would significantly affect women's suc- cess at remaining free from abusive relationships. Many reports claim that battered women are frequently isolated, therefore, they may be at risk of returning to an abusive relationship or they may have greater difficulty in adjusting to the changes they face as a result of ending such a relationship. The present study attempted to examine the amount of social support that battered women perceived to be available to them, and in addition, it provided an intervention designed to provide these women with increased access to social support. This intervention was a support group offered to women who were in the transition period after ending an abusive relationship with an intimate male partner. It has been demon- strated that support groups can provide many benefits to participants. For example, support groups can help reduce isolation by providing a sense of community and by introducing the participants to potential network members. Additionally, support groups have been shown to assist people in initiating and/or maintaining behavioral changes. Support groups can also be a 66 source of new information, and can strengthen a mem- ber's sense of self-worth by providing them with oppor- tunities to help others. Description of the Study The present study involved a support group for women who had been in an abusive relationship with an intimate male partner. The group met for two hours a week for twenty weeks. The format of the group was selected by a few interested group members and the researchers/facilitators. (A detailed description of the support group can be found in the Methods chapter.) The goal of the present support group was to help increase the availability of supportive resources to the participants. It was anticipated that this increase in resources would come from community resources that members learned about at support group meetings, and that group members would provide emo- tional and material support to each other in- and out- side of group meetings. The primary focus of the study was on documentation of group members' social support. A secondary goal was to assist group members in their efforts to remain free from abuse. The present study was designed to evaluate the ways in which and degree to which the support group reached these goals. 67 Research Questions Based on the previous discussion, the following issues were addressed by the present study. Qesgrippipn pf Sogial Suppogt Netwopks First, this study provided a description of par- ticipants' social support networks. Average network characteristics were to be examined as well as signifi- cant network patterns. Social support network descriptions were gathered from responses to the Initial Interview (see Appendix A). The following network components were included in this description: network size; amount of contact: nature of contact; adequacy of support; network compo- sition; density; geographic proximity; stability and closeness. Information on social support available to abused women in support groups is not currently available; therefore, it is difficult to describe the average net- work. Certainly, one would want to examine character- istics that might indicate isolation (e.g. network size, amount of contact). It might be expected that women who have recently left an abusive situation would have a higher number of contacts involving practical assistance. For example, help to establish a new resi- dence and follow through on legal proceedings would be necessary. Also, one may expect that networks would 68 appear less stable, as women replace network members who do not approve of the woman's decisions concerning her relationship with her abusive partner with new, more supportive friends. Some of these new friends may be women that the participant met through a battered woman's shelter, or the present support group. Comparison of variables to reveal patterns among group members was done using the responses from the Initial Interview. Network variables were compared with demographic variables, such as attendance, living situation (i.e. residing with or apart from abusive partner), employment and others. One would expect differences to emerge between women who attend meetings frequently, and those who attended only once or twice. The less frequent atten- ders may have had stronger networks (i.e. larger, more contacts, closer relationships, and higher adequacy of support), or they may have been residing with their abusive partner and were restricted from attending. One may also predict that women who appear to be more isolated (e.g. fewer contacts) and have access to fewer resources might be more likely to re-establish their relationship with their abusive partner. It would seem likely, as well, that adequacy of support would be positively related to the closeness of relationships: 69 to the amount of resources that one perceives as avail- able; and to the number of supporters in one's network. Another variable that might be related to adequacy would be the extent to which network members understand abusive situations. It was expected that network mem- bers who had themselves been abused would tend to be more understanding. 0 on nt en ' o e s' c ess Su 0 've Resou ces In addition to learning about the social support networks of women who were in the rebuilding stage fol- lowing the dissolution of an abusive relationship, the research design also included efforts to evaluate the impact of the support group interventicn. One way in which the support group could have assisted its members was by providing supportive resources. Each of the research instruments provided an indication regarding provided resources and the ways they were provided. The resources which may have been provided include: commonality, assistance, emotional support, sociability, intimacy, and influence/role modeling. These supportive resources could have been provided in a number of ways. First, resources might have been received by attending support group meetings. Second, participants could have received supportive resources from other group members during support group meetings. 70 Third, group members could have provided resources to each other outside of the support group meeting. Fourth, support group members may also have received support from the facilitators during the meetings. Finally, social support may have been provided by the facilitators outside of the weekly meetings. It was anticipated that the group members would be actively involved with each other, both during meetings and outside of meetings. It was expected that most of the discussion and offering of support during support groups meetings would take place between the partici- pants, while the facilitators would generally be less active in the discussions. As with the level of involvement in the meetings, it was hoped that many of the women would have regular contact with each other outside of the meetings and provide supportive resources to each other. It was anticipated that the facilitators would occasionally provide support to mem- bers outside of meetings, but that this would be to a lesser degree than the support exchanged between group members. Provision of social support was expected to include all of the six possible resources listed above. m ac t e Su ort Grou Interve tion 0 the i i ants' Soc a Su ort Networks In addition to offering supportive resources to group members, the support may also have had an effect 71 on the social support network of the participants. In assessing the impact of the support group on the net- works of the participants, changes over time were exam- ined. This involved a comparison of responses on the Initial and Follow-up Interviews. The major expected finding was that participants would report friendships with other group members on their Follow-up Interviews. Additionally, it was thought that the women might have less contact with less supportive network members, and more contact with those who supported their decisions concerning their relationships with their abusive partners. The find- ings of McGuire and Gottlieb (1979) that members of a support group for new parents discussed child rearing issues with their network members more often than those in a comparison group, suggested that participants in the present study would discuss the impact of their abusive situation more often than they had before join- ing the group. Explanations for this could include that the participants wanted to further explore new ideas brought up during support group meetings, or that after participating in the support group meetings they may have felt less stigma from the abusive relation- ship, and therefore felt freer to discuss their experi- ences . METHODS Support Group History Sponsoring Opganization The sponsoring organization was a shelter for women who had been in abusive relationships. It was located in a Midwestern city with a population of about 130,400. This shelter accommodated up to thirty women and children. In addition to providing safe housing, their services included a 24-hour crisis line; legal, housing and other referrals; individual counseling: support. groups; and. additional skill building ‘work- shops. The agency provided access to ex-residents' files so that the support group facilitators could con- tact those women who had indicated an interest in the Ex-residents' Support Group. crisis phone volunteers also informed callers of the option of joining the sup- port group. If a caller was interested, a message was left in the facilitators' message box. Additionally, the shelter provided a meeting room, childcare room and childcare workers for the last fourteen weeks of the twenty week data collection period. 72 73 Begearche; Involvement At the outset of the data collection period, the researcher had been a direct-service volunteer at the sponsoring organization for seventeen months. One year before data collection began, the researcher and a social work graduate student started a support group for shelter residents, ex-residents and non-residents (women who had been in abusive relationships but had never stayed at a shelter). After seven months the social work graduate student left the organization and another volunteer became the new co-facilitator. About three to four months later, in response to their obser- vations and feedback from some of the members of this "original support group," the facilitators began meet- ing with interested support group members to discuss and plan the formation of two separate support groups (described below). At the same time, the facilitators proposed a data collection procedure for the group they would be facilitating; the "Ex-residents' Support Group." The researcher met with the program director of the shelter to present the proposal for the Ex- residents' Support Group and to discuss the selection of facilitators for the other group; the "Residents' Support Group." Although the Ex-residents' Support Group did not meet at the shelter during the first six weeks of the 74 study, the researcher continued as a direct service volunteer there throughout the duration of the study. Later, it was decided to return the Ex-residents' Sup- port Group meetings to the shelter. gpopp Development In designing this support group, the facilitators consulted with members of the original support group to determine the meeting place and time, group format, advertisement, and other details. It was decided to hold the meetings at a location away from the shelter. The reasons for this were that ex-residents might feel more comfortable at a "neutral" location and the group might not want to be restricted by the shelter's poli- cies. One of the women from the original support group checked into possible meeting places, while another helped with the design of flyers to announce the forma- tion of the new group. The original support group consisted of a combina- tion of shelter residents, ex-residents, and non-resi- dents who had experienced a physically and/or emo- tionally abusive relationship with an intimate male partner. The idea of forming two separate groups came after the facilitators and some of the original support group members noticed that the women who were staying at the shelter seemed to have different needs than those who were out of the relationship and trying to 75 establish a life without their abusive partner. Although it was originally felt that the women in these different situations could learn from and help each other, most of the group's energies went to the shelter residents who were in crisis, while many of the ex- and non-residents' needs went unmet. It was hypothesized that the new "Ex-residents' Support Group" would concern itself with issues that women starting over on their own would face. Some such issues were thought to be: loneliness, raising chil- dren alone, financial problems, divorce and child cus- tody concerns, feelings for their assailants, threats and harassment from their assailants, and forming new relationships with men. The researcher anticipated that this group would have a more stable membership and that the women would bond together to form friendships and cooperative services such as babysitting, carpool- ing, etc. (This would help to replace some of the resources that they might have lost as a result of leaving their abusive partner.) The goal of the group was to provide women in this transition period with a place to discuss their concerns in a safe and under- standing environment and to increase their access to resources so that they did not feel compelled to return to their abusive relationships. 76 On the other hand, it seemed that the nature of the new "Residents' Support Group," which would meet at the shelter, would be more crisis oriented and serve as a forum for women very recently out of their abusive situation to explore their options and feelings. It would be a consciousness raising opportunity for them to begin to understand the dynamics, patterns and cycles of battering relationships. It could also be a place to experience the empowering reality that they are not the only one to experience an abusive relation- ship. From experiences in the original support group, it was expected that this group would have a high turnover in its membership and much repetition in the topics discussed. It was also feasible that upon leav- ing the shelter, some of the women would transfer from the residents' to the Ex-residents' Support Group. It seemed logical that those non-residents who had left their abusers would benefit most from the Ex- residents' Support Group, while those who were working on getting out of their abusive relationship, those trying to work things out by staying in the relation- ship, and those who were still in the process of making this decision, would find the Residents' Support Group most useful. In practice, however, the distinction between the groups was not this clear. All of the non-residents 77 from the original group who continued to attend support group meetings came to the Ex-residents' Support Group, regardless of their living situation. This may have been because they were already familiar with the facil- itators and group members who were involved with the Ex-residents' Support Group. They also may have felt that the goals of the new group would best meet their needs. (In fact, two of the non-residents in the orig- inal group who were living with their abusive partners were involved with the planning of the new group.) It was decided that new non-residents who expressed an interest in joining a support group would be given a description of each group and be allowed to choose which group matched their needs. The Ex-residents' Support Group was the focus of this study. Information concerning the support received by women outside of the shelter and the role of the support group, which members helped to design, were observed and recorded. Research Participants There was a total of eighteen women who attended the support group meetings during the twenty week data collection period. Thirteen were given the Initial Interview. One was a shelter resident who came to part of one meeting to hear the visiting speaker. The researchers were unable to gather complete interview 78 information from the remaining four women after repeated attempts were made to contact them (all attended only once and were unable to be interviewed following the meeting. Of the thirteen research participants who were interviewed in the present study, 85% were white (n = 11), while 15% were black (n = 2). Sixty-nine percent (n s 90) were ex-residents, while 31% (n = 4) were non- residents. The mean age of the participants was 34.5 (with a range of 20 to 54 years of age). Most of the women (62%, or 8 women) relied on government assistance for the majority of their income, while 23% (n = 3) were employed outside of their home, and 15% (n = 2) were financially dependent on their abusive partners. Monthly family incomes ranged from $214 to $1,700, with an average of $574 per month; family size ranged from one person to four people, with an average of 2.3 peo- ‘ple (equaling an average of about $250 per person, per month). Sixty-nine percent (n = 9) of the research participants had graduated from high school. Their level of education ranged from two to fourteen years of school completed, with an average of 10.8 years. The average number of children for support group members was 2.5 (ranging from 0 to 8). Attendance at the sup- port group meetings ranged from one to 15 meetings, with an average of 4.5 meetings attended during the 79 twenty weeks. Eighty-five percent (n = 11) of the par- ticipants had dropped-out of the group by the end of the data collection period (that is, they failed to attend any of the last four meetings). Sixty-nine percent (n = 9) were married to their abusive partner at the time of assault, and 77% (n = 10) were living away from their abusive partner at the time of their Initial Interview. The women had known their abusive partners an average of 11.6 years (with a range of 2 to 23 years). On the average, the length of abuse equaled 4.5 years, while the range was from eight months to 11 years. Eighty-five percent (n = 11) of the women experienced both physical and psychological abuse, while 15% (n = 2) reported that they had been physically, but not mentally abused. The mean number of monthly contacts that women had with their abusive partners equaled 10, with a range from "daily" to "never". women in this sample had left their abuser twice, on the average; and those who were out of the relationship had been away for an average of eight months, ranging from six days to three years. Addi- tionally, 38% (n = 5) of the women had pressed assault charges against their abusive partners. Beezpltment Support group participants were recruited in the following ways: (1) Those who transferred from the 80 original support group (46%). (2) Residents who left the shelter were called and invited to join the new group if they had indicated an interest on their shel- ter exit form. If the woman did not have a phone, and was not residing with their abusive partner, a letter of invitation was sent to her (as a safety precaution, she was not contacted by mail if she was residing with her abuser) (15%). (3) Shelter crisis phone workers were asked to give callers information on the support groups and to leave a message for the facilitators to contact any potentially interested women (8%) . (4) Approximately twenty-five flyers were posted at various locations throughout the area informing women of the existence of the support group and of a telephone num- ber to call for more information. (Locations included Department of Social Services bulletin boards, laundro- mats, grocery and convenience stores, church bulletin boards, women's fitness salon bulletin board, public bulletin boards, etc.) (0). (5) Eighty-three letters were sent to local churches, hospitals and doctor's offices asking them to include a notice about the sup- port group in their bulletin or newsletter (8% - from a church bulletin). Additionally, a number of women, who had attended meetings of the original support group returned to attend the Ex-residents' Support Group without further contact from the facilitators (23%). 81 lpltlal Qontaet When the facilitators called a potential partici- pant, they explained that: the group consisted of women in a variety of living situations, all had expe- rienced physical and/or emotional abuse, the location of the meetings must be secret, she would be asked to participate in an interview following the meeting to evaluate the support group, but that she was under no obligation to do so. She was also informed that: observers would be present at the meetings for research purposes and would not record the content of her state- ment, she had the right to request that observations of her not be recorded, everything that she said would be completely confidential, child care and transportation were available. Finally, any questions were answered. Bestpictions The only restrictions imposed on group members were that they not be under the influence of alcohol or illegal drugs and that they not reveal the location of the meetings to anyone to insure the safety of all. Generally, shelter residents were also restricted from the Ex-residents' Support Group meetings because it was felt 'that their’ concerns ‘would. be 'more effectively addressed in the Residents' Support Group. However, when the meetings were held at the shelter, residents 82 were invited to attend presentations on the occasions that special speakers were brought in. Procedure Mine The first six meetings of the Ex-residents' Support Group were held in the basement of a church. These meetings ran from 12:30 to 2:30 p.m. on a weekday afternoon. The first meeting was held in the kitchen. Folding chairs were arranged in a circle and refreshments were available. A problem arose with church staff interrupting the meeting to get coffee and so the meetings were moved to another room. The next three meetings were held in a partitioned area of a very large room. In this room, the folding chairs were arranged around a long rectangular table. As in the previous week, refreshments were available. The fifth and sixth meetings were held in yet another room. This room offered paneled walls, carpeting and greater pri- vacy. Once again, the folding chairs were arranged around a rectangular table and refreshments were avail- able. Due to low attendance at the afternoon meeting time and dissatisfaction with the churches' facilities, the remainder of the Ex-Residents' Support Group meet- ings were held at the shelter. These meetings ran from 7:00 to 9:00 p.m. in the Library. This was a large room (at least twice the size of the rooms provided by 83 the church) with two soft couches, which faced each other, two rocking chairs, a soft cushioned chair, and occasionally other padded arm chairs. A small table was located in the center of the seats. The room was carpeted, and also contained bookshelves and a fire- place. Many of the women were familiar with this room due to their stay at the shelter or from attending the original support group. Refreshments were not made available because they were rarely used. Smoking was not allowed at either the church meeting rooms or in the shelter Library. E .J.! ! The facilitators of the Ex-residents' Support Group were two white, female graduate students in Eco- logical Psychology. The researcher had previously been involved as the facilitator of the original support group at the sponsoring organization for one year prior to the formation of the Ex-residents' Support Group, while the research partner had been involved with the original support group for five months. The facilita- tors also acted as the researchers of this study. Each week the facilitators attempted to contact the women who had left the shelter during the previous week and indicated an interest in the group. The facilitators also contacted women who frequently needed rides to arrange transportation, if necessary. 84 The role of facilitators in the group discussions was usually to initiate conversation in the beginning of the meeting until the others began to carry the dis- cussion on their own. Occasionally, the facilitators would address a woman in attempt to include her in the discussion and encourage her to participate. Often, when a typically quiet woman made a comment, the facil- itators tried to reinforce her participation by acknowledging her contribution to the group. Another function of the facilitators was to help redirect the conversation back to the issues at hand when they felt that it had digressed to "trivial" subjects, or when more than one conversation was taking place at the same time. Generally, the facilitators attempted to keep their participation in the group to a minimum, believ- ing that the group members would learn more from each other. Neither of the facilitators had been assaulted by an intimate male partner, and so they did not feel that they had the practical expertise of group members. It was also hoped that the less active role of the facilitators would encourage cooperation among the women, and help to discourage dependency on the facili- tators. Despite efforts to avoid. being considered "authorities" in the group, it seemed that the facili- tators ‘were often. viewed that. way; For instance, requests for advice and information were directed 85 toward the facilitators more frequently than to the other group members. e 'n s The Ex-residents' Support Group met for twenty weeks in the winter, spring and summer of 1983. The meetings lasted for approximately two hours, and met once a week. Attendance at the meetings was completely voluntary and transportation and childcare were pro- vided free of charge so that more women could attend. The meetings usually started late because the child care workers were often late in arriving. Occa- sionally, the meetings would be delayed until at least two women were present. The meeting began with intro- ductions (first names) and then it was stated that the location of the meeting and everything that was said was to be kept confidential. In addition to the group members and facilitators, two research assistants attended each meeting. They sat apart from the rest of the group and completed a behavioral observation form (described below). If new women were present, it was explained that the behavioral observers were there to learn about support groups and that they would not write down or repeat anything that any of the women said. It was almost always the case that the behav— ioral observers seemed to be forgotten as soon as the meeting got underway. Generally, discussion was slow 86 at first. Facilitators' past experience indicated that discussion flowed more smoothly when there was no attempt to introduce set topics (except in the event that a film *was shown or a speaker ‘was present). Invariably, it seemed that the chosen topic at the pre- vious meeting was irrelevant or of little interest to the majority of women at the subsequent meeting. This was to be expected due to the attendance patterns of the group members. Therefore, it seemed most bene- ficial that the format be one in which the women had the freedom to discuss the issues currently of greatest concern. The open discussions that ensued tended to cover many topics in varying degrees of detail and intimacy (for a complete discussion of topics, see the Results section). The meetings usually ended on schedule. One of the facilitators would give notice that it was "time to rescue the child care workers." The facilitators often thanked the women for coming and wished them a good week ahead. From time to time, women would exchange phone numbers. Next, the first time attenders were asked if they were available to be interviewed and, if they agreed, arrangements were made. This procedure was also followed for women who were due to receive a Four-week Follow-up Interview. Finally, transportation 87 was provided by one of the facilitators to those need- ing a ride home, while the other completed the Group Content Form (see ZMeasures section. below) with, the research assistants. agmlplstration of Measures te ' w Following the support group meeting, first time attenders who agreed to be interviewed were introduced to a research assistant who took them to a private room to conduct the interview. When a woman was unable to be interviewed following her first meeting, every effort was made to make alternate arrangements for a research assistant to meet with her within one or two days. Typically, the women would request that the interview take place following the next week's meeting. In all cases, every effort was made to respect the women's wishes.3 Generally, the interviews took from thirty-five to forty-five minutes to complete. However, there was a range of twenty minutes to ninety minutes. Attached to the front of the Initial Interview were the Interview Consent Form and a Cover Sheet. The Cover Sheet was used to record the woman's name, address, telephone number and times to call for follow-up purposes. The women were never pressed to give any information that 88 they were uncomfortable divulging. Immediately follow- ing the interview, the Interview Consent Form and Cover Sheet were removed and separated from the Initial Interview to assure confidentiality of the responses. - 'ew It was intended that the Follow-up Interview be administered immediately following the fourth meeting that a group member attended and after every fourth meeting thereafter. These meetings need not be attended consecutively. The same procedures were used to administer the Follow-up Interview that were fol- lowed for the Initial Interview. Namely, following the meeting, the woman would be approached by an inter- viewer to request that she participate in the Follow-up Interview. Again, the options of non-participation or arranging a more convenient time and/or location were presented. The interviews were conducted in a private location and generally took between ten and thirty min- utes to complete. - u e The Drop-out Survey was conducted if a woman had missed four support group meetings. Information from the cover sheet of the Initial Interview determined whether or not the woman had agreed to be contacted and at what times it was safe to do so. If the women had a 89 phone, she was contacted by one of the research assis- tants. If she did not have a phone, the research assistant stopped by her home to arrange a time for the interview. The Drop-out Survey consisted of nine ques- tions. Its administration time ranged from approxi- mately five to twenty minutes. gzepp Earpielpetipn Form The Group Participation Form was completed by the behavioral observers during each support group meeting. Each statement that a woman made was recorded by a tally mark in one of the eight categories described below. The two observers sat on opposite sides of the room, well outside of the group circle. As a measure of’ reliability, the observers were trained to code everyone present (including the facilitators) for the first 15 minutes of the meeting. They signaled each other to indicate the beginning and end of this relia- bility period. For the remainder of the meeting, they coded the women who were on the side of the room that faced them. The observers were instructed to code each statement that was made. If a statement did not fit the definition that was given for any of the other cat- egories, it was to be classified as an "Other" state- ment. If the observer was unsure about how a particu- lar statement should be classified, she was instructed to write it down and ask the researcher or research 90 partner. The behavioral observers were instructed to be as unobtrusive as possible, and to create as little disruption as possible. The amount of time that it took to complete the Group Participation Form was nec- essarily equal to the length of the support group meet- ings. mm It took approximately five to ten minutes to com- plete a Group Content Form. The basic procedure fol- lowed was that the facilitator and behavioral observers recorded as much information as could be recalled, as in a "brainstorming session." Also, the behavioral observers often jotted down topics discussed and state- ments made during the meetings on scrap paper or in the margin of the Group Participation Form. This served to increase the accuracy and completeness of the Group Content Form. IDEEIELQWQI end Benevioral Qbserver Training Nine female undergraduate students received course credit for their participation as interviewers and behavioral observers. These research assistants also attended an initial training period and weekly supervi- sion meetings. lpipiel Tgeiping The initial training involved three, two hour meetings and included general instruction on empathy 91 skills and research methods in community psychology. Information was also presented and discussed to help provide an understanding of the dynamics of abusive relationships. This included giving statistics on the prevalence of women battering, dispelling common myths, and examining the flaws in victim blaming approaches to this problem. The research assistants were also taught inter- viewing skills and were familiarized with the instru- ments to be used in this study. Interviewing skills included presenting the questions in a conversational tone and the importance of being prepared to use empa- thy when discussing emotional subjects. They were trained to encourage the women to answer questions as thoroughly as possible without attempting to influence their responses. However, they were told to make it very clear to the women that they were not required to answer questions that they did not wish to answer and that they could choose to discontinue the interview at any time. The interviewers were also trained to inform the interviewees that there were no consequences attached to the decision to terminate the interview. They were further instructed to record exactly what was said without making any interpretations of what a statement might "really" mean. To test and practice these skills, the research assistants participated in 92 administering the interviews to each other under the supervision of the researchers. Training for behavioral observation involved a thorough explanation of the group participation form and numerous hypothetical statements which the research assistants were to categorize. They were also taught how to identify each woman in the group and to signal each other unobtrusively when the coding began and ended. It was stressed that the observers should remain silent and create as little distraction as pos- sible during the meetings. It was also stressed at each meeting that maintaining confidentiality was cru- cial. This meant that the research assistants promised not to reveal the location of the meetings, any infor- mation about the women, or what was said at the meet- ings to anyone other than those involved in the research study. W Me 'n s The researchers met with the research assistants for approximately one hour each week throughout the twenty week data collection period. During these meet- ings, interviewing and coding problems were discussed and techniques were reviewed and practiced. This was also the time at which the following week's interview- ers and behavioral observers were selected and given the necessary materials. The research assistants 93 rotated the observation and interviewing assignments based on the number of course credits that they were earning and transportation arrangements. A major focus was clarifying difficulties in categorizing statements for the Group Participation Form. Most of the research assistants experienced difficulty in making appropriate distinctions. And reminders were given at these meet- ings concerning the importance of maintaining confiden- tiality. Termination pf the Group Because group size was small and the Ex-residents' and Residents' Support Groups were more similar than originally planned, the two groups were recombined at the end of the twenty week data collection period. This was announced at the last three or four meetings so that participants would be aware of the change. Measures This study involved two types of measures: Inter- views and behavioral observations (see Appendices A-E). There were three interviews, the Initial Interview, the Follow-up Interview and the Drop-out Survey. The Ini- tial Interview was administered right after the first meeting that the research participant attended. The Follow-up Interview was given to a woman following the fourth meeting that she attended, and after each fourth subsequent meeting. These meetings did not have to be 94 attended consecutively. The Drop-out Survey was used when a woman had not been present at four consecutive meetings. The two behavioral observation measures, the Group Participation Form and the Group Content Form, were completed at every meeting. lpperviews lnigiel lpperyiew Attached to the front of the Initial Interview were the Interview Consent Form and the Cover Sheet (see Appendix A). The cover sheet was used to record the woman's name, address, telephone number and times to call for follow-up purposes. Qeveloppentleopstrpctiop. ‘ Based on feedback from the research assistants during practice administra- tions, minor revisions were made on the wording and sequencing of items in the Initial Interview. 1. Demographic items: Basic demographic items were chosen which would provide a description of the sample. This information included race, age, number and ages of children, education, employment status, amount and source of income, and relationship to and contact with the abusive partner (see items 1 to 20). 2. Attitudes/evaluation of group: Since little had been done in the way of evaluation of support groups, feedback was sought concerning the structure 95 and content of the Ex-residents' Support Group. Its purpose was to reveal those aspects that were most ben- eficial, in their own opinion, to the women in the group. The items focused on the group's structure and format and solicited the woman's expectations, com- ments, and suggestions concerning the support group. Included were such components as time, place, topics, other group members (items 22-39). 3. Social support: .As the jprevious chapter noted, women in abusive relationships are often iso- lated from both people and resources. Therefore, social support information was of key interest to the researcher. The Initial Interview attempted to collect data on various aspects of the support group members' networks. By documenting these networks, comparisons could be made for individual women at varying points in time and in varying living situations, and with women who had not experienced abusive relationships. The following items were devised in an attempt to gather information about the social support networks of the women who attended the support group meetings. It was expected that this information would be useful in describing the group and in comparative analyses. These are items 40-76, and 80 (see Appendix A). Some of these items asked about the woman's relationships in general, while others focused on her relationships and 96 their impact on her experiences of having been in an abusive relationships. It was anticipated that the responses to the items would fall into the various social support components, defined in the Introduction: commonality, assistance, intimacy, sociability, emo- tional support, and influence/role modeling. Items 40, 64, 66, 71, and 76 were adapted from Bybee (1980). A number of items provided more than one type of informa- tion. For example, items 34-39 gave useful information on the satisfaction of a respondent with certain compo- nents of the group (e.g. other members) as well as information that related to perceived support. Addi- tionally, items 35 and 36 provided information useful for comparison with follow-up responses and indicated within group friendships and support. gapegggiee. The Initial Interview consisted of 80 items; most of which were open-ended. The items in the Initial Interview were divided into the following cate- gories (see Appendix A). (Some items are in more than one category because the response contained information pertinent to more than one research question.) 1” Demographic Information (items 1-8) included race, age, SES, and data on children. 2. Relationship with Abusive Partner (items 9-21) included questions on the history of the abuse and cur- rent status of the relationship. 97 3. Evaluation and Attitudes Toward Group sought feedback on various aspects of the Ex-residents' Sup- port Group (items 21-39). 4. Social Support consisted of (a) development of supportive relationships between support group members (items 34-39); and (b) present social support network (items 40-76). This category requested information on the relationship of the woman to those in her network; gathered data on the nature, amount and feelings about her contact with those in her network; and asked about the support received from those in her network specifi- cally relating to the relationship with her abusive partner. 5. Influence of Others was assessed directly by item 80. Indirect assessment could have been provided by items 77-79 through examination of the frequency with which support group members stated that support from others was important or helpful in their decisions or actions concerning their abusive relationships. 0 w-u te 'ew The purposes of the Follow-up Interview (see Appendix B) were to get an evaluation of the support group from the participants and to follow changes over time in the social support networks of the attenders. A component part of this was to determine any impact 98 that the support group had on the social support net- works, such as the development of friendships within the group. The Follow-up Interview was very similar to the Initial Interview. Fifty-five of its 71 items were identical to those in the Initial Interview, therefore, it allowed for comparisons to be made with responses on the Initial Interview. The Follow-up Interview was also predominantly open-ended. None of the demographic items from the Initial Interview appeared on the Follow-up Interview except those concerning the abusive relationship and contact with the abusive partner. Additional items asking for an evaluation of the support group were added. It was felt that attendance at four meetings would be suffi- cient exposure for a member to give meaningful feedback and to make useful suggestions. A set of items were developed to elicit the extent and ways in which the group members felt that the group had had an impact on them. Additionally, a small number of items were added that asked about contact between group members outside of the meetings. These items were included to examine the development of any supportive relationships that emerged from the support group. All of the Social Sup- port items from the Initial Interview were in the 99 Follow-up Interview. The development of these items was described previously. gepegpziee. The Follow-up Interview contained the same categories as the Initial Interview, except for the Demographic Information category. The majority of the items were identical to those found in the Initial Interview. Added items were as follows. 1. Evaluation and Attitudes Toward Group (items 6-19, and 30). 2. Social Support: development of supportive relationships between support group members (items 20- 29). DIQR:QE§_§BI!§¥ The main purpose of the Drop-out Survey was to evaluate the Ex-residents' Support Group (see Appendix C). Of primary interest were the reasons that women stopped coming to the meetings. It was expected that this information could be used to improve the support group. Due largely to potential dangers of contacting women who are residing with their abusive partners, and the fact that in choosing to leave an abusive relation- ship, women are often cut off from the financial resources necessary for telephone services, it was often difficult to contact women who had stopped coming to support group meetings. In addition, women who had 100 attended the support group moved frequently. In some cases this was for safety reasons and in others it was due to a reconciliation or separation from the abusive partner.4 Because of these obstacles to contacting women, it was generally very difficult to get accurate informa- tion on why women stop attending support group meet- ings. The Drop-out Survey was an attempt at discover- ing some of these reasons. The researchers felt that the main reasons would be that the women were dissatis- fied with one or more aspects of the group, or that it was too difficult to get to the support group meetings. Items 1, 2, 4, and 5 provided the opportunity for the woman to express dissatisfaction with certain aspects of the support group. Items 6, 7, and 8 attempted to discover if there was something about the woman's rela- tionship with her abusive partner that affected her attendance at the support group. Another function of the Drop-out Survey was that it provided follow-up information. In fact, all except one question (item 1) were the same or similar to items on the Follow-up Interview. Ltepe. Four of the questions asked about the research. participant's opinion. of 'the support. group (items 1, 2, 4, and 5). Another questioned the women regarding contact with other women from the support 101 group (item 3). There was also a set of questions con- cerning her current living situation and contact with her abusive partner (items 6, 7 and 8). The final item requested any further comments (item 9). The inter- viewer ended by thanking the woman and by reminding her to contact the group if she needed assistance. 3 E !I . !° E The purpose of the Group Participation Form (GPF) was to provide a record of participation patterns (see Appendix D), such as who spoke and the types of state- ments that were made. This information was collected by recording the frequencies of certain types of state- ments. These categories (described below) were selected for their relevance to the research questions. Included were statements that were evaluative of the support group ("Positive Toward Group," and "Negative Toward Group"), and statements indicative of the level of“ support. provided. by' the lgroup ("Positive Toward Woman," and "Negative Toward Woman"). These categories were selected based. on the jpast experience of the facilitators regarding the types and frequency of cer- tain statements. The Group Participation Form was designed to work in cooperation with the Group Content Form (described below). While the Group Participation Form contained frequencies for each of the categories, 102 the Group Content Form contained examples of statements for most of these same categories. Deyelppmepngppepgpgpien. Due to the concern for the safety of the support group participants and the promise to insure complete confidentiality, the behav- ioral observation method was selected over other proce- dures which would have provided richer and more accu- rate data. For example, tape recordings or video-tapes of the meetings would have provided a higher quality of data, however they also involved increase risk of vio- lation of confidentiality, due to the possibility that the identities of the participants could be discovered. Additionally, the behavioral observation method that was selected allowed for the option of not collecting data on women who requested this, however, video-tapes and tape recordings could not provide this option. Deeeripgien pf Categories. 1. Member I.D.: to insure confidentiality, the women's names, except for the facilitators, were removed after the meetings and replaced with identifi- cation numbers. The behavioral observers were instructed to start with the researcher as the first person on their sheet and then proceed clockwise around the circle. This served to guarantee that each woman could be correctly identified on both Group Participa- tion Forms in the event that the behavioral observers 103 did not know a woman's name (or if there was more than one woman with the same name). 2. Positive Toward Group: favorable statements about the support group were recorded in this category. It was believed that frequency and content of such statements could provide useful evaluation information. In addition, the researchers could look at the charac- teristics of women who made frequent Positive Toward Group statements as an indication of who found the group most helpful. 3. Negative Toward Group: criticisms (or sugges- tions for improvement) made by group members were indi- cated by tally marks in this category. Such comments could assist in redesigning this or other groups. Additionally, these statements could provide insights about why women dropped out of the group. 4. Positive Toward Woman: this category included statements made by one woman toward another that were encouraging, supportive, complimentary, etc. It does not include statements made about women who were not present, even when they were members of the support group. It was expected that the data from this cate- gory would give an indication of the extent that the women gave and received support in the group. 5. Positive Feedback: this category was created to correspond directly to the "Positive Toward Woman" 104 category. Whenever a positive statement was made toward another woman, the woman that the statement was directed toward received a tally mark in her Positive Feedback box. This indicated who was receiving the most (and least) verbal support in each meeting. 6. Negative Toward Woman: statements that were classified as Negative Toward Woman included criti- cisms, insults and other unfavorable comments made by one member of the group toward another present member. 7. Negative Feedback: this category corresponded to the "Negative Toward Woman" category. Whenever a negatively coded comment was made toward another woman, the woman that the statement was directed toward received a corresponding Negative Feedback tally mark. 8. Positive Toward Self: favorable comments that a woman made about herself were included as positive toward self comments. Examples may have included accounts of how she asserted her rights, an affirmation of her own strengths or talents, self-confidence in judgements or decisions, etc. As verbal attacks and degradation almost always accompany physical violence (Roy, 1977; Star, Clark, Goetz & O'Malia, 1981), the facilitators tried to encourage the women in the sup- port group to make positive statements about them- selves. The facilitators believed that such comments made in the presence of others served to affirm the 105 self-worth of the speaker and also stood as a model for others. 9. Negative Toward Self: when a woman made a statement that criticized her own behavior, thoughts, feelings, etc. , she received a Negative Toward Self tally mark. 10. Directive/Helpful: helpful comments were those that provided information to one or more group members. They may have included sharing of knowledge about legal guidelines, community resources, etc. Helpful statements could also include offers of assis- tance, such as a ride home, etc. Directive referred to the function, usually performed by the facilitators, of maintaining smooth group interaction by beinging "irrelevant" discussions back to the general topic. Another exampLe of a Directive statement could be the announcement that the meeting was over. 11. Other: the Other category encompassed all statements that were not included in any of the above categories. These statements were recorded without regard to their content because they provided an indi- cation of the amount of input the members contributed to the group's discussions. G on on o The Group Content Form (GCF) was used as an instrument to record much of the information that 106 characterized each meeting (see Appendix E). It indi- cated the topics of concern to those present at a par- ticular meeting. It also provided information on how many women attended the meeting and their living situa- tions. An indication of the "mood" of the meeting may have been gained by examining the content of positive and negative statements made that evening. This may have revealed how a group of women felt about the sup- port group, themselves, and each other. The Group Content Form asked for descriptions of the statements that were recorded as frequencies on the Group Participation Form. The Group Content Form included the following categories: Positive Toward Group, Negative Toward Group, Positive Toward Woman, Negative Toward Woman, Positive Toward Self, Negative Toward Self, and Topics (Other on the GPF). The forms were designed to serve complementary functions in the sense that the Group Participation Form recorded the number of classified statements, while the Group Con- tent Form recorded the content of those statements. The content of statements recorded as "Other" on the Group Participation Form were included in the list of topics on the Group Content Form. The Group Content Form did not contain actual quotes. The statements were recorded after the meeting and were paraphrases of what was said. Accurate quotes were not felt to be 107 necessary because examples were sufficient. Another reason for avoiding direct quotes was the promise to the group members that their identity would remain anonymous. As stated previously, the Group Content Form was designed to be used in conjunction with the Group Par— ticipation Form. They were expected to serve comple- mentary functions in gathering group process informa- tion and in summarizing each meeting. When these "snapshot" images were examined together, patterns were expected to emerge. This information was expected to be useful in addressing the research questions, espe- cially concerning evaluation of the supportive func- tions of the Ex-residents' Support Group. RESULTS The research design of the present study sought to examine two areas. First, the research instruments gathered data to provide a description of the social support networks of women who have been in abusive relationships. Second, the effectiveness of the sup- port group in increasing access to social support was examined. Examination of Social Support Networks of Support Group Participants Avepege Netwopk Charaetepistics of Support Ggoup Earticipapps In examining the social support networks of the support group participants, average network character- istics will be discussed, as well as significant pat- terns that emerged on the network variables. The aver- age network characteristics will be presented for nine of the eleven network components outlined in the Intro- 5 This information was provided by responses duction. from the Initial Interviews and is summarized in Table 1, along with demographic characteristics. 108 1(39 ~00 c: O u ~ >n zoodza .ouoo ocqudouc: mOuooaocfi .II. :moa “meoz aa.ou ox: o o m\c o c an on cm a o a mu nausea cud: uooucoo >~zucoz m~.v h.~ n.~ h.o I: o.~ 6.0 m.~ m.o o.m o.h m.m o.a~ ~.~ mason uo meow» Oz ass 02 oz oz 02 02 oz no» no» mo> oz oz 02 oz muvmsno cue: aceoemoz mm» «mm no» no» mm» no» am» no» no» oz mo» oz mm» no» mo» magnum u30u10uo vm.v w A a N v m n ma mu m a v v cocoocouu< oz «on oz am» no» no» 02 oz Oz oz oz 02 mo» oz oz msuoum acoe>o~dem o~.~ «\c N a n N v w N o\: ~ ~ m m amusemouo no oceans mm.m ox: m.~ o.- m.~ m.o~ o.~ m.v o. o\c n. m.~ m.m o. .muou>. denmocofiuu uo cumcmu «we con om mm he co~ om ecu cod o\c cod mm o om .xbuoo: a. xueefixOLQ venousoooo 02 «No no» 02 oz oz no» no» oz no» 02 oz oz mo» oz mmLOuuoodsm oouzu ecu ca uo~omcsoo oz «cm 02 oz no» no» no» 02 oz mo> oz oz u: mo» mor «ammuuoadom omega QOu :« c0503 oomso< arm c on cm can so am cod cod ox: oo~ om oo~ om xuozuo: cw moccaum a cam oc~ cm do on ca no em cod ox: OOH com ms om .momcoamou o>eu~m0d a. uneddsm uo >oosvoo4 ans c on m sea vs so cod cod o\: co“ no oo~ om mocowuu saw: uooucou a now am cu av c on mm m~ we o\c no av om do uaaou ocozm a «Na on em no oc~ cm mm ms cm ox: mm ~m c an mucoucoo ooomuOunoomm a nm.vv mm «a so a No NA v nm O NH man 0 an mucoucoo hazucoz vm.m n b cu m n N ~ n o v o N m mean xuozuoz m and. Na. -u cod. a. a. h. o— m. v. n. N. a. meanoeuo> xuozuoz mucoQLOAuuom modumLHOuuouonU Ownooucoamo poo modnoeuo> xuozuoz co mucomfioeuuoa uo mmmcoamvm I oqnms 110 W Network size was computed by summing the responses to the questions which asked how many people the respondent considered to be close friends and support- ive relatives (items 40 and 55). The average network size was equal to 3.84 persons, with close friends com- prising 2.23 network members (or 57 percent) and rela- tives contributing 1.62 members (43 percent). W There was an average of 45 monthly contacts for women in this sample (with a range from zero to 193 contacts). Seventy-three percent of these contacts were with the respondents' friends, while 27% were with relatives. NQLQIE pf Qoptaep Of the 45 average monthly contacts, 53% (n = 24) were in-person contacts, 44% (n = 20) took place over the telephone, and participants communicated by letter 2% of the time (n = 1).6 To describe these contacts further, on the aver- age, 17 (38%) in-person contacts were made with friends, 12 (27%) telephone contacts, and less than one letter (.02%) was written to friends each month. Con- tacts with relatives were conducted in person an aver- age of seven times (16%), on the telephone eight times 111 (18%) and less than once (.02%) through written corre- spondence over the course of one month. The nature of content variable also includes the content of the interactions. Content will be described in terms of the supportive resources that appeared to have been received by the respondent from the interac- tion. The most frequent form of social support that was provided was sociability (e.g. shopping; having dinner together), which occurred in 35% of the descrip- tions of contacts that could be coded. This was fol- lowed by contacts involving commonality (e.g. partici- pating in church related activities; discussing common friends or common problems), which accounted for one- fourth of the contacts. Contacts containing intimacy (e.g. talking’ about 'the respondent's feelings) were found to occur approximately one-fifth of the time, while emotional support (e.g. encouragement from net- work members) was provided in 12% of the contacts described. Influence/role modeling (e.g. receiving advice on ways to improve one's life), and assistance (e.g. receiving rides to the store; receiving care when ill) contributed the remaining 6% and 4% respectively. e Su ort This network variable was not directly assessed by the research instruments. A general indication was 112 inferred from the number of positive, negative and neu- tral responses that were given to questions that asked how the woman felt about her contacts with her network members (items 50, 53, 59, 62, 68, and 73-75 of the Initial Interview). Of the 96% of responses that the valence could be determined, 66% were positive; 18% were neutral; and 12% were negative. WM As stated in Network Size, the mean number of close friends to be a part of the social support net- works of the participants was 2.23, while there were 1.62 supportive relatives in the average participant's network. Nearly half of the participants felt that none of their relatives were supportive. Of those women.*who listed. supportive relatives, mothers ‘were included by four respondents; fathers were included by three respondents'; a sister, brother and adult daugh- ter were each included by two respondents; and a sister-in-law and son-in-law were each included by one woman. In examining responses to the list of the top three sources of support for the respondents, 42% of the supporters were friends; 24% were relatives and; 33% were community supporters. More specifically, nine out of the thirteen participants (69%) stated that one 113 or more of their close friends had been in a relation- ship with an abusive man, and seven (54%) listed a female friend who had been in an abusive relationship as one of her top three sources of support. Additionally, 38% of the participants included a counselor or therapist among their top three sources of support, while 23% included a minister, priest or rabbi in this list. 123111111: Of the ten women who listed more than one close friend, half reported. that their friends knew each other. Two-thirds of the network members of the women interviewed in the present study lived "nearby" (in the same. or neighboring' city or 'town), while one-third lived "far away" (over a one hour drive separation). Friends tended to live nearby more often than relatives (79% to 47% respectively). §tebility The average length of relationships between par- ticipants and their close friends was four years. Fourteen percent of the friendships had lasted for ten or more years; ten percent had existed between five and ten years; 38% were between a length of one and five years; and another 38% of the friendships were less 114 than one year old. Interestingly, all of the recent friendships (less than one year) were with people that the respondent had met through the shelter (including residents, volunteers or members of the previous sup- port group). Another component of stability, the change of net- works over time, will be discussed below (see The Impact of the Support Group Intervention on the Partic- ipants' Social Support Networks). glossnsaa Five of the women (38%) perceived themselves as having at least one "very close" friend (i.e., they received a rating“ of "l," which corresponded to the following description: "There is nothing that I cannot talk to her/him about or that she/he would not do for me if she/he could."). Another four participants (31%) listed at least one "close" friend (i.e., they were given a rating of "2," defined as: "I can talk to her/him about most of my feelings and I can rely on her/his help most of the time."). This indicates that almost 70% had at least one person in their lives that they perceived as available to provide supportive resources most of the time. The average rating of closeness for all friends was 2.18 (on a four point scale, with 1 indicating the closest rating). 115 s ' at mer ed or Netwo Variables WW Statistical analyses were performed on the network variables and demographic items that were relevant to the research questions of the present study. Phi coef— ficients were calculated between dichotomous variables (see Table 2); point biserial correlations were per- formed to examine relationships between dichotomous and interval variables (see Table 3): Spearman's rank order correlations were performed for ordinal by interval relationships (see Table 4); and Pearson's product moment correlation coefficients were calculated between interval variables (see Table 5). A brief discussion follows on the variables that were found to be signifif cantly related. Employment status was found to have a significant relationship with several other variables. First, women who were employed outside of their home tended to have larger social support networks (r = -.70, p < .001), and they were also likely to have a greater num- ber of contacts with their network members (r = -.58, p < .05). Other significant findings were that women who were residing with their abusive partner generally had higher attendance rates at support group meetings (r = .66, p. < .05). Not surprisingly, women who lived with 116 00.H 0H. 00.0 mm. mm. wmmmww poo Imouo 00.H ha. mm. 00. Mddwmmmmu 00.H ma. 0H. Hmwmmmmmw cmaoz cmmsnd 00.H hm. ucm o Em A2 .1. 5 msumum udoumoua 02 u 5 mumuuommzm mmhnu no» :a uonmczou as u 5 mumuuommzm owns» 00» Ca cmaos pomsnd lg n 5 moumum HomeonQEm 84 AS u 5 mummsnm nufl3 mcfioflmmm Mwmmmd nu“; mcfimmm mofiHMMnmuomumno OH: on o m can moanmflum> uo3umz you mucwfio.uumou N manna 117 ma. ow. ma. I oo. u «*mo. ma. . Ana u :0 msuwum usonmouo ma. me. n ma. oe. om. u em. I And u :0 mumuuoamom mmnsu mow :fl Hoaomcsoo ma. mo. u mm. n ma. u «o. u mm. “NH u :0 mumuuommsm owns» 00» :H cmaos cmmsnd 0m. mm. «mm. I «£05. Hm. ms. and u :0 msumum ucmfihonEm *««mm. vo. mm. u we. a «mo. NH. 1 and u :0 mummsnm nuflz 0cfiufimmm mmmmmdvu Mwmmd mwmmwmmw mmww MMQMUQMMMQ mmmdmwmdu uomucoo no whom» >Hnucoz xuo3umz .010 mu Q 0...“qu OH on oO._-- mud. m any m> ,03 m o- mCOHH..mH.HOU HM “-m.. Hu.0.. m GHQMB .mo. v we oo.o be. on. u «mm. n ma. AHA n :0 mmmcomoHo .8 nu mumdm4\m mwmma mwmmwdmw muflm mmmmumMMMfl uomucoo mo mummy >Hnuco: xuosumz w wanna 119 Ame u :0 oo.H Momsnm nufl3 uomucoo wacucoz mo. : oo.H ANH u :0 mmsnm mo whom» He. . mm. u oo.H and n :0 muomucoo >Hnucoz he. s me. u hm. _ oo.H Ana u :0 muflm xuo3umz om. we. an. n he. 1 oo.H Ana u :0 mocmccmuu< MMWflMdVH wmmMQ wwwmmmmw wwww mmqmumMMMd uomucoo no mummy hacucoz xuozumz mmwwmwmww low mno ...mu.o.mo . n no no H.> H03»- no whoa».-- you c080: #0200 . m..om mm. m QHQMB 120 their abusive partner also reported being in contact with them more frequently (r = .95, p < .005). Another relationship that was observed for this sample was that support group participants with larger networks were more likely to describe closer relation- ships with their friends (rho = -.58, p < .05). Evaluation of the Support Group Intervention Spppprtive Besoprces Provided by phe Support Group Evidence concerning the supportive resources that appear to have been provided by the support group is presented below. This information was extracted from a number of sources including Group Content Forms, Ini- tial Interviews, Follow-up Interviews, Drop-out Inter- views, and the researcher's field notes. Further information may have been provided by the Group Partic- ipation form, however, it was not included in the anal- yses due to insufficient interrater reliability. Assistance Assistance was defined above as the provision of help *when needed, including' direct. help or’ helpful information. Assistance was one of the most frequently provided forms of social support, and was given mainly in the form of information. Over the course of the intervention, a film on woman abuse was shown and speakers on assertiveness and the response of the police to domestic violence calls visited the group. 121 Additionally, shelter volunteers who lead legal rights workshops stopped in to answer questions on five occa- sions. Along with these more formal sources of informa- tion, the support group members shared a wide range of factual and experiential knowledge among themselves. The Group Content Form provided the list of topics that were discussed throughout the twenty weeks. This list gives an indication of the variety of issues that were of concern to the participants and is presented in Table 6. The most common form of tangible assistance pro- vided by the support group was transportation. This usually involved the provision of a ride to or from support group meetings by the facilitators, but on five occasions support group members drove others to or from the meetings. In a few instances the facilitators pro- vided transportation to a member outside of the weekly meetings (e.g. bringing a woman and her sick child to a clinic, bringing a woman to enroll in an adult educa- tion program). The facilitators also called support group members a few times to furnish information on legal and employment matters. And on at least one occasion, one group member provided assistance in the form of child care for another woman. .122 Table 6 Topics Discussed at Support Group Meetings and Number of Meetings at which Topic was Discussed Topic Frequency Money and financial issues 20 Abuse general 9 role of alcohol 6 sexual abuse (rape) S psychological abuse 8 difficulty in getting out of the abusive relationship 8 societal violence 3 attempts to end the relationship 19 negative attitudes and behaviors of professionals regarding abuse 5 Abusive partners general common characteristics how to get help for them loss of love due to abuse I—‘NChU'I Children general 6 effects of abuse on them 19 custody issues 6 relationship with fathers 3 child abuse 3 Friends/Relatives general 5 issues of support 10 Society/Community legal issues 17 education divorce medical issues employment issues marriage religious issues men (other than abusive partners) H ubU'lU'IChNIWH Feelings personal growth 1 emotional support self-blame loneliness trust guilt feeling used vulnerability forgiveness suicide self-worth HruHrdHrawanLuo Support Group general 18 absent members Other 35 123 These are all examples of attempts that were made to prpyide assistance to support group members. There were also a few instances when women acknowledged that they felt that. they had. received assistance. For instance, during her Drop-out Interview, one woman stated that she had "learned a lot." Two women also reported in their Follow-up Interviews that the support group had taught them to stand up for themselves more often with their abusive partners. One of these women added that she had learned ways to get out of the abu- sive situation. Wm Emotional support is received when one's words or actions lead another to feel encouraged, understood, accepted and/or cared for. While it is possible that the forms of assistance described above may have sug- gested to the receiver that she was cared for, there were also occasions when more direct indications of emotional support were expressed. Signs of emotional support were frequently extended during support group meetings, as these paraphrased examples indicate: I understand what you mean -- it must have been very hard to press charges; I'm glad that things are looking up for you; Well, I’ll pray for you anyway; You look nice tonight; You've really gotten stronger. 124 On at least one occasion, a woman was given emo- tional support when she began to cry and another woman comforted her. Other non-verbal indicators of under- standing and encouragement came in the form of frequent nods, "uh-huh's" and other affirmations. The facilitators also provided a fair amount of emotional support outside of the meetings. For exam- ple, often during rides women discussed their situation and concerns, which would be met with additional attempts at providing emotional support. There were also at least seven telephone conversations where the facilitators provided emotional support to group mem- bers. And emotional support was provided to one group member on three occasions when she requested a meeting to discuss pressing concerns. One further attempt by a facilitator to indicate to a group member that she was cared for occurred when a letter was sent to a woman who had left her husband and moved out of state a few days after exploring her feelings and options at a group meeting. Again, evidence that some group members recognized that emotional support was being offered to them was provided through responses on interviews and para- phrased statements from the Group Content Forms. TWo women reported feeling better about themselves as a result of being part of the support group. On two 125 occasions a woman stated that she appreciated it and felt good when the facilitators called her between meetings. Examples of other such statements include: This is the only place I can get support; This has made me much stronger: You have really helped me a lot; I can lose it here because I know you all care: This group has really opened me up; and, I get support here that I don't get from my family. W While most (10 of 13) participants described expe- riencing commonality, or the feeling that others in the group had had similar experiences or values and under- stood their situations (see items 38 and 39 of Initial Interview, in the Appendix A), there were contradictory messages from a number of sources. Of those who claimed that their situations were different, one felt that some others were in "severely worse" situations, and another saw the fact that she was not married to her abusive partner as an indication that her experi- ences were "quite dissimilar." Other differences that were mentioned were that some women had been abused for much longer periods of time or had experienced mental, but not physical abuse. Furthermore, the mix of women both in and out of the relationship seemed to cause two women to question whether or not others understood their situations. One woman explained that she guessed 126 that the others understood, but that she was now in a period of adjusting to her new situation. She felt bad that she was out of the abusive situation, while others were still in it, and she felt that she could do more. On the other hand, the other woman felt that she was doing "alright," but that some women could not under- stand why she had gone back to her partner. A few statements from the Group Content Forms that seem to reflect attitudes concerning commonality are presented below. Again there were both positive and negative indications: I can relate to you: Sometimes it helps to hear others' experiences: I feel comfort- able in this group; We attract these kinds of guys because we think we deserve it: We have no self-esteem! An interesting finding involving commonality also emerged that one may not have predicted; one that the researcher has observed before. Although social com- parison theory suggests that knowing that others have gone through a similar situation can be comforting (Caplan, cited in Gottlieb, 1981), some women reported a sense of relief in discovering that others had expe- rienced a much worse situation than their own. Two women in this sample reported such feelings. Innings): It is difficult to determine if the support group provided the participants with the opportunity and 127 sense of safety to share their personal feelings. Statements from the Group Content Form can only reveal content: the degree of emotion cannot be adequately assessed from the words alone. However, a few state- ments may be safely put forth as examples of sharing of true feelings: I feel very scared to go back to school after eight years; I can lose it here because I know all of you care; I don't know who I am or where I'm going; I'm going back to visit him -- I know I'm crazy; I did it! I never though I could -- but I did it!: It scares me -- I'm losing me. Another indirect indication of the level of inti- macy provided by the support group is the range of top- ics covered (see Table 6). Although any topic can be discussed on a superficial level, some subjects are more likely to arouse emotion than others (e.g. the description of an abusive incident). InflpgpceZRole Modeling Through the sharing of experiential knowledge, participants in a support group can serve as examples or role models for each other. The participants occa- sionally expressed that they were influenced by other group members: You have really helped me a lot: You are really strong; It's great that you put your own lock on the door: She's got more patience than I have; It's great that you stood up to your lawyer. 128 Whether or not anyone was influenced or attempted to model her behavior after than of another support group member, many women provided descriptions of their efforts that others had the opportunity to learn from. A few of the many examples include: I'm proud of myself now that I've made a decision: I'm not getting divorced because I don't want to save my marriage -- I'm getting divorced because I want to save my life: I'm proud of myself for confronting my family: I trust myself: I celebrated and was good to myself: He wanted me to be scared, but I wasn't: I can do anything a man can do. 5 . J'J't In addition to to other motivations for attending support group meetings, it appears that a number of women were seeking a chance to socialize with others. In response to interview items concerning what they would like from the group (Initial Interview items 25- 28), 62% of the women's responses can be classified as indicating a desire for sociability (e.g. wanting to feel less alone; to get out of the house: to talk to others: and to form friendships). There were also at least three direct requests for social activities out- side of the usual weekly meetings. It was indicated from a Follow-up Interview that two group members did 129 engage in social activities outside of support group meetings. m a h Su or rou nte ent'on on pug Eappigipapps' Sogia; Suppopp Netwopks In examining the impact of the support group on the participants' social support networks, information from the Follow-up Interviews was to be analyzed. How- ever, only four Follow-up 'Interviews were adminis- tered.7 Information on the changes in the networks of these participants will be presented. Table 7 shows the changes that women reported from the Initial Inter- view to the Follow-up Interviews. Trends seem to indi- cate a decrease in number of monthly contacts (except for participant #5). An increase in the average length of abuse occurred because one of the women experienced a further abusive attack one and a half years following the previous attack. Another noteworthy finding is that participant #6 in her second Follow-up Interview no longer included a female friend who had been in an abusive relationship as one of her top three sources of support. It does appear that support groups can provide members with the opportunity to meet people who may become part of their social support networks, since three of the women listed people in the support group as close friends. Two additional women indicated that 1L30 .maasmpcwwum mmOHo 030 ponwuommp 0cm 00000 m.umzmw>00u:« umuu< n .000 so :0wumEHOucd 0000>000 auco usn upcmfiuw cacao usOu couponmmm 0m 0m 0m 0m 0 0m 0m 0m 0 somsbs\3 00000000 >qucoz 0.0 50.0 0.0 0.m 0.n na.0 0.0 0.m 0.5 .mumo>. 00:00 00 200:0: 0H 0H cm 0:0 :0 cm 000 cowumsuqm 00s>qq 02 oz 02 oz 02 oz 02 m00>0~050 N ~0.~ m ~ ~ m N m\c N mmocomOHO mo mcwumu 00000>4 ms. Hm.m m. 0.0 0.~ m0. 0. m\c h. .mummx we . 00000>00 asnmccmguu 0o zumcws 00H a00~ 00H 00H 00A 000A 00H m\c 00H .>00000 0. xuaeflxoum szmmuwomo 00> 00» 00> 02 mm» 02 oz admuozu uoammcsou 02 00> 02 oz 00> oz 02 0000030 no 0000200 m 000 :0 00603 000004 00H wood 00A 00A 00d 000A 00H m\: 00a xuo3uoc cw 0000000 m 00H arm Hm 00H 00H «com 00H m\0 00H ..0mou Amuusmc a + 0. 0000030 no >003000¢ 00H 000~ 00a 00a «cog mood 00H 0\: 00H mpcowuu\3 uomucoo a HH «no em 0 >0 «mm 00 m\: 50 muamo 00020 0 00a ”hm 0v N0 mm avv cm ax: mm muomucoo monuIOulmomN a a so.ms oN ~H a nn.os an o ~H muomucoo ssnucos no . N w m 00 a m.~ m 0 0v 0000 xuozuoz I1. 0. m 0‘ m0 w» (m 00 ma v: mofiumwuouomumno 3003002 ummmMMMMMMMq muCMQMOMuumm mucmmwowuumm m. asuzoauom 30u>uoucu asuzo-om 300>uoucu HawuwcH mzmq>umucu azuzoaaom ob 300>umucH Hmwuch 0;» 5000 mogumwuouomumcu 00:00000500 0:0 moanmfiue> 2003002 :0 000000000000 00 momcoomom cw mwmcmcu h mans? 131 a support group member was one of their close friends, however, these women did not meet the friend through the support group. A further clarification should be noted of the composition of networks which included friends met through the support group. Out of the five women who reported having close friends in the support group, two of these women included one or both of the facilitators. In fact, contact with facilitators com- prised 21% of all contacts for these five women: and for one member, all contacts listed were with one of the facilitators. Table 8 provides information on the support received by women from network members who were also support group members. One can see that 68% of their contacts involved the other support group mem- bers: that the women generally felt close to these net- work members: that the women generally felt close to these network members (i = 2, i.e., "I can talk to her/him about most of my feelings and I can rely of her/his help most of the time"): and they generally felt positive about these interactions (87% positive responses to items 50, 53, 59 and 62 of the Initial Interview). Feedback on the Support Group from Drop-out Surveys By the end of the data collection period, 85% (n = 11) of the participants in this study had dropped out. 132 .mucp 00uuom0ucs m0umowpcfl Anny 5000 "0002 sum I- om cos ooa ooa .mmmcommmu m>fluflmod *0 0000000 no >000000< m m H m m m mm0:0moao no mcwumu 0mmu0>< 000 mm mu 00H 00H mm 0000503 msoum nuwz uomucoo w dds Mw ma «0 4a 0002 mucmafiofiuumm mxuozu0 who am Hmwoom 0803 >0 0M>w0owmlwmmmmmw 0 warns 133 Forty-five percent (n = 5) of these participated in the Drop-out Survey. The remaining women (n = 6) could not be reached to be interviewed. All of the women who were contacted were living apart from their abusive partner. The two women who dropped out of the group while living with their abusive partner could not be contacted. None of the women who participated in Drop- out Surveys returned to the support group during the data collection period. The following reasons were given for no longer attending support group meetings (statements are para- phrased): did not like the other group members: out- grew the use for the group: had a schedule conflict: very busy; was out of the relationship and overcoming the problem and didn't know what to say anymore: family problems' and: meeting time not convenient with work schedule. In addition to supplying the reasons for not attending support group meetings, the women who partic- ipated in the Drop-out Surveys provided other valuable information. This included both positive and negative feedback, and suggestions for group improvement. All of the positive comments were given by one respondent. She reported that she enjoyed the group: that she learned a lot from it: that she laughed: that she liked the speakers: and that she liked the fact 134 that the Residents' and Ex-residents' Support Groups were separate. Indications of dissatisfactions with the support group that were not explicitly given as reasons for non-attendance included: needs were not being met by the group: felt used by another support group member: disagreed with perceived attitude of the group that all men are bad and marriage is wrong; needed more help, i.e., wanted one-to-one counseling: felt uncomfortable about being called by a group member who was not an ex- resident: was considering reconciliation with her abu- sive partner and wondered about fitting in with the group: felt that a suggestion that was offered was not taken seriously; and, felt that the group should only be for ex-residents, i.e., did not feel comfortable with non-residents attending. A number of direct and indirect recommendations for improving the Ex-residents' Support Group were offered. These included the opinions that more speak- ers would improve the group: that more women would improve the group; that "something" should be added to make the group more interesting; that it would be bet- ter to do something rather than just sitting and talk- ing; that more outside contact between group members 135 would be welcome: and that the group should be exclu- sively for ex-residents who are no longer in an abusive relationship. Summary As stated previously, the present study set out to examine the social support networks of participants in a support group for women who had been in abusive rela- tionships and the impact of the support group on the social support available to these women. Of the eigh- teen women who attended the support group during the twenty week data collection period, thirteen completed the Initial Interview. Three of these thirteen women participated in the Follow-up Interview. Out of the eleven women who left the support group, five completed the Drop-out Survey. The major findings are presented below. A summary of average network characteristics indi- cates that there were 3.84 people in the average net- work of the participants. Their mean number of con- tacts per month was 45, which most frequently involved visits or telephone calls with friends that were gen- erally social in nature. The average length of friend- ships was four years, with 76% of the friendships being less than five years old. Seventy percent of the women described at least one close friend with whom they 136 shared 'uheir feelings and relied. on for assistance "most of the time." Employment outside of the home was related to larger networks and more monthly contacts with network members. Significant correlations also indicated that women who resided with their abusive partner more fre- quently attended support group meetings. It was fur- ther found that larger networks were associated with closer friendships. Descriptive data indicated that the supportive resources of assistance, emotional support, commonal- ity, intimacy, and influence/role modeling were offered to and received by support group members. Assistance, (in the form of information), emotional support, and commonality were the most frequently provided forms of social support. Due to the small number of support group members who participated in the Follow-up Interviews, little information was available regarding the impact of the support group on the participants' social support net- works. However, it was reported that five of the sup- port group members included other group members in their social support networks. Information provided on network members who were also support group members showed that over two-thirds of their monthly contacts were with other support group members. 137 A final component of the research design was to gather follow-up data from those who dropped-out of the support group. A high percentage, 85% of those inter- viewed, dropped-out of the group. Forty-five percent were reached to supply feedback on their experience of the support group. All of the women who participated in the Drop-out Survey were living apart from their abusive partner. The respondents were asked for the reasons that they stopped attending support group meet- ings. The most frequent responses indicated that the group did not meet the woman's needs. External factors were also reported, such as schedule conflicts and fam- ily problems. There was one report of disliking other group members. Other comments included positive and negative feedback and suggestions for improvement. Generally, those who dropped-out would have preferred the meetings to contain more speakers and activities with less general discussion. DISCUSSION At the time the present study was designed, little empirical research had been published on women who had been in abusive relationships, although there was a rapidly growing collection of descriptive information on this subject. This descriptive information indi- cated that men abused women with alarming frequency, and that abuse of women crosses economic, geographic and racial lines. It was also reported that isolation and psychological abuse commonly occurred in conjunc- tion with physical abuse, and that community agencies often reinforced the victimization of women who sought their assistance. Eventually, an awareness rose of the great. needs of’ women who ‘were being' battered, and assistance was offered, mainly in the f0rm of shelter facilities and counseling services. Research on the effectiveness of these interventions is scarce and even fewer follow-up studies. have ‘been conducted on the experience of women who have received these services. The research of others indicates that having access to social support can assist those who are going through major life transitions. The present study was 138 139 designed to explore whether a support group interven- tion could provide access to social support to women who were experiencing the major life transition of end- ing an abusive relationship. More specifically, the expectations of the present study were twofold: to gather preliminary information on the nature and sources of social support that are available to women who have left abusive relationships; and to determine if a particular support group intervention was effec- tive in increasing the level of support that was avail- able to those who attended. Discussion of Findings Before the findings of the present study on the network variables are discussed, several factors must be considered. First, because this information is based on only 13 cases, it does not accurately repre- sent the majority of women who have recently left abu- sive relationships. Another major consideration is that all of the respondents were women who had attended a support group. It is not known in what ways those who have not attended support groups differ from those who have chosen to attend a support group. The remain- ing factor affecting the generalizability of the data is that although the focus of the study was on women 140 who had left an abusive relationship, close to one- third of the respondents were residing with their abu- sive partner at the time that they were interviewed. 0 V a Averaged responses on network variables are pre- sented in Table 1. As reported in the Introduction, meaningful comparisons of the findings of other researchers on various network variables are severely limited due to the lack of uniformity of operational definitions. Often the data on individual network variables have been condensed into scales or factors, and/or have been discussed only in terms of what they were found to correlate with. Therefore, even if the same instrument were employed, direct comparison would still not be possible. Despite this limitation, the findings of the network variables examined in the pre- sent study will be compared with the findings of other researchers, to the degree that such information was available. Network size. For the women who participated in the present study, the average network size was equal to 3.84 persons: 2.23 of whom were friends, and 1.62 of whom were relatives. Due to the differences in opera- tional definitions, direct comparisons could not be made, however, the figures of others seem to indicate that the participants of the present study had larger 141 networks than one group of women who had recently entered a shelter for battered women (Mitchell & Hodson, 1983), while they had smaller average networks than a sample of first year graduate students (Norbeck et al., 1983): a group of female psychiatric out- patients; and a group of female family practice patients (Silberfeld, 1978). Specifically, Mitchell and Hodson reported that 21% of new shelter residents indicated that in the previous month there had been no one with whom they could socialize or discuss personal problems. Additionally, another 40% could only list one such person. In the Norbeck et a1. sample of female first year graduate students, however, the aver- age network size was 12.39. The definition used by Norbeck et al. was, "List each significant person in your life. . .. . Consider all the persons who provide personal support for you or who are important to you now" (p. 265). The study by Silberfeld (1978) reported network sizes for two groups of women: psychiatric out-patients and family practice patients. The partic- ipants reported interactions with all of the friends and relatives they had seen in the previous week. The psychiatric group had contact with an average of seven persons, while the family practice group reported in- person contacts with an average of 11.2 friends and 142 relatives. Obviously, the variation in these defini- tions of network size must be considered when making comparisons across studies. However, when examining the scores at face value, there does appear to be sup- port for the claim that women who were in abusive rela- tionships tended to be more isolated. One explanation for the larger average networks of the support group patients as compared to the shelter residents in the Mitchell and Hodson study may be that isolation can be reduced once the women leaves the abusive relationship. Ampppt_pt_ppptapt. Again, with the amount of con- tact variable, there is no general standard against which to compare the findings of the present study. Indirectly, the average of 24 in-person monthly con- tacts for the present study can be contrasted with the shelter residents from the Mitchell and Hodson (1983) study, where over half reported one or less social con- tact unaccompanied by their abusive partner in the month previous to their separation. While in the Silberfeld (1978) study of female family practice an psychiatric out-patients, the average number of weekly in-person contacts with close friends and relatives were 18 and 16 contacts respectively. No information was reported for contacts via the telephone or written correspondence in the studies that were reviewed. 143 Natuzg pf gpptact. Actual figures on the types of contacts or resources provided through contact with network members could not be located, other than the in-person contacts discussed above. It was suggested in the Introduction that the types of contacts included may be an important consideration. For example, valu- able information may be lost by including only in-per- son contacts. Indeed, it was found that, although written contact did not appear to be a popular form of contact (only 2% of all contacts), 44% of the average monthly contacts took place through telephone calls. It was also believed that the types of resources provided could reveal important information about one's social support network. In the present study, socia- bility, commonality’ and. intimacy ‘were 'the :resources provided most frequently to the participants during their interactions with their social support network members. There are no reports from other researchers to compare these findings with because the content of interactions for other samples has not been reported. Although one may have predicted that someone going through a major life transition, such as learning to live apart from an abusive partner, may be most in need of assistance and emotional support, these were not the major resources that appear to have been provided. One possible explanation for this may be that the women's 144 interactions did not provide the resources that they needed most. However, there are other indications that suggest that inadequacy of support is not the reason for these findings. The first indication is the report that almost 70% of the participants indicated through closeness ratings that they had at least one person whom they could rely on for help most of the time; therefore, it appears that their assistance needs were being met to some degree. Another indication of a strong need for sociability was provided by feedback from support group meetings and Drop-out Surveys: specifically, many of the respondents were interested in social activities outside of group meetings. Netpptk composition. It is possible that the net- work composition variable may reveal the greatest amount of information about a person's network. For example, it can reveal the proportion of friends to relatives, and formal to informal sources of support. As a group, the participants in the present study were found to count twice as many friends as relatives as their strongest sources of support. Additionally, they reported a two-to-one preference for informal sources of support (friends and relatives) over formal (community) sources. This is in keeping with the claim of Young et al. (1982) that most people prefer to receive assistance from informal sources of support. 145 The reports from other research on the distribu- tion of friends to relatives in networks are varied. Norbeck et al. (1983) reported a greater proportion of friends to relatives in their sample of first year graduate students, however, the discrepancy was not as great as that of the present study (i.e., 44% friends and 36% relatives in the Norbeck et al. study and 57% friends and 43% relatives for the present study). Silberfeld (1978) found that the female family practice patients had a higher ratio of close relationships with relatives, while female psychiatric out-patients had a higher ratio of close relationships with friends. The findings of Straus (1980) may be most applicable to the present study. Straus reported that a higher incidence of domestic violence was correlated with high levels of stress in combination with networks that were predomi- nantly comprised of relatives. An expectation regarding network composition that was presented above was that women who have been in abusive relationships may receive a greater amount of support and understanding from those who have had simi- lar experiences. And in fact, the finding that 54% of the respondents included a "female friend who had been in an abusive relationship" as one of their top sup- porters may provide some basis for this claim. 146 My. In the present study, five of the ten women who reported more than one friend stated that their friends knew each other. Additionally, two women reported that over 90% of their contacts in the previ- ous month were with relatives. Density ratings were not presented in the studies that were reviewed, which prohibited comparisons across studies. It has been suggested that lower density networks may be beneficial for those who are attempting to make changes in their lives, and that high density networks may exert collective pressure to thwart change (Hirsch, 1980). However, almost all of the participants in the present study reported that their network members felt they should not reconcile with their abusive partner. Therefore, it seems that a low density network may have allowed a woman to make this life change, while a high density network may have encouraged it. Qeogtappig ptoximity. The results of the present study indicated that two-thirds of the participants' network members lived nearby, and that more friends lived nearby than relatives. This could account for the higher proportion of contact with friends. Infor- mation from other studies on the effects of geographic proximity or social support could not be located. Stability. Two indicators of stability were included in the present study. The first related to 147 the stability of the relationships within the networks (i.e. , the number of years that the respondents had known each of their close friends). The second indica- tor of stability related to the changes in the composi- tion of the respondents' networks which was examined by collecting information on the networks at two points in time. The average length of the friendships between the respondents and their close friends was four years. This figure is somewhat lower than those of Silberfeld (1978) for his samples of female family practice and psychiatric out-patients, whose average length of friendships were 6.5 and 6.2 years respectively. This discrepancy fits with the prediction that such a life transition as these women are facing may require that relationships with people who are unsupportive of the women's decisions be replaced with more supportive relationships. The findings of Norbeck et al. (1981) are not directly comparable because their data are not limited to friendships, but include all relationships. Additionally, they presented their data as a score, rather than in numbers of years. The average score for their sample of graduate and undergraduate nursing stu- dents was 3.29 on a five point scale. A rating of "l" on this scale equaled a relationship of less than six months, while a rating of "5" indicated that the person 148 had been known for more than five years (Definitions of interim ratings were not provided by Norbeck et al.). As stated above, another indicator of stability is the degree to which networks change over time. The design of the present study included a follow-up compo- nent to address this feature of stability. Follow-up Interviews could only be arranged with three of the six eligible group members. However, an examination of Table 7 reveals that the networks did not remain static, but rather showed changes on some variables. It was predicted that there would be change in the net- works due to the fact that the participants were under- going a major life transition which may have required a change in sources of social support to facilitate adjustment. In a study of first year graduate students, who were also likely to be facing transitions, Norbeck et al. (1983) found a significant decrease in the number of relationships and in the length of relationships at their seven month follow-up. They also observed a change in the composition of friendships, a decrease in the frequency of contact with relatives, and an increase in contacts with neighbors. C e s. The present study concentrated on gathering information on "close" relationships of the participants. An initial definition of closeness was 149 presented and then a rating for each friendship was requested of the respondents. In gathering information on relatives, respondents were asked about their "supportive" relatives without a definition being pro- vided. Closeness ratings were not elicited for rela— tives. The ratings of the respondents indicated that nearly 70% felt that someone was available to provide support to them "most of the time." Although other researchers also included closeness as a network vari- able, only one study was located which included find- ings. These findings were from Silberfeld (1978), who reported that female family practice patients had a higher proportion of close relationships with rela- tives, while female psychiatric out-patients had a higher proportion of close relationships with friends. Additionally, the family practice group tended to spend more time with their close network members than the psychiatric group did. In the present study it was found that the participants had a greater percentage of contact with friends, who constituted 57% of their net- works. It was also discovered that there was a signifi- cant correlation between closeness and the size of a woman's network, such that, women who had larger net- works also tended to report closer relationships with 150 their friends. Potential explanations for this finding are that those with larger networks may have had healthier networks that were not only larger, but that also included supporters who provided more resources to them. A similar possibility is that the skills pos- sessed by women who had the ability to form closer relationships may have also assisted them in forming supportive relationships with a larger number of peo- ple. It also seems logical that the potential for forming a very close relationship with another person increases with the number of persons that one is involved with. There were no findings reported in the studies reviewed that were related to this finding. On the average, employed women had significantly larger networks and more monthly contacts than women who were not employed outside of their homes. One of the ways in which being employed may contribute to these positive differences is that the workplace may offer employees an opportunity to form relationships with people who may become network members. This may be especially important for abused women who may other- wise be isolated from sources of support. In this sam- ple, women who were employed met 73% of their friends through work. Besides forming relationships with oth- ers who may provide supportive resources, being employed may also provide women with financial 151 resources that could allow them to have more contact with friends and relatives (i.e, they may be better able to afford a telephone, and/or automobile, and have the funds to engage in more social activities). Another significant correlations revealed the ten- dency for women who were residing with an abusive man to have higher attendance at support groups meetings. They may have come to meetings to receive additional support or possibly with hopes of broadening their net- works. Based on the major topics discussed at the meetings, it is also likely that the women who were residing with their abusive partners came to more meet— ings out of a need to explore their choices regarding their relationship with their abusive partner. Alternate Interventions Overall, it appears that the support group inter- vention in the present study did not meet the expectations that were projected. The main indicators of this were the small number of participants, low attendance, high drop-out rate and, the minimal amount of support exchanged between group members. Another way that the support group differed from prior expectations was that rather than focusing on issues that women would face when attempting to estab- lish a life without their abusive partner, the majority 152 of the discussions seemed to be directed at the con- cerns of those who were still in a relationship with their abusive partner. Even though the majority of the participants in this study were not living with their abusive partners, those who were still in the relation- ship were the ones who attended most frequently. In addition, of the ten women who were not living with their abusive partners, five had expressed that they sometimes considered resuming the relationship. This figure closely matches that of Snyder and Scheer (cited in Mitchell & Hodson, 1983) that follow-up reports on women who had left a shelter indicated that over 50% had returned to their abusive relationships after two months. It seemed that the women in the group gener- ally felt ambivalent concerning their relationship and used the support group to assist them in sorting out their options. Ambivalence also characterized the feelings of the participants in the support group developed by Rounsaville et al. (1979). Future attempts to assist women who have left abu- sive relationships could follow a number of approaches. It is likely that a needs assessment would be a wise first step because so little is known about this group. If a support group intervention is indicated by a needs assessment, one or more of the following interventions 153 may be appropriate. Responses from the Drop-out Sur- vey, Follow-up Interview and statements made during group meetings suggest that organized social activities may be successful in introducing women to potential supporters. A regular series of presentations on a variety of topics may also create a successful inter- vention as indicated by interview responses. Addition- ally, a combination. of these. approaches may' be of interest to women who may require additional support to remain free of an abusive relationship. Low attendance at the support group meetings affected the intervention in a number of ways. It was intended that the group members would be actively involved with the support group, while the facilitators would act more as resources. Ideally, the facilitators had considered withdrawing from the group and encourag- ing the members to continue on their own, however, due to the sporadic attendance levels, it did not appear that the group reached a point where it could have sus- tained itself. However, due to the recommendations of the facilitators, subsequent pairs of facilitators included at least one women who had previously been in an abusive relationship. The comments of some of the participants that they would have liked a larger number of women at the meet- ings, in addition to the disadvantages of low support 154 group attendance described above, indicates the need to incorporate additional recruitment efforts into plans for future support groups. As it happened, during the data collection period of the present study the shel- ter, which was expected to be the major source of referrals, was experiencing unusually low occupancy rates. In addition to continuing the recruitment efforts described in the Methods, letters describing the intervention could also be sent to divorce lawyers, counseling agencies and other social service agencies encouraging them to notify potentially interested clients. Public service announcements on radio and television stations and in newspapers may also reach interested women. Others have also reported low atten- dance at support groups for women in abusive relation- ships. Rounsaville et al. (1979) stated that group membership was a "serious constant problem" (p. 69) . Out of 75 women who were initially identified to par- ticipate in their support group, only ten attended even one meeting. Other support-oriented interventions that may ben- efit this group of women could be modeled after groups that were organized to assist those facing other diffi- cult situations. One such possibility would be to pair a woman who had just left her abusive partner with 155 another woman who had previously ended an abusive rela- tionship and who felt that she had made a positive adjustment. This approach would be similar to the "sponsor" component of the Alcoholics Anonymous pro- gram, and includes aspects of the volunteer linking strategy and natural helper strategy presented in the Introduction. A cooperative organization where women could exchange services such as child care and trans- portation may also be beneficial. Another potential support providing intervention may be a community advo- cacy program focusing specifically on the needs of women who have been in abusive relationships. Some of its primary concerns may consist of legal advocacy and assistance in working with social service agencies. Another consideration for those designing interventions for women who are or have been in abusive relationships is that participation in the intervention, as well as transportation and child care would be available at no charge. Besides these efforts to assist women who have been in abusive relationships, approaches can be taken by previously abused women and others to prevent rela- tionships from ever becoming abusive. Educational pro- grams and legislative initiatives may have the strongest impact. Further research on the positive relationships between employment and network size and 156 frequency of contacts may also reveal information use- ful in strengthening the networks of abused women. Due to the diversity in the population of women who have been in abusive relationships, it is likely that their needs for support would be best met through a variety of interventions rather than by one universal approach. It also seems probable that different inter- ventions could assist women at various stages in the transition process. Furthermore, it is likely that there are a number of women who do not wish to dwell on their abusive experiences and prefer not to invest fur- ther energy on the subject. Suggestions for Revisions of the Measures Based on difficulties experienced in the present study and the approaches used by other researchers, suggestions for improving the measures are presented below. n w Fo ow-u erv'ew Ngtpppk_§igg. It is believed that some valuable information may have been lost by collecting data on only three close friends and three supportive rela- tives. Therefore, it is proposed that the items that measure network size be expanded to include all persons that the respondents feel to be supportive. Ampppt pf goptact. In light of Silberfeld's (1978) finding that significant differences were 157 observed only for the length of time in contact, but not for the actual number of contacts, it is believed that the length of contact should be incorporated into the items that measure amount of contact. Natng pf Coptagt. In addition to gathering information on telephone contacts and written corre- spondence as well as in-person contacts, the present research was concerned with the supportive resources that were provided to the respondents. However, it was found that a more systematic approach was needed. The procedure used by Mitchell (1982) appears that it may be adaptable to elicit this information. This proce- dure describes activities that are representative of the resources and asks the respondent if the network member provides the resources. For example, the item for sociability may read: "Is this someone that you can be with when you want to have fun and enjoy your- self?" (p. 392). Adggpapy of support. This variable was measured by asking respondents how they felt after their con- tacts with network members. It was found that the responses that were given were vague and difficult to code. Therefore, a more direct assessment is suggested for future studies. Adequacy has been assessed in other studies in numerous ways. The approaches tend to focus on whether the respondents received enough 158 resources of various types (Duncan-Jones, 1981); whether the respondents felt that their networks were large enough (Wandersman et al., 1980): and whether the respondents felt that each network ‘member provided enough support (Sarason et al., 1983). However, it is likely that the greatest amount of information would be available if the respondents were questioned regarding their satisfaction with each resource from each network member. For example, after asking if a network member provides assistance to the respondent, she would then be questioned regarding her satisfaction with the amount of assistance received. A rating of overall satisfaction with the amount of support and the amount of contact for each network member could also be included. w o 05' i . The revision to include all supporters rather than only close friends and support- ive relatives should provide richer data on network composition. Additionally, the item which asks how the respondents met their network members should be revised to ask the nature of their relationship. my. It is recommended that future attempts to assess this variable should include items such as "which of the people that you've described know each other?": "How' much. contact do 'they have ‘with each 159 other?" (for each relationship): "How close do you feel that their relationship is?" gppp1gxity_pt_pgiatipp§pip§. A clearer indication of the roles that a person holds in the respondent's network may be provided by the suggestion that the nature of the relationship between respondents and net- work members be requested. ngipppgity. Although the present study was con- cerned with the resources that were available to the support group members to assist them through the tran- sition period, there is evidence that also collecting information on the resources that they provided to oth- ers can give an indication of the strength of their networks. For example, some studies have found a rela- tionship between the degree of reciprocity in networks and the severity of emotional problems (e.g., Hirsch, 1981). Riessman's (1965) Helper-therapy Principle also suggests that benefits are associated with relation- ships where an individual has the opportunity to pro- vide resources to others. Reciprocity could be included in the social sup- port measure by questioning whether the respondent pro- vides resources to her network members for each of the supportive resources. Geogtaphic proximity. It was believed that this variable would provide an indicator of the amount of 160 support available from network members. The rationale was that if a network member lived quite a distance away, this would restrict the amount of support pro- vided. However, there are many other factors besides physical distance that separate people. Therefore, Kaplan's (cited in Lin et al., 1981) concept of "reachability" may produce a more, accurate indication of the availability of network members. The approach used in the Lin et a1. study was to ask respondents to indicate on a five point scale how easy it was for them to contact each network member. Stgpility. In addition to the item which asks how long the respondent has known each network member, and the follow-up items that monitor change over time, it may prove very informative to include items pertaining to recent losses of supporters. Changes in network composition were expected due to the transition period that the participants were believed to be going through. The addition of an item such as Norbeck et al.'s (1981): "During the past year, have you lost any important relationship due to moving, a job change, divorce or separation, death, or some other reason?" (p. 263), followed by questions that provide the rea- sons and other details, should reveal change in net- works and the nature of such changes. 161 gigsgpggs, The rating of closeness that was employed in the present study seems to tap similar information as the items that ask about the provision of supportive resources. If further testing indicated that the closeness rating provided essentially the same information, then this item could be deleted from the measure. Once the above changes have been integrated into the Initial Interview and Follow-up Interview, further testing should be conducted to refine the items. The final measure should also be administered to a random sample of people to provide a normative reference point. Reliability and validity testing procedures should be incorporated into the pilot testing phase. Inter- rater, test-retest and internal consistency measures of reliability need to be addressed. Assessment of inter- rater reliability will determine whether the interview- ers have been sufficiently trained, and whether the intent of the items are clearly understood. An esti- mate of test-retest reliability is essential to deter- mine whether changes at follow-up should be attributed to error of measurement, or to real change in network components. Test-retest reliability will need to be assessed in a short enough time span to avoid contami- nation of real change in networks. The time frame 162 employed by Norbeck et al. (1981) was one week between administrations. Correlations between items would pro- vide an indication of internal consistency. To achieve an indicator of concurrent validity, the social support measure from the present study will need to be administered in conjunction with other mea- sures of social support. However, the measure from the present study was constructed for a particular group and therefore, the size of the correlations may be diminished due to the situation specific items. The value of situation specific measures of social support has Zbeen discussed. by others (Cauce et. al., 1982: Hirsch, 1981; Holahan & Moos, 1982; Mitchell, 1982: Mitchell & Hodson, 1983), who maintain that a great amount of important information can be obtained from questions that are exclusively relevant to the popula- tion being studied. Efforts to evaluate the construct correlating the items with measures that reportedly measure related concepts. Others researchers have com- pared their social support measures with measures of self-esteem (e.g., Mitchell & Hodson, 1983), depression (e.g., Norbeck et al., 1981) and, mastery (Mitchell & Hodson, 1983). D -ou urve The Drop-out Survey is rather short, containing only nine items. However, there was no indication that 163 this length was problematic. It appeared that the small number of items, in conjunction with the open- ended format allowed the respondents to provide as much or as little information as they wished without feeling unduly pressured. The lack of anonymity of the inter- view may have been its major drawback. H0wever, this disadvantage was believed to outweigh the alternative of mailing a questionnaire because it was expected that the return rate would be too low. In the present study, drop-out information was collected from 45% of the participants approximately one month following the last meeting that they attended. A Michigan Department of Social Services report (Carty, 1983) revealed that follow-up rates for women who had received services at 31 domestic violence assistance/shelter programs in the state were equal to 30% at one month follow-up and 21% after four’ months. .Also, some of the items were intended to provide follow-up data, which required that the identity of the respondent be known. Although the overall length of the Drop-out Survey appeared to be beneficial, the use of alternative wording of items could be evaluated through pilot testing to select those that produce the most relevant information. Such items could assess the respondent's feelings about the other group members, the facilitators, the structure of the meetings, and the topics of discussion. 164 Reliability testing of the Drop-out Survey would follow essentially the same procedures as those sug- gested for the Initial and Follow-up Interviews. These would include inter-rater, test-retest and internal consistency measures of reliability. a ' G n e o The group process measures that were employed in the present study were designed to protect the confi- dentiality of the participants. Although revisions of these measures were found to be necessary, the position has been maintained that electronically recorded tran- scripts (on audio or video tape) may present too great a real or perceived risk to the safety of some par- ticipants. Additionally, the support group was made available to all women, whether or not they wished to participate in the data collection. Therefore, elec- tronic recording of meetings was ruled out because it would not have been possible to exclude the contribu- tions of individual persons. As stated previously, one of the group process measures, the Group Participation Form, was found to be very difficult to administer and did not provide reli- able data. It is likely that numerous approaches will need to be attempted before an unobtrusive and reliable group process measure is devised that accurately describes the interactions of support group members 165 during the meetings. The validity of a group process measure may be evaluated in the piloting phase by administering a brief questionnaire after each meeting for participants to indicate the level of support that they perceived to have been offered during the meeting. It is believed that it would be too cumbersome to include weekly post-meeting evaluations beyond the piloting phase due to the participants' eagerness to depart and the expectation that some of them would be receiving Initial or Follow-up Interviews. (However, if weekly evaluations proved useful they would continue to be incorporated in. the research. design, ‘When the results of such an approach are interpreted, it would be necessary to consider the differences that are inherent in self-report and in more objective observa— tions of behavior. Kidder (1981) reports that partici- pants' ratings of effectiveness and satisfaction do not generally correlate very well with objective measures. The other group process measure, the Group Content Form, appeared to be useful in providing a general overview of meetings. It may be valuable to expand this measure and have it completed by each behavioral observer during the meetings, rather than after the meetings. In effect, the Group Participation Form and Group Content Form could be merged into one group pro- cess measure . 166 Suggestions for Evaluation of the Intervention Because of inflation of participant ratings (Kidder, 1981), objective indicators should be included to provide a more valid evaluation of the intervention. On the surface, it seems that the best predictor of the effectiveness of the support group would be attendance rates. Other objective criteria that would be expected from a successful support group format are positive changes in :networks and. other’ positive life events (such as remaining free from abusive relationships) observed at follow-up points that appear to be stimu- lated by the support group. Of course, the introduc- tion of an experimental design with random assignment would greatly increase the confidence with which any such conclusions could be drawn. It is recommended that further modifications of the support group inter- vention be introduced, based on input from partici- pants, and then if the intervention appears to promote positive changes, that a true experimental study be conducted to evaluate its effectiveness. Ultimately, the greatest impact that the support group could pro- duce would be a decrease in the incidence of violence experienced by the participants. Additional positive outcomes would include changes in network variables in the direction reported to be associated with positive 167 adjustment, such as larger networks that provide sup- port in all resource categories, higher satisfaction ratings, greater degree of reciprocity, and more com- plex relationships. It is important to consider that if there are not enough women interested in the support group to form a control group, this may indicate that the intervention is not greatly needed and that efforts should be redi- rected or abandoned. Summary It has been widely reported that access to ade- quate amounts of social support can diminish the nega- tive effects of stress. Therefore, the present study was undertaken to ihwestigate the social support net- works of women who have been in abusive intimate rela- tionships and to determine the impact of a support group intervention on ‘the amount of social support available to them. A general summary of the characteristics of the networks of the participants of a support group for women.‘who had. ended abusive relationships was pre- sented. It appeared that the networks of the partici- pants were smaller and that their contacts with sup- porters were fewer than other groups of women, which may support the accuracy of the claim that battered women tend to be more isolated. It was also found that 168 the participants' networks contained twice as many friends as relatives. They were also found to have a two-to-one preference for informal sources of support, which is consistent with the report of Young et al. (1982) that informal sources of support are generally preferred over more formal sources. It was further discovered that the networks of over half of the respondents included someone else who had been in an abusive relationship. The length of the participants' friendships were shorter than those in another study, which fits the prediction that this life transition may require the forming of new relationships with those supportive of the changes that this transition requires. The reports of others that formal employment is related to indicators of positive adjustment was replicated in the present study. For example, women who were employed outside of their homes tended to have larger networks and more frequent contacts. The support group intervention departed from the researcher's expectations in a number of ways. Most significantly, it appeared to appeal more to women who had returned to their abusive partner, rather than to women who were redesigning their lives without their abusive partners. It was not clear why the group developed this way, however, those remaining in abusive relationships may have needed more support because they 169 tended to have smaller networks than other women. Overall, attendance was much lower than expected, which had a detrimental influence on the effectiveness of the support group intervention by limiting the potential for the exchange of support between group members. Due to the difficulties experienced with the support group intervention, a number of ideas were presented to revise the format of the group. Other interventions that may be of interest to the women who have been in abusive relationships were also suggested. It was pro- posed that the results of a needs assessment could best advise interventionists on what types of programs would be most welcome. Participants in the study recommended that the support group include a larger number of attenders, more speakers, and outside social activi- ties. Other proposed intervention strategies involved advocacy programs, mutual assistance cooperatives, and prevention strategies. Finally, it was noted that certain changes could be introduced to improve the measures that were created for the present study. Revisions were suggested for network variable operationalization and for the devel- opment. of a new group jprocess measure 'which would incorporate the existing measures. In addition, methodology for the assessment of the reliability of 170 the measures and experimental evaluation of future interventions was discussed. 171 Endnotes Introduction 1These figures do not include unmarried couples who were living together. 2It is curious, however, that many men were con- sistently able to maintain enough control to inflict the injuries in places that were not ordinarily visi- ble. Methods 3There were also cases (between three and six women) when a woman came to a meeting without knowing that she would be asked to be interviewed afterward. This situation arose when a member of the original sup- port group returned after quite a while. The facilita- tors attempted to contact past members to notify them of the formation of the new group, but were not com- pletely successful. 4One third of the women who attended support group meetings moved during the twenty week data collection period. Results 5Network variables that were not directly assessed by' the measures ‘were excluded from. the statistical analyses. These included, adequacy of support: den- sity; complexity of relationships; reciprocity: and supportive resources. 6These figures do not match those in Table 1 due to rounding of individual participant's responses. 7In some cases the participants declined inter- views and in the others the researchers were not ade- quately prepared for the interviews to be administered. Frequencies of these cases were not recorded. 172 Appendices 173 Appendix A 174 INTERVIEW CONSENT FORM I have freely consented to be interviewed by (interviewer's name) under the supervision of Dr. Robin Redner of Michigan State University. The study has been explained to me and I under- stand the explanation that has been given and what my participation will involve. I understand that my participation in the study does not guarantee any beneficial results to me. I also understand that I can belong to this support group without participating in the research, with no penalty. I understand that I may be contacted if I stop coming to the group about why I stopped coming. I understand that I am free to discontinue my par- ticipation in this study at any time without penalty. I understand that the results of this study will be treated in strict confidence and that I will remain anonymous. Within these restrictions, results of the study will be made available to me at my request. I understand that, at my request, I can receive more information about the study after my parti- cipation is completed. 175 Participant's Name: Interviewer's Name: Participant's Signature: Interviewer's Signature: Date: 176 INITIAL INTERVIEW Participant's Name: Address: Phone Number: Your name will not appear with any of the answers that you give us, and this cover sheet is only for our own information. Your name, address, and phone number will not be given to anyone under any circumstances unless you desire. There may come a time in the future when we'd like to ask you some further questions. You do not have to agree to this if you don't want to. For your own safety, however, are there any times that we shouldn't call you? Times Unsafe to Call: If you have no phone, is is alright to stop by your house? I‘D. 177 Number: Interviewer's Name: Date: INITIAL INTERVIEW Indicate race of respondents: Date of birth: How many children do you have? none one two three four five or more How many boys? How many girls? How old are they? What was the last grade in school that you completed?? 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 (circle one) Are you employed If yes, doing what What is/are your source(s) of income? What is your monthly family income? I'd like to ask you some questions about why you came here today. I know these questions are personal, but could you tell me what your relationship is like that prompted your coming to this type of a support group? 178 9. Relationship to abusive man: married and living together married but not living together legally separated. divorced living together unmarried were living together unmarried previously but now are separated dating widow other: please specify 1- *How long have you been apart? 10. How long have you known him? 11. When did he first abuse you? 12. When did he last assault you? If respondent is no longer being abused, go to l4. 13. About how often does he abuse you? If respondent is still in abusive relationship, skip 14. 14. Why has he stopped assaulting you do you think? 15. How often do you see/talk to him? 16. Have you ever left him? If yes, how many times? 17. Have you ever pressed charges against him? If yes, how many times? 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 179 How would you define mental abuse? How would you define physical abuse? Do you consider yourself as having been: physically but not mentally abused mentally but not physically abused both physically and mentally abused Have you ever been to a shelter for women with abusive partners? If yes, how many times? Have you ever been to a support group for women with abusive partners? If yes, where? How did you hear about our group? How long ago? Before you came today, what all had you heard the support group offered? What do you feel is the one most important thing this group can offer you? What else would you like the group to offer? What made you decide to come tonight? 29. 30. 31. 32. 33. 34. 35. 180 Is this a good time for you to meet? If not, why not? How did you get here tonight? How long did it take you to get here? Is getting here a problem for you? If yes, why? If you have children did you bring them tonight? If no, did you know that we offer childcare? How many of the women at tonight's meeting have you met before? (If none, go to 37) Who are they and where do you know them from: NAME KNOW FROM RELATIONSHIP ** CARD 1 ** 36. How’ would you describe your relationship with each? (Place letter under "relationship" above.) A. seen her around B. talked briefly to her once or twice C. talked to her a few times, but I don't consider her a close friend D. she is a friend E. she is a close friend 37. 38. 39. 181 In general, how do you feel about the other women in the group? How do their experiences compare with your own (abusive experiences) How well do you think they understand the situa- tion that you are living in? Now I'd like to ask you some questions about your friends and friendships in general. 40. How many people do you consider close friends? A close friend is someone you can talk to about your feelings and to whom you can turn for help. *NOTE: If she cannot after thorough probing, come up with any close friends, go to 55. *NOTE: Proceed through questions 41-54 for first friend, then repeat sequence for second and third friends. NAME SEX CITY LENGTH OF F WHERE THEY MET CLOSENESS 41. 42. 43. What is the first name of your closest friend (or initials) What city does live in? (or how far away) How long have you known ? 44. ** 45. 46. 47. 48. 182 Where did you meet ? CARD 2 ** Which of the following statements best describes your relationship with ? (Read each alternative to her) 1. There is nothing that I cannot talk to about or that s/he would not do for me if s/he could. 2. I can talk to about most of my feelings and I can rely on her/his help most of the time. 3. Sometimes I can talk to about my feel- ings and I can sometimes rely on her/his help. 4. There are only a few things that I can talk to about or rely on her/him to help me with. What are the qualities that make such a good friend? (ask for each) 1. 2. 3. How many times have you seen in person in the last week month 1. 2. 3. What kinds of things do you usually do with them (each) 49. 50. 51. 52. 53. 54. 183 How do you usually feel after you've spent time with How many times have you talked on the phone to in the last week month What kind of things do you usually talk with each about? How do you usually feel after your phone calls with Do you write letters to How often (times per week or month) Which, if any, of your friends know each other? 55. 184 How many of your relatives do you consider to be supportive (IF SHE ASKS, in-laws are included) *NOTE: If she cannot after thorough probing, come up with any supportive relatives, go to 64. *NOTE: Proceed through questions 56-63 for first rela- 56. 57. 58. 59. tive, then repeat sequence for second and third. Who is your most supportive relative? (Repeat for 2nd and 3rd if applicable) NAME SEX RELATIONSHIP CITY OF RESIDENCE How many times have you seen in person in the last week month What kinds of things do you usually do with them? (each) How do you usually feel after you've spent time with 60. 61. 62. 63. 64. (Ask women what assailant's first name is) 185 How many times have you talked on the phone to in the last week month What kinds of things do you usually talk to each about? How do you usually feel after your phone calls with Do you write letters to How often (times per week or month) How many other women do you know who have been in a relationship with an abusive man? 65. 66. 186 How many of your close friends (from above) have you talked to about your (ex)husband/ (ex)boyfriend's abusiveness? If none, why not? Have any of your close friends been in a relation- ship with an abusive man? Who: * Skip 67 and 68 if close friends to NOT know about the abuse. 67. 68. 69. 70. What do your close friends (from above) think that you should do about the situation with (assailant) (Get opinion for each friend) 1. 2. 3. How do you feel about their opinion? (for each friend) Have you talked to any of your relatives about 's abusiveness? circle one: all most some few none (If answer to 69 is "none") Why not? 71. 187 Have any of them had a similar abusive experience? If yes, who? * Skip 72 and 73 if supportive relatives do NOT know about the abuse. 72. 73. 74. 75. What do your supportive relatives (from above) think that you should do about the situation with (Get opinion for each relative) How supportive do you think that your close friends will be no matter what you decide to do about your situation with ? How supportive do you think that most of your relatives will be no matter what you decide to do about your relationship with ? 188 ** CARD 3 ** 76. port 77. What are the best three sources of emotional sup- for you at_thi§_tims? A source of emotional support is someone who really cares (1 indicates most important) (Rank 1, 2 and 3) social worker (ex)husband/(ex)boyfriend (assailant -- EXPLAIN) her parents her children other relative - specify counselor or therapist minister, priest or rabbi female friend who HAS been in an abusive relationship female friend who has NOT been in an abusive relationship a male friend (other than assailant) other - specify When do you MOST often think about GOING BACK to ** OR ** When are you LEAST likely to think about LEAVING 78. When are you LEAST likely to thing about GOING BACK to ** OR ** When are you MOST likely to think about LEAVING 79. What do you do to get over the times when you feel lonely or depressed? 189 80. Do you feel that your friends or relatives have had any influence on what you've decided to do about the abuse from ? (If so) How? (If not) Why not? ** Thank her for her time and patience, etc. ** CARD #1 A. I have seen her around. B. I have talked to her briefly, once or twice. C. I have talked to her a few times, consider her a close friend. D. She is a friend. E. 190 She is a close friend. but II don't 191 CARD #2 1. There is nothing that I cannot talk to her/him about or that she/he would not do for me if she/he could. I can talk to her/him about most of my feelings and I can rely on her/his help most of the time. Sometimes I can talk to her/him about my feelings and I can sometimes rely on his/her help. There are only a few things that I can talk to her/him about or rely on her/him to help me with. 192 CARD #3 1. a social worker 2. your (ex)husband/(ex)boyfriend 3. your parents 4. your children 5. another relative 6. your counselor or therapist 7. a minister, priest or rabbi 8. a female friend who has been in an abusive relationship 9. a female friend who has p_o_t_ been in an abusive relationship 10. a male friend (who has not been abusive) 11. other 193 Appendix B I.D. 194 Number: Interviewer's Name: Date: FOLLOW-UP INTERVIEW Is this a good time for you to meet? If not, why not? How long did it take you to get here? Is getting here a problem for you? If you have children, did you bring them with you? What are some things you like most about the group? What are some things you think need to be changed? Have you ever been here for a film? If yes, how did you like it? not at all not very much neutral liked it pretty well liked it very much 195 If yes, what was the film about? If no, do you think you'd like to see a film? 9. Have you ever been here for a speaker? If yes, how did you like her? not at all not very much neutral liked pretty well liked very much If yes, what did the speaker talk about? If no, do you think you'd like a speaker to come in? 10. Have you told other women with abusive partners about this group? 11. Is this group what you expected when you first came? 12. How is it different? Now I'd like to ask you some questions about your personal life if that's okay with you. (Interviewer will use information from Initial Inter- view to guide these questions.) 13. Are you still ? (married, dating, etc.) Relationship to abusive man: married and living together married but not living together legally separated divorced living together unmarried were living together unmarried previously but are now separated dating widow other: please specify 14. (If 15. 16. 17. 18. 19. 20. 196 When last has he assaulted you? respondent is still in abusive relationship) About how often does he assault you? How often do you see/talk to him? Do you feel this group has had an impact on your relationship with your abuser? yes no does not apply If yes, how so? Do you feel this group has had an impact on your feelings about yourself? yes no does not apply If yes, how so? Do you feel this group has had an impact on any other part of your life? Your relationship with friends, family, life, etc.? If yes, how so? How many of the women at tonight's meeting have you met before? (If none go to 23) 21. Who are they and where do you know them from: NAME KNOW FROM RELATIONSHIP 197 ** CARD 1 M 22. 23. 24. 25. 26. 27. 28. How would you describe your relationship with each? (Place letter under "relationship" above.) A. see her around B. talked briefly to her once or twice C. talked to her a few times, but I don't consider her a close friend D. she is a friend E. she is a close friend In general, how do you feel about the other women in the group? How do their experiences compare with your own (abusive experiences) How well do you think they understood they under- stand the situation that you are living in? How many of the women in the support group have you talked to on the phone in the last month? How many times: (for each woman talked to) What were the reasons for the calls? How many of the women in the support group have you met with (outside of the meetings) in the last month? 198 How many times (for each) 29. What did you do together? 30. Do you have any other general comments that you'd like to make about the support group? Now I'd like to ask you some questions about your friends and friendships in general. 31. How many people do you consider close friends? A close friend is someone you can talk to about your feelings and to whom you can turn for help. *NOTE: If she cannot after thorough probing, come up with any close friends, go to 47. *NOTE: Proceed through questions 32-45 for first friend, then repeat sequence for second and third friends. NAME SEX CITY LENGTH OF F WHERE THEY MET CLOSENESS 32. What is the first name of your closest friend (or initials) 33. What city does live in? (or how far away) 34. How long have you known ? 35. 199 Where did you meet ? ** Card 2 ** 36. 37. Which of the following statements best describes your relationship with ? (Read each alternative to her) 1. There is nothing that I cannot talk to about or that s/he would not do for me if s/he could. 2. I can talk to about most of my feelings and I can rely on her/his help most of the time. 3. Sometimes I can talk about my feelings and I can sometimes rely on her/his help. 4. There are only a few things that I can talk to about or rely on her/him to help me with. What are the qualities that make such a good friend? (ask for each) 1. 2. 38. 39. 40. 41. 42. 200 How many times have you seen __ in person in the last week month 1. 2. 3. What kinds of things do you usually do with How do you usually feel after you've spent time with How many times have you talked on the phone to in the last week month What kinds of things do you usually talk with each about 43. 44. 45. 46. 201 How do you usually feel after your phone calls with Do you write letters to How often (times per week or month) Which, if any, of your friends know each other? How many of your relatives do you consider to be supportive (IF SHE ASKS, in-laws are included) *NOTE: If she cannot after thorough probing, come up with any supportive relatives, go to 55. *NOTE: Proceed through questions 47-54 for first rela- 47. tive, then repeat sequence for second and third relatives. Who is your most supportive relative? (Repeat for 2nd and 3rd if applicable) NAME SEX RELATIONSHIP CITY OF RESIDENCE 48. 49. 50. 51. 52. 202 How many times have you seen __ in person in the last week month 1. 2. 3. What kinds of things do you usually do with _____? How do you usually feel after you've spend time with How many times have you talked on the phone to in the last week month What kinds of things do you usually talk to each about 53. 54. 55. (Ask 56. 57. 203 How do you usually feel after your phone calls with Do you write letters to How often (times per week or month) How many other women do you know who have been in a relationship with an abusive man woman what assailant's first name is) How many of your close friends (from above) have you talked to about your (ex)husband's/ (ex)boyfriend's abusiveness? If none, why not? Have any of your close friends been in a relation- ship with an abusive man? Who: *Skip 58 and 59 if close friends do NOT know about abuse. 58. 59. 60. 61. 62. 204 What do your close friends (from above) think that you should do about the situation with (assailant) (Get opinion for each friend) 1. 2. 3. How do you feel about their opinion? (for each friend) 1. 2. 3. Have you talked to any of your relatives about '3 abusiveness? circle one: all most some few none (If answer to 60 is "none") Why not? Have any of them had a similar abusive experience? If yes, who? *Skip 63 and 63 if supportive relatives do NOT know about the abuse. 63. What do your supportive relatives (from above) think that you should do about the situation with (Get opinion for each relative) 64. 65. 66. 205 How do you feel about their opinions (for each)? How supportive do you think that your close friends will be no matter what you decide to do about your situation with ? How supportive do you think that most of your relatives will be no matter what you decide to do about your relationship with ? ** CARD 3 ** 67. What are the best three sources of emotional support for you AT THIS TIME? A source of emotional support is someone who really cares (1 indicates most important) (Rank 1, 2 and 3) social worker (ex)husband/(ex)boyfriend (assailant-- EXPLAIN) her parents her children other relative - specify counselor or therapist minister, priest or rabbi female friend who HAS been in an abusive relationship female friend who has NOT been in an abusive relationship a male friend (other than assailant) other - specify 206 68. When do you MOST often think about GOING BACK to ** OR ** When are you least likely to think about LEAVING 69. When are you LEAST likely to thing about GOING BACK to ** 0R ** When are you MOST likely to think about LEAVING 70. What do you do to get over the times when you feel lonely or depressed? 71. Do you feel that your friends or relatives have had any influence on what you've decided to do about the abuse from ? (If so) How? (If not) Why not? ** Thank her for her time and patience, etc. ** 207 CARD #1 A. I have seen her around. B. I have talked to her briefly, once or twice. C. I have talked to her a few times, but I don't consider her a close friend. D. She is a friend. E. She is a close friend. 208 CARD #2 1. There is nothing that I cannot talk to her/him about or that she/he would not do for me if she/he could. I can talk to her/him about most of my feelings and I can rely on her/his help most of the time. Sometimes I can talk to her/him about my feelings and I can sometimes rely on his/her help. There are only a few things that I can talk to her/him about or rely on her/him to help me with. 209 CARD #3 1. a social worker 2. your (ex)husband/(ex)boyfriend 3. your parents 4. your children 5. another relative 6. your'counselor or therapist 7. a minister, priest or rabbi 8. a female friend who has been in an abusive relationship 9. a female friend who. has not been in an abusive relationship 10. a male friend (who has not been abusive) 11. other 210 Appendix C 211 I.D. Number: Interviewer's Name: Date: Drop-out Survey Hello, this is from the support group for battered women. Is it safe for your to talk? (If not, ask if there is a better time to call and hang up.) We're trying to improve the group to better meet the needs of more women, so we're calling all of the women who haven't been back to the group in a while. Would you mind answering a few questions about what you thought of the group? It would be very helpful and any thing you say will be STRICTLY CONFIDENTIAL. (If she agrees, continue: if not, politely thank her for her time and hand up.) 1. We were wondering why you haven't been back in a while. (Possible ‘prompters: Is it something about the group? About a certain woman in the group? Time conflicts? Transportation? Unsafe? Doesn't meet her needs?. 2. What in your opinion could improve the group? 3. Do you ever see/hear from any women in the group: (Find out WHO, HOW OFTEN, WHEN LAST, and IF SHE CONTACTED MEMBER OR IF MEMBER CONTACTED HER.) 212 Were you ever here for a film? (If yes, find out which film, what it was about, and how she felt about it.) Were you ever here for a speaker? (If yes, find out who the speaker was, what the topic was, and how she felt about it.) Questions Regarding Abusive Relationship (Interviewer will use information obtained from Initial Interview to guide these questions.) 6. Are you still ? (married, single, etc.) Relationship: Are you in a dangerous situation now? Still being abused? When last did he assault you? How often do you see/talk to him? Do you have anything else you'd like to say about the support group? **Thank her very much for her time and remind her that if she ever needs anything she knows where we are and that we'll be there for her** 213 Appendix D 214 .0u00 alu¢nnflflh MDHB‘UflZ .KU‘GOfimb u>HBHmOm «Kucnaflmh fl>HB¢Ufi2 ”gotmammh fl>HBHmOm u&U¢nOflflh fi>Hh¢Ufiz .¥U¢naflmh H>H8Hmom .lU‘flONNh M>HBHBHmOm .KU‘DOHBL fl>HF£UMz .¥UH9Hm0m mmxho Batman: .fl>HhUfl¢HD Lawn 00030» N>HB¢Umz Lama ouozou fl>HhHmOm 24:03 cuazou N>Hb¢0flz 2‘20: 00030» N>H9Hm00 macaw Quezou M>HP¢UNZ ADOGU 00030» N>H9Hwom DH Gmntfl: 215 Appendix E 216 Group Content Form List names of all women present: Women who brought children; Name of women and how many kids List all topics discussed: Film title, brief summary, length (if applicable): Speakers's name, brief summary of talk, length (if applicable): List all positive comments made about the group: List all negative comments made about the group: List all positive comments made by one women toward another: 217 List all negative comments made by one women toward another: List all positive comments made by one women about herself: List all negative comments made by one women about herself: How many women are currently living with abusive man __ What are their names? How many women drove someone else to the meeting today Who drove whom? Your names: Date: 218 References 219 References Borkman, T. (1976). Experiential knowledge: A new concept for the analysis of self-help groups. §ogiai §ezyige Review, pp, 445-456. Burnell, G. M., Dworsky, W. A. & Harrington, R. L. (1972). Post-abortion group therapy. Ametican W. 12.9. 134-137- Bybee, D. I. (1980). ' thers: A n eds ssess- me t. Unpublished master's thesis, Michigan State University, East Lansing. Caldwell, R. A. & Bloom, B. L. (1982). Social Sup- port: Its structure and impact on marital disrup- tion. Ametigap Joutpai of Community Esygpglggy, i9, 647-667. Candy, S. G., Troll, L. E., & Levy, S. G. (1981). A developmental exploration of friendship functions in women. P c o o 0 Women uarter , 5(3), 456- 472. Carty, K. J. (1983). Evaluation of domestic violence setyices: Fiscal years 1979-80 and 1980-81. Plan- ning and Evaluation Division: Office of Planning, Budget and Evaluation: Michigan Department of Social Services. Cauce, A. M., Felner, R. D. & Primavera, J. (1982). Social support in high-risk adolescents: Structural components and adaptive impact. Amerigap Journal of gommunity Psychology, it, 417-428. Craven, P. & Wellman, B. (1973). The network city. Socigiogica; Ingpipy, 4;, 57-88. . Davidson, S. & Packard, T. (1981). The therapeutic value of friendship between women. Psychology of Women ngtterly, 5(3), 495-510. Dumont, M. In (1974). Self-help treatment programs. Apepigap Joutnai of Psychiatry, 131, 631-635. \ \ 220 Duncan-Jones, P. (1981) . The structure of social relationships: Analysis of a survey instrument. Sgsial_£§xshiatrx. 16. 55-61- Durman, E. C. (1976). The role of self-help in ser- vice provision. u a o d v'o a Seisnse. 12. 433-443- Eaton, W. W. (1978). Life events, social supports and psychiatric symptoms: A re-analysis of the New Haven data. na 0 e nd Soc'a ehavio , 12, 230-234. Field, T. M., Widmayer, S. M., Stringer, S. & Ignatoff, E. (1980). Teenage, lower-class, black mothers and their preterm infants: An intervention and develop- mental follow-up. Qniid Qeveiopment, 5;, 426-436. Fleming, J. B. (1979). Stepping wife ebpee. Garden City, New York: Anchor Books. Froland, C., Pancoast, D. L., Chapman, N. J., & Kimboko, P. J. (1981). e ' tw ks n uma eepyige_. Beverly Hills: Sage Pub. Gartner, A. J. & Riessman, F. (1982). Self-help and mental health. fipspitel apd Community Esyepiatm, 11, 631-635. Gelles, R. J. (1976). Abused wives: Why do they stay. ggpppel of Mapriage and the Family, 55, 659- 668. Goodstein, R. K. & Page, A. W. (1981). Battered wife syndrome: Overview of dynamics and treatment. Amsri220.122rnal_2f;£§xshiatrx. 133(8). 1036-1044- Gottlieb, B. 13. Preventive interventions involving social networks and social support. In B. H. Gottlieb (Ed.), Soeiei petwotks and eocial suppott (pp. 201-232). Beverly Hills: Sage Publications, 1981. Gray, J. D., Cutler, C. A., Dean, J. G. & Kempe, C. H. (1979). Prediction and prevention of child abuse and neglect. goutpei of Sogial Issues, 55(2), 127- 139. 221 Gray, J. 8 Kaplan, B. (1980). The lay health visitor program: An eighteen-month experience. In C. H. Kempe and R. E. Helfer (Eds.), The tattered chiig (pp. 127-139). Chicago: The University of Chicago Press. Hanks, S. E. & Rosenbaum, C. P. (1977). Battered women: A study of women who live with violent alcohol-abusing men. A e ' n our a f O - EhiQLIY: AZ, 291-305- Henderson, S. (1977). The social network, support and neurosis: The function of attachment in adult life. Briti§n_lgsrnal_gf_£§xshiatrx. 131. 185-191- Henderson, 8., Duncan-Jones, P., Byrne, D. G., & Scott, R . (1980). Measuring social relationships: The Interview Schedule for Social Interaction. Esyeno- lesisal_uedisins 19. 723-734- Hilberman , E . (1980) . Overview, The "wife-beater ' 8 wife" reconsidered. e 'ca 0 na 0 8 ch' t 137(11), 1336-1347. Hirsch, B. (1980). Natural support systems and coping with major life changes. a o mmu- nitx_E§xshgl_gx 3. 159-172. Hirsch, B. J. (1981). Social networks and the coping process: Creating personal communities. In B. H. Gottlieb (Ed.), 0 ' wo ks nd socia su t (pp. 149-170). Beverly Hills: Sage Publications. Holahan, C. J. & Moss, R. H. (1982). Social support and adjustment: Predictive benefits of social cli- mate indices. e ' an Journa o Communit Ps - EthQQY: 19. 403‘415- Katz, A. H. (1965). Self-help groups. In H. L. Lurie (Ed.), n oe i o Soc'a Work (pp. 680-683). New York: National Association of Social Workers, 1965. Kidder, L. H. (1981). Research methods in social relations. New York: Holt, Rinehart and Winston. Lin, N., Dean, A., & Ensel, W. M. (1981). Social sup- port scales: A methodological note. Schizophrenia Bulletin. 1. 73-87- 222 Linn, J. G. & McGranahan, D. A. (1980). Personal dis- ruptions, social integration, subjective well-being and predisposition toward the use of counseling ser- vices. Amsri0an_I22rD_l_Qf_§2mmuni_x_£_xsb_lsgx 87- -100. Makowsky, V. P. (1983) . Exploring the meaning of friendships with women and with men. A workshop presented at the Annual Meeting of the Association for Women in Psychology, Seattle, Washington, March 3-6, 1983. Martin, D. (1976). Satteteg gives. San Francisco: Glide Publications. Martin, D. (1978). Battered women: Society's prob- lem. In J. R. Chapman and M. Gates (Eds.), The vig- MW (pp. 111-141). Beverly H1118: Sage Publications. McGuire, J. C. & Gottlieb, B. H. (1979). Social sup- port groups among new parents: An experimental study in primary prevention. ou na 0 l' ' a eniin Esyghoiogy, 5, 111-116. Meyers-Abell, J. E. & Jansen, M. A. (1980). Assertive therapy for battered women: A case illustration. e v'ora T e a an e ime tal s - QDiQLIY. 11. 301’305- Michigan women's Commission. (1977). Domestic assault: A report on family violence in Michigan. Lansing, MI: State of Michigan. Milardo, R. M. (1982). Friendship networks in developing relationships: Converging and diverging social environments. So 'a s cho o a t 55(3), 162-172. Minde, K. , Shosenberg, N. , Marton, P. , Thompson, J. , Ripley, J., & Burns, 8. (1980). Self-help groups in a premature nursery--a controlled evaluation. Jenpnei g: Pedietpics, 55(5), 933-940. Mitchell, R. E. (1982). Social networks and psychi- atric clients: The personal and environmental con- text. Amezican Jonrnal of Community Psychoiogy, 19, 387-401. 223 Mitchell, R. E. & Hodson, C. A. (1983). Coping with domestic violence: Social support and psychological health among battered women . a o f communit¥_£§xshglegL 11. 629- -654. Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1981). The development of an instrument to measure social support. Nntsing Beseargn, 59, 264-269. Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1983). Further development of the Norbeck Social Support Questionnaire: Normative data and validity testing. aneing Reseapcn, 55, 4-9. Pagelow, M. D. (1984). W. New York: Praeger. Pancoast, D. L. (1980). Finding and enlisting neigh- bors to support families. In J. Garbarino and S. H. Stocking (Eds.), o e ' ' d d neglegt (pp. 109-132). San Francisco: Jossey-Bass, 1980. Pearson, R. E. (1983). Support groups: A conceptual- ization. The Eepeonnei end Guidange Sgnpnei, 5;, 361-364. Peterman, P. J. (1981). Parenting and environmental considerations. me 'ca 0 a o- m: 3(2) I 351-355. Polister, P. E. & Patterson, E. M. (1980). Social climates in community groups: Toward a taxonomy. Q0mm2ni12.nsntal_flsal§h_ggsrnsl, 16. 187-200- Powell, T. J. (1975). The use of self-help groups as supportive reference communities. Ametican Joutnei 9f_9rtbgssxshiatrx. 45. 756-764. Riessman, F. (1965). The "helper" therapy principle. Segiai flopk, 12. 27-32. Rousanville, B., Lifton, N., & Bieber, M. (1979). The natural history of a psychotherapy group for bat- tered women. Reygni_tty, 52, 63- 78. Roy, M. (1977). A current survey of 150 cases. In M. Roy (Ed.), Setteted women: A 05 chosoc'o o 'cal study of domestie viglence (pp. 25-44). New YCrk: Van Nostrand Reinhold. 224 Sarason, I. G., Levine, H. M., Basham, R. B. & Sarason, B. R. (1983). Assessing social support: The social support questionnaire. Soupnei pf Eepegnei- itx_and_§gsial_2§xsnglgsx. 14. 127-139- Schecter, S. (1982). Wom n n ma v'o en e. Boston: South End Press. Schwartz, M. D. (1975). Situation/transition groups: A conceptualization and review. American Journal of QIEDQEEXQDLQEIYI fiém 744-755. Silberfeld, M. (1978). Psychological symptoms and social supports. Soeiai Esycnietpy, 15, 11-17. Spiegel, D., Bloom, J. R., & Yalom, I. (1981). Group support for patients with metastatic cancer. Azgnives pf Genetei Psycnietpy, 55, 527-533. Star, B., Clark, C. G., Goetz, K. M., & O'Malia, L. (1981). Psychosocial aspects of wife battering. In E. Howell & M. Bayes (Eds.), Wgnen end nentei neaitn (pp. 426-439). New York: Basic Books. Stark, E. & Flitcraft, A. (1983). Social knowledge, social policy, and the abuse of women: The case against patriarchal benevolence. Ina D. Finkelhor, R. J. Gelles, G. T. Hotaling, & M. A. Straus (Eds.), The get]; side pf familiee (pp. 330-348). Beverly Hills: Sage. Straus, M. A. (1980). Stress and child abuse. In C. H. Kempe and R. E. Helfer (Eds.), The battered cniid (pp. 86-103). Chicago: The university of Chicago Press. Straus, M. A. (1977-78). Wife beating: How common and why? V'c ' o o : An t n t'onal ou nal, 2, Straus, M. A., Gelles, R. J., & Steinmetz, S. K. (1980). Senind giosed doots: Violence in tne Amet- iean femiiy. New York: Anchor Books. Thoits, P. A. (1982). Conceptual, methodological, and theoretical problems in studying social support as a buffer against life stress. Journal of Hea th and Secini Behaviop, 55, 145-159. 225 Tietjen, A. M. (1980). Integrating formal and infor- mal support systems: The Swedish experience. In J. Garbarino and S. H. Stocking (Eds.), Protecting WW (1313- 15-36)- San Francisco: Jossey-Bass. Turner, R. J. (1981). Social support as a contingency in psychological well being. urna 0 He 1th and figsisl_flshs_igr 22. 357- '367- Walker, L. E. (1984) . The batteped woman Syndrome. New York: Springer. Walker, L. E. (1979). The batteted woman. New York: Harper and Row. Walker, L. E. (1978). Treatment alternatives for bat- tered women. In J. R. Chapman and M. Gates (Eds.), Tne vigtimization pt wonen (pp. 143-174). Beverly Hills: Sage Publications. Walker, L. E. (1977-78). Battered women and learned helplessness. Victimology: An International Jeur- 1101. 2, 525-534. Wandersman, L. , Wandersman, A. , & Kahn, S. (1980) . Social support in the transition to parenthood. lgnpnei pf gomnunity Esyehoiogy, 5, 332-342. Wright, P. H. (1969). A model and a technique for studies of friendship. Jou na 0 x ri ent Seeiei Eeycnology, 5, 295-309. Young, C. E., Giles, D. E., & Plantz, M. C. (1982). Natural networks: Help-giving and help-seeking in two rural communities. e 'can Jou n o mmu- nity Eeyehoiogy, lg, 457-469.