UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Understanding married women's suicidal behaviour : counsellors' perceptions of self-concept and marital… White, Jennifer H. 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A8 W45.pdf [ 5.8MB ]
Metadata
JSON: 831-1.0053667.json
JSON-LD: 831-1.0053667-ld.json
RDF/XML (Pretty): 831-1.0053667-rdf.xml
RDF/JSON: 831-1.0053667-rdf.json
Turtle: 831-1.0053667-turtle.txt
N-Triples: 831-1.0053667-rdf-ntriples.txt
Original Record: 831-1.0053667-source.json
Full Text
831-1.0053667-fulltext.txt
Citation
831-1.0053667.ris

Full Text

UNDERSTANDING MARRIED WOMEN'S SUICIDAL BEHAVIOUR: COUNSELLORS' PERCEPTIONS OF SELF-CONCEPT AND MARITAL DYNAMICS by JENNIFER H. WHITE B.A., The University of Victoria, 1984 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Counselling Psychology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1991 © Jennifer H. White, 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of CCLWGpJJj /g f^jChOlOC^J The University of British Columbia Vancouver, Canada Date -5tftUf\bC/^3/3\  DE-6 (2/88) Abstract Counsellors' perceptions of married suicidal women were investigated through a qualitative analysis of counsellors' clinical records and follow-up interviews. Four suicide intervention counsellors, ranging in age from 29 to 52, who collectively had 24 years of experience at a suicide intervention counselling agency were the subjects in this study. The experiences of their clients - three suicidal women and one husband of a suicidal woman - provided the material for the analysis of counsellors' clinical reports and follow-up interviews with the investigator. The suicidal women ranged in age from 20 to 41 and the husband was 34 years old. Analysis consisted of three concurent activities: data reduction, data display, and drawing/verifying conclusions. The clients, including one husband were found to suffer from negative self-concepts, previous suicide attempts, depression, and yet they indicated a willingness to get help. According to counsellors, the three wives could be characterized as "overfunctioning", fearful of spouses, prone to minimizing abuse, and committed to the relationships "at any cost." Counsellors found that husbands of the suicidal women tended to abuse alcohol and had explosive tempers. The marital relationships were characterized by intimacy problems, communication problems, and dependency. The families of the clients were found to be plagued by alcoholism, abuse, and mental illness. Counselling approaches included personal empowerment of client, educating about abuse, use of outside resources, making links with the past, and coaching on certain skills. The findings were meaningful as they offered a preliminary framework for understanding married women's suicidal behaviour by acknowledging the social, historical, familial, marital and /nfrapersonal levels of influence. Abstract n List of Tables viii List of Figures ix Acknowledgements x Introduction 1 Suicidal Individuals as Research Subjects 2 Counsellors as Data Collectors 3 Background 5 The Human Ecosystem 7 Women and Mental Health 10 Family Systems 12 Counsellor-Client System 15 Definitions 15 Assumptions and Limitations 16 Significance of the Study 18 Review of the Literature 21 Self-Concept 21 Self-Concept and the Suicidal Person 21 Self-Concept and the Spouse of the Suicidal Person 24 Interpersonal Factors 26 Interpersonal Distress and Suicidal Behaviour 27 Suicidal Women's Intimate Relationships 29 Suicide Attempters and Their Spouses 30 Methodological Issues 33 V Qualitative Research 36 Declaration of Biases 39 Research Questions 40 Methodology 43 Preliminary Investigation 43 Original Subject Group 43 Subject Recruitment Difficulties 44 Overview of the Qualitative Research Process 46 Procedural Framework 49 Data Display 51 Drawing/Verifying Conclusions 52 Application of Procedures 53 Data Reduction 53 Data Display 56 Drawing and Verifying Conclusions 60 Data Sources 61 Counsellors 61 Description of Client Group 64 Results 68 Findings 68 Characteristics of Wives Only 69 "Overfunctioning." 69 "Intimidation and isolation." 70 "Commitment at any cost" 71 Characteristics of Husbands Only 72 "Alcoholism" 72 "Never good enough." 74 "Help-seeking." 76 Characteristics of Marital Relationship 76 "Communication problems." 76 "Dependency." 78 "Intimacy problems." 80 Characteristics of Husbands' and Wives' Families of Origin 81 "Mental illness, alcoholism, and abuse." 81 Characteristics of Counselling Perspectives and Strategies 82 "Ambivalence." 82 "Educating about abuse." 82 "Personal empowerment." 83 "Making links with the past." 84 "Coaching." 85 "Outside referrals." 86 Implications 87 Discussion 88 Research Process 88 Summary of Results 89 Discussion of Client Themes 92 v i i Self-concept 92 Family histories 94 Responsibility for the relationships 96 Dependency 97 Communication problems 99 Discussion of Counselling Process Themes 100 Recommendations for Future Research 103 Conclusion 105 References 109 Appendices 115 Appendix A 115 Methodology of Original Investigation 115 Appendix B Tables 125 v i i i List of Tables Table 1 Site by Attribute Matrix 57 Table 2 Scores of Three Suicidal Women and One Husband on SelectedScales of the TSCS and the MSI 91 Table 3 Mean Scores for Selected MSI Items Across Three Groups 125 Table 4 Mean Scores for Selected TSCS Items Across Three Groups 125 Table 5 Descriptive Characteristics of Husbands in Each of the Three Groups 126 Table 7 List of Start-up Codes 130 i x List of Figures Figure 1 Causal Network 59 Figure 2 Qualitative Analysis Documentation Form 132 X Acknowledgements To those women who, when their lives hung in a precarious balance, had the courage to reach out and share their anguish, and to those counsellors who helped them heal - with empathy, respect, and a belief in their capacity to transcend their pain. To Kim - Thank-you for your enduring support and willingness to engage in such multi-faceted roles over the past 3 years. You have been my confidant, my human-thesaurus, my comic-reliever, my sounding board, my editor, my tension-reducer, my perspective-giver, my long distance adviser, and most importantly - my partner and friend! Introduction For centuries there has been an abundance of interest and accompanying research devoted to the topic of suicidal behaviour. Undoubtedly, much of the energy fuelling this interest in suicidal behaviour has had to do with our needs to understand the complex phenomenon of deliberate self-harm so that we may control and ultimately prevent future suicide deaths. Many valuable suicide prevention programs, risk assessment tools, and crisis intervention models have arisen in response to the research findings that have been generated in the growing field of suicidology. Suicide is such a troubling facet of the human experience because of its very final nature. Researchers and mental health professionals would like to understand better some of the dynamics that are involved in an individual's choice to take his or her own life, in order to intervene and prevent such an irreversible outcome. In their efforts to try and determine a person's risk for suicide, researchers have established a number of reliable predictive criteria, some of which will be articulated below. A great deal of research has been aimed at trying to describe a profile of those individuals who are at risk for suicide by identifying those traits and characteristics that have been associated with completed suicides in the past. Some of the factors that have been identified as important dimensions for assessing an individual's risk for suicide include, but are not restricted to: age, gender, marital status, prior history 2 of attempts, depression, isolation, and levels of hopelessness (Gove, 1972; Lester, 1983; Maris, 1981). Less research has been devoted to examining the suicidal person in the context of his or her interpersonal relationships. Suicidal Individuals as Research Subjects Recruiting suicidal individuals for the purposes of research could be considered one of the major challenges inherent in investigating this particular population. Suicidal individuals are usually in crisis, very depressed, experiencing feelings of hopelessness, and constricted in their vision. These qualities are not conducive to being a research participant, a task which demands trust, willingness, energy, time, and focus. This study was originally designed as a quantitative investigation, where husbands and wives representing three types of couples -couples with suicidal wives, couples in marital counselling, and couples from the community -- were to be compared on the dimensions of self-concept and marital satisfaction, using standardized instruments. Subject recruitment of suicidal women and their spouses presented a major obstacle in this study, precipitating the need for a revised research design. Because the design was changed in the midst of the project, the challenge of presenting a smooth and consistent chronology of events became difficult. A consequence of such a major shift in methodology is that some logical internal consistency might be lost in trying to provide a narrative of events. For a detailed description of the 3 specific difficulties encountered recruiting suicidal subjects for this study, see the appendix. Given the barriers that exist in trying to access suicidal subjects for research, it has been important to capitalize on those sources of data that do not necessarily rely on the cooperation of suicidal individuals themselves, yet still offer valuable information about the nature of married suicidal women. Counsellors who work directly with suicidal individuals are in unique positions, through their ongoing therapeutic relationships, for gaining access to suicidal clients' feelings, worldviews, and experiences in relationships. Counsellors as Data Collectors Counsellors themselves could be likened to data collectors in a qualitative research study. Using Haase and Myers (1988) description of the paradigmatic assumptions of qualitiative researchers, the following tenets could also be understood to be true of the orientations of most good counsellors: understanding of the phenomenon at hand as the goal, believing reality to be multiply determined, seeing subjective experiences as legitimate, allowing for the presence of discrepancies as part of an existential reality, understanding that an interactive unity exists between researcher/ subject (counsellor/client), seeing the world as dynamic, seeing that truth exists in changing patterns, and valuing uniqueness. Already engaged in a process of "data collection" by the very nature of their work with clients, counsellors can often provide very thorough, 4 credible, and rich descriptions of their clients' lives and experiences, which are usually overlooked by researchers. The nature of the clinical relationship lends itself well to client self-disclosure, a basis on which counsellors formulate opinions and generate hypotheses about their clients' lives. Nevertheless, counsellor opinions are just that - opinions -and are not ever to be misconstrued with a belief that they represent any objective truth about their clients. The perceptions of counsellors who work with married suicidal women, as recorded in clinical records and verbally reported through the process of taped interviews, will provide the foundation for this investigation. Counsellors are privy to a wealth of information about their clients, including that material that clients overtly present, as well as the more implicit, often non-verbal information that counsellors intuit from clients. Counsellors can provide rich and valid descriptions of their clients that are based on two main sources of information: the clients' disclosures to the counsellors of what they are experiencing and the counsellors' understanding of such material. In other words, "The skilled counsellor typically processes a wide array of information from the client's cognitive, affective, and behavioral responses, as well as his or her own responses to the client" (Johnson & Heppner, 1989, p. 428). Because of their unique role and vantage point, counsellors are more likely to understand their clients' behaviour objectively, without being sidetracked by emotionality. Naturally, counsellors will not have the same type nor the same degree of emotional investment in their 5 clients' experiences as the clients themselves do, giving the counsellors perspectives that can be more process-oriented, broader in vision, and contextually based. Using data sources other than clients themselves to understand a particular phenomenon such as suicidal behaviour can offer the researcher a broader spectrum for understanding. In her discussion of realistic dependent measures for clinical use, Nelson (1981) acknowledges that therapists working in clinical settings can be a valuable resource for data collection purposes. Commenting on the goals of "openly ideological research", Lather (1986) wrote, "The methodological task is to proceed in a reciprocal, dialogic manner, empowering subjects by turning them into co-researchers" (p.73). Using counsellors' perceptions to provide a window into understanding married women's suicidal behaviour is an approach that is not without its drawbacks and the specific limitations of this study will be discussed in a later section. Nevertheless, for the purposes of this study, the researcher's main purpose is to explore how counsellors working with married suicidal women understand their clients and how these counsellors report their experiences. Background It has been consistently found that women attempt suicide at a greater rate than men do, while men kill themselves more often. This phenomenon bears itself out irrespective of age, race or marital status (Lester, 1983; Neuringer & Lettieri, 1982). The literature is replete with 6 assertions that women's suicidal behaviour often arises out of crises in their intimate relationships ( Adam et al., 1978; Beck, Lester, & Kovacs, 1973; Fieldsend & Lowenstein, 1981, Lester, 1983; Maris, 1981; Suter, 1976; Tabachnick, 1961), however few studies examine the nature of these relationships directly. Counsellors' understanding of women's suicidal behaviour, as it arises out of an interpersonal context, will be a primary focus of this study. By offering a glimpse into how counsellors might understand married suicidal women and by acknowledging the fabric of interpersonal dynamics that suicidal women live within, this inquiry seeks to enlarge the framework within which women's suicidal behaviour might be understood. By acknowledging the collective and reciprocal impact made by the intrapsychic phenomena, the marital and family dynamics, and the cultural context on women's overall levels of functioning, the possibility for a richer understanding of women's suicidal behaviour may emerge for counsellors. Having a more comprehensive understanding of women's suicidal behaviour might enrich the therapeutic possibilities and could broaden the range of intervention strategies that might be used by a helping professional in responding to a woman's suicidal crisis. For example, instead of focusing only on intrapsychic material through individual counselling, the counsellor might include family of origin work, marital counselling, and consciousness-raising about the cultural constraints facing women, as part of the overall treatment approach. 7 Traditionally, a great deal of the research on suicide that seeks to understand the nature of such self-destructive feelings has tended to focus on the suicidal individual's intrapsychic processes or personality traits. Examples of such internal dynamics, as they have been studied in relation to suicide, include: depression (Lester, 1983), self-defeating cognition (Neuringer & Lettieri, 1982), hopelessness, psychic anxiety (Fawcett, 1988) and negative self-concept (Kaplan & Pokorny, 1976; Wenz, 1976). Meanwhile, less attention has been given to the specific nature of the person's interpersonal relationships, and the role that these relationships might play in the evolution of suicidal feelings. Firstly, it is important to explore whether or not interpersonal relationships play a role in married women's suicidal behaviour according to the counsellors who work with these women. Second, if interpersonal factors are believed to be significant in the emergence of women's suicidal behaviour, then it would be important to establish a model for understanding self-destructive behaviour that would be able to capture both the intrapsychic contributions and the interpersonal levels of influence. Conceptual Framework The Human Ecosystem An interactional approach to understanding human behaviour is captured by the ecosystemic perspective which combines general systems theory with an ecological model (Jasnoski, 1984). The ecosystemic perspective is concerned with all of the levels that comprise 8 the human experience and can be figuratively represented by layers of concentric circles. Starting with the individual resting at the centre, and emanating out in layers of human influence, the ecosystem can be understood to include the following: the intrapsychic components of an individual such as physiological or psychological processes, the marital system, the family network, the community level, the cultural context, and the physical environment. It is a useful model for characterizing human behaviour because it captures both the intrapersonal as well as interpersonal nature of human existence. The individual occupies a central and primary position in the ecosystem. Studying this level of the ecosystem in relation to suicidal behaviour would involve looking at individual factors and personality traits that might influence suicide attempts, and as already noted, this level of inquiry has been frequently adopted in the literature. Another layer of the human ecosystem could be conceptualized as the level of marital system influences. Studies that report on the intimate relationships of suicidal individuals are grounded in this level of the ecosystem. In the human ecosystem, there are several other interconnected levels influencing the human experience besides the individual and interpersonal levels that cannot be overlooked if one is trying to make sense out of an event such as attempted suicide or suicidal behaviour. It is felt that without at least acknowledging the influences of the family systems, counsellor-client systems, cultural contexts, and physical 9 environments of suicidal individuals, a comprehensive understanding of the suicidal experience would not be possible. For the practical purposes of research and clinical practice though, it is necessary to put some boundaries on how many of the interacting levels of the human ecosystem can be studied at once. After reviewing some of the literature that can be accommodated within an ecosystemic model, the limits of this particular study will be more clearly articulated. Other writers capture the flavour of the ecosystemic model and confirm the importance of examining the multiple levels of influence that characterize the evolution of suicidal behaviour. Stephens (1988) writes that suicide attempting is "processual" in nature and "...ultimately suicidal behaviours are not understandable as specific acts but are rather possible outcomes of more general social processes which preceded the act and provide a matrix from which self-destructive attitudes and roles may arise" (p.74). Adam (1990) offers a model for understanding suicidal behaviour that emphasizes the role of interacting psychosocial and environmental factors in the development of suicidal behaviour. His model seems to suggest that suicidal behaviour is multi-determined and can be understood to be the culmination of certain interacting psychosocial and environmental factors that work to create and exacerbate individual vulnerabilities and susceptibilities. According to his model, there are three levels at which social and environmental factors may produce suicidal behaviour: (a) factors can be "predisposing", leaving an individual vulnerable; (b) factors may be "precipitating", acting as a trigger for predisposed persons; and (c) factors may be "contributing", acting to increase the exposure of individuals to other predisposing or precipitating circumstances. For example, using Adam's model, early family disturbance such as child abuse could leave an individual vulnerable to depression or to developing a negative self-concept. Each of these could then be seen as possible predisposing factors to later suicidal behaviour. It would not be difficult to imagine that this individual might experience some difficulty forming healthy relationship attachments as an adult. The experience of relationship difficulties in a predisposed person, such as marital conflict, could act as a precipitating factor in the development of suicidal behaviour. Meanwhile, the person might try to manage his or her experience with alcohol or drugs which, in predisposed individuals, could be seen as contributing factors to suicidal behaviour. Studies that note the relationship between early family of origin distress and recent interpersonal conflict in suicidal individuals lend credibility to Adam's model (Adam et al.,1978; Stephens, 1988). Women and Mental Health Suter (1976) and Heshusius (1980) each bring a sociocultural perspective to understanding women's suicidal behaviour that incorporates levels of the human ecosystem. These authors cite the limiting and negative influence that women's socialization process and culturally prescribed feminine role expectations have on women's self-11 concepts, sense of personal mastery, and access to resources in the environment. Suter (1976) suggests that attempted suicide has become construed as a "feminine" behaviour because it involves a sense of personal helplessness and dependency on others to be rescued, both states which are reinforced for women in the larger culture. Heshuvius (1980) notes the convergence in the psychological literature between those traits considered "feminine" and those traits related to self-injury, including: lower status of mental health, internalized aggression, subservient personal identity leading to dependency, lack of personal autonomy and competence, helplessness and hopelessness, and a strong desire for personal relationships. Pommereau and Penoutl (1987) also recognize the roles that marital conflict and societal oppression play in the evolution of suicidal behaviour in women. They suggest that a woman's suicide attempt reveals her cultural and social dependency, and any examination of a woman's suicidal behaviour must take her lack of autonomy into account. While the emphases of each of these authors remain different, they each manage to capture important aspects of the ecosystemic perspective. At the very least, they all appear to support the need for broadening our framework for understanding women's suicidal behaviour, in order to move beyond individual trait approaches. Each author uniquely recognizes the importance of acknowledging the 1 2 individual, familial, marital, and cultural contexts within which suicidal behaviour develops, for guiding our further understanding. Due to the fact that women attempt suicide at a rate consistently greater than the rate for men, it is believed that women's suicidal behaviour warrants special consideration. Building on many of the above authors' works and ideas, this present study aims to investigate a rather narrow piece of the human ecosystem - the suicidal woman in the context of her marital relationship, as perceived by the counsellor who works with her. The familial, cultural, and societal levels of influence will also be acknowledged from a theoretical perspective, as it is believed that these forces impinge upon the other levels. Specifically, the qualitative analysis of the clinical records kept by counsellors working with suicidal women, plus analysis of the verbatim transcripts of interviews with these counsellors, will provide information about the experiences of married suicidal women, using the ecosystemic model as a conceptual framework for understanding. Family systems theory will provide further theoretical grounding for this study by understanding the suicidal woman and her spouse in their marital and family contexts, and by placing the counsellor-client relationship in a systemic context as well. Family Systems It is believed that family systems theory can offer a useful perspective for expressing the importance of family of origin influences and for offering a context within which to understand marital dynamics, and as such it will be drawn on as a theoretical framework throughout the study. One of the premises of family systems theory, which guided this research, is that pathology or distress in a couple or family has as much to do with the interactions or relationships between people as it does with the internal dispositions of individual members . An example of how a systemic theorist might view suicidal behaviour in one member of a spousal system was offered by Canetto (1988), "The suicidal individual may identify with, and directly manifest, feelings of inadequacy and helplessness, while the non-suicidal partner may deal with the same feelings of inadequacy by projecting them onto the partner"(p.17). The suicidal act in this case, could be seen to arise out of the context of a stressed relationship. Both partners could be understood to be contributing, at least covertly, to its evolution. Canetto went on to suggest that an interactive or systemic approach could not explain all suicide attempts occurring in the context of a significant relationship; rather, the potential for reciprocal and interactive forces to be operating between spouses should be at least considered as a fruitful avenue for counsellor exploration. A brief overview of some of the current thinking about the family systems and marital relationships of suicidal individuals will follow. Joseph Richman (1978, 1979, 1981, 1986) has written extensively about the family dynamics of suicidal persons and he suggests that the only way to understand an individual's suicide attempt is to view it in the context of the family relationships. He writes, "Suicide is not an individual act, but is part of a collusive communication system that involves an entire family and social network" (Richman, 1986, p.147). Aldridge (1984) suggests that individuals learn how to manage their distress through cultural and familial expectations. He cites several factors that might influence a person's risk for suicide, two of which are noteworthy here: a pattern of role disturbance or role failure and a symbiotic attachment between spouses where autonomy for either partner is not tolerated. These two factors are relevant to the present investigation, as they pertain directly to the individual self-concepts and marital relationships of suicidal individuals. Stephens (1988) categorizes the early parent-child relationships of suicidal women under five main themes: non-nurturing parents, absent parents, abusive parents, mentally ill parents, and alcoholic parents. Stephens suggests that these women's negative childhood experiences are closely linked to the development of self-destructive behaviour later on. She writes that, "For them, these sad early lessons about the nature of intimate relationships - they involve betrayal, neglect, abuse, and the engendering of feelings of worthlessness - are powerfully linked with their slide into depression and suicide attempting" (p.77). The overriding purpose of this study is to explore counsellors' views of married womens' suicidal behaviour as it emerges out of a rich and varied context. By adopting a more comprehensive viewpoint for understanding women's suicidal behaviour, clinicians working with such individuals in counselling can acknowledge the presence of a complex, 1 5 systemic interplay of interpersonal, cultural and familial, as well as individual factors. Counsellor-Client System By engaging with the client in a personal relationship, a counsellor is participating in a newly created system that will be governed by its own set of rules and regulating mechanisms. Counsellors and clients mutually influence each other and like members of a couple or family, "change occurs within a context that influences and is influenced by the component parts" (Friesen, 1985, p.13). The client's experience in counselling, like the wife's experience in her marriage, cannot be understood outside of the context in which it takes place. By "joining" with the client in the early stages of therapy, the counsellor enters into a mutually created relationship, governed by its own unique rules and set of expectations. In exploring the nature of counsellors' perceptions of married suicidal women it has been essential to acknowledge the dual role of participant and observer that a counsellor plays in the overall counselling process. Out of this context of interactive reciprocity, it is believed that a unique and worthwhile perspective for understanding the nature of married suicidal women emerges. The ecosystemic perspective and family systems theory will provide the theoretical framework for comprehending the complexity that such a contextual understanding engenders. Definitions 1 6 A suicide attempt will be defined as any self-acknowledged and deliberate effort to cause self-harm regardless of the degree of lethality. Suicidal threats or crises will be defined as the experience and admission of suicidal ideation as a primary motivator in reaching out for help to deal with an emotional crisis. Counsellors who are employed as Suicide Intervention Counsellors at S.A.F.E.R. (Suicide Attempt Follow-up Education and Research), Counselling Agency, a service of Greater Vancouver Mental Health Service, will be the subjects in this study. Suicidal women who have come to the attention of S.A.F.E.R. as a result of their suicidal crises will be the focus of counsellors' written clinical records, and will be the topic of open-ended interviews conducted with counsellors, with both of these sources providing the primary data for the study. Assumptions and Limitations This study is focused on counsellors' experiences of working with those heterosexual women who have been married for at least one year, who have been in suicidal crisis, threatened suicide, or attempted suicide within the past month, and who have no evidence or history of a major mental illness, nor any present evidence of substance abuse. Any significant findings generated by this study can only be generalized to those counsellors who have worked with suicidal women sharing such characteristics. All of the suicidal women in this study who counsellors are reporting on, came to the attention of the researcher through S.A.F.E.R. These 1 7 clients were referred there through one of several possible sources, including: self, hospital, general practitioner, another social service agency, family member, school personnel, crisis centre, or psychiatrist. As mentioned, those married suicidal women who were the focus for the interviews with counsellors and the subjects of counsellors' records, were recruited from S.A.F.E.R. counselling agency which means they had already agreed to pursue counselling for themselves. This study did not include information from other sources about women who did not reach out for help following their suicidal crises. This significantly limits the generalizability of the findings. The information generated from this study, based on the perceptions of counsellors, can only apply to suicidal women and their husbands following the suicidal crisis, and after they reached out for help. It cannot specify what factors might have existed prior to the crisis; one will only be able to speculate on this. However, the experiences of marital partners as reported to the counsellors, are likely to be framed in a historical context, and evidence may emerge in counsellors' records to suggest that certain patterns of interaction among couples represent longstanding patterns. A limitation of this study is the fact that two of the husbands were not clients at S.A.F.E.R., and therefore counsellors did not have direct experience working with these particular spouses. Counsellors' perceptions about these two men were based on comments made by the suicidal women to their counsellors about their husbands and did not 1 8 emerge out of counsellors' own experiences with these men. The fact that the perceptions of only four counsellors were explored and highlighted in this study could potentially be seen as a limitation, however this is not so much an issue when doing qualitative research, as often it is small samples that researchers prefer to use in conducting these types of studies (Bailey, 1978). Counsellors' perceptions are often heavily biased and emerge out of their own clinical orientations (Snyder & Thomsen, 1988). Six counsellors faced with one client who presents identically for each counsellor will probably formulate at least six different hypotheses about the nature of the client's difficulties. A tendency for counsellors to demonstrate low levels of agreement among one another in their judgements diminishes their qualifications as objective reporters considerably (Faust, 1986; Langer & Abelson, 1974). At the same time, this study is not necessarily concerned with discovering the "objective reality" of the client but rather, it focuses on how counsellors understand, experience, filter through, and respond to all of the information and actions presented by their clients. If themes about suicidal women and their spouse are found repeatedly by more than one counsellor, the findings that emerge in this study become strengthened. Significance of the Study The present study directly addresses an issue of practical concern to mental health professionals and it is believed that the findings could provide a useful foundation for enriching the therapeutic experience for suicidal women and their husbands. If the counsellors' perceptions reflect that suicidal women are showing evidence of marital distress as well as personal distress, then treatment approaches can incorporate both of these levels. Acknowledgement of the intrapersonal as well as interpersonal levels of the human experience could produce a more potent therapeutic outcome. Throughout the literature, women's self-concepts and the quality of women's intimate relationships have been mentioned as significant variables in the development of women's suicidal behaviour. Self-concept and marital relationships represent two distinct levels of the ecosystem and both dimensions will be considered in this study as they arise in the context of the larger ecosystem, and as they are understood to be influencing factors in suicidal behaviour by counsellors. How counsellors understand, experience, and report on married suicial women and their spouses will be the guiding thrust of this study. Those studies that understand suicidal behaviour from a point of view that emphasizes individual self-concept (Kaplan & Pokorny, 1976; Neuringer & Lettieri, 1982; Wenz, 1976; Wilson et al., 1971) will be explored and discussed with that research that focuses on suicidal women's marital relations (Bhagat, 1976, 1977; Bonnar & McGee, 1977; Canetto, 1988; Fieldsend & Lowenstein,1981; Pommereau & Penoutl, 1987; Stephens, 1985,1988), to provide the foundation upon which the research questions for the present study will be formulated. A specific 20 section will be devoted to a discussion of methodological considerations for studying suicidal subjects, as the emergence of the current qualitative research design arose in response to challenges encountered in the original quantitative investigation. A review of the literature cited above follows. Review of the Literature Self-Concept Self-concept is chosen here as a focus for inquiry as it is understood to be reliably implicated in the evolution of suicidal behaviour (Kaplan & Pokorny, 1976; Neuringer & Lettieri, 1982; Schmidtke & Schaller,1989; Suter, 1976; Wenz, 1976; Wilson, Miskimins, Braucht, & Berry, 1971), and seems to represent the individual level of the ecosystem comprehensively and efficiently. The impressions that counsellors have of the self-concepts of suicidal women and their husbands can be used as a source of contrast to the opinions that they have about these couples' marital dynamics. Sharpley and Khan (1982) found that positive self-concept was correlated with high marital adjustment, as measured on two standardized instruments, the Tennessee Self-Concept Scale and the Locke-Wallace Marital Adjustment Scale. By exploring these variables in the present study, the researcher hopes to understand the relationship, if any according to counsellors, between individual self-concept and marital satisfaction in suicidal women and their husbands. Self-Concept and the Suicidal Person Wilson and his colleagues (1971) used the Miskimins Self-Goal-Other Discrepancy Scale (MSGO) to measure disparities between a person's self-concept, goal self-concept, and perceptions about the evaluations of others. Sixty-six subjects were divided into three groups 22 and were assessed on the MSGO scale: severe suicide attempters, psychiatric controls, and normal controls. Suicide attempters who were recruited from the hospital represented 22 of the total sample and were found to exhibit the highest total tension, highest anxiety, and highest levels of depression when compared with either of the other groups. Additionally, suicide attempters felt that others valued them more than they valued themselves. Contrary to the contextual, systemic framework for understanding suicidal behaviour set forth in the present study, Wilson and his colleagues found that suicidal behaviour was not influenced by the quality of interactions with others, but rather could be "based almost entirely" upon more internal sources of stress. Wenz (1976) studied the self-evaluations of three different comparison groups: those who were acutely suicidal; those considered to be low suicide-risk individuals; and non-suicidal individuals. Self-evaluation was assessed through the administration of a 10-item scale, and was defined as an attitude towards the self, either positive or negative. He found that the suicidal subjects, regardless of race, age, or gender, scored lower on self-evaluation than the non-suicidal subjects. Wenz generated some fruitful observations in his study, especially by highlighting the relationship between an individual's self-evaluation, and his or her attitudes, behaviours, and relationships to others. In particular, if the self-evaluations of suicidal persons are low, then it is likely that their attitudes, behaviours, and interactions with others will suffer accordingly, by being tainted by the same negative perceptions. Fitts (1965), would probably concur with establishing such a relationship, as he writes that, "Those people who see themselves as undesirable, worthless, or 'bad' tend to act accordingly" (p.1). Kaplan and Pokorny (1976) found that suicidal individuals exhibited a higher degree of self-rejecting attitudes than non-suicidal individuals. They studied the responses of junior high school students to a seven-item self-derogation scale and found that those students who scored higher on a measure of self-derogation (indicating greater negative attitudes about themselves) were more likely to experience suicidal ideation and suicidal behaviour, than those students who scored lower on the scale. Suter (1976) understood women's negative self-concepts in the context of the socialization process of men and women in our society, where traditionally masculine traits are typically valued over traditionally feminine traits. Suicidal behaviour in women, she contended, emerges out of women's sense of powerlessness and self-hatred, and can be seen as "...a way of punishing the self, expressing anger towards others, and asserting one's needs to others, albeit maladaptively" (p.140). The cognitive processes of suicidal women were considered by Neuringer and Lettieri (1982). Their discussion reported that suicidal women described experiencing less self-approval and less self-confidence than non-suicidal individuals. Schmidtke and Schaller (1989) hypothesized that suicidal individuals admitted to a psychiatric hospital would have more negative self-images based on their metaperspectives, or beliefs that these individuals perceived others to have of them, compared with non-suicidal patients. Suicidal subjects were made up of 28 men and 28 women who had been recently hospitalized, while the control group consisted of 52 non-suicidal men and 73 non-suicidal women. Subjects were assessed a week after admission, three weeks later, and one week prior to discharge. They found that suicidal women showed evidence of deteriorating self-concepts and metaperspectives overtime, while men showed an initial negative self-evaluation which quickly improved over time with treatment. The authors speculated that the deterioration of women's self-concepts over time may have been related to the tension these women experienced during the course of treatment, when they were forced to confront their own issues of personal responsibility. Self-Concept and the Spouse of the Suicidal Person Tabachnick (1961) described the type of individual who found himself or herself significantly involved with a suicidal person. He suggested that the non-suicidal partner was perhaps equally "dependent and masochistic", and that the role of the significant other was crucial in the development of suicidal behavior in the spouse. Further, he suggested that quite often both partners were suicidal, where, "Often, we infer a kind of race to see who will use the weapon of suicide first "(p. 18). Using the Sixteen Personality Factor Questionnaire (16PF), Bhagat (1976) investigated the following groups: (a) the personality traits of 20 men and 30 women who had attempted suicide were compared with a control group of 50 non-suicidal subjects, (b) the personality traits of the spouses of suicidal persons were compared with a control group, and (c) the personality traits of the couples in the attempted suicide group were compared with controls. Bhagat found that the husbands of suicidal wives did not appear different from control husbands or male attempters. On the other hand, his work did reveal that husbands of wives who attempted suicide appeared to share similar personality characteristics with their suicidal wives, including being mistrusting and anxious, which were traits that could have contributed to the emergence of conflict in these couples' marital relationships. Richman (1978) used the concept of a "disturbed symbiosis" to characterize the relationships that suicidal individuals were involved in at the time of their attempts. Such relationships were usually characterized by a lack of differentiation on the part of individuals, where there was no clear sense of autonomy or uniqueness for either person. Richman argued that disturbed symbiotic relationships were implicated in most suicide attempts. Richman defined symbiosis as follows: "...the term symbiosis refers to a certain kind of relationship in which the uniqueness or individuality of one member is seen as a threat and is, therefore, denied or disconfirmed" (p.141). According to Richman, when one spouse was suicidal, it was often the case that both partners were suffering from having dysfunctional symbiotic needs, and it was therefore likely that they would each lack a clear sense of self. Canetto (1988) adopted a "multi-level" approach to understanding suicidal women and their husbands that incorporated three levels of personality experience: behavioural, self-report, and intrapsychic. She was interested in exploring the object-relation dynamics in suicidal women and their spouses at two different stages of adult development, from these three perspectives. She found that at an overt, conscious level, non-suicidal husbands appeared to be asymptomatic and healthier than their wives, while at this level their suicidal wives exhibited more pathology and distress. Interestingly, Canetto's work found that at an unconscious, covert level (accessed through the Rorschach test) non-suicidal husbands and suicidal wives began to look more alike, with the suicidal women even appearing more intrapsychically healthy than their non-suicidal husbands. Interpersonal Factors While there has been little investigation into the precise nature of the interpersonal qualities of the intimate relationships of suicidal women, in particular the nature and dimensions of the conflicts and the interactive facets of the relationships, several writers have acknowledged the presence of interpersonal distress and marital discord in the lives of suicide attempters, particularly women ( Adam et al., 1978; Beck, Lester, & Kovacs, 1973; Fieldsend & Lowenstein, 1981, Hart & Williams, 1987; Lester, 1983; Maris, 1981; Tabachnick, 1961). In the present study, the qualities of the intimate relationships of suicidal women will be explored by asking counsellors who have recently worked with married suicidal women what their perceptions of these marital relationships are. An objective of this study is to understand the nature of suicidal women's marital relationships, using counsellors' understanding of their suicidal clients' marital relationships as a primary source of this information. Data sources will include: those clinical records kept by counsellors who saw suicidal married women in the context of counselling relationships, verbatim transcripts of interviews with counsellors who worked with a couple following the wife's suicide attempt, and a psychiatric report that was written following a married suicidal woman's admission to the hospital. Interpersonal Distress and Suicidal Behaviour Tabachnick's (1961) work on the self-concept of the suicidal spouse has already been discussed. Additionally, he found that the relationships of suicidal individuals were characterized by dissatisfaction and unhappiness for both partners. He suggested that these individuals stayed together despite the fact that their needs were not gratified, because the relationship served to confirm the sado-masochistic tendencies that each partner brought to the union. 28 Two hundred and forty patients admitted to the hospital following a suicide attempt were studied by Beck, Lester, and Kovacs (1973). Patients were administered the Beck Depression Inventory and a generalized expectancies scale that measured optimism-pessimism. Their results suggested that women's attempts were more likely to occur in the context of severe interpersonal friction, whereas men's attempts usually resulted from perceived performance failure such as job loss or failed goal attainment. Adam and his co-authors (1978) based their findings on the interviews of 195 patients who were admitted to the hospital following a suicide attempt. Regarding these patients' current relationships, the researchers found that over 80% of patients described themselves as having major difficulties in this area. Very few expressed any optimism over the future of their relationships, and in three-quarters of the sample, the dissolution or threatened disruption of the relationships lead to the suicide attempt in the first place. Fieldsend and Lowenstein (1981) interviewed the friends and family members of suicide attempters following their admission to hospital. All patients in this study had attempted suicide by trying to poison themselves by overdose. The patients, and those close to them, were interviewed shortly after their admission. The authors found that women's suicide attempts were most likely to have occurred following a quarrel with a significant other in the two days preceding the self-poisoning episode. Fieldsend and Lowenstein suggested that hostility was a key component in self-harm. They argued that women might more often resort to self-poisoning as a way of expressing their hostilities, while men might act out their anger in more externally directed ways, which could have accounted for the sex difference in self- poisoning attempts which they found to be four to one. Both Maris (1981) and Lester (1983) confirmed the findings of other authors in accounting for the differences in men's and women's suicidal behaviour. Consistent with other findings, they asserted that women's attempts were more often the result of interpersonal distress, while men's suicidal behaviour usually followed from more performance related losses. In a study of the interpersonal networks of suicidal patients, Hart and Williams (1987) found that suicidal individuals could be described as having interpersonal relationship deficits. They also found that the subjects themselves perceived their relationships to be unsatisfactory. The deficits and the corresponding feelings of unhappiness experienced by these individuals in their intimate relationships, were seen as factors leading to the attempt. Suicidal Women's Intimate Relationships Some very important findings about suicidal women's marital relationships have been generated by gathering women's impressions and recollections through personal accounts and interviews following a suicide attempt (Pommereau & Penoutl, 1987; Stephens, 1985, 1988). These studies have been extremely valuable in characterizing the specific dimensions of suicidal women's interpersonal relationships. Pommereau and Penoutl's (1987) work confirmed Stephens' (1985) findings. Both studies found that suicidal women's relationships were typically characterized by one of three themes: infidelity, battering, and denial of affection. In addition, Stephens found a fourth theme of "smothering love", to describe how the suicidal women had very unrealistic expectations of their men to meet their every emotional need. Stephens (1988) suggested that the most common theme characterizing suicidal women's relationships with men was their experience of being involved with uncaring, emotionally indifferent partners. These studies are unique in their focus on the specific nature of suicidal women's intimate relationships, and their contributions should not be underestimated; however, they appear limited in at least one major way. They did not take the husbands' perceptions into account. Their findings, therefore, appear one-sided and over-simplified as they neglected to understand the women's suicidal behaviour as it emerged out of a complex relationship context. Suicide Attempters and Their Spouses Only a handful of studies (Bhagat, 1976, 1977; Bonnar & McGee, 1977; Canetto, 1986; Kumler, 1964) focused exclusively on the intimate relationships of suicide attempters through the direct measurement of both partners' perceptions. Kumler (1964) explored the themes found in the verbal and non-verbal communications between suicide attempters and the significant others in their lives, through an unstructured interview format. She found that suicide attempters usually experienced feelings of isolation from others, they experienced prolonged anxiety, and became engaged in interactions with others that could be characterized by incompatible demands and reciprocal frustrations. Usually the attempt followed a series of mutual rejections between the attempter and a significant other, which served to exacerbate the attempter's feelings of isolation and anxiety. Bhagat's (1976) work found that suicidal women and their husbands shared certain common personality traits, including being mistrusting and anxious. He suggested that the interactions between these spouses, based on these traits, could contribute to conflicts in their marriages. In his later work, Bhagat (1977) elaborated on his original study, and included other dimensions for assessing suicidal individuals and their spouses. He studied the marital interactions of these couples and constructed an instrument to measure each partner's perception of the marital interaction, called the Interpersonal Relationships Questionnaire. He found that couples where a spouse had attempted suicide had more difficulties than control couples in the following areas: communication, trust, dominance-submission and the expression of hostility. Bonnar and McGee (1977), in their study of communication among couples with suicidal wives, published their results based on a small 3 2 number of subjects. They investigated the quality of marital communication of four different types of couples: (a) couples where wives had made a suicide attempt, (b) couples where wives had threatened suicide, (c) non-suicidal couples experiencing marital difficulties, and (d) couples where the wives worked at the Crisis Centre. The maximum number of couples per group was six, while the second group of couples, with wives who had threatened suicide, was comprised of only four couples. They found that as the suicidal behaviour intensified, the quality of marital communication deteriorated. Consistent with Bonnar and McGee's findings, Goldberg and Mudd (1968) also suggested that suicidal behaviour was in itself a transaction or communication between husband and wife. In describing the angry and guilty reactions of a spouse to a suicidal person, the authors explained that, "What these reactions have in common is that they indicate the great likelihood of a confused and deteriorated relationship between husband and wife, in which communication and perception have been lacking or badly distorted" (p.352). Canetto (1988) investigated the couple dynamics of suicidal women and their husbands. A segment of her study assessed the marital communication and marital strengths/problems of these couples. She found that suicidal women were less controlling of their partner's autonomy than their spouses were of theirs, and were more likely than the men to complain of insufficient closeness in the relationships. Suicidal women were more likely than men to view their marital difficulties as interactive in nature. Other noteworthy trends that were not statistically significant included: suicidal women cited more marital problems as well as strengths than men, and appeared somewhat less critical than their mates. In a study that investigated the emotional responses of husbands to their wives' suicide attempts, Bennett (1979) found that husbands of suicidal wives could be characterized as more outwardly hostile, more ambivalent in their hostility, and less affectionate than the control husbands. He suggested that the variables were descriptive of f: .conditions existing in the person or relationship context prior to the -suicidal crisis, and were not merely artifacts of the crisis. Methodological Issues Recruiting suicidal individuals to be research subjects is a challenging task in itself, but when the goal of the study is to include information from the spouses of the suicide attempters as well, the process of recruitment becomes even more daunting. Using a population such as this for research purposes creates some unique challenges for researchers that are best handled through a combination of foresight, creativity, and a willingness to be flexible. The present study, like others before it (Bonnar & McGee, 1977; Canetto, 1988) was hindered by subject recruitment difficulties. In response to such inherent difficulties, the research design was changed to obviate the problems inevitably created by a sample size that was too 3 4 small. A full description of the specifics of the original design and the methodological concerns that arose out of it can be found in Appendix A. In their study of the marital communication of suicide attempters, Bonnar and McGee (1977) acknowledged their difficulties in attaining a larger sample and explained that women who had threatened suicide were not willing to have the researchers contact them and their husbands at home, as they did not want their husbands to know they had called in crisis. The researchers recognized the problem this created in terms of getting a representative sample, and had to concede that," Therefore, it must be concluded those couples who did participate in the study probably have better marital communication than the general population of threatener couples" (p. 15). Canetto (1988) experienced similar problems in her study. She identified a total of 50 couples as appropriate for her study but was only able to recruit 18 of those couples as participants. She wrote, "...one of the more-difficult-than-expected methodological problems was subject recruitment" (p.104). She acknowledged that any findings she offered were compromised by the small sample size. In fact, several of the other studies cited above had to settle for using sample sizes of under 30 subjects, with most of these suicidal individuals being recruited from the hospitals (Wilson, et al., 1971; Bhagat, 1976; Schmidtke & Schaller, 1989). It seems clear that recruiting suicidal subjects is a difficult task, especially when subjects do not represent a captive population such as hospital admissions. As they were being recruited from a counselling agency, suicidal women in the original investigation were members of a clinical population, which in itself set up certain obstacles for the researcher. A delicate balance has to be struck between the needs of the researcher and the treatment needs of the client. What has become apparent is that conventional methodological approaches for studying human subjects, emerging out of the tradition of academic psychology, do not necessarily lend themselves well to studying phenomena in the clinical setting. Hayes (1983) suggested that this has had serious consequences for clinicians and researchers alike, including: driving a wedge between clinical practice and research, cutting the practitioner off from input into the field's knowledge base, and producing research that is not always applicable to the clinical setting. In an effort to understand married women's suicidal behaviour more holistically and in the hopes of contributing some new knowledge to practicing clinicians working with suicidal women, the researcher originally sought out information from husbands and suicidal wives about the nature of their self-concepts and marital relationships, through standardized instruments. The change in research design, which shifted from a quantitative investigation that relied on the self-report data of husbands and wives, to a qualitative investigation that relied on the perceptions of counsellors to describe the nature of suicidal women and their spouses, necessitated a shift in perspective but did not change the overall goals or intent of the study. The aim of the study - to understand how counsellors make sense out of married women's suicidal behaviour - remains faithful to the original focus. Furthermore, an unexpected benefit arising out of the methodological difficulties was the opportunity to capitalize on alternative data sources like counsellors' perceptions. In fact, such a new approach to understanding women's suicidal behaviour can be understood to be somewhat of a methodological innovation. Qualitative Research Two sources of data will be qualitatively analyzed for the purposes of this study: the clinical records documenting the impressions of clinicians working with married suicidal women, and the verbatim transcripts of interviews with counsellors who worked with suicidal women and their husbands. Using these sources of data as a method for investigation -clinical records and interviews with counsellors - raises some important considerations and each will be discussed in turn. In the present study, clinical records will be used to reveal the perspectives of counsellors about married suicidal women and their husbands. According to Bailey (1978) documents are often neglected by researchers as sources of valuable information. Bailey discusses the advantages of using documents in research, including the following: documents allow the researcher access to information on inaccessible subjects; the data collection process does not interfere with the data being collected; the approach lends itself well to qualitative analysis; and many written documents are of a very high quality and "...are written by skilled social commentators and may be much more valuable than, for example, poorly written responses to mailed questionnaires" (p.269). The primary disadvantage of using written documents as a data source has to do with the circumstances under which they were produced. What is the motivation and intent of the writer? One cannot be sure that the writer did not have ulterior motives for presenting the information in a particular light. According to Webb, Campbell, Schwartz, Sechrest and Grove (1981), by using records as a data source, one is often faced with substituting another person's "selective filter" for one's own. For the purposes here, one cannot be certain that counsellors' reports represent any objective truth about their clients. Some examples of factors that might interfere with or distort a counsellor's report-writing process include: counter-transference feelings towards the client, a desire to create a favourable impression with one's clinical supervisor, a fear that one's clinical report might become audited, poor time management, selective memory, and an overabundance of cases leading to counsellor fatigue or burnout. Despite such drawbacks, much comfort can be taken in the fact that this study does not seek to draw widely generalizable conclusions about clients through the written reports of counsellors, rather it seeks to find out just how some counsellors come to understand their married suicidal female clients and how they express this in their written reports and in their verbal descriptions. Qualitative or phenomenological methodology has an existential focus, and therefore what becomes important for study is not just what is known, but how things are known and how they might be expressed in words and behaviour (Thome, 1991). Bogdan and Biklen (1982) suggested that utilizing a combination of written records and interviews can produce some very revealing information for researchers. By analyzing clinical reports and by interviewing counsellors about their perceptions of married suicidal women, the researcher has an opportunity to check for consistency of themes and convergence amongst different counsellors. By interviewing counsellors about their perceptions, the researcher is placing the counsellor in a position of informant -"... a participant observer one selective screen away from the investigator" (Webb et al., 1981, p. 200). It Is felt that counsellors are reliable informants based on their training, experience, familiarity with their clients, ability to empathize with and articulate what the clients are feeling. A final point with regard to using written documents as a data source needs to be made. Counsellors writing their reports were unaware that their records would later become the subjects of psychological analysis, which guards against the reactive biases inherent in questionnaire studies. Furthermore, utilizing clinical records as a source of data provides the researcher with an opportunity to test her research hypotheses without the threats to validity that are common to more obtrusive measures (i.e. role selection, response sets, measurement as change agent). Through open-ended interviews with counsellors, the researcher allows the counsellors' experience of working with their clients to emerge through the counsellors' frames of reference. This is a process that allows the counsellors to reveal to the investigator their own experiences and the meaning that they have derived in working with these women, in the context of their therapeutic relationships. Declaration of Biases Prior to making the data collection and analysis procedures explicit, it is necessary to acknowledge the potential sources of bias that such an approach engenders. Firstly, the suicidal woman herself will be reporting on her current reality from her own subjective viewpoint at the time she presents herself for counselling. Secondly, the counsellor she sees will be operating from within his or her own clinical orientation which not only colours their interaction from the outset, but influences which of the client's presenting problems the counsellor will attend to most closely and later document. Thirdly, the researcher will approach the data from her own conceptually biased framework, based on her familiarity with the phenomenon and articulation of expected findings in the form of research questions. What is being studied here is not the subjective reality of the suicidal women as reported in her own language, but rather the counsellor's interpretation of what the client is experiencing based on the client's verbal and non-verbal presentation in the context of a counselling session. It is also important to declare at this point the particular biases of the investigator. Clearly, judging from the choice of topic alone, it is apparent that there is an expectation that women's marital relationships will occupy some role in the development of married women's suicidal behaviour. By adopting an ecosystemic perspective for conceptual purposes, there is a belief that multiple levels of the ecoysytem, including family of origin influences, will be operating in the development of suicidal behaviour in women. In the original design, the variables of self-concept and marital satisfaction were chosen for closer inspection for both suicidal women and their spouses. Inclusion of these specific dimensions in the study are evident in the direction that the research questions take and foretell a personal bias. Lastly, there is a personal bias of valuing the opinions and perceptions of counsellors as being legitimate and credible sources of information about suicidal women and their mates. Research Questions Miles and Huberman (1984) offer a model for conducting qualitative research that favors what they call, "front-end conceptualization." They believe that it is important to start with a conceptual framework that is general enough to allow for flexibility, and they recommend that investigators begin their studies with some initial research questions and key variables. They specifically caution that this does not mean investigators are to ignore data that contradicts original research questions. For the purposes of this study, the main global question guiding this study is: How do counsellors understand, experience, and describe married suicidal women and their spouses who have come to them for counselling? More specifically, the following list represents the particular research questions that gave the study its direction and focus: 1. How do counsellors understand their married female clients' suicidal behaviour? 2. In working with married suicidal women, what do counsellors perceive to be the most significant issues? 3. What are seen as precipitating factors to women's suicidal crises? 4. How are the spouses of suicidal married female clients perceived by counsellors? 5. How would counsellors describe the marital relationships of suicidal women and their spouses? What is most noteworthy, according to counsellors, about these relationships? 6. How would counsellors describe the family histories of suicidal married women and their spouses? 7. How do counsellors see their roles in working with suicidal women? 8. What are counsellors' clinical orientations for understanding married suicidal women? 9. How do counsellors develop treatment goals for their work with married suicidal women? 10. How would counsellors characterize the self-concepts of their suicidal married women clients and their spouses? 11. How would counsellors describe their approach to report-writing? 12. Do counsellors see any relationship between wives' and husbands' self-concepts, and their marital satisfaction? Methodology Preliminary Investigation The current study was originally conceptualized as a quantitative investigation of the dimensions along which couples with suicidal wives might have differed from couples with non-suicidal wives. Specifically, the researcher identified the self-concept and marital satisfaction scores of husbands and wives as units for study, as measured by the Tennessee Self-Concept Scale and the Marital Satisfaction Inventory, in order to compare different types of couples. Original Subject Group A sample size of 15 couples in each of the following three groups was proposed for the original study: couples where wives had made a suicide attempt, non-suicidal couples in marital counselling, and couples from the community. After eight months of gathering data, the researcher identified 12 couples with suicidal wives as appropriate, and all 12 verbally agreed to participate. Out of those 12 couples only three suicidal women and two of their spouses followed through by participating in the study . Couples in the other two groups, those in counselling and those from the community, began to make themselves available in the numbers more closely approximating those that were originally expected. For the purposes of the original investigation, only three suicidal women and two of their spouses, out of 12 couples that were identified as appropriate, comprised the subjects in the first group: couples with suicidal wives. All of these suicidal women were recruited from S.A.F.E.R. Counselling Agency, a counselling centre in Vancouver specifically established to deal with suicidal clients. Eight non-suicidal couples from marital counselling were recruited from 2 private practitioners and a counselling clinic in New Westminster and comprised the second group: couples in counselling. Nine additional couples were recruited from leisure programs offered through community centres throughout Vancouver and comprised the subjects for the third comparison group: couples in the community. For a detailed description of this portion of the investigation, including procedures, demographic information, and descriptive statistics on the TSCS and MSI for all husbands and wives who participated in the study, see Tables 1-4 in Appendix B. Subject Recruitment Difficulties The researcher contended that some of the reasons for the low return rate for subjects in the suicidal wives' group included: suicidal wives were in crisis and their motivation to fill out forms was low; husbands did not know that their wives were coming for counselling, so wives filled out the forms while their husbands did not; in some cases, the suicidal crisis precipitated a marital separation making it difficult to get responses from both spouses; and counsellors working with suicidal women felt protective of their clients and did not wish to pressure them into filling out forms while they were in precarious, often unsafe, marital relationships. 4 5 To address these issues the researcher expanded the field from which suicidal women might be recruited. Eight other agencies in the Lower Mainland servicing suicidal clients, in addition to S.A.F.E.R, were contacted about recruiting potential subjects. Unfortunately, this effort to include other agencies did not appear to offer much promise for accessing more suicidal women, as many of the staff of these agencies appeared reluctant to include their clients for participation. The study was also expanded to include women who threatened suicide, in addition to attempters. Despite such efforts to broaden the criteria for subject recruitment, no additional subjects made themselves available to the researcher. One can speculate that some of the difficulties encountered in recruiting subjects for the study may be linked to the research hypotheses. If these women were in relationships characterized by ineffective communication, poor conflict resolution, and unresolved hostilities, as predicted, it would not be surprising to discover that these couples would lack the necessary trust, commitment, or focused energy, required of research participants. Subject recruitment difficulties with suicidal subjects have proven to be an obstacle to other researchers in the past ( Bonnar & McGee, 1977; Canetto, 1988). The authors of both of these studies acknowledged that subject recruitment problems limited the size of their research samples, which in turn compromised the findings they offered. The desire to minimize the effects of such sampling limitations in the present study lead the researcher to consider alternative data sources. After repeated unsuccessful efforts to recruit suicidal women and their spouses for this study, it was necessary to identify other data sources that may have been able to offer some insight into the nature of suicidal women's intimate relationships. The process of changing the methodological approach in the midst of the project led to a serendipitous discovery - that counsellors' perceptions about suicidal women and their mates, offered through their clinical records and through follow-up interviews, offered a valuable data source for understanding the nature of married sucidal women. Counsellors who worked with suicidal women in a counselling setting were identified as valuable informants who could offer unique perspectives on the nature of suicidal women's marital relationships. While such a shift in emphasis allowed the researcher to retain the originally conceived focus on understanding suicidal women's marital relationships, the shift in perspective demanded a corresponding shift in research design. Bailey (1978) acknowledged that when a researcher undertakes a study that is of a personal or private nature, small samples are often used. He suggested that findings generated from studies of this nature lend themselves particularly well to qualitative analysis. Overview of the Qualitative Research Process 47 Qualitative research in the social sciences can be distinguished from more traditional quantitative approaches to studying human phenomena by its focus on context, and its understanding of human behaviour as it is embedded in a rich and complex setting. Qualitative researchers concern themselves with understanding processes, frames of reference, and the meaning ascribed to events by participants themselves (Marshall & Rossman, 1989). In selecting one particular qualitative research strategy for studying a unique aspect of human behaviour, Marshall and Rossman suggested that it was not necessary to use such a strategy to the exclusion of other non-qualitative techniques. Rather, they suggested, "The strategy is a road map, an overall plan for engaging the phenomenon of interest in systematic inquiry" (p.76). Such advocacy for using an array of relevant data sources gives support to the current study's use of clinical records, follow-up interviews with counsellors, and the original quantitative self-report data gathered for the original study. In addressing the issues of validity in qualitative research, Bailey (1978) suggested that records could be seen to have face validity if they represented the first-hand impressions of the person writing the document. Lather (1986) added that another way to achieve face validity is through the process of recycling the data back to the respondents for their feedback and confirmation. Reason and Rowan (1981) suggested that any good researcher "...goes back to the subject with the tentative results, and refines them in the light of the subjects' reactions" (p.248). 48 Bailey (1978) suggested that criterion validity could be achieved if the contents of the documents could be corroborated by other sources. By using the original quantitative self-report data and the follow-up interviews with counsellors, opportunities exist in the present study for conducting such validity checks. In order to achieve construct validity, Lather (1986) advocated that researchers adopt a position of "systematized reflexivity" in order to allow the original theoretical proposals to be changed by the logic of the data. She wrote that establishing construct validity in this way is crucial and "...will contribute to the growth of illuminating and change-enhancing social theory" (p.67). In the current study, rather than focusing exclusively on the self-report data of suicidal women and their husbands, based on their TSCS and MSI scores, the focus of the study shifted to a qualitative analysis of the perceptions of counsellors who worked with suicidal women and their spouses. Both the counsellors' clinical records and the content from interviews held with them were the subjects of qualitative analysis. One source of data for the qualitative analysis included the clinical reports of four different counsellors from S.A.F.E.R. Counselling Agency and one psychiatrist who worked with three married suicidal women and one spouse. Another data source was the verbatim transcripts from interviews with two counsellors who worked with a suicidal woman and her husband; each counsellor had worked separately with each spouse in individual sessions, as well as jointly, with the couple together. The reason that these two counsellors were chosen for interviews was because of their unique positions of having worked with both the suicidal woman and her husband in the context of individual sessions and jointly in marital sessions. Presumably, each counsellor was afforded the opportunity to develop an in depth understanding of his/her individual client as well as an appreciation for the couple's marital dynamics. Using counsellors' perceptions as a primary data source has proven to be a valuable methodological innovation that has not been capitalized on in previous research. One useful outcome of using such an approach is that an exploration of counsellors' clinical interventions in dealing with suicidal female clients becomes possible. Procedural Framework The content from counsellors' clinical records and the information generated from interviews with counsellors was subject to an initial coding procedure that sought to identify emerging themes. As patterns began to evolve, the researcher began grouping information together under more abstract and psychological categories. Specific strategies designed to display the data and to limit the likelihood of alternative explanations were applied to the data and will be described in detail in later sections. The model for analyzing both the clinical records and the verbatim transcripts was based on Miles and Huberman (1984) approach to qualitative analysis. These authors admit to favouring an orientation that is clearly conceptualized from the outset, starting with the following: a general conceptual framework, a set of research questions to guide the investigation, and a series of start-up codes for analyzing data as the investigation proceeds. Miles and Huberman conceive of qualitative analysis as made up of three concurrent streams of activity: data reduction, data display, and conclusion drawing/verifying. Each of these components forms an integral part of the overall analysis procedure, and each aspect will be described in turn. Following that discussion, the procedural approaches used in the present investigation will be discussed under each of the three headings. Data Reduction The purpose of this initial phase of the research is to bring all of the data together in an efficient and manageable way. Generally, the great volume of data that is generated by a qualitative study is more than most people can process all at once. Data reduction serves the purpose of synthesizing and drawing together the most salient aspects of the data, and discarding those aspects that are redundant, irrelevant, or beyond the scope of the study. Developing a conceptual framework from the outset and generating a list of start-up codes that reflects this framework, both help to keep the researcher focused and bring structure to the overall process. These early stages of data reduction provide the researcher with the necessary framework to become oriented to the data as it materializes. It is at this 51 time that the researcher starts to become familiar with the information that is being generated. Early themes begin to become evident at this stage in the process. The researcher starts to see patterns emerging among different respondents. Strategies of the researcher such as writing comments in the margins of the records or on the verbatim transcripts from interviews are known as reflective and marginal remarks. Both of these strategies allow the researcher to respond to the data personally. Noting reactions to what is being read, suggesting possibilities for later interpretations, and reflecting on possible meanings are all part of writing marginal or reflective remarks. Miles and Huberman (1984) comment on the value of such remarks, noting that "They suggest new interpretations, leads, connections with other parts of the data - and they usually point toward analytic work,... that leads further and further into analysis" (p. 65). Writing memos is another way for the researcher to respond to the data directly, but in a more conceptual way. Memoing is a way to connect isolated pieces of information in a more global manner. Memoing builds on the earlier work of marginal remarks and forces the researcher to think in more abstract, psychological terms. Data Display This aspect of the analysis serves the purpose of making the data more accessible all at once. By using creative methods to compile and display the findings the researcher can continue to see new themes emerge, and new categories for grouping items together can begin to take shape. Graphs, flow charts, tables, causal networks, and matrices are all examples of possible ways to graphically represent the data. The best data displays are those that manage to capture the richness and complexity of the phenomena being studied, while maintaining clarity and scope. Miles and Huberman (1984) wrote that," Valid analysis requires, and is driven by, displays that are as simultaneous as possible, are focused, and are as systematically arranged as the questions at hand demand" (p. 79). Drawing/Verifying Conclusions This part of the analysis is concerned with the process of how conclusions are developed. Qualitative researchers must be vigilant about guarding against bias, and Miles and Huberman (1984) offered several strategies for ensuring that conclusions are of a high quality and reflect sound decision-making processes. Included in their list of strategies were: counting, seeing plausibility, clustering, making metaphors, splitting variables, subsuming particulars into the general, factoring, noting relationships, and finding intervening variables. Each of these strategies are designed to force the researcher to make higher level abstractions, so that the data can be reported on in a way that encompasses the experiences of all of the informants. Specific strategies such as looking for negative evidence and ruling out rival explanations are both part of an overall goal of checking for 53 researcher effects. With these strategies, the researcher deliberately sets out to try and find documented instances that would contradict the purported conclusions, to be sure that such explanations can be ruled out and to ensure that personal bias is not guiding the conclusions. Application of Procedures Data Reduction This aspect of the analysis began with the development of a conceptual framework and the formulation of some initial research questions. The research questions used in this study are described at the end of Chapter 2. The clinical records of four counsellors were compiled and read with the research questions kept in mind to provide the structure for this initial data reduction phase. Following this review, three hours of open-ended, semi-structured interviews were conducted with the two counsellors who had worked with a suicidal women (Case 2) and her spouse (Case 3). The questions asked in the interviews were designed to address those issues that were highlighted in the initial formulation of the research questions. The questions were general enough to enable respondents the flexibility to guide much of the direction of the interview. Examples of questions included: how would you describe this particular married suicidal client? what was your experience working with this client? how have you come to understand her suicidal behaviour? how have you come to understand your client's relationship to his/her spouse? how 54 would you characterize this person's self-concept? how would you describe this person's family history? The contents of the clinical records of the four counsellors and one hospital psychiatrist were analyzed as they were written originally. Interviews were transcribed verbatim to facilitate coding and the search and generation of themes. The verbatim interview transcripts were subjected to the same coding and subsequent analysis procedures as the clinical records. This included an initial assignment of codes to discrete chunks of information, followed by a search for themes and patterns in order to begin making more global and generalized abstractions to capture the essence of the information contained in the reports and interviews. Start-up codes were derived directly from the research questions and conceptual framework. Each code was operationally defined and arose out of the overall conceptual structure. Many of the initial codes were changed or discarded along the way, with new ones added, as the process of analysis progressed. The following list includes the descriptions of some of the start-up codes used in this study (for a complete listing of all the codes used, see Appendix B, Table 5): counsellor opinion, framework for understanding suicidal behaviour, history of wife, history of husband, description of self-concept of wife, description of self-concept of husband, and description of marital relationship. 55 Reflective and marginal remarks were frequently written by the researcher throughout the coding procedure as a new awareness came to light and personal reactions to the material emerged. Some reflective remarks were written in response to curious or unique pieces of information in the data. Other reflective remarks were written simply as a way to stimulate further thought on a particular subject. For example, the following reflective remark was written by the researcher in response to a comment made by a counsellor that suggested the couple seemed unable to tolerate separateness, "what would it mean for this couple to separate?" Other remarks represented the early stages of later theme development, as the researcher noticed recurring ideas being expressed by the respondents. As themes began to appear, the researcher used pattern codes to identify where linkages were being reported. Pattern codes were applied to emerging themes and started to encompass more wide-ranging concepts by combining discrete pieces of information into more global abstractions. For example, the researcher was able to develop a list of clients' "rules to live by " that emerged from reading the counsellors' records, and represented a compilation of implicit and explicit expectations by which clients seemed to guide their lives by. Such rules included: "the only way to trust myself is when I am doing a perfect job," "do not let others see what is really going on inside," and "even though I am not getting my needs met here, I must remain stuck in this situation." Another technique used at this stage of data reduction was the use of memoing . Memos were written by the researcher during the initial stages of coding and continued throughout the analysis. This strategy helped the analyst to become oriented to the data in a more conceptual way by beginning to focus on the relationships between key factors. For example, a memo was used to initially link the concepts of depression and hopelessness under the more general heading of "trapped," as both of these attributes encompassed elements of being incapacitated. Memos were a way to document early impressions and they differed from remarks by suggesting more global concepts leading to higher level abstractions. Data Display Displaying the data in tables, matrices, and maps allows the researcher to view the data "at a glance" which makes it easier to draw conclusions or focus future efforts. For the purposes of displaying the data in this study, both a "site-by-attribute" matrix was constructed as well as a map of a causal network. A site-by-attribute matrix provided a way to represent the data in such a way that it became possible to differentiate between the individual sites (counsellors) on the bases of certain attributes. Groups of sites could then be combined together in families sharing the same qualities. For example, all counsellors who related that their clients called themselves stupid could be seen by looking down the column "self-concept" representing counsellors' descriptions of client self- concept (See Table 1). Table 1 Site by Attribute Matrix Descriptions Used bv Counsellors to Describe Clients counsellor early childhood self-concept marital relations process 1. p. -trauma -shy, -amazingly -concerned -abuse anxious non with violence -insecure communi- childhood -not safe -shame; low cative ;made background, -protected self-esteem assumptions abuse mom -panic -mutual history, -early dependence psychiatric attempt ; enmeshed history -bonded by -supportive early trauma counselling -sees "person-in-context" -outside referrals made; panic disorder 2. D. -abuse, pain -childlike -so little -immediacy; -survive by - feels communica- importance hiding "stupid" tion of -depression; -shame - therapeutic mood -devalued dependence relationship; swings -feels » support and -early loss of ineffective; can't get encourage-mom; early like a failure, needs met ment attempt -inadequacy -fear of -sees role of spouse's underlying anger depression -bonded by -communica-mutual tion skill abuse rehearsal -no intimacy; -outside no sexual referrals relationship Table 1 cont'd counsellor early childhood self-concept marital relations process -alcoholism -negative -physically -saw client 3. S. -mom was self-talk; battered by in terms of passive, inability to spouse "battered -never felt accept -"mind- woman special as a positive readers" syndrome" child feedback -can't talk -worked with -early -feels about internal attempt trapped feelings with dialogue -wonders if spouse -worked on she -can't see expressing deserves self leaving feelings; battering marriage validation of -calls self emotions "stupid" 4. J -alcoholism -negative -physically, -family of -child sexual view of self emotionally origin work; abuse -blames self battered genogram -parental for being -spouse tries - educated conflict battered to isolate about cycle - family -self- and control; of abuse; history of doubting - fearful of -confronted suicide high self- anger attempts to attempts criticism -intimacy; sexual minimize problems reassurance, validation -outside referrals A causal network was drawn representing counsellors' perceptions of the dynamic process and events involved in leading a married woman to experience a suicidal crisis. Feedback about the causal network was sought from the counsellors themselves and opportunities were made for them to make corrections if they felt adjustments had to be made (See Figure 1). PREDISPOSING FACTORS OS wife's family self-concept -abuse -neglect -alcoholic -mental ill husband's familv shame -low self-worth -depressed -abuse -mental ill, -neglect self-concept anxious -unworthy depressed negative Fiqure 1 PRECIPITATING FACTORS mamage drawn together by mutual abuse history -commun. problems —* -dependence -anger probs. -intimacy probs. stress/ trigger (work, $ probs. family conflict) 4, suicidal crisis RESPONSE help-seeking counselling issues(l-l) counselling could be^  too intense,leading ^ to feelings of failure OR could lead to new insights or coping -esteem, family hx, feelings, safety marital couns. -communication Drawing and Verifying Conclusions By noting recurrent patterns and themes, the process of drawing conclusions was already initiated. Clustering was a strategy that helped the analyst to actively refine the categories under which the data were placed previously. By clustering several recurrent themes together, the researcher was able to explore how counsellors' perceptions of suicidal women might be representative of more abstract and psychological concepts. The goal at this stage was not to "...search for the exhaustive and mutually exclusive categories of the statistician, but instead to identify the salient, grounded categories of meaning held by participants in the setting" ( Marshall & Rossman, 1989, p. 116) Applying metaphors to the findings was a rich and colourful approach to subsuming the particulars into the more general and abstract. In some cases the counsellors had already generated their own metaphors for describing their clients and their clinical work. In other cases the researcher used metaphors to capture an overall spirit that was communicated through the data. For example the metaphor for couples being "joined at the hip" seemed to capture the essence of their dependency and their apparent inability to tolerate separate identities. Deliberately setting out to look for negative evidence, checking out rival explanations, and ruling out spurious relations were all strategies designed to check the accuracy of the findings. In this case, alternative explanations were explored and mediating variables were also considered as possible ways to explain the data. Looking for evidence to suggest that women's suicidal behaviour was exclusively a condition of a biological, depression was one way the investigator sought out rival explanations. Cycling back the findings to the respondents themselves created another opportunity to check the validity of the findings. By allowing the counsellors themselves the chance to comment on the researcher's conceptualization of their experiences, an opportunity to include their revisions or feedback in the final report was created. In addition, the quantitative self-report data (TSCS and MSI scores) generated from the same four cases in the original study was also available to compare with this data on counsellors' perceptions. By documenting the sequence of the analysis procedures as well as recording all decision rules as the analysis proceeded the researcher was in a better position to defend the soundness of the methodology and the trustworthiness of the data. Furthermore, the likelihood that this study could be replicated by others was increased because of the systematic documentation of each phase of the research. A form for documenting each level of the analysis can be seen in Appendix B, Figure 2. Data Sources Counsellors For the purposes of the qualitative analysis, clinical records, representing the impressions of four counsellors from S.A.F.E.R. Counselling Agency and one hospital psychiatrist who worked with three suicidal woman and one husband in depth, were the subjects of analysis. Two follow-up interviews were conducted with two of the counsellors who had worked with a suicidal woman and her husband, both individually and as a couple. Each of these three suicidal women and the one husband participated in the original study, and therefore data was gathered from each of them in the form of the Tennessee Self Concept Scale and Marital Satisfaction Inventory scores. The four suicide intervention counsellors whose records were analyzed ranged in age from 29 - 52. The perceptions of four counsellors in total - three women and one man - were investigated through analyzing their clinical records, and in the case of two of the counsellors, through the process of follow-up interviews with the researcher. Their collective experience as suicide intervention counsellors spanned 24 years, with the least experienced counsellor having worked at S.A.F.E.R. for three years. Their educational experiences were predominantly at the graduate level, with three of the four counsellors having Masters' degrees, including one MSW and two MA degrees. The fourth counsellor had a BA in psychology with many years of formal training and experience. The clinical orientations of each of the suicide intervention counsellors could best be described as eclectic, with special emphasis on crisis intervention skills. The S.A.F.E.R. counsellors worked in a setting that was established to provide individual counselling and they would therefore be less likely to reflexively adopt a systemic orientation to understanding their clients than counsellors who worked in a family therapy setting. Counsellors did, however, admit to favouring certain theoretical approaches, including the following: psychodynamic therapy, gestalt therapy, process work, feminist therapy, bioenergetics, person-centred therapy, and family of origin work. Counsellor P: This counsellor was the only male counsellor and he was 38 years old. He held a Bachelor of Arts degree with a major in Psychology, and he also had a Master of Arts degree in Psychology. He had worked at S.A.F.E.R. for ten years and prior to that he had four years of related counselling experience. He describes his clinical orientation as a combination of family systems, psychodynamics, humanistic, Gestalt, and bioenergetics therapies. Counsellor D: This 41 year old woman worked as a counsellor at S.A.F.E.R. for eight years. Before she was employed at S.A.F.E.R., she worked at Planned Parenthood for one year. She worked as both a volunteer and trainer at the Vancouver Crisis Centre for three years. She held a Bachelor of Arts degree in Psychology, a Bachelor of Social Work degree and a Masters degree in Social Work. In addition, she had specialized training in sexual abuse and took a three-year training program in Gestalt therapy. Counsellor S: This woman worked at S.A.F.E.R. for just over three years and she was 52 years old. She held a Bachelor of Arts degree in Psychology and had several other training certificates in specialized areas including: sexual abuse counselling, working with battered women, group facilitation skills, family therapy, and personality disorders. In addition, she had three years of training in process oriented psychotherapy. Prior to her work at S.A.F.E.R. she worked as an adolescent suicide counsellor for two years, lead a group for survivors of sexual abuse for three years, and worked as a Child and Youth Care Counsellor for four years. Counsellor J: This woman also worked at S.A.F.E.R. for just over three years and she was 29 years old. She held a Bachelor of Arts degree in Psychology and had a Master of Arts degree in Counselling Psychology. Her orientation was a combination of family of origin therapy, feminist therapy, Gestalt therapy, and person-centred therapy. Prior to being employed at S.A.F.E.R. this woman worked for three years as a Child and Youth Care Counsellor, and she worked for one year as a family support worker. Presumably, the 4 records chosen for study document each counsellor's typical orientation to understanding clients' presenting problems and therefore the records will likely reflect a range of clinical opinions. This in itself increases the strength of the findings, for if 4 different counsellors report that their suicidal clients' experiences consistently manifest themselves in certain ways, then the likelihood of this being true of suicidal women in general is increased. Description of Client Group The three suicidal women about whom these counsellors were reporting were aged 20, 31, and 41. The one husband who was reported 65 on was married to the 31 year old suicidal woman, and he himself was 34 years old. Two of the women were Caucasian, while the youngest woman was a Native Indian. The two older women were both married for over 10 years, while the youngest woman was in a common-law relationship for just over a year. Both the middle woman and the youngest woman made drug overdose attempts, while the oldest woman came to S.A.F.E.R. threatening suicide. A more detailed description of each of the cases at the time of referral follows. Case 1: This 41 year old woman (V.) came to S.A.F.E.R. after being referred by an employee assistance counsellor from another agency where V. was being seen with her spouse. V. was feeling suicidal at the time of her referral and reported feeling depressed and "dead." V. was being battered by her spouse but did not feel she was able to share this information in the sessions with the other counsellor where this couple was being seen jointly. V. had made previous suicide attempts in the past as a teenager and as a young adult. According to the counsellor's record there did not appear to be any specific precipitator leading up to this particular suicidal crisis. This couple has been married for 20 years, and they have two adolescent children. V.'s 44 year old husband was described in the report as an alcoholic and recently began going to A.A. The main problems identified by the counsellor at the time of V.'s referral were: "battered woman syndrome," feelings of being trapped, low self-esteem, and an inability to feel. Counsellor initiated individual counselling with V. on a once a week basis. Case 2: This 31 year old woman (A.) was referred to S.A.F.E.R. by her family physician following A.'s suicide attempt by overdosing on extra-strength Tylenol and other medication two weeks earlier. A. had checked into a motel to take the overdose of medication and was subsequently found by police and taken to the hospital where she was kept for two weeks. The counsellor reported in her clinical record that the suicide attempt arose after "long-term frustration and ongoing difficulties with husband." A. and her husband have been married for 12 years and they have two pre-adolescent children. The problems that were identified by the counsellor at the time of referral were: poor communication with husband, low self-esteem, and an inability to express feelings. The counsellor agreed to see A. for individual counselling on a weekly basis, and a plan to see the couple for joint sessions was also formulated at the time of initial referral. Case 3: This client (E.) is the spouse of A., above. He is a 34 year old man and became involved with S.A.F.E.R. following his wife's suicide attempt by overdose of medication. E. initially joined his wife for a joint session following her suicide attempt and the two counsellors who were seeing this couple together felt that E. could benefit from some individual sessions as well. According to this counsellor's report, E. became very frightened by his wife's suicide attempt and realized he could lose his relationship with his wife and family. This, according to the counsellor, was what motivated E. to work on his own problems. The problems that were identified by this counsellor at the time of referral included: speech problem, shyness and social isolation, and job frustration. It was decided that this client would be seen individually every two weeks, and on alternating weeks, as a couple. Case 4: This 20 year old Native Indian woman (W.) was referred to S.A.F.E.R. by the hospital following an overdose of Tylenol. She had been fighting with her common-law spouse prior to her attempt, and according to this counsellor's record, this was the precipitator in W.'s attempt. W. and her spouse have been living together for just over a year. They do not have any children. W.'s spouse has been physically abusive to her in the past and he was identified in the report as a man with a drinking problem. He is non-Native and is several years older than W. The identified problems in this case included: relationship conflict with common-law spouse, being battered, and a sexual abuse survivor. Plans were made to see her for counselling once a week. Results Findings What is presented here are the findings that emerged from the analytical processes described in the previous chapters. Counsellors' perceptions of married suicidal women and their spouses were qualitatively analyzed through the specific techniques articulated in chapter three under the headings of data reduction, data display, and drawing/verifying conclusions. A brief summary of the overall qualitative process follows. All of the information contained in the clinical records and verbatim transcripts was coded and examined for the emergence of patterns. Recurrent patterns were grouped together under more global and psychological themes. The data was then closely examined and subject to specific scrutiny designed to minimize the likelihood of alternative interpretations. What is reported here are the final results of that coding and synthesizing process. Findings will be presented in the form of dominant themes in each of six categories. Themes about the nature of clients' experiences as reported to their counsellors emerged out of six different categories which can be described as follows: characteristics of wives only, characteristics of husbands only, characteristics of husbands and wives, characteristics of relationships, characteristics of husbands' and wives' families of origin, and characteristics of counselling perspectives and strategies. Characteristics of Wives Only "Overfunctioning." Based on counsellors' records and interviews, all of the wives could be described as overfunctioning or self-defeating in terms of the inordinate amount of responsibility each took for the quality of their relationships. Each woman had very high control needs, a desire to be "perfect", and an inability to tolerate her own limitations. One counsellor commented in an interview on her client's experience,"... the only way that she can trust herself is when she feels she is doing a perfect job and the harshly critical judgemental side,... that nothing but perfect is good enough." One woman took on the responsibility, and thus the anxiety, for resolving the conflict between her spouse and her brother, despite the fact that the conflict had nothing to do with her. In describing one suicidal woman during an interview, another counsellor commented, "I think she has a very high level of expectation that she should be self-sufficient, that she shouldn't have to rely on anyone else, that she should be able to do these things all by herself." One counsellor observed historical evidence of her client's tendency to take responsibility for others by suggesting in her report that, "She carried mom's feelings." Women often took responsibility for the abuse that was happening to them, and tended to minimize any current or past abuse by dismissing it as "not that bad" or by describing their abusers as "only [italics added] verbally and emotionally abusive." According to a counsellor's record, another woman felt "...like she was going crazy and probably made it all up." "Intimidation and isolation." Two of the suicidal women were being physically battered, and all three of them expressed fear of their husband's volatile tempers. One woman was afraid to tell the marriage counsellor, in the presence of her husband, that she was being physically battered. Another woman kept many of her feelings to herself for fear of her spouse's angry response. In describing the events that precipitated one woman's suicide attempt, her counsellor wrote in her report, "This responsibility [for looking after finances] and her constant fear that her husband was going to find out and 'blow his top' was the last straw precipitating her suicide attempt." Related to this fear of their spouses' anger was the fact that these women experienced isolation and loneliness, and generally felt alienated from other potential sources of support in their own communities. In one assessment report, a psychiatrist noted, "She describes herself as pretty lonely and she cannot name anyone that she feels close to." Another counsellor wrote, "She admits that he does try to control her and isolate her and she understands this as part of the cycle of abuse." 7 1 "Commitment at any cost" Despite the psychological damage, and in some cases overt abuse that these women experienced in their relationships, they all expressed to their counsellors an explicit desire to stay with their mates and make their relationships work. According to one counsellor's record, one woman who was being physically and emotionally battered by her spouse conveyed to her counsellor that she "... cares about him a great deal and wants to make their relationship work." Another counsellor wrote in her report about her physically abused client, "She can't see herself separating from her husband. She has a genuine desire to make 'a go' of the marriage." Another dimension common to the counsellors' perceptions of each of the suicidal women is that each woman appeared to use her counselling session to rehearse how to talk to her husband. This focus on improving communication skills in the relationships appeared related to these women's desires and efforts to save their marriages. Characteristics of Husbands Only "Alcoholism" Two of the three husbands had definite problems with alcohol according to their wives who were being seen by S.A.F.E. R. counsellors. Both of these men made efforts to control their drinking by promising to stop or by joining A.A. during the time that their wives were being seen at S.A.F.E.R. for counselling. One counsellor documented in her report that her client,"... believes that her spouse has a drinking problem and when he becomes drunk he can be quite verbally abusive." Later in the clinical record this same client was reported to have indicated to her counsellor that"... he appears to be 'really trying', e.g. not going out and getting drunk with friends." Another counsellor wrote about her client's spouse, "Husband is an alcoholic. He has gone to A.A. one week ago." Not all of the husbands married to suicidal women abused alcohol. The husband who was being seen for individual and marital counselling at S.A.F.E.R. did not have a history of alcoholism, nor was it a current problem for him. "Raging temper." As described earlier, each husband was described as having a frightening and volatile temper. A counsellor wrote about the time her client called her in crisis, describing a fight she had had with her spouse; "He became so angry at one point that he ended up smashing their answering machine that she had bought with her own money a few days earlier. She found him to be verbally abusive and he said many cruel things to her, i.e., said she was a 'dumb Indian' and said his 'friends didn't like Indians'... He physically intervened to make sure she couldn't use the phone." In describing the husband of one suicidal woman, another counsellor wrote, "...he is terrified of anger and withdraws rather than express what he is feeling - until he explodes...his anger is scary - to himself and to his wife - in part because it feels so controlled - 'and what would happen if he lost it?'" Another counsellor wrote about how her battered client felt so much more safe and relaxed when her husband was not around, "Their best time is when her husband is out of town...She has discovered that she feels 'quite o.k' a lot of the time when her husband is away; and she becomes 'quite down' and despondent on the day he returns." Characteristics of Husbands and Wives Trapped-" All the wives and the one husband communicated sensations of feeling trapped to their counsellors. These experiences of being "stuck" incorporated feelings of helplessness, hopelessness, and perceptions of being incapacitated by intense feelings of depression. One counsellor wrote of her client, "When she thinks of going ahead she begins to feel suicidal." 7 4 In describing another client's inability to reach out for help prior to her suicide attempt, her counsellor said in an interview, "...she felt totally trapped and unable to tell anybody; she fully expected to be judged negatively by that and it was kind of a last straw and that was information that she couldn't share with anyone." In characterizing this particular client's husband, another counsellor said in an interview, "So I think he feels trapped and like he's a bit of a failure." What appeared to be a propensity towards depression also contributed to marital partners' experiences of being trapped. Husbands and wives presented symptoms of clinical depression such as flat affect ("feels dead"), sleep and appetite disturbances, and decreased energy levels. Documentation from a psychiatric assessment completed after one woman's entry into the hospital following an overdose supported a clinical diagnosis of a "...longstanding depressive illness of theoretically disabling intensity." Evidence gathered from counsellors' interviews and clinical reports suggested that husbands and wives had struggled off and on with depression and suicidal feelings for many years, with many of the clients having made previous suicide attempts in adolescence or early adulthood. "Never good enough." One husband in particular and all of the wives were both reported by counsellors to have extremely negative views of themselves. Repeatedly counsellors made mention of the fact that their clients referred to themselves as "stupid." This man and these women were highly critical of themselves and often blamed themselves for events beyond their immediate control. In describing the husband of one suicidal woman, a counsellor tried to paraphrase his client's implicit beliefs about the world in an interview, "... the world is not a safe place - it is not safe for me to be here - I'm only being tolerated here, so I shouldn't take up too much space, I shouldn't be too conspicuous, I shouldn't draw too much attention to myself." In describing her client's experience of being physically battered, another counsellor wrote in her report, "She had minimized the events and felt that she probably didn't deserve better." Another client was also being abused in her relationship and her counsellor wrote about her, "She realizes that many of the things he does are abusive and intended to hurt and yet she often blames herself for much of the escalating conflict as she attributes their fights to her 'insecurities'." Worth noting is the fact that one of the husbands who was not seen in counselling, but was referred to throughout a counsellor's clinical record, did not present as a man who suffered from a negative self-concept. Rather, according to his wife and described in writing by her counsellor, he was "super competent, macho, in-charge, and self-centred." This man was also physically and emotionally abusive to his wife and it could certainly be speculated that his "macho" presentation and abusive behaviour was simply a different manifestation of underlying self-doubt and an enduring negative self-image. "Help-seeking," Husbands and wives all demonstrated a willingness to seek outside help to deal with their difficulties. While only one husband participated directly in counselling at S.A.F.E.R., other husbands were reportedly engaged in other help-seeking activities such as Alcoholics Anonymous (A. A.), parenting courses, speech therapy, and marital counselling offered through another agency. In addition to being involved in counselling at S.A.F.E.R., the women were also involved in parenting courses and support groups. Characteristics of Marital Relationship "Communication problems." Each couple was described as having major problems with communication. In an interview, one counsellor describes the couple's relationship as "amazingly non-communicative." In another interview, another counsellor said, "I don't know that I have ever had [as a client] another person in a relationship where there has been so little communication between the couple as with this particular couple..." More specifically, according to counsellors' records and interviews, the domain of expressing feelings seemed to be the most troublesome aspect of communication for couples. One counsellor wrote of her client, " V. is struggling with allowing herself to feel in the presence of her husband." In identifying the presenting problems of her client, another counsellor wrote in her report," Poor communication with husband. Inability to express feelings." Regarding couples with a suicidal spouse in general, and her own client in particular, a counsellor commented in an interview, "...all of them have real problems communicating their needs and feelings to their partners." Another facet common to the communication in these relationships was the tendency for husbands and wives to make assumptions about their spouses without checking out for sure whether their assumptions were accurate. During an interview, a counsellor described this tendency towards "mind-reading" this way, "...they would make assumptions about each other that weren't necessarily true. Assumptions about what the other was feeling and they would act on those assumptions without ever checking them out and so their communication would get more and more skewed." Another variation on this "mind-reading" theme was revealed in another client's comment to her counsellor and documented in the clinical report, "She says he only needs to look at her and she knows she's out of line." Another characteristic that was common to these suicidal women in their relationships was their apparent inability to communicate to their spouses how much psychological pain they were feeling. This appeared to be related to the pressure these women put on themselves to be perfect and always in control. In describing the events leading up to her client's suicide attempt in an interview, one counsellor commented, "...there was this profound inability to tell him that once more she had made a mistake. To go and face him with that information was more than she could bear...I think there was a wish to be saved, but not the ability to come out and communicate what was going on herself." "Dependency." In an interview, a comment offered by one of the counsellors describing the relationship dynamics of a couple with a suicidal wife captured this theme of dependency with a powerful metaphor: "So they have bonded together in this sort of insular-type, symbiotic unit to keep the terrible, fearful world out. Like little kids hiding under the bed." In attempting to capture the essence of the "double-bind" of her client, the counsellor verbalized a sense of what she imagined her client's experience to be," I am absolutely dependent on this person, who I actually in truth don't think can meet my needs and I don't think anybody else can either, but this is the person that I have and I am stuck here and it isn't really going to do the trick for me." 79 Each of the spouses seemed to express to their counsellors a sense that as husbands and wives, they could not live without each other. There was a clear sense of mutual dependency that each had on the other, and neither spouse, despite clearly expressed dissatisfaction with aspects of the relationship, could imagine going ahead without the other. One husband of a suicidal wife was described in an interview by his counsellor, "...I think E. can't see a life after A., whether she left because she was dissatisfied or whether she committed suicide, he can't see a life. He describes that he would give up on everything, he would quit his job, he would become aimless, that he just would see no hope, his life would be over." Another way to characterize these relationships would be to describe these couples as enmeshed. According to the counsellors, it seemed particularly threatening for these marital partners to be clearly differentiated from one another. One counsellor verbalized it this way, "...there is a sense of them not being very separated and part of how they keep themselves not very separated is by not telling each other who they are, what they're feeling, so that they can make these assumptions..." There is also the suggestion by counsellors that these couples may be bonded together by their mutual abuse history. One counsellor offered an interpretation to this effect, "...both of them have come from traumatic backgrounds, and what seemed to draw them to each other was the similarities they found in their backgrounds with trauma." He went on to say, "They're very close in their dilemma with the world." "Intimacy problems." The couples, as described by counsellors, had intimacy problems that were characterized by sexual difficulties in the relationships, infidelity, jealousy, and the aforementioned inability to share and communicate personal details about themselves to one another. After meeting with a suicidal woman and her spouse for several sessions over a one-month period, the counsellor wrote in her report, "However it soon became apparent that this couple have not really been talking - or playing, or having sex, sharing, or showing affection to one another - for years." She went on to describe this couple's relationship in an interview, by saying, "I think their marriage is like a fortress, but is empty inside." Another counsellor working with another suicidal woman noted the intimacy problems being manifested in her client's sexual relationship, "He accuses her of wanting to have sex too often while it appears that he himself has difficulty with intimacy and experiences anxiety around their lovemaking. He also goes out of his way to make W. aware of his attraction to other women by staring at other women and appearing entranced by them." Intense feelings of jealousy were also evident in the relationships of these suicidal women which seemed to emerge out of both the women's and men's insecurities about themselves and their relationships. Characteristics of Husbands' and Wives' Families of Origin "Mental illness, alcoholism, and abuse." According to reports and interviews, there was evidence to suggest that husbands and wives grew up in families where one or more members could have been considered abusive and/or mentally ill, and/or alcoholic. All of the suicidal women and one husband described their fathers as alcoholics. In an interview, one counsellor said of her client's husband, "Also from the descriptions that E. gives of his father later on in life, I suspect that his father is mentally ill and was subject to real rages." In writing his assessment of one suicidal women, a psychiatrist noted, "Father had depressive illness and received psychiatric treatment including ECT." One woman was sexually abused as a child by a male cousin, another woman was emotionally neglected by her parents, and both the husband and wife representing another couple were viciously beaten in their childhoods. Additional comments made by counsellors reveal important characteristics about the family histories of the suicidal women and the men they married and are worth noting. One counsellor wrote of her suicidal client, "My client was definitely socialized to be 'quiet, nice, suffer silently, defer to the male, identify herself in terms of the male'. His socialization was that he could 'take' what he wanted, even if it was by violence'". Another counsellor wrote about his male client's family history, "Rigid roles, i.e., mother does the major portion of childcare; father distant from children, cold and non-affectionate...father is the major source of the physical abuse, mother seen as victim." Characteristics of Counselling Perspectives and Strategies "Ambivalence." Each of the counsellors noted the ambivalence present in the women's suicidal behaviour. One counsellor noted in her report that even though her client was feeling very depressed and suicidal, she did not have a plan and she was willing to come in for counselling. Another counsellor reported on the significance of her client's initiation of self-rescuing by changing her mind after she had ingested many pills and getting herself to the hospital. Following an initial interview with her client after her overdose attempt, another counsellor wrote in her report that her client stated that "...she was stupid to take the pills, does not want to die." "Educating about abuse." A significant part of the work with these women and with one husband was to educate these clients about the consequences of being a victim of abuse. This included confronting clients when they had a tendency to minimize what had happened, help them to develop plans for safety, and to begin to label what happened to them as abuse. One counsellor wrote in her report, "While he did not physically hurt her I labelled his behaviour as abusive and violent and expressed my concerns to her about her safety." Commenting on the fact that her client was married to a man who suffered from extreme physical brutality, which in turn lead her client to discount what was done to her, one counsellor said in an interview, "One of the things that has been very difficult is, because there was such horrendous physical violence in E.'s family of origin, she had compared her own upbringing to that and minimized what's been done to her on account of it. But I think in terms of the verbal abuse and the raging and the put-downs and the demands and the coldness, ...she has been terribly damaged..." "Personal empowerment." Clearly, each of the counsellors put a lot of effort into building a relationship with their clients that was based on trust and provided a clear support base to their clients. Skills such as empathy, validation, and affirmation, as well as the provision of support and encouragement were used often by counsellors, and instances of these skills were documented throughout their reports and in the interviews. A counsellor who had been seeing the husband of a suicidal woman commented on his client in an interview, "E, is a pretty fragile person and a lot of his needs are of course very basic..and so I saw him as needing supportive counselling; a validation of perhaps some of the origins of some of his feelings and how he has come to be the way that he is..." In helping her client to become more self-validating, another counsellor set a goal in her sessions to examine her client's negative self-talk and helped her client to explore different ways to nurture herself. In writing about a session with her client where the woman was expressing great self-loathing for nearly hitting her daughter, the counsellor responded by focusing on "...the fact that she had been able to leave, that she hadn't lost it and praised this..." In addressing the fact that her client was having doubts about her own sexuality, her counsellor wrote in her report, "I have reassured W. that there is nothing wrong with her sexual identity..." "Making links with the past." Counsellors worked to heighten their clients' awareness about their experiences in their families in order to help them make sense out of their current situations. One way in which they did this was by helping their clients to understand the nature of "projection". In an interview, one counsellor discussed the nature of a client's projection of her father's anger on to her spouse, "I think that she saw in E. a sort of father image and even though E. was not very often expressing his anger toward A. or being disapproving, any hint of that would really terrify A., which I think would remind her of her own father, who was very critical and demanding..." In doing some family of origin work with her counsellor's help, one client got in touch with the fact that she felt some anger towards her father, "I was never special to him!", she is quoted as saying by the counsellor in her report. Furthermore, this client realized that she never received the emotional nurturing that she deserved as her mom was 85 emotionally unavailable to her and was described in the clinical report as someone who "sat on the fence." One counsellor drew a genogram with her client where the client was able to notice a strong parallel in her relationship with her mate and previous generations, including her parents' relationship: an age difference of over 15 years, family resistance to the union, previous suicide attempts by women in the family, and both her spouse and her father had daughters from previous relationships. "Coaching." The counsellors spent considerable time helping their clients to develop more effective and constructive patterns of communication and more authentic ways of expressing their feelings. The acknowledgement and expression of anger for the women was an area where counsellors would provide guidance and actively teach certain skills. During an interview, one counsellor commented on how she understood her client to be using their sessions, "My sense is that she uses her sessions with me to rehearse what she wants to say to him..." An assignment given to one couple to enhance their communication was described in this counsellor's report, "I've asked that they play at "non-assuming" - checking out the other person's feelings, thoughts, wishes.etc, rather than acting on assumption - but set it up as a low-key, non-threatening exercise." "Outside referrals." All of the counsellors seemed to favour the use of outside resources to complement the work they were doing with their clients. Among the outside referrals made by counsellors for S.A.F.E.R. clients were: speech therapy, shyness group, parenting courses, battered women support group, private psychiatrist, and Adult Children of Alcoholics (A.C.O.A.) self-help groups. Areas of Counsellor Disagreement As already noted, the primary area where counsellors disagreed was in their perceptions of husbands, specifically whether or not husbands of suicidal wives could be characterized as suffering from negative self-concepts. An explanation for this that has already been offered suggests that perhaps those two husbands who did not "present" to their wives as suffering from negative beliefs about themselves may in fact have suffered from low self-worth but they manifested it in different ways such as battering. Another related area where counsellors disagreed was in their overall descriptions of husbands. One counsellor who worked with the husband of a suicidal woman described his client as passive - "lets the wife make the major decisions for the family." The other counsellors were more likely to describe the husbands of their suicidal clients as overbearing, controlling, and dominating. These differences could be functions of individual orientations among counsellors or may reflect actual differences among husbands of suicidal wives. The underlying conditions and historical factors leading to the range of behavioural descriptions of husbands may also reveal important similarites among husbands and this would be a worthwhile area to explore in future studies. Implications The lives of these clients as described by their counsellors can provide the substance for some useful insights. The findings can serve as a stimulation point for futher investigation into the nature of the complex dynamics experienced by suicidal women in the context of their own intrapsychic conditions, their family and social histories, their marital dynamics, and external environments. The process within which counsellors interacted with their clients also provides some valuable information about how suicidal women and their spouses might be understood in a therapeutic context and what the course of their treatment might entail. A discussion of how all of these themes might be understood, and how researchers and counsellors might be able to capitalize and build on these findings is presented in the following chapter. Discussion The findings presented in the previous chapter represent a preliminary exploration of counsellors' perceptions of married suicidal women. They are in no way to be considered exhaustive of the field under study; rather, it is hoped that they can point the way for future research. The unique evolution that this particular investigation underwent, from its beginnings as a causal-comparative quantitative study to its later form as a qualitative investigation of counsellors' perceptions, necessitates that attention be given to the overall research process, as part of the general discussion of results. Research Process The findings of this study, as they emerged in their final form, were based on data collected from the reports of four different suicide intervention counsellors and one hospital psychiatrist and information gathered through interviews with two of these counsellors. It is clear that information generated from such a small sample of respondents makes the interpretations less global than the researcher would have liked. On the other hand, it is believed that this study has served some very important functions: the research process can be used to heighten awareness about the specific challenges inherent in doing research with suicidal subjects; the study was unique in its efforts to acknowledge the impact of husbands' experiences in the evolution of suicidal behaviour in their wives; by exploring counsellors' perceptions of married suicidal women, this study served the useful purpose of offering a glimpse into the actual counselling processes with suicidal female clients, including a description of intervention strategies; and by having access to the original self-report data of husbands and suicidal wives who did participate in the original study, an opportunity exists to compare that data with the counsellors' perceptions, to note if there is any convergence. Summary of Results The profile that emerged of suicidal women and their spouses as described by counsellors who worked with them was of two individuals drawn together out of their mutual experiences of early trauma and abuse; these men and women simultaneously avoided revealing themselves to their spouses and almost swallowed the other up with their strong dependency needs; they feared both their own anger and that of their spouse; they had few skills of conflict resolution or anger management to bring to the marital relationship; they both felt a great deal of shame and personal failure which they were not able to share with anyone because that might mean further self-punishment; and they were willing to enlist the help of others to deal with their difficulties, but this too, had overtones of failure at times. Results generated from specific scales of the Tennessee Self-Concept Scale and the Marital Satisfaction Scale for each of the three wives and one husband can be seen in Table 2. The two TSCS scales included in Table 2 are a total positive score (Tot P), which is an overall self-esteem score and a self-criticism score (S-C). Included in Table 2 from the Marital Satisfaction Inventory is a conventionalization score (CNV), a global distress score (GDS), a problem-solving communication score (PSC), and a family history of distress score (FAM). The results generated from the MSI and TSCS for those couples with suicidal wives and those couples in counselling and couples in the community can be found in Appendix B, Tables 3 and 4. Table 2 Scores of Three Suicidal Women and One Husband on Selected Scales of the TSCS and the MSI Subject Scale TotP S-C CNV GDS PSC FAM Measure (TSCS) (TSCS) (MSI) (MSI) (MSI) (MSI) Wife (V) 289 42 2 27 32 11 Wife (W) 291 36 1 18 30 12 Wife (A) 251 47 0 35 35 15 Husb. (E) 298 30 16 6 8 14 Mean 345.6 35.5 2.6(W)a 28.2(W) 24.6(W) 8.4(W) 3.0(M)b 26.5(M) 25.2(M) 7.8(M) Note. The higher the value on the TSCS scale, the greater the feelings of self-worth, except for S-C, which is self-criticism. The higher the value on the MSI scales, the greater the perception of distress, except for CNV, which represents social desirability. a mean score of women in marital therapy D mean score of men in marital therapy 9 2 Discussion of Client Themes Self-concept, What was most compelling about the themes that emerged from the counsellors' descriptions of their clients was the fact that both the one husband and all three wives appeared to be suffering from quite profoundly negative views of themselves. They felt a great deal of personal shame, much of it seeming to be a consequence of the horrible abuses they had suffered as children. The fact that both the husband and wives suffered from negative views of self lends credibility to the findings of other authors (Tabachnick, 1961; Bhagat, 1976; Richman, 1978; Canetto, 1988) who suggested that one spouse's suicidal behaviour was inextricably linked to the emotional health of the other spouse. Counsellors working with suicidal women whose husbands were not being seen for counselling were less likely to perceive that these husbands were suffering from negative self-concepts, based on their presentations to their wives. What has already been suggested in chapter 4 however, is that both of the husbands who were not seen for counselling were physically and emotionally abusive to their wives, and thus could still be seen as suffering from negative self-images which they manifested in their violence and abuse to others. In examining the self-report data on self-concept for husbands and suicidal wives, as measured by the Tennessee Self-Concept Scale (see Appendix B, Table 4), it becomes clear that counsellors' descriptions of their clients accurately reflected the clients' own beliefs about themselves. Suicidal wives reported the lowest total positive scores of all three groups of wives and reported the highest levels of self-criticism. Husbands of suicidal wives, while not scoring the lowest on the total positive score, did score below the mean of 345.6, and scored higher than the husbands in counselling on self-criticism. One counsellor did speculate in an interview that perhaps the wife of his client was acting out her husband's suicide attempt, by "...acting out some of the helplessness that he felt." Such a systemic hypothesis would support Canetto's (1988) view that in couples with suicidal wives, what might be transpiring is a tendency for women to '...express and act out their own feelings of inadequacy and helplessness, as well as becoming the recipient of others' projected helplessness" (p.17). What both the qualitative and self-report data seemed to confirm was the fact that the married suicidal women in this study were not married to men with high levels of personal esteem and healthy levels of self -criticism. Of course, one cannot know for sure whether these husbands confidence levels have simply eroded in response to their wives' difficulties, and are not simply a reflection of these husbands' feelings of defeat at living with emotionally unwell wives. Such a hypothesis loses its plausibility however, when one examines the quality of the family of origin experiences of husbands, as reported by themselves and as described by counsellors. Family histories. Remarkably, all of the clients reported on by counsellors described their families as being characterized by at least one of the following: alcoholism, mental illness, and abuse. All of the fathers were reported to be alcoholics. It is believed that there is a significant link between these clients' early family experiences and the later development of negative self-concepts. (Stephens, 1977; Adam, 1990). When a child is beaten, neglected, and subject to parental instability, the opportunity for him or her to develop a healthy and positive view of self becomes severely compromised. Adam (1990) suggested that "Early psychosocial and environmental influences may act as primary predisposing factors leading to a specific vulnerability to suicidal behaviour, either alone or in association with other related personality variables, such as low self-esteem, a proneness to dysphoric states, poor impulse control, and interpersonal hostility" (p.74). Because this proneness towards developing an adult depression or another form of mental illness exists for children growing up in such households, it is not surprising that husbands and wives in this study were showing evidence of clinical depressions. Findings generated from the Marital Satisfaction Inventory, regarding respondents' family histories, are also shown in Appendix B, Table 3. Suicidal wives reported the greatest perceptions of distress in their own family backgrounds and husbands of suicidal wives reported the next greatest levels of distress in their own families, compared with all other husbands and wives. These findings give strong support to the findings of Stephens (1988) who found that the major themes characterizing the parent-child relationships of suicidal women were: non-nurturing parents, absent parents, abusive parents, mentally ill, and alcoholic parents. The findings from the present study go beyond Stephen's (1988) work to suggest that the husbands of suicidal women may also come from families characterized by neglect, abuse, and mental illness. A larger sample of suicidal women and their husbands is needed to give more support to this finding. This is believed to be a very important area for future exploration, and an area that has not received much attention in previous research. Counsellors offered some astute hunches about these family of origin dynamics, suggesting that husbands and wives were drawn together by their own mutual abuse histories in order to recreate some of the earlier traumas each had suffered. Richman (1986) wrote about this tendency for couples and families to systematically recreate early loss and trauma in each successive generation. He wrote, "There is also an often unspoken agreement for the marital relationship to be either unstable or else remote and lacking in intimacy, so that both spouses can maintain their loyalties to the earlier ties" (p.34). In their discussion of suicidal women's life experiences, Kaplan and Klein (1989) indicated that in the lives of these women "... the 96 internalization of earlier disturbances in parent-child relatedness can set the stage for later vulnerability to suicidal states of aloneness, particularly when they are reinforced by subsequent relational experiences" (p.275). Responsibility for the relationships. As indicated in counsellors' comments and reports, the suicidal women in this study were typically fearful of their spouses' anger, and took more than their share of the responsibility for the quality of the relationships. Both of these facets can be understood from a model of women's development called "action-in -relationship" (Kaplan & Klein, 1989). What such a model posits is that women come to value themselves based on their ability to be actively involved in mutually-defined, growth-enhancing, reciprocally-engaging relationships. They come to understand and appreciate themselves as interacting agents in relationships and to the extent that they can participate in these types of relational experiences, they will feel fulfilled, empowered, and valued. Such a model also suggests that when these opportunities for relating are thwarted, the women begin to feel like failures as they assume the bulk of the responsibility for the relationships, and it is at such times that they can become more vulnerable to suicide. In applying such a model to the suicidal women in this study, one can begin to see that these women, who were faced with external sources of stress and generally lacked other quality relationships in their lives, might have been feeling desperate to make some meaningful connections with their spouses. Despite the fact that all of them conveyed dissatisfaction with their marital relationships to their counsellors, they all expressed a clear desire to stay and try to make the relationships work, "at any cost." Failing to get their self-esteem needs met in their most significant relationships, by not being able to "make meaningful contact" with their spouses at a time when they were feeling particularly vulnerable, these women may have resorted to suicidal behaviour as a last desperate effort to engage their husbands. Dependency. This study found that counsellors described their married suicidal clients as being involved in relationships characterized by a high degree of dependency. Other writers have commented on the symbiotic (Richman, 1978; Aldridge, 1984), sado-masochistic (Tabachnick, 1961) and "smothering love" (Stephens, 1988) relationships of suicidal persons. In an interview, one counsellor described the extreme dependency that characterized the relationship of a couple with a suicidal wife. Each spouse had his or her own idiosyncratic fear, such as fear of the dark or fear of crowds, and together they acted in silent collaboration in ways that managed each other's fears and anxieties, "...they have learned to do all kinds of things for each other within the confines of the inside of their home...it's their secret that nobody else knows about and that's been a bedrock in their relationship and yet in an adult sense, they literally have not had sexual relations for years." These couples appeared to be "joined at the hip" in their own maladaptive functioning, and any move forward by one or the other would be perceived as a threat. The suicide attempt by the wife may have represented an effort to break the symbiotic stranglehold. According to Richman (1986) symbiosis and the fear of separation is a central issue in the development of suicidal behaviours in a couple or family. Whether partners merge with one another or detach from one another, the symbiosis fuels their behaviour and prevents them from relating to each other as two independent individuals. Richman's (1986) framework for understanding suicidal behaviour in this way appears contrary to the hypothesis that Kaplan and Klein (1989) put forth, and yet they are not incompatible. The latter authors made the argument that a woman's suicide attempt may represent an effort to join or make contact, whereas the former author suggested that suicidal behaviour is often an attempt to deal with the disturbed symbiotic ties that emerge from separation anxiety in the relationship. Perhaps the suicidal woman is struggling to find a balance for herself along the precarious continuum from individuation to togetherness, leading her to experience tremendous anxiety. This struggle can become so stressful, often triggering old separation anxiety, that she begins to feel vulnerable and self-doubting. In order to try and reaffirm herself she desperately tries to seek contact with her spouse, to join with him in a mutually defined relationship in order that she may feel valuable once again. If she perceives him as unavailable, she may feel like she has failed and may find herself contemplating suicide as a way to deal with her desperate feelings. Stephens (1988) described the punishing process that suicidal women experienced in their efforts to feel valued, "As their own self-worth existed only in terms of unremitting attention and devotion from their men, any flagging of attention was immediately interpreted by them as emotional abandonment and plunged them into despair" (p.81) Communication problems. Confirming the perceptions of counsellors, the scores from husbands and wives on the Marital Satisfaction Inventory (Appendix B, Table 3) illustrate that marital communication, particularly problem-solving communication, was perceived to be the worst by suicidal wives. Husbands of suicidal wives scored very high on the social desirability scale, the highest of any group, thus their mean scores on any of the scales have to be considered suspect. Given their family histories, it is likely that these husbands and wives learned to keep their feelings to themselves in order to avoid risking censure or judgement from abusive parents. Stephens (1988) wrote about the consequences to individuals growing up in such families, "They carried the multiple burdens of families whose internal organization was severely compromised and whose members were unable to relate to each other in satisfying, ego-enhancing ways" (p.79). It is a strong possibility that the husbands who scored so high on the social desirability scale were attempting to avoid a negative evaluation and therefore concealed their true feelings. The wives on the other hand were perhaps less encumbered by such self-conscious concerns as they had already made their feelings explicit through their suicidal behaviour and help-seeking. Discussion of Counselling Process Themes What the findings of this study indicated was that counsellors working with suicidal women focused on empowering their clients by: engaging them in an empathic and supportive relationship, helping clients to understand the effects of the past on their present experiences, teaching them specific skills such as assertive communication, expressing feelings, and asking for help, and by helping them to access other resources in the community. What these findings also reveal is that counsellors' perceptions are clearly emerging out of their own clinical orientations and experiential backgrounds. All of the counsellors who worked with the suicidal women were especially sensitive to the victimization of women in our society. The female counsellors all had received some type of specialized training in counselling women, and their perceptions about the suicidal women and their spouses must be understood in this light. Each counsellor understood that violence to women was a tragic byproduct of living in a sexist society and all of the counsellors could be described as favouring a feminist perspective for dealing with issues of abuse. The strategies that these counsellors adopted in working with suicidal women also revealed something about the clients' deficit areas. 101 By working towards empowerment of the clients and by helping clients develop specific communication skills it became evident that these suicidal women were struggling with diminshed self-esteem and poorly developed communication skills. Other writers (Richman, 1986; Kaplan & Klein, 1989) have expressed support for these clinical approaches when working with suicidal individuals and have offered other useful considerations. In establishing a relationship built on mutual trust and understanding, counsellors are in a position to invite their clients to "contribute empathically to the clinical process " (Kaplan & Klein, 1989) which leads to enhanced self-valuing feelings on the part of the client. Richman (1986) offered some useful suggestions to counsellors on how to reframe suicidal behaviour in the context of an interpersonal relationship. He suggested that the "... suicidal-type communications can be reinterpreted or relabelled as expressions of frustration, exasperation, helplessness, and caring; and the love behind the rage is given an opportunity to emerge" (p.81). Counsellors in this study appeared to appreciate the complexity of suicidal behaviour. They did not attribute their clients' suicide attempts to any one singular event and appeared to understand the behaviour in the overall interpersonal and social context. Commenting in an interview on his client's suicide attempt, a counsellor explained it this way, I think her suicide attempt was indirectly related to the relationship, but probably had a lot more to do with the inner workings within her, but in an indirect way I think she was afraid of her husband's disapproval, saw herself as a failure at work and in the family and then perhaps as a spouse and then probably feeling like she couldn't cope with the relationship and the idea of -"how could I cope with not being a part of that? if I can't cope with this, what can I cope with?" Based on the findings generated from this study, it was clear that counsellors tried to understand their clients' suicidal behaviour in all its complexity and they did not try to ascribe singular and simplistic motivations to the emergence of self-destructive behaviour. A noteworthy theme that emerged from counsellors' records and interviews held with them, was the impact that historical factors had on their client's present level of functioning. At the very least, the findings from this study have implications for our suicide prevention efforts with youth and families across the family life cycle. In retrospect, the suicidal women in this study each had histories that foretold of their vulnerability to developing later mental health problems, including childhood abuse, neglect, low self-esteem, adolescent suicidal behaviour, and their choice of unstable or abusive marital partners. It appears that the presence of certain family conditions like abuse, combined with an enduring individual sense of hopelessness and unworthiness could signal that a child in a family is at risk for developing a range of mental health problems leading up to and including suicide. If we could educate potential caregivers about these early signs of vulnerability, we could all begin to be more sensitive to such signals in our children and intervene much earlier to hopefully circumvent a suicidal outcome in later adulthood. Recommendations for Future Research Given the kinds of methodological issues that plagued this project from the beginning, it is felt that the issue of subject recruitment needs to be given special attention in future research endeavours with suicidal clients. One approach would be to recruit subjects while they are in the hospital following their admission there as a result of a self-injury attempt. This would allow the researcher access to a more captive population, increasing the likelihood of their cooperation, but would inevitably necessitate the cooperation of all emergency staff and psychiatric consultants. One drawback to this approach of course, would be the fact that the sample would be restricted to only those suicide attempters who sought or required medical attention. Another methodological consideration that arose out of this study was the use of multiple data sources. To strengthen the findings of this study, a larger sample of suicidal women and their spouses would need to be used and multiple sources of data could be adopted for the purposes of triangulation including: self-reports of husbands and wives; diaries, letters, and other personal correspondence; reports from other informants including family members, physicians, and helping professionals; and direct observations of husbands' and wives' interactions. Other opportunities to explore counsellors' perceptions might have included observing counsellors in case consultation meetings regarding these particular clients. The study might have also been strengthened if husbands and wives could have been interviewed directly in the same semi-structured format used with counsellors to determine how closely counsellors' understanding of events might be to their clients' descriptions of events. Further work needs to be done in the area of exploring the differences between those couples with a suicidal spouse and those couples experiencing distress without leading one of the partners to seek a suicidal resolution. Many couples experience distress without leading one of the partners to a suicidal crisis. It is important to note which characteristics might distinguish distressed couples with a suicidal spouse from other distressed couples who manage to avoid a suicidal outcome, so that efforts at prevention and interventions can be tailored to help individuals and couples cope with crises in ways that preclude self-destructiveness. Much could be learned from those non-suicidal couples who experience distress particularly with regard to their problem-solving skills, levels of intimacy, and general satisfaction with the relationship. Couples with a suicidal wife should be compared to couples with a suicidal husband to explore whether these self-destructive behaviours might be differently manifested in the relationships of each gender. It would be useful to explore what the precipitating factors might be in the suicidal behaviour of husbands as compared to wives. The findings of this study suggest the need for multiple levels of prevention efforts and interventions at various stages in the life cycle, including: early identification of high risk children, educating and counselling children and families by promoting the mental health of all members, and offering individual, marital, and family counselling to those clients presenting with suicidal symptoms. More outcome research needs to be done that compares the effectiveness of different therapy approaches with suicidal clients, in particular, individual counselling versus marital or family counselling, or a combination of each. In order to achieve greater understanding, researchers need to place more emphasis on the ecosystemic nature of all human behaviour. Specifically in the area of suicidal behaviour, there needs to be an acknowledgement of the environmental, cultural, societal, familial and personal levels of influence. Researchers need to incorporate as many of these levels as they can into their designs in order to generate meaningful, context-based, findings. Conclusion In spite of the limitations of this study - its small sample size and shifting emphasis - this type of research is valuable if its only service has been to focus and bring attention to the very grave issue of women's suicidal behaviour. Clinicians and researchers alike need to become more collaborative in their efforts to study and understand suicidal behaviour, and each has a lot to learn from the other. By reducing the alienation between researchers and clinicians, subject recruitment difficulties like those experienced in this study may be alleviated. Suicide does not need to claim as many lives as it does each year, and researchers can play a part in reducing the suicide rate by studying the etiology of suicidal behaviour as it emerges out of a rich and varied context and by allowing their findings to be translated into suicide prevention efforts. It is believed that this study has laid some important groundwork in this regard. The findings of this study offered a conceptual framework for understanding married women's suicidal behaviour based on counsellors' perceptions. Such a framework incorporated women's early family experiences, their internal beliefs about themselves, the circumstances influencing their choice of marital partners, their relationship dynamics, and the influences that environmental and social factors had on the emergence of their suicidal behaviour. This study has been unique in its reliance on counsellors' perceptions as a primary data source. Counsellors can offer valuable information about the experiences of their clients, and this study has provided useful documentation about the process involved in accessing counsellor opinions and perceptions for research purposes. Capitalizing on the perceptions of clinicians to provide insight into the vast array of clients' presenting experiences represents a methodolgical innovation 107 that has previously been under utilized in clinical research. It is felt that such an approach can offer a unique and powerfully rich perspective for understanding the complex experiences of clients, especially when combined with other data sources. In summary, counsellors perceived that the women in this study were raised in families characterized by abuse and neglect resulting in their tendency to see themselves as unworthy and damaged. These women's negative views of themselves may have driven them to choose to marry men who came from abusive homes and felt equally negative about themselves in order to recreate dynamics that were most familiar to them. What resulted was a marriage that was characterized by conflict, distrust, dependency, and dysfunctional communication. As external pressures increased, these women felt more and more incapable, and developed strong failure identities. As they were unable to talk to their husbands about how they really felt, and as they had few other supportive relationships, they began to feel trapped and hopeless. Suicidal behaviour emerged for these women as a response to such circumstances, and out of a complex context of historical factors, present relationship dynamics, and their own beliefs about themselves. Early prevention efforts and counselling interventions could offer these women and their spouses an opportunity to negotiate the difficult challenges facing them at different life stages in a healthy manner, without leading them to choose a self-destructive resolution. It is believed that this study has heightened awareness about the complexity of suicidal behaviour and the mitigating influence that counsellors can play in helping their clients deal with life's inevitable challenges and crises. Finally, what appears to be the most valuable aspect of therapy with suicidal clients is the instillation of hope. Kaplan and Klein (1989) articulated this important aspect of therapeutic work with suicidal women, It is central, whether in the therapy hour, at home, or in the workplace, to feel that there are those with whom you can share and by whom you feel understood, whose own needs incorporate, build on, and enhance your needs. Such mutuality of affect and effort produces hope, if not change, but hope can then encourage further efforts at growth through connection and less likelihood of the feeling of disconnection which fuels suicidal behaviour in women ( p.280). References Adam, K., Bianchi, G., Hawker, F., Nairn, L., Sanford, M., & Scarr, G. (1978). Interpersonal factors in suicide attempts: A study in Christchurch. Australian and New Zealand Journal of Psychiatry. 12, 59-63. Adam, K. S. (1990). Environmental, psychosocial and psychoanalytic aspects of suicidal behaviour. In S. Blumenthal and D. Kupter (Eds.), Suicide over the life cvcle: Risk factors, assessment and  treatment of suicidal patients (pp.39-95). Washington: APA Press. Aldridge, D. (1984). Family interaction and suicidal behaviour: A brief review. Journal of Family Therapy. 6_, 309-322. Bailey, K. (1978). Methods of social research. New York: The Free Press. Beck, A., Lester, D., & Kovacs, M. (1973). Attempted suicide by males and females. Psychological Reports. 22- 965-966. Bennett, M. (1979). The emotional responses of husbands to suicide attempts by their wives. In M. Cook & G. Wilson (Eds.), Love and attraction (pp. 205-206). England: Pergamon Press. Bhagat, M. R. (1976). The spouses of attempted suicides: A personality study. British Journal of Psychiatry. 128. 44-46. Bhagat, M. R. (1977). The interpersonal context of attempted suicide. Proceedings of the Ninth International Congress on Suicide  Prevention and Crisis Intervention (pp. 445-448). Mikkeli: The Finnish Association for Mental Health. Bogdan, R. & Biklen, S. (1982). Qualitative research for education:  An introduction to theory and methods. Boston: Allyn and Bacon, Inc. Bonnar, J. W., & McGee, R. K. (1977). Suicidal behaviour as a form of communication in married couples. Suicide and Life-Threatening Behaviour. 7m. 7- 16. Canetto, S. S. (1988). Couple dynamics in young adult and adult suicidal women (Doctoral dissertation, Northwestern University, 1988). Dissertation Abstracts International. 48. 2450B-2451B. Faust, D. (1986). Research on human judgement and its applicationto clinical practice. Professional Psychology:  Research and Practice. 17. 420-430. Fawcett, J. (1988). Predictors of early suicide: Identification and appropriate intervention. Journal of Clinical Psychiatry. 49(10). 7-8. Fieldsend, R. & Lowenstein, E. (1981). Quarrels, separations and infidelity in the two days preceding self-poisoning episodes. British Journal of Medical Psychology , £4. 349-352. Fitts, W. H. (1965). Tennessee Self-Concept Scale. Los Angeles: Western Psychological Services. Friesen, J. D. (1985). Structural-strategic marriage and family  therapy. New York: Gardner Press. Goldberg, M., & Mudd, E. H. (1968). Effects of suicidal behaviour upon marriage and the family. In L. L. P. Resnik (Ed.), Suicidal  behaviours (pp. 348-356). Boston: Little, Brown, and Company. Gove, W. (1972). Sex, marital status and suicide. Journal of Health and Social Behaviour. IS, 204-213. Haase, J., & Myers, S. (1988). Reconciling paradigm assumptions of qualitative and quantitative research. Western Journal of Nursing  Research. 10(2). 128-137. Hart, E. & Williams, C. (1987). Suicidal behaviour and interpersonal network. Crisis. 8(2). 112-124 Heshusius, L. (1980). Female self-injury and suicide attempts: Culturally reinforced techniques in human relations. SexRoles. £(6), 843-857. Jasnoski, M. L. (1984). The ecosystemic perspective in clinical assessment and intervention. In W. A. O'Connor & B. Lubin (Eds.), Ecological approaches to clinical and community  psychology (pp.41-56). Toronto: John Wileyand Sons, Inc. Johnson, W., & Heppner, P. (1989). On reasoning and cognitive demands in counseling: Implications for counselor training. Journal of Counseling and Development. 67. 428-429. Kaplan, A., & Klein, R. (1989). Women and suicide. In D. Jacobs & H. Brown (Eds.), Suicide: Understanding and responding: Harvard medical school perspectives (pp. 257-282). Conneticut: International Universities Press. Kaplan, H., & Pokorny, A. (1976). Self-derogation and suicide. Social Science Medicine. 10. 113-121. Kumler, F. R. (1964). Communication between suicide attemptors and significant others: An exploratory study. Nursing Research. 13(3), 268-270. Langer, E. J. & Abelson, R. P. (1974). A patient by any other name.Clinician group difference in labelling bias. Journal of Counsulting and Clinical Psychology. 42, 4-9. Lather, P. (1986). Issues of validity in openly ideological research: Between a rock and a soft place. Interchange ,17(4), 63-84. Lester, D. (1983). Whv people kill themselves. Illinois: Charles C. Thomas Publisher. Maris, R. W. (1981). Pathways to suicide. Baltimore: The John Hopkins University Press. Marshall, C, & Rossman, G. (1989). Designing qualitative research. California: Sage Publications. Miles, M. & Huberman, A. (1984). Qualitative data analysis: A  sourcebook of new methods. Beverly Hills: Sage Publications. Nelson, R. (1981). Realistic dependent measures for clinical use. Journal of Consulting and Clinical Psychology. 49(2), 168-182. Neuringer, C , & Lettieri, D. J. (1982). Suicidal women: Their thinking and feeling patterns. New York: Gardner Press. Pommereau, X. & Penoutl, F. (1987). Suicidal women and their intimate conflicts with men. Proceedings of the Twentieth Annual  Conference for the International Association for Suicide  Prevention . 192-195. Reason, P., & Rowan, J. (1981). Issues of validity in new paradigm research. In P. Reason & J. Rowan (Eds.), Human inquiry: A  sourcebook of new paradigm research (pp.239- 262). New York: John Wiley. Richman, J. (1978). Symbiosis, empathy, suicidal behaviour and the family. Suicide and Life-Threatening Behaviour. & 139-148. Richman, J. (1979). The family therapy of attempted suicide. Family  Process. 8_, 131-142. Richman, J. (1981). Suicide and the family: Affective disturbances and their implications for understanding, diagnosis and treatment. In M. R. Lansky (Ed.), Family therapy and major psvchopathologv (pp.145-160). New York: Grune and Stratton, Inc. Richman, J. (1986). Family therapy for suicidal people. New York: Springer Publishing Company. Schmidtke, A., & Schaller, S. (1989). Sex differences in the metaperspectives of parasuicides. [Summary]. Proceedings of  the 22nd Annual Meeting of the American Association of  Suicidology. 6-7. Sharpley, C. F., & Khan, J. A. (1982). The relationship between marital adjustment and self-concept for married individuals and couples. Journal of Individual Psychology. 38(1). 62-71. Snyder, D. K. (1979). Multidimensional assessment of marital satisfaction. Journal of Marriage and the Family. 41. 813-823. Snyder, D. K. (1981). Manual for the Marital Satisfaction  Inventory. Los Angeles: Western Psychological Services. Snyder, D.K., Wills, R. M., & Keiser, T. W.(1981). Empirical validation of the Marital Satisfaction Inventory: An actuarial approach. Journal of Consulting and Clinical Psychology. 49. 262-268. Snyder, D. K. (1983). Clinical and research applications of the Marital Satisfaction Inventory. In E. E. Filsinger (Ed.), Marriage  and family assessment. Beverly Hills: Sage. Snyder, M. & Thomsen, C. J. (1988). Interactions between therapists and clients: Hypothesis testing and behavioural confirmation. In D. C. Turk and P. Salovey (Eds.), Reasoning, inference and  judgement in clinical psychology. New York: Free Press. Stephens, B. J. (1985). Suicidal women and their relationships with husbands, boyfriends and lovers. Suicide and Life-Threatening  Behaviour. 15.(2), 77-90. Stephens, B. J. (1988). The social relationships of suicidal women. In D. Lester (Ed.), Why women kill themselves. Illinois: Charles C. Thomas Publisher. Suter, B. (1976). Suicide and women. In B. B. Wolman & H. H. Krauss (Eds.), Between survival and suicide. New York: Gardner Press. Tabachnick, N. D. (1961). Interpersonal relations in suicide attempts: A psychodynamic formulation and some implications for treatment. Archives of General Psychiatry. 4, 16-21. Thorne, S. (1991). Methodological orthodoxy in qualitative nursing research: Analysis of the issues. Qualitative Health Research. 1(2), 178-199. Webb, E., Campbell D., Schwartz, R., Sechrest, L, & Grove, J. (1981). Nonreactive measures in the social sciences. Boston: Houghton MifflinCompany. Wenz, F. (1976). Self-evaluation and suicide potential. Psychological Reports. 39, 289-290. Wills, R. M. & Snyder, D. K. (1982). Clinical use of the Marital Satisfaction Inventory: Two case studies. American Journal of Family Therapy, 112(4), 17-26. Wilson, L, Miskimins, R., Braucht, G., & Berry, K. (1977). The suicide attempter and self-concept. Journal of Clinical  Psychology. 27(3), 307-309. Appendices Appendix A Methodology of Original Investigation Couples with suicidal wives were compared with two different couple comparison groups: those non-suicidal couples who recently entered marital counselling and non-clinical couples who volunteered from the community to be part of the research. An effort was made to match the women on dimensions of age, number of years married, education, and presence of children. There were limitations in this study as a result of a possible sampling bias. The counselling couples and couples with suicidal wives were unique from community couples in that they were both recruited from counselling agencies, while the community couples were volunteers who responded to an announcement asking for research subjects. Based on this characteristic alone, community couples may have been different from the other couples, but they were chosen based on the assumption that they would reflect differences on the dimensions of marital satisfaction and self-concept. Women who were either legally married or had been living in a common-law relationship for at least one year were considered "married" for the purposes of this study. No distinction was made between these two groups. 116 Hypotheses Wives' Self-Concepts Those researchers that examined the self-concepts of suicidal individuals found the following: that suicidal individuals scored lower on measures of self-evaluation than non-suicidal individuals (Schmidtke & Schaller, 1989; Wenz, 1976), that suicidal individuals frequently exhibited a high degree of self-rejecting attitudes ( Kaplan & Pokorny, 1976; Suter, 1976) and self-derogation (Wilson et al., 1971), and suicidal individuals reported experiencing less self-approval and less self-confidence than non-suicidal individuals (Neuringer & Lettieri, 1982). Based on these findings, it was predicted that suicidal women would have lower scores on the overall level of self-esteem scale, lower scores on the sub-scales of personal self and family self, and higher scores on the self-criticism scale, as measured on the TSCS, than either the women in counselling or women in the community. Husbands' Self-Concepts It was a secondary aim of this study to discover whether or not non-suicidal husbands of suicidal wives would have lower overall self-esteem scores on the TSCS than the husbands of wives in counselling or husbands of wives in the community. This was expected because of the research that noted the influence that non-suicidal spouses have on the emergence of suicidal behaviour in their mates (Bhagat, 1976; Canetto, 1988; Richman, 1978; Tabachnick, 1961). 1 1 7 Similar to the approach used with wives, husbands' self-concept scores were compared among the three types of couples, with an expectation that husbands of suicidal wives would score the lowest on the targeted subscales of the TSCS, followed by husbands in counselling, followed by husbands in the community. Finally, husbands and wives scores were compared with one another across all three groups to explore whether any significant differences were noted. Marital Satisfaction The second level of inquiry in this study was an exploration of the marital dynamics of suicidal women. The Marital Satisfaction Inventory (MSI, Snyder, 1981) was selected as a way to provide the researcher with some information about the levels of marital distress experienced and perceived by a couple following a woman's suicide attempt. It was predicted that couples with suicidal wives would score higher on the global distress scale than couples in the community; with couples in counselling scoring closer to the couples with suicidal wives. If couples with suicidal wives did show similar levels of global distress as did the couples in counselling as indicated by their scores on the MSI, then it seemed to suggest that interpersonal issues are of paramount importance in treating married suicidal women. It was on the specific MSI subscale dimensions of conventionalization, affective communication, problem-solving communication, and family history of distress, where couples with suicidal wives were expected to be distinct from the other couples. 118 These findings were expected because of the research that cited the presence of disturbed communication (Bhagat, 1977; Bonnar & McGee, 1977; Goldberg & Mudd, 1968; Kumler, 1964; Richman, 1986), unresolved hostility (Bennett, 1979; Stephens, 1985), poor conflict resolution (Fieldsend & Lowenstein, 1981), decreased satisfaction (Hart & Williams, 1987) and disturbed symbiotic relationships (Richman, 1978), in the interpersonal and family of origin relationships of individuals who engaged in suicidal behaviour. Research Design The study was a causal-comparative study yielding interval data, to be compared using analysis of variance procedures. Comparisons were made at two different levels: the individual level and the marital level, for three different types of couples: (a) suicidal wives and their husbands, (b) non-suicidal couples in counselling, and (c) non-suicidal couples who have volunteered from the community. The comparisons to be made at the individual level were as follows: wives who attempted suicide were compared with wives in counselling and wives in the community on individual self-concept measures; husbands of suicidal wives were compared with husbands in counselling and husbands in the community on individual self-concept measures. At the marital level, the comparisons were as follows: couples with suicidal wives were compared with couples in counselling and couples in the community on a marital satisfaction scale. A couple's score was 119 derived by computing a mean score based on the husband's and wife's overall satisfaction scores. Furthermore, an incompatibility score based on the discrepancy between the husband's and wife's scores on overall marital satisfaction was also generated; where low levels of discrepancy between husbands and wives scores was taken to indicate greater compatibility in their perceptions of the satisfaction in their marriage, which was understood to contribute to greater marital adjustment. The three groups of couples were also compared on various demographic variables, including: age, presence of and number of children, number of years married, income, education, and history of suicide attempts in the family. Sample Fifteen couples in each of the following three groups were to be compared on the two dimensions of self-concept and marital satisfaction: (a) couples with suicidal wives, (b) couples currently in counselling, and (c) couples who have volunteered from the community. Suicidal women and their husbands were selected from the population of clients referred to S.A.F.E.R. counselling agency. Couples in counselling were selected from local counselling agencies and private practitioners offering marital counselling. Community couples were comprised of those couples who volunteered from the community to be part of the research project. All couples were to have been married for at least one year, and no longer than ten years. There was to be no history 120 of suicide attempts for either the couples in counselling or community couples. Once a married woman was referred to S.A.F.E.R. following a suicidal crisis or attempt, she was asked by her counsellor if she and her spouse would be willing to be contacted by the researcher to be part of the study. If consent to participate in the study was given, the couple's task was for each spouse to fill out two self-report instruments and one demographics questionnaire, requiring approximately 50 minutes in total. Both the women and their husbands had the right to refuse participation and were allowed to withdraw from the study at any time. The decision to withdraw did not affect the counselling the woman wasreceiving through S.A.F.E.R. These conditions were clearly articulated on their consent forms. Couples in counselling were asked to fill out the self-report instruments and demographics questionnaire upon initial intake. Community couples were asked to fill out the same instruments once they gave their consent to participate. Again, all subjects in the project had the right to refuse participation at any time. Instrumentation The Tennessee Self-Concept Scale (TSCS, Fitts, 1965) was selected to provide a measure of self-concept for all husbands and wives participating in the study. Specifically, the form that was used was the Counselling Form, as it provided the most relevant information for the purposes of this study, including: a self-criticism score, an overall self-121 esteem score, and two subscale scores measuring personal worth and sense of worth in relation to one's family. The TSCS was chosen because of the strength of its reliability and validity and because it has withstood the test of time as a powerful measure of self-concept. Stability reliability coefficients on the TSCS range from .60 to .92. Concurrent validity has been established by comparing the counselling form of the TSCS with the MMPI (.50 to .60) and with the Taylor Anxiety Scale ( .50 to .70). Content validity was established through unanimous acceptance of items by various experts and predictive validity was also clearly established ( Fitts, 1965). The Marital Satisfaction Inventory (MSI, Snyder, 1981) was used to measure couples on 11 different marital dimensions, including: conventionalization, global distress, affective communication, problem-solving communication, time together, disagreement about finances, sexual dissatisfaction, role orientation, family history of distress, dissatisfaction with children, and conflict over childrearing. Test-retest reliability of the MSI ranges from .84 to .94. Item analyses indicate a high degree of internal consistency amongst the scales and intercorrelational and factor analyses confirm that there is clear distinction between scales. Alpha coefficients for individual scales range from .80 to .97. The strength of the MSI's discriminant validity has been clearly demonstrated and MSI scores have been shown to successfully discriminate between couples in therapy and matched controls (Snyder, 1979). Individual profiles have also been shown to be 1 2 2 related to clinicians' ratings of couples' distress across a wide range of clinical situations (Snyder, Wills, & Keiser, 1981). The MSI was chosen as an instrument because of its strong reliability and validity components, its ease of completion, the inclusion of a family of origin scale, and because it discriminates well between couples in therapy and matched controls. Analysis Comparisons The comparisons to be made were at the individual level based on self-concept scores from the TSCS and at the marital level based on marital satisfaction scores from the MSI for husbands and wives in each of the three groups. Comparisons were also made amongst husbands, wives, and couples. Additionally, demographic items were also compared across individuals and couples. On the TSCS, suicidal wives were compared with wives in counselling and wives in the community, and husbands of suicidal wives were compared with husbands in counselling and husbands in the community. Group 1 couples were compared with Group 2 couples and Group 3 couples based on their mean scores on the global distress scale of the MSI and on their incompatibility scores. Means and standard deviations were computed separately for each comparison level group. Variables In this study there were two levels of comparison: the individual level and the marital level. Three types of couples were compared : (a) 123 couples with suicidal wives, (b) counselling couples, and (c) community couples. At the individual level, husbands and wives in each of the three groups were compared separately, while at the marital level, couples from each of the three groups were compared. The independent variable in this study was the type of couple, where husbands, wives, and couples were each treated separately. The dependent variable had two dimensions: scores on the TSCS and scores on the MSI. Additionally, demographic variables were compared across groups. Statistics Descriptive statistics were computed for each of the three types of groups, for each of the husbands, wives, and couples. ANOVA procedures were to be used to ensure that the groups did not differ on demographic variables or on the dimension of married vs. common-law couples. Chi square analysis were also to be applied to the data to determine whether the observed frequencies differed significantly from the expected frequencies based on the null hypothesis. Correlation matrices were compiled in order to survey the relationships amongst the scores of individual husbands and wives by groups. For example, scores of wives in each of the three groups on the TSCS were to be compared with their scores on the MSI, husbands' scores were to be similarly compared on both measures, and husbands' and wives' scores of each couple were to be compared together to determine if there was a significant correlation. 1 2 4 ANOVA Procedures One-way ANOVA's were to be computed separately for each comparison level outlined above. Firstly, a one-way ANOVA was to be performed on wives' scores in all three groups using the TSCS scores as data. Secondly, wives' scores on the TSCS, with the demographic variables taken into account, were to be analyzed. Thirdly, wives scores on the TSCS, with the MSI scores taken into account, were to be analyzed. The same procedures were to be repeated for husbands' scores. Husbands and wives scores were to be analyzed independently from one another when analyzing the data to reduce the effects of their scores influencing one another. For this reason, multiple analysis of variance was not chosen for analyzing this data. Couple scores were alsoto be analyzed using ANOVA procedures. Couples in each of the three groups were to be compared using the MSI scores as data. Secondly, the MSI scores were to be analyzed in conjunction with the demographic data. Follow-up tests such as t-tests or Scheffe' procedures were to be conducted to determine the source of any significant differences. 125 Appendix B Tables Table 3 Mean Scores for Selected MSI Items Across Three Groups Suicide -Attempt Counselling Community scale Husband Wife Hus. Wife Hus. Wife CNV 13.0 1.0 3.3 3.5 7.9 7.3 GDS 8.5 26.7 28.5 27.0 6.3 8.8 AFC 10.0 15.3 15.6 14.9 6.6 8.2 PSC 14.5 32.3 28.1 24.8 9.2 10.7 FAM 12.0 12.7 10.9 10.3 7.7 6.0 Note. The higher the value, the greater the perception of distress, except for CNV which represents social desirability. Table 4 Mean Scores for Selected TSCS Items Across Three Groups scale Suicide Husband Wife Counselling Hus. Wife Community Hus. Wife TotP 329.0 277.0 321.3 340.1 336.2 356.8 S-C 36.0 41.7 33.4 32.4 37.3 36.0 Pers 62.0 54.0 60.4 64.6 65.7 68.3 Fam 71.0 50.7 63.5 66.5 66.6 74.0 Note. The higher the value on each scale, the greater the feelings of self-worth, except for S-C, which represents self-criticism. 1 2 6 Table 5 Descriptive Characteristics of Husbands in Each of the Three Groups VARIABLE VALUE Suicide Att. Counsel. N=2 N=8 Community N=9 FRQ. % FRQ % FRQ % 20-25 2 25 26-30 2 25 2 22.2 31-40 1 50 1 12.5 3 33.3 40+ 1 50 3 37.5 4 44.4 high school 1 50 5 62.5 post-secondary 1 50 univers./college 2 25 2 22.2 grad./doctoral 7 77.8 blank 1 12.5 less than 15,000 2 25 26,000-35,000 1 12.5 1 11.1 more than 35,000 2 100 5 62.5 8 88.9 student 1 12.5 1 11.1 homemaker clerical manager 1 11.1 professional 1 50 1 12.5 6 66.6 service 2 25 technical 1 12.5 self-employed 1 12.5 1 11.1 unemployed 1 12.5 other 1 50 1 12.5 English Canadian 5 62.5 4 44.4 French Canadian 1 11.1 European 1 50 3 37.5 2 22.2 Native Indian 1 11.1 other 1 50 1 11.1 yes 2 25 no 1 50 6 75 9 100 blank 1 50 age education income occupation ethnicity drugs/alcohol prob. self 127 Table 5 continued Suicide Counselling Comm VARIABLE VALUE FRQ °A FRQ °A FRQ °A drugs/alcohol yes 1 12.5 counselling no 1 50 7 87.5 9 100 blank 1 50 drugs/alcohol yes 1 12.5 prob. for spouse no 1 50 7 87.5 9 100 blank 1 50 drugs/alcohol yes 2 25 2 22.2 prob. for family no 3 37.5 5 55.6 blank 2 100 3 37.5 2 22.2 medication for yes 1 12.5 emotional health no 1 50 7 87.5 9 100 blank 1 50 1-1 counselling yes 2 25 2 22.2 no 1 50 2 25 7 77.8 blank 1 50 4 50 marital yes 8 100 counselling no blank ] 50 50 9 100 both yes 3 37.5 no 1 50 3 37.5 9 100 blank 1 50 2 25 history of suicide yes 1 12.5 attempts no 1 50 7 87.5 9 100 blank 1 50 family history of yes 1 50 2 22.2 suicide attempts no 7 87.5 7 77.8 blank 1 50 1 12.5 family history of yes 1 11.1 suicide deaths no 1 50 7 87.5 7 77.8 blank 1 50 1 12.5 1 11.1 128 Table 6 Descriptive Characteristics of Women in Each of the Three Groups Suicide Counsel. Community VARIABLE VALUE N=3 N=8 N=9 FRQ. % FRQ % FRQ % age 20-25 1 33.3 1 12.5 26-30 4 50 4 44.4 31-40 1 33.3 2 25 2 22.2 40+ 1 33.3 1 12.5 3 33.3 education high school 1 12.5 1 11.1 post-secondary 1 33.3 4 50 3 33.3 univers./college 1 33.3 3 37.5 3 33.3 grad./doctoral 1 33.3 2 22.2 blank income less than 15,000 1 33.3 2 25 26,000-35,000 1 33.3 2 25 1 11.1 more than 35,000 1 33.3 4 50 8 88.9 occupation student 2 66.7 2 25 1 11.1 homemaker 3 37.5 2 22.2 clerical 2 25 manager professional 1 33.3 1 12.5 5 55.5 service technical self-employed 1 12.5 1 11.1 unemployed 1 12.5 other 1 12.5 ethnicity English Canadian French Canadian 1 33.3 5 62.5 8 88.9 European 1 33.3 2 25 Native Indian 1 33.3 other 1 12.5 1 11.1 drugs/alcohol yes 1 11.1 prob. self no 3 100 7 87.5 7 77.8 blank 1 12.5 1 11.1 1 2 9 Table 6 cont'd Suicide Counselling Comm VARIABLE VALUE FRQ A FRQ A FRQ °A drugs/alcohol yes counselling no 3 100 7 87.5 9 100 blank 1 12.5 drugs/alcohol yes 2 66.7 2 25 prob. for spouse no 1 33.3 5 62.5 9 100 blank 1 12.5 drugs/alcohol yes 2 66.7 2 25 1 11.1 prob. for family no 1 33.3 3 37.5 5 55.6 blank 3 37.5 3 33.3 medication for yes 1 12.5 1 11.1 emotional health no 3 100 6 75 8 88.9 blank 1 12.5 1-1 counselling yes 3 100 2 25 no 2 25 9 100 blank 4 50 marital yes 1 33.3 7 87.5 1 11.1 counselling no 2 66.7 7 77.8 blank 1 12.5 1 11.1 both yes 1 33.3 3 37.5 no 2 66.7 1 12.5 8 88.9 blank 4 50 1 11.1 history of suicide yes 3 100 attempts no 7 87.5 9 100 blank 1 12.5 family history of yes 1 33.3 1 12.5 2 22.2 suicide attempts no 1 33.3 6 75 7 77.8 blank 1 33.3 1 12.5 family history of yes suicide deaths no 2 66.7 7 87.5 8 88.9 blank 1 33.3 1 12.5 1 11.1 130 Table 7 List of Start-up Codes 1. counsellor opinion 2. framework for understanding suicidal behaviour 3. history of wife 4. history of husband 5. description of wife (adjectives, affective, cognitive, behavioural) 6. self-statement of wife as reported by counsellor 7. description of husband (adjectives, affective, cognitive, behavioural) 8. self-statement of husband as reported to counsellor 9. description of marital relationship 10. description of relationship by husband or wife as reported to counsellor 11. description of spouse by husband or wife as reported to counsellor 12. description of self-concept of wife 13. description of self-concept of husband 14. counsellor role 15. counsellor- client relationship 16. clinical perspective 17. clinical strategy 18. help-seeking behaviour on part of wife 19. help-seeking behaviour on part of husband 20. help-seeking behaviour as a couple 21. coping strategies of wife 22. coping strategies of husband 23. wife's expression of suicidal intent 24. husband's expression of suicidal intent 25. description of current family stressors on wife 26. description of current family stressors on husband 27. current non-family stressors on wife 28. current non-family stressors on husband 29. current victim of abuse (physical, sexual, emotional) (husband/wife) 30. current perpetrator of abuse (physical, sexual, emotional) (husband/wife) Figure 2 Qualitative Analysis Documentation Form  Analyst: 1. Research Issue being explored: 2. In this analysis, what specifically are vou aiming to do? 3. Description of procedures: SPECIFIC DATA RESEARCH QUESTIONS STARTCODES SOURCES ANALYSIS OPERATIONS Data reduction Data display Drawing/Verifying Conclusions Adapted from Miles & Huberman (1984) 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0053667/manifest

Comment

Related Items