LIBRARY r Michigan State University L . e PLACE IN RETURN BOX to remove this checkout from your record. _T‘—m lil' __..—‘I .. TO AVOID FINES return on or before date due. a: DATE DUE DATE DUE DATE DUE ‘ :L___J *7 fl l—_—.— MSU is An Affirmative Action/Equal Opportunity Institution c:\circ\dMn.pm3—p.1 ADJUSTMENT OF WIVES DURING THEIR HUSBANDS' TERMINAL ILLNESSES BY David Edward Peradotto A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Psychology 1990 This study that might be r time of their h A Spouse's crises an indiv that can Predict Surviving specs! in terms of investigated f. adjuStment: ma] length of illne able to accept A retmspel and twehe completed Quest E ach Quest 10““ v | enables "hie adjustment. All 0591c e Program The inveSt ABSTRACT ADJUSTMENT OF WIVES DURING THEIR HUSBANDS' TERMINAL ILLNESSES BY David Edward Peradotto This study was undertaken to examine some of the factors that might be related to the adjustment of wives during the time of their husbands' terminal illnesses. A spouse's final illness is one of the most difficult crises an individual can face. The identification of factors that can predict the quality of the pre-loss adjustment among surviving spouses would be of considerable use to clinicians in terms of primary prevention planning. This study investigated five factors in terms of their impact on adjustment: marital. quality, wife's age, social support, length of illness, and the degree to which the husband was able to accept the terminal nature of his condition. A retrospective research design was employed. Between six and twelve months after their husbands' deaths, widows completed questionnaires that assessed nine aspects of their adjustment during the time of their husbands' final illnesses. Each questionnaire also included measures of the five variables which were predicted to have an impact on adjustment. All 47 subjects were recruited through Michigan hospice programs. The investigation uncovered no convincing evidence that Marital Qualit" capable 0f ex 4 adjustment durl However. the 33 for measuring potential effeC Significan Support and {H0 Work. Subjects support accorpi their husbands, grief issues th. support. Significa: Husbar ' .rd 3 Acci A4, aujustment. Wii Significantly n 1935 guilt tr. acceptance. TI‘ ' ‘e Import los S . adJUstmi r ESearchers b 0 Marital Quality, Wife's Age, or Length of Illness are factors capable of exerting a significant influence on a wife's adjustment during the time of her husband's terminal illness. However, the absence of psychometrically proven instruments for measuring adjustment hindered the investigation of potential effects by these variables. Significant associations were found between Social Support and two aspects of adjustment, Marital Work and Grief Work. Subjects who experienced a "high" degree of social support accomplished significantly more marital work with their husbands, and were better able to maintain awareness of grief issues than.were subjects who did not experience as much support. Significant associations were also found between the Husband's Acceptance variable and certain aspects of adjustment. Wives of "high acceptance" husbands accomplished significantly more marital work and experienced significantly less guilt than did wives whose husbands were "low" in acceptance. The importance of the issue of a surviving spouse's pre- loss adjustment warrants continued investigation by researchers, beyond the findings of this study. I am trulyr’ the completion responsible fc experience of r I am most "ETY Stage of critical think; that he was ava (a frequent occ Even more value! member who was s relatiOnship w AlthOUgh J the field ACKNOWLEDGMENTS I am truly grateful to the many people who contributed to the completion of this research project. I consider them to be responsible for making this the most positive educational experience of my years at Michigan State. I am most grateful to my chairperson, Lee Winder. At every stage of this project, I relied on Lee's insights, critical thinking and common sense, for direction. Knowing that he was available whenever I needed to discuss some idea (a frequent occurrence) was a source of great reassurance. Evenrmore valuable toime was the respect Lee always showed for my ideas, and the confidence he expressed in my abilities. I feel very fortunate to have had the support of a faculty member whO‘was sorenthusiastic about.my research and concerned about. me as a student; and. I am, proud. of the working relationship we were able to establish. Although John Schneider is acknowledged as an expert in the field of loss and grief, he approaches the topic with the humility of one who realizes that there is still much to be learned. His openness to new ideas, as well as his curiosity, enthusiasm, and willingness to share from his own grief experiences further stimulated my interest in the field. These qualities also made our discussions both valuable learning experiences and a pleasure in which to participate. More than anyone, John has influenced my thinking concerning loss and grief. He has deepened, my understanding' of the various meanings and personal significance of a major loss for an individual, and he modelled patience and respect for the way each individua; that a loss c. personal grout; loss involves 5 Anne Boga: this project. responsible for provided usefu; editing skills greatest contr; quality work, a Of this PTOject Margaret E and brought to I marital quality during the Na:- organizational c I “Ould l HlChigan hOSpi ca each individual works through his/her grief. He also taught me that a loss carries with it the potential for significant personal growth, and that true resolution and recovery from loss involves such growth. Anne Bogat made a number of significant contributions to this project. At the proposal stage, her suggestions were responsible for improvements in the study's design. She also provided useful advice about statistical methodology, and her editing skills helped me to express my ideas clearly. Anne's greatest contribution however, was her expectation that I do quality work, an attitude that influenced me in every aspect of this project. Margaret Bubolz willingly served as a committee member, and brought to this project valuable expertise in the area of marital quality. Her suggestions were particularly helpful during the planning phase of the research, when important organizational decisions needed to be made. I would like to express my appreciation to the six Michigan hospice programs who participated in this study. The staffs of these programs contributed significant amounts of time in spite of busy schedules and the fact that they had little to gain personally from their participation. Their enthusiasm for my research no doubt reflects the staffs' deep concern for the bereaved individuals who are their clients, and their commitment to providing their clients with better services. In.particu1ar'I would like to thank Marsha Zanbergen and Loretta Lind of the Hospice of Greater Grand Rapids, Howard Schaeffer and Bob Miller of the Hospice of Lansing, Nora Anderson and.Diane Junglas of Hospice Services of Western Wayne County , Jackson, Caro _ Hospice of Sou“ grief, and thu Observing the one cannot but Concern {01' Ct her friendshi] throucShout the ”Y Very greatly Eased taped me d l avenues of am Wayne County, Julie Schaeffer and Joy Frizzell of Hospice of Jackson, Carolyn Fitzpatrick—Lassen and Jan Grabinski of Hospice of Southeastern Michigan, and Jo Cunningham and Bonnie Rogers of Good Samaritan Hospice in Battle Creek. I am especially grateful to Joy Frizzell who shared with me her personal and professional experiences with loss and grief, and thus enriched my understanding of these phenomena. Observing the bond that exists between Joy and her clients, one cannot but be struck by JOy's sensitivity and genuine concern for others. She was generous with her time and also her friendship, and was a significant source of support throughout the study. My very reliable computer consultant, John Curtis, greatly eased the burden of data analysis. His advice often helped me distinguish potentially fruitful from useless avenues of analysis. It is certainly a comment on my computer expertise that setmany people provided advice and help in this area at one time or another. I am grateful to Gene Maguin, Larry Zeides, Karen.Miske, Susan Burke, Carlos Ruiz, and Keith Hoffman for their knowledge and patience. Finally, I would like to thank the widows who made this study possible through their participation, and who did so despite being in the midst of mourning. I wish them all peace. Chapter II. III. TABLE OF CONTENTS Page Chapter I. INTRODUCTION 1 Statement of Purpose 1 General Problem 1 Need for the Study 2 Theoretical Assumptions 2 Adjustment During a Spouse's Terminal Illness 5 II. REVIEW OF THE LITERATURE 12 Marital Quality 12 Social Support 25 Age 32 Length of Illness ("Sudden Death" Versus Forewarning) 37 Length of Illness (Short-Term Versus Long-Term Chronic) 42 Husband's Acceptance of Terminal Illness 47 III. DESIGN OF THE STUDY 54 Rationale for a Retrospective Research Design 54 Subject Selection 56 Hospice Services 58 Measures Used in the Study 60 Marital Adjustment 61 Social Support 62 Husband's Acceptance 63 Marital Work 65 Carrying Out Responsibilities 67 Physical Care of the Dying Spouse 68 Wife's Acceptance 69 Grief Work 70 Physical Problems 72 Guilt 76 Depression 78 Anxiety 80 Grief Intensity 83 Reliability and Validity of the Measures 85 vii RE IV. VI. APPERDIces A. TAPE c “s. Research Questions and Hypotheses Reliability Hypothesis Validity Hypothesis Experimental Hypotheses Statistical Analyses IV. RESULTS Description of Subjects Formation of the Marital Quality, Social Support, Husband's Acceptance, Age, and Length of Illness Groups Test of the Hypotheses 87 87 87 88 89 93 93 94 96 Hypotheses Related to Scale Construction 96 Hypotheses Related to Factors Influencing the Adjustment of Subjects 99 Additional Analyses Correlations Among the Dependent Variables V. DISCUSSION Sample Issues Description of Subjects Reliability of the Instruments Validity of the Instruments Factors Influencing the Adjustment of Subjects Marital Quality Social Support Husband's Acceptance Age Length of Illness Age x Length of Illness Additional Analyses Correlational Analyses VI. CONCLUSIONS APPENDICES A. TABLES Anticipatory Grief Among Surviving Spouses Measuring Adjustment Factors Related to Adjustment Marital Quality Social Support Husband's Acceptance Age Length of Illness Recommendations for Future Research viii 108 115 117 117 122 124 128 131 132 140 145 149 152 154 156 161 168 168 169 173 173 174 176 177 178 179 181 182 LIST or arsenal B. QUESTIONNAIRE PACKET C. SCORING KEY D. INTERNAL CONSISTENCY OF THE MARITAL WORK, GUILT, WIFE'S ACCEPTANCE, AND PHYSICAL PROBLEMS SCALES LIST OF REFERENCES ix 217 259 274 277 gtatement 9f f» A spouse' events, and or. face. It is th. 5‘“ aPPIOpriat Study was to e to adjustment taminal illne tetms of thej marriage, the husband's 111 Whether or CHAPTER I In ro i n Statement ef Perpeee A spouse's terminal illness is one of life's significant events, and.one of the most difficult crises an individual can face. It is therefore an important issue to»be understood, and an appropriate object of investigation. The purpose of this study was to examine some of the factors that may be related to adjustment of wives during the time of their husbands' terminal illnesses. The following factorsiwere investigated in terms of their impact on adjustment: the quality of the marriage, the wife's age, social support, the length of the husband's illness (excluding cases of sudden death), and whether or not the husband was able to acknowledge the terminal nature of his illness. General greblem This issue is part of the broader topic of bereavement among spouses. A large research literature exists that is consistent in finding bereaved spouses at risk for a variety of deleterious consequences such as :negative changes in physical health, an increased mortality rate, and lower mental health status. For many widows and widowers, the grief and stress associated with their loss continue for many years after the death occurs (Barrett & Schnewies, 1981; Parkes, 1983; Sanders, 1981; Zisook, Devaul, & Click, 1980). Need for the : Less res: surviving 5pc although this . 0f watching irreversible responsibiliti Often take an the Survival» uncertain and It is no FORM), ROQEIS widowed "Omer Either Diver likEly that a risk fOr in terms of i It Ne f h d Less research has been done on the adjustment of the surviving spouse during the time of the final illness, although this is clearly a very difficult period. During a lengthy terminal illness, the surviving spouse must not only face the prospective loss of her partner, but also the strain of watching her husband suffer from a debilitating and irreversible illness. The survivor' must assume the responsibilities previously borne by the sick partner and often take an active role in the care of the spouse. Moreover, the survivor is faced with planning for a future that is uncertain and which will not include her partner. It is not surprising that one study (Vachon, Freedman, Formo, Rogers, Lyall, & Freeman, 1977) found that 81% of newly widowed women rated the time of their husbands' deaths as either "very stressful" or "extremely stressful”. It seems likely that a terminal illness places the surviving spouse at risk for physical and psychological problems. The identification of risk factors would be helpful to clinicians in terms of primary prevention planning. h A m i n It was hypothesized that variables associated with differential outcomes after the death of one's partner would also be associated with adjustment prior to death. Similarities exist between the situation of a surviving spouse during the partner's terminal illness and that of a widowed individual during bereavement. Soon after the spouse's death, the widowed person is first confronted with powerful aspects of i studies have the death as disruption ofj Halikas, 8 14' Parkes, 19553 LYall, Lancee the "idowed pe which there i Painful 35pm: experiences 5 0! unconscio1 aspeCts of ti Commonly fol engages irr character-1m Engel. ‘961 Schneider. normally. 1 griet to a subsides P b? the Hi 3 aspects of .his/her loss. Numerous clinical and research studies have attested to these early weeks and months after the death as a time of acute psychological distress and disruption of normal functioning (Bowlby, 1961; Clayton, Halikas, & Maurice, 1971; Engel, 1961; Lindemann, 1944; Parkes, 1965a, 1965b, 1970, 1972a, 1972b; vachon, Rogers, Lyall, Lancee, Sheldon, & Freeman, 1982). During this time, the widowed person typically goes through a "coping" period in which there is only limited awareness of important, usually painful aspects of the loss. Instead, the bereaved usually experiences shock, numbness, or disbelief, and makes conscious or unconscious attempts to limit awareness of overwhelming aspects of the loss (Bowlby, 1980; Schneider, 1981). This is commonly followed by a phase in which the surviving spouse engages in the work of mourning the loss - a phase characterized by the presence of acute grief (Bowlby, 1980; Engel, 1961, 1962; Freud, 1917; Lindemann, 1944; Marris, 1975; Schneider, 1981). If the bereavement process proceeds normally, the widowed person will move from a time of intense grief to a period of gradual recovery, in which grief slowly subsides, there is acceptance of the loss, and steps are taken by the widowed person to reorganize her life and re-establish social contacts (Bowlby, 1980; Marris, 1975; Parkes, 1972b; Schneider, 1981 ) . These three phases of grief - coping, grief, and recovery - have been described by numerous researchers (Bowlby, 1980; Engel, 1961, 1962; Lindemann, 1944; Marris, 1958, 1975; Parkes, 1965a, 1972a, Parkes & Weiss, 1983). Prior to the death, the surviving spouse only experiences maintain. It h‘ BXperiences t. fantasies, an? relationship IeSponse to th a dramatic one ShOCk, dj and are cons: nevs of a part 1980; Putterm Likewise, a 4 the loss prospectively, although it is likely that she is already suffering from the loss of various functions which the dying partner previously carried out, but now is unable to maintain“ It has alsoibeen suggested that the surviving spouse experiences the immediate loss of all the hopes, dreams, fantasies, and expectations held for that person and for the relationship (Lindemann, 1944; Rando, 1984). The initial response to these prospective and real losses is likely to be a dramatic one. Shock, disbelief, and denial have been commonly observed and are considered normal reactions to first receiving the news of a partner's terminal condition (Aldrich, 1974; Carey, 1980; Futterman, Hoffman, 8- Sabshin, 1972; Kubler-Ross, 1969). Likewise, a period of great sadness or grief is expected when there is increased awareness of the reality that the partner will die, and also awareness of all the effects the loss will have on one's own life (Clayton, Halikas, Maurice, & Robins, 1973; Futterman, et. al., 1972; Kubler-Ross, 1969; Rando, 1983). One retrospective study (Clayton et. al., 1973) did in fact find substantial percentages of widows and widowers experienced symptoms characteristic of grief during the time of their spouses' final illnesses. Depressed mood, sleep disturbance, and crying were reported by over 60% of the subjects; appetite loss or weight loss, and fatigue by more than 40%; and lack of interest, concentration problems, and poor memory by more than 25%. Over the course of the illness, the surviving spouse may begin to adjust to the prospective and real losses she experiences, I of awareness. some controve: In summaz. time after the surviving spo; Progresses th; awareness of eventually and 5 experiences, with a diminishing of the duration and intensity of awareness. The length of time this takes is a matter of some controversy (Schneider, 1984). In summary, both situations — the final illness, and the time after the death - are times of overwhelming loss for the surviving spouse. In healthy bereavement, the surviving spouse progresses through a process that involves coping, limited awareness of the loss, and active grieving, which is eventually and gradually followed.by a recovery phase in which there can be acceptance and reorganization. Adjeegmen; During e Spouse'e Terminel Illneee If one is attempting to make a judgment about any individual's psychological health or "adjustment," it is important to view that person's behavior in a developmental context. Each phase in a person's life presents different challenges and requires completion of different developmental tasks. To a great extent, a person's healthy functioning or "adjustment" can be measured by the extent to which that person is successful in completing tasks that are characteristic of a particular developmental phase of his/her life and. by his/her ability to cope with the stresses associated with those tasks. Bereavement researchers have identified a number of tasks or' challenges that. characteristically face an individual during the time of the terminal illness of his/her spouse. One of these tasks is to begin to grieve the loss prior to the spouse's death (anticipatory grief). Grief can be understood as one phase of the mourning process that is that phase what has been of sadness, loneliness, h disinterest ir. (Bowlbv. 1980 '9“; Harris, Associated ph weight loss, iLindemann’ 1 EXperier SPOUSe priOr normal , heal 6 process that is the normal response to a significant loss. It is that phase of mourning in which there is full awareness of what has been lost. Grief is characterized by intense feelings of sadness, often manifested by crying. Feelings of loneliness, helplessness, hopelessness, anger, guilt, and disinterest in the outside world are also commonly experienced (Bowlby, 1980; Engel, 1961, 1962; Freud, 1917; Lindemann, 1944; Marris, 1975; Parkes, 1965a, 1972b; Schneider, 1981). Associated physical symptoms often include poor appetite, weight loss, insomnia (DSM-III, 1980), and exhaustion (Lindemann, 1944; Marris, 1975; Schneider, 1981). Experiencing grief is an important task of the surviving spouse prior to the death, first because grieving is the normal, healthy response to loss (Bowlby, 1963, 1969, 1973, 1980; Engel, 1962; Freud, 1917; Marris, 1975; Schneider, 1984), and, as previously discussed, the terminal illness is a time of loss for the surviving spouse. Secondly, most clinicians and researchers agree that the opportunity to grieve prior to the death has adaptive value (Aldrich, 1974; Carey, 1980; Gerber, 1974; Gut, 1974; Parkes, 1972b, 1975; Rando, 1984; Vachon, Rogers, et. al., 1982). It is suggested that "anticipatory grief" is beneficial because it allows the surviving spouse, to gradually absorb the reality of the loss and plan for the future (Rando, 1984). During the terminal illness, the surviving spouse is also faced with the task of completing marital work with the dying partner. This might include discussions of past and present marital problems that remain, and attempts to resolve these problems. The and "sum up" wishes, the Lastly, the 4 another. In order Process of gr that her part BY refusing t awareness of SPOUSe fails work is there grief work If important ma; diSCUSsing ff 7 problems. The couple may reminisce about times spent together and "sum up" their marriage. If the dying partner has any last wishes, the surviving spouse will try to carry them out. Lastly, the couple needs to say their final good-byes to one another. In order to complete marital work, and also to begin the process of grieving, the surviving spouse must first recognize that.her partner's illness will result in the partner's death. By refusing to accept the diagnosis or by attempting to avoid awareness of the terminal nature of the illness, the surviving spouse fails to experience the reality of the loss, and grief work is therefore delayed" If denial lasts too long, essential grief work may be missed and remain incomplete. Completing important marital business such as carrying out last wishes, discussing feelings about the partner's death and dying, and saying final good-byes, all become problematic if the surviving spouse does not, sooner or later during this terminal phase, recognize that her partner is dying. Awareness and recognition of the terminal nature of her partner's illness is thus a crucial task for the surviving spouse during the time of the final illness. Providing for the physical care of the dying person is an important responsibility of the surviving spouse throughout the terminal illness. Often the surviving spouse can provide this care directly. At times when she is unable to provide needed care, the surviving spouse is responsible for arranging for others to do so. It is also the responsibility of the surviving spouse to assure that her partner receives the proper nedica as possible . Spouse is amp with the job. the survivin r"ESPOIISibilit tasks Will he done by the (3 One of t cope With the with this Der or illneSs. E grief and cor have als° be during the t 8 proper medical care and is kept as comfortable and pain-free as possible. During the terminal illness, the surviving spouse is likely to have other important responsibilities in addition to providing for her partner's physical care. If the surviving spouse is employed, there will be responsibilities associated with the job. If there are dependent children living at home, the surviving spouse typically will have the major responsibility for their care. Finally, numerous household tasks will need to be performed, some of which may have been done by the dying partner. One of the challenges facing the surviving spouse is to cope with the significant stress that is typically associated with this period without developing serious physical symptoms or illness. Some physical symptomatology - sleep and appetite problems, and loss of energy - are commonly associated with grief and considered a normal concomitant. These difficulties have also been found to be prevalent among surviving spouses during the time of the final illness (Clayton, Desmarais, & Winokur, 1968) and can be considered a normal response to the stress of the period. It is suggested that physical symptoms might become a problematic aspect of adjustment if their severity interferes with the completion of necessary tasks and responsibilities. Similarly, the surviving spouse must find ways to c0pe with stress without developing serious psychiatric problems such as depression or severe anxiety. It is important here to distinguish a depressive response to loss (reactive depression) f normal respc: reintegrative' (Bowlby, 198:. The presence I I a significant healthy proce Similari and reactive C' 1055- Freud ( world, a loss actiVity . Smith (15 state, an ina 9 depression) from normal grief. As noted earlier, grief is the normal response to significant loss, and has a healing or reintegrative function for the individual who is bereaved (Bowlby, 1980; Engel, 1962; Marris, 1975; Schneider, 1984). The presence of grief in a person.who.has recently experienced a significant loss is a positive sign and suggests that a healthy process is occurring that normally leads to:recovery. Similarities can be noted between uncomplicated mourning and reactive depression, each of which involves a recognizable loss. Freud (1917) observed that mourning and melancholia or depression both involve a lack of interest in the outside world, a loss of the capacity to love, and an inhibition of activity. Smith (1975) also found these similarities: lowered mood state, an inability to find meaning or interest in anything other than the object lost, and a difficulty believing that things will ever get better. Beck (1972) reported that fatigue, sleepvdisturbance, and appetite change are common physical manifestations among depressed subjects he studied. As previously discussed, similar physical responses are common among bereaved persons. Deutsch (1982) found a significant correlation and overlap between depression, as measured by the Beck Depression Inventory (BDI) (Beck, 1974) and scores on an inventory of grief, the Response to Loss Instrument (RLI) (Schneider 3. Deutsch, 1982). In her study, moderate levels of depression were correlated with moderate grief. Scores on the BDI that indicated severe depression were not significantly correlated with grief similaritie more severe Freud observing t expectation but are uncc “0W3 a numb. greater aCC imaginatiOn; others, 1’81; characteI-iZEC. 105" In deep: there is a worthlesS ' 10 with grief scores, suggesting that there might be few similarities between grief and depression when depression is more severe. Freud (1917) first distinguished depression from grief by observing that lowered self-regard, self-reproach, and the expectation of punishment are frequent aspects of depression but are uncommon in grief. In summarizing the literature comparing grief with depression, as well as his own observations, Schneider (1980) noted a number of differences, including, for normal grief, a greater range and variability of mood and activity states; greater access to material from dreams, fantasy, and imagination; and greater variability in responsiveness to others, relative to depression. Normal grief also is characterized by "pining" - a preoccupation with the object lost. In depression, however, the focus is on oneself, and there is a strong tendency to view oneself as bad or worthless, often reflected in suicidal thoughts or feelings. Schneider concludes that depression represents a disruption of normal grief, preventing recovery from loss, and requiring therapeutic or medical intervention. One task of the surviving spouse then is to avoid a depressive reaction to her prospective loss and instead begin the process of normal grief. Dealing with guilt feelings is another challenge to the surviving spouse during the final illness. In depressive reactions to loss, guilt is often manifested in a severely diminished self-concept, while in normal grief, guilt, though common (DSM- typically as: frequent. sou_ spouse that partner's ill better care f grief proces "al that din: from diminis.t In sunma relevant aspe a SurviVing illness: ' BeQin 2- Comp; 3- Acknor iline: ° Prov1, Darth, . Carry Chila. ‘ AVQid 11 common (DSM-III, 1980; Marris, 1975; Parkes, 1970), is not typically associated with feelings of worthlessness. Among the frequent sources of guilt are any belief by the surviving spouse that she could have done something to prevent her partner's illness and death, and that she should have provided better care for her partner. One of the important tasks of the grief process is the awareness and expression of guilt in a way that diminishes it over time or prevents guilt feelings from diminishing the survivor's sense of self-worth. In summary, the following tasks have been identified as relevant aspects of adjustment and characteristically faced by a surviving spouse during the time of the partner's final illness: 1. Begin to grieve the loss 2. Complete marital work with the dying partner 3. Acknowledge the terminal nature of the partner's illness 4. Provide for, or assist in the care of the dying partner 5. Carry out other important responsibilities i.e. job, child—care, household tasks 6. Avoid developing serious physical symptoms oneself 7. a 8. Avoid developing serious psychiatric problems i.e. depression or severe anxiety 9. Resolve guilt feelings. Marigal ual Researc relationship during the t has been don Survivor at These studie the death of marital rela aSSOCiated w Carey ( 28‘70’ thirti folmd a Curvi happiness CHAPTER I I R vi w f h Li r r Emmi QuLlth Researchers have yet to thoroughly examine the relationship between marital quality and the wife's adjustment during the terminal illness of her spouse. However, research has been done concerning marital quality and adjustment of the survivor at various points in time after the spouse's death. These studies strongly suggest that differential responses to the death of a spouse are partially a function of the prior marital relationship. In general, better marital quality is associated with better coping by the surviving spouse. Carey (1980) interviewed 119 widows and widowers, aged 28-70, thirteen to sixteen months after they were widowed. He found a curvilinear relationship between self-reported marital happiness and scores on an 8-item measure of adjustment- depression. Those who reported that their whole marriage had been happy were "better adjusted" (less depressed) than those who said that there had been some period of unhappiness in the marriage. Widowed persons who reported chronic and serious problems in the marriage were not depressed and appeared to be functioning well. The author speculated that this last group of people may have experienced relief that their severe marital problems were finally over. Singer (1982) found that widows who reported a large number of marital problems were also experiencing significantly more grief 10-16 months after their husband's 12 death than H Winn (‘5‘- anxiety, stre and widowers indicative 0' associated u freerrent mar; relationship (p<.05). In an ir 0f grief, t} widows and w; interviews bereaVEment . issues or ma; as either "H 13 death than were widows reporting fewer marital difficulties. Winn (1981) obtained self-ratings of physical health, anxiety, stress, depression, fatigue, and guilt from 49 widows and widowers after one year of their bereavement. Scores indicative of poor adjustment on each of these measures were associated with reports on the Dyadic Adjustment Scale of frequent marital quarrels, low satisfaction with the marital relationship, and disagreements by the couple about sex (p<.05). In an investigation of factors that influence the course of grief, the Harvard Bereavement Study followed 68 young widows and widowers (under age 45) longitudinally, conducting interviews at 3 weeks, 6 weeks, and 13 months following bereavement. Parkes (1975) reported some findings relevant to issues of marital quality and adjustment. Marriages were rated as either "High" or "Low'Conflict" based upon responses of the surviving spouse to eleven questions assessing various aspects of the marital relationship. Persons who reported repeated problems in two or more areas were considered to be in the "High Conflict" group. Subjects in the "Low Conflict" group reported less than two problem areas. An overall outcome score was obtained by combining assessments of a number of aspects of adjustment, such as physical health, anxiety/depression, self-evaluation, guilt, social withdrawal, and acceptance of reality. Parkes found that at 13 months following bereavement 61% of the "Low Conflict" group could be rated as having a "good outcome" compared to 29% of the "High Conflict" group (p(.05). At follow-up, two to four years after bereavement, subjects in t depressed, I physical he. comparison t< provide stat subjects who: were rated a HadlSOn informat 10m marriage alwa SurviVing 14 subjects in the "High Conflict" group were significantly more depressed, reported more anxiety and guilt, were in worse physical health, and were more socially withdrawn in comparison to the "Low Conflict" group. Although he did not provide statistical data, Parkes reported that bereaved subjects whose marriages had been openly hostile generally were rated as having "good" outcomes. Madison (1968) concluded (without providing statistical information) that one should not assume that a disturbed marriage always results in a bad bereavement outcome for the surviving spouse. Questionnaire information was obtained concerning the physical and mental health of 132 widows during the 13 months following bereavement. Twenty widows who were rated "bad outcome" were compared with 20 "good outcome" widows. The author found that some women who were functioning well came from marriages that totally lacked intimacy or which were openly recognized as failures. However he also reported that the poor outcomes of some widows seemed to be related to previous problems in the marriage. In conclusion, it appears that surviving spouses of marriages that were relatively happy and low in conflict generally are functioning well in terms of physical and psychological health one year after the bereavement, and beyond. Although the results are less clear for widowed from problematic marriages, there seems to be a greater risk among this group for suffering a "bad outcome". However, widowed persons from marriages with extreme difficulties generally do appear to at death . Parkes the poor lo: marriages we marriages ma- °£ “0138 for ' that Persons peeple Who, bu“dine sati others of re. in Construct unsatisfying in oneself < social conta HOWeveI feelings t 15 appear to attain "good" outcomes one year after the partner's death. Parkes (1983) discusses a number of theories to explain the poor long-term outcomes seen among the bereaved whose marriages were "conflictual". Widowed persons from problematic marriages may not only mourn the deceased, but also the loss of hape for improvement in the relationship. It might also be that persons who form and remain in troubled marriages are people who, because of earlier experiences, have problems building satisfying attachments. They may be less capable than others of recovering from loss by forming new attachments or in constructing a supportive social network. Similarly, an unsatisfying marriage may teach or reinforce a lack of trust in oneself or others, making the widowed person fearful of social contacts and hindering recovery. However, Parkes believes that the survivor's ambivalent feelings toward the lost partner are a key to understanding the relatively "poor" long—term outcomes found among widows and widowers whose marriages had been high in conflict. He argues that in marriages with many problems, anger and the desire t0pportunity for anticipatory grief to occur. In the case of a sudden de surviving unpredic relation a <. relatio' functic focused adjust. more t SPOUSe theSe adjus diffj illnq 1932 whet Gem Qt. cgr be1 th. "1 me Dr 40 sudden death, there is no chance to prepare for the loss. The surviving spouse may henceforth perceive the world to be an unpredictable and risky place; and she may feel as if relationships are insecure and dangerous (Parkes, 1983). A good deal of research has been done investigating the relationship between the length of a terminal illness and the functioning of the surviving spouse. Most of the studies have focused on comparing adjustment following "sudden death" with adjustment to a death following a final illness that lasted more than a few days, and one could assume that the surviving spouse had some warning and time to prepare. The results of these studies are inconclusive. Some studies find that adjustment to bereavement following a "sudden death" is more difficult than adjustment to a death following a more lengthy illness (Carey, 1980; Sanders, 1983; Vachon, Rogers et. al., 1982). Other studies however, found little or no difference, whether or not the surviving spouse had time to anticipate the death (Bornstein et. al., 1973; Clayton et. al., 1971; Clayton et. al., 1972; Clayton et. al., 1973; Madison & Walker, 1967). As Ball (1977) and. Balkwell (1981) point out, the contradictory findings may be explained by an interaction between age and mode of death (sudden versus forewarning) . All the previous research cited employed samples that included a wide age range of subjects, and did not analyze adjustment to mode of death differentially by age. Researchers who studied only the younger (under age 46) widowed.consistent1y find that having had some period of forewarning prior to the death is associated with better adjustment at all points throughout widow? 1984al Blancl adjus that: young forev cons surv fore Arkj ver: dif Sur lit be] ap] ad me Cc 8: 41 widowhood (Ball, 1977; Glick, Weiss, & Parkes, 1974; Lundin, 1984a, 1984b; Parkes, 1975). One exception (Blanchard, Blanchard, & Becker, 1976) reported few differences in adjustment whether or not there was forewarning, but did note that suicidal ideationwwas significantly'more common.among the young widowed after a "sudden death" than when there was forewarning (p<.03). In studies of the older bereaved, findings are also consistent. Few differences are observed in adjustment among survivors of a sudden death, as compared to those who had forewarning (Parkes, 1964; Gerber, Rusalem, Hannon, Battin, & Arkin, 1975). It is suggested therefore that mode of death (sudden versus forewarning) affects response to bereavement differentially depending on the agewof the survivor. For older survivors, having some forewarning of the death seems to have little effect on their later adjustment. Among the young bereaved however, having some time to anticipate their loss appears to have a significant, positive effect on later adjustment. The only study that compared the impact of age as well as mode of death on adjustment (Ball, 1977) obtained results consistent with these conclusions. For younger widows, having some period of time to anticipate their spouses death was associated with a significantly more mild grief reaction, 6-9 months following bereavement than was the case for widows c0ping'with the "sudden.death" of‘a spouse (p<.05). No similar relationship was found for older widows. Num- the loss severely living, loss is shocking adjustme forevarn some tim of 1055 used the With the their 8; It not as s the dea suddenly 42 Numerous studies (previously cited) clearly indicate that the loss of a spouse generally impacts a young person very severely. There typically' is great disruption. of :normal living, much more so than for an older individual. When the loss is also a sudden one, it most commonly is a truly shocking event for the young spouse. The relatively better adjustment observed among the young bereaved who had some forewarning of their spouse's death, indicates that having some time to prepare is beneficial, possibly because the shock of loss is reduced somewhat. The young bereaved may also have used the time prior to the death to begin to come to terms with the loss and accept its reality, and also to resolve with their spouse any problematic marital issues. It is likely that among the aged, the loss of a spouse is not as shocking or as disruptive an event, possibly because the death is never really unexpected, even if it occurs suddenly. It has been suggested that older people prepare themselves psychologically for the loss. of their spouse (Cumming & Henry, 1961) and for the role of widowhood (Heyman & Gianturco, 1973) before the event actually occurs. There would therefore be less need among the aged for a period of time prior to the death to prepare for the loss. -.- , . _ n~:_ . - - ~rr r _0n°--'su 1 . Few researchers have compared adjustment to short-term, chronic illnesses - illnesses lasting a few weeks to a few months - and adjustment to long-term, chronic illnesses, in which the final illness lasts more than a few months. It has been speculated (Aldrich, 1974; Ball, 1977; Gerber, 1974; Randc prepa lengi for emot here illn surx stra adj' lon Hid tel In 43 Rando, 1984; Sanders, 1982) that.at some point, increasing the preparation time prior to a death will cease to be helpful. A lengthy chronic illness, rather than helping a spouse prepare for widowhood, might strain or deplete the survivor's emotional and physical resources, leading to a negative post- bereavement outcome. It would be expected that a lengthy illness would have its greatest negative impact on aged survivors, who might be less physically able to handle the strain. In general, researchers have found better post-death adjustment to be associated with short-term, as opposed to long-term final illnesses. In one study (Gerber et. al., 1975) of older widows and widowers (mean age: 67 years old), adjustment was measured in terms of physical health during the 6 months following bereavement. It was found that those bereaved whose spouses died after a lengthy chronic illness (more than 6 months duration) did worse on all three medical variables (# of physician visits, # of times ill without contacting a physician, and # of psychotropic medications taken) than did widows and widowers whose spouses died after illnesses lasting between 2 and 6 months. The authors concluded that, at least for the aged bereaved, an extended illness prior to the spouse's death may place the surviving spouse at relatively greater risk for negative health outcomes. Similar results were found in a sample of older surviving relatives, some of whom had lost a relative other than a spouse, when grief reaction was measured one year after the death (Schwab, Chalmers, Conroy, Farris, & Markush, 1975). When il relativ lasted reactic Tt outcome (Sande: age<§ a pare grief again death mOnths those mOnth: more , than grief chr0n °°mp1 re1at f°llc innE disc] whic1 44 When illnesses lasted more than one year, 68% of surviving relatives showed. intense ‘grief' reactions. When illnesses lasted less than one year, only 30% showed a severe grief reaction. The most recent investigation comparing bereavement outcome in short- versus long-term chronic illness survivors (Sanders, 1983) employed a relatively young sample (average age < 55 years old) and included individuals who had also lost a parent or child, as well as some widows and widowers. When grief reaction and adjustment were measured at 2 months and again at 18 months following the death, the survivors of a death following a short-term chronic illness (less than 6 months duration) appeared on average to be doing better than those bereaved following an illness lasting more than 6 months. At two months, the "long-term chronic" group suffered more social isolation, loss of energy, and physical symptoms than did the "short—term chronic” group, although levels of grief experienced were similar. At 18 months, the "long-term chronic" group was more depressed, had more somatic complaints, and less energy than did the "short-term chronic" group. The three studies cited are consistent in finding relatively negative outcomes to be associated with bereavement following a long-term chronic illness, as opposed to a chronic illness of shorter duration. One study that obtained results discrepant from these studies was Clayton et. a1. (1973), in which the authors compared the frequency of a variety of psychiatric symptoms in a sample of 81 widows and widowers. Some illnes spouse than 1 diffe: the t irrit among (p<.0 widow betwe 9r0u; "lens grou] You: phyS 0f 5 0f t itsE Has assc mgrQ the the 45 Some widowed persons lost their spouses after "lengthy" illnesses (more than 6 months duration), and some lost their spouses after illnesses lasting more than 5 days, but less than 6 months ("short" illnesses). The authors found little difference between the two groups in symptoms occurring during the terminal illness. Only 1 of 20 psychiatric symptoms -— irritability - was reported significantly more frequently among those whose spouses died after more lengthy illnesses (p<.01). The authors concluded that prevalence of symptoms in widows and widowers was unrelated to length of illness. However, a comparison of the subject characteristics between the two groups indicates that the "short illness" group, on average*was significantly older than subjects in the "lengthy illness" group (p<.01). Although subjects in both groups were relatively aged, it may have been that the "younger" subjects in the "lengthy illness" group were physically more able t0*withstand the strain of a long illness than an older sample might have been. There is a need for further research to assess the impact of short— versus long—term chronic illness on the adjustment of the surviving spouse during the time of the final illness itself. The present research made this comparison. First, it was expected that relatively' better adjustment. would Zbe associated with a short-term chronic illness, as opposed to a more lengthy chronic one. This prediction was consistent with the results (previously cited) of studies of the adjustment.of the surviving spouse following bereavement. These studies found negative outcomes to be associated with a lengthy termi has prolo and p their was e simil survi anxie COpin Partn the a illne evide. dePre. young. liter. 46 terminal illness, as opposed to a short-term chronic one. It has been suggested that the strain of coping with the prolonged illness of one's partner can deplete the physical and psychological resources of the widowed person, increasing their vulnerability to physical and psychiatric problems. It was expected that the strain of a prolonged illness would have similar effects during the final illness itself, with surviving spouses having more problems with depression, anxiety, guilt, and physical symptoms, relative to spouses coping with chronic illnesses that are of more brief duration. An interaction was expected between the length of the partner's illness ("short-term" or "long-term" chronic) and the age of the surviving spouse. For "short-term" chronic illnesses, it *was Ihypothesized that. older spouses ‘would evidence better adjustment, in terms of the presence of depression, anxiety, guilt, and physical symptoms, than would younger spouses. This prediction was consistent with the literature (previously cited) which clearly indicates that in general, younger spouses suffer more severe reactions to the loss of their partners than do older spouses. However, for "long-term" chronic illnesses, opposite results were predicted. It was hypothesized that younger spouses would evidence better adjustment - fewer problems with depression, anxiety, guilt, and physical symptoms - than would older spouses. It was expected that younger spouses, who are likely to be physically stronger, would be better able to withstand the strain of coping with a partner's prolonged illness, more so than would older spouses. resx hav- att adj rel ac< of th 47 n ' A n f T in l I ln 5 Prior to this investigation, the relationship between a wife's adjustment during her husband's terminal illness and her husband's attitude toward his illness - whether or not he is able to acknowledge the likelihood that the illness will result in his death.- had.not been investigated. A few studies have examined the relationship between the dying spouse's attitude toward his/her illness and the surviving partner's adjustment following the death. These studies suggest that relatively better adjustment is associated with an attitude of acceptance by the dying partner regarding the terminal nature of his/her illness, and acknowledgement of this attitude to the surviving partner. The process by which the existence of a terminal illness is or is not acknowledged and discussed by a couple is almost surely one of mutual influence between husband and wife. If the surviving spouse is able to acknowledge the likelihood of the partner's death, believes that the open.discussion.of this issue with the partner is important, and feels comfortable doing so, it seems likely that the dying partner might be influenced in the direction of greater acceptance and acknowledgement of his condition. Similarly, the dying spouse who can accept the likelihood of his own death and communicates this to his partner, will likely influence the surviving spouse toward greater acceptance of the reality of the illness. Since this research is concerned with adjustment during the terminal illness of the surviving spouse, and factors that may influence adjustment, this discussion will focus c dying possib A not co becaus inforn (1981) Progre evasiv likel: Patie1 someo: infor 0ther hatUr "arni theiI With 48 focus only on the possible influence that the attitudes of the dying partner may have on the survivor, rather than on any possible reciprocal influence. Among the terminally ill, there are individuals who do not communicate awareness of their condition to their spouse because they are themselves unaware, having never been informed.by anyonelof the seriousness of their illness. Hinton (1981) interviewed. 80 patients in the latter stages of progressive cancer and found that a number had received evasive or overly optimistic prognoses from staff. It is likely common for medical personnel, who seek to protect a patient or are uncomfortable with the idea of informing someone that he/she is terminally ill, to withhold such information, or to provide a vague or inaccurate prognosis. Other terminally ill people have been informed regarding the nature of their illnesses, but have refused to accept the warnings, They deny that their illnesses will likely result in their deaths. They do not discuss their prospective deaths with their partners. In the Hinton study, 18 of 80 patients were unable to demonstrate any awareness of the terminal nature of their conditions. It is unclear, however, how many patients were unaware because they had not been warned, and how many had been informed of their terminal prognoses, but refused to accept the warnings. Finally, it appears that some individuals who are aware that their illnesses will result in their deaths, choose not to communicate this awareness with their spouses. Among the 62 subjects in Hinton's study who were aware of the nature of their conditions, 19 never discus: E‘ avaren guaran likeli patie1 illne mutua like] Patt1 more main Chal to ( whi. the was ac) me we tt 49 discussed their awareness with their partners. Even when the dying spouse does decide to communicate awareness of his/her condition to his/her partner, it does not guarantee that a realistic discussion concerning the likelihood of death will occur. Among the 43 couples in which patients did communicate some awareness of their terminal illnesses to their spouses, in only 22 cases was there a mutual acknowledgement between the couple that the patient was likely to die. The other 21 couples seemed to fall into 3 patterns: 1. the patient's comments were vague, and little more was said. 2. the patient was clear but the spouse maintained ignorance. 3. the patient was clear but the spouse challenged the patient's appraisal. These results were similar to those found by Vachon et. al. (1977) among 162 couples in which the husband was terminally ill with cancer. In 61% of the couples, no discussion concerning the prospective death was ever held. Among 29% of the couples, there was a mutual acknowledgement that death was likely. Krant and Johnston (1978) found that 22% of the family members of terminally ill patients (about one-third of whom were spouses) had discussed dying with the patients prior to the deaths. Clearly, a variety of communication patterns characterizes couples coping with a terminal illness in one of its members. In only a minority is their a mutual discussion and acknowledgement of the likelihood of death occurring. It appears from Vachon et. al's (1977) study that post-death adjustment may be facilitated by a couple's discussion of dying p a discw report bereavt dying 1 bereav discus 1 0f dea may be survi‘ ways the d it mo the acknc Coup' "mar mari carI N 90c in) Dre tha his 11119 Guy 50 dying prior to the death. Widows who reported having had such a discussion with their husbands were very likely to also report that their discussion helped them with their bereavement. Conversely, many widows who did not talk about dying with their husbands during the illnesses believed that bereavement was made more difficult by the lack of such a discussion. If the dying partner fails to acknowledge the likelihood of death resulting from his/her illness, the surviving spouse may be inclined also»to deny that the illness is terminal. The surviving spouse might then fail to prepare for the death in ways that might reduce its impact (i.e. by grieving prior to the death, by formulating an explanation of the loss to make it more understandable and acceptable, or by making plans for the future after the spouse's death). The lack of acknowledgement and conununication about dying between the couple could lead to a failure to successfully perform "marital tasks" of this period - resolving any conflictual marital issues, reviewing and summing up the marriage, carrying out any final wishes of the dying spouse, and saying "good-bye". The current research assessed the possible impact of the dyinngartner's attitude towards his terminal illness upon the pre—death adjustment of the surviving spouse. It was expected that a dying person who acknowledged the terminal nature of his illness and communicated this attitude to his partner, might have facilitated the adjustment of the surviving spouse during the pre-death.period. There is some evidence to support this 1 havin more getti berea data other mutui a di: trie like John him/ memt (p<. true Pati imp, 'P< att 39o and 91‘s 51 this prediction: those bereaved in Hinton's study who reported having had a discussion with their partners about dying, were more likely to rate the relationship with their spouses as getting closer since the illness began, than were those bereaved who reported limited or no discussion (p<.05). This data does not indicate however, whether one spouse or the other initiated the discussion or if the discussion was mutually initiated. Important though is the evidence that such a discussion might be beneficial. If the dying partner acknowledges his/her condition and tries to communicate this to the surviving spouse, the likelihood of such a discussion is increased. Krant and Johnston (1978) found that the more a patient perceived him/herself’as very sick.or dying, thermore likely were family members to talk together' with that patient about. dying (p<.001). Hinton found that the reciprocal situation was also true - when the surviving spouse believed that informing the patient of the terminal nature of his/her illness was important, discussions of dying were much more likely (p<.001). It was expected that the pre—death adjustment of the surviving spouse would be influenced by the dying partner's attitude towards his terminal illness. Specifically, it was hypothesized that relatively better adjustment among surviving spouses prior to the deaths - greater acceptance of the terminal nature of the illnesses, less depression, anxiety, and guilt; fewer and less severe physical problems, and greater completion of marital tasks and meaningful grief work - woul termi: ackno‘ More] relat dying termi of t SHIV: Part1 Near impo PEIi task by 1 meat SPOI Sur. and fin Suf ass Qui ack “it fir 52 - would be associated with an attitude of acceptance of the terminal nature of the illness by the dying spouse, and acknowledgement of this attitude to the surviving partner. More problematic adjustment would be associated with a marital relationship in which there was no acknowledgement by the dying spouse that he was suffering from an illness that was terminal in nature. As previously discussed, the dying partner's acceptance of the probability of death is likely to influence the surviving spouse to develop a similar attitude regarding her partner's illness. Mutual acknowledgement.by the couplezof the nearness of death is likely to facilitate their awareness of important marital issues that are characteristic of this period, ultimately leading to greater completion of marital tasks. In addition, acknowledgement of the terminal condition by the surviving spouse might facilitate the performance of meaningful anticipatory grief work, because of the surviving spouse's heightened awareness of the prospective loss. If the surviving spouse is able to work through some of the emotions and issues associated with the loss during the time of the final illness, that spouse should be at diminished risk for suffering a number of negative outcomes that have been associated with the failure to grieve: depression, anxiety, guilt, and physical problems. It was also hypothesized that the dying partner's acknowledgement of his terminal condition would be associated with more intense grief by the surviving spouse during the final illness, relative to spouses whose partners did not acknow surviv nature aware losses to m01 inten: illne: are d: 53 acknowledge that they were dying. It is suggested that as surviving spouses attain greater acceptance of the terminal nature of their partners' conditions, they also become more aware of feelings and issues associated with the prospective losses. Their greater awareness of the losses will likely lead to more working through of grief issues, but also to a more intense anticipatory grief reactions during the final illnesses, relative to spouses who deny that their partners are dying and who are less aware of their prospective losses. TI design hospic the by l retro: inves ades illne spous PrOs] here aCut con, the. the for Six QHAETER I I I Deeign 9f the §tudy This chapter contains. a :rationale for the) research design, and descriptions of the subject sample, the nature of hospice services, the measures used, validation procedures, the hypotheses, and the statistical analyses conducted. R i n l f r R r e iv R r h D i n Because of practical and ethical concerns, a retrospective research design was chosen for this investigation over a design in which data on the survivor's adjustment would be obtained during the final illness itself. As previously discussed, the time of the husband's final illness typically is a very stressful period for the surviving spouse, who often experiences an acute grief reaction to her prospective loss. Similarly, the early months following bereavement have been described (previously) as a time of acute psychological distress and disruption of normal functioning. In consideration of their feelings, widows were not contacted during their husbands' final illnesses; nor were they approached until a minimum of six months had passed since their husbands' deaths. Grief researchers generally agree that for most bereaved persons, acute grief has subsided by the sixth month of bereavement (Ball, 1977; Clayton et. al., 1968; Krupp, 1962; Lindemann, 1944; Parkes, 1965a). Because some time had elapsed since the husband's final 54 ir 51' di EV Hi fo an au si fe Si “h re1 ad, ho, and pd: pd:- 55 illness occurred, and this is the period about.which the widow was asked to report, a methodological issue arises as to the validity of the self-report data obtained in this investigation. It might be speculated that as time passes since the loss, the intensity of symptoms experienced may diminish in a widow's memory. There is however, research evidence that this is unlikely to occur. Blanchard et. al. (1976) asked 30 widows to recall how often they experienced each of 20 depressive symptoms soon after their husbands' deaths. At the time of the interview, the mean time that had elapsed since the death was 7.1 years, with.a range of 1 to 25 years. No significant relationship was found between the frequency of depressive symptoms reported and the length of time that had passed since the death. The authors speculated that "the death of husband is such a significant crisis event for a widow that her memory of her feelings of each stage remains clear and vivid in her mind" (p.396). Because a husband's final illness also represents a significant crisis, it is suggested by this investigator that a widow is also likely to have a clear memory of experiences which occurred during that period. One important practical consideration in choosing a retrospective research design was the availability of an adequate subject sample. For this investigation, Michigan hospices were the only feasible source of potential subjects, and hospice workers are understandably protective of participants in their programs. One program refused participation on these grounds. Because the time of a husl for hos sub occ ili de IL” 56 husband's terminal illness is typically one of extreme stress for the surviving spouse, it was considered unlikely that hospices would permit the investigator to contact potential subjects during the time of the final illnesses. Moreover, it was considered probable that a higher rate of refusal would occur among potential subjects during the time of the final illness itself, as opposed to six to twelve months after the death, when acute grief was likely to have subsided. For the reasons discussed, a retrospective research design was chosen for this investigation. ms; The subject sample consisted of a recently—widowed group of women. The criteria for inclusion as a subject was as follows: an individual must be female and have become widowed no less than six and no more than twelve months prior to completing the research materials; the husband's final illness must have been.of at least one month's duration; and.a subject must have participated in a hospice program. Hospice programs were attractive sources of potential subjects because hospices typically can access significant numbers of widowed individuals from among their clients. In addition, it was considered likely that widows who participated in hospice programs might be more cooperative about taking part in a research project than would widows who were not hospice-involved. The former group might agree to participate as subjects in gratitude for the help the hospices have provided them. Moreover, a widow might be more likely to participate in a research project that has been introduced to her sue] the Gra Jac Mic im th to Ho ea Hi Tl 57 her by someone with whom she is familiar and whom she trusts, such as a hospice worker. In order to obtain subjects, the investigator contacted the following groups: (a) Hospice of Lansing (b) Hospice of Grand Rapids (c) Hospice of Battle Creek (d) Hospice of Jackson (e) Hospice of Garden City (f) Hospice of Southeastern Michigan. These organizations agreed to provide the investigator with access to potential subjects. Procedures differed somewhat among the hospices as to how the subject questionnaire was to be administered. The following procedures applied to all the hospices except the Hospice of Southeastern Michigan. Between six and twelve months after the spouse's death, each subject was contacted by the hospice volunteer working with the family, or by the hospice bereavement coordinator. The subject was informed that the investigator was conducting a study of a spouse's response to her partner's terminal illness. The hospice worker also informed the subject that participation in the study involved the completion of a self- report questionnaire packet requiring about 1-2 hours of the subject's time. If the subject agreed, the hospice worker scheduled an appointment for the subject to complete the written questionnaire. Four hospices preferred to have the hospice worker go to the subject's home with the questionnaire. The Grand Rapids hospice preferred that subjects come to the hospice to complete the questionnaire and do so as a group. In either’ case, prior’ to completing' the questionnaire, the 58 subject was given an explanatory letter (see Appendix B pgs. 218-219) describing the study and what was expected of the subject should she agree to participate. A hospice worker was available to answer any questions the subject may have had about the study. If she agreed to participate, the subject signed the research consent form (see Appendix B, pg. 220) required by the Human Subjects Committee, and then completed the written questionnaire (see Appendix B, pgs. 221-251). The Hospice of Southeastern Michigan preferred to make the initial contact with subjects by letter (see Appendix B, pgs. 256-258) rather than by phone. Women who were willing to participate filled out a post card (provided) and returned it to the investigator. After receiving the post card, the investigator contacted the widow by phone to set up an appointment to complete the consent form and the questionnaire. Depending on the subject's preference, the questionnaire was either administered at the subject's home or at the hospice. In all cases, the investigator was the person who administered the questionnaire, without anyone from the hospice being present. H i rvi The 47 women who completed the subject questionnaires are, in at least one respect, atypical from most women who have lost their husbands - every widow in this study participated in a hospice program during the time of her husband's final illness. Since this investigation is concerned with the adjustment of surviving spouses, and factors which influence adjustment, and because participation in a hospice pro of nat pro num imp den prc the con COIT Prc the for Suz abi ill 59 program itself may be a significant factor in the adjustment of a surviving spouse, it is important to understand the nature of the services provided to families by hospices. During the time a family participates in a hospice program prior to a spouse's death, the family is offered a number of services that potentially, can have a positive impact on the ability of a surviving spouse to cope with the demands and stresses of the final illness. Hospices typically provide families with significant medical assistance during the illnesses, particularly in terms of pain management and control, with the goal of maximizing the patient's physical comfort. Medical care is provided by a nurse who is assigned to the family, and whose work is supplemented by a home health aid who assists in many non-medical aspects of physical care, such as physically moving, bathing, or shaving the patient. Volunteer para-professionals are available to provide respite help to family members who care for the patient. A volunteer might help with housekeeping or by staying with the patient while the spouse shops or goes to her job. When a family first begins participation in a hospice program, a social worker conducts an in-depth assessment of the psychosocial needs of various family members, and formulates a plan for meeting those needs. With regard to the surviving spouse, the social worker will assess the wife's ability to perform important tasks of the dying phase of the illness, as well as factors which might impact the spouse's capability of carrying out these tasks (e.g. social support, financial problems, other crises, the marital relationship, reli the 35S pa: mu on it 60 religious and cultural attitudes, past losses). When needed, the social worker will offer help, including counseling, to assist the surviving spouse in accomplishing necessary tasks. The degree to which a hospice is involved varies greatly among families. Much depends on when a family decides to participate in a hospice program. Obviously a hospice will be much less involved with a family that seeks hospice services only during the latter stages.of a terminal illness, than with a family that begins participation at the beginning of a lengthy illness. Once a family does decide to participate, level of involvement by the hospiceris determined.by the needs of the patient and the family, and by the family's desire for particular services, both of which may vary greatly among families. Moreover, hospice involvement often varies a great deal within a particular family, as the final illness progresses. Typically, there is a greater need for hospice services as the death approaches. The comprehensive nature of the services offered by hospices suggests that participation in a hospice program ought to be considered as an important factor in evaluating the adjustment of women to the terminal illnesses of their husbands. W The thirteen measures used in this study are discussed in this section. Measures of three of the independent variables - marital adjustment, social support, and the dying spouse's attitude towards his terminal illness - will be reviewed fir: use- var con Spc ac CC 61 first. Next, there is a discussion of the nine instruments used to measure the various aspects of the main outcome variable - adjustment of the surviving spouse. These are: completion of marital business, ability to care for the dying spouse, ability to carry out other responsibilities, acceptance/denial of the husband's terminal condition, completion of meaningful grief work, presence and severity of physical symptomatology, intensity of guilt, level of depression, and severity of anxiety. Finally, the measure of thelother'outcome'variable, grief intensity, will be reviewed. M i A u m n The Dyadic Adjustment Scale (DAS) (Spanier, 1976) was selected as the measure of marital quality for this investigation. The DAS is widely used, and is considered to be a psychometrically strong instrument. The measure consists of 32 items, selected to provide a broad range of criteria related to marital adjustment. The DAS yields a total score that can vary from 0 to 151, with higher scores indicating better marital adjustment. Factor analysis of the DAS (Spanier, 1976) resulted in the following four factor scales: 1. Dyadic Cohesion, 2. Dyadic Consensus, 3. Dyadic Satisfaction, 4. Affectional Expression. However, Sharpley and Cross (1982), in their own factor analytic study of DAS items, found only one underlying "adjustment" dimension. For a sample of 312 subjects, Spanier (1976) found the overall reliability of the DAS to be .96, using Cronbach's Coefficient Alpha as the measure of internal consistency. 0t) £01 pr: in ma it p1 62 Other studies (Sharpley'a Cross, 1982; Spanier’a Thomas, 1982) found the overall reliability to be .91 and .96 respectively, providing further evidence that the DAS is a highly reliable instrument. Concurrent validity was established with samples of 218 married and 94 divorced subjects (Spanier, 1976). For each item of the DAS, the mean score differed significantly in the predicted direction between the married and divorced samples (p<.001). In addition, the mean total score was significantly different between the two groups (p<.001). Using the same samples, construct validity was tested by correlating the DAS with the Locke-Wallace Marital Adjustment Scale (Locke 3. Wallace, 1959). The two measures correlated .86 for the married subjects, .88 for the divorced subjects, and .93 for the total sample (p<.001). The DAS is listed in the Questionnaire Packet (Appendix B) on pages 247-251, items 1-32. Scoring for the DAS is shown on pages 267-269 of the Scoring Key (Appendix C). Wager: Emotional support (empathy, understanding, and sympathy) has been identified as the primary need of a surviving spouse during the time of her husband's final illness. Therefore, the measurement.of social support.in.this investigation.focused.on the degree to which emotional support was available to the subject. Two aspects of emotional support were identified by Schneider, Winder, and Peradotto: whether the subject had someone in whom she could confide about herself and problems, ar Th WI re: the 260 F3 :3" /. 63 and.more specifically, whether there was anyone the surviving spouse talked with about her husband's illness and his dying. The Social Support measure consists of the following two written, self-report items: 1. There was someone particular (other than my husband) in whom I confided or talked to about myself or my problems. 2. There was someone particular (other than my husband) in whom I confided or talked to about my husband's illness or his dying. A subject was asked to rate each item based on her response during the entire time of the final illness, using the following system: 0 2 did not describe me 1 2 described me somewhat 2 - mostly described me 3 = accurately described me Scoring for the Social Support measure is shown on page 260 of the Scoring Key (Appendix C). h- D n- ..u~-'_ A i d- Tow.ro: Hi T-rmin. .n. i-n It was hypothesized that relatively better adjustment by the surviving spouse might be associated with acceptance and acknowledgement by the dying spouse of the terminal nature of his condition. In order to assess the surviving spouse's perception of her husband's acceptance/denial, two written, self-report items were generated: 1. By that time, my husband believed that he would die from his illness. He had little or no hope for recovery. The subject was asked to rate her husband's 64 acceptance/denial according to the following system: did not describe him described him somewhat mostly described him accurately described him I did not know whether my husband believed he would survive the illness. IbUN-‘O IIIIIIIII 2. My husband encouraged me to discuss with him my feelings about his dying. The subject was asked to rate her response using the following system: 0 = did not describe him 1 a described him somewhat 2 a mostly described him 3 - accurately described him It is not considered necessary that the dying spouse accept and acknowledge his terminal condition throughout the entire course of his illness, in order to have a positive influence on his wife's adjustment. In some cases, a husband might not acknowledge the possibility of his own death until fairly near its occurrence. However, if the husband's acceptance of his terminal condition is to help move his wife towards greater acceptance of that fact, leading to her accomplishing important marital and grief work prior to the death, then the husband's acceptancelacknowledgement must occur at a point in the illness that allows time for his wife to perform these tasks. In other words, the husband's acknowledgement of his terminal condition must occur sufficiently before his death if it is to have any significant positive influence on his wife's adjustment. Consequently, each subject was asked to report her husband's attitude towards his illness - his acceptance/denial of his condition - 65 for the time two weeks prior to his death. Scoring for the Husband's Acceptance Scale is shown on page 267 of the Scoring Key (Appendix C). Qempleting Merieel Werk The Marital Work measure was developed in order to assess the degree to which the surviving spouse was able to complete important marital work with her husband prior to his death. Initially, Winder, Schneider, and Peradotto identified a number of significant marital tasks which are thought to be characteristically faced by a couple when one of the spouses has a terminal illness: resolution of any remaining marital problems, sharing awareness.of and feelings about the spouse's terminal condition, reminiscing about and summing up the marriage, planning for the surviving spouse's future, carrying out any last wishes of the dying partner, and saying final "good-byes". Thirteen written self-report items were generated from these concepts. The subject was asked to rate each based on her response during the entire time of her husband's final illness. The exception is item #1, which asks the surviving spouse to rate the extent to which there were unresolved marital issues just prior to her husband's final illness. Eight items are related to the possibility of a discussion by the couple of each of the following issues: resolution of marital problems, sharing awareness of the husband's terminal condition, planning for the future, and reminiscing/summing up the marriage. A discussion by the couple is likely to be indicative of the completion of marital 66 work around the particular issue discussed. Therefore, for each of the issues listed, the widowed spouse was asked whether or not aidiscussion.by the couple occurred, and if so, who mostly spoke. Finally, the subject was asked how she felt the discussion went. In regard to the issue of the resolution of marital problems, it is recognized that in some cases, a surviving spouse might tend to idealize her relationship with her husband. A subject might report that there were no marital problems to be discussed, when in actuality, unresolved marital issues did exist. To assess the possibility that a surviving spouse might have failed to discuss marital issues with her husband because she denied the existence of real problems, the following two items were chosen from the Dyadic Adjustment Scale (Spanier, 1976) as a brief measure of denial: 21. How often did you and your husband quarrel? 22. How often did you and your husband "get on each other's nerves"? A subject indicated.her response to these items using the following system: 0 never rarely occasionally more often than not most of the time all the time Ululwa-fi IIIIIIII The investigator suggests that is extremely unlikely for a couple never to have quarreled or never to have "gotten on each other's nerves" during the entire marriage. It is further suggested that any subject who answers "never" to both items, can be seen as engaging in denial regarding the existence of pr ot wc IE 67 problems in her marriage. It was decided that any subject who obtained a score of 0 on both of the above-mentioned DAS items would receive a score of 0 for any items having to do with the resolution of marital issues. Two items concerned the issue of the husband's last wishes. It is important to understand the reason a surviving spouse might have been unable to carry out her husband's final wishes before his death. The dying partner may not have had any last wishes. In other cases, the carrying out of any last wishes might not have been possible until after the husband's death. Finally, the surviving spouse might have failed to carry out her husband's last wishes because she felt that his requests were unreasonable. Therefore, those subjects who did not report having carried out their husband's last wishes, were asked why they were unable to do so. The Marital Work Scale is listed in the Questionnaire Packet (Appendix B) on page 223 and pages 225-228, items 1-5a, and items 8-11. Scoring for this scale is shown on pages 260- 262 of the Scoring Key (Appendix C). n R n i i i As discussed in a previous section, one aspect of adjustment during a partner's final illness is the performance by the surviving spouse of various tasks for which she is responsible. The Carrying Out Responsibilities measure was developed to assess the subject's ability to perform the following three responsibilities: care of the children (if there were dependent children living at home), job performance (if the surviving spouse held a job), and performance of 68 household tasks. For each area of responsibility that was applicable, a subject.was asked to rate her performance during the entire time of her husband's final illness using the following system: A. I was able to perform these tasks better than prior to his illness. B. My level of performing these tasks was about the same as before this illness C. My performance on these tasks was somewhat less during the illness than prior to the illness. D. My performancerof these tasks was significantly less during my husband's illness than prior to the illness. E. During my husband's illness, I was able to do almost nothing in the way of performing the tasks that.I did prior to his illness. The three items of the Carrying Out Responsibilities Scale are listed in the Questionnaire Packet (Appendix B) on pages 228-229, items 12-14. Scoring for this scale is shown on page 262 of the Scoring Key (Appendix C). £hx§igel gere gf ghe Dying SQQQ§e The Physical Care measure consists of one written, self- report item which asks the subject to rate her ability to meet responsibilities related to the physical needs of her husband during his entire final illness, using the following system: A. I was always able to carry out any responsibilities I had for the physical care of my husband. B. I was mostly able to carry out my responsibilities for his physical care. C. I was often unable to carry out.my responsibilities for his physical care. D. I was never or almost never able to meet my responsibilities for the physical care of my husband. E. At times when I was unable to provide any of the physical care my husband needed, I found other people who could provide the care. The Physical Care measure is listed as item 15, on page 229 of the Questionnaire Packet in Appendix B. Scoring for il re co Sc cc su hu al av Ir be De ea 69 this item is shown on pages 262-263 of the Scoring Key (Appendix C). h W f ' A n deni l f Her H n ' T rmin l Cordillera As stated previously, a surviving spouse must, at some point, recognize that her partner is likely to die from his illness if she is to begin to grief work and complete remaining marital business prior to her husband's death. To construct a measure of the subject's acceptance/denial, Schneider, Winder, and Peradotto first identified four concepts related to acceptance of the prospective loss by the surviving spouse: acknowledgement of the probability of her husband's death, accepting that his death cannot be reversed, allowing thoughts about losing her husband into conscious awareness, and accepting the loss as a real part of her life. The following four written, self-report items were generated from these concepts: 1. By that time, I knew that my husband would die from his illness. I had little or no hope that he would recover. 2. I could accept that he would never return. 3. At that time, I could accept that this loss was real. 4. I did things to avoid thinking about the possibility of losing my husband. Some of these items were taken from the Response to Loss Instrument (RLI) (Schneider & Deutsch, 1982) and modified to be relevant to the experiencerof a surviving spouse during her partner's final illness. A subject indicated her response to each item using the following system: 70 did not describe me described me somewhat mostly described me accurately described me UN-‘O “III In order to begin significant anticipatory grief work and complete final important marital tasks, it was not considered necessary for the surviving spouse to accept the reality of her husband's terminal condition throughout his entire final illness. However, if a subject only acknowledged the likelihood of her husband's death a few days before the death occurred, she was unlikely to complete significant grief and marital work in the short time remaining. Therefore, two*weeks prior to the death was chosen as the time a widowed subject was asked to recall the degree tO‘WhiCh she was able to accept the probability of the loss. Scoring for the Wife's Acceptance Scale is shown on page 263 of the Scoring Key (Appendix C). grief Werk The development of a scale to measure the completion of anticipatory grief work was based on two assumptions concerning the nature of the grieving process. The first assumption is that for meaningful grief work to be accomplished, the bereaved.must allow into awareness thoughts and feelings related to the loss. The second assumption is that feelings which are brought into awareness will diminish in intensity as they are worked through over time. In this investigation, "grief work" was therefore measured in terms of the subject's awareness of important thoughts and feelings regarding the prospective loss whenever they occurred during 71 the final illness, and also in terms of progress made by the subject in the resolution of those thoughts and feelings. A number of items from the RLI were considered appropriate as meaningful grief items because they involve cognitive and emotional responses that occur frequently among individuals who have experienced a significant loss. Additional items, considered relevant to the experience of a spouse suffering the prospective loss of her husband, were generated by Schneider, Winder, and Peradotto. All 35 of the Meaningful Grief items are also part of the larger Grief Intensity Measure, which was completed by every subject. Each subject was asked to respond to the items using the following system: 0 2 did not describe me 1 = described me somewhat 2 a mostly described me 3 = accurately described me In order to assess the extent to which progress was made in the resolution of these grief experiences, the subject was asked to rate the degree to which she experienced each aspect of her prospective loss during the early and middle stages of her husband's illness, and again, during the last two»weeks of her husband's illness. It is suggested that if a particular aspect of the loss experience is reported to have decreased in intensity by the last two weeks prior to the death, relative to earlier in the illness, that this change is indicative of some resolution of the issue by the surviving spouse. If a particular aspect of a subject's grief response is reported to have increased in intensity by the last two weeks If”. tc Pr 8y ( 5 6x] ma) egt fro awa tas} 72 prior to the death, the change is also thought to be indicative of progress, since it suggests increased awareness of grief issues. A subject's total Grief Work score represents the sum of the two component scores: 1. the sum of the change scores for grief items that decreased in intensity from the initial rating by the subject, to her rating for the last two weeks prior to the death (Less Intense); 2. the absolute sum of the change scores for grief items that increased in intensity over the same period (More Intense). The Grief Work Scale is listed in the Questionnaire Packet (Appendix B) on pages 242-244, items 59-93. Scoring for the Grief Work Scale is shown on pages 264 of the Scoring Key (Appendix C). Phyeiee; greglems A measure of physical problems was constructed in order to assess the extent to which the stresses associated with a spouse's terminal illness and the grief related to that prospective loss might have been expressed through somatic symptomatology. It is recognized that some physical problems (sleep and appetite problems, and loss of energy are an expected and normal part of anticipatory grief. However, there may be some subjects for whom somatic symptoms are an equivalent of grief, keeping important grief-related issues from awareness. Because of a severe physical response to grief awareness, other subjects may be unable to complete important tasks of the pre-death period such.as providing care for one's husband or carrying out other responsibilities. The Physical P1 51 73 Problems measure attempts therefore, to distinguish those subjects who suffered relatively severe somatic difficulties from subjects whose experience of physical problems fell within a range of severity that is considered normal and expected for persons coping with a major loss. First, Schneider, Winder, and Peradotto identified a number of common physical responses to a prospective loss. These include difficulties in the areas of sleep, appetite, physical energy, and bodily tension/pain. The following eight written self- report items were generated to assess these areas of functioning: 19. My stomach really churned when I thought about this loss. 20. I had difficulty getting to sleep. 21. I awakened during the night more often than I did before the illness. 22. I slept poorly. 23. I slept too much. 24. I was eating too much. 25. When I was reminded of what I was losing, my whole body felt heavy. 26. I had more aches and pains than before the illness. Most of these questions were adopted from items used on the RLI. For each item, the subject was asked to rate her physical response using the following system: 0 - did not describe me 1 - described me somewhat 2 - mostly described me 3 s accurately described me In addition, two items (#‘s 46 s 48) from the instrument used to measure depression - The Beck Depression Inventory CO!‘ GEE Prc 3?] me (3 of be 74 (BDI) (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) - were considered relevant to the assessment of a subject's physical energy, and were therefore included as part of the Physical Problems Scale. A third BDI item which assesses a subject's appetite was also included. Finally, one item (If 96) from the instrument used to measure anxiety - the State-Trait Anxiety Inventory (Spielberger, Gorsuch, & Lushene, 1970) - was included because of its relevance to a subject's physical response in terms of bodily tension. If an individual responds to her husband's illness primarily by developing physical symptomatology, resulting in the failure to complete important tasks of this period, then that individual can be considered to be manifesting relatively poor adjustment. However, during a lengthy final illness, an individual might initially have difficulty coping with the stresses of the illness and the prospective loss, instead suffering significant physical problems. Later, she may begin to adjust, with physical problems decreasing in severity. She may also possess an improved ability to accomplish work on important tasks prior to the death. Therefore, adjustment, in terms of somatic problems, was defined for this investigation, as the subject's ability to be relatively free from physical symptomatology in time to complete developmental tasks that are characteristic of this period. A subject was asked to respond to the Physical Problems items only in terms of her experience during the last two weeks prior to her husband's death. 75 In addition to identifying problems that are typically experienced, Schneider, Winder, and Peradotto identified a number of other aspects of a spouse's physical condition that are relevant to) her' response to Iher partner's terminal illness. They include accidents or injuries, sexual behavior, the presence of symptoms similar to those of the dying spouse, and health promotion activities. The following eight written, self-report items were generated these areas of functioning: 27. During that time, I had symptoms like my husband had. 28. I had more accidents or injuries than before the illness. 29. I was exercising. 30. I was neglecting or punishing my body. 31. I was more active in caring for myself than I was before the illness. 32. What I ate was healthy. 33. I could understand what my body needed better than before the illness. 34. Touching and being touched were important to me. A number of these items were taken from the RLI and modified to be relevant to the experience of a surviving spouselduring her husband's final illness. A subject.was asked to rate her response to each item using the following system: 0 - did not describe me 1 - described me somewhat 2 - mostly described me 3 - accurately described me. In addition, one item (# 52) from the BDI was considered relevant to the assessment of a subject's sexual behavior and was therefore included in the Physical Problems Scale. The surviving spouse's smoking habits were assessed thi as) be: bei as: 3?! 26 no lo af ad Ch fc be 1] ht Ce fc at S) 76 through one written self-report item (# 35). The subject was asked to choose the one statement among the following that best described her smoking behavior during the last two weeks before her husband's death: was a non-smoker. smoked less than prior to his illness. smoked about as much as prior to his illness. smoked somewhat more than prior to his illness. smoked much more than prior to his illness. HHHHH Finally, one self-report item.(# 45), taken from the BDI, assessed the surviving spouse's perception of her own physical appearance during the last two weeks of her husband's illness. Scoring for the Physical Problems Scaleris shown.on pages 265-267 of the Scoring Key (Appendix C). Quilt Experiencing some feelings of guilt is both a common and normal aspect of an individual's response to a significant loss, as an anticipatory reaction prior to the loss, and also after the loss occurs. Guilt becomes a problematic aspect of adjustment if it is excessive or if it persists and becomes chronic. Winder, Schneider, and Peradotto identified the following sources of guilt, each having some likelihood of being experienced by a spouse during her husband's final illness: feeling that she caused or failed to prevent her husband's illness, feeling that she did not provide enough care for her husband during his illness, feeling responsible for marital problems, living while her husband dies, feeling anger at her spouse for dying, feeling hostility towards her spouse, feeling that she should have appreciated her spouse more during their marriage, having thoughts about current and fut bur The f01 qu 77 future problems, feeling that her husband's illness is a burden and wishing that it would be over, and wishing to have a new, close relationship again after her husband is gone. These potential sources of guilt generated the following fourteen written, self-report items which are listed in the questionnaire packet (Appendix B) on pages 235-236, items 5- 18: 5. I felt as if I caused or should have prevented my 9. 10. 11. 12. 13. 14. husband's illness. I believed that it would help if someone could forgive me for my involvement in this illness. . I felt guilty for not always being able to give my husband enough care during his illness. I felt bad when I thought about my responsibility for problems in my marriage. At that time, I felt guilty about enjoying myself. I felt guilty that my life went on while my husband's would soon end. I felt guilty for feeling anger towards my husband for dying. I felt guilty when I sometimes felt that there was some justice in my husband's illness. I wished that.I had appreciated my husband.more throughout our marriage. It was hard to think about any problems without feeling guilty. 15. When I started to think about or plan for a future without 16. 17. 18. my husband, I felt guilty. I felt guilty about my anger towards my husband for my having to endure his illness. I felt guilty when I wished my husband would hurry up and die and get it over with. When I thought of the possibility of being close to someone again, after my husband is gone, I felt guilty. 78 For each item, the subject.was asked to:rate the level of guilt she experienced using the following system: did not describe me described me somewhat mostly described me accurately described me wN-‘O I H II II Because guilt feelings are a normal concomitant of grief, the presence of guilt during the final illness should not be considered an indication of problematic adjustment. However, when the husband's final illness lasts more than a few weeks, as it does in the case of each of this study's subjects, there is an opportunity for the surviving spouse to work through some of the guilt aroused by the prospective loss. Near the time of the death, guilt feelings that remain are likely to’be somewhat chronic and more problematic in terms of adjustment. Therefore, the time period for which.each subject.was asked to report on feelings of guilt was the two weeks prior to the husband's death. Scoring for the Guilt Scale is shown on page 263 of the Scoring Key (Appendix C). Meier; Because of its proven reliability and validity, and frequent use among psychological researchers, the Beck Depression Inventory (BDI) (Beck et. al., 1961) was chosen as the instrument to measure depression in this investigation. The BDI consists of 21 self-report items which assess the cognitive, emotional, attitudinal, and physical aspects of depression. Each item is comprised of four self-evaluative statements, in order of increasing severity. The subject was 79 asked to choose the alternative that best described her experience, during the last two weeks prior to her husband's death. This time period represented the final opportunity for the surviving spouse to overcome maladaptive coping strategies and engage in meaningful grief work prior to the death. As discussed in a previous section, depression represents a pathological response to a loss in that it disrupts normal grief, prevents recovery, and is characterized by lowered self-regard. An individual whose primary response to her prospective loss is a depressive one for the entire time of her husband's illness, may be considered to have had a relatively poor adjustment during that time. However, another individual might initially have evidenced a depressive reaction, but may have recovered and completed meaningful grief work prior to the death. Therefore, adjustment in terms of depression, was only assessed during the last two weeks of the final illness. In a study evaluating the internal consistency of the measure, Beck and Beamesderfer (1974) found the BDI to have a split-half reliability coefficient of .86 for a sample of 97 depressed subjects. A.number of studies comparing clinicians' ratings of depth.of depression with scores on the BDI, provide support for the measure's concurrent validity. The following correlations were found between the BDI and clinicians' ratings of depression: .65 (Beck et. al., 1961); .66 (Nussbaum, Wittig, Hanlon, & Kurland, 1963); .62 (Metcalfe & Goldman, 1965). Nussbaum et. al. (1963) also found that ini' Sca pro mea EX SE ar 80 initial and final correlations between the BDI and the‘MMPI D— Scale to be .75 and .69, respectively. Strong support for the construct validity of the BDI was provided by studies in.which the BDI was used as the criterion measure of depression, and a number of hypotheses regarding depression were confirmed. Among the hypotheses finding support were that depressed subjects are likely to 1. have a "masochistic" content to their dreams; 2. have a negative self-concept; 3. identify with the "loser" on projective tests; 4. have a history of deprivation; 5. respond to experimentally-induced failure with a disproportionate drop in self-esteem and increased hopelessness; 6. show a subjective and objective improvement in depression following a success experience; and 7. show a high correlation between intensity of depression and suicidal intent (Beck & Beamesderfer, 1974) . The BDI is listed in the Questionnaire Packet (Appendix B) on pages 239-241, items 38-58. Scoring for the BDI is shown on pages 269-272 of the Scoring Key (Appendix C). Anxiety The A-State scale of the State-Trait Anxiety Inventory (STAI) (Spielberger, Gorsuch, & Lushene, 1970) was chosen as the measure of anxiety for this investigation. The STAI is composed of two scales - the Trait Anxiety or A-Trait Scale, and the State Anxiety or Arstate Scale. The A-Trait Scale assesses what is thought to be a relatively stable aspect of personality - an individual's proneness to respond to a wide range of situations with anxiety. The A-State Scale measures a person's level of anxiety for a proscribed period of time. Fl 0v (1 te i1 81 Fluctuations in an individual's A-State scores are expected over time and across situations. According to Spielberger (1972), the STAI was constructed to measure the qualities of tension, apprehension, and nervousness, as these feelings vary in intensity. The STAI A-State Scale consists of 20 statements which reflect these concepts. The subject responds to each item by rating his/her feelings (for the particular time indicated) according to the following system: 0 a Not at all 1 2 Somewhat 2 = Moderately So 3 = Very Much So In the present investigation, a subject.was asked to rate the level of anxiety she experienced.during the last two*weeks prior to her husband's death. Similar to the assessment of depression, this time period was chosen because it represented the surviving spouse's last chance to overcome debilitating anxiety and complete meaningful grief work prior to her husband's death. It is suggested that a subject who suffered relatively severe anxiety throughout.most.of the final illness might still be considered to have adjusted well if she was able to recover prior to the death, in time to perform important anticipatory grief work. Unlike the STAI A-Trait Scale, which has been found to have high test-retest reliability, the A-State Scale's test- retest reliability is relatively low, with a median correlation coefficient of .3 across a number of studies (Spielberger et. al., 1970). Since the purpose of the A-State Scale is to measure anxiety as a transitory emotional state, 82 it is consistent with the concept of the scale that an individual's scores for any two points in time should be poorly correlated. However, when reliability was assessed in terms of internal consistency, high alpha coefficients were found for both the A-State and A-Trait Scales (Spielberger et. al., 1970). Support for the concurrent validity of the measure was provided by a study in which scores on the A-State Scale were compared.with scores on another measure of state anxiety - the Zuckerman AACL Today Form (Zuckerman 8- Lubin, 1965) , and found to be highly correlated (Spielberger et. al., 1970). In terms of the measure's construct validity, scores on the A-State Scale have been found to increase in response to various stresses, including: the prospect of surgery (Auerbach, 1973), threat of shock (Hodges, 1967; Hodges & Spielberger, 1966), making a speech (Lamb, 1969), pregnancy (Edwards, 1969), receiving negative feedback (McAdoo, 1969), and having to view a film of painful accidents (Lushene, 1970; Spielberger et. al., 1970). Finally, a decrease in A-State scores has been found to occur in situations designed to produce a more relaxed or calm state: systematic desensitization (Allen, 1971; Anton, 1975; Bedell, 1975) and relaxation training (Bedell, 1975). The STAI (A-State Scale) is listed in the Questionnaire Packet (Appendix B) on page 245, items 94-113. Scoring for this measure is shown.on page 269 of the Scoring Key (Appendix C). 83 Mama: The Response to Loss Instrument (Schneider & Deutsch, 1982) is psychometrically strong in terms of its reliability and validity (Deutsch, 1982). For this reason, it was chosen, in a modified and expanded form, as the instrument to measure intensity of grief. The RLI consists of 37 items which assess the cognitive, emotional, physical, imaginative, and spiritual aspects of an individual's response to a significant loss. Four items from the original RLI measure were eliminated for this investigation because they lack relevance to the situation of an individual experiencing a prospective loss. Nine new items were generated by Schneider, Winder, and Peradotto in order to assess other potentially important aspects of an individual's grief experience: 34. I looked for some way to fit the loss of my husband into the rest of my life. 35. Almost anything could remind me of what I was losing. 36. In comparison.to*what.I*was losing, everything else seemed trivial and meaningless. 37. I felt empty, like a shell, like I was just existing. 38. I felt confused and disoriented. 39. I felt numb. 40. There were parts of the illness or loss experience that disgusted me. 41. I felt so lonely and alone when I thought of my husband being gone. 42. I felt as if luck, fate, or God must be against me; I felt like a victim. Subjects were asked to respond to each item using the following system: 84 u did not describe me . described me somewhat c mostly described me = accurately described me (ON-5° In this investigation, intensity of grief refers to that time during the husband's final illness when the surviving spouse experienced her strongest response to each aspect of grief being assessed. Therefore, the subject was asked to respond to each item during the early and middle stages of her husband's final illness, and again, during the last two weeks of his illness. For a sample of 152 adult subjects, all of whom had experienced a significant loss within the previous two year period, Deutsch (1982) assessed the reliability of the RLI. Reliability was tested in terms of the measure's internal consistency using Chronbach's Coefficient Alpha. For the total scale, the alpha coefficient obtained was .95. Split—half reliability was also assessed and a reliability coefficient of .95 was found. Support for the construct validity of the instrument was provided by testing a number of hypotheses, using the RLI as the criterion measure of grief. Among the hypotheses that found support were that 1. grief intensity was significantly correlated with the amount of time since the loss for non- depressed subjects; 2. for depressed subjects, no significant relationship was found between grief and time since the loss; 3. depressed subjects could be distinguished from non- depressed subjects by the farmer's significantly higher score on the physical response scale of the RLI; 4. depressed 85 subjects could be distinguished from non-depressed subjects by the greater discrepancy between the emotional and cognitive scale scores for depressed subjects, relative to non-depressed ones; 5. depressed subjects scored higher on the RLI than those subjects who were not seriously depressed. The Grief Intensity Scale is listed in the Questionnaire Packet (Appendix B) on pages 231-233, items 1-42. Scoring for this measure is shown on pages 263-264 of the Scoring Key (Appendix C). i ' i n V i i f h M ur Because a number of the measures were first developed for use in this investigation, information on their reliability and validity was lacking until this study was conducted. To insure that each. newly-developed scale 'was measuring' an homogenous construct, reliability was assessed using a test of internal consistency. Information was also gathered relevant to the concurrent validity of these measures. Concurrent validity was assessed by comparing the subject's score on each measure of adjustment with a corresponding rating of adjustment which was provided by the hospice worker who was the primary contact with the surviving spouse. The hospice worker was asked to provide the following ratings of adjustment: Wife's Acceptance (the extent to which the surviving spouse was able to accept the terminal nature of her husband's illness during the final two weeks of the illness), (completion of) Marital Work (the wife's ability to accomplish important marital work prior to the death), 86 Physical Care (of the dying spouse) (the subject's ability to provide for her husband's physical care throughout the illness), Carrying Out Responsibilities (how well the surviving spouse was able to meet any responsibilities she may have had during the final illness in terms of employment, care of dependent children, and household tasks), Physical Problems (the survivor's physical response to her prospective loss during the two weeks prior to the death), Guilt (experienced during the final two weeks of the illness), Grief Work (the extent to which the surviving spouse had gained awareness of and worked through important issues related to her prospective loss, prior to the death). In addition, the hospice worker provided a rating of the two other aspects of adjustment examined in this investigation (depression and anxiety). Adequate information (previously cited) already exists regarding the validity of the instruments used to assess depression and anxiety - the BDI (Beck et. al., 1961) and the STAI (Spielberger et. al., 1970), respectively. The ratings obtained. in the current study represented an additional check of these measures' construct validity. Similarly, the instrument used as a measure of grief intensity (the RLI) has already been shown to have adequate validity (Deutsch, 1982). The hospice worker's rating of the intensity of the surviving spouse's grief response provided additional information regarding the concurrent validity of this instrument. Finally, no data existed on the validity of the 87 instrument used to measure the husband's attitude regarding his illness, Husband's Acceptance. Therefore, as an assessment of concurrent validity, a rating was obtained from the hospice‘worker on the degree to which the husband.was able to accept the terminal nature of his illness, and acknowledge this fact to»his wife during the twO'weeks prior to his death. The Hospice Worker Questionnaire is listed in Appendix B on pages 252-255. h ' n n h This study investigated the relationship between a woman's adjustment during the time of the terminal illness of her husband, and a number of factors that might exert an influence on her adjustment: quality of the marriage, social support, her age, length.of‘her husband's illness, and whether or not her husband was able to.acknowledge the terminal nature of his illness. flxpettee;e_;;, The internal consistency of the total items of the Marital Work, Guilt, Physical Problems, Carrying Out Responsibilities, Wife's Acceptance, Social Support, Husband's Acceptance, and Grief Work self-report scales will be sufficiently high, to infer that each scale is measuring an homogenous concept. fixpetteeie_1;; A significant positive correlation will be found between ratings provided by the hospice worker, and the subject's scores on the following self-report measures: Marital Work, Grief Work, Physical Problems, Guilt, Wife's Acceptance, Depression, Anxiety, Grief Intensity, and 88 Husband's Acceptance. gypgtheeis 1;;; Subjects from marriages that were "low" in quality will score significantly lower in intensity of grief response than will subjects from marriages that were "high" in quality. Hypgtheeis 12; Subjects from marriages that were "low" in quality will score significantly higher on measures of depression, anxiety, guilt, and physical problems; and significantly lower on measures of the completion of marital work and the performance of meaningful grief work than will subjects from "high quality" marriages. Hypetheeis V; Subjects who were "high" in social support during the time of their husbands' terminal illnesses will score significantly lower on measures of depression, anxiety, guilt, and physical problems; and significantly higher on the measures of grief work, acceptance of the terminal nature of their husbands' illnesses, and the accomplishment of marital work than will spouses who were "low" in social support. H h V ° Subjects who were "high" in social support will score significantly higher on a measure of grief intensity than will subjects who were "low" in social support. i 12' Subjects whose husbands were "high" in their ability to accept the terminal nature of their illnesses will score significantly lower on.measures of depression, anxiety, guilt, and physical problems; and significantly higher on measures of marital work, wife's acceptance of the husband's terminal condition, and the performance of grief work than will spouses whose husbands were "low" in acceptance. 89 H h V 11' Subjects whose husbands were "high" in acceptance of their terminal conditions will score significantly higher on a measure of grief intensity than will subjects whose husbands were "low" in acceptance. fiypethesis IX: Older subjects will score significantly lower on measures of depression, anxiety, guilt, and physical problems than will younger subjects. fixpethe§i§_ze, Older subjects will score significantly lower on a measure of grief intensity than will younger subjects. W Subjects whose husbands died after long-term chronic illnesses will score significantly higher on measures of depression, anxiety, guilt, and physical problems than will subjects whose husbands died after short-term chronic illnesses. Hypetheeie 311; .A significant interaction effect will be found between subject's age and length of the final illness. For short-term chronic illnesses, younger subjects will score significantly higher on measures of depression, anxiety, guilt, and physical problems than will older subjects. For long-term chronic illnesses, older subjects will score significantly higher on measures of depression, anxiety, guilt, and physical problems than will younger subjects. MW Procedures used to test the eleven hypotheses are discussed in this section. Hypothesis I addresses the question of the reliability of the measures used in this investigation. Reliability refers to 90 the homogeneity of an instrument - the extent to which a single construct is being measured. The most common assessment of a scale's reliability involves analyzing the items of the scale for internal consistency. Reliabilities of the Marital Work, Guilt, Physical Problems, Wife's Acceptance, and Carrying Out Responsibilities scales were tested by computing Cronbach's Coefficient Alpha (Cronbach, 1951) on the total items of each scale. A value of at least .70 was considered necessary for the items of a scale to be judged internally consistent. For the Social Support and Husband's Acceptance scales, each of which is composed of two items, reliabilities were estimated by computing a Pearson Product-Moment Correlation Coefficient for each scale. Because the Grief Work scale is composed of two subscales, More Intense and Less Intense, and the subscales are each considered to be measuring aspects of grief work, the reliability of this scale was estimated by computing a Pearson Correlation Coefficient between the subscale totals. Hypothesis II addresses the question of the validity of the measures used in this investigation. The validity of an instrument refers to the extent to which the instrument measures what it is supposed to measure. Tests of concurrent validity, one type of validity, refer tO‘whether'or'not scores on an instrument are related to some currently available criterion measure. In this investigation, the subject's scores on the following self-report measures - Physical Care, Marital Work, Carrying Out Responsibilities, Grief Work, 91 Physical Problems, Guilt, Wife's Acceptance - were correlated with a criterion measure, the hospice worker's ratings of the same variables. Correlations between self-report and hospice worker ratings were considered significant at the .05 level. Hypotheses IV, V, VII, Ix, XI, and XII were intended to determine whether a number of variables have a significant influence upon the adjustment of a surviving spouse during the time of her husband's final illness. Analysis of variance was chosen as the appropriate test of significance for group differences for these data. Therefore, ANOVA's tested for main effects of marital condition, social support, husband's attitude (toward his illness), wife's age, and length of illness, as well.as interaction effects between.age and length of illness, on the following dependent variables: Physical Care, Marital Work, Carrying Out Responsibilities, Physical Problems, Guilt, Wife's Acceptance, Depression, and Anxiety. Main effects and interaction effects were considered significant at the .05 level. Analysis of Co-variance was utilized to test for main effects of these independent variables, as well as for interaction effects between age and length of illness, on the dependent variable Grief Work. This type of analysis was chosen in order to control for differences among subjects in terms of the intensity of their initial (early and middle stages of their husbands' illnesses) grief responses. Therefore, the co—variate was the subject's initial score on the measure of Grief Intensity. Main effects and interaction effects were considered significant at the .05 level. hypot impact of surviving version 0 her husba of merit wife's a effects interact level. 92 Hypotheses III, VI, VIII, and x assessed the possible impact of the independent variables on the intensity of the surviving spouse's grief response (as measured by the expanded version.of the Response to Loss Instrument) during the time of her husband's final illness. ANOVA's tested for main effects of marital condition, social support, husband's attitude, wife's age, and the length of illness, and any interaction effects between age and length of illness. Main effects and interaction effects were considered significant at the .05 level. QHAPIfER IV RE§QLT§ 2W5 The final sample consisted of 47 subjects, all of whom were widowed females. Subjects ranged in age from 25 to 84 with a mean age for the total sample of 62.79 years (etges 12.69). Their husbands' terminal illnesses lasted an average of 11.56 months (stgt= 8.94) with a range of 2 to 48 months. In 24 of the cases, the illnesses lasted more than one month but less than 6 months, and in 23 cases, the husbands' illnesses were longer than 6 months. Twenty subjects (42.6%) were employed outside their homes during the time of their husbands' final illnesses. A summary of the total sample's performance on the various measures employed in this study is provided in Table 1 The frequency of occurrence in the sample of various symptoms is reported in Table 2 . A symptom was considered present if the surviving spouse reported to have experienced the symptom either during the early and middle stages of her husband's illness, or during the last two weeks of the illness (see Appendix C for scoring of Symptom Frequency, pgs. 272- 273). High rates of occurrence were found for symptoms which are typically associated with a normal grief state. Depressed mood and sleep disturbance were each experienced by over 90% of the sample. At least 70% of the sample reported crying, loss of interest, and fatigue. Irritability and appetite loss 93 94 occurred in more than half the sample. A substantial minority of subjects (43%) reported experiencing significant guilt feelings. Table 3 reports the number of symptoms experienced by subjects. Ninety-five percent of the subjects reported having experienced 3 or more of the 8 symptoms during the time of their husbands' illnesses; and 71% reported 5 or more symptoms. Five subjects (11%) experienced all 8 of the symptoms about which they were questioned. One other study (Clayton et. al., 1973) reported the frequency' of 'various symptoms among Ibereaved. individuals during the time of their spouses' final illnesses, and these frequencies are provided in Table 2 for comparison. All eight symptoms were more common among subjects in the current study than among Clayton et. al.'s subjects. For 5 of the 8 symptoms, differences between the two samples were found.to be statistically significant by chi-square analysis (depressed mood: x2 = 3.85, p<.05; loss of interest: x2 = 35.59, p<.001; fatigue: x2 = 9.77, 2<.o1; irritability: x2 = 31.11, p<.001; guilt: x2 = 98.66, p<.001). _orm: -on . h- M. i :1 0 -17 . ._ ...r H _-.n.'- A n A n L n h f Illn r Testing of the hypotheses required the formation of two groups for each of the five independent variables. For the Marital Quality variable, each subject was classified as having had either a "high" or a "low" quality marriage, based on the subject's DAS score. If a subject's score was above the median score for all subjects 95 (median: 118), that subject was classified in the "High" Marital Quality group (g: 22). Subjects whose scores fell at or below 118 were classified in the "Low" Marital Quality group (a: 25). The same method was used for classifying subjects in terms of their social support: subjects who scored at or above the median score for all subjects on the Social Support scale (median: 4) comprised the "High" Social Support group (a; 26), while the "Low" Social Support group consisted of subjects whose scores fell below the median score (g: 21). For the Husband's Acceptance variable, the median score (medians 3) was again used to classify subjects into "High" and "Low" groups. Subjects whose scores were at or above the median were classified as being "High" in Husband's Acceptance (g: 25), and subjects whose scores fell below the median were considered to be in the "Low" Husband's Acceptance group (n= 22). In classifying subjects according to age, it was decided to use age 65 as the cutoff, dividing "Young" and "Old" subjects, since 65 continues to be somewhat of a societal standard for aging. Thus, 24 subjects who were less than 65 years of age comprised the "Young" group, and 23 subjects who were 65 or older comprised the "Old" age group. Finally, for the Length of Illness variable, surviving spouses whose husbands' illnesses lasted at least one month, but less than six months, were considered to be in the "Short- term" chronic illness group (a: 16). The "Long-term" illness group consisted of surviving spouses whose husbands died after 96 illnesses lasting six months or more (a: 31). W W: The internal consistency of the total items of the Marital Work, Guilt, Physical Problems, Carrying Out Responsibilities, Wife's Acceptance, Social Support, Husband's Acceptance, and Grief Work scales will be sufficiently high to infer that each scale is measuring an homogenous concept. The internal consistency of the Marital Work, Guilt, Physical Problems, and Wife's Acceptance scales was determined by computing Cronbach's Coefficient Alpha. After removing items that failed.to.achieve item-total correlations of .30 or above, three of the self-report scales did achieve reliability coefficients greater than .70: Marital Work (.82), Guilt (.81), and Physical Problems (.89). The exception was the Wife's Acceptance scale, which attained a coefficient alpha of .63. Given the small number of items in this measure (four), its coefficient alpha suggests that the scale is sufficiently reliable to proceed with further analysis. No coefficient alpha was calculated for the Carrying Out Responsibilities scale. This measure contains only three items, two of which - Carrying out Job Responsibilities, and Caring for Dependent Children - were completed by small numbers of subjects (as 20 and a: 9; respectively), resulting in a great deal of missing data when calculating a Responsibilities total score. It was therefore decided that for all further analyses, no overall score on Carrying Out Responsibilities would be used; instead, Household Tasks and Jobs would.be analyzed as separate variables, with the item1on 97 caring for children eliminated because of small sample size. For the Social Support and Husband's Acceptance scales, that were each composed of two items, Pearson Product-Moment Correlation Coefficients were computed. For Social Support, £2 .78; for Husband's Acceptance, E? .39. Reliability data was also available for a scale previously found to be reliable, the Response to Loss Instrument (Grief Intensity). This scale attained a coefficient alpha of .93. Information regarding the reliability of the Grief Work scale was obtained by computing the correlation between Grief Work's two component scales, More Intense and Less Intense. The two subscales were found to be correlated -.11. Because Grief Work was not found to be a sufficiently reliable instrument, no other analyses were performed using this scale. All reliability tests of the scales used in this investigation are summarized in Table 4. Final item-total correlations for the four scales developed for this investigation. (with those items removed.‘which failed to achieve an item-total correlation of at least .30) are listed in.Appendix B, Tables 1-4. All analyses with these scales were performed without the items which lacked an acceptable level of reliability. W: A significant positive correlation will be found between ratings provided by the hospice worker, and the subject's scores on the following self-report variables: Marital Work, Grief Work, Physical Problems, Guilt, Wife's Acceptance, Depression, Anxiety, Husband's Acceptance, and Grief Intensity. 98 Hypothesis II addressed the question of the concurrent validity of the subject self-report measures. Correlations between.hospice worker ratings and total scores on each of the self-report scales were determined by computing Pearson Product-Moment Correlations. In only one instance did the correlation between subject and hospice worker ratings attain statistical significance (see Table 5). The self-report Grief Intensity score correlated significantly with the hospice worker's rating of the same dimension (£3 .44; p: .01). Because little agreement was found between self-report and hospice worker ratings of the various aspects of adjustment, the question arose as to which type of rating represented the more accurate or valid measure of these variables. Based on available evidence (to be discussed later) it was determined that the more accurate ratings were the subjects' self-reports. Therefore, in all further analyses, variables related to the surviving spouse's adjustment are represented by self-report rather than hospice worker ratings. 99 g, h-‘-- --,. -. . . or~ nf -n inc - A- m-n of Subjects from marriages that were "low" in quality will score significantly higher on measures of depression, anxiety, guilt, and.physical problems; and significantly lower on measures of the completion of marital work and the performance of grief work than will subjects from "high" quality marriages. This hypothesis was tested by computing six ANOVA's, each employing Marital Quality as the independent variable, and one of the five adjustment scales as the dependent variable. Relevant means and standard deviations are listed in.Table 6. The results did not support the hypothesis that subjects from "low quality" marriages would score significantly less well on measures of adjustment than would "high quality" marriage subjects. In fact, the "High Quality" group scored slightly higher (as doing more poorly) on measures of Physical Problems ("High Quality" 3: 20.14; "Low Quality" 5: 15.84) (£2 3.05; p< .10) and Depression ("High Quality" m: 17.00); "Low Quality m: 13.20) (E: 2.99; p< .10). These results are presented in Tables 7 and 8. No significant main effects were found for level of Marital Quality. Pearson correlations were also obtained between subjects' Marital Quality scores and their scores on the five measures of adjustment, and are presented in Table 9. In general, the results are consistent with those of the ANOVA's. No significant correlation was found between.Marital Quality and two of the dependent variables, Guilt and Anxiety. Higher Marital Quality was associated with greater depression 100 and more physical problems. In the case of Depression, this relationship was shown as a trend (;= .24; p= .10); while for Physical Problems, the correlation with Marital Quality reached statistical significance (£= .32 ; p< .05). The exception to the findings presented thus far is the relationship between Marital Quality and Marital Work. As predicted, subjects who accomplished more marital work prior to their husbands' deaths, tended to be from "higher" quality marriages. This relationship appeared as a trend, not quite reaching statistical significance (g: .28; p< .10). Hypetheeie Ill: Subjects from marriages that were "low" in quality will score significantly lower in intensity of grief response than will subjects from marriages that were "high" in quality. This hypothesis was tested by computing an analysis of variance for level of Marital Quality on Grief Intensity. Relevant means and standard deviations are listed in Table 6, and results of the ANOVA are shown in Table 10. The prediction was supported by the results of the analysis. The "High Quality" group scored higher on average (m: 85.78) on the Grief Intensity measure than did the "Low Quality" group (_n_1_= 73.96). The difference did not reach statistical significance however, and appeared as a trend (is 3.03; p< .10). Correlational analysis also found a positive relationship between Marital Quality and Grief Intensity (see Table 9). In this case, the relationship between the two variables did attain statistical significance (£= .32; p< .05). 101 The analyses discussed thus far revealed that to some degree, an association exists between Marital Quality and four of the dependent variables i.e., Marital Work, Physical Problems, Depression, and Grief Intensity. Additional analyses were undertaken to clarify which aspects of the marital relationship might contribute to Marital Quality's association with the dependent variables. Previous research (Spanier, 1976) reliably identified the following four factor subscales within the DAS: a) Dyadic Cohesion (the degree to which a couple engages in activities together), b) Dyadic Consensus (the degree to which a couple agrees on matters of importance to the relationship), c) Dyadic Satisfaction (the degree to which a spouse is satisfied with the relationship), and d) Affectional Expression (the degree to which a spouse is satisfied with the expression of sex and affection in the relationship). To investigate the possibility that one or more of these factors might be most responsible for the main effects found for Marital Quality, separate DAS subscale scores were computed for each subject. For each of the four DAS subscale variables, subjects whose scores were at or above the median score for that subscale were placed in the "High" group for that variable; while subjects whose scores were below the median score comprised the "Low" group. Median scores for each DAS subscale were as follows: Dyadic Cohesion: 16; Dyadic Consensus: 54; Dyadic Satisfaction: 32; Affectional Expression: 9. Sixteen ANOVA's were performed to test for main effects of the four DAS variables on the four dependent 102 measures - Marital Work, Physical Problems, Depression, and Grief Intensity. A significant main effect was found for Dyadic Consensus on Physical Problems and Grief Intensity. Means and standard deviations concerning Dyadic Consensus are presented in Table 11; and the results of the ANOVA's are shown in Tables 12 and 13. Subjects in the "High" Dyadic Consensus group scored significantly higher on the measures of Physical Problems ("High" m; 21.13; "Low" ms 14.71) (E? 7.46; pr .01), and Grief Intensity ("High" ma 87.65; "Low" :3: 71.58) (_1=_'= 6.08; 2< .05) than did subjects from the "Low Dyadic Consensus" group. In addition, a trend was found for the "High" Dyadic Consensus group to score as more depressed (m: 17.09) on the BDI than did the "Low" Dyadic Consensus group (m= 12.96) (E: 3.59; p< .10). The results of this ANOVA are presented in Table 14. No significant relationship was found between Dyadic Consensus and the Marital Work variable. For the Affectional Expression subscale, a main effect was found for one of the dependent variables - Grief Intensity. Subjects who reported relatively greater satisfaction with the expression of sex and affection in their marriages (the "High" Affectional Expression group) scored significantly higher on Grief Intensity on average that did subjects who were less satisfied with this aspect of their marriages (the "Low" Affectional Expression group) ("High" m: 84.47; "Low" 3: 70.59) (E= 4.02; p< .05). No significant associations were found between Affectional Expression and the other three dependent variables. Relevant means and standard 103 deviations are listed in Table 15. Results of the ANOVA concerning Grief Intensity are shown in Table 16. Neither the Dyadic Satisfaction or the Dyadic Cohesion scales were found to be significantly associated with scores on any of the dependent variables. Means and standard deviations concerning Dyadic Satisfaction and Dyadic Cohesion are presented in Tables 17 and 18 respectively. amnesia: Subjects who were "high" in social support during the time of their husbands' terminal illnesses will score significantly lower on measures of depression, anxiety, guilt, and physical problems; and significantly higher on measures of grief work, acceptance of the terminal nature of their husbands' illnesses, and the accomplishment of marital work than will subjects who were "low" in social support. To test this hypothesis, an ANOVA.was performed for each of five adjustment variables, with Social Support as the independent variable. Means and standard deviations are listed in Table 19. Women who experienced relatively "high" levels of social support scored significantly higher on the Marital Work variable (m: 20.54) than did women who had "low" levels of social support (m: 15.90) (E= 7.22; p< .01). Results of this ANOVA are shown in Table 20. No significant differences were found between the two social support groups for any of the other five dependent variables. Pearson correlations between Social Support scores and scores on the six adjustment variables are presented in Table 21. Results of these analyses confirmed a positive relationship between Social Support and Marital Work (Is .37; 104 p< .01). H h V : Subjects who were "high" in social support will score significantly higher on a measure of grief intensity than will subjects who were "low" in social support. An analysis of variance was computed to test the effect of level of social support on the Grief Intensity scale. Means and standard deviations are reported in Table 19. No significant main effect was found for Social Support group on Grief Intensity. Similarly, a correlational analysis (see Table 21) was unable t01detect.a significant relationship between the two variables. h i V I: Subjects whose husbands were "high" in their ability to accept the terminal nature of their illnesses will score significantly lower on measures of depression, anxiety, guilt, and physical problems; and significantly higher on measures of marital work, wife's acceptance of the terminal condition, and performance of grief work than will spouses whose husbands were "low" in acceptance. To test this hypothesis, ANOVA's were performed for six of the adjustment variables, with Husband's Acceptance as the independent variable. Means and standard deviations are listed in Table 22. The hypothesis was supported for two of the six adjustment variables tested. Subjects in the "High" Husband's Acceptance group scored higher on the Marital Work scale (ms 20.68) than did subjects in the "Low" Husband's Acceptance" group (m= 15.95) (E= 7.62; p< .01). The "High" group also scored significantly lower on Guilt (m; 3.88) than did the "Low" group (as 6.82) (E: 4.80; p< .05). The results of these ANOVA's are presented in Tables 23 and 24. For the 105 Physical Problems, Depression, Anxiety, and.Wife's Acceptance scales, the ANOVA's revealed no significant differences between the two Husband's Acceptance groups. Results of the correlational analyses concerning Husband's Acceptance and adjustment are consistent with the analyses presented above (See Table 25). A positive correlation was found, significant at the .01 level, for Husband's Acceptance and Marital Work (1;: .38). Husband's Acceptance scores were negatively correlated with scores on the Guilt measure (g: -.40; p< .01). Expothssis_YIII: Subjects whose husbands were "high" in acceptance of their terminal conditions will score significantly higher on a measure of grief intensity than will subjects whose husbands were "low" in acceptance. An analysis of variance was computed to test the effect of level of Husband's Acceptance on the Grief Intensity scale. Means and standard deviations relevant to this issue are presented in Table 22. The prediction was not supported by the results of the analysis. Subjects from the "High" Husband's Acceptance group did not score higher on the Grief Intensity scale than did subjects from the "Low" Husband's Acceptance group. Nor did a correlational analysis find a significant relationship between Husband's Acceptance and Grief Intensity (see Table 25). 106 H229thsaia.ll= Older subjects will score significantly lower on measures of depression, anxiety, guilt, and physical problems than will younger subjects. To test the effect of Age level, an analysis of variance was performed for each of the four adjustment variables, with Age as the independent variable. Relevant means and standard deviations are listed in Table 26. The predicted relationship between age and the adjustment variables was not confirmed by the results. No difference was found between "young" and "old" subjects on any of the adjustment variables. The above results regarding the effects of age' on adjustment were confirmed by correlational analyses. No significant correlation was found between age and any of the four adjustment variables (see Table 27). H h x: Older subjects will score significantly lower on a measure of grief intensity than will younger subjects. This hypothesis was tested by computing an analysis of variance with Age as the independent variable. Means and standard deviations are listed in Table 26. The hypothesis was not confirmed by the results of the analysis. No significant difference was found between "old" and "young" subjects in terms of their Grief Intensity scores. Nor did correlational analysis find a significant relationship between subject's Age and Grief Intensity (see Table 27). 107 W: Subjects whose husbands died after long-term chronic illnesses will score significantly higher on measures of depression, anxiety, guilt, and physical problems than will subjects whose husbands died after short-term chronic illnesses. To test this hypothesis, an analysis of variance was performed for each of the adjustment variables, with Length.of Illness as the independent variable. Relevant. means and standard deviations are presented in Table 28. The analyses failed to find a main effect for Length of Illness for three of the four adjustment variables. The one exception was the Guilt variable. A trend was found for subjects in the "Long-term" chronic illness group, to score lower on the Guilt scale (m: 4.35) than did subjects in the "Short-term" group (m= 7.00) (E: 3.18; p< .10). This result was opposite to that predicted and is presented in Table 29. Correlational analyses failed to find a significant relationship between Length of Illness and scores on any of the four adjustment variables. Results of these analyses are presented in Table 30. W1: A significant interaction will be found between subject's age and length of final illness. For short-term chronic illnesses, younger subjects will score significantly higher on measures of depression, anxiety, guilt, and physical problems than will older subjects. For long-term chronic illnesses, older subjects will score significantly higher on.measures of depression, anxiety, guilt, and physical problems than will younger subjects. To investigate this question, a 2 x 2 ANOVA was computed for the two Age groups ("Young" and "Old") and the two Length of Illness groups ("Short-term" chronic and "Long-term" 108 chronic) on each of the four adjustment variables. Relevant means and standard deviations are presented in Table 31. The results failed to support Hypothesis XII. No significant interaction was found between Age and Length of Illness on any of the adjustment variables. W The results of the analyses discussed thus far suggest that the issue of what constitutes healthy adjustment is a complex one. For three of the adjustment variables (i.e., Physical Problems, Depression, and Anxiety), few strong associations were found with any of the independent variables. When a significant relationship was found for Physical Problems and Marital Quality, it was opposite to that which was predicted. It had been assumed that adjustment ought to be measured as if it were a linear variable. That is, high scores on such variables as depression and anxiety represent less optimal coping or adjustment than do lower scores. However, it has previously been noted that manifestations of grief overlap to some: extent. with. certain. psychiatric conditions such. as depression and somatoform disorders. A high score on the Beck Depression Inventory might not necessarily indicate problematic adjustment, but rather only that a surviving spouse was engaged in a normal grief process. Instead of simply measuring coping or adjustment, the Physical Problems, Depression, and Anxiety scales might also be measuring normal grief. This could be the reason that few significant associations were found between these variables 109 and the independent variables. For a woman who is coping with the terminal illness of her husband, the relationship between Physical Problems, Depression, or Anxiety and adjustment might not be linear, but rather curvilinear. "Good adjustment" might best be represented. by’ mid-range scores on. measures of Physical Problems, Depression, and Anxiety, indicating that the surviving spouse is engaged in anticipatory grief work. Relatively "poor adjustment" might be represented by either very low or very high scores on the same measures. Low scores would indicate the absence of normal grief, while high scores would suggest an extreme grief reaction. To test this hypothesis, it was decided to once again study the association between Physical Problems, Depression, and Anxiety; and the predictive variables, investigating whether extreme scores (either high or low) on the dependent measures differed from mid-range scores in terms of their association with the independent variables. It was expected that "high" quality marriages would be associated with better coping or adjustment, as indicated by mid-range scores on measures of physical problems, depression, and anxiety. "Low" quality marriages would be associated with less Optimal adjustment, as indicated.by extreme scores on the dependent variables. Similarly, it was predicted that subjects with high levels of social support would score in the mid- range on the dependent variables. Subjects experiencing lower levels of social support were expected on average to have extreme scores on measures of physical problems, depression, and anxiety. Older subjects were alsolexpected to score in the 110 mid-range on the dependent variables, more so than would younger subjects. Similarly it was predicted that subjects whose husbands were able to accept the terminal nature of their illnesses would be more inclined to score in the mid- range on the dependent variables than would subjects of "less accepting" husbands. Subjects whose husbands died after relatively brief chronic illnesses were expected to score more in the mid-range on measures of physical problems, depression, and anxiety than.would subjects whose husbands died.after more long-term chronic illnesses. Lastly, an interaction was predicted for the subject's age and the length of the husband's illness. Older subjects whose husbands died after "short-term" chronic illnesses were expected to score more in the mid-range on the dependent variables than were younger "short-term" chronic illness subjects. For "long-term" chronic illnesses, however, younger subjects were predicted to score more in the mid-range than were older subjects. Subjects' scores on the Physical Problems, Depression, and Anxiety scales were first converted to absolute z-scores. This procedure equated high. and low ‘values within each measure, and produced new distributions in which converted mid-range scores had the lowest values. For each of the five independent variables, ANOVA's were performed on the converted variables - z-Physical Problems, z-Depression, and z-Anxiety. A 2 x 2 ANOVA tested for an interaction between the two Age groups ("Young" and "Old") and the 2 Length of Illness groups ("Short-term" chronic and "Long-term" chronic). For the independent variable Marital Quality, the analy for r adjus signi Z-De} sign qual scoz Resx SUp deg ra1 SO 111 analyses provided no support for the hypothesis (see Table 32 for relevant means and standard deviations). For two of the adjustment variables, z-Physical Problems and z-Anxiety, no significant main effect was found for Marital Quality. For the z-Depression variable, a trend was found, not quite significant, opposite to that predicted. "High" marital quality was associated with somewhat more extreme Depression scores than was "low" marital quality (1.5: 3.85; p< .10) . Results of this analysis are presented in Table 33. No significant main effects were found for the Social Support variable. Subjects who experienced a relatively "high" degree of social support did not tend to score in the mid- range on the dependent variables any more than did "low" social support subjects. Relevant means and standard deviations are listed in Table 34. Similarly, level of Husband's Acceptance was not differentially associated with scores on the z-Physical Problems, z-Depression, or z-Anxiety measures. The "High" Husband's Acceptance group did not score more in the mid-range on the dependent variables, relative to the "Low" Husband's Acceptance group (see Table 35 for relevant means and standard deviations). No significant main effects were found for Age on the dependent variables. Relevant means and standard deviations are listed in Table 36. "Old" and "young" subjects were not significantly different in termslof their tendency to score in the mid-range or the extremes on the dependent variables. Neither were main effects found for the Length of Illness 112 variable. Subjects in the "Short-term" chronic group did not score more in the mid—range on the dependent variables than did "long-term" chronic subjects. Relevant means and standard deviations are listed in Table 37. Lastly, no significant interaction was found between subject's Age and the Length of Illness variables for any of the dependent variables (see Table 38 for means and standard deviations). One other attempt was made to distinguish normal grief from a more pathological condition, depression. Six items were selected from the BDI because they are characteristic of clinical depression, but are not thought to be typically associated with the experience of normal grief. The items chosen are as follows: 40. I did not feel like a failure I felt that I had failed more than the average As I looked back on my life, all I could see was a lot of failures I felt that I was a complete failure as a person 42. didn't feel particularly guilty felt guilty a good part of the time felt quite guilty most of the time felt guilty all the time HHHH 43. didn't feel I was being punished felt that I may be punished expected to be punished felt I was being punished HHHH 44. didn't feel disappointed in myself was disappointed in myself was disgusted with myself hated myself HHHH mig? pro min on var Mai De‘ 301 del Su- re Su 011 De Wh hu re 6f; 113 45. I didn't feel I was any worse than anybody else I was critical of myself for my weaknesses or mistakes I blamed myself all the time for my faults I blamed myself for everything bad that happened 46. _____ I didn't have any thoughts of killing myself I had thoughts of killing myself, but I would not carry them out I would have liked to kill myself I would have killed myself if I had the chance It was thought that a total score on this group of items might be a more valid measure of a depressive response to a prospective loss since the contribution of normal grief is minimized. A total Depression-2 score was therefore computed on the six items for each subject. Hypotheses concerning the association between this new variable and the independent variables were identical to the original hypotheses. Higher Marital Quality was expected to be associated with lower Depression-2 scores. Similarly, it was predicted that "High" Social Support subjects would score lower on the new depression variable than would subjects in the "Low" Social Support group. Older subjects, subjects whose husbands were relatively high in acceptance of their terminal illnesses, and subjects whose husbands' illnesses were "short-term" chronic ones were all expected to score lower on average on the Depression-2 variable than would younger subjects, subjects whose husbands were "less accepting", and subjects whose husbands died following "long-term" chronic illnesses, respectively. Five ANOVA's were performed, each of which used Depression-2 as the dependent variable, to test for main effects of Marital Quality, Social Support, Husband's 114 Acceptance, Age, and Length of Illness. Results of the analyses did not support the hypotheses discussed above. None of the independent variables showed a significant main effect on the Depression-2 variable. Relevant means and standard deviations are shown in Table 39. Regarding the Grief Work scale, results of the analysis to determine the scale's reliability indicated that Grief Work is not a measure of a surviving spouse's performance of anticipatory grief. However, since each subject's Grief Work score represents the sum of two subscale scores (More Intense and Less Intense), one of these component scores might be significantly associated with the independent variables. This possibility was explored for each of the three independent variables for which predictions concerning Grief Work were made, Marital Quality, Social Support, and Husband's Acceptance. For each independent variable, two covariate analyses were performed, with a subject's Initial Grief Intensity score representing the covariate (Initial Grief Intensity is the subject's total score on the RLI for the early and middle stages of the husband's illness), and the variables More Intense and Less Intense as the dependent measures. For the independent variable Marital Quality, no significant relationship was found with either the More Intense or Less Intense variables. Relevant means and standard deviations are listed in Table 40. A significant main effect was found for the Social Support variable on Less Intense (E: 5.19, p< .05). Subjects 115 in the "High" Social Support group showed less of a decrease in the intensity of their grief response on average than did subjects in the "Low" Social Support group. Relevant means and standard deviations are presented in Table 41, and results of the analysis concerning Less Intense are shown in Table 42. For the Husband's Acceptance variable, a trend was found for a main effect on More Intense (E3 3.41 , p< .10) . Surviving spouses in the "High" Husband's Acceptance group did not experience as great an increase in the intensity of their grief responses as did spouses in the "Low" Husband's Acceptance group. Means and standard deviations are listed in Table 43, and results of the analysis concerning More Intense are presented in Table 44. i n Am n h D n n i 1 Although no specific hypothesis were constructed concerning relationships among the dependent variables, Pearson Product-Moment Coefficients were calculated to explore these associations. Results of the analyses are presented in correlation matrix form in Table 45. With regard to the Marital Work variable, significant correlations were found with Guilt (;= -.46, p< .001), Jobs (1: -.52, p< .05), and Household Tasks (E: .31, p< .05). A trend was also found for Marital Work to be associated with Wife's Acceptance (£= .27, p< .10). In addition to its association with Marital Work, Guilt was found to be significantly correlated with Depression (1:.40, p< .01). Guilt also tended to be associated with Physical Problems, although this relationship did not quite reach SCOI( Phys COII Hous 116 reach statistical significance (£= .27, p< .10). A very strong relationship was found between subjects' scores on the Depression and Anxiety measures (g: .67, p< .001). Depression was also very strongly associated with Physical Problems (;= .81, p< .001). Lastly, a significant correlation was found between Depression and scores on the Household Tasks variable (t: -.43, p< .01). In addition to the associations previously described, significant correlations were found between Anxiety and Household Tasks (5: -.50, p< .001) and.with Physical Problems (;= .61, p< .001). Regarding the Household Tasks variable, significant correlations were found with Physical Problems (;2 -.32, p< .05), Jobs (g: -.57, p< .01), and Wife's Acceptance (:2 .28, pa .05). Strong associations were found between Grief Intensity and the Guilt (;= .31, p< .05), Depression (;= .66, pr .001), Anxiety (2: .58, p< .001), and Physical Problems (£= .68, p< .001) variables. Sample obtair discu: had a may 1 it i acce on Per: Pro ter to ma QHAPIIER V W Because the sample was not obtained randomly, but was obtained from six Michigan hospices, it is important to discuss aspects of the sampling procedures which might have had an impact on the results of the investigation, or which may limit the generalizability of the findings. By sampling only from among hospice-involved individuals, it is probable that subjects were characterized by greater acceptance of the terminal nature of their spouses' illnesses, on average, than is the general population of bereaved persons. In order for a family to participate in a hospice program, one of the family members must be experiencing a terminal illness. Therefore, it.would.be difficult for a woman to maintain a strong state of denial while, at the same time, make a decision with her husband to become involved with a hospice. Similarly, husbands of women who participated in this study might have been more accepting of their terminal conditions, on average, than is the general population of dying spouses. Since many dying spouses were likely to have been involved in the decisions to join.hospice programs, it is improbable that such choices could have been made while engaging in significant denial concerning their illnesses. Hospice programs may well attract people who are relatively more willing to deal with their grief- and 117 loss-r likely a rela a pro expec indiv deal: 90911 part diff some sub; ave stu 6-1 age av. “o tt tt Pr he loss-related issues in general. If this is the case, it is likely that membership in a hospice might be associated with a relatively low probability of a surviving spouse utilizing a problematic approach to coping with her loss. It would be expected therefore that a sample of hospice-involved individuals would manifest better adjustment, on average, in dealing with their prospective losses, relative to the general population of people who are anticipating similar losses. A surviving spouse's decision whether or not to participate as a :research subject. may itself have: been differentially associated.with coping or adjustment. There is some likelihood that women who agreed to participate as subjects experienced fewer problems in their bereavement on average, than did women who chose not to be a part of the study. Some surviving spouses who were having difficulties 6-12 months after the husbands' deaths may have decided against participating in the study because they wished to avoid what they perceived to be an additional stressor. These women, who were experiencing relatively greater problems after their husbands' deaths, may also have had greater difficulties ‘with.adjustment during the final illnesses. To»the extent that this was true, the sample of surviving spouses was somewhat unrepresentative of the population of surviving spouses as a whole. In some cases, an individual's decision to participate in the study, and previously, to become involved in a hospice program, may have been made because that person was seeking Ihelp>for problems in coping or adjustment. This might suggest 118 that among indi1 Hove' invo like from vhic rel: to her do 5P hi fa We 119 that a higher incidence of adjustment difficulties existed among persons who joined hospice programs, and also among individuals who agreed to participate as research subjects. However, it the investigator's impression that a hospice- involved individual who agreed to become a subject was more likely to have had a high recognition of her need for help from others, rather than having had serious difficulties, which required others' help. This suggests the presence of relatively greater coping skills among the sample, as compared to the general population of surviving spouses. Participation in a hospice program involves significant responsibility for the physical care of the dying family member, more so than is likely to be the case in families who do not choose to request hospice services. The fact that a spouse has expressed a willingness or desire to care for his/her dying partner suggests that the couple experiences a fairly high degree of comfort in their interactions, even under stressful circumstances. One would expect such a pattern of behavior to be associated with a relationship that is relatively high in quality. Moreover, it would be expected that a spouse would experience the prospective loss of such a relationship to be a particularly severe one. It is possible therefore that the choice to participate in a hospice program is most likely to be made by persons whose marriages are relatively high in quality and satisfaction, and by persons who are responding to intense loss experiences. However, there islevidence that indicates that the sample was not atypical in terms of the quality of their marriages. The 1 the ‘ The by t per: stuc mea: for cor sii ha 31' 'A In (no 120 The subjects attained a mean score of 117.7 (gtet- 16.0) on the Dyadic Adjustment Scale, the measure of marital quality. The sample mean was not appreciably different (as determined by t-test) from the mean attained by a married sample of 218 persons who completed the questionnaire as part of a validity study of the DAS (Spanier, 1976). Spanier's sample achieved a mean score of 114.8 (gtet- 17.8) on the DAS. The fact that hospice-involved wives have chosen to care for their dying spouses indicates that these wives have some confidence in their own ability to cope with crisis situations. It also suggests that hospice-involved wives may have more positive self-images on average, than do wives who are not participating in hOSpice programs. In all likelihood, the sample's involvement with hospice programs insured that this group experienced a relatively high degree of social support, more so perhaps than was the experience of surviving spouses generally. It is characteristic of hospices to make emotional and more tangible forms of support readily available to their client families. Moreover, an individual who has become involved with a hospice program has shown some capability of seeking out help from others when it is needed. This indicates that the individual's social skills are at least reasonably well-developed. It also suggests that other sources of social support (in addition to the hospice) might be available. In obtaining the sample, surviving spouses were included only if their husbands' final illnesses were of at least one month's duration. Thus, none of the subjects experienced the 121 loss of a spouse following a very brief illness, nor was any subject coping with the sudden death of a spouse. Since an accident or sudden, short illness is more likely to be the causerof death.in younger persons, it.is probable that younger surviving spouses tended to be excluded from the study. Evidence has previously been discussed linking both younger age and sudden death (versus a more lengthy illness) with heightened stress and greater difficulties in adjustment for the surviving spouse. It is likely therefore that, on average, surviving spouses in the sample might have experienced fewer problems coping with their prospective losses than did the general population of surviving spouses in dealing with their husbands' prospective deaths. Primarily because of the concern about obtaining an adequate sample of male surviving spouses, only women were included in the pool of potential subjects. There is little consensus among researchers as to whether bereavement has a bigger impact upon men or women. Some studies report that men are more greatly affected (Gerber et al, 1975; Stein 8- Susser, 1969; Stroebe a Stroebe, 1983); others report the opposite (Carey, 1980; Parkes, 1964; Sanders, 1981); and some studies report finding no differences (Bornstein et al, 1973; Clayton et al, 1971; Gallagher et al, 1983). For this reason, it is difficult to predict how the exclusion of male surviving spouses might affect the findings in this investigation. It does suggest that caution should be exercised in making generalizations from the findings beyond the population of female surviving spouses. can: he dii Si: in1 re. re to 122 Lastly, although sample size is not relevant to issues concerning the randomness of the sample, it is an important factor in determining the power of the study to detect differences between groups when such differences exist. The size of the current sample (a: 47) suggests that in this investigation, tests to measure interaction effects might be relatively insensitive. The increased number of cells involved restricts the number of subjects per cell, reducing the power to detect differences between cells. D ri t n of ub ec ° m m 1 In the current sample, the experience of a surviving spouse during the time of her husband's final illness appears typically to have been a severe one. Depressed mood and sleeping problems were each reported by more than 90% of the subjects. Crying, fatigue, loss of interest in pleasurable activities, diminished appetite, and irritability were also common, each being reported by more than 50% of the subjects. Moreover, subjects most often experienced a number of these symptoms; 95% of the subjects reported having experienced at least 3 of 8 symptoms about which they were questioned . One explanation for the high level of symptomatology experienced by the sample, is that subjects were reacting to the stress associated with having to care for their seriously ill spouses. The investigator has no doubt that this task was both depressing and physically stressful for all subjects, and that to some degree, symptoms either resulted from or were exacerbated by this stress. However, it is the investigator's 123 belief that the symptoms reported by subjects can best be understood as resulting from the process of anticipatory grief. Each of the eight symptoms which were found to be frequent in the sample are also common aspects of a normal grief experience. The great majority of subjects were likely engaged in the work of mourning their prospective losses. The frequency of symptoms among subjects in the current study was compared with frequencies found among Clayton et. al.'s (1973) subjects. All eight symptoms occurred more frequently in the current sample, with differences achieving statistical significance in five cases - depressed mood, loss of interest, fatigue, irritability, and guilt. One explanation for these results is that Clayton et. al. may have used more stringent criteria than did the current investigator in rating subjects' symptomatic behavior. This could not be determined since little information is available from published reports regarding the way in which Clayton et. al. classified subjects with regard to symptomatology. A second possibility is that the inclusion of males as subjects by Clayton et. al. could account for differences found between the two studies. It might be the case that various grief-related symptoms occur less frequently in men than among women. Since males comprised approximately one- fourth of Clayton et. al.'s subjects, they may have lowered the overall frequency of occurrence of some symptoms. It is the investigator's belief however, that another difference between the two studies in subject selection procedure might best explain differences in symptom freq recr rec1 cer sam mig rel an of me vi '1: th 3C1 124 frequencies. Unlike the current study in which subjects were recruited through a network of hospices, Clayton et. al. recruited subjects through obituary notices and death certificates. It is suggested by this investigator that a sample composed entirely of hospice-involved individuals might, on average, have greater awareness of grief and loss- related issues, and would therefore be more likely to experience grief-related symptomatology, than would a sample of individuals not specifically selected for hospice membership. Heightened awareness of grief is likely to be associated with hospice-involvement for at least two reasons. First, it is difficult for an individual tolengage in significant.denial concerning the terminal nature of a spouse's illness, and at the same time participate in a program in which it is assumed that the spouse is terminally ill. Second, the social support typically provided by hospice programs creates an atmosphere which encourages both awareness and resolution of grief. R i 1i h In r m n It was predicted (Hypothesis I) that the scales constructed for this investigation would achieve acceptable levels of reliability. A test of internal consistency, Cronbach's Alpha, was performed on four of the scales, Marital Work, Guilt, Physical Problems, and Wife's Acceptance. Total scale alphas were greater than .80 for three scales, Marital Work, Guilt, and Physical Problems, indicating that each of these scales is reasonably consistent, and that each scale is accurately measuring one concept. level .63. asse expl for con; P05 the l h In th C01 125 The Wife's Acceptance scale failed to reach the criterion level of reliability of .70, attaining a coefficient alpha of .63. One explanation is that the items in this scale might be assessing too wide a range of behaviors. Another more likely explanation is that an insufficient range of scores existed for some items, reducing their item-total correlations, and consequently, the total scale alpha. Evidence for this possibility is the fact that in response to the statement "By that time, I knew that my husband would die from his illness. I had little or no hope that he would recover", more than 85% of the sample reported that this statement "mostly" or "accurately" described them. It is apparent that the sample was characterized by a fairly high degree of acceptance regarding their husbands' terminal conditions. This is not surprising given the subjects' involvement in hospice programs. A high degree of acceptance was also likely since each subject was asked to recall her response during the last two*weeks prior to her husband's death, a time when there must have been many cues that the death was imminent. If in fact most subjects were high in acceptance, Wife's Acceptance scores were probably restricted in range, and this could account for the low alpha coefficient. In spite of the fact that the Wife's Acceptance scale did not display as high a degree of internal consistency as might be desired, the total scale alpha of .63 indicates that some degree of internal consistency does exist, particularly since the scale is composed of only four items. Moreover, the alpha coefficient indicates that there is a good deal of variability with pm the: suf ICS as: co: Th IE 5!. an 126 within the sample of Wife's Acceptance scores, notwithstanding previously discussed limitations on the range of scores. For these reasons, it has been concluded that the scale is sufficiently reliable that meaningful interpretations of results involving Wife's Acceptance can proceed. The reliability of the Grief Intensity scale was also assessed, using Chronbach's Alpha as the test of internal consistency. The scale achieved a coefficient alpha of .93. This result attests to the capability of the instrument to reliably' measure an individual's «grief response, and is consistent with the 1982 research of Deutsch, who obtained a coefficient alpha of .95 in her sample. To assess the reliability of the Social Support and Husband's Acceptance scales, which are each composed of two items, a Pearson Correlation Coefficient was computed on the items within each scale. The Social Support scale achieved a correlation coefficient of .78, while the Husband's Acceptance scale attained a correlation coefficient of .39. The high correlation between the two Social Support items suggests that this scale is measuring a unitary concept. It also indicates considerable overlap between items, in terms of what is being measured. One of the items assesses the extent to which a subject was able to confide in someone other than her husband about herself or her problems. The other item is (concerned.with the subject's ability to discuss her husband's illness or his dying with someone else. It appears that subjects who talked with others about their husbands' deaths and dying generally were also able to discuss issues and 127 problems less directly related to their husbands' terminal conditions. Lastly, the high inter-item correlation indicates that substantial variability exists within the sample on the Social Support variable, in spite of subjects' hospice involvement, and the likelihood that this was a relatively highly supported group. This variability among subjects could reflect actual differences in the availability of social support, and/or differences in the willingness of subjects to utilize existing sources of support. The moderate correlation between items of the Husband's Acceptance scale suggests that the scale is measuring one construct. It also indicates that the items in this scale are measuring aspects of a husband's attitude towards his terminal condition which are somewhat distinct from one another. Some concern has previously been expressed that few differences in Husband's Acceptance might be found within the sample, given the subjects' involvement with hospice programs. The correlation coefficient of .39 indicates that some variability does exist among the husbands. An examination of the frequencies of scores for the two items reveals substantial variability in terms of the husbands' beliefs concerning the likelihood of their illnesses resulting in their deaths, but much less variation in the degree to which husbands encouraged their wives to discuss feelings about the jprospective deaths. Less than 30% of the husbands provided any encouragement to their wives to discuss this issue. The relative lack of range for this item likely attenuated the correlation between the two Husband's Acceptance items. 128 W Hypothesis II was tested to determine whether or not the scales used is this investigation were measuring what they were designed to measure. It was expected that significant positive correlations would be found between total scores on the nine self-report scales and the hospice worker ratings of the same dimensions. In fact, scores on only one of the self- report. measures, Grief Intensity, were .significantly correlated with hospice worker ratings. This finding provided little evidence to support the validity of most of the measures used in this investigation. It raised the question as to which source of information was the more accurate - the subjects' self-reports or the hospice worker ratings. For a number of reasons, it is suggested that the self- reports were the-more valid.ratings of the subjects' responses during the final illnesses. First, the validity of the self- report data derived from the Depression (BDI) and the Anxiety (STAI) scales has repeatedly been demonstrated in studies which have previously been discussed. However, in the present investigation, correlations between independent, hospice worker ratings of depression and anxiety, and self-report data obtained from the BDI and STAI, respectively, failed to reach statistical significance. Second, the hospice worker ratings of the various aspects of adjustment were found to be highly intercorrelated (see firable 46). This strongly suggests that hospice workers tended to perceive the adjustment of widows in a global manner; at .least this was the case when.the hospice workers were asked to 129 recall the coping ability of widows during the time of their husbands' final illnesses. A number of significant correlations were also found among subject self-report ratings (see Table 45). HDwever, subjects were able to achieve some degree of independence in their ratings of the different aspects of adjustment, much more so than.were hospice workers. As will be discussed later, the fact that significant correlations were obtained among self-report variables reflects not so much a tendency for subjects to perceive themselves globally, but rather reflects a process (grief) which affects some of the variables in a similar manner. Third, in most cases, there is reason to question whether the hospice worker who provided the ratings was qualified to make an accurate assessment of the subject's state of adjustment. Prior to the husband's death, a social worker was primarily responsible for assessing the psychosocial needs of the family, including the needs of the surviving spouse. In the great majority of cases, it.was also the social worker who provided the surviving spouse with counseling for any grief- related issues prior to the death. Although exact statistics are not available, approximately 2/3 of all hospice worker questionnaires were completed by nurses. The remaining questionnaires were completed by social workers or by hospice bereavement coordinators. This occurred because at two of the hospices, Hospice of Jackson and Hospice of Grand Rapids, where the majority of hospice worker questionnaires were obtained, social workers who had been in contact with the 130 family, were unavailable to complete the questionnaires. Although the nurses were likely to have had frequent contact with the surviving spouses prior to the husbands' deaths, especially during the last few weeks of the illnesses (as often as 3-5 times per week), the nurses' primary responsibility was pain management and control. The nurses did not typically participate in the formal psychosocial assessment of the families, nor were they likely to have been involved in providing counseling services to surviving spouses. As a result, opportunities for the nurses to be exposed to information relevant to issues concerning the adjustment of the surviving spouses, were limited. Moreover, at the time the hospice worker questionnaires were completed, between six months and one year had elapsed since the husbands' deaths. In the intervening period, nurses as well as social workers had been in contact with a great many other families. It is probable that for both nurses and social workers, these two factors - lengthy time since the deaths, and large number of other client families seen - resulted in difficulties recalling particular aspects of the adjustment of individual surviving spouses. It is likely that these factors also contributed to a tendency by hospice workers to have "global" perceptions of subjects' adjustment. Although it has been concluded that the subjects' self- report data is preferable, in terms of its validity, to independent hospice worker ratings, self-report data has several potential sources of bias. The need by some subjects to respond in a socially desirable manner may have influenced 131 them to minimize past difficulties in coping or adjustment, or to exaggerate the quality of their marriages. Moreover, a subject might have failed to admit to problems because of the need to«defend against awareness of her loss, in particular to protect herself from the painful feelings associated with the loss. To some extent, these factors are likely to affect subjects' responses, resulting is some distortion of the data, and contributing to an increase in error variance. However, evidence exists which attests to the basic validity of the self-report information. The relatively high intercorrelations among some of the dependent variables indicates a significant degree of variability among subjects' responses. The high correlations would not.be found if few'subjects were reporting problems with adjustment. In addition, the fact that all subjects were involved with hospice programs suggests that subjects may have been more rather than less willing to admit to difficulties in coping with their losses. Lastly, as described previously, the sample mean score on the DAS was only slightly higher than the mean attained by a large sample of married individuals, none of whom were widowed or involved with hospice programs. This indicates that subjects engaged in relatively little distortion in evaluating their marriages, and suggests that in general, information provided by subjects might be quite accurate and valid. n n h m n The primary purpose of six hypotheses (IV, V, VII, IX, XI, XII) was to determine whether a number of variables have 132 a significant influence on the adjustment of a surviving spouse during the time of her husband's final illness. ANOVA's were performed to test for main effects of Marital Quality, Social Support, Husband's Acceptance, Age, and Length of Illness on relevant adjustment variables. In addition, Pearson correlation coefficients were computed for the six predictive variables with each of the measures of adjustment. Four hypotheses (III, VI, VIII, and X) predicted a significant association between Grief Intensity and the following independent variables: Marital Quality, Social Support, Husband's Acceptance, and Age. Again, ANOVA's tested for main effects of the four predictive variables. Pearson correlation coefficients were also computed between these four variables and Grief Intensity. M r 1 It was predicted (Hypothesis IV) that surviving spouses of "high" quality marriages would evidence superior coping or adjustment on measures of guilt, depression, anxiety, physical problems, and the accomplishment of marital work, relative to surviving spouses of "low" quality marriages. For the Marital Work variable, the analysis of variance failed to reveal a main effect for level of Marital Quality. However, a trend towards a positive correlation was found between the two variables (;= .28, p< .10). As predicted, higher marital quality was associated with the accomplishment of more marital work. Further analyses were undertaken to determine which aspect of Marital Quality might be responsible for the 133 association between the two variables. Separate ANOVA's were performed for each of the four factors of the DAS, Cohesion, Consensus, Satisfaction, and Affectional Expression. with Marital.Work.as the1dependent variable. The analyses failed.to find a significant association between Marital Work and any of the four DAS factors. One possible explanation for the association between Marital Work and Marital Quality is suggested by an examination of the items in the Marital Work scale. Every Marital Work item involves the performance of some marital task that is characteristically faced by a couple when one of the spouses is terminally ill. Each of these tasks is likely to be emotionally laden, and appears to require a high degree of communication between the spouses for successful performance. It might be that some couples, who possess superior communication skills, are better prepared to accomplish these tasks. The ability of spouses to communicate well might also be associated with the creation and maintenance by the couple of a successful marriage. For the Depression and Physical Problems variables, it was found that subjects from marriages that were "high" in quality experienced greater depression and more physical problems on average, than did subjects whose marriages were relatively "low" quality ones. For Depression, the association with Marital Quality appeared as a trend in both the analysis of variance and the correlational analysis. For Physical Problems, the correlation with Marital Quality did attain statistical significance at the .05 level. These findings 134 failed to support Hypothesis IV and were opposite to the predictions. One explanation for the results is that the surviving spouse of a "high" quality marriage might experience the relationship with her husband to be more gratifying on average, than does the surviving spouse of a marriage that is "low" in quality. Consequently, the impact of the husband's prospective death might be more stressful for the "high" marital quality spouse. Because she experiences greater stress, the spouse from the "high" quality marriageemight also be at relatively greater risk for developing a pathological grief response to her loss, a response which includes depression and physical problems. An alternative explanation for these findings is that a surviving spouse from a "high" quality marriage who scored high on the Depression and Physical Problems scales, was not exhibiting a problematic response to her loss; rather she was showing an intense grief response as a normal reaction to experiencing a severe loss. This explanation assumes that, for the population of individuals who are coping with the prospective (deaths of their spouses, the» Depression. and Physical Problems scales are most often measuring the intensity of normal grief rather than pathological responses to loss. Significant correlations were found between the measure of Grief Intensity (the RLI) and both the Depression and Physical Problems scales. More intense grief responses were associated with greater depression (£8 .66, p< .001) and more severe physical problems (r= .68, p< .001). 135 One possibility is that the RLI is not measuring normal grief, but is actually measuring depression. However, the RLI was developed for the purpose of measuring normal grief in populations of bereaved individuals, while the BDI was not developed to distinguish depression from grief in such populations. If one adopts Beck and Burmeister's (1974) criterion for diagnosing severe depression (a score of 16 or greater on the BDI), 22 of 47 subjects would achieve this diagnosis, 46.8% of the sample. It seems clear that when the BDI is administered to bereaved persons, the instrument must be measuring something more than clinical depression. Since a strong association has been found between the BDI and the Response to Loss Instrument, it is probable that the additional construct being measured is normal grief. If a husband's prospective death represented a more severe loss to a spouse whose marriage was "high" in quality, that spouse would be expected to exhibit a grief response of greater intensity than that shown by a surviving spouse from a "low" quality marriage. In fact, a significant correlation was found between Marital Quality and Grief Intensity, as measured by the RLI (;= .32, p< .05). As predicted (Hypothesis III), "higher" marital quality was associated with more intense grief. To the extent that the Depression and Physical Problems scales are also measuring normal grief intensity, this could account for the associations found between these variables and Marital Quality. In the analyses to determine which aspects of the marital relationship might contribute to the associations between 136 Marital Quality and the Depression, Physical Problems, and Grief Intensity variables, two DAS factors stood out, Dyadic Consensus and Affectional Expression. Women whose marriages were marked by greater agreement on important marital matters (the "High" Dyadic Consensus group) experienced significantly more intense grief than did women whose marriages were characterized by less agreement on such issues. In addition, a trend was found for women whose marriages were "high' in Dyadic Consensus to experience greater depression relative to women in the "Low" Dyadic Consensus group. A spouse whose marriage was marked by agreement rather than conflict over important marital issues likely experienced the prospective loss to be a greater one that did a spouse from a more conflictual marriage. The former would be expected to exhibit a more severe grief response, reflected by higher scores on instruments such as the RLI, BDI, and the Physical Problems scales, all of which appear to be sensitive to the intensity of an individual's normal grief response. More intense grief, as measured by the RLI, was also associated with marriages in which the surviving spouse was relatively happy with the way affection and sex were expressed. Again, it seems that a spouse who was satisfied with this important aspect of the marital relationship likely experienced the prospective loss to be a greater one than did a less satisfied spouse. It is not surprising that the former spouse also experienced a more intense grief response to her loss. For the Guilt variable, it was predicted that surviving 137 spouses whose marriages were "high" in quality would experience less guilt, relative to surviving spouses from "low" quality marriages. This prediction was based on psychodynamic and existential theories of grief, both of which argue that the loss of a relationship that is ambivalent or highly conflictual generally leads to more guilt for the surviving spouse than does the loss of a happier, less conflictual relationship. The analyses however, failed to discover a significant association between Marital Quality and Guilt. One explanation for this result is that the Guilt scale might not be a valid measure of the construct. Prior to this investigation, no validity data existed on the scale. However, in this study, a significant and meaningful association was found between Guilt and one of the independent variables, Husband's Acceptance, lending support to the construct validity of the instrument. Another possibility is that during the time of the husband's final illness, the severity of guilt experienced by the surviving spouse may not have a strong association with the quality of the marital relationship. Issues which are related to the husband's illness might have the greatest potential to»induce guilt, and these issues may have little to do with longstanding marital concerns. Two such issues might be the degree to which the husband is able to accept the prospect of his own death; and the ability of the surviving spouse to perform tasks related to her husband's care. It is possible that differences did exist between 138 surviving spouses of "high" and "low" quality marriages, in terms of their guilt experiences. A surviving spouse whose marriage had been a successful one might have evaluated herself more positively with regard to her performance in the marriage, and therefore had fewer guilt feelings to deal with than did a spouse from a less successful marriage. At the same time, the "high" marital quality spouse might have had greater awareness of her guilt feelings, relative to the spouse from the "low" quality marriage. Perhaps people who succeed in their marriages have better access to all their feelings, including guilt. Alternatively, a woman who has an overall sense of satisfaction concerning her marriage might be better able to consider her own shortcomings without experiencing a threat to the evaluation of herself as a good marital partner. One factor which is likely to have had a significant impact on a person's awareness of guilt feelings is the severity of the loss. More severe losses may well intensify all of the emotional experiences of an individual, including a person's awareness of guilt. If spouses from "high" quality marriages had fewer guilt-related issues to deal with, but had greater awareness of these issues, this would explain the similarity of the scores on the Guilt measure of the "High" and "Low" Marital Quality groups. In this investigation, a significant relationship was found between the Guilt and Grief Intensity scales (I‘ .31, p< .05). The finding indicates that the Guilt scale might be measuring normal, grief-related guilt, as well as neurotic or pathological guilt. It also suggests that an individual's 139 awareness of guilt feelings is to some extent a reflection of the severity of the loss suffered by that person. To the extent that a surviving spouse from a "high" quality marriage experienced a more severe loss than did a surviving spouse from a marriage "lower" in quality, the former spouse might have greater awareness of guilt feelings. For the Anxiety variable, itwwas predicted.that surviving spouses of "high" quality marriages would experience less anxiety than would surviving spouses whose marriages were "lower" in quality. Results of the analyses did not support the prediction. No significant association was found between Anxiety and Marital Quality. Subjects were asked to recall the degree of anxiety they experienced during the last two weeks of their husbands' illnesses. This was likely to have been a highly stressful time for all the surviving spouses, as they anticipated their husbands' deaths. Not surprisingly, spouses from both "high" and "low" quality marriages attained fairly high scores on the measure of anxiety (41.36 and 40.68, respectively). A comparison with normative data on the STAI (Spielberger, 1983) reveals that both "high" and "low" marital quality spouses scored approximately one standard deviation higher than Spielberger's sample of 50-69 year-old women. It seems reasonable to conclude that, similar to the Depression and Physical Problems scales, the Anxiety scale is, at least in part a measure of normal grief. Further support for this conclusion is provided by the finding that scores on the STAI correlated .58 (p< .01) with the RLI, the measure of grief 140 intensity. To the extent that high levels of anxiety were common among subjects in the current sample, any differences among subjects' scores on the Anxiety measure which could be attributed to differences in Marital Quality would be attenuated. An alternative or additional explanation for the finding of no differences in Anxiety is that "high" and "low" quality marriages have different issues associated with the loss of a spouse, but for each group, the issues tend to increase anxiety for the surviving spouses. For a surviving spouse of a "high" quality marriage, the husband's prospective death likely represented.a particularly severe loss. Anxiety might have been heightened for this spouse as she anticipated the death. For a surviving spouse whose marriage was "lower" in quality, the husband's death could have represented the loss of any opportunities to resolve marital difficulties. In addition, she was likely to have felt some responsibility for the lack of success of her marriage. Therefore, the loss of that relationship might have been associated with the threat of significant guilt and diminished self-esteem. This surviving spouse might have experienced a high degree of anxiety prior to her husband's death as she anticipated having to deal with these issues. W Hypothesis V predicted that surviving spouses who experienced a "high" degree of social support during the time of their husbands' final illnesses would exhibit better 141 adjustment on measures of depression, anxiety, physical problems, guilt, the accomplishment of marital work, and acceptance of the terminal nature of their husbands' illnesses than would surviving spouses who experienced a "low" level of social support during that time. The results of the analyses involving the Marital Work variable provided some support for this prediction. "High" Social Support subjects accomplished significantly more marital work with their husbands than did subjects who were "low" in Social Support. One explanation for this finding is that the social networks of women in the "High" Social Support group created an atmosphere which encouraged these women to accomplish marital work with their husbands. In some cases, a friend or confidante might have directly encouraged a spouse to deal with marital issues. Other times, a spouse might have felt able to undertake difficult marital work because of the emotional support she experienced while discussing these issues with another person. An alternative or additional explanation is that women whose interpersonal skills allowed them t03 ('6 6 9%‘<%, q? ab 6? 4%. fig 23. I had more accidents or injuries than before the o 1 2 3 illness. 29. I was exercising. O l 2 3__ 30. I was neglecting or punishing my body. 0 l1 ‘3 3 31. I was more active in caring for myself physically O l 2 3 than I was before the illness. 32. What I ate was healthy. 0 l 2 3 33. I could understand what my body needed better than 0 l 2 3 before the illness. 34. Touching and being touched were important to me. 0 1 2 3 35. Choose the one statement that best describes your smoking habits during the lag; gwg week; before vgg; husband's death: _____ I was a non-smoker I was smoking less than prior to his illness _____,I was smoking about as much as prior to his illness I was smoking somewhat more than prior to his illness I was smoking much more than prior to his illness 238 ***Answer the following questions in terms of your husband'sresoonse‘during'the‘z weeks before his death. Pick out the statement that best describes your husband's response during that time. Place an "x" next to that statement. 36. By that time. my husband believed that he would die from his illness. He had little or no hope for recovery. did not describe him described him somewhat mostly described him accurately described him I did not know whether my husband believed he would survive the illness. 37. Hy husband encouraged me to discuss with him my feelings about his dying. did not describe him described him somewhat mostly described him accurately described him 239 On this questionnaire are groups of statements. Please read each group of statements carefully. Then pick out the one statement in each group that best describes the way you were feeling during the 2 weeks just before your husband died. Place an"x” next to that statement. 38. 39. 40. 41 42 43 44 I did not feel sad. I felt sad. I felt sad all the time and couldn't snap out of it. I was so sad or unhappy that I couldn’t stand it. I was not particularly discouraged about the future. I felt discouraged about the future. I felt I had nothing to look forward to. I felt that the future was hopeless and that things could not improve. I did not feel like a failure. I felt that I had failed more than the average. As I looked back on my life. all I could see was a lot of failures. I felt that I was a complete failure as a person. I got as much satisfaction out of things as I used to. I didn't enjoy things the way I used to. I didn't get real satisfaction out of anything anymore. . I was dissatisfied or bored with everything. didn't feel particularly guilty. felt guilty a good part of the time. felt quite guilty most of the time. felt guilty all the time. HHHH didn't feel I was being punished. felt that I may be punished. expected to be punished. felt I was being punished. didn't feel disappointed in myself. was disappointed in myself. was disgusted with myself. hated myself. HHHH HHHH 45. 240 I didn't feel I was any worse than anybody else. I was critical of myself for my weaknesses or mistakes. I blamed myself all the time for my faults. I blamed myself for everything bad that happened. 46. I didn't have any thoughts of killing myself. I had thoughts of killing myself. but I would not carry them out. I would have liked to kill myself. I would have killed myself if I had the chance. 47. I didn't try anymore than usual. I cried more than I used to. I cried all the time. I used to be able to cry. but then I couldn't cry even though I wanted to. 48._____ I was no more irritated than I ever am. I got annoyed or irritated more easily than I used to. I felt irritated all the time. I didn't get irritated at all by the things that used to irritate me. 49._____ I had not lost interest in other people. I was less interested in other people than I used to be. I had lost most of my interest in other people. I had lost all of my interest in other people. 50. I made decisions about as well as I ever could. I put off making decisions more than I used to. I had greater difficulty making decisions than before. I couldn't make decisions at all anymore. 51. I didn't feel I looked any worse than I used to. I was worried that I was looking old or unattractive. I felt that there were permanent changes in my appearance that made me look unattractive. I believed that I looked ugly. 52. I could work about as well as before. It took extra effort to get started at doing anything. I had to push myself very hard to do anything. I couldn't do any work at all. 53. 54. HHHH 241 I could sleep as well as usual. I didn't sleep as well as I used to. I awakened 1-2 hours earlier than usual and found it hard to get back to sleep. I awakened several hours earlier than I used to and could not get back to sleep. I didn't get more tired than usual. I got tired more easily than I used to. I got tired from doing almost anything. I was too tired to do anything. My appetite was no worse than usual. My appetite was not as good as it used to be. My appetite was much worse. I had no appetite at all anymore. hadn't lost much weight. if any lately. had lost more than 5 pounds. had lost more than 10 pounds. had lost more than 15 pounds. I was purposely trying to lose weight by eating less. Yes No 57. 58. I was no more worried about my health than usual. I was worried about physical problems such as aches and pains; or upset stomach: or constipation. I was very worried about physical problems and it was hard to think of much else. I was so worried about my physical problems. that I couldn't think about anything else. I had not noticed any recent change in my interest in sex. I was less interested in sex than I used to be. I was much less interested in sex. I had lost interest in sex completely. 242 *** Answer the items below in terms of your response only during the last 2 weeks before your husband's death. Put an "x“ in the box under the statement of your choice. 6? O’qh (b q? 6; ?%\‘§? fib <5 ‘z’ Qp a Q, J- .A 0' 0' «a 0" *0 0“ 9" 9% \r c; Q. e‘ 0 " e‘ 6 <9 6 .5 Q 0‘. 9? a o '5 9' 0 lg, 1% a 59. When I thought about what I was losing. I felt that O l ‘2 3 I had nOthing to look forward to. 60. I had many feelings about what I was losing. 0 ’1 l2 3 61. I daydreamed about scenes from before my husband's 0 l ,2 ’3 illness. 62. I was less frightened about dying than I was before 0 1 ’2 3 my husband's illness. 63. I thought about what.I was losing and I thought about. a 1 2 3 how my life was being affected. 64. I often wept or sobbed about what I was losing. 0 1 3 3 65. I felt as if losing my husband was a reminder of the O 1 2 3 limitations of my human power. 66. I was aware of what would never again be part of my 0 1 2 3 life because of what I was losing. . 67. I felt angry about some of the consequences of what a 1 2 3 I was losing. 68. My beliefs gave me as much comfort as they did 0 1 3 3 before my husband's illness. 69. I thought about what I was losing. a 1 2 3 70. I felt sadness whenever I was reminded of what I was 2 3 losing. 0 1 2243 *6 or (3 a we 0" $0 96 9' 0A u" (0 I 2 9° '56 J. J- “. r r. a a, 45 Q; Q; Q;,~%, {3 e, <% 1%k a? %’ 4 9 9 9g 96 O G 71. My dreams about what I was losing seemed to help me o 1 '- 3 understand and accept it. ' ‘ 72. Losing my husband challenged some of my most 0 1 "2 3 cherished beliefs. 73. I knew that my husband would never return. 0 l 2 3 74. I was angry with some people associated with my 0 1 I: 3 husband's illness. 75. My faith was shaken by what I was losing. 0 1 I: 3 76. When I admitted it to myself. I felt sad about what a 1 2 3 I was losing. ' 77. I spent time sifting through past experiences related 0 1 l2 3 I to what I was losing. 78. My tears were hard to stop. 0 1 3 3 79. I had vivid dreams about people and places related 0 1 2 3 to my husband. 80. Before my husband's illness. I believed that I was- 0. 1 2 3 special and nothing bad would happen to me; I continued to believe this. 81. I knew that I was helpless to change the situation o ‘ 2 3 and bring back what I was losing. * 82. I felt guilty about some things I did or did not do 0 l 2 3 before or during my husband's illness. 83. I found myself longing for the past with‘my husband. a 1 2 3 84. Many more people irritated me than did before the O l 2 3 illness. 85. I looked for some way to fit the loss of my husband 0 l 2 3 into the rest of my life. 86. Almost anything could remind.me of what I was losing; 0 1 2 3 87. 88. 89. 90. 91. 92. 93. 244 o' g; 99 Q5 0’ Q: ¢)g§ p a Q- q; .2 f; o; a“ <7 m o m Q3 f. Q~_ ac 5? “do ‘3 4 ‘7 43 fi‘ 29 ‘2 In comparison to what I was losing. everything else 0 ll '2 3 seemed trivial and meaningless. I felt empty. like a shell. like I was just existingu o ‘ "2 [3 I I felt confused and disoriented. O '1 '2 '3 I felt numb. O '1 '2 3 There were parts of the illness or loss experience 0 '1 '2 3 that disgusted me. ___ I felt so lonely and alone when I thought of my a ,1 '2 3 husband being gone. I felt as if luck. fate. or God must be against me: 0 "1 "2 '1‘] I felt like a victim. ’ ' 1245 Directions: A number of statements which people have used to describe themselves are given below. Read each statement and then make an ”x" in the appropriate box to , the right of the statement to indicate how you felt’v "04 ’3; during Egg lagt two gggkg bgfiggg ygg; hugbggg'g gggsh. 9, my 9; *4 There are no right or wrong answers. Do not spend too 7). ’6, ’31, ’6‘. much time on any one statement but give the answer which 1*? ‘32,) “b ‘9 best describes the way you felt. ‘5 O 94. I felt calm.. i. __ 95. I felt secure. _ - 96. I felt tense.-. 97. I felt strained. 98. I felt at ease —_-——- 99. I felt upset. g.- 100. I was worried over possible misfortunes. '4 101. felt satisfied. . - .4 “-a. - . I 102. I felt frightened. I 103. felt comfortable. -.-‘._-—.---- ... --- - —-~ ... 104. I felt self-confident 105. I felt nervous was jittery. felt indecisive." felt content._ —- . . .-..... . ..- I I 108. I was relaxed-.. _ .....H.. I I was worried._“____-m."___.w"_-.“m I felt confusedh__n_ .4 F‘ 04 w» P' r» 04 P' F‘ " " 112. I felt steady.__.,-w i 113. felt pleasanta,. O 246 The questions you are about to answer have to do with your response during the entire time of your marriage. 2247 Host persons have disagreements in their relationships. Please indicate below the approximate extent there was agreement or disagreement between you and your husband for each item on the following list during your entire marriage. Put an 'x" in the box under the statement of your choice. 9’ on has. 9, r0 ”5% o/QQCOJ‘ t/ng' 2L%1;‘L o 2. wihiiii‘i s. 9 O 3 o,,.9.fi,fi ‘90:”,9‘; . 0 abeeee 0“ o %°%Q%%v%o°o 0 1’2 aiais 1. Handling family finances i t o 1 1 3 4 s 2. Matters of recreation ' O l 2 3‘4 5 3. Religious matters 1 O 1 4. Demonstrations of affection 2 3 ‘ 5 o 1 2 3i4 s 5. Friends i l r o 1 2'3 4 s 6. Sex relations 7. Conventionality (correct or proper behavior) 0 1 2 3" 5 i————— O 8. Philosophy of 11:. 1'2 3 ‘ 5 O 1 2 3 4 5 9. Ways of dealing with parents or in-laws i O 1 2 10. Aims, goals. and things believed important 3 4 5 O 1 2 3 4 5 11. Amount of time spent together 248 5 5 5 i— 5 4 4 4 4 3 3 3 3 i‘ . I 2 2 2 2 1 1 1 1 O O O O s Q .1 t .1 V .1 t c a d n & s s n t o s .1 e 3 r .1 e c t s e s n n d k .... o s .1 r 3 O 3 O t m .1 .3 .1 c I d t C m 1 d o e g h I r n C u G i s s e .K H .1 r C O C C H H L C 2 3 4 5 1 l 1 1 249 ***Answer the items below in terms of your reSponse during the entire time of your marriage. Indicate your choice by putting an "x" in the box. 16. How often did. you. discuss or did you consider divorce. separation. or terminating your relationship? 17. How often did you or your mate leave the house after 18. 19. a fight? In general. how often did you think that things between you and your partner were going well? Did you confide in your husband? Did you ever regret that you married? How often did you and your husband quarrel? how often did you and your husband "get on each other's nerves”? For the following items. indicate your choice by placing an ”x” next to that statement. 23. Did you kiss your husband? never rarely occasionally almost every day every day 250 24. Did you and your husband engage in outside interests together? all of them most of them some of them very few of them none of them How often would you say the following events occurred between you and your husband? 25. Bad a stimulating exchange of ideas. never less than once a month once or twice a month once or twice a week once a day more often N 6. Laughed together. never less than once a month once or twice a month once or twice a week once a day more often lllli 27. Calmly discussed something. never less than once a month once or twice a month once or twice a week once a day more often N 8. Worked together on a project. never less than once a month once or twice a month once or twice a week once a day more often 251. These are things about which couples sometimes agree and sometimes disagree. Indicate if either item below caused differences of opinions or were problems in your relationship. Indicate your choice (yes or no) by putting an "x" in that box. 29. Being too tired for sex. Y" N° Yes No 30. Not showing love. 31. The dots on the following line represent different degrees of happiness in your relationship. The middle point. ”happy" represents the degree of happiness of most relationships. Please circle the dot which best described the degree of happiness. all things considered in your relationship. annfly away Auunue lung may Dmnnfly Bum“: Mm mm mm 32. Which of the following statements best describes how you felt about your relationship? Indicate your choice by placing an "x" next to that statement. I wanted desperately for my relationship to succeed. and went to almost any length to see that it did. ' I wanted very much for my relationship to succeed. and did all I could to see that it did. I wanted very much for my relationship to succeed. and did my fair share to see that it did. It would have been nice if my relationship had succeeded. but I couldn't do much more than I did to help it succeed. It would have been nice if my relationship had succeeded. but I refused to do any more than I did to help it succeed. My relationship could never have succeeded. and there was no more that I could have done to keep the relationship going. 252 u" r o'W.r The following questions refer to different aspects of coping or adjustment by a surviving spouse during the time of her husband's final illness. For each item. please rate the adjustment of . Indicate your rating by making an "x" on the line next to that statement. Skip any item for which you feel unable to make a rating. 1. During her husband's illness. rate the ability of the surviving spouse to carry'out.her responsibilities for the physical care of her husband. showed exceptional ability to carry out her responsibilities. was very able to carry out her responsibilities. was mostly able to carry out her responsibilities but sometimes was unable to do so. showed only limited ability to carry out her responsibilities. was never or almost never able to carry out her responsibilities. 2. During the final illness. rate the ability of the surviving spouse to complete important marital work with her husband (This may include resolution of past or present marital problems. reminiscing about times spent together and summing up the marriage. sharing her feelings about.the prospective death. planning for her future. carrying out any last wishes. and saying final good-byes). showed exceptional ability to complete marital work. was very able to complete marital work. was mostly able to complete marital work. but sometimes was unable to do so. showed only limited ability to complete marital work. showed little or no ability to complete marital work. 253 3. During the final illness. rate the ability of the surviving spouse to meet any responsibilities she may have had in terms of employment. care of dependent children. and household tasks. showed exceptional ability to carry out her responsibilities. was very able to carry out her responsibilities. was mostly able to carry out her responsibilities. but sometimes was unable to do so. showed only limited ability to carry out her responsibilities. was never or almost never able to carry out her responsibilities. 4. During the final illness. rate the extent to which the surviving spouse was able to begin to grieve the prospective loss of her husband. showed exceptional ability in completing grief work. was very able to complete grief work. was mostly able to complete grief work. but sometimes was unable to do so. showed only limited ability to complete grief work. was never or almost never able to complete grief work. 5. Durigg the 1a§t gwg wgekg begogg hg; nugbgng': death. rate the extent to which the surviving spouse experienced problems with her physical health i.e. problems with sleep»or appetite. loss of energy. aches and.pains. substance use. accidents/injuries. or other illnesses. very severe physical problems severe physical problems moderately severe physical problems mild physical problems very mild or absence of physical problems 0‘ 5 wow ' , rate the extent to which the surviving spouse experienced feelings of guilt. very severe guilt severe guilt moderately severe guilt mild guilt very mild or absence of guilt 254 7. 'n t wo w ks e e h a ' d t . rate the ability of the surviving spouse to accept that her husband's illness was a terminal one. showed exceptional ability to accept the terminal nature of her husband's illness. was very able to accept the terminal nature of her husband's illness. was mostly able to accept the terminal nature of her husband's illness. but sometimes was unable to do so. was mostly unable to accept the terminal nature of her husband's illness. was never or almost never able to accept the terminal nature of her husband's illness. 8. 'n w we k e e a ' t . rate the extent to which the surviving spouse experienced depression (as distinguished from normal grief). very severe depression severe depression moderately severe depression mild depression very mild or absence of depression 9. e a wo w ' rate the extent to which the surviving spouse was experiencing feelings of anxiety. very severe anxiety severe anxiety moderately severe anxiety mild anxiety very mild or absence of anxiety 255 10. In your opinion. rate the overall adjustment of the surviving spouse during the time of her husband's final illness. showed exceptional coping ability and adjustment coped or adjusted very well coping or adjustment was mostly good showed only limited ability to cope or adjust showed very poor coping ability or adjustment 0 11. During the final illness. rate the intensity of the grief response of the surviving spouse to the prospective loss of her husband. extremely intense grief response intense grief response moderately intense grief response mild grief response very mild or absence of grief response The following item refers to the response of the husband to his terminal illness. 12. thing the 1a§t two wgghg hgfiggg hi; dgagh. rate the extent to which the husband was able to accept the terminal nature of his illness and acknowledge this to his wife. showed exceptional ability to accept the terminal nature of his illness. was very able to accept the terminal nature of his illness. was mostly able to accept the terminal nature of his illness. but sometimes was unable to do so. was mostly unable to accept the terminal nature of his illness. was never or almost never able to accept the terminal nature of his illness. 256 Dear It is our sincere hope that you are in the process of emocional healing as you recover from the death of your husband. As you can understand. we are always seeking new insight regarding the needs of our patients and families. With that thought in mind. our hospice would like to participate in a bereavement research survey. The accompanying letter from Mr. Peradotto explains the intent of the survey. I underscore the words voluntary and confidential. Your participation. and that of others like you. will facilitate the efforts of other hospices and those professionals working with the bereaved to be more aware of their needs and more effeccive. If you are willing to participate. please complete the enclosed post card and return it to us through the mail. Mr. PeradOtto will then contaCt you by phone to schedule a time for you to complete the survey. Sincerely. Janet D. Grabinski. B.S.. M.R.E. Director of Volunteer Services 257 MICHIGAN STATE UNIVERSITY Department of Psychology Dear I want you to know how much I appreciate your willingness to consider participation in the bereavement research survey which I am conducting for my doCtorel research project. It is hoped that the results of the survey will provide increased knowledge and awareness of the bereavement process of widows. and make it possible to provide more effeCtive services for those who are grieving the loss of a spouse. I. of course. hope that enough widows will make a decision to participate in the study in order to complete the research. But please understand that your participation in this study is completely voluntary. therefore you are free to decide REE to participate. In addition. you may discontinue your participation at any time during the survey. without pressure to continue. Participation would require the completion of a written questionnaire that takes about 1 to 2 hours to complete. Host of the questions revolve around the possible responses a wife has to her husband's terminal illness. Some have to do with thoughts or feelings you might have experienced during your husband's illness. Others relate to your actions at that time. A few questions pertain to your husband's attitude toward his illness. Finally. there are questions concerning various aspects of your marriage. The hospice worker is also asked to complete a brief questionnaire. Your identity and that of other subjects will be protected and your answers to all questions kept in strict confidence. Within these restrictions. the results of the study will be made available to you at your request. It is important that you be aware that participating in the survey may involve some emotional distress for you. The questionnaire asks you to recall a period of your life that understandably was a highly stressful one for you - the time of your husband's final illness. If at any time you wish to discuss thoughts or feelings that occur as a result of responding to an item. I will be there with you to facilitate your expression of your feelings. 258 It is hoped that you will benefit from your participation in this study as you re-examine an important period of your life. Although this cannot be promised. you can be proud and perhaps find comfort knowing that your participation will increase knowledge and awareness that may be helpful to other widows. If you have any questions about this research project. please feel free to discuss them with me or with the hospice staff. Should you decide to participate as a subject in this study. please read and sign the enclose post card. Thank you very much. Sincerely. David Peradotto. M.A. Bereavement Research Project Hichigan State University APPENDIX C SCORING KEY 259 260 Social Support The scoring for both.Social Support items was as follows: did not describe me . 0 described me somewhat c 1 mostly described me a 2 accurately described me = 3 Marital Businggg 1. Just before your husband's final illness, to what extent were there marital issues that had not been thoroughly discussed. If the subject endorsed item # 1 (reported that there were unfinished marital issues at the time of her husband's final illness), she was asked if a discussion took place between the couple concerning marital issues, and if so. who mostly spoke. The subject was also asked how she felt the discussion went. These items were scored as follows: 2. During the entire time of his illness, when you and your husband discussed past issues from your marriage that had never been thoroughly discussed, who mostly talked: Scoring: neither of us - 0 mostly my husband - 1 mostly me a 2 both of us about equally a 3 2a. Which of the following words would best describe how you felt these discussions went: Scoring: unfinished = 0 mostly unfinished = 1 mostly finished - 2 finished a 3 If the subject did not endorse item # 1 (denied the existence of unresolved martial issues). but she received a "high denial" score on items 21 and 22 of the DAS (received a score of 5 on each item) = O; the subject denied the existence of unresolved marital issues and did not receive a "high denial" score . 6. 8. 261 I was able to carry out my husband's last wishes before he died - 3 I was mostly able to carry out my husband's last wishes before he died a 2 I was mostly unable to carry out my husband's last wishes before he died = 1 I was unable to carry out my husband's last wishes before he died 2 0 If you were not able to carry out all your husband's last wishes before he died, please explain why you were unable to do so. Scoring: The husband's last wishes were carried out after the husband's death a 3; the husband did have last wishes but there was a good reason why the subject did not carry them.out (as assessed by the interviewer) = 3; the husband had last wishes which were not carried out by the subject . 0. 3. The following items were scored identically: When you and your husband discussed your awareness that his illness would most likely result in his death, who mostly talked. When you and your husband discussed thoughts and plans for your future, who mostly talked. . When you and your husband.discussed memories of times spent together, who mostly talked. Scoring: neither - 0 mostly him a 1 mostly me = 2 both of us about equally - 3 The following items were scored identically: 3a, 4a, and 5a Scoring: unfinished a 0 mostly unfinished a 1 mostly finished c 2 finished - 3 262 10. I said my good-byes to my husband. Scoring: did not describe me . 0 described me somewhat a 1 mostly described me = 2 accurately described me .3 11. Choose the statement that.best describes how'important you felt it was to try to be with your husband at the moment of his death Scoring: very important = 3 important - 2 somewhat important not important . 0 rr in R n i The scoring for all three Carrying Out Responsibilities items were as follows: I was able to perform these tasks better than prior to his illness 3 4 My level of performing these tasks was about the same as before this illness . 3 My performance of these tasks was somewhat less during the illness than prior to the illness = 2 My performance of these tasks was significantly less during my husband's illness than prior to the illness = 1 During my husband's illness, I was able to do almost nothing in the way of performing the tasks that I did prior to his illness - O. h i l The item which asked the subject to rate her ability to meet her responsibilities for the physical care of her husband was scored as follows: I was always able to carry out any responsibilities I had for the physical care of my husband - 3 I was mostly able to carry out my responsibilities for his physical care a 2 I was often unable to carry out my responsibilities for his physical care a 1 263 I was never or almost never able to meet my responsibilities for the physical care of my husband s 0 At times when.I was unable to provide any of the physical care my husband needed, I found other people who could provide the care a 3. h: Wif-': : -. ;n - .- . of 1‘ H ..n.'; -rm n- m The following items were scored identically: 1. By that time, I knew that my husband would die from his illness. I had little or no hope that he would recover. 2. I could accept that he would never return. 3. At that time, I could accept that this loss was real. Scoring: did not describe me a 0 described me somewhat - 1 mostly described me a 2 accurately described me . 3 4. I did things to avoid thinking about the possibility of losing my husband. Scoring: did not describe me = 3 described me somewhat = 2 mostly described me = 1 accurately described me = 0 9.11111; The scoring for all fourteen Guilt items was as follows: did not describe me - 0 described me somewhat s 1 mostly described me . 2 accurately described me = 3 grief Intensity All of the questionnaire items included in the Grief Intensity measure (except items 12, 13, and.29) were scored.as follows: did not describe me a 0 described me somewhat s 1 mostly described me x 2 accurately described me x 3 264 Items 12, 13, and 29 were scored as follows: did not describe me . 3 described me somewhat a 2 mostly described me a 1 accurately described me = 0 m l n f W rk Progress in the resolution of grief issues was defined as any decrease or increase in the intensity with which an item was endorsed, from the initial rating by the subject, to her rating for the last two weeks before the death. The intensity of the subject's response was computed for the initial rating and for the rating of the last two weeks prior to the death, using the following system for all items except items 68 and 80: did not describe me - 0 described me somewhat s 1 mostly described me a 2 accurately described me . 3 Items 68 and 80 were scored as follows: did not describe me = 3 described me somewhat . 2 mostly described me = 1 accurately described me = 0 The subject's score for each item was the absolute difference between the two ratings: the subject's initial score minus the subject's score for the two*weeks prior to the husband's death. To summarize, a subject received 1-3 points for any item in which the intensity of the subject's response changed over the course of her husband's illness. 265 W The following questionnaire items were scored identically: 19. My stomach really churned when I thought about this loss. 20. 21. 22. 23. 24. 25. 26. 27. 28. 30. I had difficulty getting to sleep. I awakened during the night more often than I did before the illness. I slept poorly I slept too much. I was eating too much. When I was reminded of what I was losing, my whole body felt heavy. I had more aches and pains than before the illness. During that time, I had symptoms like my husband had. I had more accidents or injuries than before the illness. I was neglecting or punishing my body. Scoring: did not describe me = O 29. 31. 32. 33. 34. described me somewhat . 1 mostly described me a 2 accurately described me a 3 The following items were scored identically: I was exercising. I was more active in caring for myself physically than I was before the illness. What I ate was healthy. I could understand what my body needed better than before the illness. Touching and being touched were important to me. Scoring: did not describe me = 3 described me somewhat a 2 mostly described me a 1 accurately described me 2 0 Five items from the Beck Depression Inventory were 266 considered relevant to a subject's physical functioning and also included as part of the Physical Problems Scale. The scoring of these items was as follows: 45. 46. 48. 49. 52. I didn't feel I looked any worse than I use to a 0 I was worried that I was looking old or unattractive . 1 I felt that there were permanent changes in my appearance that made me look unattractive - 2 I believed that I looked ugly . 3 I could work about as well as before a 0 It took extra effort to get started at doing anything a 1 I had to push myself very hard to do anything - 2 I couldn't do any work at all - 3 didn't get more tired than usual - 0 got tired more easily than I used to s 1 got tired from doing almost anything 2 2 was too tired to do anything. HHHH My appetite was no worse than usual . 0 My appetite was not as good as it used to be = 1 My appetite was much worse a 2 I had no appetite at all any more a 3 I had not noticed any recent change in my interest in sex a O I was less interested in sex than I used to be a 1 I was much less interested in sex = 2 I had lost interest in sex completely . 3 Because of its relevance to a subject's physical functioning, one item from the State—Trait Anxiety Inventory was also included as part of the Physical Problems Scale. The scoring of this item was as follows: 96. I felt tense. Scoring: not at all x 0 somewhat a 1 moderately so a 2 very much so - 3 267 35. Choose the one statement that best describes your smoking habits during the last two weeks before your husband's death. I was a non-smoker = -1 I was smoking less than prior to his illness = -1 I was smoking about as much as prior to his illness s 0 I was smoking somewhat more than prior to his illness s 1 I was smoking much more than prior to his illness 2 2 Scoring: Th' H ‘o;no': ' _ 0’ o‘.r0f Hi: ‘rm . on. -.n 1. By that time, my husband believed that he would die from illness. He had little or no hope for recovery. Scoring: did not describe him . 0 described him somewhat - 1 mostly described him - 2 accurately described him = 3 I did not know whether my husband believed he would survive the illness =0 2. My husband encouraged me to discuss with him my feelings about his dying. Scoring: did not describe him a 0 described him somewhat s 1 mostly described him - 2 accurately described him a 3 M ri 1 li Items 1-15 of the DAS were scored as follows: always agree 2 5 almost always agree =4 occasionally disagree a 3 frequently disagree . 2 almost always disagree 8 1 always disagree - 0 Items 16, 17, and 20-22 were scored as follows: all the time a 0 most of the time - 1 more often than not a 2 occasionally = 3 rarely a 4 never a 5 268 Items 18 and 19 were scored as follows: all the time s 5 most of the time s 4 more often than not - 3 occasionally . 2 rarely a 1 never - O 23. Did you kiss your mate? Scoring: every day =4 almost every day = 3 occasionally s 2 rarely - 1 never . O 24. Did you and your mate engage in outside interests together? Scoring: all of them - 4 most of them a 3 some of them = 2 very few of them a 1 none of them a 0 Items 25-28 were scored as follows: never 2 0 less than once a month a 1 once or twice a month . 2 once or twice a week = 3 once a day a 4 more often a 5 Items 29 and 30 were scored as follows: yes s 0 no - 1 Item 31, which asks the subject to rate the degree of happiness in her relationship‘with.her husband, was scored as follows: extremely unhappy s 0 fairly unhappy - 1 a little unhappy = 2 happy - 3 very happy - 4 extremely happy = 5 perfect - 6 269 32. Which of the following statements best describes how you feel about the future of your relationship? Scoring: I wanted desperately for my relationship to succeed, and went to almost any length to see that it did 2 5 I wanted very much for my relationship to succeed, and did all I could to see that it did - 4 I wanted very much for my relationship to succeed, and did my fair share to see that it did u 3 It would have been nice if my relationship succeeded, but I couldn't do much more than I did to help it succeed - 2 It would have been nice if it succeeded, but I refused to do any more than I did to keep the relationship going a 1 My relationship could never have succeeded, and there was no more that I could have done to keep the relationship going a 0 Anxiety The following items from the STAI were scored identically: 94, 95, 98, 101, 103, 104, 108, 109, 112, and 113. Scoring: not at all . 3 somewhat = 2 moderately so . 1 very much so = 0 The following items from the STAI were scored identically: 96, 97, 99, 100, 102, 105, 106, 107, 110, and 111. Scoring: not at all s 0 somewhat s 1 moderately so = 2 very much so a 3 Regression 38. I did not feel sad . 0 I felt sad = 1 I felt sad all the time and couldn't snap out of it = 2 I was so sad or unhappy that I couldn't stand it = 3 39. 40. 41. 42. 43. 44. 45. 46. 47. 270 I was not particularly discouraged about the future = 0 I felt discouraged about the future - 1 I felt I had nothing to look forward to s 2 I felt that the future was hopeless and that things could not improve - 3 I did not feel like a failure a 0 I felt that I had failed more than the average - 1 As I looked back on my life, all I could see was a lot of failures 2 2 I felt that I was a complete failure as a person s 3 I got as much satisfaction out of things as I used to = O I didn't enjoy things the way I used to s 1 I didn't get real satisfaction out of anything any more - 2 I was dissatisfied or bored with everything - 3 didn't feel particularly guilty 3 O felt guilty a good part of the time . 1 felt quite guilty most of the time - 2 felt guilty all the time a 3 thhiH didn't feel I was being punished = O felt that I may be punished s 1 expected to be punished a 2 felt I was being punished . 3 HHHH didn't feel disappointed in myself = 0 was disappointed in myself a 1 was disgusted with myself - 2 hated myself a 3 HHHH I didn't feel I was any worse than anybody else = O I was critical of myself for my weaknesses or mistakes - 1 I blamed myself all the time for my faults = 2 I blamed myself for everything bad that happened - 3 I didn't have any thoughts of killing myself a 0 I had thoughts of killing myself, but I would not carry them out . 1 would have liked to kill myself . 2 would have killed myself if I had the chance = 3 HH I didn't cry anymore than usual 2 0 I cried more than I used to . 1 I cried all the time - 2 I used to be able to cry, but then I couldn't cry even though I wanted to - 3 48. 49. 50. 51. 52. 53. 54. 55. 56. 271 I was no more irritated than I ever am. I got annoyed or irritated more easily than I used to . 1 I felt irritated all the time = 2 I didn't get irritated at all by the things that used to irritate me a 3 I had not lost interest in other people 2 0 I was less interested in other people than I used to be - 1 I had lost most of my interest in other people . 2 I had lost all of my interest in other people - 3 made decisions about as well as I ever could a 0 put off making decisions more than I used to - 1 had greater difficulty making decisions than before a 2 couldn't make decisions at all anymore s 3 HHHH didn't feel I looked any worse than I used to a 0 was worried that I was looking old or unattractive = 1 felt that there were permanent changes in my appearance that made me look unattractive . 2 I believed that I looked ugly = 3 HHH I could work about as well as before a 0 It took extra effort to get started doing anything a 1 I had to push myself very hard to do anything a 2 I couldn't do any work at all s 3 I could sleep as well as usual s 0 I didn't sleep as well as I used to a 1 I awakened 1-2 hours earlier than usual and found it hard to get back to sleep . 2 I awakened several hours earlier than I used to and could not get back to sleep a 3 I didn't get more tired than usual - 0 I got tired more easily than I used to . 1 I got tired from doing almost anything = 2 I was too tired to do anything 2 3 My appetite was no worse than usual 2 0 My appetite was not as good as it used to be a 1 My appetite was much worse a 2 I had no appetite at all anymore . 3 hadn't lost much weight, if any lately a 0 had lost more than 5 pounds - 1 had lost more than 10 pounds . 2 had lost more than 15 pounds 2 3 was purposefully trying to lose weight by eating less s 0 (even if subject has also checked another alternative) HHHHH 272 57. I was no more worried about my health than usual - 0 I was worried about physical problems such as aches and pains; or upset stomach; or constipation = 1 I was very worried about physical problems and.it was hard to think of much else . 2 I was so worried about my physical problems, that I couldn't think about anything else - 3 58. I had not noticed any recent change in my interest in sex a O I was less interested in sex than I used to be a 1 I was much less interested in sex a 2 I had lost interest in sex completely . 3 MW Depressed Mood: was considered to have been present if a subject responded to the item "I felt sadness whenever I was reminded of what. I was losing," as having "mostly" or "accurately" described her during the early and middle stages, or during the last two weeks of her husband's illness. Crying: was considered present if a subject responded to the item "I often wept or sobbed about what I was losing," as having "mostly" or "accurately" described her during the early and middle stages, or during the last two weeks of her husband's illness. Sleep Disturbance: was considered present if a subject responded to the item "I slept as well as I did before my husband's illness, as describing her "somewhat" or as not describing" her during the early and middle stages of the illness. This symptom was also considered present if the subject failed.totendorse the statement "I could sleep as*well as usual," during the last two‘weeks of her husband's illness. Loss of Appetite: was considered present if the subject failed totendorse the statement "My appetite was nO‘worse than 273 usual," during the last two*weeks of her husband's illness (no information concerning appetite loss was available during the earlier parts of the illness). Loss of Interest: was considered present if the subject responded to the item "In comparison to what I was losing, everything else seemed trivial and meaningless," as having "mostly" or "accurately" described her during the early and middle stages, or during the last two weeks of her husband's illness. Fatigue: was considered present if the subject responded to the item "I was easily exhausted by any effort," as having "mostly" or "accurately" described her during the early and middle stages of her husband's illness. This symptom was also considered present if the subject failed to endorse the statement "I could work about as well as before," during the last two weeks of the illness. Irritability: was considered present if the subject responded to the item "Many more people irritated me than did before the illness," as having' "mostly" or "accurately" described her during the early and middle stages, or during the last two weeks of her husband's illness. Guilt: was considered present if the subject responded to the item "I felt guilty about some of the things I did or did not do before or during my husband's illness" as having "mostly" or "accurately described her during the early and middle stages, or during the last two weeks of the illness. APPENDIX D INTERNAL CONSISTENCY OF THE MARITAL WORK, GUILT, WIFE'S ACCEPTANCE, AND PHYSICAL PROBLEMS SCALES 274 275 Table D-1 _n rn -_-n of h M 1 Work . or -, ed Itam- Tatal gar rra lati an a an nd gr rgabagh' 5 Alpha (3: 47) Corrected Item # Item-total Correlation 2 .53 2a .67 3 .36 3a .66 4 .50 4a .55 5 .35 5a .51 10 .51 Cranbagh's Alpha: .82 with items 1, 8, 9, and 11 removed Table D-2 In rn 1 ion 1 -n of h il _ :l orr- -. I -m- . .1 Correlations an ng Qr 9n nbagh's Alpha (n= 47) Corrected Item # Item-total Correlation 7 .48 8 056 9 .50 10 .59 13 .37 14 .65 15 .65 grgnbagh'a Alpha: .81 with items 5, 6, 11, 12, 16, 17, and 18 removed 276 Table D-3 In r . on of ' l Itam-tptal garralati Ian 5 an 5a gr an npagh 'a Alpha (n= 47) Corrected Item # Item-total Correlation 1 .25 2 .46 3 .65 4 .28 grpnpagh'a Alpha: .63 Table D-4 In rn l on-i -n of h Ph - .J r-ol m -_r- Iggm-tgtgl Qggggl atipng an Dd Cr 91'! bag_.'_§__lp_§A h (2347) Corrected Item # Item-total Correlation 19 .63 20 .69 21 .66 22 .74 25 .61 26 .56 30 .41 46 (BDI) .52 48 (BDI) .51 49 (BDI) .64 52 (BDI) .57 96 (STAI) .52 r n ' h = .89 with items 23, 24, 27-29, 31-35, and 45 (BDI) removed LI ST QF REFERENCES 277 List of References Aldrich, C. K. Some dynamics of anticipatory grief. In B. Schoenberg, A. C. Carr, A. H. Kutschen, D. Peretz, & I. K. Goldberg (Eds. ). Aptigipatprx_§;1afi. New York: Columbia University Press, 1974. Allen, G. J. Effectiveness of study counseling and desensitization in alleviating test anxiety in college students. WW. 1971. 11. 282-289. American Psychiatric Association” .Q1ag_p§tig_apd_atati§tigal manpal pf manta; Qi sprdara (3rd ed. ). Washington, D.C.: Author, 1980. Anton, W. D. An evaluation of process and outcome variables in the systematic desensitization of test anxiety. Unpublished doctoral dissertation, University of South Florida, 1975. Auerbach, S. M. Trait-state anxiety and adjustment to surgery. rn l f n n l ni l l , 1973, 4Q, 264-271. Balkwell, C. Transition to widowhood: A review of the literature. Eamily Ralatipps, 1981, 19, 117-127. Ball, J. F. Widow's grief: The impact of age and mode of death. Qmaga, 1976-77, 1, 307-333. Barrett, C. J., & Schneweis, K. M. An empirical search for stages of widowhood. Qmaga, 1980-81, 11, 97-104. Beck. A. '1'. WW. Philadelphia: University of Pennsylvania Press, 1972. Beck, A. T. Assessment of depression: The Depression Inventory. WW. 1974. 1, 151-169. Beck, A. T., & Beamesderfer, A. Assessment of Depression: The Depression Inventory. In P. Pichot a R. Oliver-Martin (Eds. ) WM paya_pppa;paaplpgy New York: S. Karger, 1974. Beck, A. T., Ward, C. H., Mendelson, M., Mock, J. E., & Erbaugh, J. An inventory for measuring depression. Ar h f l P hi r , 1961, 4, 561-571. 278 279 Beckman, L. J., 8 Mouser, B. B. The consequences of childlessness on the social-psychological well-being of older women. lgnrsal_9f_§ergntglesx. 1982..11. 243-250. Bedell, J. The role of suggestion in the desensitization and relaxation treatments of test anxiety. Unpublished doctoral dissertation, University of South Florida, 1975. Berardo, F. M. Widowhood status in the United States: Perspective on a neglected aspect of the family life-cycle. Eamilx_£99rdinat2r. 1968. 11. 191-203. Berardo, F. M. Survivorship and social isolation: The case of the aged widower. £amily_§pp;fi1pa§p;, 1970, 12, 11-25. Blanchard, C. G., Blanchard, E. B., & Becker, J} V3 The young widow: Depressive symptomatology throughout the grief process. ggyapiatpy, 1976, 32, 394-399. Bornstein, P. 8., Clayton, P. J., Halikas, J. A., Maurice, W. L., & Robins, E. The depression of widowhood after thirteen months. Britiah.£92rnal.9f_gszghiatrx. 1973. 122, 561-566. Bowlby, J. Processes of mourning. n rn Psxshgzanalxsis. 1961. 22. 317-340- Bowlby, J. Pathological mourning and childhood mourning. - rn. - 11 _.~‘ roan-H i» 0 o ,1963, 1—1.’ 500-541. Bowlby, J. A chmen n . Vol. I: attagpmapt. New York: Basic Books, Inc., 1969. Bowlby, J. Attaghment_asd_lgss. Vol. 11: Separation. New York: Basic Books, Inc., 1973. Bowlby, J. attagpmapt_apd_lpa§. Vol. III: L953. New York: Basic Books, Inc., 1980. Caplan. G. sn229rt_s1atema_asd_29mmnnitx.mentel_healthi v 1 m . New York: Behavioral Publications, 1974. Carey, R. G. Widowed: A year later. Qpappal_pfi_gpp§pltipg zsxghglssx. 1977. 21. 125-131- Carey, R. G. Weathering widowhood: Problems and adjustment of the widowed during the first year. M, 1979-1980, 1_Q, 163-174. Clayton, P. J. The effect of living alone on bereavement symptoms. Ameri2an.isnrsal_9f_zsxshiatrx. 1975..132. 133-137. 280 Clayton, P., Desmarais, L., & Winokur, G. A study of normal bereavement. AmeIi2an.192rnal_9f_zaxshiatrx. 1968. 125. 168-178. Clayton, P. J., Halikas, J. A., & Maurice, W. L. The bereavement of the widowed. piaaaaaa_pfi_tpa_§apypaa_ fixatam, 1971, 1;, 597-604. Clayton, F. J., Halikas, J. A., 8 Maurice, W. L. The depression of widowhood" Britigh Qppgnal pf gayghiatry, 1972, 129, 71—78. Clayton, P. J., Halikas, J. A., Maurice, W. L., & Robins, E. Anticipatory grief and widowhood. Britiap_lparpal_pf Esxshiatrx. 1973. 122. 47-51. Cronbach, L. J. Coefficient alpha and the internal structure of tests. gayappmatgika, 1951, 1a, 297-334. Cumming, E., 5 Henry, W. E. r wi . New York: Basic Books, 1961. Deutsch, D. The development, reliability, and validity of an instrument designed to measure grief. Unpublished doctoral dissertation, Michigan State University, 1982. Edwards, K. R. Psychological changes associated with pregnancy and obstetric complications. Unpublished doctoral dissertation, University of Miami, Florida, 1969. Engel, G. L. Is grief a disease? Psxshgsgmatis_uedigine. 1961, 2;, 18-23. Engel, G. L. h l i l v m diaaaaa. Philadelphia: W. B. Saunders Co., 1962. Fenichel. 0. 2a22h9aaalxti2_thegrx_9f_the_nenrgai§. New York: Norton Books, 1948. Freud, S. Mourning and melancholia. 1917. mm. London: Hogarth Press, 1961. Futterman, E. H., Hoffman, I., & Sabshin, M. Parental anticipatory mourning. In B. Schoenberg, A. C. Carr, D. Peretz, & A. H. Kutscher (Eds.). BMW tarminal para. New York: Columbia University Press, 1972. Gallagher, D. E., Breckenridge, J. N., Thompson, L. W., & Peterson, J. A. Effects of bereavement on indicators of mental health in elderly widows and widowers. Qparnal pf fiergntglggx. 1983. 18. 565-571. 281 Gallagher, D. E., Thompson, L. W., 8 Peterson, J. A. Psychosocial factors affecting adaptation to bereavement in the elderlyu n rn n r l f.A in n H m Dexelgpment. 1981-82. 11. 79-95. Gerber, I. Anticipatory bereavement. In B. Schoenberg, A. C. Carr, A. H. Kutschen, D. Peretz, 8 I. K. Goldberg (Eds.). Aptigipatp;1_§riaf. New York: Columbia University Press, 1974. Gerber, I., Rusalem, R., Hannon, N., Battin, D., 8 Arkin, A. Anticipatory grief and aged widows and widowers. 12329.11. 9f_§ergntglggx. 1975. 32. 225-229. Glick, I. 0., Weiss, R. S., 8 Parkes, C. M. papaayamapt. New York: John Wiley 8 Sons, 1974. Gut, E. Some aspects of adult mourning. Qmaga, 1974, 5, 323-342. Heyman, D. R., 8 Gianturco, D. T. Long-term adaptation by the elderly to bereavement. Qpagpal_p§_§a;pptplpgy, 1973, 28, 359-362. Hinton, J. Sharing or withholding awareness of dying between husband and wife. q9urnal_Qf_2sxshgsgmati2_3eseargh. 1981, 15, 337-343. Hodges, W. F. The effects of success, threat of shock, and failure on anxiety. Unpublished doctoral dissertation, Vanderbilt University, 1967. Hodges, W. F., 8 Spielberger, C. D. The effects of threat of shock on heart rate for subjects who differ in manifest anxiety and fear of shock. zaxgppphxaiplpgy, 1966, 2, 287-294. Krant, M. J., 8 Johnston, L. Family members' perceptions of communications in late stage cancer. n I_ternatignal Q9uraal_9f_2§x9hiatrx_in_uedisine. 1977-78. 8. 203-216. Kraus, A. S., 8 Lilienfeld, A. M. Some epidemiologic aspects of the high mortality rate in the young widowed group. 19uInal_9f_§hrgsis_niseases. 1959. 19. 207-217. Krupp, G. The bereavement reaction. ‘zazghpapalytig_fita§1_p§ 529111.21: 1962: 2.: 42-74. Kubler-Ross, E. Qp_§aa§p_apd_gyiag. New York: Macmillan, 1969. Lamb, D. H. The effects of public speaking on self-report, physiological, and behavioral measures of anxiety. Unpublished doctoral dissertation, Florida State University, 1969. 282 Lindemann, E. Symptomatology and management of acute grief. i n n hi , 1944, 191, 141-148. Locke, H. J., 8 Wallace, K. M. Short marital adjustment and prediction tests: Their reliability and validity. Magpiaga and Family Laying, 1959, 21, 251-255. Lopata, H. K. Loneliness: Forms and components. fippia; Prpplamg, 1969, 11, 248-261. Lowenthal, M. F., 8 Haven, C. Interaction and adaptation: Intimacy as a critical variable. Ama;19ap_§ppiplpgigal m: 1968: 3.1: 20’31- Lundin, T. Morbidity following sudden and unexpected bereavement. Br r , 1984, 144, 84-88. (a) Lundin, T. Long-term outcome of bereavement. i rn pf Payghiatpy, 1984, 145, 424-428. (D) Lushene, R. E. The effects of physical and psychological threat on the autonomic, motoric, and ideational components of state anxiety. Unpublished doctoral dissertation, Florida State University, 1970. Madison, D. The relevance of conjugal bereavement for preventive psychiatry. British_iournal_of_nedisal Payphology. 1968. 41. 223-233. Madison, D. , 8 Walker, W. L. Factors affecting the outcome of conjugal bereavement. BriIish.iournal.of_£syshiatry. 1967, 113, 1057-1067. Marris, P. w n h r 11 . London: Routlege and Kegan Paul, 1958. Marris, P. Lp§§_apg_ppapga. Garden City, N.J.: Anchor Press, 1975. McAdoo, W. G. The effects of success and failure feedback on A-State for subjects who differ in A-Trait. Unpublished doctoral dissertation, Florida State University, 1969. Metcalfe, M., 8 Goldman, E. Validation of an inventory for measuring depression. British.iournal.of_2§yshiatry. 1965, 111, 240-242. Morgan, L. A. A re—examination of widowhood and morale. rn r n , 1976,.11, 687-695. Nussbaum, R., Wittig, B. A., Hanlon, T. E., 8 Kurland, A. A. Intravenous nialamide in the treatment of depressed female patients. gamprahanaiva gayahiatgy, 1963, 4, 105-116. 283 Parkes, C. M. Effects of bereavement on physical and mental health - A study of the medical records of widows. MW. 1964. 2. 274-279. Parkes, C. M. Bereavement and mental illness, Part I. A clinical study of the grief of bereaved psychiatric patients. ri h rn i h , 1965, 38, 1-12. (a) Parkes, C. M. Bereavement and mental illness, Part II. A classification of bereavement reactions. MW Q£_!§Q1£§l_£§Y£thQQY: 1955: 15. 13-26. (h) Parkes, C. M. The first year of bereavement. Payphiappy, 1970, 31, 444-467. Parkes, C. M. Components of the reaction to loss of a limb, spouse or home. WW. 1972. 1a, 343-349. (a) Parkes, C. M. r v m ° ' ' in London: Tavistock Publication, 1972. (b) Parkes, C. M. Determinants of outcome following bereavement. Qmaga, 1975, 6, 303-323. Parkes. c. M., a Weiss. R. s. W. New York: Basic Books, Inc., 1983. Rando, T. A. An investigation of grief and adaptation in parents when children have died from cancer. ‘Qppppal_p§ i i P h ' 1983’ a, 3-20. Rando, T. a. Wfieflb. ChMpaign. 11.: Research Press Co., 1984. Rees, W. D., 8 Lutkins, S. G. Mortality of bereavement. W. 1967. :1. 13-16. Robertson, N. C. The relationship between.marital status and the risk of psychiatric referral. B;1§;sh_fipp;pal_pf WI 1974: 1.11: 191-202- Sanders, C. M. Comparison of younger and older spouses in bereavement outcome. Qpaga, 1980-81, 11, 217-232. Sanders, C. M. Effects of sudden vs. chronic illness death on bereavement outcome. Qmaga, 1982-83, 11, 227-241. Schneider, J. Clinically significant differences between grief, pathological grief and depression. 2mm WW: 1930. 3., 151-159. 284 Schneider, J. Growth from bereavement. In P. Pegg 8 E. Metze (Eds. ) . WW. London: Pitman Books, Ltd., 1981. Schneider, J. r n r . Rockville, Md.: Aspen Systems Co., 1984. Schneider. J.. & Deutsch. D. W. Unpublished manuscr,ipt Michigan State University, 1982. Schwab, J. J., Chalmers, J. M., Conroy, S. J., Farris, P., 8 Markush, R. Studies in grief: A preliminary report. In B. Schoenberg, I Gerber, A. Wiener, A. Kutscher, D. Perets. a 15.. Carr (Eds.). W W, New York: Columbia University Press, 1975. Sharpley, C. F., 8 Cross, D. G. A psychometric evaluation of the Spanier Dyadic Adjustment Scale. W n h F il , 1982, 739-741. Singer, C. Correlates of grief in widowhood. Unpublished doctoral dissertation, university of San Diego, 1982. Smith, J. H. On the work of mourning. In B. Schoenberg, I. Gerber, A. Wiener, A. Kutscher, D. Perets, 8 A. Carr (Eds.). W. New York: Columbia University Press, 1975. Spanier, G. B. Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. l f rri n h F 11 , 1976, 15-28. Spanier, G. B., 8 Thomas, L. A confirmatory analysis of the Dyadic Adjustment Scale. rn M r i h Family, 1982, 731-738. Spielberger, C. D. Anxiety as an emotional state. In C. D. Spielberger (Ed. ) . Anaiaty; gargant tpandfi ip thappy and reggazgh (Vol. I). New York: Academic Press, 1972. Spielberger, C. D. Gorsuch, R. L. , 8 Lushene, R. E. Mapaa; h- .i L; -n or r ,- v. 1 ., W. Palo Alto, Ca. . Consulting Psychologists Press, 1970. Stein, M. The relationship of psychological and physical symptomatology and conditions relating to widowhood. Unpublished doctoral dissertation, Kent State University, 1983. Stein, 2., 8 Susser, M. W. Widowhood and mental illness. Br 1 r l f v n M , 1969, 21, 106-110. 285 Stern, R., Williams, G. M., 8 Prados, M. Grief reactions in later life. r n n 1 hi r , 1951, lgfi, 289-294. Stroebe, M. S., 8 Stroebe, W. Who suffers more? Sex differences in health risks of the widowed. Psychological_nulletin. 1983. 23. 279—301. Vachon, M. L. S., Freedman, R., Formo, A., Rogers, J., Lyall, W. A. L., 8 Freeman, S. J. J. The final illness in cancer: The widow's perspective. n l n n Medical_és§ociation. 1977. 111. 1151-1154. Vachon, M. L. S., Rogers, J., Lyall, W. A., Lancee, W. J., Sheldon, A. R., 8 Freeman, 5. J. J. Predictors and correlates of adaptation to conjugal bereavement. Am ri n rn l P hi , 1982, 132, 998-1002. Vachon, M. L. S., Sheldon, A. R., Lancee, W. J., Lyall, W. A. L., Rogers, J., 8 Freeman, 8. J. J. Correlates of enduring distress patterns following bereavement: Social network, life situation and personality. Payppplpgiaal Eedisine. 1982. 12. 783-788. Walker, K. N., MacBride, A., 8 Vachon, M. L. S. Social support networks and the crisis of bereavement. Spaia; Scien£e_asd_uedicine. 1977. 11. 35-41. Winn, R. Retrospective evaluations of marital interaction and post bereavement adjustment of widowed individuals. Unpublished doctoral dissertation, Northwestern University, 1981. Young, M., Benjamin, B., 8 Wallis, C. The mortality of widowers. Lapaat, 1963, 2, 454-456. Zisook, S., Devaul, R. A., 8 Click, M. A. Measuring symptoms of grief and bereavement. Amaripan Qpprpal pf gaygpiagpy, 1982, 132, 1590-1593. Zuckerman, M., 8 Lubin, B. l i l Af h . San Diego, Ca.: Educational and Industrial Testing Service, 1965. "7'1? “11411471541”Wills