THE FEAR OF DEATH AND RESPONSES TO OTHERS' CONCERNS ABOUT DEATH flissert‘atisn for the degree sf PhD. MECHEG N STATE UNEVERSET’I’ SUSAN MARGARETHA PARRY 1977 HEM-V / ll; if” '“ This is to certify that the fit thesis entitled The Fear of Death and Responses to Other's Concerns about Death presented by Susan Margaretha Parry has been accepted towards fulfillment of the requirements for Ph.D. Counseling degree in Major professor Date 2‘4/‘7? 0-7 639 5'9 d“, 1"! ‘ W .rnsu’ hrj~4_ - y’i‘. a; ~ ...... _ ’21 MJ ’1 & ufiafiflw i ‘l-d....' UV}, 403 A051 *wmw ABSTRACT THE FEAR OF DEATH AND RESPONSES TO OTHERS' CONCERNS ABOUT DEATH By Susan Margaretha Parry A 67-item Death Experience Questionnaire was constructed, con- taining demographic items, questions about experiences with death, and the scales of Collett and Lester's1 Fear of Death Inventory (fear of death of self, death of others, dying of self, and dying of others). Only the death of self scale achieved satisfactory reliability on this sample. Undergraduates, graduate students in education, and first and second year medical students were compared in fear of death. Scale scores were correlated with demographic variables. Medical students, who were predicted to fear the death and dying of others more than the other groups, scored lower in fear of dying of self and others. There were no differences in fear of death of self or others. Females scored higher than males in fear of death of others and dying of self. Less religious people showed more fear of death and dy- ing of self than the more religious. Belief in afterlife was positively related to fear of death of others and negatively to fear of death of self. Medical students who had suffered a recent loss feared death of self more than those who had not. There was a positive correlation be- tween direct admission of fear of death and scores on the fear of death of self and others scale. Fifty-six medical students who had completed the DEQ responded to six taped simulated dying patients, imagining themselves as the patients' Parry physician. Those who scored high in fear of death and dying of others were expected to show a greater response lag and more emotional responses, to report more anxiety while responding, and to make less direct references to death than those with less fear. These hypotheses were generally not supported. No significant correlations were found between fear of death and response lag or directness of death response. Emotionality was sig- nificantly correlated only with fear of dying of others. There was a negative correlation between response lag and both emotionality of re- sponse and directness of death reference. Subjects who scored high in fear of dying of self gave lower anxiety ratings to the episodes than those with lower fear. All subjects said they would want to know if they themselves were dying, so that they could change their lifestyles, put their affairs in order, or feel in control. Most indicated that some patients should not be told they were dying, usually for psychological reasons. The majority believed that the physician, alone or in consultation with others, should both decide whether to tell patients of their impending death and break the news to them. These results must be interpreted with caution; the development of fear of death scales of greater reliability is needed. Assumptions about the influence of physicians' fears of death on their ability to help dying patients need to be re-examined. Physicians were expected to differ from other groups particularly in fear of death and dying of others, but it appears that it is in their attitudes about dying, rather than about death, that differences are likely to occur. 1Lora-Jean Collett and David Lester, "The fear of death and the fear of dying," The Journal of Psychology 72, l969, 179-18l. THE FEAR OF DEATH AND RESPONSES TO OTHERS' CONCERNS ABOUT DEATH BY SUSAN MARGARETHA PARRY A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services, and Educational Psychology 1977 (9 Copyright by Susan Margaretha Parry l977 To Doug and Imogen ACKNOWLEDGMENTS I would like to thank my chairman, Dr. William W. Farquhar, for his assistance and encouragement in the completion of this research, and for his willingness to be available when needed. His constructive criticism and his help in developing the ideas behind the research have made it a much more pleasant and valuable task than it might otherwise have been. Dr. Sam Plyler was of great assistance during the planning stage of the dissertation. I particularly value discussions with him which aided in formulating and clarifying the concepts behind the research. Dr. John Schneider and Dr. Robert Zucker were of assistance with questions of design and methodology. Dr. Robert Wilson gave valuable and expert statistical and computer ad- vice, and taught me a great deal in the process. Dr. Douglas Miller was an extremely valuable consultant in planning the design of the re- search; I particularly appreciate his critical reading of the proposal and the many valuable suggestions he made at that time, as well as his help with statistical questions which arose. Other individuals gave generously of their time to assist in this re- search. Diane Plyler, Jean Schwartz, Peg Geggie, and Mary Nowak spent many hours interviewing subjects. Jean Schwartz, Carol Weinberg, Dr. David Inman, Nancy Stockton, Sharon Robinson, and James Novosel all assisted in rating of tapes and collation of the data. I would like to thank the subjects who participated in the research, the instructors who volunteered their classes as potential subjects, and the large number of individuals who were willing to spend time in discussion with me during the course of the research. Finally, I must acknowledge the support, love and confidence of the Newton Street Ladies Gang, Jean Schwartz and Joan Stieber, who lived with me through this whole project. TABLE OF CONTENTS Page Chapter I Introduction ............................................... 1 Need for the Study ......................................... 2 Purpose of the Study ....................................... 4 Research Hypotheses ........................................ 7 Theory ..................................................... 8 Chapter 11 Introduction ............................................... 14 Sex Differences ............................................ 15 Preoccupation with death .............................. 15 Content of thoughts about death ....................... l7 Fear of death ......................................... 20 Religious Differences ...................................... 23 Denominational preferences ............................ 23 Religiosity ........................................... 28 Belief in afterlife ................................... 33 Age Differences ............................................ 35 Health and Personality ..................................... 38 Mental and physical health ............................ 38 Learning of death anxiety ............................. 47 Other Demographic Variables ................................ 56 Race .................................................. 56 Occupation ............................................ 56 Living situation ...................................... 58 Miscellaneous ......................................... 59 Attitudes of Medical Personnel ............................. 61 Fear of Death Measures ..................................... 68 Fear of death inventories ............................. 72 Collett-Lester scale .................................. 77 Chapter III Collection of Questionnaire Data ........................... 84 The Samples ................................................ 85 Undergraduates ........................................ 85 Graduate students ..................................... 92 Medical students ...................................... 95 Experimental sample ................................... 97 Construction of the Death Experience Questionnaire ......... 102 Demographic items ..................................... 102 Fear of death items ................................... 102 Reliability of the Fear of Death Scales .................... 103 Factor Analysis of Scale Items ............................. l08 Construction and Format of Stimulus Tapes .................. lll Conduct of the Experiment and Interview .................... llB iv Page Hypotheses ................................................. 116 Survey hypotheses ..................................... 116 Experimental hypotheses ............................... 117 Dependent Measures ......................................... 118 Response lag .......................................... 118 Directness of death reference ......................... 120 Emotionality .......................................... 123 Reliability of dependent measures ..................... 125 Design and Analysis ........................................ 125 Summary .................................................... 126 Chapter IV Introduction ............................................... 128 Death Experience Questionnaire Results ..................... 129 Sample differences in fear of death ................... 129 Relationships between fear of death and other variables 134 Undergraduates ................................... 137 Graduate students ................................ 138 Medical students ................................. 139 Experimental subjects ............................ 141 Testing of survey hypotheses .......................... 142 Anxiety Ratings of Vignettes ............................... 145 Correlations between anxiety and fear of death ........ 148 Correlations between anxiety and demographic and experience variables ..................... 149 Discussion ............................................ 150 Reasons for highest and lowest anxiety ratings ........ 151 Dependent Measures ......................................... 157 Intercorrelations among dependent measures ............ 157 Correlation of fear of death and dependent measures ... 158 Correlations between demographic and experience variables and dependent measures ......... 159 Testing of experimental hypotheses .................... 164 Interview Data ............................................. 166 Knowledge of own impending death ...................... 166 Informing patients of impending death ................. 168 Summary .................................................... 174 Chapter V Overview of the Study ...................................... 177 Conclusions ................................................ 179 Discussion ................................................. 181 Reliability of Collett-Lester scales .................. 182 Fear of death and responses to simulated patients ..... 184 Differences among groups .............................. 190 Suggestions for Future Research ............................ 190 Use of format in training ............................. 191 Conclusion ................................................. 192 Table .b 05 «ooowmmbwm abnnawwwwwwwwwww —l—-l-—l—l Ndo 01-th LIST OF TABLES Intercorrelations between Collett-Lester Subscales Marital Status Academic Major Religion of Origin Current Religious Preference Religious Self-Rating Mother's Employment Father's Employment Fear of Death Intercollections between Collett-Lester Subscales Subscale Reliabilities in Three Different Populations References to Death Reliability of Dependent Measures Analyses of Variance, Fear of Death Scales by Samples Contact with the Dying Presence at Death of Others Interscale Correlations (Experimental Sample) Intercorrelations of Fear of Death and Demographic Variables (Whole Sample) Collett-Lester Subscale Scores and Admitted Fear of Death Intercorretations between Fear of Death Scales and Demo- graphic Variables (Undergraduate Sample) Intercorrelations among Fear of Death Scales and Demo- graphic Variables (Graduate Student Sample) Intercorrelations of Fear of Death Scales and Demographic Variables (Medical Student Group) vi Table 4.10 4.11 4.14 4.15 4.17 4.18 Intercorrelations between Subscale Scores and Demo- graphic Variables (Experimental Subjects) Anxiety Rating Means and Standard Deviations Number of Subjects Rating Each Vignette Highest and Lowest in Anxiety Intercorrelations among Dependent Measures Intercorrelations of Scale Scores and Dependent Measures Demographic Variables and Response Lag, Death Directness, Emotionality, and Anxiety Significant Correlations between Demographic Items and Dependent Variables Reasons for Wanting to Know of Own Impending Death Reasons for Not Telling Patients of Their Impending Death Who Should Decide Whether a Patient is to be Told of Impending Death? Who Should Be Responsible for Telling the Patient of Impending Death? vii CHAPTER I With social and technological changes during this century have come changes in what individuals can expect to face during the time they are dying and in the problems which arise for those who come into contact with dying patients. People no longer usually die at home with family, but in the hospital, in which their primary contacts are with medical personnel. Advances in medical technology have made it possible for dying to be prolonged, and death may even be post- poned indefinitely at the will of family, doctor, or judge, as the Quinlan case illustrated.1 Decisions regarding the management of the patient are made by the physician rather than by the family. Perhaps the major change is that dying, once a private affair, is now a matter handled almost entirely by professionals. While physicians have always attended those among the dying who could afford their services, not until recently has attendance upon the dying and regulation of the last days of life fallen so entirely within the sphere of professional, rather than private, relationships. This change has created stresses in institutions which deal with the sick (both curable and incurable) and in the legal profession, and new demands upon individual physicians and other medical personnel, and upon members of the dying patient's family, who often are able to exercise little or no control over what happens to that patient. As a result, research 1 , "Ideas and Trends," New York Times, November 2, 1975. 1 interest in the treatment of the dying, in the kinds of problems faced by their families during the period of bereavement, and in the psycho- logical processes of dying patients has grown considerably in the past ten or fifteen years. Attention is also beginning to be paid to the attitudes and feelings of those who work directly with the dying in a professional capacity: medical personnel, mental health professionals, clergy, and students of any of these fields. Some of the recent 2’ 3’ 4 focusing on work in this area has been of a sociological nature, institutions which deal with the dying, and other studies have examined the feelings and attitudes of individual professionals. The latter focus is taken in this study. Need For the Study It is particularly important that the attitudes of medical professionals be understood. Responsibility for dealing directly with the dying patient is incurred or assumed by the physician most directly responsible for his or her care, and by the nurses who have daily patient contact. In addition, the physician often must respond to the needs of the patient's family and the rest of the medical staff, all of whom are under stress when a person is dying. (The attitudes and behavior of the individual physician influence also the way in which 2Arnold Toynbee, et al., Man's Concern With Death, London, Hodder and Stoughton, 1968. 3William A. Faunce and Robert L. Fulton, "The sociology of death: a neglected area of research," Social Forces, 1958, 36-205-09. 4 Renee Fox, Experiment Perilous (Glencoe, 111.: Free Press, 1959). 3 the institution within which s/he works responds to the dying patient, but such considerations are beyond the scope of this study.) Cramond5 points out that the patient is an important object for the physician, and that interaction with him or her may tap into both consciously held attitudes and unconscious processes of the physician. He argues that physicians in training, as well as other medical personnel, need opportunities to explore their feelings so that they do not hinder their future interactions with their patients. Bouton6 concludes, from informal discussions with medical students, that some of the concerns they themselves feel a need to discuss include: personal experiences of death in the past; viewing the dead body as both person and thing, and the feelings that arouses; and what to tell the patient who is dying. It is assumed fer purposes of this research that interaction with the dying patient is always, to some degree, stressful for the physician involved. Medical students may receive better training in coping with this stress and in being helpful to such patients if it is known what attitudes and feelings of theirs influence how they interact with those patients. While there exists a substantial body of research on correlates of attitudes toward death, little of it is focused directly on characteristics of physician or medical students. 5W. A. Cramond, "Psychotherapy of the dying patient," British Medical Journal, 3 (5719), 1970, 389-393. 6David Bouton, "The need for including instruction on death and dying in the medical curriculum," Journal of Medical Education 47 (3), 1972, 169-175. Purpose of the Study There are two main purposes of this study: 1. To cross-validate a widely-used fear of death inventory, that of Collett and Lester,7' on a large group of undergraduates, graduate students, and medical students. This self-report question- naire contains four separate subscales, measuring the fear of death of self, dying of self, death of others, and dying of others. It is intended to learn whether these scales are well constructed and reliable, and whether there are differences among the three subject populations in scores, and to discover whether demographic and personal experience variables have any relationship to these various kinds of fear of death. This study will be fbcused on attitudes towards the deaths of others as well as on feelings about one's own death, since its main concern is with how physicians respond to the concerns of others. 2. The second purpose is to examine more closely the attitudes and feelings of medical students in the preclinical years of training, most of whom have had little or no contact with patients. The focus is on consciously held attitudes, those which may be most directly affected by training programs administered in an academic setting. Relationships between scores on the fear of death scales and responses to demographic items, and responses to a series of simulated patients are examined. 7Lora-Jean Collett and David Lester, "The fear of death and the fear of dying," The Journal of Psychology 72, 1969, 179-181. Two main questions are addressed in this study: whether medical students are different in their attitudes than their peers in other areas of study, and whether there are identifiable attitudes or feelings about death which influence how they would respond to a dying patient. While it is not the intent of this study to investi- gate unconsciously held feelings, it is still insufficient to simply ask subjects how they think they would respond to dying patients. Therefore, subjects are asked to respond directly to a taped simula- tion of a dying patient, imagining that he is a real patient. Since the focus of medical education is on working with patients, subjects are asked to respond to simulated patients, in an initial attempt to investigate those factors which may influence responses they might make to real patients. Hopefully, the relationship between pre- viously held attitudes and the way a student may be expected to function in a clinical setting may begin to be clarified. Feifel et a1.8 feund that medical students were intermediate in their attitudes towards death, between the general public and doctors who have already completed medical school. The attitudes of medical students approached those of doctors more closely the nearer the student was to the end of his or her training, suggesting that a process or socialization occurs. Thus, changes in a student's attitudes may not be explainable totally in terms of his or her experiences with the dying. While it is outside the scope of this study to explain changes in attitudes toward death, it will be assumed that experience with the 8Herman Feifel, et a1, "Physicians consider death," Proceedings, 75th Annual Convention of the American Psychological Association, 1967, 201-202. 6 dying, and with others in the medical field, contributes to the forma- tion of a student's attitudes. Thus, it was decided to limit the population studied to students in the early years of their training, when exposure to both patients and other medical professionals is more limited than in succeeding years. Knowledge of students' attitudes at the beginning of their medical school career may or may not enable one to predict the attitudes they will have once they are actually practicing medicine. In the past few years, particularly with the advent of affirma- tive action programs, medical schools have admitted more women, older students, and members of minority groups. Although there is conflicting evidence on the influence of demographic variables on attitudes toward death, the question must be raised whether these changes in the composi- tion of a student's peer group will have an effect on the socialization process and on the attitudes students will express by the end of their training. In addition, it may be the case that different kinds of students are being attracted to medical schools. In addition, there have been social changes in the past 15 years (during which time most of the research on death attitudes has been done) which make it necessary to be cautious in making comparisons between past and present studies. Some of these changes include: new attention given by both scholars and the media to issues of death and dying, liberation movements, leading to new systems of power relations among various groups (women, blacks, etc.), new emphasis on training in human relations for professionals. Much previous research in the area of attitudes of medical personnel has asked specifically about the doctor's perception of his or her role and relationships with patients. It appears that it is also 7 useful to ask the same questions of medical people which are asked of others. This procedure will not only facilitate comparison with other groups, but will also tend to elicit attitudes which are not expressed in medical language, and which may therefore be less contaminated by medical training than would responses to questions about how they see themselves functioning as doctors. Therefore, in the questionnaire portion of the study, the same questions about the fear of death are asked of all groups, while in the interview, the medical students are asked specifically about their attitudes, perceptions, and opinions which fecus on themselves as physicians. Research Hypotheses 1. Previous research on the Collett-Lester inventory has shown that people differ, not only in total fear of death, but in their fears of different deaths; thus, they achieve different scores on the subscales of the inventory. This result will be replicated here, and the fear of death will be shown not to be a unitary phenomenon, but to have separate, although related, components. 2. There will be differences in fear of death based on sexual and religious characteristics. Females will score higher than males. Highly religious subjects will score lower in fear of death of self. 3. Medical students will score higher on all subscales than either undergraduate or other graduate students, and particularly on fear of death and dying of others. While there may be other differences across or between groups, previous research is too conflicting in results to allow for specific predictions of the nature and direction of differences. 8 4. There will be a negative relationship between the medical student's fear of death (as expressed on the Collett-Lester inventory particularly on the fear of death and dying of others scales) and the quality of his or her responses to simulated dying patients. Those who score high on those scales will take a longer time before responding to the patient, and will express the most anxiety over having to do so. They will respond with less direct references, or no references, to death. They will give emotionally laden responses. 5. Those patients which are rated as arousing more anxiety will be responded to with fewer direct death references and with a longer time lag before response by all subjects, regardless of their reported fear of death. Theory The fear of death, like other fears, results from the person's experiences in life, and his or her interactions with significant others in the environment. There is not, however, enough evidence to allow one to specify exactly what features of a person's experience contribute to a high or low fear of death; what evidence exists is reviewed in Chapter 11. However, at the most general level, there are several con- ditions which jinfluence the fear of death. The physical environment, to the extent that it is experienced as dangerous to the well-being, or even the survival, of the organism 9 10 must play a role, ’ if only to determine the amount of attention the 9Gene Lester and David Lester, "The fear of death, the fear of dying, and threshold differences for death words and neutral words," Omega 1, 1970, 175-179. 10John Bowlby, Attachment and Loss, Vol. II, Separation, London, Hogarth Press, 1973. 9 threat of death must be given on a daily basis. However: . knowledge of the 'external' degree of threat along seems to be an insufficient basis on which to predict with any certainty how a person will react to it. The person's character structure--the type of person he is-- may sometimes be more important than the death-threat stimulus itself in determining reactions. 12, 13 There are those, of course, who argue that it is features of the psychological make-up common to everyone which accounts for the existence of the fear of death. Until recently, the psychoanalytic coneptualization of attitudes toward and fear of death as derivative events was dominant. Sentiments about death were essentially manifestations of a more ultimate reality, that is, separation anxiety or conflicts about castra- tion. Doubtless, such clinical displacement does occur. Death fears can be secondary phenomena. Nevertheless, incoming data increasingly suggests that the reverse may be more to the point. Apprehensiveness over bodily annihilation and concerns about finitude themselves assume dissembling guises. The depressed mood, fears of loss, sundry psychosomatic symptoms, and varying psychological disturbances all have evidenced afinity to anxieties concerning death. The fear of death thus need not be seen as a derivative phenomenon, to be explained in terms of more basic psychological functioning, but may fruitfully be regarded as existing in its own right. 1lHerman Feifel, "Attitudes toward death in some normal and mentally ill populations," in H. Feifel, ed., The Meanigg of Death, (New York: McGraw-Hill, 1959), 114-130. 12Sigmund Freud, Beyond the Pleasure Principle, (New York: Liveright, 1961), (Tr. by James Strackey). 13Melanie Klein, "A contribution to the theory of anxiety and guilt," International Journal of Psychoanalysis, 29, 1948, 114-123. 14Herman Feifel, "Attitudes toward death: A psychological perspective," Journal of Consultinggand Clinical ngchology 33 (3), 1969, 292-295. 12‘) 1(1). 10 While there are psychological factors that undoubtedly contribute to high or low fear of death, it appears more fruitful to focus on those characteristics of the individual which are more clearly measurable, and more clearly may be tied to specific aspects of his or her upbringing, such as religion, the presence or absence of specific others in the family, socioeconomic status. To do so will allow prediction of fear based on general kinds of experiences available within the society, rather than necessitating fecusing on individual intrapsychic processes. Finally, there may be specific kinds of life experiences, occurring at specific times in the life of the individual, which are influential in producing fear of death, and in producing different kinds of fear in different individuals. For instance, the individual who has had a close escape from death should be expected to attend to thoughts and feelings about his or her own death, while one whose most impactful experiences have been of the loss of highly significant others (particularly ones which whom s/he had strong but ambivalent relationships) may be expected to show more affect regarding the death of others. Similarly, there should be identifiable experiences in the life of the individual which contribute to a focus of fear either on being dead, or on the process of dying. The cultural milieu in which a person grows up, to the extent that it determines the kind of experiences available to the individual, 15 must also have an effect. It should make a difference, for instance, what role the dead are expected to play in the life of the society, 15David Lester, "Antecedents of the fear of the dead," E§ychological Reports 19, 1966, 741-742. 11 what is expected of individuals and families which have experienced a loss, and how much actual contact with death one normally experiences. In addition to general cultural factors, the more intimate social environment of the individual (family relationships, the presence or absence of certain significant others, the experiences of being sheltered or abandoned, the kind of interaction experienced with significant others, and the extent and timing of the person's exposure to the deaths of others)should all be of relevance. The assumption is made in this study that some such factors contribute to making the individual afraid of death to a greater or lesser degree. If attitudes toward death and dying are learned through a combination of various kinds of experiences, they should influence the behavior of the individual faced with a situation in which death may occur. One's attitudes should also influence the readiness of the individual to place him or herself in situations in which there is the possibility of his or her own death or that of others. It is possible that a high fear of death would incline a person to avoid such situa- tions altogether. 0n the other hand, a counter-phobic response may be elicited, in which the individual seeks out such situations or environ- ments so as to conquer that which s/he fears. One such situation, which involves the constant possibility of the death of others, is the practice of medicine. Weisman . . . suggests that since most medical personnel have not truly accepted their own mortality, they have difficulty facing rationally the death and dying of others, which remind them not only of their own mor- tality, but also of their powerlessness in saving lives. Weisman states further that because of the conflicts and pressures of caring for critically ill patients, medical staff members deny the realities and problems of dying and encourage patients to deny the gravity of 12 their situation and to behave so as not to disrupt the staff's denial.16 If this is true, then medical personnel may be expected to be more afraid of death than people in other professions, but, because of their own anxiety, to interact with their patients in such a way as to inhibit the patient from facing the physician with the reality of the patient's impending death. Livingston and Zimet17 portray the physician as experiencing each patient's death as a personal failure; it may be that physicians become more afraid of death as they ex- perience the deaths of more patients, and that the fear of death is, in that case, a reflection of a deeper fear of failure. However, if the physician's entry into the practice of medicine was already partially determined by a high fear of death, the experience of more deaths of patients could serve to decondition the physician somewhat to that fear. For this reason, this study is focused upon students, who have made the choice of a career in medicine, but have not yet experienced contact with patients which could, during their profes- sional life, influence their fear of death. Medical students, if they experience a higher fear of death and students in other fields, may be expected to respond to simulated patients with some of the same characteristic responses as physicians would make to real patients. They may respond in such a way as to l6Lisa R. Shusterman, and Lee Sechrest, "Attitudes of regis- tered nurses toward death in a general hospital," Psychiatry in Medicine 4 (4), 1973, 411-425. 17Peter B. Livingston, and Carl N. Zimet, "Death anxiety, authoritarianism, and choice of specialty in medical students," Journal of Nervous and Mental Diseases 140 (3), 1965, 222-230. 13 attempt to stop them from talking about death; this may be accomplished in many cases, simply by not mentioning the topic, especially in response to a patient's explicit reference to it. The impression of unwillingness to discuss the topic may also be given by taking a longer time to respond to the patient, by appearing hesitant. It is not necessary that these behaviors be conscious or intentional, only that, when they occur, the impression be given that the patient's impending death is not an acceptable topic of discussion. There should, therefore, even in medical students who have not yet had contact with actual patients, be an association between their expressed fear of death and the amount of time it takes them to respond to simulated patients, and a negative association between fear of death and the direct mention of death in response to the patient's presented concern about it. In addition, if the respondent is experiencing anxiety during the interaction with the supposed dying patient, the amount of emotion, either expressed or denied, in what s/he says, should be higher than when there is no felt anxiety. CHAPTER II In the literature on correlates of attitudes toward death, results reported by one researcher are often not duplicated by others. This fact motivates the use of an extensive questionnaire in the current study. It also accounts for the general lack of clear directional hypotheses about the nature of relationships to be feund between demographic variables and fear of death. On the general theory that attitudes toward death are learned, the task becomes to specify how they are learned, and this may vary a great deal from one individual to another. It will be clear from this review that sex, religion, and the attitudes of parents are likely to be related to attitudes toward death; other relationships are even less clear. This chapter is organized into sections describing separately the major demographic and experiential variables which have been in- vestigated for relevance to attitudes toward death, the attitudes of physicians and other medical personnel, and the instruments which have been most commonly used in such research. Many studies investigate only one or two variables in death attitudes, or fecus on limited populations, or investigate variables which stand in no clear theoretical relationship to each other. Chasin1 argues that single-variable approaches to the understanding of 1Barbara Chasin, "Neglected variables in the study of death attitudes," Sociological Quarterly_12, 1971, 107-113. 14 15 death attitudes leave out too much potential infermation, and that it is the interaction among variables which must be understood. This principle has guided the design of the current research. Caution must be used in interpreting reports of difference or lack of difference based on investigation of a single variable. Evidence bearing on the relevance of each particular variable to attitudes toward death will be summarized separately. The main ones to be considered are sex, religious preference, religiosity, age, and resemblance to parental attitudes. Sex Differences Several kinds of attitudes towards death are considered in the literature, including amount of thought about death (preoccupation with death), the nature of the thoughts expressed, and the amount of fear attached to those thoughts. In regard to sex differences, each of these is considered separately. Some differences may be related to sex roles and gender identity rather than to biological gender. 2 Preoccupation with death. Feifel, cautioning that frequency of thought about death has no necessary relationship to the fear of death, reported that women, in both normal and mentally ill populations, admitted thinking more frequently about death than did men. Lester,3’4 2Herman Feifel, "Attitudes toward death in some normal and mentally ill populations," in Herman Feifel, ed., The Meaning of Death (New York: McGraw-Hill, 1959), 114-130. 3David Lester, "Re-examination of Middleton's data: Sex differences in death attitudes," Psychological Reports 27, 1970, 136. 4David Lester, "Sex differences in attitudes toward death: A replication," _§ychologjcal Reports 28, 1971, 754. 16 on the other hand, found that male college students thought about their own death and pictured themselves as dead or dying more often than females. Selvey5 found no sex differences in reported preoccupation with death, although in her Study, scores on Dickstein and Blatt's preoccupation with death scale were significantly correlated with scores on Boyar's Fear of Death Scale in both sexes. This is a particularly clear example of the conflicting evidence on sex differences in death attitudes. Durlak6 found no relationship between fear of death and subjects' estimates of the frequency with which they thought about their own death, nor with whether the individual had thought about his or her own death within the last two days. Cameron7 found that those who had thought of their own death within the last two days gave a significantly larger estimate of the probability Of their own death within the next year than did those who had not thought about it. 8 feund that males see themselves individually living Tolor and Murphy longer than they think the average man will live; this did not happen among females. 5Carole L. Selvey, "Concerns about death in relation to sex, dependency, guilt about hostility, and feelings of powerlessness," Omega 4 (3), 209- 219. 6Joseph A. Durlak, "Relationship between various measures of death concern and fear of death," Journal of Consulting_and Clinical Psychology 41 (l), 1973, 162. 7Paul Cameron, "The imminency of death,“ Journal of Consulting and Clinical Psychology 32 (4), 1968, 479-481. 8Alexander Tolor and Vincent M. Murphy, "Some psychological correlates of subjective life expectancy," Journal of Clinical Psychology 23, 1967, 21-26. 17 Content of thoughts about death. Lester9 reported that men were less likely than women to say they were depressed by funerals, cemeteries, and stories about death, and also10 that they less often reported ever having wished that they were dead. In both studies, he found men more often wanting to know for sure whether there was a life 1] that men more after death, but less likely to believe in one, and often said they would change their manner of living if they knew fer sure that there was a life after death. On a variety of other questions about their thoughts about death, there were no sex differences. Lester's conclusion from these studies is that males think more about death, but are less likely to have or express a negative affective reaction to it. Several studies of the content of thoughts about death have yielded results which might reflect differences associated with learned sex roles. Diggory and Rothman hypothesize that: To the extent that the goals a person values highly depend on his social status, his fear of various consequences of his own death should vary with his status or role, whether defined by age, sex, social class, religion, or marital condition. They asked a varied group of subjects which of several consequences of their own death they feared most. Women, significantly more than men, feared both what would happen to their bodies after death and the possible pain involved in dying. Women's higher level of concern fer 9David Lester, 1970, op. cit. 1O 11 12James C. Diggory and Dorreen Z. Rothman, "Values destroyed by death," Journal of Abnormal and Social Psychology 63 (l), 1961, 205. David Lester, 1971, op. cit. David Lester, 1970, op. cit. 18 the fate of their bodies may reflect the extent to which the condition of their bodies is seen as being of importance in the perfor- mance of their traditional sex role. Men were concerned significantly more than women that they could no longer care for their dependents once they were dead, possibly a reflection of the male's traditional role as having dependents, contrasted with the woman's role as being dependent. Similarly, Schneidman13 found that undergraduate (mostly single) women expressed a preference to die with or before their putative spouse significantly more often than men, who preferred to die after their spouse. He suggests that women wished to avoid grief and loneliness and men to spare their partner that suffering, and that this finding may be the result of social expectations that the man will be the caretaker and the woman the recipient of caretaking. ‘4 ‘5 asked men and women to construct stories Lowry and Selvey about death to TAT cards, including one specially constructed for the study. In men's stories, the person who died was usually a male, and the theme was one of violence and mutilation. In women's stories, usually a male died also, but the theme was more often one of loss. These results may reflect another aspect of the same phenomenon: that males' thoughts about death tended to be self-oriented, active, and independent, or to reflect the male's assumption of responsibility for 13Edwin S. Schneidman, "0n the deromanticization of death," American Journal of Psychotherapy_25 (1), 1971, 4-17. 14Richard J. Lowry, Male-Female Differences in Attitudes Toward Death, Ph.D. Dissertation, Brandeis University, 19652 15 Carole L. Selvey, op. cit. 19 providing for his dependents, while women think of losing their source of support and are concerned with the condition of that which, at least in part, gains them that support--their bodies. 16 found, however, that women gave practical, Kahana and Kahana interpersonal, and caretaking (other-directed) reasons fer wanting to know ahead of time if they were going to die, while men's reasons for wanting that information had to do with mastery, facing reality, and personal need fin‘awareness. Here, it is the caretaking role of the woman, rather than of the man, which is reflected in attitudes towards death. It may be that this reflects the traditional expectation that the woman will think of the needs of others before her own, while men might think first of their own abilities and capacities, which only secondarily will be used for meeting the needs of others. It may also be that the kinds of stimuli given in the two experiments elicited different facets of attitudes towards death, related to different per- ceptions of sex roles. In research which relies entirely on self-report, differences may appear which result from certain attitudes or feelings being more acceptably expressed by one sex than the other. Thus, the expression of thoughts about death related to one's role as a caretaker appears to be equally socially desirable for both men and women, although it may take different forms. But a fear of loss, for instance, may be thought 'more acceptable when expressed by a woman, and less so when a man feels it. 16Boaz Kahana and Eva Kahana, "Attitudes of young men and women toward awareness of death," Omega 3, 1972, 37-44. 20 Fear of death. Several researchers have found sex differences in amount and content of fear of death, using various measures developed for the purpose of measuring fear of death. 17, 18 Templer and Ruff feund that females scored significantly 19 higher than males on Templer's Death Anxiety Scale, in populations of adolescents, their parents, and a group of residents of an upper- middle-class apartment house. In groups of psychiatric patients and low-income psychiatric aides, the difference was in the same direction, but was not significant. (They do not address the question of whether sex difference are influenced by differences in socioeconomic status.) 20 Ray and Najman found a slight but significant tendency fer female undergraduates to be more death anxious than males, using the same instrument, but the sexes did not differ on a measure of death accep- tance which was negatively correlated with the Death Anxiety Scale. 21 Selvey reported that women scored significantly higher than 22 men on Boyar's Fear of Death Scale. In males, fear of death was 17Donald I. Templer and Carol F. Ruff, "Death anxiety: age, sex, and parental resemblance in diverse populations," Developmental Psychology 4 (l), 1971, 108. 18Donald I. and Carol F. Ruff, "Death Anxiety Scale means, standard deviations, and embedding," Psychological Reports 29 (l), 1971, 173. 19Donald 1. Templer, The Construction and Validation of a Death Anxiety Scale, Ph.D. Dissertation, UniVersity of Kentucky,’1967. 20J. J. Ray and J. Najman, "Death anxiety and death acceptance: A preliminary approach," Omega 5 (4), 1974, 311-315. 21 Carole L. Selvey, op. cit. 22Jerome I. Boyar, The Construction and Partial Validation of a Scale for the Measurement of the Fear of Death, Ph.D.—Dissertation, University of Rochester, 1963. 21 correlated with scores on the Fordyce Dependency Scale, and preoccupa- tion with death with guilt about hostility on the Mosher Guilt Scale, but there were no such correlations among females. Lester,23 using the scale of Collett and Lester (which is used in this study also), found that women had a higher fear of dying of self, death of self, and death of others, but there was no difference between the sexes in fear of dying of others, consistency of death attitudes, or semantic differential rating of the concept of death. (Evaluative ratings were significantly related to the fear of death.) 24 Handal found that females with an unrealistically low sub- jective life expectancy (the age they expected to live to, compared with the actuarial life expectancy for their age and sex) had signi- ficantly higher death anxiety scores on a revision of Livingston and 25 Zimet's scale than did those with realistic or unrealistically high subjective life expectancy. There were no significant differences among males in this respect. 26 Feldman and Hersen found that women expressed more conscious death concern than men. 23David Lester, "Studies in death attitudes: Part two," Psychological Reports 30, 1972, 440. 24Paul J. Handal, "The relationship between subjective life expectancy, death anxiety, and general anxiety," Journal of Clinical Peychology 25 (l), 1969, 39-42. 25Peter B. Livingston and Carl N. Zimet, "Death anxiety, authori- tarianism, and choice of Specialty in medical students," Journal of Nervous and Mental Disease 140 (3), 1965, 222-230. 26Marvin J. Feldman and Michel Hersen, "Attitudes toward death in nightmare subjects," Journal of Abnormal Psychology_72 (5), 1967, 421-425. 22 In several studies, no sex differences have been shown. None 27 were found by Pandey and Templer in a group of college students, Lucas28 in a sample of surgical and dialysis patients and their wives, Templer and Dotson29 in a study of people of different religious persuasions, or by Feifel and Branscomb30 in a group of both healthy 31 and ill people. Rhudick and Dibner found no sex differences among the elderly in the number of references to death given in stories told to TAT cards (taken as a measure of unconscious death concern). Swenson32 feund no differences in a sample of normal aged people, nor did Reynolds and Kalish33 34 in a group of young, middle-aged and elderly people. Dickstein fbund no significant differences in responses to his 27R. E. Pandey and Donald 1. Templer, "Use of the Death Anxiety Scale in an inter-racial setting," mega 3 (2), 1972, 127-130. 28Richard A. Lucas, "A comparative study of measures of general anxiety and death anxiety among three medical groups including patient and wife," Omega 5 (3), 1974, 233-243. 29Donald I. Templer and Elsie Dotson, "Religious correlates and death anxiety, Psychological Reports 26, 1970, 895-897. 30Herman Feifel and Allan B. Branscomb, "Who's afraid of death?" Journal of Abnormal Psychology 81 (3), 1973, 282-288. 31Paul J. Rhudick and Andrew S. Dibner, "Age personality, and health correlates of death concerns in normal aged individuals," Journal of Gerontology 16 (l), 1961, 44-49. 32Wendell M. Swenson, A Study of Death Attitudes in the Gerontic Population and Their Relationship to Certain Measurable—Physicalfand' Social CharacteristTEs, Pth. Dissertation,_Ufiiversity 6fiMinnesota, 1958. 33David K. Reynolds and Richard A. Kalish, "Anticipation of futurity as a function of ethnicity and age," Journal of Gerontology 29 (2), 1974, 224-231. 34L0U15 5- Dickstein, "Death concern: Measurement and correlates;' Peychologjpal Reports 20, 1972, 563-571. 23 death concern scale, and found differing correlations between death concern and personality variables among males and females. Hooper and Spil ka35 found no relationship between sex and positive or negative perception of death in a group of college students. In conclusion, it appears that sex differences may be related to other variables, which may explain the fact that some researchers have found differences while others have not. In those studies in which sex differences have been found, in all cases the difference has been in the direction of women being more afraid of death than men, a finding which remains unexplained. In the current study, it is hoped that, should such differences be found, they may be explainable, at least partly, in terms of other differences between groups. Religious Differences Denominational preference. Martin and Wrightsman36’ 37 criticize existing studies of religious differences in death attitudes for not focusing on a broad enough spectrum of churchgoers. However, they recommend limiting such research to church-going adults, who they expect to exhibit more differences among themselves than populations of college students (the people most often used as subjects in such 35Thornton Hooper and Barnard Spilka, "Future time and death among college students," Omega 1, 1970, 49-56. 36David Martin and Lawrence S. Wrightsman, Religion and fears about death: A critical review of research," Religious Education 59 (2), 1964, 174-176. 37David Martin and Lawrence S. Wrightsman, "The relationship between religious behavior and concern about death," Journal of Social Psychology 65, 1965, 317-323. 24 studies). Vernon38 argues to the contrary that to use 'none' as a unitary category in asking about religious affiliation is to obscure differences which may exist among those who are not members of estab- lished religious denominations. He suggests that that category be further divided into such subcategories as 'religious independent' or 'no religion at all,' or whatever categories appear relevant to the question at hand. By implication, he contends that a population of church-going adults is too limited to diSplay the full range of possible differences in attitude, since they may be more like each other than like those who do not belong to churches. In this study, an attempt is made to ask about a full range of possible religious preferences, in order not to obscure real differences. Most studies of death attitudes do focus on people who are members of major American religious groups, a fact which might account 39 found no for the minimal differences often found. Templer and Dotson relationship between religious belief or religious behavior and scores on Templer's Death Anxiety Scale--but only one Jew and two nonbelievers were included in the predominantly Catholic and Protestant sample. They theorize that, rather than being an over-all determinant of death anxiety, religious upbringing may interact with other personality variables to result in an abnormally high or low fear of death. It may, however, be the case that their sample consisted of people who were too much like each other in religious conviction fer real differences to emerge. 38Glenn M. Vernon, "The religious 'nones:' A neglected category," Journal for the Scientific Study of Religion 7, 1968, 219-229. 39 Donald 1. Templer and Elsie Dotson, op. cit. 25 In a sample of college students, such as is used in the current study, it is likely that there will be a fair amount of diversity of religious conviction and membership. Many subjects may have changed their religious membership or beliefs during adolescence or young adult- hood, or may have given up religion entirely, either temporarily or permanently. On the hypothesis that there may be small differences among people which are relevant to their attitudes toward death, and that changes in religious conviction may also influence (or be the result of) attitudes toward death, it is attempted here to ask about religious membership in enough detail so that differences may emerge. Respondents are asked what religion, if any, they were brought up in, what their current religious persuasion is, how religious they con- sider themselves to be, and whether they believe in an afterlife. Faunce and Fulton40 gave college students a sentence completion questionnaire about death, and arrived at a measure of the consistency of each respondent's answers. Among high consistency scorers, there were two general orientations: temporal (concerned with what happens to the body, funeral customs, the finality of death, disbelief in an afterlife, concern for the bereaved, separation from family and friends, and emphasis on this life as having been satisfying) and spiritual (death as a transition to another life, an end to one's 4William A. Faunce and Robert L. Fulton, "The sociology of death: A neglected area of research," Social Forces 36, 1958, 205-209. 26 41, 42 43 troubles, or a prelude to judgment). (Feifel and Shrut arrived at these same general categories of attitudes.) Catholics had a significantly higher mean consistency score than non-Catholics, accounted for by a higher proportion of Catholics with a consistently Spiritual orientation. The orientation of fundamentalist Protestants was more like that of Catholics than like that of more liberal Protestants, a result which suggests that denominational differences, although a convenient a priori way to categorize people, may not be very relevant in looking for religious differences in attitudes toward death. Some studies have found denominational differences however. In interviewing Catholic and non-Catholic nursing home personnel (a group which, working in a setting in which death is a regular occurrence, may be expected to have faced the necessity of finding some way of 44 found c0ping with uncomfortable feelings about it), Pearlman et al. that Catholics viewed death in a positive light, as the beginning of another life, and found their religion comforting in dealing with the eventuality of their own death more often than non-Catholics, who viewed death negatively, as the end of life, and did not find comfort from their religion in confronting the idea of death. Non-Catholics said they would deal with difficult feelings about death by avoiding 41Herman Feifel, "Attitudes of mentally ill patients toward death," Journal of Nervous and Mental Diseases 122, 1955, 375-380. 42Herman Feifel, "Older persons look at death," Geriatrics 1956, 127-130. 4353NU€1 0- Shrut. "Attitudes toward old age and death," Mental Hygiene 42, 1958, 259-266. 44Joel Pearlman, et al., "Attitudes toward death among nursing home personnel," Journal of Genetic Peychology_ll4, 1969, 63-75. 27 thinking about it, keeping busy, and crying, while Catholics said they would remain composed, talk about it, and pray. Diggory and Rothman,45 examining the extent to which different consequences of their own death were feared by members of different groups, fbund that no longer being able to have any experiences, and the ending of all one's plans and projects, was near the top of what was feared in all groups, but ranked lower for Catholics than for others. They suggest that Catholics regard death less as a termination of experiences than do non-Catholics. Fear of what might happen to the body after death was least in the Other-None religious group, and most among Catholics, which they suggest might reflect Catholics' concrete- ness of belief about an afterlife (and therefore concreteness of possible bad things that might happen after death). The possible pain involved in dying was feared most by Protestants, least by Jews, with Catholics a middle group. Again, the consistently more spiritual orientation of Catholics is suggested. 46 feund that In a group of college students, Golburgh et a1. Jews reported belief in an afterlife significantly less often than either Catholics or Protestants, but significantly more often said that their belief in an afterlife (or lack thereof) influenced their feelings about personal death. Jews were also less often willing to say they would die for a cause than the other two groups. 45 46Stephen L. Golburgh et al., "Attitudes of college students toward personal death," Adolescence 2 (6), 1967, 212-229. James C. Diggory and Dorreen Z. Rothman, op. cit. 28 In the only study in which specifically fear of death was 47 feund no differ- examined in different denominational groups, Lester ences among Catholics, Protestants, and Jews in fear of death of self, dying of self, death of others, or dying of others. Religiosity. 'Religiosity' is defined by Williams and Cole as 48 "the magnitude of religious activity reported by the subjects." Several studies have examined this variable, generally conceived of as the extent of religious activity, or number of religious activities, people report involvement in. The term is also sometimes used to refer to the extent of commitment a person expresses to a religious belief system, or to the importance of that belief or commitment to him or her. These two dimensions of religious behavior overlap, and since the term 'religiosity' is variously operationalized, they will be treated together here, all definitions of 'religiosity' may be treated as referring to a dimension of religious involvement. Williams and Cole49 found that religiosity‘ was not related to GSR reactivity to death-related words. Subjects of all levels of religiosity showed greater reactivity to death words than to affectively neutral words. 50 Lester found a conflicting relationship between religiosity and fear of death. Less religious subjects scored significantly higher 47David Leter, "Religious behavior and the fear of death," Omega 1, 1970, 181-188. 48Robert L. Williams and Spurgeon Cole, "Religiosity, general- ized anxiety, and apprehension concerning death," Journal of Social Eeychology 75, 1968, lll-117 (quote page 111). 49 Ibid. 50Lester, op. cit. 29 on his fear of death scale than did more religious people. On Collett and Lester's scale (which contains most of the items of Lester's scale), less religious subjects scored significantly higher on fear of dying of self, but not on the other subscales (this would be consistent with a more temporal orientation towards death). Less religious subjects had greater fear for self than for others, while in more religious subjects that relationship was reversed. Less religious subjects were also less inconsistent in their attitudes toward death. Faunce and Fulton found that those who attended church more frequently had more consistent attitudes towards death, and more often a spiritual atitutde than a temporal one. The relationship between fear of death and consistency of orientation remains to be further investi- gated, but emotionally oriented responses to the sentence-completion task given to subjects in this study, suggesting higher fear of death, were more common among spiritually-oriented people. In addition, people with incomes over $10,000 and under $3,000 were more consistently spiritual in orientation than were middle income groups. . . it would appear to be consistent with Veblen's suggestion that the general orientation toward life of the middle class is temporal, means-oriented, and secular, while that of the upper and lower classes is more agt to be trans-temporal and fate or luck oriented. 1 52 found that college students who had a strong attach- Templer ment to their religious belief system, attended religious functions more frequently, were certain of the existence of an afterlife, believed in a 51 52Donald 1. Templer, "Death anxiety in religiously very involved persons," Psychological Reports 31, 1972, 361-362. Faunce and Fulton, op. cit., p. 208. 3O literal interpretation of the Bible, and considered themselves strong in religious conviction compared with others, had lower scores on his Death Anxiety Scale. Scale means in this religiously very involved group were lower than in any other group to which the instrument had been given. No relationship was found between Death Anxiety Scale score and denominational affiliation, whether the person was still in the religion s/he had been brought up in, or whether the main reason given for being religious was so as to have the possibility of a life after death. Templer suggests that even those college students who are religiously very involved may not be especially religious compared with the general population rather than with other college students, and that the college period is one in which religion plays a small part in a person's life, so that more differences would be found between very religious and nonreligious people in the general population than were found in this group. However, it may also be that college students who do define themselves as very religious are likely to be even more so than those in the general population who so define themselves, since peOple in this environment and of this age frequently ferm fairly intense attachment to belief systems. Several studies have compared self-defined religious and non- religious people. Usually this means that relatively_more or less religious people have been compared with each other, not that a special effort has been made to seek out genuinely non- or antireligious people. Often, the effect is that people of relatively different degrees of religious involvement from within a given religious group are compared with each other. 31 Alexander and Adlerstein53 divided middle-class, male, Protestant undergraduates into religious and nonreligious groups. They found that eighty percent of the religious subjects reported having become aware of the existence of death before age six, compared with only thirty percent of the nonreligious subjects, and the religious subject tended to report clearer (although earlier) memories of their experiences with death, and more feelings about death and burial. At the end of the experiment, manifest anxiety levels in the two groups were approximately equal, but there was some indication that general anxiety was aroused more easily by death-related stimuli in the nonreligious subjects. 54 found, contrary to expectation, that in a group of Adlerstein male undergraduates, the religious group showed a significantly greater increase in manifest anxiety as a result of completing a questionnaire and interview on death attitudes. Religious and nonreligious people did not differ in their semantic differential ratings of death words, judging the concept bad and potent. Feifelss found religious pe0p1e to be personally more afraid of death, especially in an OREr'population, in which religious people held a more negative orientation toward the later years of life than did the nonreligious. Swenson56 found that elderly people with more fundamentalist 53Irving E. Alexander and Arthur M. Adlerstein, "Death and Religion,")in H. Feifel, ed., The Meaning_9f Death (New York: McGraw- Hill, 1959 . 54Arthur M. Adlerstein, The Relationship between Religious Belief and Death Affect, Ph.D. Dissertatibnj‘Princeton‘Univarsity, 1958. 55Herman Feifel, "Attitudes toward death in some normal and mentally ill populations," in H. Feifel, ed., The Meaning of Death, (New York: McGraw-Hill, 1959), 114-130. 56 Swenson, op. cit. 32 religious beliefs and habits looked fbrward to death more than those with less fundamentalist beliefs. Ray and Najman57 found that religious un- believers were more acceptant of death than believers, in a group of undergraduates, but there was no significant correlation between religious belief and death anxiety. 58 feund that those who indicated that they Feifel and Branscomb did not fear death "because it is God's will" rated themselves more religious. The more religious subjects also had more positive fantasy imagery about death, and more often rated their own death as "clean," "fair," "kind," and "sociable" than less religious subjects. 59 60 and Feifels] have shown no differ- Studies by Blake, Kalish, ences between believers and unbelievers. However, Blake notes that attitudes toward death expressed in religious language did not correlate with the same attitudes expressed in secular language, and suggests that the use of religious language may confuse the interpretation of the expressed attitudes. He contends that religious belief may offer the opportunity for the individual to deny the reality of death. Similar- ly Adlerstein suggests that the evidence does not support the hypothesis that strong religious belief effectively reduces negative affect toward death: 57 . . Ray and Najman, op. c1t. 58Feifel and Branscomb, op. cit. 59 Robert R. Blake, Attitudes Toward Death as a Function of Developmental Stages, Ph.D. Dissertation, Northwestern University, 1969. 60Richard A. Kalish, "Some variables in death attitudes," Journal of Social Psychology 59, 1963, 137-145. 61Herman Feifel, "Religious conviction and fear of death among the healthy and the terminally ill," Journal for the Scientific Study of Religion 13, 1974, 535-560. 33 The evidence that does emerge suggests that religious and nonreligious people use different methods fer bind- ing their negative affect and anxiety toward death. The nonreligious subjects handle their affect by suppres- sing memories and feelings about death. The religious subjects handle their affect by denying the reality of physical death, focusing instead on_the afterlife. 2 Belief in afterlife. Belief in an afterlife is a particular content of religious belief which is often thought to serve the purpose 63 of reducing fear of death. Middleton in 1936, in one of the earliest studies of death attitudes, reported that sixty-six percent of his 64’ 65 found college student subjects believed in an afterlife. Lester that males (in Middleton's 1936 sample and a similar one studied in 1970) more often than females reported wanting to know fer sure whether there is an afterlife, and said that they would change their manner of living if they knew for sure that there was one, but that they were less likely to believe in one. Lester66 found that students in 1970 were less likely to express a wish to live after death, more likely to want to know for sure whether there is a life after death, and less likely to believe in one, than were students in 1936. 62Adlerstein, op cit., abstract. 63Warren C. Middleton, "Some reactions toward death among college students," Journal of Abnormal and Social Peychology 31, 1936, 165-173. 64David Lester, “Re-examination of Middleton's data: Sex differ- ences in death attitudes," Psychological Reports 27, 1970, 136. 65David Lester, "Sex differences in attitudes toward death: A replication," Psychological Reports 28, 1971, 754. 66David Lester, "Attitudes toward death today and thirty-five years ago," mega 2, 1971, 168-173. 34 67 Jeffers et al. found that, in people over 60, fear of death was associated with less belief in an afterlife, as well as with less fre- 68 administered stimuli quent reading of the Bible. Osarchuk and Tatz which intensified fear of death temporarily, and fOund that belief in afterlife increased in those in whom it was already high, but not in those whose initial belief was low. Belief in afterlife, they conclude, is reinfbrcing as a means of fear reduction in some people. Berman69 found no association between a subject having experi- enced a life-threatening situation and whether s/he believed in an after- life, but significantly more religiously inactive subjects reported at least one life-threatening experience than religiously inactive subjects. h70 fbund no significant correlation between fear of death and Kalis belief in an afterlife or belief in God. In summary, the evidence appears to suggest that there is some religious influence on the fear of death and other attitudes toward death, or some influence of those feelings on a person's religious con- victions, but the nature of the relationship is far from clear. Evidence has been found that religious affiliation, religiosity, and belief in an afterlife both are and are not related to the fear of death. It is 67Frances C. Jeffers, et al., "Attitudes of older persons toward death: A preliminary study," Journal of Gerontology 16 (l), 1961, 53-56. 68Michael Osarchuk and Sherman J. Tatz, "Effect of induced fear of death on belief in afterlife," Journal of Personality and Social Eeychology 27 (2), 1973, 256-260. 69Alan L. Berman, "Belief in afterlife, religion, religiosity, and life-threatening experiences," Omega 5 (2), 1974, 127-135. 70 Kalish, 0p. cit. 35 probable, therefore that some combination of variables will predict fear of death better than any single variable, at least of a religious nature.71 Age Differences Studies which have examined relationships between age and the fear of death have produced conflicting evidence. 72 73 Golburgh et a1. and Hooper and Spilka found no relationship between the age of college students and whether they held positive or 74 negative attitudes toward death. Rhudick and Dibner found that the number of death references given in stories told to TAT cards (a measure of unconscious death concern) by elderly people did not vary with age, 75 and Swenson also found no differences among the elderly. Templer and Ruff76’ 77 found no age differences in fear of death in a sample which ranged in age from 19 to 85 years, and similar results are reported by Lester.78 Only in three studies have clear age-related differences in the 79 fear of death been shown. Blake found that older people reported less 71 72 Chasin, op. cit. Golburgh, et al., op. cit. 73Hooper and Spilka, op. cit. 74Rhudick and Dibner, op. cit. 75Swenson, op. cit. 76Templer and Ruff, Developmental Psychology 4(1). Op- Cit- 77Templer and Ruff, P§ychologi§a1 REPOVtS 29 (1): OP- Cit- 78Lester, 1972, op. cit. 7931ake, op. cit. 36 80 found that the elderly fear than adolescents. Reynolds and Kalish reported less fear than either the middle aged or the young, although there was no relationship found between fear of death and the expecta- tion of or wish for a long life. Feifel and Branscomb8] found that those who admitted fear of death were significantly younger than those who denied it, and that subjects in the 59 to 70 year old group denied fear of death more often than those in other age groups. Older people more often rated their own death as "clean," "fair," "kind," and "sociable" than did younger subjects, and, on a word association task, took longer to react to death-related words than younger people. This suggests that, while the elderly were less afraid at a conscious level (or less willing to express fear), they feared death at an unconscious level more than younger persons did. Cameron82 found that people tend to grossly over-estimate the probability of their own death within the next year, but that older people more closely approximate the actuarial probability of their own 83 found that people death than do younger ones. But Tolor and Murphy under 30 were more realistic in estimating their life expectancy than were people over 30. Diggory and Rothman84 found that older (40-54 years of age) people were more concerned with their inability to provide 80Reynolds and Kalish, op. cit. 8lFeifel and Branscomb, op. cit. 82Cameron, op. cit. 83Alexander Tolor and Vincent M. Murphy, "Some psychological correlates of subjective life expectancy," Journal of Clinical Psychology_ 23, 1967, 21-26. 84 Diggory and Rothman, op. cit. 37 for their dependents as a result of their death than were those in the 15-19 year old group, a predictable result since the younger age group is unlikely to have had any experience of having and caring for dependents. 85 found lowered skin resistance (GSR) Alexander and Adlerstein to death-related words on a word-association task in a group of five to 16 year old boys, but this effect was least in the nine to twelve year old group. The authors suggest that the latency period is less stressful to the ego than other ages, and that therefore people in this age group respond to stressful stimuli with less anxiety. 86 Bromberg and Shilder contend that subjectively held attitudes usually are similar to the attitudes the person attributes to others. Fei fel 87 asked mentally ill patients at what times of life they thought people most feared death. Many chose childhood and old age as the times during which people were either most or least afraid of death. There were no differences in responses from people under or over thirty years old. Feifel88 found that patients and older people saw old age as the time of life when people most fear death, while younger normals saw the forties and fifties as the times when people are most afraid. 85Irving E. Alexander and Arthur M. Adlerstein, "Affective re- sponses to the concept of death in a population of children and early adolescents," Journal of Genetic Psychology 93, 1958, 167-177. 86W. Bromberg and P. Shilder, "The attitudes of psychoneurotics toward death," Psychoanalytic Review 20, 1933, 133-185. 87Herman Feifel, "Attitudes of mentally ill patients toward death," Journal of Nervous and Mental Diseases 122, 1955, 375-380. 88Herman Feifel, "Attitudes toward death in some normal and mentally ill populations," in H. Feifel, ed., The Meaning of Death (New York: McGraw-Hill, 1959), 114-130. 38 However, evidence from other studies suggests that, if there are age differences, they are in the direction of younger people being more fearful of death. Since the evidence is scant, it is difficult to tell whether this discrepancy indicates that there really are no general age differences in fear of death, or that the attribution of attitudes to others is not a good method of assessing the respondent's own attitudes. Peripherally, patients ranked childhood as second to the seventies as the age in which people most fear death, while normal groups said that childhood is the time when death is feared least. Feifel suggests that childhood ideas of death have to do with deprivation, and that the difference may be accounted for if one assumes that the patient group came from more deprived childhood environments. Health and Personality Several conditions of mental and physical health have been examined for relevance to feelings about death, and psychiatric and medical patients have often served as subjects fer such research. In this section, research which has a bearing on mental and physical health, and personality variables, is summarized. 89 Mental and physical health. Feifel found that psychiatric patients chose old age or childhood as the times in which people most or least feared death, with the tendency to select old age more often. He also suggests that they see childhood as a less secure time than nonpatients. Some saw death as occurring violently, although most said they saw it as the natural end of the life process, and some found it so anxiety-arousing that they denied having any thoughts about it at all. 89Ibid. 39 The conjecture is that a violent conception of death mirrors self-held feelings of aggressiveness toward others as well as toward oneself.90 Asked what they would do if faced with imminent death, patients tended to choose social and religious activities, in contrast to normal groups who emphasized personal pleasures and gratifications. The degree of mental illness had little effect on attitudes toward death. Neither neurosis nor psychosis produces attitudes toward death which cannot also be found in normal subjects.91 Lester92 found no significant differences in fear of death between high and low scorers on the neuroticism scale of the Maudsley Personality Inventory. Templer and Ruff93 summarize research on the Death Anxiety Scale which shows that psychiatric patients score con- sistently higher than nonpatients. Thus, the nature and direction of differences between patients and nonpatients are far from clear. Research on the effect of physical health and illness on death anxiety has also produced conflicting results. Templer94 sent the Death Anxiety Scale, the 0 scale of the MMPI, and the Cornell Medical Inventory (a report of one's own opinions of the physical and psychiatric symptoms from which one suffers) to a group of 250 retired persons, of which 75 were returned. He found that death anxiety was positively goIbid. 91 92David Lester, "Religious behavior and the fear of death," Omega 1, 1970, 181-188. 93 94Donald I. Templer, "Death anxiety as related to depression and health of older persons," Journal of Gerontolegy_26 (4), 1971, 521-523. Ibid. Templer and Ruff, op. cit. 40 related to depression, and to self-report of a high number of psychia- tric symptoms. Removing the influence of the psychiatric scale on the somatic scale, death anxiety was negatively related to a high number of somatic symptoms. Templer contends that death anxiety is part of a depressive syndrome in the elderly, and may be treated in part by treating depression, but that death anxiety should not be regarded as a cause of depression in the elderly. A negative relationship is suggested between ill health and death anxiety; when psychiatric disturbance was controlled for, physically healthy people had more death anxiety than 95 found that older people in poor health tended sicker ones. Swenson to have positive attitudes towards death, while those in good health tended to avoid the idea of death. (He found no relationship between death anxiety and any MMPI scales.) Rhudick and Dibner96 gave the MMPI, the Cornell Medical Inventory, and a set of TAT cards to elderly men and women. High death concern was associated with high scores on Hypochondriasis, Hysteria, Dependency (Navran) and Impulsivity (Gough), but there were no significant correlations with any other scales. Individuals who reported a higher number of physical and/or psychiatric complaints showed higher death concern than those who reported fewer symptoms. In this study, it is the sicker individuals who show greater death concern. The measure of death concern used in this study is intended to reflect unconscious death concern, in contrast to Templer's Death Anxiety Scale, which measures consciously held attitudes. It may be that sickness and 95Swenson, op. cit. 96Rhudick and Dibner, op. cit. 41 health bear a different relationship to death anxiety at different levels of awareness. The lack of relationship of high death concern to anxiety and psychasthenia suggests that those who have preconscious death concern do not have anxiety of the free-floating or obsessive variety; rather, the anxiety tends to be attached to bodily symptoms. This statement seems to be partly corroborated by the finding on the CMI data, i.e., the more frequent the reporting of physical symptoms, the higher the death concern. It is important to note that the CMI elicits attitude toward health, not necessarily the actual physical status of the respondent. The strongest relationships are suggested by the subjects who admit to many physical symptoms on the CMI and also score high on the Hypochondriasis scale of the MMPI. 98 Christ found that elderly psychiatric patients in better health were less afraid of death than those with poorer health. Lucas99 found no differences in either general anxiety or death anxiety among groups of dialysis and surgical patients and their respective wives, and found that they scored very similarly to normal groups on Templer's Death Anxiety Scale, suggesting that serious physical illness does not heighten death anxiety. ‘00 10] found no differences between Munro and Hopkinson and Reed depressed psychiatric patients and normals in the rate at which they had lost a parent or a sibling by age 16. Munro, however, found that twice 97Ibid. 98Adolph E. Christ, "Attitudes toward death among a group of acute geriatric psychiatric patients," Journal of Gerontology l6 (1), 1961, 56-59. 99 100Alistair Munro, "Parental deprivation in depressive patients," British Journal of Peychiatry_112, 1966, 443-457. 1016. Hopkinson and G. F. Reed, "Bereavement in childhood and depressive psychosis," British Journal of Peyphiatry_112, 1966, 459-463. Lucas, op. cit. 42 as many severe as moderate depressives had lost a parent in childhood, 102 feund that psychiatric patients who usually their mother. Greer had attempted suicide had lost.eetg_parents significantly more often than patients of the same age who had not made suicide attempts. There was no significant difference between the two groups in the sex of the parent lost when only one had been lost, or in the cause of loss, ‘03 found that patients whether separation or death. Cash and Kooker who had attempted suicide were more inconsistent in their attitudes toward death than either patients who had not made attempts or graduate and undergraduate students. These results may mean that the suicidal patients were more ambivalent in their attitudes towards death. The relationship between death anxiety and general anxiety is positive, but general anxiety measures correlate more highly among themselves than they do with measures of death anxiety, suggesting that death anxiety, while increased by general anxiety level, also 104, 105 has other components. Templer found moderate correlations between scores on his Death Anxiety Scale and scores on three MMPI scales of general anxiety, correlations which were lower than those 106 among the MMPI scales themselves. Dickstein found that, in females, 102S. Greer, “Parental loss and attempted suicide, a further report," British Journal of Psyehiatgy_112, 1966, 465-470. 103Larry M. Cash and Earl W. Kooker, "Attitudes toward death of neurOpsychiatric patients who have attempted suicide," _eychological Reports 26, 1970, 879-882. 104Donald 1. Templer, The Construction and Validation of a Death Anxiety Scale, Ph.D. Dissertation, UnTVersity of Kentucky, 1967. 105Donald 1. Templer, "The construction and validation of a death anxiety scale," Journal of General Psychology 82, 1970, 165-177. 106 Dickstein, op. cit. 43 death concern was positively correlated with state anxiety and train anxiety, and with manifest anxiety in both males and females. He reports about 13 percent common variance between death concern and 107 manifest anxiety. Nogas et al. found that general anxiety accounted for about 16 percent of the variance of death anxiety, and 108 Handal obtained low but significant correlations between the Livingston and Zimet fear of death scale and a measure of general anxiety. Only 109 Williams and Cole obtained results which indicated no relationship between death anxiety and general anxiety. Tolor and Reznikoffno gave subjects a revision of Livingston and Zimet's scale, Byrne's repression-sensitization scale, and Rotter's scale of internal and external locus of control. Sensitizers and internals were significantly more death-anxious than repressors and externals. While the first difference was in the predicted direction, the finding that internals were more death-anxious than externals was contrary to expectations. The reasoning behind the prediction was as fellows: 107Catherine Nogas, et al., "An investigation of death anxiety, sense of competence, and need for achievement," Omega 5 (3), 1974, 245-255. 108Paul J. Handal, "The relationship between subjective life expectancy, death anxiety, and general anxiety," Journal of Clinical Psychology 25 (l), 1969, 39-42. 109 Williams and Cole, op. cit. noAlexander Tolor and Marvin Reznikoff, "Relation between insight, repression-sensitization, internal-external control, and death anxiety," Journal of Nervous and Mental Diseases 140, 1965, 222- 230. 44 Since the concept of death and the process of dying re- present in our society phenomena that are potentially threatening to many individuals, the degree of overt anxiety experienced in this area may be regarded as one indication of a characteristic response to extremely threatening stimuli. Sensitization should therefore be related to a heightened degree of overt death anxiety as compared to repression. Similarly, since the person with an external orientation believes that factors beyond his control determine his failures, external expectancies should be more associated with overt death anxiety than the belief in internal control of reinforcement. 1 It may be supposed, however, that the person who believes in internal control of reinforcement, when faced with a situation like death in which that control is not there, would suffer more anxiety than one who already believes that what happens is under the control of external forces. In another study, Templer112 found that sensitizers scored significantly higher on his Death Anxiety Scale. There were low but significant correlations between Death Anxiety Scale scores and GSR readings taken as subjects gave associations to death-related words, but there was no correlation between GSR readings and repression- sensitization. He concludes that autonomically measured death anxiety is independent of repression-sensitization, while consciously reported death anxiety may not be. The research on the relationship between death anxiety and other personality variables is peripheral to the current study. Many studies show no such relationships to exist, particularly among females. No relationships have been found, for either sex, between death anxiety 1”Ibid., p. 223. 112Donald I. Templer, "The relationship between verbalized and nonverbalized death anxiety," Journal of Genetic Peychology_ll9, 1971, 211-214. 45 113, 114 115, 116 external locus of control, fll7 and need for achievement, (as noted above, Tolor and Reznikof found contrary results), fear 118 119 of failure, or denial. Fear of death has been found to be correlated with needs fer heterosexuality and succorance on the Edwards Personal Preference 120 121 Schedule, with dependency and guilt about hostility (in males only), with lower sense of purpose and meaning in life,122 123 and with present- Jeffers et a1.124 as opposed to future orientation. reported that fear of death, in people over 60 years of age, was correlated with feelings of rejection and depression, lower full-scale IQ, lower performance IQ, and fewer Rorshach responses; and Paris and H3Nogas et a1. op. cit. 114Ray and Najman, op. cit. 115$elvey, op. cit. H6Dickstein, op. cit. H7Tolor and Reznikoff, op. cit. 118Ronald J. Cohen and Christian Parker, "Fear of failure and death," ngchological Reports 34, 1974, 54. 119 120 Ray and Najman, op. cit. Dickstein, op. cit. 12‘Selvey, 0p. cit. 122Joseph A. Durlak, "Relationship between individual attitudes toward life and death," Journal of Consultingiand Clinical Psychology, 38 (3), 1972. 463. 123Kahana and Kahana, op. cit. 124Jeffers et al., 0p. cit. 46 Goodstein125 obtained tentative results indicating that women felt some sexual arousal upon reading death-related literary material. Several researchers have investigated dreams with themes of death, and have looked for connections between dreaming and the fear of death. Handal and Rychlak126 found that college students who scored either high or low on Handal's Death Anxiety Scale127 reported more un- pleasant dreams and more dreams of death than those who scored in the middle range, possibly indicating that that scale measures repression, at least in part. Feldman and Hersen theorize that nightmares "mirror . areas of conflict with which the dreamer feels especially help- less and unable to cope,"128 and that death may be this kind of conflict for some people. Undergraduates who reported more frequent night- mares scored higher on their scale of conscious death doncern. Women reported more nightmares and death concern than men. Those who had more nightmares had lost an important other befbre the age of nine significantly more often, and those who experienced fewer nightmares reported fewer and later losses of significant others. Lesterlzg’130 125Joyce Paris and Leonard D. Goodstein, "Responses to death and sex stimulus materials as a function of repression-sensitization," Psychological Reports 19, 1966, 1283-1291. 126Paul J. Handal and Joseph F. Rychlak, "Curvilinearity between dream content and death anxiety, and the relationship of death anxiety to repression-sensitization," Journal of Abnormal Psychology_77 (l), 197], 1.1-16. 127Handal, op. cit. 128Feldman and Hersen, op. cit., p. 421. 129David Lester, "Fear of death and nightmare experiences," Eeychological Reports 25, 1969, 437-438. 130David Lester, "The fear of death of those who have nightmares," Journal of PsyChology 69, 1968, 245-247. 47 failed to replicate that study, finding no association between frequency of dreams or nightmares with fear of death measured on a variety of scales, although there was an association between poor memory of dreams and lower fear of death, contrary to prediction. ‘3] studied the dreams of death of Mexican-American Roll et a1. and Anglo-American students. Mexican women reported more dreams of death than Mexican men, while among Anglo students that relationship was reversed. Clearly this implies that caution must be exercised in generalizing findings about death attitudes and other variables outside the cultural milieu in which they were obtained. Learning of death anxiety, Death anxiety may be determined by personality variables and intrapsychic dynamics, may be determined by early unresolved conflicts, may be instinctual, may be a survival device for the species as a whole, or may be a learned phenomenon which is amenable to change. If it is subject to change, the question arises of how resistant to change it is, whether it is learned in early childhood and remains fairly stable thereafter, or whether it is susceptible of influence by current events in the life of the indivi- dual. Finally, there is the question of what, if any, experiences and life situations contribute to the fear of death. The question to be discussed in this section is the more general one of how stable or changeable death attitudes are. This is a question of both theoretical and practical importance, since it is the 1315amue1 Roll, et al., "Dreams of death: Mexican-Americans vs. Anglo-Americans," Revista Interamericana de Psicologia 8 (1-2), 1974, lll-115. ‘ 48 overall aim of this study to produce information which may be useful in the designing of training programs for medical students which may make impacts on their attitudes which will be useful to them in improving their interactions with their patients. 132 Freud believed that the unconscious did not have the idea of death, and could not contain the conception of its own extinction. Klein133 argues to the contrary that anxiety of all kinds originates in the fear of death, rather than the fear of death being, as Freud believed, a censcious reflection of some other unconscious fear (such as fear of castration). Klein argues that the unconscious does have the idea of death, the death instinct, and hence has the fear of its own annihilation. The struggle between life and death instincts continues throughout life and contributes to all anxiety situations, because it is the basis for anxiety. However, the research which is reviewed in this chapter suggests that different life experiences affect the feelings an individual has about death, and while a purely correlational study will not explain the origin of death anxiety, it will perhaps allow a picture to be drawn of the kind of individual who is likely to experience a heightened level of fear of death. Research in the area of parent-child resemblances in death attitudes is relevant to the general theoretical question of how death attitudes are learned and from whom, and how stable they are. 132Sigmund Freud, Beyond the Pleasure Principle,in Standard Edition, Vol. XVIII,(London: Hogarth Press, 1955) (Original, I920). 133 Melanie Klein, "A contribution to the theory of anxiety and guilt," International Journal of Psychoanalysis, 1948, 29. 49 Lester and Templer134 compared the fear of death of males and females aged 13 through 19 with that of their parents. The magnitude of the correlation between mother and daughter, and between father and daughter, decreased with the age of the child. The mother/daughter correlation in the 13-14 year old group was significantly higher than in any other age group. The mother/son and father/son correlations increased with the age of the son, with those for the 18-19 year old group being significantly higher than all others. Lester135 found a significant correlation between scores of mothers and those of their undergraduate daughters on the Collett and Lester scale. Daughters reported significantly more fear of death of self than their mothers but did not differ from them in fear of dying of self, death of others, or dying of others. There were no significant differences on any of these scales between fathers and daughters. Templer and Ruff found positive correlations, ranging from .34 to .51 between Death Anxiety Scale scores of parents and their male and female children. Scores of the adolescents correlated most highly With those of the same sex parent, and scores of the parents were also significantly correlated with each other, .59. These data suggest that explanations based upon principles of learning account better for the observed parent-child correlations than explanations invoking genetic similarities. The substantial correlations between Death Anxiety Scale scores of parents further suggests that death anxiety is far from being dependent largely upon early childhood experiences and basic personality structure. It appears 134David Lester and Donald I. Templer, "Resemblance of parent- child death anxiety as a function of age and sex of child," Psychologi- cal Reports 31, 1972, 750. 135David Lester, "Relation of fear of death in subjects to fear of death in their parents," isychological Record 20, 1970, 541-543. 50 that death anxiety is not so much a fixed entity as a state that is sensitive to environmental events in general and to the impact of intimate interpersonal relationships in particular.136 An alternate hypothesis is expressed by Golburgh et al., who asked college students about their relationships with their parents and the extent to which they had discussed death with them. Those who were uncertain whether their parents had ever discussed personal death with them were significantly less afraid of death than those who knew for sure that their parents either had or had not done so. Those who had had no discussion of death with their parents tended to discuss it less with others, possibly indicating that they had learned to deny its reality. Those who had discussed it with their parents had more feelings that they could die comfortably under some conditions. Many of those who did not fear death described their current relationship with their parents as "very poor." The number of contacts with death were about equal in those who said they fear, do not fear, or do not think about death. Contacts with the deaths of others did not significantly influence attitudes toward death. While close contact with the death of another undoubtedly influences one's feelings about death on a temporary basis, long-term changes in attitudes appear to be unlikely. An equilibrium seems to be re-established and the original attitudes remain intact. This might support the authors' hypothesis that attitudes toward death are developed at early stages in development and external infggences thereafter do not significantly influence them. 136Donald I. Templer and Carol F. Ruff, "Death anxiety: age, sex and parental resemblance in diverse populations," Developmental Psychology4 (l), 1971 p. 108. - 137Stephen L. Golburgh, et al., "Attitudes of college students toward personal death," Adolescence 2 (6), 1967, 212-229. 51 The question is how much an individual's attitudes toward death are subject to change--how much they are influenced by early experiences and how much by current events in the person's life. A few studies of the effects of recent loss or recent contact with death are relevant here. Bruhn et a1.138 had nurses in a coronary care unit periodically rate patientis levels of anxiety, unaware that those ratings of concern were the ones made close to the time of the death of a fellow patient. Ratings of anxiety made by the nurses, and systolic blood pressure levels as well, were higher than normal immediately fellowing the death of a patient with whom the individual had shared a room. Lester and Kam questioned female undergraduates, half of whom had experienced the death of a close friend or relative in the past five years and half of whom had not, about their thoughts about death. Those who had experienced a recent loss thought of their own death more frequently, were more inclined to entertain thoughts of dying from some specific disease, more likely to picture death as horribly painful, and reported more often being depressed by cemeteries. Those who had experienced a recent loss were generally more preoccupied with thoughts about death and more depressed and fearful. The existence of a weak, but nonetheless consistent, effect from experiences of a recent loss suggests that recent loss serves only to modify very slightly already existing attitudes toward death. It is more probable, therefore, that.the determinants of atti udes toward death must be sought in earlier experiences. 138John G. Bruhn, et al., "Patients reactions to death in.a coronary care unit," Journal of Psychosomatic Research 14 (1), 1970, 65-70. 139David Lester and Elizabeth G. Kam, "Effect of a friend dying upon attitudes toward death," Journal of Social Psychology 83, 1971, 149-150,p. 150. 52 However, what one chooses to count as "recent" is somewhat 140 arbitrary. Selvey fbund that women who had lost someone in the last two years reported more fear of death than those who had not, '4‘ found that both but this relationship was not found among men. Boyar males and females who had experienced a death in their family had significantly higher fear of death than those who had not, and the recency of the death was not taken into account at all. Tolor and 142 Murphy found more unrealistic subjective life expectancy in subjects who had experienced the death of a spouse, close friend, or relative. On the other hand, Durlak143 found no relationship between scores on Lester's fear of death scale and whether the individual had ever experienced the death of a close friend or family member, or had ever been in a situation in which s/he had thought that his or her own 144 found that fear of death death was imminent. Feifel and Branscomb was not related to having had the recent experience of the death of a close person. None of these studies have considered the nature of the death that was experienced, whether it was premature or not, whether 140Carole L. Selvey, "Concerns about death in relation to sex, dependency, guilt about hostility, and feelings of powerlessness," Omega 4 (3), 1973, 209-219. 14lJerome I. Boyar, The Construction and Partial Validation of a Scale for the Measurement of the Fear of Death, PHID. Dissertatidn. UnTVersity of Rochester,71963. 142Alexander Tolor and Vincent M. Murphy, OSome psychological correlates of subjective life expectancy," Journal of Clinical Psychology 23, 1967, 21-26. 143Joseph A. Durlak, "Relationship between various measures of death concern and fear of death," Journal of Consultingyand Clinical Psychology 41 (1), 1973, 162. 144Herman Feifel and Allan B. Branscomb, "Who's afraid of death?" Journal of Abnormal Psychology 81 (3), 1973, 282-288. 53 it was sudden, unexpected, or long-awaited, or how close the subject 145 and Reynolds and Kalish146 felt to the person who died. Weisman argue that the problems of adjusting to different sorts of deaths may be different, and that there is no reason to expect that different experiences of death will have the same kind of impact. Feifel147 points out that knowledge of the external degree of threat of a death-related stimulus is insufficient to allow prediction of how a person will react to it. It is necessary to know about the person, to make an accurate prediction. Feifel contends that it is the person's character structure that must be known; it should be clear from the above that many factors must be considered. However, the work situation of a physician, except in certain specialties like psychiatry, must be considered a continual situation of threat of death of another, and of recent loss of persons whose significance to the physician may or may not be great personally, but into whose well being s/he has invested effort and perhaps some self- esteem. In this context, a passage from Diggory and Rothman is relevant: . . . utility of the self corresponds to the pro- bability that a person, by his own efforts, can achieve objectives that are important to him. The larger the number of important objectives for which one's probability of achievement is high, the greater his self-esteem. Loss of ability or skill reduces probability of achieve- ment, and with it, self-esteem. Elimination of 145Avery D. Weisman, "Coping with untimely death," Esychiatry 36 (4), 1973, 366-378. '45oavid K. Reynolds and Richard A. Kalish, "Work roles in death- related occupations," Journal of Vocational Behavior 4, 1974, 223-235. 147Herman Feifel, "Attitudes toward death in some normal and mentally ill populations," in H. Feifel, ed., The Meaning of Death (New York: McGraw-Hill, 1959), 114-130. 54 opportunity for the exercise of skills or abilities makes them, in effect, worthless, even though they are undamaged. We try to preserve or extend objects we value highly, but those of low value are treated with indifference or des- troyed. Thus, a person who values himself highly should be more afraid of death than one whose self-esteem is low, because death is the limiting case of loss or destruction of the self. . . . To the extent that the goals a person values highly depend on his social status, his fear of various consequences of his own death should vary with his status or role, wheiher defined by age, sex, social class, religion, or marital condition.14 The physician is in the position of investing self-esteem in the exercise of his or her skill at keeping people alive and healthy, so that the death of a patient may also represent the loss of self-esteem due to the loss of both an opportunity for the exercise of a skill and of the evidence of the worth of that skill. A death is a failure, an evidence of the inevitable failure of the physician's skill; it should then be the case that physicians, over the course of their careers, would develop different attitudes toward death than those held by the general popula- tion, since death has additional professional meaning fer them over and above the personal meaning it has. It may also be asked whether those who choose such a profession in the first place come into it with dif- ferent attitudes than students who enter other fields, that is, whether there is a reason connected with their attitudes toward death which in- clined them to enter a profession in which much of their self-esteem re- volves about successful battles against it. To summarize the issues involved, it is a matter of dispute whether death attitudes are formed by early childhood experiences and 148James C. Diggory and Dorreen Z. Rothman, "Values destroyed by death," Journal of Abnormal and Social Psychology_63 (l), 1961,205-210, p. 205. 55 are relatively stable, of whether they are susceptible to influence by later experiences and, if so, whether the resulting change is permanent or necessarily only temporary. Finally, what kinds of later experiences influence a person's attitudes towards death, or could this differ for different people? The results of the few studies which have addressed this issue are conflicting, suggesting that the nature of the death that was experienced and the nature of the individual's relationship to the deceased, and the emotional meaning of that death to the individual, in- fluence whether than particular death has a lasting impact on his or her attitudes and fears. It appears, on the basis of scant evidence, that recent loss may have a greater impact on females than on males, and that stable personal- ity traits have a greater influence on the attitudes and fears of males. Perhaps the development of fears of death takes a different course in males and females. The attitudes of fiemales resemble those of their parents less as they pass through adolescence, while those of males come to resemble those of their parents more over the same period. Perhaps this indicates that recent experiences have more impact on women, while males become more firmly set in attitudes to which they have been ex- posed all their lives. In this study, then, it is expected that differences will be found between the sexes, and among people in different fields depending on the amount of contact with death they have. And differences may be related to the recency of contact with death. 56 Other Demogrephic Variables~ 149 150 Race. Pandey and Tenpler and Reynolds and Kalish found no significant correlations between race and fear of death. 151 Occupation. Rhudick and Dibner found no significant associa- tion between death concern and occupational status, marital status, or educational level. Ford et a1.152 looked for differences among police- men, mail carriers, and undergraduates, predicting that job stress (assumed to be highest for patrolmen) would be positively associated with fear of death; there were no differences among the three groups. Lester153 compared staff members of a suicide prevention center with members of the same occupational groups working in a different setting. There were no differences in fear of death between the two groups, but the center staff were more consistent in their death attié tudes. Lester suggests that such consistency is to be expected from pe0p1e who work in a setting in which the possibility of death is a regular occurrence. .154 Magn1 found some indicating that theology students who planned to be parish priests identified tachistoscopically presented 149Pandey and Templer, op. cit. 150 151 152Robert E. Ford, et al., "Fear of death of those in a high stress occupation," Psychological Reports 29, 1971, 502. Reynolds and Kalish, op. cit. Rhudick and Dibner, op. cit. 153David Lester, "Attitudes toward death held by staff of a suicide prevention center," Psychological Reports 28, 1971, 650. 154Klas G. Magni, "Reactions to death stimuli amon theology students," Journal for the Scientific Study of Religion 9 3), 1970, 247-248. 57 pictures of corpses faster than those who did not plan on serving a parish. Eiiot‘55 classified the attitudes of soldiers in training during World War II and of men in battle; these ranged from realistic estimates of the chances of survival, to fatalistic acceptance of the probability that one will die, to various magical attempts to control fate. ‘56 in an important study, examined the pro- Reynolds and Kalish, blems and coping strategies of people in death-related occupations: funeral directors, deputy coroners, and terminal ward personnel. They discuss the roles these people fulfill, the role conflicts they experi- ence, the attractions of that type of job, and the differences between public and intra-professional presentations of job and self. Attitudes reported by subjects include: the perception that death is an ordinary and expected event, fatalism about one's own death, humor and cynicism serving a defensive purpose, and a view of self as more realistic and objective about death than one's customers. The extent to which a given death is upsetting to people in these fields seems to have to do with whether there has been a prior nonprofessional relationship with the deceased, the cause of death, whether death was sudden or expected, the life situation of the deceased (age, social status) and other prior ex- periences of the person involved. They illustrate the complexity of the task of determining what factors are related to death attitudes: 155Thomas 0. Eliot, ". . . of the shadow of death," Annals of the American Academy of Political and Social Science 229, 1943, 87-99. 156David K. Reynolds and Richard A. Kalish, "Work roles in death- related occupations," Journal of Vocational Behavior 4, 1974, 223-235. 58 These persons carry with them not only characteristic psychological dilemmas (revolving around role conflicts and the anxiety associated with reminders of one's own mortality), but also incorporate unique social problems of stigmatiza- tions, and the difficulties of interacting with emotionally upset clients. Living situation. Three researchers have examined the relation- ship between the kind of place an elderly person lives in that his or her attitudes toward death. Shrut158 looked at residents of an old-age home which contained both an institutional set-up, and facilities for apartment living. Those who lived in apartments were less afraid of death than those who lived in the institutional facility, who were more dependent on the institution for everything they needed and who lived in an environment which was quite dissimilar to that of their previous place of residence. The apartment dwellers retained more independence and lived in a more familiar environment. Swenson159’160 found that positive, ferward-looking attitudes toward death were associated with living in an old-age home, rather than some other environment. Fear of death was associated with living alone. Roberts et al.16] interviewed 57 nursing home inmates, and fbund that only nine reported any fear of death, while eleven said they looked forward to dying. Thirty-five percent said they never or almost never think about death and dying, 23 percent said they often do, and 5 percent 1571bid., p. 224. 158Samuel D. Shrut, "Attitudes toward old age and death," Mental Hygiene 42, 1958, 259-266. 159Swenson, op. cit., 1958. 160 16lJean Roberts et al., "How aged in nursing homes view dying and death," Geriatrics 25 (4), 1970, 115-119. Swenson, op. cit., 1961. 59 said they think about death all the time. Feifel and Branscomb162 found no relationship between fear of death and whether the individual resided in an institution, but in a study by Templer and Ruff163 institutionalized psychiatric patients scored higher in death anxiety than noninstitutionalized normals. It may be that living in an institution, particularly an unpleasant one, inclines one to look more positively at anything, even death, which would take one out of that setting. 164 Miscellaneous. Hooper and Spilka found no relationship between positive or negative attitudes towards death, and marital status, whether the person was employed or not, age, sex, parental in- come, political preference, dogmatism, college grade point average, or tendency to respond in a socially desirable manner. Feifel and Brans- comb165 found no relationship between conscious fear of death and educa- tional level, socioeconomic status, personal nearness to death, sex, marital status, number of children, or institutionalization. In a study 166’167 active avoidance of the idea of death, in a group by Swenson, of elderly people, was associated with lower educational level, engage- ment in a large number of activities, and good health. No relationship was found between death attitudes and age, sex, socioeconomic status, or living in a rural vs. an urban environment. 162Feifel and Branscomb, op. cit. 163 165 Templer and Ruff, op. cit. Hooper and Spilka, op. cit. 166Swenson, op. cit., 1958. 167Swenson, op. cit., 1961. 60 168 Kalish and Reynolds found no differences between widows and 169 nonwidows in death attitudes, and Lester found that place in a sib- 170 ship had no influence on fear of death. Blatt and Quinlan reported that students who were punctual in fulfilling a course requirement had significantly higher death concern scores than those who fulfilled it late. Alexander and Lester171 looked for differences among parachute jumpers, and found only that the less experienced jumpers showed a 172 greater fear of death. Shilder examined the death attitudes of convicted murders, but drew no general conclusions. In two studies, the relationship between death attitudes and 173 other attitudes has been examined. Lester reported no association, in a population of college students, between attitudes towards suicide ‘74 found that the fear of death was correlated and towards death. Kalish significantly negatively with approval of abortion (in 1963, before safe abortions became available). There was no signficant association with 168Richard A. Kalish and David K. Reynolds, "Widows view death: A brief research note," Omega 5 (2), 1974, 187-192. 169David Lester, "Studies on death-attitude scales," Psycholo- gical Reports 24, 1969, 182. 170Sidney J. Blatt and Paul Quinlan, "Punctual and procrastina- ting students: A study of temporal parameters," Journal of Consulting Psychology 31 (2), 1967, 169-174. 171Michelle Alexander and David Lester, "Fear of death in parachute jumpers," Perceptual and Motor Skills 34 (l), 1972, 338. 172Paul Shilder, "The attitudes of murderers towards death,“ Journal of Abnormal and Social Psyohology_3l, 1936, 348-363. 173David Lester, "Attitudes toward death and suicide in a non- disturbed population," Psychological Reports 29, 1971, 386. 174Richard A. Kalish, "Some variables in death attitudes," Journal of Social Psychology 59, 1963, 137-145. 61 approval of birth control, euthanasia, war-time killing, or capital punishment. Attitudes of Medical Personnel Although there is no reason to assume a priori that medical professionals will be different in their attitudes towards death than the general public, it is relevant to review separately those studies which have focused on people in the medical field, since they form the sub- ject population of interest in the current study. It is conventionally thought that physicians enter the medical profession partly because they are more afraid of death than most people, or because they cannot accept the reality of death. This point of view is well-expressed by Livingston and Zimet: (The almost dead) . . . are particularly disturbing for the physician; deSpite his sophisticated understanding of the gross limitations of medern medicine, each dying patient represents a failure and a disappointment. The physician would like to believe himself omnipotent, an intrepid healer with access to many life-sustaining techniques, fluids, powders, and pi11s.175 The research evidence supporting or contradicting this conventional wisdom is reviewed here, and in this study an attempt is made to compare the attitudes of medical students with those of students in other fields. Caldwell and Mishara approached 73 hospital doctors for a brief interview on their attitudes towards death, and were only able to complete interviews with thirteen of them. Many others originally consented to be interviewed, but refused upon learning the subject of it, while others refused an interview on any subject. 175Livingston and Zimet, op. cit. 62 The most typical reason fer not consenting to the inter- view was that the S felt he would be a less effective doctor if he concerned himself with feelings involved in this area, since he felt his own emotional reSponse would inter- fere with his effective practice of medicine.”6 Craniond‘77 points out that the patient is an important object for the physicians, and that interaction with the patient may tap into the phy- sician's unconscious narcissism or omnipotent fantasies. Feelings which the doctor has may interfere with his or her interaction with patients, or may be overly-controlled by the individual out of the fear that those feelings may make him or her a less effective practitioner. It must not, however, be assumed that the physician's beliefs or attitudes toward? death make him or her either more or less effective in interactions with patients, without evidence to support this. Degner178 fbund that whether a doctor believedianod or in an afterlife made no differences in whether s/he would elect to use life-prolonging measures in simulated situations. The majority favored withdrawing life support from terminally ill patients in two out of three cases presented; there was no difference in treatment proposed for patients of differing pre- sumed social status. It may be that while medical treatment of patients would not differ in physicians with differing attitudes about death, interactions with the patient might differ considerably. It is this possibility that this study is designed to examine. 176Diane Caldwell and Brian L. Mishara, "Research on attitudes of medical doctors toward the dying patient: A methodological problem," Omega 3 (4), 1972, 341-346. 1”N. A. Cramond, "Psychotherapy of the dying patient," British Medical Journal 3 (5719), 1970, 389-393. 178Leslie Degner, "The relationship between some beliefs held by physicians and their life-prolonging decisions," Omega 5 (3), 1974, 223-232. 63 Kram and Caldwell179 compared psychiatrists, other physicians, Jewish and Protestant clergyman, and attorneys in respect of their views on the treatment of dying patients. Large majorities of all groups said that they would want to know if they themselves were dying, that the patient should be told of his or her impending death, and that it is primarily the physician's responsibility to tell the patient that s/he is dying. The majority also believed that it was possible that the patient might deny the reality of approaching death, and that s/he would do that mainly fer his or her own comfbrt, then for the comfbrt of the family, and only lastly to protect the physician from uncomfortable feelings. (This result might indicate a lack of knowledge on the part of these professionals who deal with the dying of the social realities the patient has to face or of the force of intimate and professional relations on the behavior and feelings of the patient.) Feifel et al.180 compared groups of internists, surgeons, and psychiatrists, with medical students, seriously and terminally ill patients, and healthy normal individuals, on a variety of attitudes to- wards death. (In all groups, subjects were above average in intelli- gence and socioeconomic status; most were male, Protestant, and married with children.) Significantly more physicians said they first became aware of the existence of death between the ages of l and 5, although in all 179Charles Kram and John M. Caldwell, "The dying patient," Psychosomatics lO (5), 1969, 293-295. 180Herman Feifel et al., "Physicians consider death," Proceed- ipgs, 75th Annual Convention of the American Psychological Association, 1967, PP. 201-202. 64 groups, most individuals said this first happened to them between the ages of 6 and 12. Physicians showed significantly more negative ver- bal death imagery than patients, and were unable to come up with any imagery in this area more often than the normal group. Physicians were less religious and had a less religious orientation toward personal fate after death than either patients or normals. The overwhelming majority of physicians said they would want to be informed if they had an incurable disease, but indicated significantly less willingness than patients to provide that information to someone else in that situa- tion. PSYchiatrists expressed more anxiety about having to tell a patient s/he was dying than other physicians. Surgeons said they less often thought about death than the other two groups--rarely, as opposed to occasionally. Medical students generally were a middle group: they were more afraid of death than patients and healthy nor- mals, but less so than physicians. They were more religious than phy- sicians, less so than others. They admitted to thinking about death significantly less often than all other groups, and significantly more medical students than patients said they would prefer to die at night, so as to be unaware that it was happening. In the only other study of medical student attitudes, male students who scored high on the California F Scale (authoritarianism) were compared with nonauthoritarian students by Livingston and Zimet.181 Those who indicated a preference for specialization in psychiatry were significantly lower in authoritarianism and higher in death anxiety than those who chose surgery, medicine, or pediatrics. Future surgeons were 181Livingston and Zimet, op. cit. 65 less death-anxious than the other three groups. There was in all groups a significant negative correlation between death anxiety and authori- tarianism. The theory was that authoritarian medical students are better defended against unconscious processes in general and therefore against unconscious fear of death. Less authoritarian students should have fewer or less rigid defenses against unconscious death anxiety and would therefbre be likely to choose a specialty in which their contact with death would be less, such as psychiatry. This prediction was born out by the results. Death anxiety was significantly higher in third and fourth year students, although authoritarianism did not change over time, thus indi- cating that experience has some influence on the fear of death indepen- dent of this personality characteristic. Psychiatry (low authoritarian- ism/high death anxiety) and surgery (high authoritarianism/low death anxiety) were seldom chosen as second-choice specialties; both were either the person's top choice or very low on their rank ordering of ' preferences, additional evidence that fear of death may influence the choices a medical student makes of Specialties. Studies of the attitudes of nurses and nursing students are also relevant. Although future nurses and physicians are being pre- pared for very different roles in medical institutions, there is suf- ficient overlap in the kinds of problems they have to face, since both will have to interact face to face with dying patients, that, while it is not desirable to generalize too freely from one group to the other, there may be some similarities. 66 Snyder et a1.182 asked freshman and senior nursing students about the frequency with which they had various thoughts about death. The freshmen were at the beginning of their training, while the seniors had completed a course of study which included lecture, group experi- ences, and clinical experience of death. Seniors reported signifi- cantly less frequent thoughts and dreams about, and wishes for, death of self, friends, or relatives. There was no difference in the reported frequency of reading death-related material or in the frequency with which they experienced a strong fear of death. Subjects in this group were almost all female, working class, Catholic, urban, and single, and ranged in age from 18 to 24. This description of the subject population points up another reason for being careful of generalizing from one group of medical personnel to another: the characteristics of this group of nursing students are very different from those of a typical group of medical students, which is more likely to be pre- dominantly male, with more members from upper classes, of greater re- ligious diversity, and older. Lester et al.183 found that, in a group of nursing under- graduate and graduate students and faculty, fear of death decreased with educational level (and thus with age). The only exception to this general trend was that first year graduate students (who had spent a number of years in the nursing field before returning to school) scored higher than senior students. The intervening years of clinical 182M. Snyder, et al., "Changes in nursing students attitudes toward death and dying: a measurement of curriculum inte ration effec- tiveness," International JOurnal of Social Psychiatry_l9 3-4), 1973, 294-298. ‘ 183David Lester, et al., "Attitudes of nursing students and nursing faculty toward death," Nursing Research 23 (l), 1974, 50-53. 67 experience, away from the academic environment, may have led to in- creased fears of death due to increased contact with dying patients. All groups indicated no fear of the dying of others, and the extent to which it was not feared increased with the level of experience. Although psychiatric nurses tended overall to have higher scores than those in other specialties, there were no significant differences on any measures used, among those in different nursing specialties. These two studies indicate a tendency for nurses to become less preoccupied and/or less fearful of death with years of training, a tendency opposed to that found among medical students. It is not clear whether there is a difference in attitudes towards death which those in the two groups bring with them to their training, in how they handle their feelings about death, or in their training which increases fear in medical students and decreases it in nursing students. There are a number of differences between students in the two fields which would make comparison in this area harder. Golub and Reznikoff184 examined differences between nursing students and graduate nurses, in attitudes related to the care of dying patients, which did not differ among various nursing specialties even though those specialties differed in their amount of contact with death. Student nurses more often said that all possible efforts should be made to keep a seriously ill person alive, although in both groups the most popular response was that "reasonable effort" should be made. Graduates were more often approving of an autopsy being 184Sharon Golub and Marvin Reznikoff, "Attitudes toward death: a comparison of nursing students and graduate nurses," Nursing Re- search 20 (6), 1971, 503-508. 68 performed on themselves, while students more often said they did not care whether one was done or not. Graduate nurses more often said that they believed that psychological factors could influence or cause death than students, but it was mainly the younger and less experienced nurses who held that opinion. Students most often said they did not know whether this could happen or not. This is the extent of research into attitudes of medical professionals and students about death. There is very little in the way of clear results indicating who fears or does not fear death. This facet of death attitudes has not been examined, except in a few sociological studies of institutions which deal with dying patients. Fear of Death Measures The fear of death, both conscious and unconscious, has been measured in a variety of ways. The most commonly used measures are summarizedin this section, and those which are most similar to the one used here are described in detail. One approach to below-conscious fear of death has to do with .185 perceptual sensitization or defense. Magn1 showed subjects tachis- toscopically presented pictures of corpses, but found that they were not identified consistently more rapidly or more slowly than neutral 186 pictures. Golding et al. found that subjects took longer to iden- tify death-words than neutral words, but there was no relationship to 185Klas G. Magni, “Reactions to death stimuli among theology students," Journal for the Scientific Study of Religion 9 (3), 1970, 247-248.‘ 186Stephen L. Golding, George E. Atwood, and Richard A. Goodman, "Anxiety and two cognitive forms of resistance to the idea of death," Psychological Reports 18, 1966, 359-364. 69 187 self-reported fear of death. Lester and Lester gave carbon copies, in varying degrees of clarity, of the same lists of words, and found that subjects identified death-words at a lesser level of clarity than neutral words, suggesting that it is in the survival interest of the organism to recognize such potentially dangerous stimuli as rapidly as 188 possible. Alexander and Adlerstein, and Alexander, Colley, and 189 Adlerstein, on the other hand, found that subjects took longer to associate to death-words than to neutral words, but not longer than to other affectively laden words. Both they and Williams and 190 Cole found greater GSR reactivity to death than to neutral words in an association task. In the use of word association tasks, Alexander and Adler- 191 192 stein, Alexander, Colley and Adlerstein and Feifel et a1.193 looked at the length of time it takes subjects to respond to death words and neutral words. The latter authors also used the Color Word 187Gene Lester and David Lester, "The fear of death, the fear of dying, and threshold differences for death words and neutral words," Omega 1, 1970, 175-179. 188Irving E. Alexander, and Arthur M. Adlerstein, "Affective responses to the concept of death in a population of children and early adolescents," Journal of Genetic Psyohology_93, 1958, 167-177. 189Irving E. Alexander, Randolph S. Colley, and Arthur M. Adlerstein, "Is death a matter of indifference?", Journal of Psychology 43, 1957, 277-283. 190Robert L. Williams, .and Spurgeon Cole, "Religiosity, generalized anxiety, and apprehension concerning death," Journal of Social Psychology 75, 1968, lll-117. 191 Alexander and Adlerstein, op. cit. 192Alexander, Colley and Adlerstein, op. cit. 193Herman Feifel, Jeffrey Frelich, and Lawrence J. Hermann, "Death fear in dying heart and cancer patients," Journal of Psychoso- matic Research 17, 1973, 161-166. 70 Interference Test, and fbund that subjects took longer to identify the color words were printed in, when the word was a death-related word, than when it was a neutral word. In an attempt to find a projective test of unconscious death 194 195 attitudes, Lowry and Rhudick and Dibner asked subjects to make up stories about death to TAT cards, and examined the themes in those 1.196 stories. Bruhn et a measured systolic blood pressure in patients whose hospital roommate had just died. Most research in the area of fear of death has consisted of construction and administration of self-report questionnaires, and of 197 interviews. Jeffers et a1. asked the direct question: are you afraid to die, which is also used in this current study. Questionnaires covering a broad range of death attitudes have 198 199 been used by Adlerstein, Faunce and Fulton, 200 Diggory and Rothman, 194Richard J. Lowry, Male-Female Differences in Attitudes towards Death, Ph.D. Dissertation, Brandeis UfiiverSity, 1965. 195Paul J. Rhudick, and Andrew S. Dibner. "A99, personality, 60d health correlates of death concerns in normal aged individuals," Journal of Gerontolegy l6 (1), 1961, 44-49. 196John G. Bruhn, A. Eugene Thurman, Betty C. Chandler, and Thomas A. Bruce, "Patients' reactions to death in a coronary care unit," Journal of Psyohosomatic Research 14 (l), 1970, 65-70. 197Frances C. Jeffers, Claude R. Nichols, and Carl Eisdorfer, "Attitudes of older persons toward death: a preliminary study," Journal of Gerontology 16 (l), 1961, 53-56. 198Arthur M. Adlerstein, The relationship between religious belief and death affect, Ph.D. Dissertation, Princeton University,51958. 199 200 Faunce and Fulton, op. cit. Diggory and Rothman, op. cit. 71 201 202, 203 204 205 Feifel et al., and Middleton. 206 Lester, Lester and Kam, Feifel tasked subjects about the attitudes they thought others held, on the assumption that the attitudes people attribute to others are 207 208 likely to be similar to their own. Adlerstein and Lester used 209 a semantic differential technique. Maurer asked adolescent subjects to write essays on the subject of death. Open-ended interviews have been employed by Reynolds and 210 211 Kalidi, with a population of death professionals, Bouton with 212 medical students, Eliot with soldiers during World War II, 2O'Feifei, Frelich and Hermann. op. cit- 202David Lester, "Attitudes toward death today and thirty-five years ago," Omega 2, 1971, 168-173. 203David Lester, "Sex differences in attitudes toward death: a replication," Psyohological Reports 28, 1971, 754. 204 Lester and Kam, op. cit. 205Warren C. Middleton, "Some reactions toward death among college students," Journal of Abnormal and Social Psyohology 31, 1936, 165-173. 206Herman Feifel, "Attitudes of mentally ill patients toward death," Journal of Nervous and Mental Disease 122, 1955, 375-380. 207Adlerstein, op. cit. 208David Lester, "Studies in death attitudes: Part two," Psychological Reports 30, 1972, 440. 209Adah Maurer, "Adolescent attitudes toward death," Journal of Genetic Psychology_lOS, 1964, 75-90. 210David K. Reynolds, and Richard A. Kalish, "Work roles in death-related occupations," Journal of Vocational Behavior 4, 1974, 223-235. 2”David Bouton, "The need for including instruction on death and dying in the medical curriculum, JOurnal of Medical Education 47 (3), 1972, 169-175. 212Thomas 0. Eliot, ". . . of the Shadow of death," Annals of the American Academy of Political and Social Science 229, 1943, 87-99T5 72 213 214 215 Caldwell and Mishara, Roberts et a1.,216 and Shiider.2'7 Feifel, Pearlman et al., Fear of Death Inventories Several fear of death inventories which have not been widely used by other than their authors will only be mentioned here; their 218 findings have already been reviewed. Dickstein and Blatt developed a death concern and preoccupation scale, which has been used by 219 220 221 222 Dickstein, Blatt and Quinlan, and Selvey. Sarnoff and Corwin 213Diane Caldwell and Brian L. Mishara, "Research on attitudes of medical doctors toward the dying patient: a methodological problem," Omega 3 (4), 1972, 341-346. 214Herman Feifel, "Attitudes toward death in some normal and mentally ill populations," in Herman Feifel, ed., The Meaning of Death (New York: McGraw-Hill, 1959), 114-130. 215Joel Pearlman, Bernard A. Stotsky, and Joan R. Dominick, "Attitudes toward death among nursing home personnel," Journal of Gene- tic Psychology_ll4, 1969, 63-75. 216Jean L. Roberts, Larry R. Kimsey, Daniel L. Logan and Gordon Shaw, "How aged in nursing homes view dying and death," Geriatrics 25 (4). 1970. 115-119. 217Paul Shilder, "The attitudes of murderers towards death," Journal of Abnormal and Social Psychology 31, 1936, 348-363. 218Louis S. Dickstein and Sidney J. Blatt, "Death concern, futurity, and anticipation," Journal of Consulting Psychology 30, 1966, 11-17. 219Louis S. Dickstein, "Death concern: measurement and corre- lates," Psychological Reports, 1972, 563-571. 220Sidney J. Blatt and Paul Quinlan, "Punctual and procrasti- nating sutdents: a study of temporal parameters," Journal of Consul- ting Psyohology 31 (2), 1967, 169-174. 221 222Irving Sarnoff and Seth M. Corwin, "Castration anxiety and the fear of death,“ J0urnal of Personality 27, 1959, 374-385. Selvey, op. cit. 73 developed a scale based on the theoretical notion that death anxiety 223 stems from unconscious castration anxiety. Golding et al. used this scale in looking for a relationship to tachistoscopically presented death stimuli. 224 Livingston and Zimet developed a scale for use with medical students, investigating the relationship of fear of death to authori- 225 tarianism. This scale was also used by Tolor and Reznikoff with the result that sensitizers and externals scored higher than repressors and internals. Handal and Rychlak226 227 228 used a scale developed by Handal. Boyar, on the theory that the fear of death is not unitary in nature, developed a multi-factor fear of death scale. There were no significant correlations between scores on this instrument and demographic variables, although subjects who had experienced a death in their family had significantly higher scores than those who had not. There were low but significant correlations between fear of death and scores on the K, At, and Pt scales of the MMPI. Attempting to use this instrument to measure change in the fear of death after presenta- tion of a fear-arousing stimulus, Boyar found that the scale failed to retain internal consistency, and he speculates that there are aspects 223Go1ding, Atwood and Goodman, op. cit. 224Livingston and Zimet, op. cit. 225 226 Tolor and Reznikoff, op. cit. Handal and Rychlak, op. cit. 227Paul J. Handal, "The relationship between subjective life expectancy, death anxiety, and general anxiety," Jaurnal of Clinical Psychology 25 (l), 1969, 39- 42. 228Boyar, op. cit. 74 229 of fear of death which are not tapped by that scale. Selvey found that females scored significantly higher than males on Boyar's scale, 230 found a difference not found in Boyar's original research. Lester no association between high frequency of dreams or nightmares with scores on Boyar's scale. 23' developed a widely used Death Anxiety Scale. which Templer correlates .74 with Boyar's Fear of Death Scale. This 15-item scale was validated on populations of college students and psychiatric patients. Death Anxiety Scale scores were not correlated with responses to the Marlowe-Crowne Social Desirability Scale, nor with the Couch- Keniston scale of agreeing response tendency, but were significantly correlated with scores on the Manifest Anxiety Scale and the Welsh Anxiety Scale (but not as highly as those MMPI scales of general anxiety correlated among themselves). In addition, among the psychia- tric patients tested, there was a significant correlation with the Sc. Pt, and D scales of the MMPI, but none with any other MMPI scales. Templer asked his college student subjects to give ten associa- tions to both 'death' and nondeath-related words. Those with higher scores on the Death Anxiety Scale gave significantly more words des- criptive of emotions in association to 'death.‘ Templer hypothesizes that the appearance of emotionally-laden words is an indicator of death anxiety. This construct is further discussed in Chapter III. 229Selvey, op. cit. 230David Lester, “The fear of death of those who have night- mares," Journal of Psychology 69, 1968, 245-247. 231Donald 1. Templer, The construction and validation of a death anxiety scale, Ph.D. Dissertation, University of’Kentucky, 1967} 75 In other research on the Death Anxiety Scale, Templer and Ruff232 found that psychiatric patients scored higher than nonpatients, 233 and women scored higher than men. Ray and Najman and Templer and Ruff234 also found that women scored higher than men, and in the latter study, a significant correlation was found between scores of 235 parents and their children, and between spouses. Templer found that very religious subjects scored lower on the average than those who saw themselves as less religious, although there was no relationship between fear of death score and either religious affiliation or change 236 in affiliation. Templer found a positive relationship between DAS scores and GSR ratings of anxiety during a death-related experiment, and found that sensitizers scored higher than repressors. 237 Templer found that persons who reported a large number of psychiatric symptoms scored higher, while those who reported a larger number of physical symptoms tended to have lower scores; however, 238 Lucas feund no relationship between health and DAS score, and no sex 232Donald I. Templer and Carol F. Ruff, "Death Anxiety Scale gegns, standard deviations, and embedding,"__sychological Reports 29 l . 1971, 173. 233 234 235Donald I. Templer, "Death Anxiety in religiously very involved persons," isychological Reports 31, 1972, 361-362. 236Donald I. Templer, "The relationship between verbalized and nonverbalized death anxiety," Journal of Genetic Psychology_ll9, 1971, 211-214. ' 237Donald I. Tenpler, Death anxiety as related to depression and health of older persons," J0urnal of Gerontology 26 (4), 1971, 521-523. “ 238 Ray and Najman, op. cit. Templer and Ruff, op. cit. Lucas, op. cit. 76 239 differences in scores. Templer and Dotson feund no relationship with either sex or religion, and Ray and Najman240 found no relationship with authoritarianism (denial) or achievement motivation. The items of the Death Anxiety Scale are embedded in items of the MMPI, facilitating investigation of relationships with various 241 MMPI scales. Templer found a positive correlation with the D scale, and Templer and Lester242 found significant relationships between OAS scales. and individual MMPI items related to health and introversion. Lester243 developed a questionnaire scored on a six-point scale from Strongly Disagree to Strongly Agree. Although this instru- ment itself has not been extensively studied, much of the research on the instrument devised by Collett and Lester (which is discussed in the next section) is relevant to the Lester scale, since both scales are comprised of essentially the same items. Lester244 found that responses to this scale which were returned incomplete were not significantly different than those which were returned completely filled out. Also, there was no significant association between consistency of attitude 239Templer and Dotson, op. cit. 240 241 242Donald 1. Templer and David Lester, "An MMPI scale for assessing death anxiety," Psychological Reports 34, 1974, 238. 243David Lester "The construction of a fear of death scale: its consistency, validity and use," unpublished manuscript, Brandeis University, 1966. 244David Lester, "Studies on death attitude scales," Psycholo- gical Reports 24, 1969, 182. ' Ray and Najman, op. cit. Templer, op. cit. 77 and strength of fear. Durlak245 found no relationship between scores on the Lester scale and the social desirability rating of the response. Scores on the Lester scale have not been found to be related to whether the subject had ever experienced the possibility of imminent death, the death of a close other, or the recent thought of his or her own death, nor to the subject's estimate of either the probability of her or his own death in the next year or of the frequency with which 46 247 found no association of scores 248 s/he thinks about death.2 Lester with frequency of dreams or nightmares, nor with occupation. 249 Durlak found that those who expressed a high sense of meaning and purpose in their lives scored lower on the Lester scale, while Lester250 found that those who were less religious were more consistent in their attitudes towards death, though expressing more fear. Collett and Lester Scale. The scale used in this study, devised 251 by Collett and Lester is a derivative of Lester's scale. Items of 245Joseph A. Durlak, "Relationship between various measures of death concern and fear of death," Journal of Consulting and Clinical Psychology_4l (1), 1973, 162. 246 247 op. cit. 248David Lester, "Attitudes toward death held by staff of a suicide prevention center," Psychological Reports 28, 1971, 650. 249Joseph A. Durlak, "Relationship between individual attitudes toward life and death," Journal of Consulting and Clinical Psychology_ 38 (3), 1972, 463. 250David Lester, "Religious behavior and the fear of death," Omega 1, 1970, 181-188. 25ILora-Jean Collett, and David Lester, "The fear of death and the fear of dying," The Journal Of‘PsycholOgy_72, 1969, 179-181. Joseph Durlak, Ibidu David Lester, "The fear of death of those who have nightmares," 78 the Lester scale were categorized into subscales on the basis of inter- item correlations. The theoretical construct underlying this method of construction is that, while different kinds of fear of death are related, the fear of death is not a unitary phenomenon. The scales are constructed to measure fear of death of self, death of others, dying of self, and dying of others. Items of the Lester scale which did not achieve a satisfactory level of correlation with any subscale were dropped from the final version of the test. '(Items of the Collett- Lester inventory are found in Appendix A.) Durlak investigated the concurrent validity of the fear of death scales developed by Sarnoff and Corwin, Boyar, Tolor, and Lester. The correlations between these four tests ranged from .40 to .65, (p.5 .01). Each test was also correlated with each of the four sub- scales of the Collett-Lester inventory. All of the above inventories correlated most highly (range: .47-.78) with the Death of Self (DS) subscale. Ranges of correlations with the other three subscales were: Dying of Self’flwS), .46 to .58; Death of Others (00), .31 to .46; and Dying of Others (DyO), .36 to .40. The results for males and females were similar. Durlak concludes from these data that the inventories under examination "measure attitudes toward personal death and dying, rather than generalized fears and feelings about death."252 The DO and DyO subscales of the Collett-Lester, he believes, measure this general fear of death, even though all the correlations between these subscales 252Joseph A. Durlak, "Measurement of the fear of death: an examination of some existing scales," Journal of Clinical Psychology 28 (4), 1972, 545-547, p. 547. ‘ 79 and the other inventories were significant. He concludes that: If C-L is accepted on a face validity basis, the data indicate that the death scales are relatively stronger and better measures of personal fears and anxieties about death and dying (i.e., when the self is the referent) than they are measures of generalized fears or anxieties about death (i.e., when the other is the referent). 254 fbund low correlations between the Collett and Lester four subscales (in 2 samples of 25 undergraduate women each) ranging from .03 between DyS and 00, to .58 (in one sample) between 05 and DyO. Subscale intercorrelations are shown in Table 2.1. Table 2.1 Intercorrelations Between Collett-Lester Subscales . Sample 1 Sample 2 Subscale Comparisons (n = 25) (n = 25) Death of Self - Death of Others .22 .26 Death of Self - Dying of Self .24 .41* Death of Self - Dying of Others .09 .58* Death of Others - Dying of Self .03 -.07 Death of Others - Dying of Others .46* .40* Death of Self - Dying of Others .28 .40* * = significant at .05 level (two-tailed test) Source: Collett and Lester, 1969, p. 180. All but 6 items of Lester's scale correlated at least .26 (p §_.10) with the test as a whole. Those six items were dropped from the Collett-Lester version of the test in order to improve the internal consistency of the instrument. 253Ibid.. p. 547. 254Collett and Lester, op. cit. 80 The table above reveals discrepancies between the two samples in the correlations between DyO and OS, DyO and DyS, and between DS and DyS, and no information is given about the samples which might explain the differences. Collett and Lester take the low to moderate inter-scale correlations as evidence that the four kinds of fear are distinguishable, but they do not attempt to account for the signifi- cant correlations which were found. There are several possible explanations for the differences in result between the samples: 1. Demographic differences between the samples. Although the evidence on the effects of demographic variables on the fear of, death is conflicting, the studies which are reviewed in this chapter indicate at least that such differences cannot be discounted in attempting to explain differences among groups in amount and kind of fear of death. 2. Other affect which is stronger than fear. Studies dis- cussed in this chapter have shown that recent loss may have an effect on the fear of death. One possible explanation of this phenomenon is that the effect of a recent loss is to create feelings about death which are different from fear, and which are experienced more strongly than fear, to that whatever fear does exist does not get expressed as readily. 3. The stimulus value of the questions comprising the inven- tory may not be high enough to elicit what fear does exist, or may elicit it from some individuals and not others. Collett and Lester report that subjects showed a significantly greater fear when the self was the referent than when the other was (p less than .01). In view 81 of the correlations with other fear of death measures reported by Durlak, it may be that this difference reflects a real difference in the strength of the fear. However, it may also mean that the questions asked are not an adequate stimulus for eliciting real fear of death and dying of others. 255 used the Collett and Lester inven- Shusterman and Sechrest tory with a group of registered nurses, to examine the relationship of measured fear of death and dying of others to nurses expressed satisfac- tion and comfort with their role in caring for dying patients. As in this current study, they selected this inventory because it looks at the relationship between fear of death and dying of self and that of others, even though the reliability and validity of some other scales are better established. They found no significant differences in mean scores on each subscale, between nurses working on different hOSpital services. Age and amount of experience as a nurse were nega- tively correlated with fear of death of others; there were no corre- lations between scores on any of the subscales and any personality variables. Nurses' satisfaction with the standard care given to dying patients was unrelated to any aspect of fear of death. Correlations between the subscales were computed, and only four significant inter-scale correlations were fbund. Death of self and death of others correlated .42; death of self and dying of self, .24; death of others and dying of self, .25; and dying of others and dying of self, .24 (all p's less than .05). There were no other 255Lisa Shusterman, and Lee Sechrest, "Attitudes of registered nurses toward death in a general hospital," Psychiatry_in Medicine 4 (4). 1973. 411-425. 82 significant intercorrelations among the scales, suggesting that the nursing sample was more inconsistent in their attitudes, or that different fears were better differentiated in the nursing sample than in the college student samples on which the inventory had previously been used. In addition, many items of the fear of dying of self scale did not correlate well with that subscale when used with this sample. Lester256 fbund no significant differences in any subscale scores among Catholics, Protestants and Jews. Less religious subjects showed a greater fear for self than for others, while more religious subjects feared Significantly more for others than for self. Less religious subjects Showed a higher fear of dying of self than the more religious. Lester257 feund that scores bore no relationship to the sub— jecthsage. There were no sex differences in fear of dying of others or overall fear of death, but females scored higher than males in fear of death of self, dying of others, and death of others. Lester258 also found that scores on the inventory decreased with educational level in a group of nurses and nursing students, the same result found by Shusterman and Sechrest. Other studies which have used this inventory have either focused on papulations very different from the one used in this study, or have examined variables which are not of interest here. Those studies are reviewed in other sections of this chapter. 256Lester, op. cit. 257David Lester, "Studies in death attitudes: part two," op. cit. 258David Lester, Cathleen Getty, and Carol Ren Kneisl, "Attitudes of nursing students and nursing faculty toward death," Nursing Research 23 (l), 1974, 50-53. 83 The present research is designed to investigate more fully the reliability of the scales of this instrument, and to look for associa- tions of scores with a variety of demographic variables (sex, age, religious preference, religiosity, etc.) with which previous research indicates relationships may exist. A much larger sample is used than has been used in any previous study, and students in different fields at different levels are compared, particularly for the purpose of looking for differences between medical students and those in nonmedical fields. CHAPTER III In this chapter, the three different samples are described, followed by a description of the construction and administration of the questionnaire, and the procedures involved in the simulation study and interview. Reliability data on the Collett-Lester fear of death instru- ment are given. Collection of Questionnaire Data Sample size of at least 100 was desired in each of the three groups studied: undergraduates, graduate students in education, and medical students. Since this sample size was achieved with ease, only those questionnaires which were completely filled out were included in the analysis. Lester1 showed that responses to the Collett-Lester in- ventory which were received incomplete did not differ significantly from those which were entirely completed; to facilitate analysis of the data, incomplete questionnaires were eliminated from the sample. Ten unusable questionnaires were received from undergraduates, six from graduate stu- dents, and two from medical students; in addition, 14 non-undergraduates were eliminated from the undergraduate sample (probably discussion sec- tion instructors who also filled out questionnaires) and one undergraduate was eliminated from the graduate student sample. One exception to the rule of eliminating incomplete responses was made: it was discovered, 1David Lester, "Studies on death-attitude scales," Psychological Reports, 1969, 24, 182. 84 85 after the interview had been done, that the questionnaire returned from one interviewee had omitted a response to one demographic item. Rather than lose a whole interview, it was decided to include that subject even though an item had been omitted, since it did not appear that this omis- sion would greatly affect the characteristics of the sample as a whole. Since the questionnaire data was collected in slightly different manner in each of the three samples, the description of the manner in which the data was collected is included in the description of the characteristics of each sample. The Samples Undergraduates.- The undergraduate sample consisted of 172 under- graduates enrolled in Education 200 (a class of several hundred students, meeting in small discussion sections) at Michigan State University, during fall quarter, 1975. All discussion section instructors were asked to distribute the questionnaire in class; enough agreed so that the desired sample size (100 or more) was achieved. There was no evidence that stu- dents whose instructors were unwilling to distribute the questionnaire differed in any way from those whose instructors agreed to do so. Some students completed the questionnaire during class; others took it home and returned it to their instructor. It was stressed to all potential subjects that participation was entirely voluntary. In the undergraduate group, there were 45 males (26.2%) and 127 females (73.8%), probably a reflection of the fact that female students predominate in education classes. The majority (65.7%) were under twenty years old, to be expected in a class taken primarily by sophomores, and 83.1% were single. A complete account of the marital status of the group 86 is shown in Table 3.1. (To facilitate comparison among the three groups, characteristics of all three groups are shown together in each table.) Table 3.1 Marital Status of the Three Samples Groups Under Graduate Medical graduates Students Students Marital Status n % n % n % Never Married 143 83.1 21 21.0 64 49.2 Married 13 7.6 67 67.0 48 36.9 Separated or Divorced 2 1.2 6 6.0 10 7.7 Cohabiting (same or opposite sex partner) 14 8.1 5 5.0 8 6.2 Most (82.1%) had no children. Twenty of them (11.6%) still lived with their family of origin, 25 (14.5%) with their spouse or partner and/ or children, five (2.9%) lived alone, and 122 (70.9%) with roommates. Reflecting the fact that the course in which they were enrolled at the time is oriented particularly toward education majors and people earning a teaching certificate, the majority (59.3%) were social science majors, a category which specifically included education. Other cate- gories of academic major represented by sizeable groups were natural sciences (14%) and arts and letters (19.2%). bution of academic majors is shown. In Table 3.2, the distri- 87 Table 3.2 Academic Major Groups Under- Graduate Medical graduates Students Students Major n % n % n % Medicine 4 2.3 129 99.2 Natural Sciences 24 14.0 1 1.0 l 0.8 Social Sciences (in- cluding education) 102 59.3 84 84.0 Arts and Letters 33 19.2 4 4.0 Business 6 3.5 4 4.0 Technical (including agriculture and engineering) 2 No Preference 1 0—1 05“) N Nm CO An account of the religious upbringing and current religious pre- ference of the groups studied is shown in Tables 3.3 and 3.4. Among the undergraduates, the majority were brought up in either a major Protestant denomination (not including Baptists or other fundamentalist groups) (34.9%) or in the Roman Catholic or Orthodox Church (36.6%). Fewer sub- jects currently belonged to either of these groups than were brought up in them, but they were still the largest groups, with 20.3% currently preferring membership in a major Protestant denomination, and 26.7% Catholic or Orthodox. Subjects were also asked how religious they con- sidered themselves, using any definition of 'religious' they felt applied to them personally. The majority rated themselves as "about average (42.4%) or "not very religious." Table 3.5 contains a summary of these religious self-ratings. 88 Table 3.3 Religion of Origin Groups Under- Graduate Medical graduates Students Students Religion n z I n % h 24 Roman Catholic or Orthodox 63 36.6 34 35.0 40 30.8 Major Protestant 60 34.9 40 40.0 42 32.3 Fundamentalist Protestant 16 9.3 11 11.0 16 12.3 Jewish 9 5.2 5 5.0 22 16.9 Eastern or Moslem O 0.0 O 0.0 O 0.0 No Formal Affiliation 9 5.2 3 3.0 l 0.8 No Religious Upbringing 7 4.1 5 5.0 5 3.8 Anti-religious Up- bringing l 0.6 0 0.0 O 0.0 Other 7 4.1 l 1.0 4 3.1 Table 3.4 Current Religious Preference 1 Groups Under- Graduate Medical graduates Students Students Religion ’ n % n %v n i % Roman Catholic or : Orthodox 46 26.7 21 21.0 20 1 15.4 Major Protestant 35 20.3 23 23.0 17 1 13.1 Fundamentalist 2 Protestant 13 7.6 14 14.0 9 : 6.9 Jewish 8 4.7 4 4.0 19 1 14.6 Eastern or Moslem O 0.0 0 0.0 O i 0.0 No Formal Affiliation 29 16.9 18 18.0 21 1 16.2 Agnostic 19 11.0 13 13.0 22 1 16.9 Atheist 6 3.5 4 4.0 7 1 5.3 Other 16 9.3 3 3.0 15 1 11.5 i 89 Table 3.5 Religious Self-rating Defining 'religious" in any way you feel applies to you, how religious would you say you are? Groups Under- Graduate Medical graduates Students Students Rating n % n % n % Very Religious 31 7181.0 I 20 20.05 28 21.5 About Average 73 42.4 37 37.0 52 40.0 Not Very Religious 48 27.9 22 22.0 22 16.9 Non-religious 18 10.5 21 21.0 24 18.5 Anti-religious 2 1.2 0 0.0 4 3.1 In this group, all had grown up with a mother, or mother-substitute, in the home, and in about half the cases (45.3%) she had never worked out- side the home, or had done so only briefly. In 49.4% of the cases, the mother was currently or usually employed outside the home. The largest numbers of those who worked were employed in a professional or business capacity. (See Table 3.6.) All but two subjects in this group had grown up with a father or father-substitute in the home. Most (85.4%) reported that their fathers were currently or usually employed outside the home, in numbers about 90 Table 3.6 Mothers' Empjoyment Groups Under- Graduate Medical graduates Students Students Mother Employed n % n % n % Never or Only Briefly 78 45.3 48 48.0 54 41.5 Currently Employed 73 42.4 31 31.0 57 43.8 Usually Employed, Cur- rently Unemployed 12 7.0 2 2.0 3 2.3 Incapacitated l 0.6 2 2.0 3 2.3 Retired 4 2.3 11 11.0 5 3.8 Deceased 4 2.3 6 6.0 7 5.4 (No Response) 1 0.7 Level of Employment Professional 35 20.3 13 13.0 31 24.2 Business 28 16.3 17 17.0 21 16.4 Skilled Worker 19 11.0 10 10.0 14 10.9 Unskilled Worker 16 9.3 17 17.0 10 7.8 Total Employed 98 56.9 57 57.0 76 59.3 equally divided among professional, business, and skilled labor positions, and a smaller group of unskilled laborers. (See Table 3.7.) 91 Table 3.7 Fathers' Employment Groups Under- Graduate Medical graduates Students Students Father Employed n % n % n % Never or Only Briefly 3 1.7 0 0.0 l 0.7 Currently Employed 143 83.1 53 53.0 93 71.5 Usually Employed, Cur- rently Unemployed 4 2.3 2 2.0 2 1.5 Incapacitated 3 1.7 2 2.0 2 1.5 Retired 10 5.8 23 23.0 12 9.2 Deceased 7 4.1 17 17.0 16 12.3 No Father in Home 2 1.2 3 3.0 4 3.1 Level of Employment Professional 55 32.0 26 26.0 51 39.8 Business 54 31.4 26 26 0 35 27.3 Skilled Worker 50 29.1 36 36.0 27 21.1 Unskilled Worker 11 6.4 11 11.0 12 9.4 Total Employed 170 98.9 99 99.0 125 97.5 Respondents were also asked about their experiences with the deaths of others, particularly close family and friends, and whether, to their knowledge, they were suffering from a terminal illness themselves. Nine said that they were. It is possible that some or all of those who responded positively to that question did so considering "life" as a terminal ill- ness. This cannot be assessed, Since respondents were not asked the specific nature of their illness. Four members of this group had lost their mother to death, nine their father, seventeen had lost one or more siblings, and one person had lost more than one child. None had been widowed. Twenty-seven (15.7%) 92 had lost a significant relative within the past year, 12 (7.0%) a friend, and 11 (6.4%) had lost more than one significant person within the year. Only seven (4.1%) said that that person's death had ever been mentioned in conversation between them, although in most cases death had been sud- den, affording no chance to talk about it. Twenty-six subjects said there was a relative important to them who was expected to die soon, five had a friend who was dying, and one person expected soon to lose more than one important person. Twenty said they had talked with that person about their impending death, while thirteen had not. Thirteen of the undergraduates (7.6%) had never attended a funeral, while 24 (14.0%) had been to nine or more. Fourteen (8.1%) said they had been with a person at the moment of his or her death, while 96 (55.8%) had never been with a person who was terminally ill, even one who was not very close to death. Fourteen reported having talked about death with a person whom they knew to be dying. Finally, they were asked directly whether they were afraid of their own death. A summary of those responses is shown in Table 3.8. Discussion of that question is deferred to another section of this chap- ter, since it is directly related to the question of the reliability of the fear of death scales. Graduate students.. The second sample consisted of 100 graduate students enrolled in graduate level courses in educational administration, counseling, and teaching, at Michigan State University, during the 1975- 76 academic year. They received the questionnaire during their class meeting. Some filled it out during class; others took it home and re- turned it at the next class meeting. It was described to them, as to all subjects, as a questionnaire about their experiences with and attitudes 93 Table 3.8 Fear of Death l r Groups Under- Graduate Medical graduates Students Students Attitude n % n % n % Afraid 13 7.6 6 6.0 16 12.3 Sometimes Afraid 100 58.1 57 57.0 79 60.8 Not Afraid 52 30.2 25 25.0 33 25.4 Don't Think About It 7 4.1 12 12.0 2 1.5 about death. It was stressed that their responses would be anonymous and that participation was voluntary. In all cases, good cooperation was re- ceived from those who were approached, and no one was pressured in any way to participate. Forty-eight subjects in this group were male, and 52 female. (Since sample size was exactly 100, it is not necessary to report both frequencies and percentages.) In age, 26 were between 21 and 25 years old, 42 between 26 and 30, 25 between 31 and 40, and 7 were 41 or older. There were 21 single people, 67 who were married, 6 who were separated or divorced, and 5 cohabiting with a partner of the same or opposite sex. Forty-two had children. Only 6 lived with their family of origin, while 70 lived with their spouse or partner and/or children, 11 with roommates, and 13 alone. Since this group was limited to students enrolled in grad- uate courses in education, the majority (84%) reported a major in the social sciences (including education). (See Table 3.2.) The religious upbringing and current religious preference of mem- bers of this group are shown in Tables 3.3 and 3.4. The distribution of 94 membership is approximately the same as in the undergraduate group. The majority were brought up in either a major Protestant denomination (40%) or in the Roman Catholic or Orthodox church (35%). In current preference, these two religious categories still held the largest number of members (23% and 21% respectively), but a sizeable group (18%) reported no cur- rent affiliation with any religious group. In religious self-rating, the largest number rated themselves as "about average'I (37%) (see Table 3.5), while no one described him or herself as "anti-religious." In Table 3.6 the employment level of the mothers of these sub- jects is shown and in Table 3.7 that of their fathers. All the graduate students grew up with a mother or mother-substitute in the home, while three said they had had no father present during their childhood. Five subjects in this group reported currently suffering from a terminal illness. Within their families, eight had lost their mother and eighteen their father. Eight had lost one or more siblings, one had lost a child, and none had been widowed. Twenty had lost a relative who was important to them within the past year, five a friend, and seven had suffered more than one loss. Nine said that they had talked with that person about his or her impending death, while in 17 cases death had been sudden. Six people in this group reported that they expected a relative to die soon, and four of them had talked about this person's impending death with him or her. Only four people had never attended a funeral, 31 had been to between one and four funerals, 30 to between five and eight, and 35 had attended nine or more--the largest single group. The increase in the proportion of subjects who had attended a larger number of funerals may be attributable to the greater age of members of this group. Fifteen 95 said they had been with a person at the moment of her or his death, while 41 had never had contact with a terminally ill person. Subjects' atti- tudes toward their own deaths are summarized in Table 3.8; the largest number indicated that they are sometimes afraid and sometimes not afraid of their own death. Medical students. The third sample consisted of 72 first year and 58 second year medical students in the College of Human Medicine at Michigan State University, a total of 130. All were in the pre-clinical phase of their training, although some had experienced contact with pa- tients in other positions they had held before entering medical school. All 200 students in the first and second year classes were con- tacted by a letter requesting their participation in the study, and in- forming them that they would receive the questionnaire through their of- fice mail boxes in the medical school (Appendix B). Since, in the second part of the study, only medical students were used as subjects, and it was doubtful whether they would be willing to participate, the additional contact was intended to increase their awareness of the existence of the study and hopefully their inclination to participate. This may not have been necessary; responses to the questionnaire from medical students far exceeded expectations. Within a week after they received the first letter, the questionnaire was distributed, and they were asked to return it through campus mail. At the end of the questionnaire, they were asked to indicate whether they would be willing to participate in the second part of the study, for which they would be paid $3.00. A total of 130 usable responses were received. Among the medical students there were 73 males (56.2%) and 57 females (43.8%). Only one was under 20 years of age, while 117 (90%) 96 were between 21 and 30 years old, and 12 were between 31 and 40 (9.2%). There were 64 single students (49.2%), 48 who were married (36.9%), 10 (7.7%) separated or divorced, and 8 (6.2%) cohabiting with a partner of the same or opposite sex. Thirty-three (25.4%) had children. Three (2.3%) lived with their family of origin, 53 (41.4%) with their spouse or partner and/or children, 50 with roommates (39.1%) and 23 (17.7%) alone. One subject listed as academic major the field of natural sciences; all others named medicine, but all were enrolled in the medical school. As in the other two groups, the majority were brought up either in a major Protestant denomination (32.3%) or in the Roman Catholic or Orthodox church (30.8%). Unlike the other groups, a sizable proportion (16.9%) had been raised Jewish. (See Table 3.3.) In current religious persuasion, this group differed from the others in that the largest num- bers listed themselves as either agnostic (16.2%) or as religious, but with no formal affiliation (16.2%). Those groups which held the largest numbers in the undergraduate and graduate student groups, the Roman Catholic and major Protestant denominations, attracted 15.4% and 13.1% of the medical students respectively. (See Table 3.4.) Table 3.5 shows the religious self-ratings of this group; the largest number rated them- selves as "about average" in religiosity. All subjects in this group had grown up with a mother or mother- substitute present in the home, but four reported having had no father during their childhood. The employment status and levels of the parents of these students are reported in Tables 3.6 and 3.7. A slightly larger percentage of students in this group reported that their fathers had been employed as professionals than in the other two groups. No one in this group reported being terminally ill. Eight had 97 lost their mother (6.2%), and 17 (13.1%) their father. In all groups, fathers had died more often than mothers. This result is to be expected, due to the fact that men are generally older than their wives, and have a lower life span, and are also more likely to be involved in a hazardous occupation or military combat. Six subjects had lost one or more siblings, one had lost more than one child; none were widows. Nineteen (14.6%) had lost an important relative in the past year, nine (6.9%) a friend, and three had lost more than one important person recently. Seventeen of these said that they had discussed with that person their impending death, while in ten cases there had been no opportunity to do so due to the sud- denness of death. Sixteen had a relative who was expected to die soon, one a friend, and two had more than one person they expected to lose soon. Twelve said they had talked with that person about his or her impending death. Eight of the medical students had never attended a funeral. Twenty- seven (20.8%) had attended nine or more; the remainder had been to between one and eight. Thirty-seven (28.5%) said they had been with a person at the moment of his or her death, while 36 had had no contact at all with any terminally ill person. Twenty-eight (21.5%) said they had talked about death with a dying person. Their attitudes towards their own death are summarized in Table 3.8; the majority (60.8%) said they were some- times afraid and sometimes not afraid of death. Experimental sample. Thirty-one male and twenty-five female mem- bers of the group of medical students who completed the questionnaire served as the experimental sample. They responded to the simulation tapes, and were interviewed extensively about their experiences with death and their attitudes about death and the role of the physician in 98 working with the dying patient. All but nine of those who completed the questionnaire indicated a willingness to participate in this part of the study as well. Those who had returned an incomplete questionnaire were eliminated from the potential subject pool. The remainder were called by telephone, and asked to participate in the 45 minute experiment and inter- view. A few refused to participate further due to lack of time or illness, a few could not be scheduled due to conflicts between their schedules and those of the interviewers, and many could not be contacted by phone at all. Fifty-eight subjects were interviewed, of which 56 produced usable material. Of the two which could not be used, one became upset early in the experiment and asked not to complete it. The interviewer spent the remaining time talking with him about his distress, rather than following the interview schedule. One other subject's responses were lest due to the failure of the tape recorder to function. Subjects were offered $3.00 for their participation. About half of those interviewed said, at the instigation of two of the subjects, that they wanted their payment donated to the Multiple Sclerosis Association for research purposes. The others either were paid or refused money al- together, saying that they were glad to help, or that they had partici- pated out of interest and the hope that they might learn something. Many, when offered money at the end of the interview, indicated that they had entirely forgotten that payment had been offered for participation. Money, therefore, appears not to have been a major factor motivating subject's participation. There were 36 first year students and 20 second year students in this group. No effort was made to contact an even number of first and second year students, and they were treated as a single group. While 99 there is some evidence2’3’4’5 that attitudes towards death change over the course of medical training, it appears unlikely that measurable changes would occur while a student was still in the pre-clinical phase of training. For most of these students, the issue of how they will handle dying patients is one with which they have not yet been faced in their professional life. In the experimental group of medical students, 31 were male (55%) and 25 female (45%). One was under 20, 28 (50%) were between 21 and 25 years old, 19 (34%) were between 26 and 30, and eight (14%) between 31 and 40. Thirty-two (57%) were single, 20 (36%) married or cohabiting, and four (7%) were separated or divorced. Twelve (21%) had children. Two subjects lived with their family of origin, eighteen (32%) with spouse or partner and/or children, 27 (48%) with friends, and nine lived alone (16%). Fourteen (25%) said that their mothers never or al- most never worked outside the home; 34 had mothers who currently or usually were employed (62%) and seven (13%) said that their mothers were incapaci- tated, retired, or deceased. Among the mothers who worked, twenty (36%) were professionals, six (11%) were in business, and sixteen (29%) were skilled or unskilled workers. Forty-one (73%) said that their fathers zHerman Feifel, etif|-."Physicians consider death," Proceedings, 75th Annual Convention of the American Psychological Association, 1967, 201-202. 3Sharon Golub and Marvin Reznikoff, "Attitudes toward death: a comparison of nursing students and graduate nurses," Nursing Research, 1971, 20 (6), 503-508. 4David Lester, etifl-,"Attitudes of nursing students and nursing faculty toward death," Nursing Research, 1974, 23 (1), 50-53. 5Peter B. Livingston and Carl N. Zimet, "Death anxiety, authori- tarianism, and choice of specialty in medical students," Journal of Nervous and Mental Disease, 1965, 140 (3), 222-230. 100 usually or currently were employed, two said they had grown up without a father in the home, and thirteen said that their fathers either never were employed outside the home, or were retired, incapacitated or de- ceased. Among those fathers who were currently or previously employed, twenty-five (45%) were professionals, fourteen were in business (25%), eleven (20%) were skilled workers, and five (9%) unskilled workers. Eighteen of these subjects were brought up in the Roman Catholic or Orthodox church (32%). Twenty-one (38%) were raised in a major Protestant denomination, and seven (13%) in a fundamentalist Protestant church. Seven (13%) were raised Jewish, and three in some other religious group. In current religious preference, eleven (20%) were Catholic or Orthodox, seven Protestant (13%), five (9%) fundamentalist Protestant, eight (14%) Jewish, thirteen (23%) said they were pro-religion with no formal affiliation or members of some other religious group, and twelve (21%) were atheists or agnostics. Ten (18%) rated themselves as "very religious," 21 (38%) as "about average," twelve (21%) as "not very reli- gious,“ and 13 (23%) as "non-religious" or "anti-religious." Thirty- five (62%) expressed some degree of belief in an afterlife, eighteen (32%) strongly believed in an afterlife while eleven (20%) strongly disbelieved in one. No one in this group was widowed, and none had ever lost a child. Forty-two (75%) said that they had not experienced the death of a signifi- cant other person within the past year. Five had lost a grandparent, one a parent, five another close relative or friend, and three had suffered more than one loss. Seventeen (30%) said that they had been with some person at the moment of his or her death, fourteen (25%) had been with a person close to the moment of death, eleven (20%) had been with a person 101 who was terminally ill, but not close to death, and fourteen (25%) had never had any contact with a terminally ill person at all. Eleven (20%) said they had experienced a conversation about death with a person who was dying. The students were asked what their first and second choice medi- cal specialty would be if they had to choose today. The most popular first choice speciality was family practice, specified by 25 students, followed by internal medicine, chosen by 11. Five people chose pediatrics; psychiatry and emergency medicine were each chosen by three. Other specialties named, each by only one person, were oncology, anaesthesiology, neurology, physiatry (work with the chronically disabled), ob/gyn, radiol- ogy, surgery, and hematology. One person was unable to give a preference. Most indicated that they really did not know what specialty they would eventually choose, and would not have expressed a preference except that they were urged to do so. They were also asked what their second choice would be at the current time. Family practice was again the top choice, given by 16 students, followed by internal medicine and pediatrics, each chosen by eight, ob-gyn by six, psychiatry by three, and emergency and public health/ preventative medicine by two each. Others, mentioned by only one person, were surgery, cardiology, neurology, radiology, anaesthesiology, geria- trics, endocrinology, and health center work. Three students refused to specify a second choice. Again, most students stressed that their choices were very much open to change. Subjects were asked to describe in detail their attitudes about telling a patient that S/he is dying, whether they themselves would want to know if they were dying, and why. These topics are discussed in Chapter 4. 102 Construction of the Death Experience Qgestionnaire In the description of the various samples, the manner in which the questionnaire data was collected is described. In this section, the construction of the questionnaire used in this study is described, along with the results of reliability studies on the Collett-Lester subscales, which comprise part of the Death Experience Questionnaire (which will be referred to, for convenience sake, as the DEQ). (Appendix C.) Demographic Items. Research reviewed in Chapter II has produced conflicting results on the relevance of various demographic characteris- tics to the development of fear of death. A comprehensive set of demo- graphic items is included in the DEQ in order to determine what results of previous studies might be borne out in the present one. The content of the demographic items is drawn from previous research, and includes questions on sex, age, narital status, employment history of parents (socio-economic status), academic major, religion, and previous experience with death. Only those items which were scalable were examined for pos- sible correlations with fear of death scale scores; the remainder are used in describing the samples in the first section of this chapter. Fear of Death Items. The remainder of the DEQ is composed of the items of the Collett-Lester fear of death inventory, which they have subdivided into four scales measuring fear of death of self (05), death of others (00), dying of self (DyS), and dying of others (DyO). The items of this scale are all from Lester's6 scale; those items of Lester's original scale which did not show clear correlations with one subscale and low or zero correlations with others were deleted by Collett and 6David Lester, unpublished Ph.D. Dissertation, 1967. 103 Lester from the final form of the inventory. In this questionnaire, the items are all included in the order in which they appear in Collett and Lester's scale; all appear after the demographic items. The scoring of these items was changed to conform with that of other items of the DEQ. The original scoring was on a six-point scale from -3 (strong disagree- ment) to +3 (strong agreement). Here, possible scores on each item run from 0 (strong disagreement) to 5 (strong agreement). No neutral choice is allowed. Half the items are worded so that a high score indicates high fear, while the other half are worded so that a high score indicates low fear. Scoring on the negatively worded items was reversed in analysis so that high scores consistently indicate higher fear. Reliability of the Fear of Death Scales The four scales of the Collett-Lester inventory consist of 10 items each, with the exception of Scale 3, Fear of Dying of Self, which has eight items. Using data from the combined sample of undergraduates, graduate students, and medical students (n = 402), item and scale means and standard deviations were computed, along with inter-item correlations, and an estimate, by analysis of variance, of the reliability of the scale. (See Appendix A for items.) Scale 1., Fear of death of self. Scale 1 contains items which measure the individual's fear of his/her own death. As on all scales, half the items are positively and half negatively worded. Item content has to do with knowing what happens on earth after one's death, the shortness of life, death as the end of what one experiences, and knowing what being dead is like. Item means for this sample run from a low of 2.02 on Item 108 (Not knowing what it feels like to be dead does not 104 bother me) to a high of 4.09 on Item 103 (I would not mind dying young). Scale mean was 27.21, standard deviation 10.15, possible range from 0-50. Reliability of Scale 1, calculated by analysis of variance, was .796. The poorest item on the scale was Item 105 (I view death as a relief from earthly suffering). Inter-item correlations were the lowest for this item (the highest only .25, with Item 103). Reliability of the scale could be improved to .803 if Item 105 were deleted. This item is the only one deletion of which would improve the reliability of the scale. (Inter-item correlation matrices are found in Appendix O.) Scale 2., Fear of death of others. Scale 2 contains items mea- suring the individual's fear of the death of other people. Item content has to do with loss, accepting the death of another as the end (Hi his/her life on earth, communication with the dead, and sadness about what the dead person would be missing. For the current sample, item means ranged from a high of 4.69 for Items 201 and 209 (I would experience a great loss if someone close to me died; If someone close to me died, I would miss him/her very much), to a low of 1.13 for Item 202 (I accept the death of others as the end of their life on earth.) Mean score on Scale 2 was 28.75, standard devia- tion 6.60, possible range 0-50. Reliability of Scale 2 was .52, which could have been appreciably improved by the deletion of Item 208 (I do not think of dead people as having an existence of some kind). Reliability without this item would have been .58. Scale 3., Fear of dyingyof self. Scale 3 measured the individual's fear of going through the process of dying, in contrast to Scale 1 which measures the fear of being dead. Item content has to do with deteriora- tion, pain, the limitation of one's experiences during dying, talking 105 about death, knowing one is dying, and preference for a sudden death. Item means ranged from a low of .43 for Item 306 (If I had a fatal di- sease, I would like to be told) to a high of 3.82 for Item 301 (I am disturbed by the physical degeneration involved in a slow death). Scale mean was 18.26, standard deviation 4.54, possible range 0-40. As in all previous studies which have used this scale, Scale 3 was the least homogeneous; reliability here was only .20. However, reliability could have been greatly increased, to .40 with the dele- tion of Item 307 (I would rather die suddenly than of a slow death). It is possible to speculate as to the reasons for the lack of homogeneity in Scale 3. The fear of physical and intellectual deteriora- tion may have little to do with fear of dying, and may vary independently of it. Thus, individuals who scored high on one kind of item may score low on the other. In fact, the only inter-item correlations in this scale which are greater than .30 are between Items 301, 305, and 308, the three items whose content has clearly to do with deterioration. All other inter-item correlations are low and many are close to zero, even among items which appear on the face of it to be related to fear of dy- ing. Scale 4.,7Fear of dying of others. Scale 4 measures fear of the process of dying of other people, rather than of the consequences of others being dead. Item content has to do with Spending time with a dying person, anxiety in their presence, being willing to talk about death with a dying person, preference that a friend die suddenly rather than slowly. Item means ranged from a low of .67 for Item 410 (If a friend were dying, I would not want to be told) to 3.83 for Item 402 (I would prefer that someone close to me diedaisudden death rather than a slow death). 106 Scale mean for this scale was 15.64, standard deviation was 6.15, possible range 0-50. Reliability was .58. The best reliability that could have been achieved would have been through deletion of Item 402 (I would prefer that someone close to me died a sudden death rather than a slow death), which would have resulted in a reliability of .60. Reliability of total Scale. .Reliability of the total scale was also calculated to be .77. The total scale is an addition of the four subscales. Relationships among_Sca1es. There were significant correlations among all subscales, a result which has also been found in previous re- search. (See Table 3.9.) Highest correlations were between Death of Self and Dying of Self (.35), Death of Self and Death of Others (.38), and Dying of Self and Dying of Others (.35). Table 3.9 Intercorrelations between Collett-Lester Subscales S2 S3 $4 $1 .38 .35 .16 s = .001 s = .001 s = .001 $2 .22 .12 s = .001 s = .008 53 .35 s = .001 Inter-scale correlations were computed for each group (under- graduates, graduate students, and medical students) separately as well. As in the total group, all correlations were significant at the .05 107 level or better, with the exception of the correlation between Scales 2 and 4 in the graduate students group. The lowest correlations were be- tween Scales l and 4, and Scales 2 and 4. (The correlations between all scales were significantly different from zero, but the correlations are so low that the percent of common variance among the scales is not very large.) While fear of death of self and dying of others might not be ex- pected to be very highly correlated, it is more difficult to explain the low correlations between fear of death of others and dying of others. Most of the intercorrelations are not very high, the strongest being less than .40, which suggests that, while these fears overlap somewhat, different constructs actually are being measured by the four subscales. Either the constructs are not very clearly distinct, or the scales do not measure them very well, or both. The last interpretation seems the most likely. Reliabilities of each scale, computed for each group separately, are shown in Table 3.10, showing that the instrument functions similarly in the different populations. Table 3.10 Subscale Reliabilities in Three Different Populations Under- Graduate Medical graduates Students Students n = 172 n = 100 n = 130 Scale 1 .80 .78 .81 Scale 2 .50 .57 .51 Scale 3 .28 .09 .12 Scale 4 .56 .56 .54 108 It is worth noting that Scale 3, which consistently functions least well, worked better in the undergraduate population than in the others. This finding may reflect the fact that the instrument was ori- ginally developed on an undergraduate population, and may indicate that its generalizability is limited. It is not clear what causes this limi- tation of its usefulness, but even at best the scale functions poorly. Aside from this exception, the reliabilities are fairly consistent across populations. Factor Analysis of Scale Items Although the scale of Collett and Lester was not constructed by factor analysis, such an analysis was performed to test the extent to which the scales would empirically factor as written, and to determine whether any additional factors representative of clear constructs would emerge. Using Guttman's lower bound theorem, twelve factors emerged. An additional factor analysis was performed to force the items into four factors, since there were four scales on the original inventory. In the original twelve-factor analysis, seven of the ten items of Scale 1 (Fear of death of self) loaded on Factor 1, along with one item each from Scales 2 and 3. Four of the eight items of Scale 3 loaded on Factor 2, along with one item from Scale 4. No more than three items of any one scale loaded together on the same factor in any other case. (See Ap- pendix E for items and factor loadings.) In the four-factor analysis, a Varimax rotation with Kaiser normalization was performed, and it is that rotation which is discussed here. Items of the four-factor analysis were examined for comparisons with the grouping of items in the original subscales, and also for any 109 constructs they might exhibit which did not correspond with the constructs used in the CL scale. Each factor is discussed individually here. Factor 1. All items of Scale 1, Fear of death of self, loaded on the first factor (Appendix E), along with two items from Scale 2, Fear of death of others. Examining those items with loadings higher than .40, it appears that most of them have to do with wanting to know what it is like to be dead. The exceptions are Items 104 (I am disturbed by the shortness of life) and 205 (If someone close to me died, I would miss him/her very much) and 106 (I would not mind dying young). Three items (103, 205, and 208) appear to be ones which could easily be answered in the same way by either high or low fear individuals, and it is difficult to interpret the meaning of these items in the context of fear of death. Factor 2. Factor 2 is composed primarily of items from the two scales which purport to measure fear of the process of dying, as opposed to the state of being dead. The better items have to do with avoidance of the pain of dying or of the experience of dying, either one's own or that of others, and avoidance therefore of dying persons and even of the knowledge that someone else is dying. (Items having to do with knowledge of one's own impending death