OVERDUE FINES ARE 25¢ pER DAY PER ITEM Return to book drop to remove this checkout from your record. Copyright by KAREN MARODA 1979 CONCEPTUAL DEVELOPMENT AND MEASUREMENT OF ADULT SEPARATION ANXIETY BY Karen Maroda 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 1979 ABSTRACT CONCEPTUAL DEVELOPMENT AND MEASUREMENT OF ADULT SEPARATION ANXIETY BY Karen Maroda The purpose of this study was to develop a valid theoretical concept of adult separation anxiety, drawn from the existing research pertaining mostly to children, and attempt to measure this set of dynamics through construction of a personality test. It was postulated that significant results from such a test, in the predicted direction, would give evidence for both the validity of the test itself and for the validity of the theoretical constructs on which it was built. An objective format was chosen for the test of separation anxiety, and items were formulated in accordance with established procedures regarding face validity, mean- ing, and readability. The response code chosen was a four- point true—false continuum. The final form of the test was given to two groups, accompanied by a background questionnaire asking about family history, personal habits, physical problems and current attitudes that were related to separation anxiety. Karen Maroda The first group given the test and questionnaire was a diverse but not random group of 400 adults, who were approached in various settings and agreed to participate. This group was labeled the General Sample and was surveyed for the purpose of acquiring a sample that was large enough to provide sufficient data for a reliable cluster analysis. Because both samples were sufficiently large, cluster analyses were done on the sets of responses to test items for both groups. This was done in order to determine which items were related to one another and would therefore be retained and scored; whether or not subscales would be formed, and, if produced, what their content would be. Results of these analyses produced six subscales, labeled Overprotection, General Separation Anxiety, Self-acceptance, Mother Hostility, Father and Basic Personality Type. The last two of these subscales were not scored for separation anxiety because they were exploratory in nature and not intended for that purpose. The second sample was chosen because they were in the midst of a separation experience. Ninety-four percent of the first-term freshmen women living in Rather Hall, on the Michigan State University campus, were given the separation anxiety test and background questionnaire in an attempt to study the validity of the test. It was hypoth- esized that, of the 300 women participating, those who were the most homesick and anxious, as determined by Karen Maroda self-report, would also score significantly higher on the test of separation anxiety. Comparisons were then made between groups within each sample, on the basis of their responses to question— naire items. Mean subscale scores were compared through the use of the t—test procedure, with the probability of significance set at p < .05. All hypotheses were supported by at least one of the four subscales scored for separation anxiety. The 16 different group comparisons made under all of the hypotheses tests yielded 64 opportunities for the subscales to be found significant in the predicted direction, significant in the Opposite direction, or not significant. Of this number, there were 33 cases of significant differ- ences in the predicted direction, one case of significance in the Opposite direction, and 30 cases of no significant differences between subscale mean scores. DEDICATION To my parents, Jean and Frank Maroda, my sister, Liz, and my brother, Greg. iii ACKNOWLEDGMENTS I discovered in the process of completing this dissertation that any worthwhile research project requires the skills and talent of more than one person; that the task is a very complex and demanding one; and that the support of friends and family is invaluable. Therefore, to the following colleagues, consultants and friends who gave me the help I needed, I would like to express my deep appreciation: Jill Hirt, who provided consultation throughout every aspect of this study, worked extremely hard in giving practical assistance, and continually offered the moral support and generous spirit of true friendship. Sandy Simmons, who graciously assisted in both data collection efforts, giving freely of her time when she had much work of her own. Her constant encouragement often kept me going when I felt like quitting. Bill Brown, of the MSU Computer Center, who wrote all of the major computer programs for analyzing the data in this study, and was very generous with his time and expert computer skills. iv Robert L. Green, Dean of the College of Urban Development and committee member, who employed me as a research associate for three years and helped me to stay in touch with the problems and rewards of the world outside academia; gave me the opportunity to become involved in professional work that deve10ped my writing and research skills; and continually lent his personal and professional support. Norman Stewart, committee chairman, for his helpful review and editing of the manuscript. Albert Rabin, committee member, for his editing of the literature review and for introducing me to the work of John Bowlby. Marlene Pringle, committee member, for her encourage— ment. Finally, I would like to thank all of the peeple working every shift in the Michigan State University Com- puter Laboratory. In the three months that I spent working there I was continually amazed at how efficient, friendly, and helpful these people were, even when confronted con- stantly with the problems and requests of dozens of inexperi- enced users like myself. Their help greatly facilitated the analysis of this data and was greatly appreciated. LIST OF CHAPTER 1. TABLE OF CONTENTS TABLES O O O O O O O O O O O 0 THE PROBLEM . . . . . . . . . . Need for the Study . . . . . . Purpose . . . . . . . . . . . . Hypotheses . . . . . . . . . . Overview . . . . . . . . . . . Definition of Terms . . . . . . Summary . . . . . . . . . . . . THEORETICAL CONSIDERATIONS OF EARLY SEPARATION ANXIETY . . . . . . Introduction . . . . . . . . . Theoretical Considerations . . Summary . . . . . . . . . . . . REVIEW OF THE LITERATURE . . . Introduction . . . . . . . . . Major Theoretical Approaches to Separation Anxiety . . . . . The Work of Bowlby . . . . Protest, Detachment, Despair Supporting Evidence . . The Work of Mahler . . . . Link Between Mahler and Bowlby Separation-Individuation Process . . . . . . . Summary . . . . . . . . . . School Phobia . . . . . . . vi Page xi H OQO\UIU1I—' ll 12 20 21 21 23 23 23 24 27 28 29 37 37 CHAPTER Page Relationship to Separation Anxiety . . . . . . . . . . . . . 38 Summary . . . . . . . . . . . . . . 39 Separation Anxiety: Etiology and Dynamics . . . . . . . . . . . . . . . . 40 Mother-Child Relationship . . . . . . . 40 Repetition of Relationship Through Generations . . . . . . . 42 Hostility . . . . . . . . . . . . . 43 Symbiosis and Aggression . . . . . . . 45 Hostility and Rejection . . . . . . . . 46 Summary . . . . . . . . . . . . . . 47 Role of the Father . . . . . . . . . . 47 View of Father's Role by Mahler and Others . . . . . . . . 48 Position of Father in Relation- ship to School Phobia . . . . . . 49 Father's Typical Role in Family . . . . . . . . . . . . . 51 Father as Facilitator of Individuation . . . . . . . . . . 53 Summary . . . . . . . . . . . . . . 54 Siblings and Related Deve10pment . . . 54 Sibling Rivalry . . . . . . . . . . 54 Oedipal Conflict . . . . . . . . . 56 Summary . . . . . . . . . . . . . . 57 Specific Aspects of Separation-Anxiety Psychodynamics . . . . . . . . . . . . . 58 Orality . . . . . . . . . . . . . . . . 58 Change in Appetite as a Response to Separation . . . . . . . . . . 58 Symbolic Meaning of Food . . . . . 60 Summary . . . . . . . . . . . . . . 60 Body-Image, Sexual Identity, and Sexual Behavior . . . . . . . . . . . 61 vii CHAPTER Page The Deve10pment of Body-Image . . . . 61 Receptivity to Physical Contact . . . 63 Consequences of a Poor Body Image . . . . . . . . . . . . . . . 64 Sex and the Need to Be Held . . . . . 66 Summary . . . . . . . . . . . . . . . 66 Sex Differences . . . . . . . . . . . . . 67 Motility and Differentiation . . . . 67 Possible Influential Factors . . . . 68 Observed Differences . . . . . . . . 70 Summary . . . . . . . . . . . . . . . 70 Psychosomatics . . . . . . . . . . . . . 71 Purpose of Illness . . . . . . . . . 71 Most Prevalent Problems . . . . . . . 72 Characteristics of Underlying Conflicts . . . . . . . . . . . . . 74 Summary . . . . . . . . . . . . . . . 75 Adult Separation Anxiety . . . . . . . . . . 76 Introduction . . . . . . . . . . . . . . 76 Separation Anxiety: Continuation into Adulthood . . . . . . . . . . . . 77 Mitigating Factors . . . . . . . . . . . 81 Separation-Individuation as a Life-long Process . . . . . . . . . . . 84 Summary . . . . . . . . . . . . . . . . . 87 4. TEST CONSTRUCTION . . . . . . . . . . . . . . 89 Format . . . . . . . . . . . . . . . . . . . 89 Steps in Test Construction . . . . . . . . . 92 Item Choice . . . . . . . . . . . . . . . 93 Response Codes . . . . . . . . . . . . . . . 95 First Test Administration . . . . . . . . . . 97 Type of Sample . . . . . . . . . . . . . . . 98 Cluster Analysis . . . . . . . . . . . . . . 100 Resulting Subscales . . . . . . . . . . . . . 102 Summary . . . . . . . . . . . . . . . . . . . 105 viii CHAPTER Page 5. EXPERIMENTAL DESIGN AND PROCEDURES . . . . 106 Overview . . . . . . . . . . . . . . . . . 106 Samples . . . . . . . . . . . . . . . . . . 106 Hypotheses . . . . . . . . . . . . . . . . 111 Hypotheses Applying to Both Samples . . . . . . . . . . . . . 111 Hypotheses Applying Only to the General Sample . . . . . . . . . 114 Hypotheses Applying Only to the Freshmen Sample . . . . . . . . . 114 Analysis . . . . . . . . . . . . . . . . . 115 6. RESULTS . . . . . . . . . . . . . . . . . . 118 Hypotheses Tests . . . . . . . . . . . . . 122 Hypotheses Tested for Freshmen Sample Only . . . . . . . . . . . . . 139 Summary . . . . . . . . . . . . . . . . . 142 Additional Results . . . . . . . . . . . . 145 Frequency of Calls and/or Visits to Mother . . . . . . . . . . 151 Separated from Mother as a Child . . . 151 Living with Parents or Not . . . . 154 Marital Status . . . . . . . . . . . 154 Race . . . . . . . . . . . . . . . . 157 College Adjustment Self- -report . . . . 160 Age . . . . . . . . . . . . . . . . . 160 Religion . . . . . . . . . . . . . . . 162 Number of Pe0ple Close to Emotionally At One Time . . . . . . . 164 Summary . . . . . . . . . . . . . . . . 167 7. SUMMARY AND CONCLUSIONS . . . . . . . . . . 168 Summary . . . . . . . . . . . . . . . . . . 168 Conclusions and Discussion . . . . . . . . 171 Hypotheses . . . . . . . . . . . . . . 182 Discussion of Additional Results . . . 186 Limitations . . . . . . . . . . . . . . . . 188 ix CHAPTER Page Generalization of Findings . . . . . . 188 Item Selection . . . . . . . . . . . . 189 Delineation of Criterion Groups . . . . 190 Implications for Future Research . . . . . 190 Implications for Counseling Practice . . . 192 Conclusion . . . . . . . . . . . . . . . . 195 APPENDICES A. Separation Anxiety Test, Form A . . . . . . 196 B. Background Questionnaire for General Sample . . . . . . . . . . . . . . . . . 203 C. Instructions to Research Assistants . . . . 206 D. New Test Items Added to Original 127 Items . . . . . . . . . . . . . . . . . . 208 E. Separation Anxiety Test, Form B . . . . . . 209 F. Significant Test Items for School Phobia (General Sample) . . . . . . . . . . . . 217 G. Subscales and Correlation Matrices . . . . 219 H. Correlation Matrix for Clusters . . . . . . 225 I. Background Questionnaire for Freshmen sample 0 O O O O O O O I O O O I O O O O 226 REFERENCES . . . . . . . . . . . . . . . . . . . . 233 Table 5.1. 5.2. 6.1. 6.6. LIST OF TABLES Demographic Data for General Sample . Demographic Data for Freshmen Sample Comparison of Mean Subscale Scores by Degree of Fear of Going to School as a Child for General Sample . . Comparison of Mean Subscale Scores by Degree of Fear of Going to School as a Child for Freshmen Sample . . Comparison of Mean Subscale Scores by Degree of Orality (Weight) for General Sample . . . . . . . . . . Comparison of Mean Subscale Scores by Degree of Orality (Weight) for Freshmen Sample . . . . . . . . . . Comparison of Mean Subscale Scores by Degree of Orality (Smoking) for General Sample . . . . . . . . . Comparison of Mean Subscale Scores by Degree of Orality (Smoking) for Freshmen Sample . . . . . . . . . . Comparison of Mean Subscale Scores by Degree of Orality (Weight and Smoking Behavior) for General Sample . . . Comparison of Mean Subscale Scores by Degree of Orality (Weight and Smoking Behavior) for Freshmen Sample . . . Comparison of Mean Subscale Scores on Present Health Problems for General sample 0 O O O O O O O O O O O O 0 xi Page 109 112 123 124 126 127 129 130 131 132 133 Table 6.10. 6.16. 6.22. Comparison of Mean Subscale Scores on Present Health Problems for Fresh- men Sample . . . . . . . . . . . . . Comparison of Mean Subscale Scores by Opinion About Living Alone for General Sample . . . . . . . . . . . Comparison of Mean Subscale Scores by Opinion About Living Alone for Freshmen Sample . . . . . . . . . . . Comparison of Mean Subscale Scores by Sex for General Sample . . . . . . . Comparison of Mean Subscale Scores by Degree of Homesickness for Freshmen Sample . . . . . . . . . . . . . . Comparison of Mean Subscale Scores by Degree of Separation Anxiety for Freshmen Sample . . . . . . . . . . . Comparison of Mean Subscale Scores by Close Relationships for Freshmen Sample . . . . . . . . . . . . . . . Summary of Results for General Sample . Summary of Results for Freshmen Sample Results for Father Subscale (Freshmen Sample) . . . . . . . . . . . . . . . Comparison of Mean Subscale Scores by Childhood Absence from School for General Sample . . . . . . . . . . . Comparison of Mean Subscale Scores by Childhood Absence from School for Freshmen Sample . . . . . . . . . . . Comparison of Mean Subscale Scores by Frequency of Contact with Mother for General Sample . . . . . . . . . Comparison of Mean Subscale Scores by Long Childhood Separation from Mother for General Sample . . . . . . xii Page 135 137 138 140 141 143 144 146 147 148 150 152 153 155 Table Page 6.24. Comparison of Mean Subscale Scores by Parental Residence for General Population . . . . . . . . . . . . . . . 156 6.25. Comparison of Mean Subscale Scores by Marital Status for General Sample . . . . 158 6.26. Comparison of Mean Subscale Scores by Race for General Sample . . . . . . . . . 159 6.27. Comparison of Mean Subscale Scores by Adjustment to College for Freshmen sample 0 O O O O O O O O O O O O O O O O 161 6.28. Analysis of Variance for General Sample on Age . . . . . . . . . . . . . . . . . 163 6.29. Analysis of Variance for Freshmen Sample on Religious Background . . . . . . . . . 165 6.30. Analysis of Variance for Freshmen Sample on Number of Close Relationships at One Time . . . . . . . . . . . . . . . . 166 7.1. Summary of Results by Criterion Group on Test Subscales for the General Sample . . 172 7.2. Summary of Results by Criterion Group on Test Subscales for the Freshmen Sample . . . . . . . . . . . . . . . . . 173 xiii CHAPTER 1 THE PROBLEM Need for the Study Separation anxiety is a psychological phenomenon that has been given little attention in modern psychology. During and after World War II, a handful of researchers, most of them in England, wrote about the effects on children of physical separation from their parents (Bowlby, 1960; Freud & Burlingham, 1944). Others, such as Fairbairn (1952), observed and wrote about the effects of wartime separation on soldiers. Still others were convinced that an individual's future development was dependent on his ability to become truly separate from his mother while maintaining a loving attachment. Then Mahler (1955) and her followers observed and labelled the stages of what they called the "normal separation-individuation" process. Following these developments, based primarily on the works of the British psychoanalytic school, came more systematic studies. Researchers began to study infants and small children in nursery settings and noted their reactions when left by their mothers and later reunited with them. Assessment in these studies was limited solely to the researcher's observations of mother and child, usually in a restricted environment outside the home. This was probably due to the fact that no instruments had been devised that could measure separation anxiety. This dependence on mother—child observations resulted in two basic research trends: 1. Focusing chiefly on limited interactions between mother and child to assess separation anxiety and developing theories regarding its nature and origin. Although some studies, such as Mahler's (1975) included home observations, this is not typical. And even Mahler's research excluded the effects of siblings and totally ignored the role of the father in the separation-individuation process. 2. The exclusion of adolescents and adults from the body of research concerning separation anxiety. The only exception to this pertains to adolescents and has been done in the area of school phobia, a generally accepted manifestation of separation anxiety. Hansberg (1972) has deve10ped a projective method for assessing separation anxiety that is designed for use with 11-17 year-olds. However, this instrument has no published validity information and to date has not been used to any extent as a research tool. More recently, Klagsbrun and Bowlby (1977) have devised an instrument based on Hansberg's test for use with young children, ages four to seven years. But again, little has been done to confirm the validity of this measure. Regardless of these instruments' validity, neither one has been designed for use with an adult population. Thus the study of separation anxiety terminates with the school-age child, since there is neither an appro- priate analogue for adult behavior nor a device for the assessment of separation anxiety in adulthood. Therefore, all of the questions that might be asked regarding the per- sistence of this dynamic into adulthood remain unanswered. Do children "grow out" of separation anxiety? If not, what are the consequences for an adult with this problem? How does he cope with the problem, if it continues into later years? How is separation anxiety manifested in adulthood? Do theoretical constructs translate into behavior and atti- tudes? What are they? And what part does separation anxiety play in the current schema of both normal function- ing and pathological states? To answer these and other research questions, it is obviously necessary to study adults. Randomly sampling and studying adults who are currently going through a major separation experience is an unfeasible method for deter— mining the extent of an individual's separation anxiety. To do so would require the screening of literally hundreds of persons in order to obtain a homogeneous sample of the appropriate size. Then this process would most likely have to be repeated, in order to obtain a comparable control group of adults who are not in the midst of a separation experience or reaction. A more efficient research method must be available that utilizes the standard random sampling procedures or examination of existing groups. Only when this practical concern is attended to, can the problem of separation anxiety be studied systematically. Then the application of the knowledge gained can be made to adult relationships, and to separation experiences such as moving away from home, ending a relationship, or losing a loved one. Bowlby (1973) points out the importance of such research: In this complex scene the place of separation anxiety is still unclear. In particular, it remains uncertain how large a contributor it is to sources of neurosis in comparison with anxieties and fears of different origin. . . . A pile of clinical reports . . . shows that experiences of separation and loss, occurring recently or years before, play a weighty role in the origin of many clinical conditions. These are grounds enough for concentrating attention on the problem. (p. 30) The problem remains of how to assess separation anxiety in adults, particularly for research purposes. Naturally, the more objective and brief, the better. But the measurement device must also have the capability of assessing a rather complicated set of psychodynamics that is largely unconscious and frequently denied. Thus, this thesis deals with the problem: Can adult separation anxiety be assessed by an objective personality measure? Purpose The purpose of this study is twofold: (a) to develop an objective measure to assess separation anxiety in adults; and (b) to determine the validity of the hypothesized theoretical constructs regarding the psychodynamics of separation anxiety. This was accomplished through the creation of an original objective measure. The items were formulated chiefly from theoretical material and published case studies. Item content was also derived from interviews with persons who identified themselves as having had life-long diffi- culties with separations, several of whom had been school refusers. When test construction was completed, the test was administered to different adult samples and groups were differentiated on the basis of current theoretical premises. Then the groups were compared for differences on their test scores. Hypotheses All of the hypotheses stated have their basis in two primary hypotheses: (a) that adult separation anxiety is a real phenomenon that can be measured with a paper-and- pencil test, and (b) that the test devised for this study is a valid measure of this phenomenon. More specific hypotheses will be formulated on the following research questions: 1. Do children who are afraid to go to school actually fear being separated from their mothers? (Hypothesis 1) 2. Are those with oral traits higher on separation anxiety than those without them? (Hypothesis 2) 3. Are certain psychosomatic disorders indicative of separation anxiety? (Hypothesis 3) 4. Do peOple who lost one of their parents during childhood maintain an anxious attachment to peOple in general, as a result of this experience? (Hypothesis 4) 5. Are persons with separation anxiety less likely to enjoy the prospect of living alone? (Hypothesis 5) 6. Are women more likely to be separation-anxious than men? (Hypothesis 6) 7. Is an extremely close family situation indicative of a fear of losing loved ones? (Hypotheses 7, 8 and 9) Overview The theoretical basis for this study is rather com- plex. For this reason it will not be presented in this chapter. Instead it will be presented entirely in Chapter 2. Chapter 3 contains a review of the various studies done relative to separation reactions and separation anxiety. This also includes more specific aspects of the general psychodynamic underlying the tendency toward experiencing separation anxiety. Chapter 4 is devoted to the step-by-step process involved in creating the test of separation anxiety that was conceived as part of this study. This section includes a report of the results of the cluster analysis performed on the test and a discussion of the rationale for the final selection of items. The design of the study for the validity of this test of separation anxiety follows in chapter 5. The sample used in the validity study is also described. Chapter 6 contains a report of the results of the comparisons made between groups, while the contents of chapter 7 include a summary, discussion of the results, conclusions, and impli- cations for future research. Definition of Terms Defining the terms used in the various theoretical discussions of separation and separation anxiety poses a problem discussed by Bowlby (1973). He states that when observing infants and children it becomes difficult to ascertain when a separation is "short" or "long," because the younger the child, the more slowly time seems to pass. It is also difficult to know when a separation defined as "temporary" appears to the child to be permanent. Thus, the meaning of many terms used in theory and research is relative. Additionally, Bowlby (1973) takes exception to the very use of the term "separation anxiety," on the grounds that it does not very accurately describe the condition that is present. . . . the term "separation anxiety" is not ideal. A better way to describe the condition is to term it "anxious attachment" or "insecure attachment." This makes it clear that the heart of the condition is appre- hension lest attachment figures be inaccessible and/or unresponsive. For these reasons, therefore, and especially because it can be expected to enlist our sympathy, anxious attachment is the term to be used. It respects the person's natural desire for a close relationship with an attachment figure, and recognizes that he is apprehensive lest the relationship be ended. (p. 23) While Bowlby's point is well taken, the term "separation anxiety" will be used throughout this thesis for the usual reason of desiring continuity. Another problem that presents itself when terms are to be defined is the different usage made of the same terms by certain theorists. When Bowlby discusses separation he is referring to physical separation. When Mahler uses the same word she is usually referring to the extent of intra— psychic differentiation achieved between mother and child. However, when they speak of "separation anxiety" and describe it as a psychOlogical phenomenon, they appear to be essenti- ally saying the same thing. Thus, a few basic definitions are presented here to aid the reader and prepare him or her for the theoretical discussion that follows. Separation (as used by Bowlby)--temporary inaccessibility of the subject's attachment figure (Bowlby, 1973). Loss--permanent inaccessibility of the subject's attachment figure (Bowlby, 1973). Separation (as used by Mahler)--refers to "the intrapsychic achievement of a sense of separateness from mother and, through that, from the world at large. . . . This sense of separateness gradually leads to clear intrapsychic represen- tations of the self as distinguished from the representa- tions of the object world" (Mahler et al., 1975). Infantile separation anxiety--the infant's fear of losing the mother. Childhood separation anxiety--"Fear, anxiety, etc. occasioned by the threat or actuality of separation from mother and home. School phobia is a type of separation anxiety" (Hinsie & Campbell, 1970). Adult separation anxiety--the adult's fear of being alone, without being attached to someone in a dependency relation- ship. Body image--"The picture or mental representation one has of his own body at rest or in motion at any moment. It is derived from internal sensations, postural changes, contact with outside objects and people, emotional experiences, and fantasies" (English & English, 1958). Dual unity--"The symbiotic unit of mother and child, imbued by the infant with omnipotent qualities, in which there is a vague sensing of the symbiotic half of the self (Spitz's 'external ego')" (Mahler et al., 1975). Adult symbiosis--"(E. Fromm) a condition in which a person depends upon others, not for cooperative mutual support and affection but for exploitation and the satisfaction of neurotic needs . . ." (English & English, 1958). 10 Hostile dependent--"Basic mother-child dynamic in separation anxiety. The mother is hostile due to ambivalance regarding motherhood. At the same time she binds the child to her and encourages symbiotic dependence as a defense against these hostile impulses and as a way of gratifying her own unsatisfied needs. The circle is completed by the subse- quent development of an almost identical dynamic in the child" (Estes et al., 1956). Object+-"The object of an instict is that in or through which it can achieve its aim" (Hinsie & Campbell, 1970). Object constancy--the extent to which an object has been introjected and becomes a permanent representation. Summary The lack of scientific investigation pertaining to the psychological phenomenon of separation anxiety in adults has been cited. The need for such an investigation has been shown, in that an understanding of adult separation anxiety has major consequences as it relates to forming relationships, functioning in a mobile society, and coping with divorce, aging and death. To accomplish this inquiry, a criterion for determining separation anxiety is necessary. Therefore, this thesis is concerned with the development of both the concept of adult separation anxiety, and an objec- tive test to measure it. This will provide a base for future research and, perhaps, a new perspective on psy- chological functioning. CHAPTER 2 THEORETICAL CONSIDERATIONS OF EARLY SEPARATION ANXIETY Introduction To facilitate the reader's understanding of the concepts of separation anxiety and its concommitants, and so that the basis for the content of test items is clear, a review of the underlying theory will be presented. The concept of separation anxiety is not currently a part of the mainstream of American psychological thought and practice, and remains even more obscure in the research. In fact, most of the writers in this area are British or European and most are adherents of the "object relations" school of psychoanalytic thought. While the writings of this school are gaining in popularity, they are not as well-known or popular as those of the classic Freudian school. Thus, it seems appropriate to provide an overview so that the reader, regardless of discipline, may better understand the theoretical basis for this thesis. After this general theoretical material is presented, more specific theoretical constructs, along with their research evidence, will be presented in chapter 3--Review 11 12 of the Literature. The infonmation provided in chapter 3 will shed light on the specific psychodynamic structure and behavioral manifestations that develop as a result of early separation from attachment figures or fear of such a separation occurring. Thus a foundation will be provided for later assumptions regarding adult manifestations of separation anxiety. Finally, the available literature concerning the continuance of separation anxiety into adulthood will be reviewed, along with a discussion of the life-long impli- cations of the separation-individuation developmental pro- cess. Theoretical Considerations Bowlby (1960) outlines the theoretical basis for the original experience of separation anxiety, as seen by the most prominent theorists who speculated about the nature of anxiety. He also laments the gross neglect of separation anxiety in theory and research which results in a dearth of solid theoretical formulations. He further states that this situation has prevailed additionally due to the lack of concensus among theoreticians on what anxiety is and means. As a result of this state of affairs we find, when we come to consider how analysts conceive of separation anxiety, some widely differing formulations; for each formulation is strongly influenced by the particular outlook regarding the nature and origin of anxiety which the analyst happens to have. Moreover, the place given to separation anxiety within the wider theory of anxiety varies greatly. For some, like Hermann and 13 Fairbairn, separation anxiety is the most important primary anxiety; for others, like Melanie Klein and her associates, separation anxiety is deemed to be secondary to and of less consequence than other and more primitive anxieties. (p. 92) In reviewing the major existing theories, Bowlby does not concur with most of the premises set forth by others. He disagrees with Rank and Greenacre, both of whom believe that the first experience of separation anxiety occurs as part of the "birth trauma" and is subsequently reproduced from this prototype. He states that research in the area of early infancy has shown that babies are protected during birth and early life by a remarkable "in- sensitivity" to psychic trauma. These findings obviously are at odds with any theory regarding "birth trauma." And Bowlby has concluded that it is unlikely that the infant can experience separation anxiety of any type before the age of about 28 weeks. Jones and Spitz, in accordance with Freud's signal theory, believe that the infants' experience of anxiety upon losing contact with its mother serves as a safety device. This mechanism is activated by the infant's aware- ness of the mother's absence, incites a cry of protest, and thereby functionally attempts to insure that she will not be gone for long. While Bowlby feels that the initial anxiety might well come to serve secondarily as such a warn- ing or "signal" of something worse, he feels that it is not and cannot be its primary function. 14 Instead, it is thought of as an elemental experi- ence and one which, if it reaches a certain degree of intensity, is linked directly with the onset of defence mechanisms. It is because of this, and because I wish to distinguiSh it sharply from states of anxiety depen- dent on foresight, that I have termed it Primary Anxiety. (p. 93) Referring to the work of Melanie Klein, separation anxiety is seen as resulting either from the infant's belief that he or she has eaten or otherwise destroyed the mother when she is not present (Depressive Anxiety); or from the infant's belief that the mother has left because she is angry and wishes to punish him or her (Persecutory Anxiety). Bowlby agrees with Klein regarding the importance of per- secutory and depressive anxieties, but states that they are not primary--rather they serve to intensify existing separa- tion anxiety. . . . for anything which leads the child to believe he either has destroyed or alienated his mother, or may do so, cannot fail to exacerbate his expectant anxiety of temporary or permanent separation. (p. 95). He qualifies this position, however, by stating that this exacerbatory effect is only possible later, when the infant has developed the capacity for using symbols and is capable of foresight. Bowlby's own theoretical position regarding the origin of separation anxiety states that it appears as part of an instinctual response system for survival, just as attachment components do. . . . separation anxiety is the inescapable corollary of attachment behaviour--the other side of the coin. As soon as the instinctual response systems mediating 15 such behaviour have matured and, by a process of learning of a simple kind, become oriented towards any object whatsoever, the child will become prone to experience primary anxiety at separation from it. (1960, p. 102) He also says that as the individual matures there is a "diminution in the readiness with which the response systems are activated" and the person becomes more ready to accept substitute figures and is less vulnerable. He further states that he has concluded from his observations that the human infant reaches a stage of decreased vunerability to this primary anxiety at about two years, nine months-- although this varies with individuals. (Note that this coincides with Mahler's timetable for the cessation of the normal separation-individuation process, which will be dis- cussed later.) Regarding the first experience of anxiety as the infant's response to separation from the mother, Mahler has addressed this issue minimally. That is, she does not con- cern herself with the function of early anxiety. Rather she focuses primarily on developments that are initiated following the first few months of life. However, she does briefly discuss early infant functioning, accenting the necessity for the infant's adapta— tion to the mother. Such adaption, says Mahler, is a pre- requisite for a harmonious mother-child relationship. (Con- versely, she states that incapacity for such adaption characterizes the child who will later be labeled psychotic.) Mahler states that this molding of the infant to the mother 16 is the most important early function, and that most mechanisms operating at this time are biological rather than psychologi- cal, such as early anxiety, and serve to facilitate this essential adaptation. Metapsychologically, the focus of the dynamic point of view--the conflict between impulse and defense--is far less important in the earliest months of life than it will come to be later on, when structuralization of the personality will render intra- and intersystematic conflicts of paramount importance. Tension, traumatic anxiety, biological hunger, ego apparatus, and homeo- stasis are near-biological concepts that are relevant in the earliest months and are the precursors, respec- tively, of anxiety with psychic content, signal anxiety, oral or other drives, ego functions, and internal regula- tory mechanisms (defense and character traits). (1975, P- 5) While Mahler et a1. use the qualifying term near- biological, their views appear to be congruent with Bowlby's regarding the operation of an instinctual response system during this early period. Organismic survival is therefore seen as the purpose of anxiety as a response to early infant separation experiences. While Freud is noticeably absent heretofore in this theoretical review, and is not a major contributor to separa- tion anxiety theory, he did make reference to the origin and primary nature of this phenomenon. In The Problem of Anxiety (1936/1963) he describes the infant's longing for the mother and anxiety as a manifestation of this longing and an ex- pression of helplessness. Anxiety thus seems to be a reaction to the per- ception of the absence of the object (mother), and there at once springs to mind the analogies that castra- tion anxiety has also separation from a highly valued 17 object as its content and that the most basic anxiety of all, the "primal anxiety" of birth, arises in con- nection with separation from the mother. (p. 75) It is unfortunate that Freud chose to overlook this insight and, instead, concentrates so heavily on the later Oedipal period. Here he noted the similarities between separation and castration anxieties, but failed in his writings on the Oedipal conflict to make note of separation anxiety as the precursor to later anxieties. In reviewing biographical material, it seems probable that Freud did not explore this further because of his personal preoccupation with his own unresolved Oedipal conflicts. This researcher would be interested to know how he would have reacted if someone had told him that the Oedipal stage becomes proble- matic as part of a general family configuration that includes lack of separateness between mother and child, subsequent eroticization of the relationship; a father who is unable to express his feelings, and a poor relationship between husband and wife. While the basis for these statements will be dis- cussed in chapter 3, the point is made now so that the reader can be thinking about how this material fits in with classical Freudian thought; and to consider the strong possibility that without early pathological separation anxiety there is no significant Oedipal conflict--the Oedipal conflict is secondary to the more basic problem of early attachment and differentiation and deve10ps later as a manifestation of an existing problem in this area. 18 One of the earliest proponents of object-relations theory, Fairbairn, presents a unique explanation of separa- tion anxiety that focuses on the aspect of early dependence. He states that from very early in life the child needs to know that he is loved by his parents and that they also can accept his love. In the absence of such assurance his relationship to his objects is fraught with too much anxiety over separation to enable him to renounce the attitude of infantile dependence, for such a renunication would be equivalent in his eyes to forfeiting all hope of ever obtaining the satisfaction of his unsatisfied emotional needs. (1952, p. 39) Thus Fairbairn sees separation anxiety as a manifestation of infantile dependence that was originally aimed at the mother figure, but if maintained, transfers to the father and subsequently to all later identifications. He continues to relate in a symbiotic fashion; you are part of me, and vice versa. And its function continues to be the same: to provide an attachment figure for identification and need gratification purposes--to have a sense of self and provide for survival. In accordance with Fairbairn's theory the individual, whether toddler or adult, could never move from this position of infantile dependence until he establishes an individual identity and believes that he is loved. Guntrip (1969) also stresses the need for an established sense of unity and security between mother and child, essential especially during the first year. He refers to the infant's early vulnerability and alludes 19 to a symbiotic stage in infancy as being fundamental to the organism for survival. We begin as literally a part of another person, of "nature," and only if that basic sense of unity and the security it gives persists as a permanent unconscious foundation, can we sustain separateness and individu- ality without anxiety. The psychic factor that under- lies everything else, the experience of identification, of oneness, of the sense of belonging, of not being an "isolate," is the psychological replacement for the security of the original literal organic oneness with the mother and "nature." (As an interesting aside, Guntrip states that this is the basis of religious belief and practice.) Guntrip calls the loss of this sense of unity, either through maternal desertion or psychological with- drawal, "deprivation trauma." He says the infant's ego is too fragile to withstand this and psychic damage is inevit- able. Such damage is manifested later in fears of deperson- alization and annihilation and a sense of unreality. Guntrip makes an especially significant contribution when he speaks of separation anxiety as a "negative con- cept." He notes that what it describes is what happens when relationships fail--but says nothing about the nature of positive early relationships and development. This void, of course, has been partially filled by Mahler's theory of the normal separation-individuation process. But, nonetheless, separation anxiety has been a vaguely defined and illusory concept since its invention. Guntrip's point is well-taken when he states that it does not describe 20 basic attachment mechanisms. Nor does it currently describe which mechanisms have failed or how. Summary Basic theoretical positions regarding the nature of separation anxiety as seen by Bowlby, Mahler, Spitz, Klein, Freud, Fairbairn and Guntrip have been presented. While some of these writers believe that separation anxiety has a "primary" function and others see it as "secondary," there is agreement regarding the basic nature of this phenomenon, i.e., it is an integral part of early mother- child relations and trust-building in the organism. The extent and intensity of separation anxiety and its crucial ramifications for identity establishment are especially emphasized by Fairbairn and Guntrip, who maintain that the repercussions of early developmental failures can be ob- served throughout an individual's life. CHAPTER 3 REVIEW OF THE LITERATURE INTRODUCTION Both of the major theorists on the topic of separa- tion, Mahler and Bowlby, View actual absence(s) of the attachment figure during critical periods, threats of separation or abandonment, and rejection by the attachment figure as being precursors of separation anxiety. Each has focused primarily on a single etiological factor--Bowlby on physical separation and Mahler on the early relationship between mother and child. While each of them seems reluctant to acknowledge the overlap between their theoretical positions, it seems evident that their work has been parallel to a great extent. This is especially true with regard to the discussion of separation anxiety symptoms. When Bowlby (1960), Spitz (1955), and Heinicke and Westheimer (1965) discuss the behaviors and beliefs operative in a child who has been deprived temporarily of his mother's presence, they are also describing the symptoms that Mahler (1952, 1955, 1961, 1963, 1971, 1972) says results from an inadequacy in attachment and/or the separation-individuation process. Each of these 21 22 two approaches will be reviewed and discussed in the first section of this chapter. Additionally, the concept of school phobia will be reviewed briefly and selectively. This is not a literature review of this topic, since this would go far beyond the issue of separation anxiety. Its mention here is for the purpose of establishing the link between school phobia and separation anxiety, so that this relationship is clear to the reader. Case histories of school phobic children and their families contribute to the practical understanding of separation anxiety and are a valuable source of knowledge regarding common attitudes, behaviors and somatic complaints. The second section of this chapter deals more speci- fically with the psychodynamics of separation anxiety, rather than concentrating on the theoretical components of attachment and separation. The type of family interrela- tionships that produce separation-anxious individuals are discussed here, as well as crucial aspects of development and specific characteristics of the separation-anxious per- sonality. The third section of this chapter contains a dis— cussion of the consequences of childhood separation anxiety for adult adjustment. This includes a consideration of the likelihood of the continuation of early separation anxiety into adulthood, some possible mitigating influences, and speculation regarding the translation of childhood symptoms into adult behavior. The word speculation is used since 23 three is no research data available currently on this topic. Finally, the concept of separation-individuation as a life- long process will be discussed. MAJOR THEORETICAL APPROACHES TO SEPARATION ANXIETY The Work of Bowlby Bowlby is unquestionably the foremost authority on the effects of maternal deprivation and one of the earliest writers on the topic of separation anxiety. Bowlby's re- search has been devoted to the study of the effects of physical separation of mother and child. His focus has been on early childhood, particularly the first three years of life. Protest! Detachmenti Despair As a result of extensive observations of young children in institutional settings Bowlby (1958, 1973) formulated a triadic sequence of responses that inevitably occur when a child is separated from its mother for a pro- longed period of time. The first stage is one of protest, which lasts from a few hours to a week or more. In this stage the child screams and cries and is generally incon- solable. Following the end of the protest period, the child becomes withdrawn, depressed, and hOpeless. Bowlby has named this stage despair; and it is essentially assumed to be the period of grief and mourning for the absent mother figure. 24 The next stage, detachment, says Bowlby, ". . . sooner or later succeeds protest and despair and is often welcomed as a sign of recovery." (1960, p. 90) In this stage the child is again able to interact with others, and show an interest in the world outside himself. However, he has accomplished this through utilization of the defense mechanism of denial. That is, he does not allow himself to consciously feel anything for his mother so that he can avoid the experience of pain. Children who are reunited with their mothers after having reached this stage typically act indifferently and apathetically toward them. This stage of detachment gives way to renewed or new attachments, of course, if such a possibility presents itself. Bowlby points out, though, that repeated abandonment by attachment figures can result in a child who is more or less permanently "detached." The existence of the stages of protest, despair, and detachment postulated by Bowlby is based to some extent on the work of Freud and Burlingham (1944), who observed infants and children placed in nurseries in wartime England. Many of these children experienced repeated changes of mother figures and their resultant anxieties and depressions were described in case study form. Supporting Evidence These stages formulated by Bowlby have been con- firmed by other researchers. Spitz's classic article on anaclitic depression is a study of the reaction of children 25 to separation from the mother between the sixth and eighth month of life (this period coincides with Mahler's "symbiotic stage of development"). In almost all cases the mother and child were separated for a continuous period of three months. The very obvious "anaclitic depression" and its symptoms noted by Spitz (1946a) compare well to Bowlby's stage of "despair" and can thus be viewed as essentially the same phenomenon. Further substantiation can be noted in Heinicke and Westheimer's book, Brief Separations (1966). They describe the behavior of ten children who have been temporarily placed in nurseries, most frequently because of the birth of a sibling. They were further observed upon their return to their parents, adding new information regarding "recovery" from a separation experience. While particulars of this study will be noted later, the results served to confirm Bowlby's observations, and additionally contributed more specific information regarding individual differences. Robertson and Robertson (1971) attempted to deter- mine if the stages of protest, despair, and detachment could be prevented through effective intervention by a mother- substitute. Their findings showed, however, that these are necessary consequences of separation and could be ameliorated but not eliminated. That is, the children in their study showed various signs of all three stages, even with a caring mother substitute. 26 Additional evidence for Bowlby's sequential response to separation is seen in the study done by Ainsworth and Bell (1970). In addition to supporting Bowlby's formula- tions regarding attachment and reactions to separation, they documented the "search" aspect of the protest stage. That is, they systematically observed that children old enough to possess physical motility actively search for the departed mother as well as cry in protest. Bowlby (1973) responded to those who generally accepted his hypotheses and the research results, but who doubted the universality of the phenomenon. At one time it was supposed by critics of our thesis concerning separation that distress seen in a child during a period away from mother, and increased ambiva- lence and anxiety seen after it, must betoken an un- favourable relationship between child and mother before the event, or reflect perhaps a child's anxiety about his mother's pregnancy or illness. Yet observations of healthy children from thoroughly satisfactory homes, who are separated from mother for one of many differ- ent reasons, show that, whatever contribution other variables may make, when a young child is in a strange place with strange peOple and with mother absent, pro— test, despair, and detachment still occur. (pp. 15-16) Regardless of theoretical position, researchers such as Bowlby, Ainsworth, Heinicke, and Freud and Burlingham, have confirmed the existence of attachment beginning in early infancy and shown that separation from the mother during approximately the first three years of life causes separation anxiety and its concommitants to appear and can lead to extreme psychopathology if not ameliorated within a reason- able time span. Also, Harlow (1958) has produced similar reactions in monkeys through physical removal of the mother 27 and documented the nature of attachment through his experi— ments with mother surrogates. Harlow's famous studies regarding the quality of mothering provided evidence for the essential nature of normal attachment and consequences of separation during critical developmental periods. Not only will prolonged physical separation from the mother cause developmental problems for the child, but shortcomings and inadequacies in the day-to-day process of attaching, learning to relate, and becoming independent, can do the same. .The Work of Mahler Pursing this thought leads to the problem of separa- tion anxiety as presented by Mahler (1961, 1963, 1972, 1974, 1975). While Bowlby has concentrated on the effects of sustained physical separation of mother and child, Mahler has focused on the nature of their early relationship. She postulates that the child goes through stages of attaching to the mother and then separating from her to achieve an individual identity; this is part of every child's develop- ment and is referred to as the "normal separation- individuation process." Any significant breakdown in this process, which gggld include the actual loss of the mother, results in pathological separation anxiety. It can be seen, however, that Mahler's primary emphasis is on normal development rather than pathology, even in light of her extensive attention to the problem of child psychosis (1952, 1955, 1971). In terms of pathological 28 states that do result from inadequacies in the separation- individuation process, she is more concerned with aspects of the unhealthy mother-child relationship rather than with the incidence of physical separation. Link Between Mahler and Bowlby Thus it would seem on the surface that Bowlby and Mahler have little in common. But Bowlby (1973) has acknowledged sources of separation anxiety other than actual loss of the attachment figure, even though they receive secondary attention. . . . an excess of separation anxiety may be due either to an experience of actual separation or to threats of separation, rejection, or loss of love, (this) can be adopted with confidence. (p. 106) In the same publication he also notes that it is not uncommon for mothers to use the threat of separation or abandonment as a disciplinary tool. This is often seen in the form of threatened love withdrawal as a type of punish- ment, which Bowlby equates with threat of separation, since in the post-toddler age bracket they have essentially the same meaning for the child. That mothers more often use this unconsciously, rather than consciously, is noted as well. Bowlby summarizes what he believes to be the major determinants of separation anxiety, adding that this list is by no means exhaustive. 1. One determinant is undoubtedly the actual experi- ence of separation. 2. Another determinant is the excessive use by parents of threats of separation or withdrawal of love as sanctions. 29 3. Another is the child's experience of rejection by the mother, especially where her positive feelings are mixed with unconscious hostility. 4. Another is any actual event, such as a parent's or sibling's illness or death, for which the child has come to feel responsible and, therefore, guilty and unloved. (p. 107) Determinants two and three especially emphasize the nature of the mother-child relationship and illustrate the connection between Bowlby and Mahler. While Mahler pays little attention to actual separation in her discussion, the studies conducted by her and her co-workers (1975) utilize nursery setting observations of mother-child brief separations and reunions. This is virtually indistinguish- able from the research framework used by Bowlby's colleagues, Heinicke (1965) and Ainsworth (1970). Thus it can be readily seen why their observations and conclusions regarding separation anxiety are so similar. In fact, the main ob- stacle between these two groups of researchers appears to be their reluctance to acknowledge one another and integrate their findings. (This will be attempted in the second section of this chapter.) Separation-Individuation Process Having established the link between these approaches, it would be expedient to present the theory espoused by Mahler et a1. (1975) in further detail. To begin, "separa- tion and individuation are conceived of as two complementary developments: separation consists of the child's emergence from a symbiotic fusion with the mother and individuation 30 consists of those achievements marking the child's assumption of his own individual characteristics." (p. 4) According to Mahler, this process spans from about the fifth or sixth month of life to the thirtieth to thirty-sixth month, and is represented in four stages. Preliminary phases of development. The four to five months prior to the onset of the separation-individuation process are divided into the "autistic" and "symbiotic" phases. The autistic stage occurs during the first few weeks of life when the infant spends a good deal of his time sleeping and eating. Mahler et a1. (1975) says that "physiological rather than psychological processes are dominant" during this phase, with primary emphasis on responding to internal rather than external stimuli. This is followed, at approximately the second month, by a dim awareness of the mother's existence, but not as a separate entity. This is called the "symbiotic" phase by Mahler because "the infant behaves and functions as though he and his mother were an omnipotent system--a dual unity within one common boundary." Thus the child is absolutely dependent and does not differentiate between himself and his mother. The infant's body is very sensitive during this period and important experiences of being touched and held lay the foundation for formulation of the body image in later stages of development. What both the autistic and symbiotic phases have in common is that they are periods when the infant is 31 "undifferentiated"--when he has no real sense of self or others as independent beings. He experiences everything globally in terms of need satisfaction--or frustration. At the peak of symbiosis with the mother (approxi- mately four to five months of age), the child enters the first of four subphases that comprise the normal separation- individuation process: Differentiation and the Deve10pment of the Body Image. The first subphase, "differentiation." The first subphase (approximately the fifth to tenth months), "differ- entiation," begins at the peak of normal symbiosis and is marked by a decrease of physical dependence. The child begins independent locomotion, and engages in sensorimotor exploration. Beginning awareness of the mother as a separate person as well as critical "bonding" to her are demonstrated by the increased desire for observation and the specific smile in response to her (Bowlby, 1958). It is during this period that what Mahler et a1. (1975) calls the "hatching process" begins and develops. At around six months of age the child can be observed pulling at the mother's hair and clothes, and exploring her facial features with his hands. He is also likely to pull away slightly from her body when being held--so that he can get a good look at her. The child's curiosity extends to his own body as well, of course, and is illustrated by the familiar sight of an infant discovering his own toes, realizing that they belong to him, and squealing with 32 delight at the prospect. Through these repeated visual and physical contacts, the child establishes a separate and individual body image. When not in close physical proximity to the mother, the child will often satisfy his desire for contact with her through acquisition of what Winnicott (1953) has labeled a "transitional object." Items falling into this category are the familiar security blanket, Teddy bear, or other soft, pliable objects. It is referred to as "transitional" because it symbolically offers the comfort of the mother's body and closeness while the child ventures away from her. At some point during this first subphase the child also shows a pronounced curiosity in response to other-than- mother persons. He may also alternately experience a keen "stranger anxiety" or "eight months anxiety," as observed by Spitz (1950). However, Mahler (1975) takes exception to Spitz's conclusion that this is a universal experience and always occurs at around eight months of age. Her obser- vations have revealed a wide variety of individual differ- ences, both in the age at the onset of anxiety and in its intensity. She notes that in some children it is hardly noticeable at all (pp. 56-57). Though she does acknowledge that for some other children, expressed fear at being approached by a stranger is the most marked behavioral manifestation occurring in this subphase. The second subphase, "practicing." The second sub- phase (approximately 10 to 15 months) is referred to by 33 Mahler as the "practicing" period and is characterized by the development of independent locomotion and the first departures from the mother's side. Toddling, exploring, yet constantly returning to touch base with the mother during this period has been termed "emotional refueling" by Furer (unpublished, cited by Mahler, et al., 1975). It is easy to understand the significance of walking as it pertains to the separation-individuation process. Mastery of this task enables the child to make his 9gp decisions about what direction he wants to go in, and who he wants to approach--or avoid. Mahler, et a1. (1975) made an interesting discovery regarding the child's entrance into the world of locomotion. Quite late in our study we came to realize that it is the rule rather than the exception that the first unaided steps taken by the infant are in a direction gyay (emphasis mine) from the mother or during her absence; this contradicts the popular belief . . . that the first steps are taken toward the mother. The signi- fiance of this phenomenon bears further study. (p. 73) The third subphase, "rapproachment." The third sub- phase (approximately 16 to 21 months) is the "rapprochment" period, and the one that is deemed the most critical by Mahler. During this stage the child's awareness of his separateness is heightened by his mastery of walking and basic speech. While he is becoming increasingly indepen- dent, he also becomes very concerned with his mother's presence and approval of his new status. He learns almost simultaneously to say "no" to her and to "woo" her. The ambivalance inherent in this stage is further illustrated 34 by the toddler's habit of first "shadowing," then darting away, with the expectation that he be chased and caught. It is during this developmental period that the child learns (or fails to learn) that separation is not equivalent to loss of love. This is accomplished through parental accep- tance and promotion of separateness, while still maintaining a loving relationship. The confusion experienced by both mother and child during this time has been noted by Mahler et al. The mother is confused by the child's seemingly contradictory behavior of being more independent, yet also being very demanding and insisting on sharing all his experiences with her. (This can be seen in the "I want it (or you) and I want it now" phenomenon.) For the child, the confusion stems from his desire to venture independently into the world, while still maintaining his former Utopian "oyster," inhabited blissfully by only he and his mother. The junior toddler gradually realizes that his love objects (his parents) are separate individuals with their own personal interests. He must gradually and painfully give up the delusion of his own grandeur, often by way of dramatic fights with mother--less so, it seemed to us, with father. This is the crossroads that we term the "rapproachment crisis." (1975, p. 79) Mahler notes that the formation of specific gender identity and expansion of social relationships to include the father and others, are also characteristic of develop- ment during this subphase. These can be seen as natural consequences of observing and becoming involved in the 'world outside the mother-child duality. This, along with 35 awareness of his own body, leads to an awareness of sex differences that is incorporated into the child's forming sense of self. Awareness of others is also the first step in re- lating to them. So the child is ready to see and experience what can be offered by other-than-mother figures. The importance of the mother's acceptance of this new indepen- dence and involvement cannot be overstated. It is here that she either encourages the child to "spread his wings" or binds him to her out of her own need. The fourth subphase, "individuation." The fourth subphase (approximately 22 to 36 months) is the period of individuation and object constancy. The child learns to unify the concepts of "good" and "bad"; begins to actively relate to adults other than his mother and to peers; experi- ments with separation, and learns to express himself ver- bally. The child additionally engages in more purposeful play and enters the world of make-believe through a develop- ing ability to fantasize. In fact, the attainment of cogni- tive skills is a hallmark of this subphase. A sense of time (and also spatial relations) begins to develop and with it, an increased capacity to toler- ate the delay of gratification and to endure separation. Such concepts as "later" or "tomorrow" are not only understood but also used by the child of this age: they are experimented with, polarized by his mother's comings and goings. (Mahler, et al., 1975, p. 116) However, the most important psychic development of this period is the attainment of emotional object constancy. 36 This refers to the extent that the child has internalized a "good mother" image, providing him with the security that comes from believing that he is loved and that his needs will be met--even when the actual mother is absent. Mahler says the "essential prior determinants" to achieving this are the develOpment of "trust and confidence." That is, if the child's basic physical and emotional needs have been promptly and continually met, he will have no reason to doubt that this will continue, and he will approach the world with this belief. Regarding this relationship between mothering and trust-building, Erickson (1963) states: The infant's first social achievement, then, is his willingness to let the mother out of sight without undue anxiety or rage, because she has become an inner cer- tainty as well as an outer predictability. . . . But let it be said here that the amount of trust derived from earliest infantile experience does not seem to depend on absolute quantities of food or demonstration of love, but rather on the quality of the maternal relationship. (pp. 247-249) Mahler (1955) stresses that all of the four sub- phases described here are crucial to later develOpment, as are the preceding autistic and symbiotic phases. A strong and adequate symbiotic phase is a pre- requisite for subsequent successful disengagement of the human infant as well. Only if symbiosis has been adequate, is he ready to enter the phase of gradual separation and individuation. The aim and successful outcome of this individuation process is a stable image of the self. (PP. 196-197) The model presented by Mahler of the separation- individuation process is an additive one, in that the necessary prerequisite for healthy development in each stage is successful completion of the preceding one. Failures or 37 breakdowns in earlier stages consequently are graver than those occurring in later stages. However, this does not minimize the importance of the entire process in forming a healthy individual. Inadequacies in this develOpment produce the manifestations of separation anxiety and the character— istic psychodynamics that underly it. Discussion of specific aspects of this personality and etiological factors involved will appear in the second section of this chapter, following a discussion of the final body of literature devoted to the t0pic of separation anxiety --that dealing with school phobia. Summary The work of Bowlby, Mahler, and others has been reviewed. Bowlby reports certain symptomatic reactions that occur during early separation experiences that coincide with many behaviors cited by Mahler as being indicative of a problem in the normal separation-individuation process. Additionally, both of these major theorists define the period of attachment and attainment of individual identity as being from birth through approximately the third year of life. Separations, threats of separation, or obstacles to satisfying attachment during this period are said to have a major negative impact on normal development. School Phobia The final approach to the topic of separation anxiety stems not from child development research or theory, but 38 primarily from everyday practice. As stated previously, this refers to the literature concerned with school phobia. What this literature lacks in theoretical foundation, it compensates for in providing informative family histories and behavioral observations on children from nursery age to high school. Relationship to Separation Anxiety School phobia has for some time been generally accepted as a manifestation of separation anxiety, except in cases where the child's strong fear is based in some real physical or emotional trauma. In fact this was esta- blished in the literature and emphasized quite heavily in publications from the 1950s, when this was made and reported. Estes, et a1. (1956) explain that Because separation anxiety is manifested so fre- quently by refusal to attend school, the term school phobia often has been used to designate this condition. This term is still in popular use. However, the term emphasizes a common symptom rather than the underlying true nature of the disorder, which is anxiety associated with leaving the mother for any reason, rather than simply a fear of school. (p. 682) Vaughn (1954) and Waldfogel et al. (1956) make similar statements regarding the nature of the school phobic child's fears. McDonald and Shepard (1976) reviewed the psycho- analytic view of school phobia as a "consequence of a close symbiotic mother-child relationship in which the mother is overly protective and the child is excessively dependent" (p. 297). Suttenfield (1954) notes that this dynamic can 39 be seen as it is recreated in the classroom. That is, children who are able to manage their anxiety and remain in school, or those who return to school, often attempt to duplicate this type of mother-child relationship with the teacher. They are typically demanding and possessive, they interpret the teacher's attention to other students as rejection, and predictably end up feeling unwanted and unloved. Greenbaum (1964) approaches the link between separa- tion anxiety and school phobia somewhat differently. He states that the child uses the school as the focus of his projection because it is one of two places where he cannot leave without it seeming strange or odd. (Greenbaum says the other place is the movies.) Thus his anxiety about something like achievement, which is common to school phobic children, may lead secondarily to a school phobic reaction because the child cannot easily leave. No matter what he is anxious about, he is not free to go home for comfort and solace. Greenbaum does not speculate, however, as to why these children feel the need to go home while most children do not. Such speculation would seem to return to the fact of a stronger-than—usual desire for contact and proximity to the mother. Summary Thus, school phobia can be seen as a symptomatic reaction of children who fear being away from their mothers. 40 The basis of this fear, according to Estes, et al., reflects the Kleinian concepts of persecutory and depressive anxities. That is, the child fears that his mother may leave because she is angry with him or does not want him—-and he fears that his hostility toward her for not allowing him to be separate and independent may destroy her while he is at school. While it might be argued that such motivations have not been proven, it is generally acknowledged that the basic fear experienced by the school phobic child is that of being separated from his mother. SEPARATION ANXIETY: ETIOLOGY AND DYNAMICS While the concept of separation anxiety as a psy- chological entity has been established, the particulars of the mother-child relationship and general family con- stellation have not. These will now be discussed, as well as specific aspects of the separation-anxiety psychodynamic. Mother-Child Relationship The typical relation between mother and child seen in cases of separation anxiety is said to be a hostile- dependent one. The mother is described as being dependency- prone, having failed to separate from her own mother. And her personal struggle to be independent, combined with her unmet dependency needs, result in a subsequent attitude of ambivalence regarding motherhood. She wants the emotional closeness and satisfaction that can be derived from motherhood, but she also wants 41 someone to take care of her--thus, she often feels burdened and overwhelmed by parental responsibilities. The maternal behavior that results from these attitudes is a clinging dependency on the child-—a reluctance or refusal to allow him to separate--and a demand that he take on the responsi- bility of satisfying her needs and making her happy. Both this infantalizing and pressure into undertaking an abnormal amount of adult responsibility are aspects of a "symbiotic" relationship, where the child is not allowed to individuate. While this sounds very destructive, and in many cases is, it does not mean that these mothers do not care for their children. It simply means that they are not capable of a relationship that is not based on need. And the negative consequences of this on a child's forming ego are considerable. Mahler (1963), however, acknowledges the difficulty inherent in supplying closeness, while at the same time giving the child room to grow. Striking a balance between mothering without undue frustration on the one hand, and without intrusion or stifling of the infant's individual, inborn rhythym of needs on the other hand, is a task not easily achieved by the average mother in our culture. (p. 310) In accordance with this statement, no mother ever completely succeeds in facilitating true differentiation of the child from her. But in cases of debilitating separa- tion anxiety, such as school phobia, the lack of differ- entiation between mother and child is much more pronounced than usual. In the most extreme cases where the child remains symbiotically bound to the mother, borderline 42 syndrome or actual psychosis are the result (Mahler, 1974; Masterson, 1973). Repetition of Relationship Through Generations The aspect of a repeated cycle of mother-child dyna- mics, through generations, is of particular interest in examining the problem of separation anxiety. Benedek (1949) states that: The behavior manifestations which are usually accessible to psychoanalysis reveal that the woman's identification with her mother motivates her attitude toward motherhood and determines her behavior toward her own children. (p. 642) The specific form of this maternal cycle has been noted by Masterson (1973),who has written extensively on the border- line syndrome and the problem of separation anxiety at its core. The mother of the borderline patient suffered from a borderline syndrome herself. Having been unable to separate from her mother, she clung to her child to foster the symbiotic relationship, discouraging moves towards individuation by withdrawing her support. The child needed her supplies to grow; however, if he grew the supplies were withdrawn. (p. 331) Wold (1973), in reporting the family structure in cases where separation anxiety was the primary dynamic, states: "The mothers were unable to separate from their own mothers and could not refuse them favors" (p. 1396). Admittedly, these examples deal with cases of fairly extreme psychopathology. Nonetheless, they illustrate the repetitive and constraining nature of the basic relation- ship that engenders separation anxiety. 43 What is especially important to remember is that the mother in this type of relationship is usually quite "devoted" to her children and often very solicitous towards them. In fact, in less severe cases she may often be quite nurturant. It is part of the nature of the problem that the ambivalence she feels regarding motherhood, the feelings of hostility and desire for dependency, are largely repressed --and therefore, unconscious. Mahler (1974) states that: . . . the optimal evolution of the infant's partial ego functions . . . is either facilitated or hindered by the conscious and, more particularly, the unconscious attitudes of the mother. (p. 310). Hostility What has been determined to be the most crucial element of this unconscious attitude is hostility. This repressed hostility pervades the interactions between mother and child and is characteristic of their interactions with others, as well. Waldfogel, et a1. (1957) point out that the mother denies this hostility and often overcompensates by being overly protective and doting on the child. But they note that the child has a harder time dealing with his hostility. The child, however, must not only deny his mother's anger, but his own as well. In an atmosphere where rage always carries with it a destructive connotation, and where he is so totally dependent upon his mother for support and protection, he is left little alterna- tive but to displace his anger onto some other object or person. Thus, by displacing his anger, he protects himself from the terrifying prospect of being alone and helpless. 44 In studying infantile separation anxiety, Tennes and- Lampl (1966) became aware of the tremendous significance of hostility in the separation-anxiety dynamic. In reporting the results of their study of children and their mothers they state: The best predictors of the intensity of infantile separation anxiety were found to be the mother's inhibi- tion of the child's aggression and the mother's hosti- lity toward the child. (p. 436) Additionally, researchers report aggressive reactions to separation from the mother as typical (Burlingham & Freud, 1944; Robertson, 1958; Bowlby, 1960; and Heinicke, 1966). In each of these studies the children who were separated from their mothers were significantly more aggressive (hostile) in their behavior toward other children and adults, than were those who were not separated. Feshback (1970), in reviewing the literature on aggressiveness in children, reported the following: Comparisons of family backgrounds of overinhibited and aggressive children . . . reveal a history of severe maternal rejection in the aggressive children. (p. 216) Thus both physical separation and a mother-child relationship characterized by inhibited hostility and dependency are shown to produce separation anxiety in the child. But what do these two situations have in common other than their stimulus value? Both threaten the child with helplessness. If a young child is separated from his mother, he naturally feels helpless and afraid. He knows that he cannot take care of himself and fears for his 45 emotional and physical well-being. The same could be said for a child locked into a symbiotic union where his needs are not met; he experiences his mother's ambivalence toward him, and his sense of self is precarious. He is thus con- stantly threatened by feelings of helplessness--even fear of annihilation. In either instance, these feelings of anxiety and loss of self are countered by defensive aggres- sion. Symbiosis and Aggression The link between symbiosis and aggression, in parti- cular, was noted by Greenacre (1952): "The longer the child is treated as though only a part of the parent, the greater will be the aggression against the parent" (p. 171). Bowlby (1973) cites a dramatic case study of matricide reported by Burnham. In this instance the adolescent who murdered his mother said afterwards, "I couldn't stand to have her leave me" (p. 251). It is interesting to note further that even the physical manifestation of a symbiotic mother-child relation- ship has consequences in terms of generating aggression in the child. This refers to the habit of some mothers of constraining the child by holding him and cuddling him, even when he has progressed chronologically to the stage where he would normally be crawling and exploring. Or it can be seen in the case of the parent who carries a toddler almost everywhere, even when he is fully capable of walking. 46 The relationship between this physical restraint and aggres- sion is postulated by Spitz (1946) in his explanation Of the circumstances leading to depression. From the dynamic point Of View locomotion and motility in general fulfills the important task Of offering a necessary channel Of release for the aggres- sive drive. When motor anxiety is inhibited in in- fancy, all normal outlets Of the aggressive drive are blocked. In this case only one alternative remains for dealing with the aggressive drive; that is, to direct it against the self. (p. 334) Thus, Spitz demonstrates the significance Of aggression, even in infancy. Hostility and Rejection Bowlby (1960) has also discussed the place Of aggression in separation anxious children, noting that rejection or the expectation Of rejection is a powerful stimulus to hostility. He goes on to say that if the child expresses this hostility toward the parent this only in- creases the likelihood that he will be rejected, thereby creating a vicious cycle. His Opinion regarding the rela- tionship between hostility and separation anxiety is clear: "In my View both an excess Of separation anxiety and an excess Of hostility are very commonly provoked by the same experience" (p. 107). Finally, regarding aggression, Greenbaum (1964), in discussing the treatment Of school phobia, reports that "for six Of the seven patients who returned to school, the return followed shortly after the expression and analysis Of the anger toward the mother" (p. 626). 47 Summary In summary, it can be seen the typical mother-child relationship that precipitates separation anxiety is one Of hostile-dependence. Additionally, there is a pattern Of the mother having received the same treatment from her own mother, and being fatefully doomed to repeat it. Mothers in these situations are said to be immature, to have their own unresolved dependency needs; to be ambivalent about having and caring for children; and to have failed in separating from their own mothers. The relationship between the mother and father in cases Of separation anxiety will be discussed following an examination Of the role Of the father. Role Of the Father SO far it would seem that separation anxiety is totally a matter between mother and child. But this is not the case. The father has not yet been mentioned for two main reasons: (1) In almost all cases the person to which the child bonds in early life is the mother, and (2) the father has received scant attention in the literature, leaving little tO report. What is Often said about the father is that he must be important in the life Of the child, but no one knows exactly how or why. Most of the material that is available deals with the father's influence on achievement and sexual identity, particularly with regard to daughters. While 48 this does seem to accurately reflect society's historical preoccupation with men as purveyors Of power and sex, it sheds little light on the father's role in child development --both normal and abnormal. Naturally, this includes the topic at hand--separation anxiety. View of Father's Role by Mahler and Others The place Of the father in the separation- individuation process has also been largely ignored by Mahler. However, the discussion that has been directed to this topic has a general theme running through it: that the father mediates between the mother-child pair and the world outside. Mahler and Gosliner (1973) state: We believe the stable image Of a father or of another substitute Of the mother, beyond the eighteen months mark and even earlier, is beneficial and perhaps a necessary prerequisite to neutralize and tO counter- act the age-characteristic oversensibility Of the toddler to the threat of re-engulfment by the mother. (p. 209) Abelin (1971) has done the only study Of the role Of the father in the separation-individuation process as outlined by Mahler. He expresses agreement with Mahler as to the father's function and further adds that the father is Often seen by the toddler as someone who will rescue him from the "bad mother." The conclusions Of Abelin's study Of toddlers and their parents also support Mahler's hypothesis that the relationship with the father begins in the symbiotic phase, although it occurs somewhat later than with the mother or 49 siblings. The attachment to the father progressively in- creases during the subsequent differentiation subphases, but is most conspicuous at the beginning Of the practicing subphase. Abeline also noted that in his subjects the girls made the attachment to the father somewhat earlier than the boys did. Of those writing on this topic Abelin attaches the greatest importance to the father's role. He states: . . . from the fact that the vital task Of a child at that time is to achieve individuation through a process Of intrapsychic separation from the symbiotic mother, it does not necessarily follow that this is an affair between just mother and child. Quite the con- trary, the task might be impossible for either Of them to master without their having the father tO turn tO. (p. 248) While this may be overstating the case, there is no question that the father holds an essential place in the family constellation as it pertains to the separation- individuation process. Abelin's statement Of Loewald's (1951) view is particularly apt. "This early identification with the positive father figure preceded and prepares the way for the oedipus complex" (p. 249). Position Of Father in Relationship to School Phobia Other discussions pertaining to the father are chiefly products Of the school phobia literature and thus relate to the father's position in this pathological dynamic. The general theme here is Of the father as a rival to the child 50 and a man unsure Of his own identity. Waldfogel et a1. (1957) provide a discussion Of the father in relation to the school phobic child. We usually find that the fathers also play a signi- cant role in this pathological constellation. Here our information is more limited, but the cental fact that emerges is that the father, because Of his Own uncer- tain sexual identification, is unable to define clearly his paternal position. He shares the mother's anxious concern for the child and Often vies subtly with her for the maternal role. At times, it is as though the child had two anxious mothers to contend with instead of one. Actually, the mother and father are both looking to each other for gratification Of their own dependency needs. In relation to the child, the father's search for dependent gratification usually takes the form Of over- identification with the child, but sometimes we see him competing as a sibling for mother's affection. (p. 758) Thus fathers are seen to have as great a difficulty with the adult parent role as the mothers Of these children do, and perhaps, greater. This may be related to certain existing social role expectations Of this culture which permit men tO be aloof from their wives and families (especially if it is rationalized by time spent "getting ahead"), and to be demanding Of attention and service; but expects women to face up tO their responsibilities and be self-sacrificing, especially with regard to their children. A study done by Talbot (1957) describes the detached, child-like father as follows: Fathers represented here do not seem to be emancipated from their own families either. One father was so apprehensive that his mother might break her leg or die while alone that he made his daughter leave her play every 15 minutes to lOOk after the grandmother. Fre- quently the mothers inform us that their husbands place them and their children second to the paternal parents and siblings. (p. 288) 51 Father's Typical Role in Family Yamazaki (1973) provides further information on the role Of the father in families with school phobic children. His study consisted Of conducting extensive home and school interviews, and administration of the Roschach Inkblot Test to both mothers and fathers. His conclusions are therefore based on these interview Observations and test interpreta- tion. It should be noted when considering the results Of this study that the school-phobic subjects were all adoles- cents whose problems in this area had persisted for at least four years--indicating an unusually severe problem. Of additional importance is that the study was done in Japan, which may limit the generalizability Of the results Of this culture. In general, Yamazaki found the fathers to these school-phobic children to be reluctant to take any responsi- bilities in the home other than financial ones. They left all Of the major decision—making to their wives and were viewed by their families as ineffectual, particularly in a crisis. They tended to be absent from the home Often, taking on overtime work so that they could avoid domestic responsibilities. This description Of fathers Of school- phobic children was also made by ChOi (1961). Personally, he found them to be unsociable, talking very little and having few or no friends. In the inter- views they were reluctant tO reveal any personal details or 52 express any feelings. In fact, many of the fathers would not come to school for the scheduled interviews, so Yamazaki had to make home visits in order to interview and test them. He also found them to be very secretive about their family life and defensive. The Rorschach protocols reveal the fathers to be highly anxious, with intense underlying feelings Of inferi- ority, repressed dependency, and deep-seated hostility. They also revealed an overall weak identity--particularly with respect to their masculinity. Yamazaki also reported how the wives and children of these men perceived them. The children tended tO see their father as weak, even a figure of sympathy. But they also found him to usually be gentle and good-natured in his limited dealings with them. (However, a rather small per- centage Of these men were less repressed with respect to their hostility and were prone to violent outbursts.) The wives Of these men tended to lOOk down on them and belittle them in front Of the children. They felt abandoned and let down by their husbands, seeking gratifi- cation from their children instead. They were usually over- protective toward them, and tried to infantalize them-- keeping them tied to home. The mothers in these families were also termed egocentric, usually did not get along well with their neighbors, and expressed resentment at being burdened with household responsibilities. 53 Father as Facilitator of Individuation In yet another study Of the fathers Of school-phobic children, Choi (1961) outlined l7 adolescent female cases, emphasizing the father-daughter relationship. Her descrip- tion Of the father runs parallel to what Yamazaki and Wald- fogel reported, but she Offers unique insights into the particulars Of the father—daughter dynamic. She Observed that the fathers Of these school-phobic girls failed to help facilitate their independence, and noted that Goldberg (1953) had made the same Observation: This writer notes, tOO, that none Of the girls in her sample received enough security from their relation- ships with their fathers tO help them resolve their ties to the mothers. (p. 160) Choi also indicated that when the fathers and daughters did make contact, it was Often Of an inappropriate type. That is, fathers Often initiated social activities with their daughters that are typically shared with peers; fathers encouraged their daughters to sleep with the parents when they were young; and fathers Often seemed tO over- eroticize the relationship with their daughters, presumably seeking the attention that was not forthcoming from their wives. In line with these Observations Choi reported that " . . . in more than 50 percent Of the cases the girls' main problems were diagnosed as being Oedipal ones" (p. 177), and in all Of the other cases it had been listed as one Of the significant problems the girls' experienced. 54 Summary In summary, the desired role Of the father in separation-individuation appears to be one Of mediator between mother and child, and supplier Of emotional needs to the mother. In families where separation anxiety is prevalent in the children, the parents typically have a poor relationship (including unsatisfactory sex life); the mother turns to the child to have her needs met; and the father is an emotional isolate who is not available to the child at the time Of individuation when an other-than-mother figure is needed to facilitate the child's independence. Siblings and Related Development Mahler et a1. (1975) noted that siblings are accepted by the individuating toddler prior to the father figure, and can contribute to the individuation process as an avail- able other-than-mother figure. However, in families where separation anxiety exists, the ideal Of this sibling growth facilitation is not met. Sibling Rivalry In reviewing case study material and conducting interviews, it has become apparent that siblings in the separation-anxiety prone family relate poorly to One another. Two patterns have emerged. One is Of intense sibling rivalry, Often to the extreme Of physical fighting into adolescence and early adulthood. The other pattern is one Of symbiotic dependency, a reflection Of the initial 55 attachment relationship. Finally, these two patterns some- times alternate or overlap in the context Of the same rela- tionship; i.e., the siblings may be excessively "devoted" to one another at one moment and hostile, or even physically abusive, the next moment. This hostile-dependent relation has been explained previously in the discussion Of the mother—child relation— ship. The children interact with one another in this fashion also as a result Of early learning that generalizes from the first relationship with the mother. This is the only mode Of interaction that they know. Other factors that bring siblings together symbioti- cally are the birth Of a new child who receives all Of the mother's attention, or some other alienation from the mother. Under these circumstances the siblings use one another as parent substitutes. However, the hostility between them is usually great because each Of them prizes acceptance and love from the mother most highly--and they are rivals for this attention. In many cases this rivalry is fed by the mother in an attempt to manipulate the children; e.g., "Susie is wonderful because she gives Mommy what she wants and makes her happy." Also, there is a tendency for mothers tO cling tO their sons and favor them over their daughters, as sons come tO substitute for the emotionally absent husband. This sets the stage for an intense boy-girl rivalry. The daughter may then turn to the father for support, but is usually 56 disappointed because he has such difficulty with intimacy. Thus she is likely to experience further rejection and feel increased hostility toward the Opposite sex. (Psychological developments such as this in women are Often Observed and erroneously determined to constitute "penis envy.") And if she does succeed in forming a gratifying relation with her father, this is likely to be met by jealousy and further rejection by the mother. Oedipal Conflict For the boy in this family situation, there is always solace to be taken in knowing that he is mother's favorite. He thus does not need tO pursue the father for support, as the girl does. However, there results from this mother-son union a strong father-son rivalry and resentment. (Naturally, this does not facilitate identification and the boy may feel lost, when he needs a male figure with whom to relate.) This creates the Oedipal crisis, as the boy struggles to maintain his relation with his mother while desiring to confirm his maleness through identification with his father. This becomes increasingly difficult, however, since the father has difficulty relating tO anyone, and probably is resentful at the prospect Of being displaced by his son in his wife's affections. On the son's part, he is also aware Of having "illegitimately" taken his father's place, feels guilty, 57 and subsequently fears retaliation. This is Often in— accurately described as castration anxiety. While in some cases it might manifest itself in this form, it is not the basis of the fear. Nor does it seem to be an inherent aspect of normal development. Rather, it is a reflection of the degree Of unhealthy relationships within the family, as has been described. Summary Summarizing, the family constellation providing the etiology for separation anxiety has been presented. Children in these families usually feel unloved and rejected by the parents, the extent of this varying with the degree of parental pathology. The children often turn to each other to have their needs met, but since everyone in the family feels deprived and is desperate for relief, there is little hOpe that they can consistently nurture one another. Rivalry for the parents' affection is often intense, setting the stage for hostility and resentment between siblings. Rivalries between child and same sex parent are also common, resulting in traditional Oedipal or Electra conflicts. The commonly seen "seductive behavior on the part of a parent toward a child of the opposite sex" (Estes & Haylett, 1956, p. 686) further exacerbates these con- flicts. Thus, while the children turn to one another at times for a substitute-parent, the parents turn to the children for substitute-partners. 58 SPECIFIC ASPECTS OF SEPARATION- ANXIETY PSYCHODYNAMICS Orality Orality has been much discussed as an integral part of the separation anxiety personal orientation. The focus on this as a key aspect is due in part to the frequent observations made regarding separation and eating. Naturally, an explanation for these appetite changes or abnormalities was sought. Change in Appetite as a Response to Separation For example, Heinicke (1965) found in his study of children separated from their parents that ". . . the separated children became greedier, especially for sweets, the longer they stayed in the nursery" (p. 252). However, he states that this behavior followed an initial lgss of appetite during the first few days of separation. Bowlby (1973), citing a study done by Spencer-Boothe and Hinde (1970) with monkeys, noted the occurrence of the same behaviors. "On the first day of separation one infant ate hardly at all. Subsequently all four showed a tendency to eat more" (p. 67). Spitz (1953) also reported a loss of appetite in infants during the initial stages of separation, and theori- zed regarding its origin. An attempt to explain the loss of appetite and the loss of weight has to take as its starting point the libidinal stage at which the infant is at this period. 59 It is the oral stage; one of the attributes of the lost love object is the gratification of the oral zone. The mother is the very source of food, and, psychologically speaking, food itself. When the infant is deprived of this love Object, the libidinal and the aggressive drives are denied the Opportunity for discharge. They are dammed up and turned against the self. After a brief period of transition we can observe that the infant withdraws and rejects everybody who lacks the attributes of the love Object. Similarly food alone lacks these attributes and will be rejected. Loss of appetite would then represent a behavior of withdrawal and rejection; loss of weight its consequences. (p. 134) The explanation for later over—eating can be given, follow- ing Spitz's line of thought, if viewed in Bowlby's concep- tual framework. Spitz has essentially described the loss of appetite in terms of Bowlby's protest and despair reac— tions to loss of the mother. Despair is especially empha- sized, as aggression is turned inward on the self. Accord- ing to Bowlby's paradigm, the next stage is that of detach- ment, the chief characteristic of this stage being the use of denial. Now, since the typical behavior that accompanies denial is exaggerated and overcompensatory in nature, and if this is applied to partaking of food, the logical result is overeating. Returning to the reports in the literature, Buxbaum (1958), in discussing the problems of separation, stated that specific physical manifestations Of oral components have been noted by many researchers. Feelings of discomfort in the intestinal tract, the desire to eat without being hungry is one Of the physical reactions to separation which Anna Freud, Spitz, and others described. (p. 8) 60 Symbolic Meaning of Food Additionally, more than one researcher has noticed the preference for sweets demonstrated by those suffering from separation anxiety. As previously noted, Heinicke (1965) reported "greediness," particularly with respect to sweets; and Kestenbergh (1971) attempted to explain the sig- nificance this has with regard to separation-individuation: Food as the heir of symbiotic dual unity preserves mother-child unity despite separation. It is looked upon as a bridge to the oral image of the mother, a bridge that maintains the integrity of the oral organ- object image as a functional unit. Throughout life, a drink or sweet can restore the long-lost well-being intrinsic in oral organ-object unity. (p. 78) This concept of achieving restoration of the mother- child unity through eating is an interesting one. Evidence for this hypothesis is provided in Mahler's observations of young children separated from their mothers (1974). Here she noted the predominance of eating over other oral acti- vities. If an appeal for comfort or reaching out for contact goes unheeded, the child seeks substitutions. In our study, we observed that the substitution used most fre- quently was eating, rather than autoerotic sucking activities. (p. 318) Summary In summary, orality is commonly seen as a predomin- ant orientation in persons suffering separation anxiety. While eating has been discussed as the activity-of-choice, it seems likely that substitutions such as talking, or smoking might be made later in life. The aim of this oral 61 activity is seen as restoring the mother-child dual unity. Such activity in adults is therefore regressive, or as Krout and Tabin (1954) would say, it constitutes a "run back to the nursery" (p. 297). This is due to an inability to deal with frustration, resulting from an inadequate early mother-child relation. ' Body-Image, Sexual Identity, and Sexual Behavior The Deve10pment of Body-Image As an infant develops, he gradually learns to differ- entiate between his internal and external experiences. As part of this he comes to the realization that he and his mother are not a single unit, but rather are separate entities, each with its own boundaries. This development results through a combination of simple maturation and various sense experiences. Hoffer (1950a) emphasized the essential aspects of touch in the boundary formation process as well as the importance of the libidinization of the infant's body by the mother. The nature, type and extent of such contacts determine the child's degree of physical differentiation and subsequent body image. Mahler (1952) emphasizes the importance of this interchange as it pertains to separation-individuation. Bodily contact with the mother, that is, fondling and cuddling, is an integral prerequisite for the demarcation of the body ego from the nonself within 62 the stage of somatopsychic symbiosis of the mother- infant dual unity. The infant discovers first the mother's body--then his own--touching and mouthing, and comparing the two. The fact that they are different and separate gradually becomes clearer as ego differentiation becomes clearer. (p. 287) This may seem to imply that the more a child is touched as an infant and toddler, the more differentiated he or she will be. But this is not the case. As with most things, moderation is the ideal. Just as too little contact pro- duces a child unsure of his own physical boundaries, lacking a true sense of self in the world; excessive physical con- tact, particularly as the child matures, can produce over- eroticization of the relationship. Another form of destructive contact is hugging the infant or child excessively and forcing him into physical passivity. This also impedes differentiation and will negatively affect body image. Most of the discussion in the literature regarding body-image is illustrated by cases of psychosis, where the distortion is more visible and observable. While these examples are not descriptive of the typical separation- anxious individual, they give some indication of the nature and type of distortions that are possible. For example, Fliess writes in Ego and Body Ego (1961) about tendencies toward self-stimulation, the delu- sion that the body is made of mechanical parts, the fears of disintegration, and the more commonplace tendency toward weight problems. (Note that this also relates to the 63 aforementioned topic of orality.) He also speaks a great deal of the important capacity to give and receive physical affection. He cites Mahler's example of symbiotic psychotic children and quotes her: . . . [they] crave body contact and seem to want to crawl into you--yet they Often shriek at such body contacts or overt demonstrations of affection on the part of the adult, even though they themselves may have asked or insisted on being kissed, cuddled, and "loved." (1952, p. 57). Receptivity to Physical Contact While this illustrates the common fear of engulfment in psychotic children, it is not hard to think of a "normal" adult who is reluctant to be touched, who stiffens or backs away from an affectionate greeting. This discomfort with affection or sexual contact is a prime aspect of body image, just as is the desire for extensive physical contact. Another aspect of a poorly developed body image is reflected in the fear of disappearing, or the self- description of emptiness often reported by borderline patients. This concept is well-illustrated by Furer's description of a pre-school symbiotic psychotic boy: The psychotic child suffers from extreme panic and anxiety, and at first cannot be comforted. The source for this anxiety is not always clear, though the fre- quency of self-stimulating and body—defining activities such as rubbing the body with sand, or head banging, as well as the wild aggressive outbursts, have led many to speculate that such behavior has to do with the fear of loss of body boundaries and with the lack of capacity for binding aggression. (1964, p. 495) Again, this is an extreme that would ordinarily not be observable in the normal or "normal-neurotic" personality. 64 But, extrapolating, it does suggest something about people who require a great deal of physical contact in their every- day life. This discussion has alluded to the place of sexual behavior, as it related to body-image and separation. Before becoming specific with regard to sexual behavior per se, there is a point to make pertaining to the deve10p- ment of sexual identity in this schema. On this topic, Greenacre (1952) has made some rather astute Observations. Certainly, so it seems to me, children with a delayed sense of separateness from the mother, may as part of this general condition, have also a delayed and uncertain appreciation of sexual differences, the pro- longed confusion on this score being part of the broader based disturbance of reality differentation between what is I or what belongs to me and what is his/hers, or belongs to him/her, which includes others than the mother. (p. 174) The only conclusion that can be drawn from this statement is that the greater the lack of separateness, and subse- quently, separation anxiety, the greater the overall identity confusion, including sexual identity. Consequences of a Poor Body Image Since this concept is rather complex, a review of the statements on body image and separation may be helpful before continuing further. 1. Too little or too much physical contact results in lack of emotional and physical differentiation--a poorly formed sense of self. 65 This, in turn, leads to a distorted view of one's own body, that may or may not correspond with what others see. Anxieties that arise from this poor physical self- image and the early experiences that formed it, result in problems handling physical contact. In this regard, there is a noted tendency to go to extremes. Extended symbiosis, in particular, impedes differ- entiation and makes it difficult to discern exactly who is what. This results in sexual identity con- fusion. Now, depending on the severity of the separation problem, you would expect to see any one or more of the following problems. 1. 2. Little attention to physical appearance. Weight problems. (Anorexia nervosa, in particular, has been linked to separation anxiety [Wold, 1973].) Sexual identity confusion. (In more extreme cases this would, of course, produce homosexuality.) Difficulties in expressing and receiving physical affection. (This could include sexual dysfunction.) Tendency to crave body contact to the extreme. (This could be revealed in the nymphomania and Don Juan syndromes.) 66 Sex and the Need to Be Held Regarding this last point, Hollender (1970) examined "The Need to be Held" through extensive interviews with women. He found a significant trend and stated his major conclusion as follows: Clearly the wish to be held and the wish for sexual gratification can be placed on a gradient with one or the other dominant in a given women or in a given woman at a particular time. . . . The woman with a strong craving to be held often barters sex, giving the man what he desires (coitus) for what she desires (cuddling). Indeed, this craving is sometimes the key determinant of promiscuity. (p. 448) He also noted that there existed for one woman he interviewed a desire for sex when she was depressed. He said this was understandable if "recognized as a disguised expression of the wish to be held or cuddled" (p. 448). Finally, Hollender suggested what might be used as common substitutes when it was not possible to be held, or have sex. He said that activities such as eating, drinking, smoking, or bundling up in a warm sweater or blanket were quite common--as were receiving vicarious pleasure from holding children or pets. Summary In summary, body-image is a basic aspect of separa— tion and later identity formation. Poor body image is a segment of the total self-image and has particular conse- quences for the formation of satisfying adult relationships. 67 Sex Differences The issue of sex differences and separation anxiety has not been systematically studied and Opinions vary from one writer to the next. Mahler et al. (1975) Observed that during the third subphase Of the separation-individuation process (18-22 months), there was a noticeable difference between the boys and girls that she studied in the nursery setting. In our comparatively small sample of cases, the boys, if given a reasonable chance, showed a tendency to disengage themselves from mother and to enjoy their functioning in the widening world. The girls, on the other hand, seemed to become more engrossed with mother in her presence; they demanded greater closeness and were more persistently enmeshed in the ambivalent aspects of the relationship. (p. 102) Motility and Differentiation In contrast to this tentative observation, Mahler, speaking later in the same work, states more specifically that the "boy seems better able to cope with 'symbiosis anxiety,‘ and to disidentify with the mother (Greenson, 1968), to avoid her or at least to resist her in a more covert way" (p. 215). Mahler also emphasized that motility is of great importance in achieving separation; since boys, whether through natural inclination or social learning, are more physically active than girls, it would seem likely that this would serve to facilitate an earlier, if not more complete, separation. This argument is logical and is probably true in the instance of normal development. But in the family that 68 produces separation-anxious children, as mentioned earlier, there exists a mother-son relationship that stifles this typical early male independence. That is, the tendency for the mother to infantalize and eroticize her relationship with a male child may well be enough to counteract the advantage acquired through early motility. Or, depending on the extent of symbiosis, the boy may be discouraged from such physical activity in order to satisfy the mother's needs for physical contact. Possible Influential Factors Thus the issue of sex differences may well be more complex than it appears on the surface. There could be a curvelinear relationship between sex and separation, with boys being fairly separate in a "normal" pOpulation, and the degree of individuation declining with the extent of the mother's separation anxiety. This would create a curve with boys higher than girls on separateness on one end, while lower than girls at the other end. However, this is purely speculative, and is stated for the purpose of illus- trating the possible complexities involved. Other con- founding factors could be the extent of parent sex- stereotyping, birth order, age of the child, and the influ- ence exerted by the father in raising the child. Since literally none of the aforementioned factors have been controlled in the studies reported in the literature, it is understandable that there are so many contradictions. 69 Continuing in this review, Abelin (1971), like Mahler, studied toddlers as they proceeded through the normal separation-individuation process. In his findings he reported that "on the whole, they [girls] maintained closer and more specific ties to both parents and were more guarded toward strangers than were the boys" (p. 242). In Bowlby's discussion of sex differences reported in the separation anxiety literature, he states that, ". . . neither Ainsworth with her one-year-olds nor Macoby and Feldman with their two-and-three-year-olds found sex differences of any magnitude" (1973, p. 51). Spitz (1946) came to the same conclusion following his observation of 123 unselected infants who stayed in the nursery from the fourteenth day through the first year of life. Regarding the influence of race and sex on the degree of manifested separation anxiety, he reported that "the factors of color and sex do not appear to exert demonstrable influence on the incidence of the syndrome" (p. 318). Choi (1961) reviewed the issue of sex differences in the incidence of school phobia, and concluded that there is no agreement in the field regarding this issue. She cited some examples of this lack of continuity. Lippman and Johnson note that the symptom appears more frequently in girls than in boys. Lippman says that Klein also has this view. Two studies by Jacob- sen and Talbot indicate that in their research settings there were more girls with a school phobia than boys, although for the setting as a whole boys far outnumbered girls. On the other hand, Goldberg's study sample con- tained more boys than girls. Thompson notes that school phobia is equally likely to Occur in boys or girls. (pp. 154-155) 7O Choi also notes that in the work at the Judge Baker Guidance Center, her findings reveal that "boys and girls seem to be equally susceptible to developing school phobia" (p. 155). Observed Differences Finally, in an attempt to clarify this issue some- what, there has been some agreement on the behaviors that are typical for each sex, among children, when their mothers are not present. These Observations have typically been made in nurseries, where children have been left for the morning or the day. Bowlby (1973) summarizes these ten- dencies as follows: In some studies and at some ages no differences are observed in the behaviour of boys and girls. In so far as any differences are Observed, boys tend to explore more in mother's presence and to be more vigorous in their attempts to reach her when she has gone; girls tend to keep closer to mother and also to make friends more readily with the stranger (observer or nursery attendant). (p. 52) While these observations are useful, particularly in that they have been made on more than one Occasion, they do not reveal anything about separation anxiety per se. They merely show that there are differences in behavior, based on sex. The meaning of these behaviors, as they relate to intensity of separation anxiety, remains to be seen. Summary In summary, there appears to be more opinion than fact regarding the impact of gender on the separation- individuation process and/or the intensity of separation 71 anxiety. It is evident that the many contradictory findings and theories will prevail until systematic studies are con- ducted in this area. Psychosomatics The separation anxiety literature naturally includes a discussion of somatic symptoms that appear in response to a separation, as well as typical psychosomatic developments in the "separation-anxiety personality." Most of the infor- mation reported on this topic comes from the school phobia literature. This is not surprising, in light of the common development of somatic complaints to justify school ab- sences, and the usual Observing and recording of children's illnesses that is part of the school's administrative routine. Purpose of Illness Sperling (1967) explained the fundamental purpose of psychosomatic manifestations and stated that they are frequent substitutes for school phobia. She states that in families where the development of school phobia would be "unacceptable," illnesses are often developed instead. Sperling listed ulcerative colitis and bronchial asthma as typical disorders accompanying separation anxiety. She added that many children who have these problems, as well as others, and who are subsequently home from school frequently, are actually achieving their aim of being in close proxi- mity to the mother, through illness. 72 Naturally, in order for there to be this secondary gain, the development of an illness must be "rewarded" by maternal closeness and attention. Thus, a child who had a working mother, or a mother who was punitive, or otherwise rejecting of the child when he was sick, would not be as likely to develop psychosomatic disorders. If these assumptions are correct, then there are a great many more separation-anxious children than would be evident on the basis of using school phobia as the sole criterion. Many of those who remain home frequently, or for long periods of time, especially with gastrointestinal or respiratory disorders, may do so out of fear of being away from the mother. Most Prevalent Problems That there are two main body areas where separation- anxiety problems appear is confirmed by Clyne (1966) in his study of school phobic children. He found their most pre- valent somatic symptoms to be either gastrointestinal or respiratory in nature, with abdominal pain being the single most frequent complaint. Others include diarrhea, anorexia, cough, colds, sore throat, earache, and asthma. In Bowlby's brief discussion of school phobia he notes that: Not infrequently the condition is accompanied by, or masked by, psychosomatic symptoms of one kind or another--for example, anorexia, nausea, abdominal pain, feeling faint. (1973, p. 261) 73 Further evidence support of this view is seen in Szyrynski's report on the treatment of school phobia, where he summarized the somatic complaints most typically seen. There may be headaches, dizziness, palpitations or shortness of breath, pseudorheumatic pains in the muscles or joints; but most frequently the "organ- language" of the gastro-intestinal tract is employed with abdominal pains or cramps, nausea, vomiting or diarrhea, refusal of food, (and) choking . . . (1976, p. 168) Similar symptoms have also been noted in the study of infants and children in the midst of a separation experi- ence (where anxiety is a normal response). Spitz (1946) observed that the infants he studied commonly responded to the loss of their mothers with loss of appetite, refusal to eat, and loss of weight (anorexia), as well as insomnia, in the early stages. Heinicke (1965) additionally noted that children separated from their mothers were more likely to become ill than those who were not, with eight out of ten children in his study developing colds shortly after the separation took place. . . . more of the separated children became ill then would be expected if they lived at home. Since the children were healthy when they entered the nursery, it is unlikely that previous ill health accounts for the finding. (p. 254) Migraine headache is also included here for dis- cussion, although it has received less attention as a mani- festation of separation anxiety than have gastrointestinal and respiratory problems. This may be the case, however, because migraine is rarer than either Of these other types of disorders. Fenichel (1945) cites the work of Knopf (1935) 74 and Trowbridge et a1. (1943) regarding the nature of the psychic conflicts underlying the development of migraine headaches. Investigation Of the personalities of patients suffering from migraine shows that they most regularly be classi- fied as "neurotic character" of marked emotional insta- bility. They frequently have an intense attachment to their parents. (1945, p. 253) Characteristics of Underlying Conflicts Other researchers have similarly attempted to describe the conflicts most Often accompanying the develop- ment of the aforementioned respiratory and gastrointestinal problems. Williams (1975), in a comprehensive discussion of basic psychic conflicts associated with asthma, states: The asthma conflict appears as a clash between two tendencies--one, driving the patient away from the dependency upon the mother and the other, a fearful regressive tendency toward a dependent, infantile attitude towards the mother. (p. 199) In the study conducted by Williams, he sampled normal children as well as children displaying asthmatic severity at three different levels. In comparing these children, he concluded that: a) there is a closer bond between the asthmatic child and his mother than in the normal mother-child situa- tion; b) the threat of separation from mother appears in the stress situation which stimulated aggressive behavior in the asthmatic child. (p. 214) . And in general, The asthmatic children demonstrated an excessive dependence-independence conflict with an intensive mother-child bond and core anxiety around the threat of separation. (p. 215) 75 Fenichel (1945) made similar statements regarding the asthmatic's proclivity toward dependency, citing French and Alexander, who stated that the task of "mastering the fear of being left alone governs the patient's whole life." Ulcers, the most common gastrointestinal disorder in adults, presents a somewhat different psychodynamic picture. Fenichel, in the same work, gives the following description of the "ulcer personality." Persons with a chronically frustrated oral-receptive demanding attitude, who have repressed this attitude and often manifest very active behavior of the reaction- formation type, are, unconsciously, "hungry for love." (p. 245) Summary Recapitulating, psychosomatic disorders of the respiratory and gastrointestinal symptoms, as well as migraine headache, have been associated with separation anxiety and early separation reactions. The most common of these problems--ulcer, asthma, and migraine--have been highlighted here, for the purpose of aiding in the indenti- fication of separation-anxious adults. Each disorder is said to be an indicator of a parti- cular intrapsychic conflict associated with separation anxiety, with migraine reflecting intense attachment and repressed hostility toward a parent; asthma reflecting intense ambivalance over dependency on the mother; and ulcer reflecting an often-repressed oral frustration and subse- quent desire for oral passivity. 76 ADULT SEPARATION ANXIETY Introduction Virtually all of the literature reviewed thus far has dealt with separation anxiety in childhood, with adoles- cents being the oldest group mentioned. Since the topic of this thesis is agglg separation anxiety, there must be a basis for assuming that this childhood syndrome is likely to continue into adulthood—-in one form or another. NO real evidence can be provided for this position, of course, since there are no studies reported in the literature on adults. Instead, the probability that early separation anxiety defines a personality style that is likely to prevail throughout maturation will be discussed in this section. Possible mitigating factors will also be reviewed. That is, developments taking place during childhood and adolescence that could serve to ameliorate the effects of early separation anxiety will be presented. The reader is reminded that the initial separation anxiety produced in childhood is caused by actual separation from the mother during the first few years of life; threats of separation; an inadequate mother-child relationship that retards or otherwise impedes the separation-individuation process; or any combination of these. Also, separation anxiety is deemed to be more than a simple anxiety reaction to a life experience. The term additionally describes the entire 77 resultant psychodynamics—-a way of relating others, a basic set of behaviors and attitudes, a way of seeing the world. Finally, separation-individuation as a life-long process will be discussed. Crucial stages and life experi- ences will be noted that serve to add to or detract from the achievement of individuation in adulthood. Separation Anxiety: Continuation into Adulthood Regarding the aftereffects of early separation anxiety, Guntrip (1969) speaks most articulately and speci- fically. He discusses the depth of the problem, as revealed in the process of psychoanalysis. How deep is the ego-weakness left by very early failure of a good mother-infant relationship, and how long it can take to grow out of it, may well emerge in a protracted analysis, long after specific "illness" has been overcome. Analysis has to go very deep indeed to enable the patient to feel basically safe against the fear of breakdown in an environment that feels empty of support. When analysis has been begun because of speci- fic illness, that illness can be overcome in due course, most of the patient's anxieties relieved, and his practical life rehabilitated. Yet there can still be dependent characteristics and vulnerabilities to anxiety, after the patient has held his gains for a number of years, to show that there remains a deep hidden core of infantile trauma, insecurity and need for the mother who failed the baby. This deprivation trauma, the loss of the good mother of the first year, sets up an un- alleviated need for mother-substitues and a liability to separation-anxiety and depressive despondency, which profoundly affects adult living. (p. 106) Guntrip's position is clear: that it is impossible to sustain separation anxiety as a young child without the effects carrying over into adulthood. 78 Separation anxiety may be a ubiquitous and pervasive phenomenon, once acquired. Fairbairn (1952) describes how the separation-anxious person continues to relate inappro- priately to others. . . . separation anxiety is a characteristic product Of the tendency of individuals who have remained in a state of infantile dependence to make identification the basis of their emotional relationships with those upon whom they depend. The figure with whom the dependent indi- vidual originally identified is, of course, his mother; and whilst it is not long before he begins to identify himself with other figures, particularly his father, the original identification persists underneath all others subsequently made. (p. 276) According to Fairbairn, then, it is the fate of the separation-anxious person to transfer the original inade- quate mother-child psychodynamic onto all future relation- ships. He also states that the early failures in identity formation and fixation in infantile dependence make forma- tion of healthy adult relationships an improbability. Describing the tendency of a separation-anxious person to "join" others through the processes of introjec- tive and projective identification, Fairbairn states: He tends to feel that he is part of them, and equally that they are part of him. In their absence his very personality tends to be diminished--and, in extreme cases, even his sense of personal identity may be com- promised. (p. 277) Amrith (1975) confirms this tendency in his explanation of the purpose of projective identification: Now what are the motives underlying projective identification? I think, the main reason is intoler- ance of separation. Whenever there is a massive separa- tion, the child goes into the mother, and becomes the mother, rather than tolerate the idea of separation. p. 5) 79 Assuming that Fairbairn and Guntrip are correct in their assessment of psychological damage done by early separation anxiety, the difficulties in establishing rela- tionships and functioning normally can be seen. The chief reason for this seems to be that the individual is not fully capable of relating independently. ”He feels compelled to "fuse" with other people, preventing the formation of the healthy independent liaisons that he requires to attain his individuality. Ironically, the healthy person who would ‘refuse to participate on this level would probably be per- ceived as rejecting, and no further relationship would be possible. What is being said here is that in order to relate positively to others as an adult, a person must have achieved a certain degree of object constancy, as Mahler et a1. (1975) outline its development in the final subphase of the separation-individuation process. Without this, there can be no independent adult. While the person may have matured chronologically, emotionally he is still an anxious child who is not fully able to discriminate between himself and others. Winnicott (1958) stresses the essential aspects of object constancy in his discussion of "The Capacity to be Alone." The capacity to be alone depends on the existence in the psychic reality of the individual of a good object. The good internal breast or penis or the good internal relationships are well enough set up and defended for the individual (at any rate for the time being) to feel confident about the present and the future. The rela- tionship of the individual to his or her internal objects, 80 along with confidence in regard to internal relation- ships, provides Of itself a sufficiency of living, so that temporarily he or she is able to rest contented even in the absence of external Objects and stimuli. Maturity and the capacity to be alone implies that the individual has had the chance through good-enough mothering to build up a belief in a benign environment. This belief is built up through a repetition of satis- factory instinctual gratifications. (p. 417) Without such internal assurance, the separation-anXious adult is too afraid to tolerate existing alone, in any sense. What he does most Often is to take refuge in others, through the aforementioned process of identification. Angel (1972) offers further illumination on this topic in his discussion of the adult patient who has not succeeded in becoming intrapsychically separate: The self and Object representations have remained fused to a greater or lesser degree. This does not mean that these patients have never separated from their real mothers. Nor does it mean that these patients duplicate what the two—year-old child does. . . . These patients cannot help learning certain things by having lived to adulthood. But the struggle between the wish to merge with (or be devoured by) the Object and the fear of merging goes on constantly even during relatively simple activities. (p. 308) Parens (1971) stresses the same point, emphasizing that the good symbiotic partner who is identified with must, at some time, be internalized for separateness to be achieved. How and when this occurs, and whether or not it is possible in adulthood, is unknown. How long does the symbiosis with the mother proceed? How much residual symbiosis remains? As Mahler (1963, 1965) points out, Freud held that a nucleus of the relation to the mother remains with man throughout his life. Jacobson (1964) concurs with this by stating that the wish to re-establish the lost mother-child unit ". . . probably never ceases to play a part in our emotional life" (p. 39). (p. 109) 81 Mitigating Factors Assuming that at least some degree of mitigation of separation anxiety is possible, under what circumstances is this likely to occur? Bowlby writes that: Among the conditions known to mitigate the intensity of responses of young children separated from mother the two most effective appear to be: --a familiar companion and/or familiar possessions --mothering care from a substitute mother. (1973, p. 16) This suggests that reduction of immediate anxiety can be accomplished through the means of remaining in familiar surroundings, or relating to a substitute mother figure. While this offers only temporary or superficial mitigation, it has been proven to work with children and may well be the chief defense used by adults. What is proposed then, is that for most people, early separation anxiety continues into adulthood but is masked through a system of anxiety management similar to the one described above. That is, separation-anxious adults would live near their parents and childhood friends, never venturing far from home base, so that "refueling" would always be possible. Another possibility, and one that is probably most often used, is to create a symbiotic marital union that will substitute for the former relation to the parents. As can be seen in the previous literature review, such a union is likely to produce new symbiotic partners, as children are born and raised. Others, whose socialization is less complete, would be more likely to remain single, possibly remaining in the 82 family home with other adult siblings and/or aging parents. It is these persons who are likely to be noticed as "odd" or perceived to have a relationship problem. However, there are many peOple who appear to be well-adjusted, having jobs, marriage partners and children, who continue to function psychologically from a position of infantile dependence and to be anxious at the thought of being alone. Since it is commonplace in our society for married individuals to speak constantly of "we" rather than "I," this anxiety and dependency often go unnoticed. Overt anxiety is observable when some threat pene- trates the defense system. Examples of these are a break- down in the marital relationship resulting in feelings of rejection; an actual loss of partner through death or divorce; the loss of children serving as substitute partners through maturation; or an unwilling move to new surroundings. Naturally, there are numerous internal threats, more subtle in nature, that are possible. But those that are listed here relate directly to separation anxiety and are predict- able as catalysts for a separation anxiety attack. While this discussion has focused on the use of familiar persons and possessions as superficial mitigators of separation anxiety, this is not to say that amelioration in a more substantial and permanent sense, is not possible. This is especially true in the case of a person who lost either the mother of both parents at an early age and then acquired a new attachment figure or figures. In this 83 instance, assuming that some early attachment had taken place, the damage caused by the loss could be repaired sub- stantially, provided that the new attachment figure could establish the necessary restorative conditions in the rela- tionship. Ironically, in the case where the family is intact and the separation anxiety results from a poor mother-child relationship, the prognosis is not good. The reasons for this can be seen in an examination of the possible events in later development listed by Mahler et a1. (1975) as most crucial to individuation, following the initial separation- individuation period. These are: 1. development toward libidinal Object constancy; 2. later stress; 3. possible shock; 4. degree of castration anxiety; 5. resolution of Oedipus complex; 6. resolution of developmental crisis of adolescence. The second and third stages deal with the issue of possible psychic trauma, which would serve to exacerbate existing separation anxiety. Numbers four and five are concerned with relations with the father which, as noted previously, are not likely to be much more favorable than the existing relationship with the mother. In fact, the father is typi- cally more distant emotionally, less socially skilled, and therefore an unlikely person to provide compensation for problems in the mother-child relationships. The first stage 84 refers to a process which essentially goes on between child and parents--leaving it unlikely that any substantial changes would occur in this area. Finally, stage six is the "resolution of develop- mental crises Of adolescence." This period of development is emphasized by Mahler as a crucial one--especially with regard to identity. It seems probable that any ameliorating influences would be most effective at this time. This is chiefly due to the fact that the adolescent spends more time away from home (which is not always a healthy atmos- phere), is capable of higher levels of cognition, and has an Opportunity to make use of mother substitutes. Middle- school and high school teachers and counselors Often assume this role during school hours and may have a substantial effect in some cases. In summary, it can be seen that the prevailing Opinion presented here is that early separation anxiety is not likely to be altered significantly as development pro- ceeds. In order for this to be accomplished, a substantial intervention by a substitute "good parent" must be made. While this is not impossible, neither is it probable. Separation-Individuation as a Life-long Process The concept of pathological separation anxiety has been presented in great detail. However, little attention has been paid to the normal separation-individuation 85 process. This is important in terms of gaining a perspec- tive on the issue of separation and separation reactions. Bowlby (1973) speaks of the natural grieving that occurs upon losing or being separated from a loved one. In so far as attachments to loved figures are an integral part of our lives, a potential to feel distress on separation from them and anxiety at the prospect of separation is so also. (p. 56) It is only the prolonged and exaggerated distress, perceived actually as a loss of self, that is indicative of separation anxiety. In fact, it is through the life-long process of forming love relationships, and then leaving those loved ones--feeling the pain of separation, and recovering--that individuation is continually confirmed. The issue of separa- tion, then, is not limited to the first few years of life. Sternschein (1973) reported the panel discussion on this topic held at a meeting of the American Psychoanalytic Association. The panel members were Mahler, Fleming, Pine, Neugarten, Ross, and Sternschein. They all seemed to agree that separation-individuation is indeed a life-long process. But the ease or difficulty with which an individual pro- gresses through key separation experiences and learns from them has its basis in the initial birth-to-three-years period. Sternschein states that: . . . separation anxiety is an essential force that propels man along the path of object finding and object relating, and toward a dynamically responsive self- concept--that is, toward an individuated, well- differentiated, but not unmodifiable self-image. But developmental progression does not occur without 86 transient regression, as the present interacts with memories of the past. Even well after the separation- individuation phase of childhood has been traversed, self- and object-images and their mental representations continue to be modifiable at critical affect-laden junctures such as puberty, adolescent disengagement, marriage, parenthood, grandparenthood, the climacterium, retirement, senescence, and during psychoanalytic treat- ment, as well as socio-pathologically with exposure to massive psychic treatment. (p. 634) Mahler states that the first life test of the indi— vidual's success in separating occurs in adolescence, which she and other panel members view as critical to resolution Of earlier separation-individuation problems. Many of these problems are worked through as the adolescent goes through what is essentially a second "rapproachment" stage with his parents. The stages listed previously by Sternschein Offer the same opportunity, although with decreasing likelihood of early separation problem resolution. While the other life stages, especially parenthood, are of great importance, the later life stages present unique problems. The aging person must come to terms with death as separation and the possibility Of exchanging roles with his children. The effects of aging on body image is still another phenomena to be dealt with and is discussed by Sternschein: There is a new integrative challenge when illness, injury, and signs of somatic aging are of such a nature and occur at such a time as to alter the body image. Kernberg suggests that symptom-free narcissistic per- sonalities should be analyzed as a means of strength- ening them so that they can deal with their aging with- out psychic breakdown. (p. 637) 87 Thus, the results of the early separation-individuation process must be dealt with, and can be modified as life progresses. In fact, Mahler states that "one could regard the entire life cycle as constituting a more or less success- ful process of distancing from and introjection of the lost symbiotic mother" (Sternschein, p. 639). Summary Separation anxiety has been discussed as a psycho- logical problem that is established in the first three years of life and most likely continues into adolescence and adulthood. While maturation alters the form and overt behavioral manifestations of this problem, it is not likely that any major amelioration of the initial effects takes place. Exceptions to this occur as the result of the mean- ingful intervention of a person who essentially provides the stable, healthy relationship that did not exist between parent and child. Examples of this type Of intervention include (a) actual replacement of the parent with someone who is capable of establishing a constructive relationship (such as adoptive, foster, or step-parents), (b) extensive time spent with an effective mother substitute (Often seen in attachments to grandparents, teachers, or other available adult figures), and (c) long-term psychotherapy, preferably occurring in adolescence rather than adulthood. It has also been established that normal events in life such as leaving home, getting married or divorced, 88 losing a loved one, aging, and facing death, all represent challenges to anyone's individuation and carry the threat of separation anxiety. All persons face this threat, since no one has achieved complete separateness. Therefore, the mark of healthy functioning is not the absence of anxiety or grief at separation, but the ability to cope with these feelings effectively. CHAPTER 4 TEST CONSTRUCTION Format In this chapter an explanation is given of how the preceding theoretical material was integrated with inter- views and case histories to provide a basis for constructing a test of adult separation anxiety. This chapter will include a discussion of how the objective format was selected, how the items were written, how the instrument was used on a pilot sample, and how the cluster analyses and final selection Of items were completed. In the construction of the instrument, an initial decision had to be made regarding the format. A projective format, modeled upon either the Rorschach Inkblot or Thematic Apperception tests was considered. An objective format, similar to that of the Minnesota Multi-phasic Personality Inventory was also deemed to be a plausible choice. Because separation anxiety involves a great deal of repression, and is therefore largely unconscious in adults, the use of a projective method at first seemed most appropriate and desirable. However, the design of a new projective method presented several problems. Assuming that the use of visual 89 90 stimuli is optimal, the pictures or figures shown to subjects must be relevant to the psychodynamic that is to be assessed, but must still be sufficiently ambiguous to allow for indi- vidual "projective" responding. This issue was discussed by Rabin (1968), who states: . . . if additional elements besides the variable rele- vant are permitted to intrude upon the subject's sensory apparatus then various competing responses may interact and often inhibit the relevant one which is of primary interest to the investigator. (p. 623) Another problem that is often cited in the use of any projective technique, whether old or new, is the sub- jective nature of the scoring and interpretation. While it is generally agreed that highly competent diagnosticians make very similar interpretations of a given protocol, it is not realistic to assume that this consistency will prevail in general usage. Thus, the margin of error that exists when using projective techniques presents an obstacle to syste- matic research, particularly when a large number of subjects is used. The final problem in the use of projective methods in research is concerned with practicality. The time re- quired for the administration, scoring, and interpreting of a projective test renders it prohibitive for use in any large-scale research project. The required number of diag- nosticians to complete such a task, as well as the number of protocols that would be generated and scored, inflate the margin of error in measurement well beyond acceptable standards. Additionally, the time demands on subjects and 91 researchers alike make the use of projective techniques, for extensive research, neither practical nor expedient. Since the major purpose for creating a measure of separation anxiety is to provide a research tool that would facilitate exploration of this tOpic, it seemed advantageous to use some other assessment procedure. Considering that objective personality tests are most often used in systematic studies, and that this method avoids many of the pitfalls of projective techniques, such a format seemed to hold the most promise. Objective instruments can be scored quickly and uniformly by machine, and the time involved in administra- tion is usually short--making it easier to gain the OOOpera- tion of potential subjects and to generally facilitate completion of the study. This is not to say, however, that the creation and use of an objective measure is problem-free. The greatest difficulty is formulating items that transcend individual differences and assess the psychodynamics of interest. To accomplish this when dealing with unconscious attitudes and feelings, as in separation anxiety, presents additional difficulties. A high degree of sophistication is required to tap this unconscious material without also stimulating anxiety in the subject. If such anxiety is stimulated, defenses will be aroused and the true response will be re- pressed. The response that is likely to result in this instance will no doubt reflect the subject's denial, this being the Opposite of what he or she actually feels or thinks. 92 Producing a sizable number of valid items, that adequately sample the area of concern, is therefore diffi- cult. Items must be written that reflect the attitudes, behaviors and life circumstances associated with separation anxiety; and, contrary to the goal of ambiguity in the invention of projective stimuli, the meaning of objective items must be perceived similarly across persons in order to obtain meaningful results. If Person A believes that one statement is being made by a particular item, and Person B believes that the same item is saying something entirely different, comparing their responses to that item would be meaningless. Steps in Test Construction With these factors in mind, the construction of an Objective measure of adult separation anxiety was undertaken. The steps outlined by Sarnoff (1971) for construction of a paper-and-pencil personality measure were used as a guide- line in this process. 1. The initial articulation of items in terms of theoretical aptness. 2. The wording of items in such a way as to maximize their comprehensibility and unidimensionality, while minimizing the possibility of their generating a response set. 3. The composition of a sufficient number of items to enhance the ultimate reliability of the measure, while requiring minimal time and effort for their administration. 4. The composition of clear instructions concerning the way in which subjects are asked to respond to the items. 5. The checking of items and instructions with collea- gues and a sample of potential subjects. 93 6. The formal administration of the tentatively selected items to a large enough population to permit a factor analysis. 7. Factor analysis to determine the extent to which the a priori theoretical formulation of the tentatively selected items actually circumscribes an empirically common cluster. 8. The item-analysis of the factor-analyzed items, eliminating those that fail to discriminate between subjects at Opposing ends of the distribution on the total scale--or that fail to correlate significantly with the aggregate score. 9. Assessment of the internal consistency reliability of the purified items, yielding an estimate of the unitary character of the scale with respect to what- ever dimension it purports to measure. 10. The ordering of itemsiin the final measure in a manner designed to offset the emergence of a response set and to minimize awareness by subjects of the variable actually defined by the measure. (pp. 141-142) Item Choice Regarding the "initial articulation of items," published case material and extensive interviewing with several separation-anxious adults supplemented the theoreti- cal material as a basis for item content. Items were gen- erated to represent the major psychodynamic aspects and family history variables reported in each of these sources. However, there were some exceptions. For example, no items dealing with specific sexual behaviors or pertaining to sexual identity confusion were formulated, in the belief that the social desirability factor in responding would be too great. That is, it seemed probable that most persons would answer in the direction of socially approved behavior. In this case the items would have little value as psychologi- cal discriminators. Socially unacceptable behaviors deemed 94 to be rooted in early oral frustration, such as alcoholism and drug-taking, were excluded for the same reason. The wording Of items was carefully done, then, in an attempt to insure that subjects would not tend to answer in a particular way that represented what seemed most acceptable to society, to the researchers, or to themselves. As stated by Kline (1973), this is essentially what defines an objective test. An Objective test is a procedure for obtaining an individual difference score, based on the responses to a specific set of stimuli or sequences of stimuli, such that either the correct implication of the response in question is unknown to the subject or the nature of the response is such that the subject cannot readily modify his response in some desired direction . . . (p. 68) It is important to remember, however, that the goal in item writing is not necessarily to provide tools for discovering the "truth" about a person. The problem at hand is not to discern what separation-anxious people do or feel that is different from others. The goal is to write items that will tap what separation anxious persons tend to say about themselves--whether or not their self-perceptions are true. Meehl (1967) discusses this issue in relation to the psychopathic deviate scale of the MMPI: An allied item is "I have been quite independent and free from family rule" which psychOpaths tend to answer false--almost certainly Opposite to what is actually the case for the great majority of them. (p. 519) Items were also developed so that all factors believed to comprise separation anxiety were represented by more then one item. In this way, if one item fails to 95 discriminate adequately, another may prove workable. This assures that all factors will be included in the final selection of items. Also, a sufficient quantity of items was created to insure adequate length for reliability. It was expected that no more than half of the original items would be retained, because of failure to discriminate, poor wording, etc. Having formulated the items with these principles in mind, the test was then informally given to approximately 20 different people, many of whom had identified themselves as being separation-anxious. They were interviewed following their written response to the items, and asked to comment regarding item meaning, readability, and response code. Items were then re-written in accordance with the criticisms received from this group, and from a few colleagues who reviewed the test. It should also be noted that many existing person- ality inventories were reviewed to aid in composing items, such as the Minnesota Multiphasic Personality Inventory, F-Scale, Rokeach Dogmatism Scale, and Tennessee Self— Concept Scale. Additionally, Cattell's research on per- sonality factors (1941) and Edwards' discussion of social desirability and response sets (1957) were reviewed. Response Codes Regarding response codes, three different ones were used at various times during the pilot administrations. 96 First, a simple true-false code was used, but this was criticized as being too limited. Many people stated that they felt restricted by this type of "all or nothing" response choice. Then a five-point true-false continuum was implemented--such as the one used in the Tennessee Self- Concept Scale. The choices in this format are as follows: 1 2 3 4 5 True Mostly Partly true Mostly False true partly false false However, whether due to the nature Of the items or to the type of format, persons taking the test demonstrated a marked tendency to answer with response three--part1y true, partly false. Since this does not qualify as an effective discriminator, a new response format was sought. After due consideration of the subjects' criticisms was made, a four-point true-false continuum, which eliminated the middle-of-the-road number three response, was decided upon. Items were then compiled, with care taken to insure that highly related items were not grouped together. Items were alternated in relation to the true-false direction of a separation-anxiety response, to ensure that no series of true or false answers would occur. However approximately 65 percent of the items, scored for separation anxiety, would be answered "true." An attempt was made to create a 50-50 split on true and false answers, but some of the items 97 could not be reworded to effect this without either distort- ing their meaning or detracting from their readability. The test in this initial form appears in Appendix A, is labeled as Form A, and contains 127 items. First Test Administration The next step in developing the instrument was to administer it to a large and diverse group of people. Responses from such a group would provide the necessary basis for a factor analysis, and would reflect the prevalence of separation anxiety in the general population. It also seemed desirable to administer a questionnaire at the same time, so that hypotheses could later be tested and group differences discerned. Such a questionnaire (see Appendix B) was created and given with the test of separation anxiety, to 409 per- sons. Numerical codes were used and all subjects remained totally anonymous. While a random sample ideally should have been taken, it was not possible to collect data on over 400 diverse people in this way. However, the large number of subjects partially compensates for this lack of random selection, and results of hypotheses tests are reported in light of this sampling procedure (chapter 6). All of the persons participating in this phase of the study were approached in person by the researcher or one of her assisting colleagues. Three of these research assistants collected data from other parts of the country, 98 but were given a verbal orientation and a set of written instructions to follow (Appendix C). In all locations and instances the refusal rate was negligible and almost always attributed to a lack of time. (Thirty to forty minutes were required to complete the test and questionnaire.) All data was collected during July and August of 1978, and will be referred to throughout this thesis as the General Sample. The breakdown of this sample is as follows: N Johnson Controls, Milwaukee, Wis. 35 Detroit ghetto 15 Towar Gardens public housing, E. Lansing 35 Counseling classes, Univ. of Louisville, Kentucky 62 Lansing airport 45 Clerical personnel, Erickson Hall, MSU 30 Sienna Heights College, MI. 25 Meridian Mall, Okemos 30 Kings Point East Apts., E. Lansing 50 College of Urban Development (MSU) faculty and staff 20 Librarians, MSU Library 15 Physicians, nurses, technicians, MSU Health Center 20 Miscellaneous 21 Sub-total 409 Removal of bad cases ._:2 TOTAL 4 0 0 Type Of Sample This sample was quite diverse, including college professors and administrators, students, office workers, technicians, laborers, secretaries, doctors, nurses, busines- persons, welfare mothers, engineers, salespersons, assembly- line workers, housewives and nuns. The only segments of 99 society that seemed to lack representation were the groups at either extreme end of the socioeconomic continuum. The purpose of administering the test to this group was to collect sufficient data for a factor analysis, run such an analysis, eliminate unworkable items, and conduct a validity study on the resulting instrument. However, time problems encountered in collecting the General Sample, as well as in gaining access to the residence hall sample for the validity study, created an overlap in events. That is, by the time all of the data from the General Sample had been collected, keypunched, fed into the computer, and edited, the incoming freshmen targeted for study were arriving on campus. This issue was further complicated by the fact that, at that time, it was discovered that no factor analy- sis program available could handle 127 variables at one time. For this reason, the data from the General Sample was not analyzed prior to the data acquired in the subse- quent validation study. Therefore, essentially the same test was given both times, with no elimination of items taking place. Some new items were added, however, at the request of the dissertation committee. They noted that there were no items pertaining to the role of the father in separation anxiety. Therefore, items parallel in content to the existent "mother" items were written and added to the test for the second 100 administration described in chapter 5. These items relating to the father (see Appendix D) were added to the new test, form B (see Appendix E). Cluster Analysis Returning to the discussion of the General Sample, because no factor analysis program was available, an alterna- tive cluster analysis program (PACKAGE) was decided upon. While the program itself was capable of completing the analy- sis, the university computer was not. The required core memory exceeded the computer's limits, making a simultaneous analysis of the items impossible. The upper limit was deter- mined to be 96 items. Therefore, it was necessary to randomly select 96 items representing all theoretical areas for the first cluster analysis. This was done with both samples, the aforemen- tioned General Sample, and the validity study Freshmen Sample, described in chapter 5. As items clustered together well, they were removed from consideration, and remaining items were inserted. This was done a total of four times for each data sample, with item removal occurring at each stage to allow for the insertion of previously excluded items. There were several criteria used for deciding which items should remain and which should be eliminated. The major criterion, of course, was how well the item clustered with other items. Another criterion was whether or not the 101 item content was meaningfully compatible with other items in its cluster. The final criterion was based on the results of a chi-square test run on the item responses of "school phobics" versus the item responses of "non-school phobics." School phobics were defined here as those persons in the General Sample who responded with "sometimes" or "often" to statement number 19 on the background questionnaire, which reads "I was afraid to go to school as a child." Non-school phobics were those in the General Sample who responded with "hardly ever" to this same statement. A listing of the items that were determined to have been answered significantly differ- ent when the groups were compared appears as Appendix F. The alpha level was set at .05. An attempt was made to retain all of these items for the final form of the test because of the relationship said to exist between separation anxiety and the manifestation of school phobia. (It should be noted, however, that using this final criterion for item selection essentially guarantees that the school phobic group will score significantly higher on the test of separation anxiety than others in the General Sample. But it does not guarantee high scores for any other criterion group. These points must be considered when examining the hypotheses tests executed on this data and reported in chapter 6.) When all of the cluster analyses on both samples, General and Freshmen, had been completed, further elimination 102 of weak items was made and basic cluster content was deter- mined. Several small changes were subsequently made, how— ever, as items were shifted from one cluster to another. This was done to determine the best "fit" and achieve the highest alpha coefficient for each cluster. The final clustering of items into subscales and the correlation matrix for each cluster are reported in Appen- dix G. The names and alpha coefficients associated with each cluster will be presented here, along with a discussion of each subscale's content. The alpha coefficient describes the extent to which each item in a given subscale measures the same trait and, as such, is a measure of internal con- sistency or reliability. Resulting Subscales Cluster 1, with an alpha coefficient of .67, contains items dealing with dependency and inhibition. This scale is not scored for separation anxiety per se and was created as an indicator of personality type. It has been labeled "Basic Personality Types," and has as its purpose, to differ- entiate between those who are passive and receptive, and those who are maintaining a strong defensive reaction to these traits. The only relationship anticipated between scores on this scale and separation anxiety would occur when scores are very high. This indicates a rather passive- dependent personality, associated theoretically with separa- tion anxiety. 103 Cluster 2, with an alpha coefficient of .64, is the first subscale that represents a direct measure of some aspect Of separation anxiety. This subscale has been named "Overprotection," because most of the items refer specifi- cally to the over-involvement of parents in the person's life. Cluster 3 contains items describing the nature of a person's relationship with his or her father. This subscale has the second highest alpha coefficient--.78, and is labeled simply "Father," since the items here were written as parallel to some of the "Mother" items and has no hypothe- sized outcome. They were all scored in a negative direction, a high score thus reflecting hostility, estrangement, and even fear of the father. A low score reflects what might be called an over-attachment and degree of symbiosis in the relationship. Viewed in light of the theoretical material pertaining to the role of the father, it would seem that extreme scores at either end would be indicative of separa- tion anxiety. But this is speculative, and the results per- taining to this subscale should be considered as exploratory in nature. Cluster 4, named "General Separation Anxiety," has the highest degree of internal consistency, with an alpha coefficient Of .82. This subscale is comprised of general symbiosis items that clustered well together but defied a more specific descriptor. This is by far the largest 104 subscale in content, with 18 items, and appears to be the most reliable assessor of separation anxiety. Cluster 5 has been named "Self-Acceptance" and contains a variety of self-descriptive items. Exceptions to this are seen in items 18, 36, 37, 43, and 141. Rather than being directly self-descriptive, these items reflect the person's level of trust in self and others. However, all of these items did cluster together, showing an alpha coefficient of .62. This is the lowest consistency associ- ated with any Of the subscales and apparently reflects the diversity of scale content. Future revisions of the test would probably be improved by the splitting of this sub- scale into "Self-Concept" and "Trust," with the addition of new items to strengthen each subscale as an independent unit. Cluster 6, with an alpha coefficient of .64, is the subscale that deals with the nature of a person's relation- ship with his or her mother. Since a high score on this subscale reflects angry estrangement, it has been named "Mother Hostility." Interestingly, several of the school phobia items clustered here, providing evidence for the hypothesis regarding the hostile-dependent relationship between mother and child in cases of school phobia. Three additional subscales are reported and named "Colds," "Migraine," and "Ulcer." These are one-item scales, and therefore have no alpha coefficient associated with 105 them. They were included for the purpose of examining the relationship between separation anxiety and certain psycho- somatic disorders. As would be expected, the relationship between sub- scales is minimal, reflecting a high degree of independence. All subscales shared a mildly positive relation, with one exception. Clusters 3 and 4, which became the "Father" and "General Separation Anxiety" subscales respectively, showed a slight negative relationship between them (-.21). This coefficient is not high enough to be regarded as significant and no hypothesis will be made regarding this. The corre- lation matrix demonstrating the relationship between the six subscales is reported in Appendix H. Summary In summary, constructing a test of adult separation anxiety was begun by examining the various measurement methods available. Following this, an objective test with a four-point true-false response code was devised. Items were written in accordance with recognized principles. Then data was collected from two samples, the General Sample (pé400)and the Freshmen Sample (gs300, and described in chapter 5). Cluster analyses were executed, with the final clusters being determined by the Freshmen Sample, due to the addition of items pertaining to the role of the father. The six subscales that resulted were described, listed, and measures of internal consistency were reported here. CHAPTER 5 EXPERIMENTAL DESIGN AND PROCEDURES 0_vs_r_v_isr The purpose of this study was to devise an instrument for the measurement of adult separation anxiety and test its validity. Conjointly, this process also served as a test of the theoretical constructs pertaining to this set of psycho- dynamics. An objective personality test was created and administered to a diverse, but not random, sample of adults. It was then given to the entire freshman population of a Michigan State University dormitory. The responses of both these samples were analyzed by a cluster analysis computer program. Subscales were formed from the resultant clusters and scored for both samples. Finally, the mean scores of separation anxiety criterion groups, determined by answers given on a background question- naire by each person, were compared within each sample. Samples The first group, labeled the General Sample, repre- sents numerous societal subgroups. This sample was obtained through various means and chiefly from the East Lansing, 106 107 Michigan, area. (Geographic locations other than this and breakdown of collection areas are listed in chapter 4.) In all cases, persons were approached by the researcher or assistant to the researcher, and asked if they would be willing to participate. The majority of persons in the General Sample are women, and almost half of this group is married. Sixty per- cent of the sample hold at least a college degree, while 0.5 percent do not have a high school education. All adult age groups are represented, with a slight majority being under 30 years of age. The major work areas of persons in this sample are business, education and social science. However, many students and housewives are also included. In order to compare separation anxiety scores of blacks and whites, an attempt was made to include a large number of blacks; the final result was a sample that was approximately 15 percent black. This group is not completely representative of blacks in this society. While a portion of this sample was collected in a Detroit ghetto area, most of the blacks participating were from Michigan State Univer- sity and either held or were pursuing advanced degrees. This General Sample resembles most university- obtained ones, in that most of the participants are middle- class and many students are included. It differs from this stereotype, however, in that the majority of persons taking part in the study are not students, but are working outside the home or raising families at home. Many people answered 108 the test and questionnaire items while waiting for planes, watching their children; while taking a coffee break in their Offices, or at home at the end of a day of teaching, treating patients, or advising clients. Others were approached on weekends and took time out from their shopping or relaxing at home to participate. Therefore, while this sample is not truly random, it is representative of most segments of society. For more detailed information describ- ing this group, see Table 5.1. The second group participating in this study was a group of freshmen women at Michigan State University, residing in Rather Hall. This group was chosen to test the validity Of the separation anxiety test because they were a relatively homogeneous population and were simultaneously in the midst of their first major separation experience-- leaving home to attend college and live on campus. A totally female group was chosen because they are a homogeneous group, and because it has been stated that women are more likely to experience separation anxiety than men. An addi— tional reason for sampling women is that most of the case material, both published and that which was gathered by the researcher, described and analyzed the psychodynamics opera- tive in women rather than men. Almost all of the women participating in this study were 18 years old. While a small portion of the residents of Rather Hall were over this age, they were upperclassmen who were excluded from the study. The only persons beyond .OGHOCOOH ou moo ucmoumm OOH on ppm #0: was mHmuOB .>HOOODMO comm How cow u so 109 o.H OOSOOAS m.~ omumummmm m.mH Omouo>flo m.om mamahm Mo¢v UTflHHMS monmum Hmuwumz m.m~ OOHOOO mumoomuw m.mm moumop OOOHHOO N.Hm momaaoo OEOm v.m muoopmum Hoonom cmflm m.n moowcoHHmomHz m.o Hoocom omen nonwommwws o.H pmaaflxm . om o.H ucmEmOHOmcm m.vw muHcS m.~ O>Huoumxm H.ma xomam o.m OOH>HOm doom Anemone mowmmo . m mam o.m .umaamu game. H Hmowumao m.wm Hmamm m.m Omaaflxmcs xom H.h ucmEmmmcmE mappflz H.m HMOHGEOOB m.o +omIHm. Aumflmmu xumao N.~H omlaq ~.vH .wumumnommv HH Hmowumao v.m~ owlam m.mH ucoosum «.mv omumm m.vm Hmcowmmmwoum m.vH NNImH coflummoooo mod mnemouwm Ofiumfiumuomumco opcwoumm Owumwumuomnmnu mHmEmm Hmumcmo MOM memo Dandmumoema H.m OHQMB 110 the 18-19 age range who were included were those who had attended a junior college while living at home, and then transfered to Michigan State. In this way, all of the per- sons sampled were living away from home for the first time. The total number of Rather Hall residents who met this criterion was 335. Final number of test forms and questionnaires collected was 315, for a return rate of 94 percent. Most of the 20 persons who did not wish to participate stated that they were too busy with coursework and did not have the time. Approximately six people refused because they were not interested. This sample is therefore representative of those persons attending a large, public Midwestern University and living away from home for the first time. Regarding the participants' choice of living area, the vast majority of them did not choose to live in an all-female dormitory. Assignment to dormitories is random, with the exception of a few persons who may request special placement. The only choice that is given all potential residents appears on a card that is sent to them and returned prior to their arrival on campus. The card reads: Which type lifestyle Option are you most interested in? (Number your choices, 1, 2, 3.) limited visitation unlimited visitation quiet house 111 Rather Hall was placed into the "limited visitation" cate- gory, indicating that residents placed there requested an environment which provides some quite study time. For further descriptive information on this sample, see Table 5.2. Hypotheses Groups were separated for comparison on the basis of their responses to items on the background question- naire (see Appendices B and I), which they filled out at the time of test administration. However, the question- naires for the General and Freshmen samples differ in accordance with the slightly different sets of hypotheses. The hypotheses that apply to both samples will be listed first, along with the background questionnaire item(s) used to define each criterion group. Then the hypotheses that apply to only one of the samples will be listed. (The General sample will be abbreviated as GS and the Freshmen sample as FS.) Hypotheses Applying to Both Samples H01: There will be no difference between the mean separation anxiety test scores of persons indicating they were afraid to go to school and those who did not. (GS item 19, F5 item 15) Hal: Those persons indicating they were afraid to go to school will score significantly higher on the separa- tion anxiety test than those who did not. 112 Table 5.2 Demographic Data for Freshmen Sample Characteristic Percenta Age 18 87.4 19 5.8 20 4.1 21+ 2.7 Race Black 4.4 White 93.5 Other 2.0 Religious background Catholic 38.8 Protestant 41.5 Jewish 4.4 Other 14.3 College major Agriculture and Natural 14.3 Resources Arts and Letters 11.9 Business 22.8 Communication Arts and 5.8 Sciences Education 5.8 Engineering 3.7 Human Ecology 1.0 Human Medicine 6.1 Natural Science 10.9 Social Science 6.8 Urban Development 0.3 Veterinary Medicine 6.1 No preference 4.4 100 an = 300 for each category. percefit due to rounding. Totals may not add to 113 H02: There will be no difference between the mean test scores of those indicating oral traits and those who did not. (GS items 11 and 14, F8 items 8 and 10) Haz: Those persons indicating oral traits will score signi- ficantly higher on the test than those who did not. H03: There will be no difference between the mean.test scores of those indicating respiratory or gastrointestinal problems and those who did not. (GS item 13, F8 item 32) Ha3 : Those persons indicating respiratory or gastrointestinal problems will score significantly higher on the test than those who did not. H04: There will be no difference between the mean test scores of those who lost a parent, as a child, and those who did not. (GS items 20, 21, 22, F8 items 16, 17, 18) Ha4: Those persons who lost a parent, as a child, will score significantly higher on the test than those who did not. HOS: There will be no difference between the mean test scores of those who stated that they disliked the idea of living alone and those who did not. (GS item 10, F8 item 28) Has: Those persons indicating they disliked the idea Of living alone will score significantly higher on the test than those who did not. 114 Hypotheses Applying Only to the General Sample H06: There will be no difference between the mean test scores of women and men. (GS item 2) Ha6: Women will score significantly higher on the test than men. Hypotheses Applying Only to the Freshmen Sample H07: There will be no difference between the mean test scores of those who state that they are homesick and those who state that they are not. (FS item 59) Ha7: Those who stated that they are homesick will score signifi- cantly higher on the test than those who state that they are not. H08: There will be no difference between the mean test scores of those who state that they are separation anxious and those who state that they are not. (FS item 26) Ha8: Those who stated that they are separation anxious will score significantly higher on the test than those who state that they are not. H09: There will be no difference between the mean test scores of those who indicate that most of their close relationships have been with family members and those who do not. (FS items 24 & 25) Hag: Those persons indicating that they have had most of their close relationships with family members will score signi- ficantly higher on the test than those who did not. 115 Analysis The first analysis completed on the data collected during this study was the aforementioned procedure con- cerning the test items (see chapter 4). In lieu of a factor analysis program, a cluster analysis program named PACKAGE was used. This program was designed by Dr. John Hunter of the Michigan State University psychology department, for use in analyzing social science data. As mentioned previously, this program was used because no factor analysis program was available that was capable of adequately handling the large number of variables present in this study. Both factor and cluster analyses reveal the rela- tionships between items, and indicate which items appear to be assessing the same personality trait. The chief differ- ence between the two forms of analysis is that the resultant groupings of test items are independent in factor analysis, but not in cluster analysis. For example, it was reported in chapter 4 that slight positive relationships were found to exist between most of the clusters that were formed. This means that the clusters or, in this case, subscales, are relatively independent. However, if a factor analysis program had been used, the relationship between subscales would be zero, indicating absolute independence. In either case the analysis yields valuable infor- mation concerning the effectiveness of test items and how they should be scored. Items that do not adequately cluster or factor with other items are not likely to be valid and 116 should not generally be retained. Those items that con- sistently relate to Other items, on the other hand, can be viewed as potential valid indicators of a specific trait. The nature of that trait, and whether it is part Of what the test intended to assess, however, is not determined by cluster or factor analyses. What it does provide is a set of subscales that are relatively independent and are scored and examined indepen- dently. Whether any or all of the subscales are valid assessors of separation anxiety can only be determined by studies designed for this purpose. For this reason, the sampling of the first-term dormitory freshmen was done. In this way their current attitudes, feelings, and physical reactions to separation could be compared to their scores on each of the separation anxiety subscales. Results described in chapter 6 indi- cated that some of these subscales might be valid assessors of separation anxiety. However, Cronbach and Meehl (1967) point out that construct validity is never proven by a single validation study, even those that include a factor or cluster analysis. This is only accomplished through repeated studies and cross-validation efforts. The second analysis done on the collected data involved the method for comparing the subgroups in the Freshmen Sample, as mentioned above. For example, "home- sick" students were compared to students who were not home- sick. Since this required a test of the difference between 117 the mean subscale scores of each Of two groups, the E-test procedure was used. One-tailed tests, compatible with the stated directional hypotheses, were run with the alpha level set at .05. Similar tests were conducted on the General Sample data. The assumptions that must be met to justify the use of a p—distribution to test differences between means are: (a) the samples each have a normal distribution, and (b) the variances of compared groups is approximately the same. The assumption of normality is said to be met in this case because of the large sample sizes. In these instances, 21:400, p2=300, it is extremely unlikely that the distri- bution would not be normal. Regarding the assumption of homogenity Of variance, a study such as this one cannot violate this assumption. Since the subgroups being compared are from the same sample, their variances are identical. Therefore, both assumptions are met and the use of the E-test is appropriate. The results of all of the p—tests are reported in the next chapter. CHAPTER 6 RESULTS The separation anxiety test was administered to two separate groups, the General Sample in July-August, 1978, and the Freshmen Sample in October, 1978. The former was a diverse group sampled for the purpose of executing a cluster analysis on the test items. The latter group was used for a separate validity study with the major hypotheses involving the comparison of homesick and non-homesick college freshmen in their first term at the university. The extent of home- sickness is thus used as a criterion for separation anxiety. Homesickness, as well as other hypothesized concepts, was determined by subjects' responses to a background ques- tionnaire which was administered with the test. The General Sample and Freshmen Sample received different background questionnaires (see Appendices B and I), but the content of the two is very similar. Therefore, most of the hypotheses apply to both samples. However, the E-tests were used and reported on each sample independently. Once subgroups had been delineated by their responses to a particular questionnaire item, their responses were compared to the responses of subjects answering in the 118 119 opposite direction. For example, subjects in the General Sample answering "hardly ever" to background item #19, "I was afraid to go to school as a child," were compared to those who answered "sometimes" or "often." Hypotheses were formulated on the basis of mean test scores (see Chapter 5), rather than subscale scores because it was not known at the advent of the study whether sub- scales would be formed. Because the cluster analysis was successful, creating six subscales (four of which are scored for separation anxiety), the mean differences of each of these subscale scores were computed. Then a E-test was used to assess mean differences, with the probability of signifi- cance set at the p < .05 level. In the following report of the results of these t-tests, comparisons are noted for each of the six subscales formed by the cluster analysis. The subscales scored for separation anxiety, Overprotection, General Separation Anxiety, Self-acceptance, and Mother Hostility are said to support the hypotheses listed if the criterion group's mean scores were significantly higher than those of the control group. A high score on Overprotection indicates insufficient individuation. On the General Separation Anxiety subscale a high score is self—explanatory, as is the meaning of the Mother Hostility score. For the Self-acceptance subscale a high score is indicative of 19! self-esteem. While this may seem confusing, it simplified the implementation of a computer program for scoring. 120 The two remaining subscales, Basic Personality Type and Father, were not scored for separation anxiety and were not expected to be significant in only one direction. There- fore both of these scales were assessed by comparing mean differences with a two-tailed p—test. The probability of significance was again set at the p < .05 level. High scores on the Basic Personality Type subscale are said to indicate a more aggressive independent or pseudo-independent personality type, while low scores indicate the passive— dependent type of personality associated with higher levels of separation anxiety. This scale was created to separate these two types, once separation anxiety had been indicated on one or more of the other scales. It would thus only be expected to show significance in cases of intense separation anxiety, when the repressed, passive-dependent personality would be in evidence. The Father subscale, as stated previously, is com- prised of parallel "mother" items and exists only in the test form given to the Freshmen Sample. A high score on this scale indicates a rather hostile attachment to the father, while a low score reflects an intense positive attachment. Results pertaining to this subscale are reported for all tests conducted on the Freshmen Sample. The scores pertaining to the psychosomatic items are reported for both samples. These items asked about the incidence of ulcer and migraine headache for the subject or his family, and about the frequency of colds for only the 121 subjects themselves. Although these are also somewhat exploratory in nature it was expected that subjects scoring significantly higher on the separation anxiety subscales would also report a greater incidence of one or more of these physical problems. Finally, background information from the question- naires was used to compare groups on the basis of certain facts that are not part of the hypotheses for this study. For example, results were compared and analyzed on the basis of age and religion. These comparisons were made as a way of further exploring the realm and concomitants Of separa- tion anxiety. Because these are purely speculative and are not an integral part of the study, they are reported in an addendum to this chapter, entitled, "Additional Results." Tables are presented here following a statement of each alternative hypothesis. Hypotheses relevant to both samples followed by tables relating to each. However, before examining the means stated in the tables, some explanatory notes are necessary. The true-false continuum response code used with the General Sample ranged from false to true. This allowed easier scoring for separation anxiety, since more items went in the true direction. However, subjects stated that they found this to be confusing and the response code used with the Freshmen Sample was changed to a more standard true—false. 122 The computer program was adjusted appropriately so that higher scores always indicate a greater degree of separation anxiety--on all of the subscales and for both samples. Thus, in reading the accompanying tables, the criterion group mean was expected to be greater than the control group mean, indicating a greater degree of separation anxiety. The reader is also reminded that a high score on the Self-acceptance subscale indicates low self-esteem. Hypotheses Tests I331 Those persons indicating they were afraid to go to school will score significantly higher on the separa- tion anxiety test than those who did not. This hypothesis was supported for the General Sample for subscales Overprotection, General Separation Anxiety, Self-acceptance, and Mother Hostility, as well as the psycho— somatic item Migraine. Additionally, a significant differ- ence was found between compared groups on the Basic Person- ality Type subscale, with the subjects who feared school scoring on the high, dependency-oriented end of the scale (see Table 6.1). Regarding the Freshmen Sample, the hypothesis was supported for the General Separation Anxiety, Self- Acceptance, and Mother Hostility subscales. There was also a significant difference when groups were compared on the Father subscale, with the group expressing a fear of school demonstrating a hostile, estranged relationship (see Table 6.2). 123 Table 6.1 Comparison of Mean Subscale Scores by Degree of Fear of Going to School as a Child for General Sample (Hypothesis 1) Background questionnaire item 19 Group 1 = sometimes or often afraid to go to school as a child (£7103) Group 2 = hardly ever afraid to go to school as a child (pé292) Subscale R S.D. Efvalue pa Basic Personality Group 1 1.9119 .390 4.86 .000**** Group 2 1.6913 .398 Overprotection Group 1 1.8005 .580 4.97 .OOO**** Group 2 1.4981 .512 General Separation Anxiety Group 1 1.7201 .396 6.30 .000**** Group 2 1.4316 .401 Self-acceptance Group 1 1.7201 .396 6.30 .000**** Group 2 1.9433 .300 Mother Hostility Group 1 1.3022 .505 6.57 .000**** Group 2 .9477 .458 Colds Group 1 2.0291 .880 .71 .240 Group 2 1.9521 .973 Migraine Group 1 2.2427 1.445 2.19 .015* Group 2 1.8973 1.353 Ulcer Group 1 1.7864 1.303 -.25 .400 Group 2 1.8253 1.345 aA one-tailed Eftest was used for all subscales, except for Basic Personality, in which a two-tailed Eftest was used. * ** p<,05 p< *** p < .005 **** p < .001 124 Table 6.2 Comparison of Mean Subscale Scores by Degree of Fear of Going to School as a Child for Freshmen Sample (Hypothesis 1) Background questionnaire item 15 Group 1 = afraid to go to school as a child (2&124) Group 2 = not afraid to go to school as a child (pél68) Subscale R S.D. Efvalue pa Basic Personality Group 1 1.8055 .378 1.60 .112 Group 2 1.7338 .378 Overprotection Group 1 1.8814 .400 .86 .195 Group 2 1.8400 .12 General Separation Anxiety Group 1 2.6192 .420 3.49 .001**** Group 2 2.4389 .447 Self-acceptance Group 1 1.3576 .347 4.87 .000**** Group 2 1.1613 .335 Mother Hostility Group 1 2.2181 .427 4.76 .000**** Group 2 1.9647 .466 Colds Group 1 1.4435 .904 .36 .359 Group 2 1.4048 .904 Migraine Group 1 1.9113 1.385 .04 .484 Group 2 1.9048 1.350 Ulcer Group 1 1.6613 1.242 .90 .186 Group 2 1.5357 1.137 Father Group 1 1.9068 .588 1.76 .040* Group 2 1.7852 .579 aA one-tailed Eftest was used for all subscales, except for Father and Basic Personality, in which two-tailed Eftests were used. * p < .05 ** p < .01 *** p < .005 **** p< .001 125 E32 Those persons indicating oral traits will score signifi- cantly higher on the test than those who did not. This hypothesis was tested through various means. In each sample the following groups were compared: (a) over- weight or underweight vs. not; (b) smoker vs. nonsmoker; and (c) overweight and smoker vs. neither. The first comparison was made on the assumption that the denial of an oral fixa— tion would be indicated by underweight, while a more overt oral orientation would be evidenced by overweight. The second comparison was made for the obvious reason that smoking is an oral activity. The third comparison was made to see what type of difference, if any, would be seen when comparing those who indulged in these two forms of overt oral behaviors as compared to those who did not indulge in either. Therefore, three sets of results are presented for each sample. The hypothesis was supported pertaining to the first comparison regarding weight, in the General Sample, for the Overprotection and Self-acceptance subscales. In the Fresh- men Sample, it was supported for the Self-acceptance sub- scale (see Tables 6.3 and 6.4). The hypothesis was supported for the second compari- son, based on smoking, for the Self-acceptance and Mother Hostility subscales, in both the General and Freshmen samples. Also, in the Freshmen Sample only, there was a significantly 126 Table 6.3 Comparison of Mean Subscale Scores by Degree of Orality (Weight) for General Sample (Hypothesis 2) Background questionnaire items 11 and 12 Group 1 = overweight or underweight (nfl48) Group 2 - not overweight and not underweight (2&247) Subscale i S.D. tfvalue pa Basic Personality Group 1 1.7613 .431 .47 .639 Group 2 1.7414 .393 Overprotection Group 1 1.6699 .543 2.64 .005*** Group 2 1.5213 .542 General Separation Anxiety Group 1 1.5159 .409 .33 .369 Group 2 1.5013 .425 Self-acceptance Group 1 2.0870 .328 3.94 .000**** Group 2 1.9582 .306 Mother Hostility Group 1 1.0408 .489 .02 .492 Group 2 1.0398 .499 Colds Group 1 1.9932 .944 .34 .367 Group 2 1.9595 .953 Migraine Group 1 1.9865 1.395 -.01 .496 Group 2 1.9879 1.381 Ulcer Group 1 1.9189 1.378 1.20 .116 Group 2 1.7530 1.303 a . A one tailed tftest was used for all subscales, except for Basic Personality, in which a two-tailed tftest was used. * ** p < .05 p < .01 *** p < .005 **** p< .001 127 Table 6.4 Comparison of Mean Subscale Scores by Degree of Orality (Weight) for Freshmen Sample (Hypothesis 2) Background questionnaire items 8 and 9 Group 1 - overweight or underweight (5&96) Group 2 = not overweight and not underweight (gell2) Subscale i S.D. tfvalue pa . . b Ba31c Personality Group 1 1.8178 .385 1.19 .236 Group 2 1.7536 .391 Overprotection Group 1 1.8276 .445 -1.23 .111 Group 2 1.8963 .363 General Separation Anxiety Group 1 2.4708 .449 -1.44 .075 Group 2 2.5595 .438 Self-acceptance Group 1 1.3507 .382 3.42 .001**** Group 2 1.1801 .336 Mother Hostility Group 1 2.1097 .470 .49 .313 Group 2 2.0774 .477 Colds Group 1 1.4792 .962 .39 .347 Group 2 1.4286 .887 Migraine Group 1 1.8125 1.324 -1.12 .132 Group 2 2.0268 1.411 Ulcer Group 1 1.5312 1.151 .14 .444 Group 2 1.5089 1.107 Fatherb Group 1 1.9038 .593 .39 .697 Group 2 1.8717 .588 a . A one-tailed tftest was used for all subscales, except for Father and Basic Personality in which two-tailed Eftests were used. b . . Exploratory subscale, not scored for separation anxiety. * ** p < .05 p < .01 *** p < .005 **** p< .001 128 greater family incidence of migraine headache (see Tables 6.5 and 6.6). Finally, in the third comparison based on orality (overweight and smoker vs. neither), the hypothesis was sup- ported for the Self-acceptance and Mother Hostility subscales in the General Sample. In the Freshmen Sample, the hypoth- esis was supported for the Overprotection and Self-acceptance subscales (see Tables 6.7 and 6.8). 333 Those persons indicating respiratory or gastrointestinal problems will score significantly higher on the test than those who did not. This hypothesis is based on theoretical material suggesting that these physical ailments are more prevalent among people with separation anxiety. For the General Sample the background questionnaire asked for a report of present health problems. Due to the small number of reported problems, respiratory and gastrointestinal were grouped together. Using this method, the hypothesis was supported for the General Separation Anxiety and Mother Hostility sub- scales. As expected, significant differences were also found for the colds and migraine items (see Table 6.9). For the Freshmen Sample, a slightly different approach was taken. The background questionnaire examined current somatic symptoms that might relate to the separation experience, rather than ongoing physical problems. This precluded respiratory problems for most people and was not 129 Table 6.5 Comparison of Mean Subscale Scores by Degree of Orality (Hypothesis 2) (Smoking) for General Sample Background questionnaire item 14 Group 1 = smoker (2?130) Group 2 = nonsmoker (2?26l) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.7403 .417 -.35 .730 Group 2 1.7554 .405 Overprotection Group 1 1.5250 .522 -1.37 .085 Group 2 1.6056 .557 General Separation Anxiety Group 1 1.5113 .442 .21 .418 Group 2 1.5020 .405 Self—acceptance Group 1 2.0534 .326 2.10 .019* Group 2 1.9815 .317 Mother Hostility Group 1 1.1132 .536 2.20 .014* Group 2 .9972 .467 Colds Group 1 2.0538 .959 1.21 .114 Group 2 1.9310 .942 Migraine Group 1 2.0642 1.413 .49 .313 Group 2 1.9732 1.377 Ulcer Group 1 1.8923 1.365 .80 .212 Group 2 1.7778 1.317 a . A one-tailed Eftest was used for all subscales, except for Basic Personality, in which a two-tailed tftest was used. b . Exploratory subscale, not scored for separation anXiety. * p < .05 ** p<.01 *** p < .005 **** p < .001 130 Table 6.6 Comparison of Mean Subscale Scores by Degree of Orality (Hypothesis 2) (Smoking) for Freshmen Sample Background questionnaire item 10 Group 1 = smoker (pf47) Group 2 nonsmoker (2&245) Subscale i S.D. Efvalue pa . . b BaSic Personality Group 1 1.7116 .361 -1.02 .300 Group 2 1.7731 .383 Overprotection Group 1 1.9421 .357 1.50 .068 Group 2 1.8450 .417 General Separation Anxiety Group 1 2.5804 .412 1.10 .135 Group 2 2.5021 .451 Self-acceptance Group 1 1.3599 .390 2.44 .003*** Group 2 1.2226 .345 Mother Hostility Group 1 2.2199 .510 2.31 .011* Group 2 2.0481 .458 Colds Group 1 1.7894 .804 .57 .286 Group 2 1.4082 .917 Migraine Group 1 2.3830 1.497 2.71 .004*** Group 2 1.8041 1.313 Ulcer Group 1 1.6596 1.238 .45 .323 Group 2 1.5755 1.173 Fatherb Group 1 1.7834 .478 -.73 .461 Group 2 1.8517 .603 a . A one-tailed tftest was used for all subscales, except for Father and Basic Personality, in which two-tailed tftests were used. Exploratory subscale, not scored for separation anxiety. * ** P < .05 p < .01 *** p < .005 **** p < .001 131 Table 6.7 Comparison of Mean Subscale Scores by Degree of Orality (Weight and Smoking Behavior) for General Sample (Hypothesis 2) Background questionnaire items 11 and 14 Group 1 = overweight and smoker (gf43) Group 2 = not overweight and nonsmoker (2&198) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.6884 .380 -1.08 .284 Group 2 1.7576 .383 Overprotection Group 1 1.6663 .448 .86 .196 Group 2 1.5896 .547 General Separation Anxiety Group 1 1.5282 .376 .19 .425 Group 2 1.5155 .405 Self-acceptance Group 1 2.1873 .309 4.68 .000**** Group 2 1.9500 .299 Mother Hostility Group 1 1.1657 .518 1.93 .027* Group 2 1.0113 .465 Colds Group 1 2.1163 .956 1.07 .144 Group 2 1.9444 .957 Migraine Group 1 1.2558 1.465 .93 .177 Group 2 1.0354 1.394 Ulcer Group 1 1.6977 1.282 -.14 .446 Group 2 1.7273 1.293 a . A one-tailed Eftest was used for all subscales, except for Basic Personality, in which a two-tailed tftest was used. b . . Exploratory subscale, not scored for separation anXiety. * ** p < .05 p < .01 *** p < .005 **** p < .001 132 Table 6.8 Comparison of Mean Subscale Scores by Degree of Orality (Weight and Smoking Behavior) for Freshmen Sample (Hypothesis 2) Background questionnaire items 8 and 10 Group 1 more than ten pounds overweight and smoker (gell) Group 2 = not overweight and a nonsmoker (n?112) Subscale H S.D Efvalue pa . . b BaSic Personality Group 1 1.8131 .315 .30 .762 Group 2 1.7756 .396 Overprotection Group 1 2.0292 .466 1.36 .038* Group 2 1.8632 .378 General Separation Anxiety Group 1 2.5758 .313 .46 .323 Group 2 2.5134 .436 Self-acceptance Group 1 1.5682 .271 4.12 .000**** Group 2 1.1638 .314 Mother Hostility Group 1 2.2222 .453 1.00 .161 Group 2 2.0784 Colds Group 1 1.2727 .905 -.61 .273 Group 2 1.4464 .909 Migraine Group 1 1.8182 1.401 .25 .401 Group 2 1.9286 1.380 Ulcer Group 1 1.5455 1.214 .05 .480 Group 2 1.5268 1.131 Fatherb Group 1 1.9610 .546 .37 .356 Group 2 1.8895 .617 a . A one-tailed tftest was used for all subscales, except for Basic Personality and Father, in which a two-tailed Eftest was used. b . . Exploratory subscale, not scored for separation anxiety. * ** p < .05 p < .01 *** p < .005 **** p < .001 133 Table 6.9 Comparison of Mean Subscale Scores on Present Health Problems for General Sample (Hypothesis 3) Background questionnaire item 13 Group 1 - respiratory or gastrointestinal health problems (2§29) Group 2 = no health problems (2e302) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.7793 .446 .34 .735 Group 2 1.7526 .401 Overprotection Group 1 1.6828 .592 1.13 .131 Group 2 1.5674 .521 General Separation Anxiety Group 1 1.6379 .435 1.98 .025* Group 2 1.4838 .398 Self-acceptance Group 1 2.0603 .302 1.23 .110 Group 2 1.9857 .314 Mother Hostility Group 1 1.1767 .418 1.65 .050* Group 2 1.0199 .496 Colds Group 1 2.7931 1.013 5.18 .000**** Group 2 1.8808 .896 Migraine Group 1 1.9310 1.412 -.01 .496 Group 2 1.9338 1.360 Ulcer Group 1 2.4483 1.526 2.90 .002*** Group 2 1.7152 1.278 a . A one-tailed Eftest was used for all subscales, except for Basic Personality, in which a two-tailed Eftest was used. b . Exploratory subscale, not scored for separation anx1ety. * p < .05 ** p < .01 *** p < .005 **** p < .001 134 asked in the questionnaire. Upset stomach was one of the symptoms that could be checked, however, and the gastro- intestinal category is based on a comparison of those experiencing this symptom and those who did not. The hypothesis was supported for the Overprotection, General Separation Anxiety, Self-acceptance and Mother Hostility subscales. In addition, significant differences were found in the expected direction on the migraine and ulcer items (see Table 6.10). -H—a4 Those persons who lost a parent as a child will score significantly higher on the test than those who did not. This hypothesis was not tested because there was an insufficient number of subjects in this category for both samples. However, a related comparison was made on the basis of recalled childhood separation from mother. This was done on the General Sample and is reported in Additional Results. 219.5 Those persons indicating they disliked the idea of living alone will score significantly higher on the test than those who did not. Since living alone may preclude participation in a symbiotic type relationship and would be predicted to arouse anxiety in a person with separation anxiety, this question was included for comparison. The question does differ for each sample, however. In the General Sample the subjects 135 Table 6.10 Comparison of Mean Subscale Scores on Present Health Problems for Freshmen Sample (Hypothesis 3) Background questionnaire item 32 Group 1 ' present health problem is upset stomach (gé38) Group 2 = no present health problems (£f140) Subscale i S.D. Efvalue pa . . b BaSic Personality Group 1 1.7263 .392 -.71 .479 Group 2 1.7749 .369 Overprotection Group 1 1.9605 .450 2.07 .020* Group 2 1.8032 .406 General Separation Anxiety Group 1 2.7719 .465 4.63 .000**** Group 2 2.4028 .428 Self—acceptance Group 1 1.3882 .442 3.09 .001**** Group 2 1.1925 .315 Mother Hostility Group 1 2.1637 .480 2.68 .004*** Group 2 1.9391 .453 Colds Group 1 1.5000 1.133 .80 .213 Group 2 1.3643 .867 Migraine Group 1 2.5789 1.518 3.37 .001**** Group 2 1.7571 1.280 Ulcer Group 1 2.0526 1.432 3.37 .001**** Group 2 1.3786 .985 b Father Group 1 1.8008 .622 -.08 .935 Group 2 1.8096 .581 a . A one-tailed Eftest was used for all subscales, except for Father and Basic Personality, in which a two-tailed Eftest was used. b . . Exploratory subscale, not scored for separation anx1ety. * ** p < .05 p < .01 *** p < .005 **** p < .001 136 were asked whether or not they had lived alone. If they answered "Yes," then they were asked to rate how well they enjoyed the experience. But some persons answering the questionnaire answered that item even if they hadn't ever lived alone and simply expressed their Opinion about the idea. But in the Freshmen Sample the item on the question- naire asked each person to respond to the idea of living alone and its appeal to the him or her. Therefore, the results for each sample do not represent the same subject response and should be viewed accordingly.- In the General Sample, the hypothesis was supported for the Overprotection and General Separation Anxiety sub- scales (see Table 6.11). In the Freshmen Sample, the hypothesis was supported for the General Separation Anxiety subscale. The Mother Hostility subscale score was also significantly different between groups, but the criterion group scored lower than the control group, contrary to the hypothesis. Significant differences were also seen on the Father subscale and one of the psychosomatic items, with those persons who disliked the idea of living alone indi- cating a more intense Father attachment and a greater incidence of colds (see Table 6.12). 536 Women will score significantly higher on the test than men. 137 Table 6.11 Comparison of Mean Subscale Scores by Opinion About Living Alone for General Sample (Hypothesis 5) Background questionnaire item 10 Group 1 = those who had lived alone and did not like it (gf44) Group 2 = those who had lived alone and liked it (2f80) Subscale i S.D. Efvalue pa . . b BaSic Personality Group 1 1.7492 .479 .34 .735 Group 2 1.7224 .388 Overprotection Group 1 1.7000 .538 1.67 .049* Group 2 1.5219 .585 General Separation Anxiety Group 1 1.5781 .451 2.56 .012* Group 2 1.3480 .494 Self-acceptance Group 1 2.0715 .327 .67 .253 Group 2 2.0294 .341 Mother Hostility Group 1 1.1165 .500 .35 .363 Group 2 1.0828 .516 Colds Group 1 2.1136 .895 .50 .309 Group 2 2.0250 .968 Migraine Group 1 1.0455 1.430 -.06 .475 Group 2 1.0625 1.408 Ulcer Group 1 1.9091 1.378 .28 .390 Group 2 1.8375 1.345 a . A one-tailed Eftest was used for all subscales except for Basic Personality in which a two-tailed test was used. b . . Exploratory subscale, not scored for separation anXiety. * ** p < .05 p < *** p < .005 **** p < .001 138 Table 6.12 Comparison of Mean Subscale Scores by Opinion About Living Alone for Freshmen Sample (Hypothesis 5) Background questionnaire item 28 Group 1 = those who liked the idea of living alone "not at all" (2&81) Group 2 = those who liked the idea of living alone "very much" (EfSO) Subscale i S.D. tfvalue pa . . b BaSlC Personality Group 1 1.8218 .362 1.75 .083 Group 2 1.6984 .438 Overprotection Group 1 1.8821 .366 .25 .401 Group 2 1.8625 .520 General Separation Anxiety Group 1 2.6448 .427 3.53 .001**** Group 2 2.3388 .560 Self-acceptance Group 1 1.2472 .298 -.99 .161 Group 2 1.3133 .463 Mother Hostility Group 1 1.9739 .495 -3.80 .000**** Group 2 2.2978 .437 Colds Group 1 1.6049 .971 2.00 .024* Group 2 1.2800 .784 Migraine Group 1 1.7901 1.301 .04 .483 Group 2 1.7800 1.329 Ulcer Group 1 1.6049 1.201 .60 .277 Group 2 1.4800 1.111 Fatherb Group 1 1.6587 .475 -4.09 .000**** Group 2 2.0519 .620 a . A one-tailed tftest was used for all subscales except for Basic Personality and Father, in which a two-tailed tftest was used. b . . Exploratory subscale, not scored for separation anXiety. 1: *1: p<,05 p< *** p < .005 * p< .001 139 This hypothesis relates to the aforementioned dis- cussion of sex differences (see Chapter III) and was sup- ported for the Overprotection, General Separation Anxiety and Self-acceptance subscales. This hypothesis was tested only for the General Sample because the Freshmen Sample was all-female (see Table 6.13). Hypotheses Tested for Freshmen Sample Only 3‘37 Those who state that they are homesick will score significantly higher on the test than those who state that they are not. This hypothesis was tested by comparing the low and high homesick groups, as determined by their self-report on the background questionnaire. The group stating that their homesickness was "Nonexistent" was excluded from the com- parison on the assumption that some or all of these persons may have character disorders or even be sociopathic person- alities. The hypothesis was supported for the General Separa- tion Anxiety subscale. The more homesick group also evi- denced a significantly greater incidence of ulcer, and a more intense father attachment (see Table 6.14). £11'18 Those who state that they are separation anxious will score significantly higher on the test than those who do not. 140 Table 6.13 Comparison of Mean Subscale Scores by Sex for General Sample (Hypothesis 6) Background questionnaire item 2 Group 1 = female (2e297) Group 2 = male (gf98) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.7489 .412 .00 .999 Group 2 1.7488 .393 Overprotection Group 1 1.6030 .562 1.65 .050* Group 2 1.4980 .489 General Separation Anxiety Group 1 1.5432 .423 3.04 .002*** Group 2 1.3965 .386 Self-acceptance Group 1 2.0313 .317 2.70 .004*** Group 2 1.9314 .322 Mother Hostility Group 1 1.0429 .490 .19 .425 Group 2 1.0319 .513 Colds Group 1 1.9798 .937 .28 .391 Group 2 1.9490 .988 Migraine Group 1 2.0269 1.400 .99 .162 Group 2 1.8673 1.337 Ulcer Group 1 1.8182 1.328 .08 .469 Group 2 1.8061 1.352 a . A one-tailed tftest was used for all subscales, except for Basic Personality, in which a two-tailed Eftest was used. b . . Exploratory subscale, not scored for separation anxiety. * p < .05 ** p < .01 *** .005 **** p< .001 141 Table 6.14 Comparison of Mean Subscale Scores by Degree of Homesickness Background questionnaire item 30 for Freshmen Sample (Hypothesis 7) Group 1 = noticeably so or very homesick (2&42) Group 2 = not at all homesick (2&68) Subscale i S.D. Efvalue pa . . b BaSic Personality Group 1 1.8357 .392 1.76 .080 Group 2 1.7047 .372 Overprotection Group 1 1.8119 .472 -.70 .243 Group 2 1.8768 .475 General Separation Anxiety Group 1 2.7632 .472 4.62 .000**** Group 2 2.3186 .501 Self-acceptance Group 1 1.3135 .384 1.40 .082 Group 2 1.2071 .388 Mother Hostility Group 1 2.0000 .367 -.85 .199 Group 2 2.0764 .506 Colds Group 1 1.5952 1.014 1.42 .080 Group 2 1.3235 .953 Migraine Group 1 1.8571 1.372 -.26 .399 Group 2 1.9265 1.375 Ulcer Group 1 1.7857 1.335 .87 .194 Group 2 1.5735 1.188 Fatherb Group 1 1.7265 .581 -2.57 .012* Group 2 2.0252 .600 a . (A one—tailed Eftest was used for all subscales, except for Basic Personality and Father, in which a two-tailed Eftest was used. * p<.os *** p < .005 **** b . . Exploratory subscale, not scored for separation anXiety. p < .001 142 This hypothesis was supported for the General Separa- tion Anxiety subscale. The group who described themselves as anxious was also significantly higher on the Basic Personality Type subscale, scoring in the direction of dependency, and reported a significantly greater family incidence of ulcer (see Table 6.15). £129 Those persons indicating that they have had most of their close relationships with family members will score significantly higher on the test than those who did not. This hypothesis was formulated on the premise that symbiotic relationships within the family, that discourage intimacy with anyone who is not a family member, are typi- cally seen in cases of separation anxiety. This hypothesis was supported for the General Separation Anxiety subscale. Also, the criterion group scored significantly higher on the Father subscale, indicating an estranged, hostile rela- tionship with the father (see Table 6.16). Summary All of the hypotheses were supported by scores on at least one of the four separation anxiety subscales. The 16 different group comparisons made under all of the hypoth- eses tests yielded 64 opportunities for the subscales to be found significant in the predicted direction, significant in the opposite direction, or not significant. Of this number, there were 33 cases of significant differences in 143 Table 6.15 Comparison of Mean Subscale Scores by Degree of Separation Anxiety for Freshmen Sample (Hypothesis 8) Background questionnaire item 26 Group 1 high or very high anxiety over being separated from home (Eé33) Group 2 = low anxiety over being separated from home (gellS) Subscale H S.D. tfvalue pa . . b BaSic Personality Group 1 1.8758 .398 2.65 .010** Group 2 1.6834 .358 Overprotection Group 1 1.8236 .447 -.88 .190 Group 2 1.8590 .400 General Separation Anxiety Group 1 2.7357 .405 4.24 .000**** Group 2 2.4084 .386 Self-acceptance Group 1 1.3636 .407 1.80 .037* Group 2 1.2393 .331 Mother Hostility Group 1 2.0774 .428 .37 .356 Group 2 2.0444 .458 Colds Group 1 1.4848 .939 .53 .298 Group 2 1.3913 .876 Migraine Group 1 1.8870 1.466 .76 .226 Group 2 2.0909 1.336 Ulcer Group 1 1.9091 1.400 1.93 .028* Group 2 1.4696 1.071 Fatherb Group 1 1.7439 .592 -1.60 .111 Group 2 1.9373 .616 a . A one-tailed tftest was used for all subscales, except for Basic Personality and Father, in which a two-tailed Eftest was used. b . . Exploratory subscale, not scored for separation anx1ety. * p < .05 ** p < .01 *** p < .005 **** p < .001 144 Table 6.16 Comparison of Mean Subscale Scores by Close Relationships for Freshmen Sample (Hypothesis 9) Background questionnaire item 24 Group 1 - most close relationships with family members (gf63) Group 2 = most close relationships with no one, same or opposite sex peers, equally with both sexes, or older adults (27231) Subscale i S.D. Efvalue pa . . b BaSic Personality Group 1 1.7952 .400 .72 .474 Group 2 1.7565 .374 Overprotection Group 1 1.8135 .336 -1.01 .157 Group 2 1.8719 .425 General Separation Anxiety Group 1 2.6668 .452 3.04 .002*** Group 2 2.4772 .436 Self-acceptance Group 1 1.2027 .319 -1.07 .144 Group 2 1.2565 .364 Mother Hostility Group 1 1.9912 .427 -1.58 .058 Group 2 2.0962 .478 Colds Group 1 1.4127 .775 —.12 .451 Group 2 1.4286 .934 Migraine Group 1 2.0000 1.403 .65 .259 Group 2 1.8745 1.350 Ulcer Group 1 1.4603 1.060 -.95 .172 Group 2 1.6190 1.210 Fatherb Group 1 1.6682 .617 ~2.63 .009** Group 2 1.8848 .568 a . A one-tailed Eftest was used for all subscales, except for Basic Personality and Father, in which a two-tailed Eftest was used. b . . Exploratory subscale, not scored for separation anXiety. * ** *** **** p < .05 p < .01 p < .005 p < .001 145 the predicted direction, one case of significance in the opposite direction, and 30 cases of no significant differ- ences between subscale mean scores. The numerical breakdown of these for each sample appears as Table 6.17 and Table 6.18, and includes similar information on the psychosomatic items as well. Results for the Father subscale are summar- ized in Table 6.19. Additional Results A great deal of information was gathered in both of the samples used in this study, through use of the back- ground questionnaires. Some of this information served to delineate groups for the hypothesis tests already reported. But other comparisons, more speculative in nature, were possible with the use of other questionnaire items. Some of the more interesting results of these comparisons are therefore reported here. The same Eetest procedure, described in the main body of this chapter, was utilized in obtaining these addi- tional results. Also, for comparisons done on age and religion, more than two groups were compared and the appro- priate one-way analysis of variance (ANOVA) procedure was used. The probability of significance for the ANOVA was again set at the p < .05 level, and the conservative Scheffé Multiple Range Test was chosen as the post-hoc procedure to determine the nature of any significant differences found. 146 Table 6.17 Summary of Results for General Samplea Subscales and Number Significant Number Not Psychosomatic Items in Predicted Direction Significant Separation Anxiety Subscales Overprotection 4 3 General Separation 4 3 AnXiety Self-acceptance 5 2 Mother Hostility 4 3 Psychosomatic Items Colds 1 6 Migraine l 6 Ulcer 1 6 aThere were no subscales for the General sample that were significant in the opposite direction. 147 N Honda m msflmumwz a mpaou mEmuH oeumfiomosommm H m sumafiumom upsuoz o mocmummoomumamm humaxc< m coflumuwmmm Hmnmcmo N coapomuoumuw>o mmamombsm mumflxc< cowumummmm pamoHMHcmHm uoz umbesz cowuomuflo mnemommo CH cofluomuflo pmuoflpmum Ga mamuH oeumEOmonowmm uGMUAMHcmHm Hwbfioz ucmoflmacmflm HwbEDz pom moamombsm mHQEmm cwenmwum Mom muHSmmm mo mumEEsm ma.m OHQMB 148 manombsm a H 4 uwnumm ucmoamacmam uoz lucmsnomuum mafiumonv lucmsnomuum mumcofluummmmv mmsHH do wmnssz ppm coax um unmoHMAcmflm cam 304 um unmoHMHcmHm . mmefle mo HmbEsz moses mo HmQEoz AmHmEmm swabmmumv manombsm monumm How muasmmm mH.w OHQMB 149 The following factors were used as bases for com— paring group mean test scores: 1. Rate of school absence (GS item #17, F8 item #12). 2. Frequency of contacts with mother (GS item #16). 3. Report of separation from mother as a child (GS item #24). 4. Living with parents or not (GS item #15). 5. Marital status (GS item #4). 6. Race (GS item #3). 7. Self-report of satisfaction with college adjustment (FS item #31). 8. Age (GS item #1). 9. Religion (FS item #3). 10. Number of peOple close to emotionally at one time (FS item #23). The first comparison, made on rate of school absence, was predicted to show significantly higher test scores for the group who stated that they missed school quite often, as Opposed to those who did not. This relates to theoretical constructs on the subject of school phobia and its psycho- dynamics. Results showed that in the General Sample the Self-acceptance and Mother Hostility subscale scores were significantly higher for the "missed school" group. Also, they reported a significantly greater incidence of migraine headache and colds (see Table 6.20). 150 Table 6.20 Comparison of Mean Subscale Scores by Childhood Absence from School for General Sample Background questionnaire item 17 Group 1 Group 2 children (2&356) as a child missed school more than most children (gf39) as a child missed school an average amount or less than most Subscale i S.D. Efvalue pa . . b BaSlC Personality Group 1 1.7806 .468 .51 .608 Group 2 1.7454 .400 Overprotection Group 1 1.6705 .689 1.13 .261 Group 2 1.5667 .529 General Separation Anxiety Group 1 1.5400 .563 .52 .404 Group 2 1.5032 .401 Self-acceptance Group 1 2.2749 .352 5.73 .000**** Group 2 1.9771 .303 Mother Hostility Group 1 1.3365 .604 4.01 .000**** Group 2 1.0077 .471 Colds Group 1 2.2821 1.050 2.16 .032* Group 2 1.9832 .932 Migraine Group 1 2.5128 1.467 2.51 .012* Group 2 1.9298 1.365 Ulcer Group 1 1.7949 1.321 —.10 .460 Group 2 1.8174 1.335 aA two-tailed Eftest was used for all subscales. b . . Exploratory subscale, not scored for separation anx1ety. * ** p < .05 p < .01 *** p < .005 **** p< .001 151 In the Freshmen Sample, those who missed school more often scored significantly higher on the Father, Self- acceptance, and Mother Hostility subscales. They also reported a greater incidence of migraine headache and colds than did the persons who did not miss school very often (see Table 6.21). Frequency of Calls and/or Visits to Mother The second comparison was made on the basis of reported number of calls and/or visits that the subjects in the General Sample made to their mothers. Those who fre- quently made these contacts scored significantly higher on the Overprotection and General Separation Anxiety subscales than those who did not (see Table 6.22). This informal hypothesis was based on the assumption that peOple who call and visit their mothers much more often than most tend to have more separation anxiety. Separated from Mother as a Child It was expected that persons who stated that they had been separated from their mothers when they were children would be somewhat higher on separation anxiety. Great dif— ferences were not expected, however, because the age at separation, circumstances, and duration of separation are not known. Results showed only one significant difference. 152 Table 6.21 Comparison of Mean Subscale Scores by Childhood Absence from School for Freshmen Sample Background questionnaire item 12 Group 1 as a child missed at least one week of school a year (pf58) Group 2 = as a child almost never absent from school (2&79) Subscale i S.D. Efvalue pa . . b BaSlC Personality Group 1 1.7236 .389 -.82 .413 Group 2 1.7797 .399 Overprotection Group 1 1.8604 .429 -.25 .806 Group 2 1.8781 .405 General Separation Anxiety Group 1 2.4690 .502 -.40 .690 Group 2 2.4994 .384 Self-acceptance Group 1 1.3593 .393 3.42 .002**** Group 2 1.1501 .323 Mother Hostility Group 1 2.1327 .459 2.28 .024 Group 2 1.9568 .437 Colds Group 1 1.8793 .860 2.77 .006*** Group 2 1.4557 .903 Migraine Group 1 2.1724 1.440 2.48 .006*** Group 2 1.6076 1.224 Ulcer Group 1 1.7586 1.288 1.02 .308 Group 2 1.5443 1.152 Fatherb Group 1 1.9087 .599 1.98 .050* Group 2 1.7040 .599 a A two-tailed tftest was used for all subscales. b . . Exploratory subscale, not scored for separation anXiety. * p < .05 ** p < .01 *** p < .005 **** p < .001 153 Table 6.22 Comparison of Mean Subscale Scores by Frequency of Contact with Mother for General Sample Background questionnaire item 16 Group 1 = call or visit mother every day (gf4l) Group 2 = call or visit mother a few times a year (E?35) Subscale i S.D. Efvalue pa . . b BaSlC Personality Group 1 1.7642 .429 .59 .560 Group 2 1.7054 .446 Overprotection Group 1 1.8780 .546 3.05 .004*** Group 2 1.4686 .623 General Separation Anxiety Group 1 1.8175 .441 5.38 .000**** Group 2 1.2654 .451 Self-acceptance Group 1 2.0725 .299 1.45 .152 Group 2 1.9729 .299 Mother Hostility Group 1 2.0854 .543 1.01 .218 Group 2 1.9714 .422 Colds Group 1 2.2195 1.037 1.68 .098 Group 2 1.8286 .985 Migraine Group 1 1.9268 1.385 .04 .968 Group 2 1.9143 1.337 Ulcer Group 1 1.5610 1.163 -l.63 .108 Group 2 2.0571 1.494 a . A two-tailed‘tftest was used for all subscales. b . . Exploratory subscale, not scored for separation anXiety. * p < .05 ** p < .01 *** p < .005 *‘k‘k * p < .001 154 Subjects in the General Sample who stated that they had been separated from their mothers scored significantly higher on the Mother Hostility subscale (see Table 6.23). A compari- son of groups from the Freshmen Sample was not made because there was not a sufficient number of subjects in this cate- gory. Living with Parents or Not For the General Sample a comparison was made between the test scores of peOple who were living with their parents and those who were living in a different part of the country than their parents. Even though some of the subjects who reported that they lived with their parents were under 21 years of age, a significantly higher score was anticipated when compared with the mean score of the group who had moved away. This was based on the assumption that adults who remain at home with their parents are doing so because they are symbiotically tied to them and are afraid to go out on their own. This informal hypothesis was supported by the Overprotection, General Separation Anxiety and Mother Hostility subscale scores (see Table 6.24). Marital Status Marital status was considered as a possible pre- dictor of separation anxiety, in that it is difficult to maintain a child-like self-image with unusually strong parental attachments while being married. Also, the sexual 155 Table 6.23 Comparison of Mean Subscale Scores by Long Childhood Separation from Mother for General Sample Background questionnaire item 24 Group 1 = separated from mother for long time as a child (2733) Group 2 = not separated from mother for a long time as a child (2e362) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.6545 .385 -1.39 .165 Group 2 1.7574 .408 Overprotection Group 1 1.4955 .570 -.89 .742 Group 2 1.5844 .544 General Separation Anxiety Group 1 1.4773 .508 -.42 1.346 Group 2 1.5095 .410 Self-acceptance Group 1 1.9894 .391 —.32 1.500 Group 2 2.0080 .314 Mother Hostility Group 1 1.2646 .517 2.74 .012* Group 2 1.0197 .489 Colds Group 1 2.0000 1.000 .18 1.720 Group 2 1.9696 .945 Migraine Group 1 2.2727 1.506 1.24 .432 Group 2 1.9613 1.372 Ulcer Group 1 2.2424 1.458 1.93 .108 Group 2 1.7762 1.316 a A two-tailed Eftest was used for all subscales. b . . Exploratory subscale, not scored for separation anxiety. * p < .05 ** p < .01 *** p < .005 *** p < .001 156 Table 6.24 Comparison of Mean Subscale Scores by Parental Residence for General Population Background questionnaire item 15 Group 1 - parents live in the same house as respondent (pf4l) Group 2 = parents live in another part of the country (p764) Subscale i S.D. ‘tfvalue pa . . b BaSic Personality Group 1 1.7447 .468 .43 .667 Group 2 1.7055 .445 Overprotection Group 1 1.8829 .615 3.46 .002* Group 2 1.4992 .511 General Separation Anxiety Group 1 1.7058 .410 3.63 .000**** Group 2 1.4138 .397 Self-acceptance Group 1 2.0405 .292 1.16 .248 Group 2 1.9615 .367 Mother Hostility Group 1 1.2561 .430 2.04 .044* Group 2 1.0723 .464 Colds Group 1 2.2439 .888 1.43 .156 Group 2 1.9531 1.090 Migraine Group 1 1.9268 1.385 -.43 .666 Group 2 2.0469 1.385 Ulcer Group 1 1.8293 1.395 -.39 .698 Group 2 1.9375 1.379 aA two-tailed tftest was used for all subscales. bExploratory subscale, not scored for separation anxiety. * ** p < .05 p < .01 *** p < .005 **** p < .001 157 identity confusion and/or body-image problems seen in more extreme cases of separation anxiety are contraindicative of adequate heterosexual adjustment. Therefore, while all single persons are not said to be classifiable in this manner, or all married persons in the Opposite manner, it seemed likely that a comparison of group means would reveal higher scores for the single group. This prediction held true for the General Sample, where persons over the age of 23 were compared on the basis of marital status. The singles group scored significantly higher on all of the subscales scored for separation anxiety; Overprotection, General Separation Anxiety, Self-acceptance, and Mother Hostility (see Table 6.25). m Racial differences pertaining to mental health are often cited, usually noting poorer mental health among blacks if there is no control of socioeconomic status in the study, and no differences found when this variable is controlled. Since the socioeconomic status of blacks and whites in the General Sample appeared to be approximately equal, no dif- ferences were expected. The results confirmed this for all of the test subscales and the psychosomatic items. 158 Table 6.25 Comparison of Mean Subscale Scores by Marital Status for General Sample Background questionnaire items 1 and 4 Group 1 = over 23 years of age and not married (né172) Group 2 = over 23 years of age and married (2&165) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.7450 .421 -.75 .453 Group 2 1.7776 .375 Overprotection Group 1 1.6445 .567 2.15 .032* Group 2 1.5167 .523 General Separation Anxiety Group 1 1.5229 .446 2.06 .040* Group 2 1.4313 .365 Self-acceptance Group 1 2.0386 .325 1.96 .050* Group 2 1.9715 .302 Mother Hostility Group 1 2.0698 .482 2.77 .006*** Group 2 1.9287 .453 Colds Group 1 1.9593 .874 -.04 .988 Group 2 1.9636 1.023 Migraine Group 1 2.0465 1.409 .75 .446 Group 2 1.9333 1.367 Ulcer Group 1 1.9302 1.379 1.84 .066 Group 2 1.6667 1.241 aA two-tailed Eftest was used for all subscales. b . . Exploratory subscale, not scored for separation anx1ety. * ** *** **** p < .05 p < .01 p < .005 p < .001 159 Table 6.26 Comparison of Mean Subscale Scores by Race Background questionnaire item 3 Group 1 = white (gf336) Group 2 = black (2?57) for General Sample Subscale i S.D. Efvalue pa Basic Personality Group 1 1.7581 .415 1.00 .319 Group 2 1.6998 .360 Overprotection Group 1 1.5735 .557 -.61 .544 Group 2 1.6211 .475 General Separation Anxiety Group 1 1.4975 .420 -.96 .240 Group 2 1.5550 .415 Self-acceptance Group 1 2.0093 .326 .35 .728 Group 2 1.9932 .289 Mother Hostility Group 1 1.0215 .492 -1.41 .160 Group 2 1.1206 .489 Colds Group 1 1.9524 .926 -1.12 .262 Group 2 2.1053 1.080 Migraine Group 1 1.9375 1.373 -1.65 .100 Group 2 2.2632 1.421 Ulcer Group 1 1.8006 1.320 -.31 .748 Group 2 2.8596 1.407 a . A one-tailed tftest was used for all subscales, except for Basic Personality, in which a two-tailed‘tftest was used. b . . Exploratory subscale, not scored for separation anXiety. * ** p < .05 p < .01 *** p < .005 ** p < .001 160 College Adjustment Self-report For the Freshmen Sample, a comparison was made between those who rated their adjustment to college as satisfactory and those who rated it as unsatisfactory. It was assumed that students who were experiencing separa- tion anxiety would be unhappy and more likely to indicate that they had not adjusted well in their first six weeks of college. This informal hypothesis was supported for the General Separation Anxiety and Self-acceptance subscales, where the unsatisfied group scored significantly higher. They also showed a difference on the Basic Personality Types subscale, which was designed to measure tendencies toward either passive-dependency or aggressive pseudo-independence when other subscale scores indicated the presence of separa- tion anxiety. The group reporting dissatisfaction scored higher on the passive-dependent end of the scale (see Table 6.27). Age A comparison was made on the basis of age, with no prediction made. Rather, an answer was sought regarding the intensity and duration of separation anxiety throughout life. A significant difference was found between age groups in the General Sample, using the aforementioned ANOVA. However, the computer program used could not analyze the age groups on more than four variables. Therefore the separa- tion anxiety subscales; Overprotection, General Separation 161 Table 6.27 Comparison of Mean Subscale Scores by Adjustment to College for Freshmen Sample Background questionnaire item 31 Group 1 = unsatisfactory adjustment to college (nf30) Group 2 very satisfactory adjustment to college (2f141) Subscale i S.D. tfvalue pa . . b BaSic Personality Group 1 1.9322 .462 3.99 .000**** Group 2 1.6308 .356 Overprotection Group 1 1.8208 .384 -.28 .784 Group 2 1.8449 .444 General Separation Anxiety Group 1 2.6342 .455 2.64 .010** Group 2 2.4112 .413 Self-acceptance Group 1 1.3861 .370 3.36 .002**** Group 2 1.1515 .342 Mother Hostility Group 1 2.0120 .409 -.63 .538 Group 2 2.0679 .446 Colds Group 1 1.5000 1.137 1.04 .298 Group 2 1.3121 .838 Migraine Group 1 2.1000 1.423 1.05 .296 Group 2 1.8156 1.334 Ulcer Group 1 2.0333 1.426 1.62 .108 Group 2 1.6241 1.216 b Father Group 1 1.8476 .584 -.05 .961 Group 2 1.8533 .571 aA two-tailed Eftest was used for all subscales. b . . Exploratory subscale, not scored for separation anxiety. * p < .05 ** p < .01 *** p < .005 **** p< .001 162 Anxiety, Self-acceptance, and Mother Hostility were chosen for analysis. The range of each age group is as follows: Group 1: 18-22 years Group 2: 23-30 years Group 3: 31—40 years Group 4: 41-50 years Group 5: 51-60 years Significant differences were found on three out of the four subscales. 0n the Overprotection subscale Group 1 scored significantly higher than Groups 4 and 5. 0n the General Separation Anxiety subscale, Groups 1 and 2 scored significantly higher than Groups 3, 4, and 5. 0n the Mother Hostility subscale, Group 1 scored significantly higher than Groups 2, 3, 4 and 5. All results related to age show a significantly higher rate of separation anxiety in the youngest age group when compared with the others (see Table 6.28). Religion The other ANOVA was calculated on a comparison of groups delineated on the basis of stated religious back— ground. These groups, in the Freshmen Sample, are: Group 1: Catholic Group 2: Protestant Group 3: Jewish Group 4: Other 163 H0O. v m moo. v m «*«« ««« .omm .v u Eocmmum mo momumma “muoz «...¢ooo. m¢v~.m ommm. m~m~.H mafiaflumo: umnuoz wows. «mom. amoa. ammo. mocmummoomumamm . . . . mumflxca ««««oooo wmho OH wood NhHm H cowumhmmmm Hmumcmu ..«mmoo. «bmm.m nmmm. mmma.a cofiuomuouaum>o m 04 mu: mmumsvm mmumsvm mm mom 5 .u. h Gmwz HOHHW Gmmz Gmmsfibm H n— m AH EmuH pcsoumxommv was so mHmEmm Hmuocmw How mUGMAHm> mo mammamcd mN.o wanme 164 No prediction was made on the basis of religion, this comparison being entirely exploratory in nature. How- ever, a difference between groups was found on one of the four subscales. 0n the Mother Hostility subscale, Group 3 scored significantly higher than Group 2 (see Table 6.29). Number of People Close to Emotionally At One Time The participants in the Freshmen Sample were asked about the number of close relationships that they usually had established at any one given time. The groups were delineated as follows: Group 1: Close to only one person at a time. Group 2: Close to a couple of people at a time. Group 3: Close to several different peOple at a time. Group 4: Close to many different people at a time. It was expected that persons in Group 1 would score signifi- cantly higher on the test than those in the other groups, reflecting a tendency toward exclusive, symbiotic relation— ships typically established by people who experience patho- logical separation anxiety. The results of the ANOVA showed that Group 1 scored significantly higher on the Self- acceptance subscale than those in Groups 3 and 4, but was not significantly different from Group 2. 0n the Mother Hostility subscale, Group 1 scored significantly higher than Groups 2, 3 and 4 (see Table 6.30). 165 mo.vd k. .hmm .m u Eocmmum mo mmmummo "muoz «memo. mmmm.~ mmam. «mam. suflauumom umnuoz Heme. moaq.~ mews. Gama. mocmuamoomumawm . . . . wumfixc¢ Name mass m mqma maom coflumumamm Hmumcmo mama. mmmm. mama. Home. cofluomuoudum>o m oaumuum mwumsvm moundvm mm mom 5m . £602 .HOHHW Emmi Gwmwfiwmm H D AN EwuH ccsonmxommv pcsoumxomm msoflmflamm no mHmEmm cmfinmmum How mUCMflum> mo mammamc¢ mN.w magma mo. v a t. .mmm .m u Eocmmum mo mwmummo "wuoz 166 «mafia. masm.m «van. mmmb. sugauumon umnuoz .mmeo. a~m5.~ HvNH. mavm. mocmuamoomuuamm . . . . sumuxca ommo «Goa N emma Hmmv :oflumumamm Hmumcmo mmmm. smmo.a mama. Hana. acuuomuoumum>o a oHumunm mmumswm mmumsvm mmamombsm . :mmz Houum cam: cmmzumm Amm EmuH pcsoumxommv mafia one am mmflnmcoflumHmm wmoHU mo Hmbadz co mHmEmm cmenmmum How moccaum> mo mflmwamcfi om.m manme 167 Summary Additional results showed significant differences in the predicted directions on some of the test subscale scores. Comparisons were made on the basis of school absence rate, frequency of contacts with mother, separation from mother as a child, whether living with parents or not, marital status, report of satisfaction with adjustment to college, and the number of relationships established at one time. Further results were reported on the basis of race (no differences), age (young people scoring higher on some subscales), and religion (Jews scoring higher on Mother Hostility than Protestants). CHAPTER 7 SUMMARY AND CONCLUSIONS In this chapter the thesis is summarized, the results are discussed, and conclusions are stated. The limitations of each aspect of this study, the implications for future research, and the implications of the findings for counseling practice are also reviewed. Summary The purpose of this study was to develop a valid theoretical concept of adult separation anxiety, drawn from the existing research that pertains mostly to children, and an attempt was made to measure this set of psychodynamics through construction of a personality test. It was postu- lated that significant results from such a test, in the pre- dicted direction, would give evidence for both the validity of the test itself and for the validity of the theoretical constructs on which it was built. An objective format was chosen for the test of separation anxiety and items were created on the basis of theoretical material, published case studies, and interviews conducted by the researcher. These items were checked for 168 169 readability, face validity, and clarity of meaning through pilot efforts and consultation with colleagues. Items were also worded in such a way as to minimize the effects of social desirability and other response sets. Following the piloting and critiquing of items, any that were obviously unworkable were either removed or reworded. A four-point true-false continuum was chosen as the response code to eliminate middle-of-the-road responses and the direction of true-false coding was varied in order to interrupt response set. When the items were put in their final order, they were also separated on the basis of content to minimize the effects of the effects of tendencies toward response carry-over. The final form of the separation anxiety test was given to two groups, along with a background questionnaire asking about family history, personal habits, physical prob- lems, and current attitudes relevant to separation anxiety. However, there were slight differences in the questionnaires, with adjustments made to suit the characteristics of each group. Also, 16 items were added to the test for the second administration. The first group given the test and questionnaire was a diverse, but not random, group of 400 adults who were approached in various settings and agreed to participate. This group was labeled the General Sample and was surveyed for the purpose of acquiring a sample that was large enough to provide sufficient data for a reliable cluster analysis. 170 The second sample was chosen because they were in the midst of a separation experience. Ninety-four percent of the first-term freshmen women living in Rather Hall, on the Michigan State University campus, were given the separation anxiety test and background questionnaire in an attempt to study the validity of the test. It was hypothesized that, of the 300 women participating, those who were the most homesick and anxious, as determined by self-report, would also score significantly higher on the test of separation anxiety. Because both samples were sufficiently large, cluster analyses were done on the sets of responses to test items for both groups. This was done in order to determine which items were related to one another and would therefore be retained and scored; whether or not subscales would be formed; and, if so, what their content would be. Results of these analyses produced six subscales, labeled Over- protection, General Separation Anxiety, Self-acceptance, Mother Hostility, Father, and Basic Personality Type. The last two of these subscales were not scored for separation anxiety because they were exploratory and not intended for that purpose. For example, the Basic Personality Type sub- scale was created to distinguish two basic personality types associated with separation anxiety rather than measuring the anxiety itself. 1 The cluster analysis program used in this study, PACKAGE, also reported the measure of internal consistency 171 for each subscale. These standard score alpha coefficients are as follows: Basic Personality Type .67 Overprotection .64 Father .78 General Separation Anxiety .82 Self-acceptance .62 Mother Hostility .64 Comparisons were then made between groups within each sample. Background questionnaire items were used to delineate groups on the basis of homesicknesses, childhood fear of school, physical complaints, and other character- istics said to be related to separation anxiety. In most instances two group means were compared and a t—test was thus utilized, with the probability of significance set at the p < .05 level. Where multiple group comparisons were made, an analysis of variance was done, with the probability also set at the p < .05 level. Summaries of these results, by hypotheses content, are stated in Tables 7.1 and 7.2. Conclusions and Discussion First, conclusions that were drawn regarding the test itself will be discussed. The cluster analysis resulted in eliminating approximately one-half of the items originally proposed. Therefore, it would be apprOpriate to examine the content and scoring results of the subscales that were formed (see Appendix G for reference). 172 mm H m h m Hmuoa novao can own m x x mappfia .m> uasvm maso> .MH uo: .m> umnuoe m x cums muomucoo ucmsvmuu .NH 90: .m> H x menace EOuu pmumummwm .HH you can .m> m x x :mumo Hoonom pmmmwe .oa v x x x vwfluume .m> mavcfim .m uoc m x x .m> wucmuma :ufl3 m>wa .m m x x mama .m> mameww .5 x pap .m> ocean mcw>wa m mo nova mxwa u.ccwo .0 ac: .m> HmcflumquH m x nouummm Mo >uoumufimmwu .m uoc .m> m x x umeEm can unmflm3um>o .v N x x uoc .m> umeEm .m uoc .m> ucofiw3 N x x . lumps: no unowmzum>o N m x x x x uo: .m> Hoesom no ummw .H mm>e >ufiafiumom wocmummoom xumaxcd :ofiuomu quHmCOmuwm onmm umguoz luawm :ofiumummwm Ioumum>o Hmuoe . . Hmumcwo masouu cofluwufluu manomndm >uoumuonxm mwamumndm >uwaxc< coflumummmm deEmm Hmuwcwo on» uom wmamombsm umwe co msouu COAHmuqu >3 muasmmm mo xumsfism H.h wanna 1f73 .cofluomuwp pmuufipwum Ga unmofiuficmflm uozm Om m Hmuoe duos no 03» .m> 08w» m an COmHGQ wco ou mmoHU mucmumwuoum .m> mzwn vmfiumauMm .m> ucoEumsflom wowa Ifloo cum: vmfimmaummcs uo: .m> >Hucosv noun Hoonom commas uo: .m> mumnEwE >Hfl56m o» wmoHo >HumOE uo: .m> kuuomwu >uwflxcm coflumummmm uo: .m> pmuuommu mmmcxowmoeon vac .m> ocean mcw>wa mo meow mxwa u.ccwo x uoc .m> somEOum puma: uo: .m> umeEm cam unoflw3uw>o uo: .m> uoxOEm uoc .m> unuflw3 uumpcs Ho usmflmzuw>o uoc .m> Hoocom Ou om Ou nflmumm .MH .NH .HH .oa Hmuoa mm>a >uwamc0mumm uwcumm oflmmm xuflawumom mocmummoom umcuoz nuamm >uwwxc< coflumummow Hmuwcmo sawuomu noum m u >0 mmaouo cofiumufluo moamomnsm >uoumuonxm mmamomndm wuwwxc< coflumummwm mHQEmm cmenmmum on» now mwamomnsm puma co msouu coHuwuHuU >3 muasmmm mo >umEEdm N.h mHDMB 174 1. Overprotection. The content of this subscale dealt with instances of parental hovering and reluctance to allow the individual to be independent as a child. This related to the general parent-child symbiosis seen in families with separation anxiety. If the child believes he is not capable of acting on his own, it is unlikely that he will ever stray very far from the parent on whom he depends for guidance. This subscale was significant in the predicted direction for eight hypotheses tested in the General Sample, but only for two in the Freshmen Sample (see Tables 7.1 and 7.2). This discrepancy may have resulted either because of scale unreliability or because of differences between the two sample groups; both of these factors must be weighed when considering the value of this subscale. The measure of reliability for this subscale, the standard score coefficient alpha, was among the lowest for the test at .62. However, it does appear to have differenti- ated well between groups in the General Sample. Regarding the second factor, sample differences, Tables 7.1 and 7.2 reveal that the overall results for both samples were very similar, in spite of the differences in the nature of the groups. The only major differences were the aforementioned ones seen in response to the Overprotection subscale, sug- gesting that while the scale may have some validity with its current item content, it should be revised to strengthen its effectiveness across groups. 175 2. General Separation Anxiety. Most of the items clustering together in this subscale dealt directly with the fear of losing the affection or presence of loved ones. Other items pertained to the extent of family closeness. Again, this reflected the theme of family symbiosis typical in separation anxiety. Perceptions of the family unit as an emotional "safe house," free of the demands and responsi- bilities of the outside world, are common. This, of course, is taught by the parents who bind the children to them by inducing a fear of,.andinability to deal effectively with the world outside the family. This subscale is the most direct measure of what is thought of as separation anxiety and also has the greatest internal consistency. When examining the content of the formal and informal hypotheses (Tables 7.1 and 7.2), it can be seen that this subscale consistently showed significant differences in the predicted direction for criterion groups most obviously reflecting components of separation anxiety, such as the self-reported homesick, separation anxiety, and fear of school groups. This subscale appears to be quite stable and valid as a measure of overt anxiety responses to the idea of separation. Future research on the instrument devised in this study should therefore retain this subscale in its present form, and attempt to gain further evidence of its validity. 176 3. Self-acceptance. The content of this subscale reflected a mixture of items related to self-esteem. Included were items pertaining to body image, trust, ego strength, and the related issue of willingness to attend school. This scale is theoretically tied to the concepts of body-image, trust, self-perception, and identity dis- cussed in Chapter 3. In general, it can be stated that the high and continuous levels of identification as a mode of relating seen in separation anxiety results in an incomplete and negative sense of self. The person with separation anxiety sees himself as an extension of others, is dependent on them, and subsequently reacts with confusion or disdain when assessing himself. While this subscale was significant almost as often as the others, its level of internal consistency was the lowest. Since it was frequently successful in discriminating between groups, it should be retained. However, both the reliability and validity could no doubt be improved by splitting the subscale into two separate subscales repre- senting the major content themes: self-image and trust. New items could be added to each of these to provide for adequate reliability and to strengthen each scale as a measure of a particular aspect of separation anxiety. 4. Mother Hostility. The content of this subscale dealt primarily with overt expressions of anger toward the mother figure. The exceptions to this were the two items related to school phobia. As mentioned in Chapter 3, a 177 component of school phobia is guilt based on hostility toward the mother. The child fears that his hostile wishes may magically destroy his mother while he is away at school. Thus, the presence of school phobia items is not a theor— etical contradiction. The original experience of hostility is said to be a natural outgrowth of inhibition and forced dependency, as well as a response to the mother's rejection. Because this hostile-dependent relationship between mother and child is theorized as the core of separation anxiety, this subscale was expected to achieve significant results. This expecta- tion was met, with the Mother Hostility subscale revealing significant differences with the greatest frequency. In one instance, however, this subscale showed sig- nificant differences between groups in a direction Opposite of that which was hypothesized. This occurred in the Fresh— men Sample, where subjects who stated that they liked the idea of living alone scored significantly higher on the Mother Hostility subscale than those who disliked the idea. There were two possible explanations for this: (a) the subscale was unreliable when used with this group, or (b) the responses to the background item did not clearly distinguish between separation anxiety and control groups. In light of the other results on this subscale, it seems likely that the latter explanation is most plausible. That is, the responses to the background item about living alone may not have separated subjects solely on the 178 basis of separation anxiety, but rather included some other variable. Since almost all of the subjects in this sample were late adolescents, they had not ever lived alone or even been away from their families very often, if at all. Therefore, it seems possible that the group desiring most to live alone included those persons who were struggling the hardest to be independent from their families, and possibly involved in the mother-daughter hostile inter- actions typically seen during this late adolescent period. (The results of the analysis of variance done on age (see Chapter 6) also confirm this explanation.) Thus, it seems that the higher score on Mother Hostility seen in the group who did like the idea of living alone may be a function of the background item and the age of the sample used, rather than a function of the reliability of the subscale itself. The Mother Hostility subscale is therefore considered to be a valid measure of an essential aspect of separation anxiety. It is recommended that it be retained for future research in its present form. 5. Father. All of these items were parallel to the "mother" items, and resulted in a cluster with the second highest reliability coefficient. Scoring in the low direction on this subscale portrays a close, somewhat sym- biotic relationship with the father. Scoring in the high direction portrays a hostile, estranged relationship with the father. The results for this subscale indicated that it was significant four out of five times in the direction 179 Of a close affectionate attachment. It was significant in the Opposite direction, i.e., a hostile attachment, in only one instance. This occurred for the "fear of school" group. While no direction was hypothesized for this subscale the frequency of significance in the close attachment direc- tion is consistent with the theory pertaining to separation anxiety, which states that the father is Often a nurturing second "mother" figure to whom the child turns when the actual mother is rejecting. The problem in this, Of course, is that the father also engages in a symbiotic-type relation- ship rather than facilitating individuation. The exception to this, scoring in the direction Of hostile attachment, seen in the "fear Of school" group, may represent the smaller percentage of fathers reported in school phobia cases, who are Openly hostile rather than repressing their hostility (see Role of the Father, Chapter 3). Thus, the results for this subscale were not inconsis- tent with the literature, and it appears to have potential as a valid indicator of separation anxiety. However, a problem is presented for the scoring, in that significant scores in either direction are possible, even though there is a marked tendency toward lower scores for criterion groups. This problem is left to be solved by future research efforts, which could perhaps be resolved by proper weighting with other subscale scores. 6. Basic Personality Types. This subscale was designed to determine the personality style of individuals 180 who have separation anxiety. In Chapter 3 the personality style of those with separation anxiety was described as essentially passive-dependent. However, Fairbairn (1954) noted that, usually in less severe cases, a personality style of aggressive pseudo-independence was observable. This personality style is said to represent a denial of passive-dependent wishes, and is seen in persons whose ego is strong enough to support such a defensive system. Items in this subscale were all related to-independence-dependence, gregariousness-inhibition, and admission of anxiety-denial. of anxiety. The purpose of the Basic Personality Type subscale, then, is to differentiate between these two personality styles. In comparing groups with moderate or little separa- tion anxiety there should be no significant differences on this subscale. But when a comparison is made between a criterion group said to be high on separation anxiety, such as school phobics, and those who are low on separation anxiety, the passive-dependent personality should be in evidence significantly more Often in the former group. The results that were Obtained in this study showed that more extreme separation anxiety groups did score higher than others in the passive-dependent direction. And moderate or minimal separation anxiety criterion groups scored almost identically when compared with each other, indicating that this subscale appears to be functioning as it was designed. 181 To accomplish the task of noting and reporting ten- dencies toward either end of this subscale, a scoring system that measures tendencies away from the middle is necessary. Future use Of the instrument should employ such a method. Finally, pertaining to the test content, there remains the issue of inclusion or exclusion of the psycho- somatic items. While these items were never significant in any but the hypothesized direction, they did not reveal significant differences very Often. This suggests that colds, ulcer, and migraine are probably related to separa- tion anxiety, but do not occur often enough to be valuable as a separate scale or scales. Instead, they should be included as regular test items and incorporated into present clusters, or reworded to represent the physical symptoms, but not necessarily the disorder itself. For example, an item dealing with the frequency Of headaches might prove more valuable than an item pertaining to migraine, and an item about gastrointestinal distress might prove to be a good replacement for the ulcer item. In summary, it seems that much Of the test has proven to have validity in this preliminary stage of deve1— Opment. The Basic Personality, General Separation Anxiety, and Mother Hostility subscales seem to be adequately measur- ing what they were designed to measure. The Overprotection and Self-acceptance subscales are in need of some revision, but are workable, in many instances, in their present form. Certain psychosomatic items were shown to be related to 182 separation anxiety and should be incorporated, along with other items, in future versions of the instrument. Hypotheses The summary Of hypotheses test results, by separa- tion anxiety subscale, can be seen in the aforementioned Tables 7.1 and 7.2. Both formal and informal hypotheses are included as reported in the results and additional results in Chapter 6. This is to aid the reader in the following discussion organized by hypothesis content. School phobia. Of the hypotheses formally stated for this study, those concerning fear of school (or school phobia) produced the greatest number of significantly high subscale scores. While this would appear to provide clear evidence for the validity of the test, since school phobia is the most generally acknowledged expression of separation anxiety, this is not quite the case. As stated previously in Chapter 4, an item-by-item significance test was done on the school phobia group responses in the General Sample, as part Of the process for final item selection. Therefore, significantly higher scores by persons in this group was partially guaranteed by the choice of items. However, this did not guarantee a similarly high score by the fear of school group in the Freshmen Sample, which did, in fact, score higher than the control group on all Of the major subscales except Overprotection. Thus the fear of school groups for both samples scored quite high on 183 the test, even though the items were only preselected in the case of the General Sample. These findings suggest that a valid test of school phobia has been produced, which may or may not include all forms of separation anxiety. This would be a logical con- clusion if it were not for the similar results achieved by tests conducted on other hypotheses that did not deal with childhood school reactions. For example, the results for "upset stomach" in the Freshmen Sample and "marital status" and "living with parents or not" in the General Sample showed similar significant differences in the expected direction. Therefore, the school phobia related results serve to confirm the validity of the test, just as the test results in other areas confirm school phobic reactions as a manifestation of separation anxiety. Orality. Hypotheses tests related to orality pro- duced fairly consistent significant results on some sub- scales. The significant differences between "Oral" and "non- oral" groups regularly showed the "oral" groups to be higher on the Self-acceptance subscale (indicating lower self- esteem) and frequently to be higher on the Mother Hostility subscales. Also, smoking produced significant differences on the test more Often than did the overweight or under- weight criteria. Significant differences On only two subscales indi- cate either that an oral orientation is common to persons who do not have separation anxiety as well as those who do; 184 or that smoking and weight problems do not adequately define an oral orientation (for example, talking was not included); or that such an orientation is only one aspect of separation anxiety and should not be expected to produce differences on all of the subscales. It may well be that all of these reasons have relevance concerning the outcome based on oral determinants. Therefore, it seems premature to draw any conclusions on this matter until further research is done with the separation anxiety instrument. Psychosomatics. Hypotheses tests pertaining to psychosomatic illness and symptoms produced significant results in the expected direction. The significant differ- ences on all major subscales for subjects in the Freshmen Sample reporting “upset stomach" gives strong support for theory linking gastrointestinal distress with separation anxiety, as well as helping to establish the validity of the test . Unfortunately, not enough subjects were found in either sample to test the "asthma theory" separately, but persons with respiratory problems were included in the psychosomatic significance test done on the General Sample, which produced predicted results. Desire to live alone. Tests concerning the subjects' responses to the idea of living alone produced mixed results. Since persons in the General Sample were to answer this item only if they actually had lived alone, the results for this group seemed to hold promise, in spite of the reliance on 185 attitude self-report. But numerous persons answered this item even though they had not lived alone, thereby contami- nating the sample. In the Freshmen Sample, as stated previously, atti— tudes on this subject seemed influenced by the fact that none of the research participants had ever lived alone and many were in the throes of their late-adolescent rebellion against parents and the idea of living with them. The typical reaction to this situation is to state a desire to live alone, whether or not the person would actually be comfortable in this situation. Because Of these criterion problems there was no consistency between groups in each sample, as there was with the other hypotheses, and there were only minimal differ- ences found between criterion and control groups. In retro- spect, it appears that this question is not a reliable dis- criminator and should not be utilized in future studies. Sex differences. The hypothesis test done on the basis of sex was only for the General Sample, since the Freshmen Sample was all female. Although this was collected in a manner that did not insure the comparability of the male and female groups, the large number of subjects used reduces the error associated with unequal and nonrandomly selected groups. Therefore, in light Of the preliminary validity Of the test shown by other results, it seems likely that women do experience more separation anxiety than men. 186 Homesickness, Separation Anxiety, and Number of Close Relationships. Results on all three of these hypothesis tests showed significant differences on the General Separa- tion Anxiety subscale for each, and also on the Self- acceptance subscale for the test based on self-report of separation anxiety. These limited differences, found almost exclusively on the General Separation Anxiety subscale, suggests that either the background questionnaire items did not adequately discriminate between groups on the basis of separation anxiety, or the majority of the test subscales were not effective. In View of the fact that the criteria used in this study that did not rely on attitude self-report, such as school absence, produced significantly different groups; it seems likely that the less dramatic differences seen in the homesick and other groups is a function of social desira- bility effects on self-report. Also, the term "homesick" was not defined and group membership was thus determined by the relative assessment of each individual. Discussion of Additional Results Additional results showed that marital status was the single best predictor of separation anxiety, as measured by the instrument devised for this study. That is, subjects over the age of 23 and single scored significantly higher on every subscale than subjects of the same age who were married. However this may have been affected by other mental 187 health variables related to being single or married. The single group may be an extreme one that would be likely to score significantly higher on any test of psychOpathology, regardless of its relationship to separation anxiety. Further investigations regarding the importance of separa— tion anxiety as it pertains to marital status are therefore necessary. (For mention of theoretical relevance the reader is referred to the "marital status" section in Chapter 6.) Other comparisons made that were also related to independent adult functioning were based on whether or not subjects lived with their parents, and the frequency of contacts made with the mother. These revealed that persons living with their parents scored significantly higher on all the major subscales, except Self—acceptance, than those who did not. And persons who frequently contacted their mothers scored significantly higher on the Overprotection and Gene- ral Separation Anxiety subscales than those who did not make such contacts. Additional results based on school attendance and report of satisfaction with college adjustment showed similar differences in the predicted direction. Therefore, the criteria most related to the actual experience of separation anxiety, such as school attendance and marital status, did produce significant differences on the test. This established the existence of separation anxiety in adults and the potential for the test as a valid indicator of this psychological phenomenon. 188 Criterion not related, such as age, religion, and race showed little or no differences, as expected. The differences that appeared regarding age are not surprising, especially when viewed in the context of separation as a life-long process. Young adults, currently attempting to become truly independent for the first time, are predictably more anxious about separation than their elders. NO differences at all were expected on the basis of race or religion. Therefore, the one difference found per- taining to Jews and Protestants on Mother Hostility was not anticipated. It would be interesting to make such compari- sons in the future to see if these results would be repeated. However, overall, the differences found in religious back- ground were extremely minimal. Limitations The limitations of this study concern generalization of findings, test item selection, and delineation Of cri- terion groups. Generalization of Findings The sampling procedure used in selecting the Fresh- men Sample virtually had no flaws, since the entire pOpula- tion Of a given dormitory was used. Therefore, the general- ization of findings from this sample can legitimately be made to any comparable female group of freshmen attending a large Midwestern public university and living on campus. For more accurate generalization Of results the reader is 189 referred to the table of descriptive data on the Freshmen Sample reported in Chapter 5, Table 5.1. For the General Sample the procedure was not random or complete and the group is therefore not truly represen- tative of the pOpulation in general. While the sampling procedure did not rely on volunteers, since all subjects were requested to participate, there is no way of knowing exactly how representative this group is. Therefore, the reader is advised to use caution in generalizing the results beyond the sample used in this study. Again, the use of the demographic data on this sample, reported in Table 5.2, may be helpful in the consideration of wider application of these findings. Item Selection The cluster analysis and selection of items on one sample with hypothesis testing on another, totally inde- pendent, sample would be ideal. This was not done in this study because the second possibility for a cluster analysis with a large, homogeneous group presented by the Freshmen Sample, seemed too essential to the test develOpment to be ignored. Furthermore, items were only rejected on the basis of their failure to cluster and not on any other criterion for this sample. However, as previously stated, significant school phobia items from the General Sample were used as a cri- terion for final item selection, which served to bias the 190 school phobia and school attendance hypotheses for that group. But other hypotheses tests for this sample were not considered to be affected by this bias. Delineation of Criterion Groups The Optimal method for selecting groups of persons who appear to have separation anxiety is through behavioral observation, which was not utilized in this study. Rather, criterion groups were determined by their responses to back- ground questionnaire items. Originally, this study had been planned in a way that would have used behavioral observa- tions, but this involved judgments by residence hall assis- tants regarding the symptomatic behavior of freshmen living in the dormitory. This was determined to be a violation of the individuals' rights to privacy and was therefore aban- doned in favor of the questionnaire format. This limited the study in that self-report regarding attitudes is usually unreliable to some extent, depending on what is being asked. Generally, however, these limitations are minimal and do not substantially effect the outcome of the study. The most important limitation deals with the preselection of school phobia items for the General Sample, and strongly biases the hypothesis test done on that criterion for that group. Implications for Future Research Issues in adult separation anxiety could be explored and documented through the use of the separation anxiety 191 instrument as either the major or ancillary method of assess- ment. Because the nature and number of possible studies relating to the test and/or separation anxiety are quite numerous, only some of the more important avenues of investi- gation related to this study are listed. 1. Studies that involve the comparison of test scores for criterion and control groups using different scoring approaches, to determine the most accurate method of assessing item responses. 2. Further validation efforts concentrating on continued testing with separation anxiety criterion groups, such as mothers and fathers of school-phobic children. 3. Investigations of the relationship between the separation anxiety test and related psychological measures, such as the inhibited hostility subscale of the MMPI. 4. The gathering of normative data on the test through administration to various target groups. 5. Studies devoted to defining and testing a particular aspect of theory, such as orality, psychosomatic illness, or passivendependent personality. Possible cri- terion groups are alcoholics, compulsive talkers, compulsive eaters, persons with chronic ulcers or colitis, or adults living with their parents or siblings, who could be given the separation anxiety instrument and compared to an appro- priate control group. 6. Relationship between depression and separation anxiety. Guntrip (1976) speaks of "depressive dependency" 192 that results when a good mother is lacking. Is this depres- sion chronic when there has been no good mother? 7. Studies of racial differences based on family structure and social roles. For example, would Mexican- Americans score differently, sharing a different pattern from what was found for the predominantly white sample used here? Or would blacks score differently when socio- economic subgroups were compared, e.g., blue-collar vs. white-collar professionals? Implications for Counseling Practice The results of this study strongly suggest that separation anxiety does continue into adulthood and may be a major factor in any moderate to severe psychOpathology. This also suggests that the process of attaching and indivi- duating is a crucial aspect of the therapeutic relationship, for both the therapist and the client. This obviously has ramifications for the way in which the therapeutic relation- ship is established, continued, and terminated. Especially for those who experience a great deal Of separation anxiety, the wisdom of the currently pOpular short-term therapy is brought into question. The commonly allotted 10 to 12 sessions funded by government agencies for the poor are undoubtedly insufficient for establishing a therapeutic relationship and resolving conflicts. Indeed, if any degree of attachment does occur during this brief period, the act of termination would serve to exacerbate 193 the client's existing separation anxiety. Particularly for those who suffered early childhood psychic trauma, longer term psychotherapy is indicated. The issue of client dependency is also closely related to separation anxiety. Therapists who refuse to assume any responsibility for their clients at any time may fail to facilitate the necessary transference relationship, while those who indefinitely indulge their dependent clients' demands repeat the destructive parental overprotection that originally helped to create separation anxiety. Ultimately, regardless of therapeutic approach, individuation attained by the client in therapy is linked to the therapist's own sense of separateness. The counselor or therapist must have achieved an adequate degree of separation in order to deal effectively with the demands, expectations, and criticisms Often voiced by clients with separation anxiety who interpret all expres- sions of therapist separateness as rejection. The client's attempt to induce responsibility and guilt for his or her own feelings of rejection and worthlessness is common, and requires the therapist to have firm psychological boundaries in order to avert the acceptance of responsibility, or the defensive response of anger. Naturally, termination of therapy is an extremely important and critical event and must be handled in such a way that it is seen by the client as a facilitation of individuation rather than a separation based on rejection. 194 Also, relative to termination, the client practice of ending therapy before conflicts have been resolved, commonly referred to as a "flight into health," may actually repre- sent the client's growing attachment to the therapist, and subsequent fear of becoming too dependent and being engulfed by the therapist. As a defense against this fear, the client denies that there is anything wrong, or that there is any need to continue therapy, thereby providing an escape from the threat of attachment. Finally, awareness of separation anxiety and its concomitants are essential for counselors and therapists. In addition to the therapy-related issues already mentioned, the role of separation anxiety and the indicative behaviors in neuroses and psychoses should be part of the therapist's consciousness. For example, awareness of the role of separation anxiety in Oedipal conflicts; exaggerated dependency and related problems such as alcoholism, Obesity, and drug addiction; identity conflicts resulting from a symbiotic mother-child relationship, sexual problems, certain phobias, and numerous other presenting problems would obviously contribute to a more complete understanding of the client and a more effective therapeutic treatment. Therefore, it can be seen that therapist awareness of separation anxiety as it pertains to both moderate and severe problems is desirable as part of a network of thera— peutic insights and skills. What is even more important, however, is that the therapist has attained a relatively 195 high degree of individuation, so that he or she is capable of facilitating this with clients. Conclusion This study has highlighted the ubiquitous character of separation anxiety, demonstrating that the nature Of the first major attachment has repercussions for interpersonal relations throughout life. It seems that the impact of these early experiences has been ignored in research and theory, largely due to the influence of the Freudian psycho- analytic school, which stresses later developmental periods as the most crucial ones. This study sought to fill some of the void pertain- ing to theoretical formulations and research practices con— cerning separation anxiety in adults. Since systematic investigations in this area have been impeded by the lack of an apprOpriate assessment device, the task of creating and validating such an instrument was undertaken. Results indicated that this test of separation anxiety may well serve the purpose of defining this psychodynamic framework in adults, thus providing a valuable research tool. Future use of the instrument could result in greater insights into normal and pathological functioning, and provide part of the missing link between childhood and adult psychological development. APPENDICES APPENDIX A SEPARATION ANXIETY TEST, FORM A APPENDIX A SEPARATION ANXIETY TEST, FORM A Please respond to the following statements as quickly and as accurately as you can. Please do not omit any items. Read each statement, then select one of the five responses listed below, according to whether you think the statement is completely false, mostly false, mostly true, or completely true. If you were raised primarily by someone other than your mother, for example, by a stepmother, grandmother, older sister, or aunt, please answer the "mother" items with that person in mind. Completely Mostly Mostly Completely False False True True 1. In most situations there is no clear right or wrong. 2. When I was growing up I didn't know if my parents had sexual relations or not. 3. I am not a jealous person. 4. When I was a child I felt nervous about returning to school after a vacation, even when I really wanted to go. 5. I am not very concerned with whether peOple like me or not. 6. If someone I care about very much is upset or hurt, I seem to feel their pain almost as if it were my own. 7. When I was a child I didn't have to check everything with my parents before I did it. 8. My brother(s) or sister(s) and I fought quite Often when we were kids. 9. I have had Older women friends or relatives who were like mothers to me. 196 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 197 When I was a child I sucked my thumb. When someone I like doesn't like me or want to be around me I try to figure out why. I can't bear the thought.of failing at anything. I think I was tied to my mother's apron strings. My parents were so busy working and raising a family that they didn't go out much on weekends. My mother or father is my best friend. I usually know when to quit. I am neither too tall nor too short. I feel guilty when I leave someone. I don't hide very much about myself from others. No one in my family has asthma. (Answer either 1 or 4) I am not sure of what I want to do with my life. I hate to make mistakes, because I am so hard on myself when I do. I don't feel guilty as often as most pe0ple do. I don't think I should take care of other people. I do not catch colds easily. I have often felt that my mother needed me as much as I needed her. I'm slow to let peOple know how much I like them because I'm afraid I'll come on too strong. I am not a stubborn person. I don't mind taking orders from my superiors. My mother listens and doesn't get upset when I tell her my problems. As a child, I never felt uneasy about going to the movies. PeOple think I am just like my mother or father. 198 33. The thought of committing suicide over failing at some- thing does not occur to me. 34. From the very first day of kindergarten I enjoyed school and couldn't wait to go. 35. I usually express my anger or irritation when I feel it. 36. I feel guilty when I think negative things about my mother. 37. I am a capable person. 38. My parents encouraged me to bring along my brother or sister when I went out with my friends. (If only child, leave blank) 39. I get frightened when one of my parents gets sick. 40. Most peOple will not let you down. 41. I like the way my face and body look. 42. I always felt that my parents would be very unhappy if I moved away from home when I grew up. 43. I have never been shy. 44. If I had to choose, I would rather be taken care of than take care of others. 45. Some people think I am aggressive. 46. I never really give up, even when I should. 47. I like less physical affection than most peOple. 48. My mother was less critical of me than my father. 49. I eat when I feel upset. 50. As a child I was told more about sex than most children my age. 51. I think I am lovable. 52. My parents think it is awful if I get into a fight with my brother(s) or sister(s). (If only child, leave blank) 53. I used to regularly have big fights with my mother. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 199 It doesn't really bother me when peOple refuse to admit they are wrong. I enjoy spending time in the house alone. I feel scared when someone I am close to goes away. I feel foolish when I cry. As a child I was not absent from school very Often. I am a talkative person. I have never had trouble with my weight. I enjoy sex more than most peOple. I get attached to peOple easily. My mother likes to tell me what to do. When I am involved in a romantic relationship, I do not usually think that my partner will cheat on me. I expect much more of peOple in authority than of people in general. ' I am a self-conscious person. Most peOple do not take advantage of others. When my parents raised me, they hardly ever yelled at me. I can take care of myself in any situation. My expectations for myself are tremendous. You should take care of yourself more than other peOple. When I was a child and got into arguments and fights, my parents didn't get involved unless someone might get physically hurt. I am afraid of my mother. I want more affection from peOple than I have a right to expect. I get jealous when I see a close friend of mine having a really good time with someone else. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 200 I am a calm and relaxed person. It is possible to feel both love and hate for the same person. I threw tantrums when I was younger if I didn't get what I wanted. I have sometimes been homesick. As a teenager I picked my own clothes. I am not very critical of myself. Eating usually makes me feel better. I am never positive that I have done a good job until someone else tells me that I have. I get very frightened when I think that someday my parents will die. I have never smoked a pipe, cigarette, or cigar regularly. I have an active social life. I enjoy it when my friends and family are physically affectionate towards me. I am very close to my mother. When I was younger I carried around a teddy bear, blanket, or favorite toy. I like my looks just the way they are. Caring for others makes me feel good. I am a quiet person. Someone in my family gets migraine headaches. (Answer 1 or 4) I am often afraid that someone I care about very much will find another person who they will like better than me. My parents encouraged me to make my own decisions. I was a late bloomer. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 201 Ending a relationship is not particularly difficult for me. I have said terrible things to my mother. My mother is not overprotective towards me. Sometimes I think I need more love than anyone could possibly give me. I used to wish I had a different mother. I used to get an upset stomach before I left for school in the morning. When I was in elementary school one of my parents would want to take me to school. I like the idea of living alone. I like my friends to be like me. When someone rejects me I get very angry inside. It doesn't bother me if someone close to me disagrees with me on an important matter. If I ever let out all the anger in me I might really damage someone. As a child in school, it was very important to me to please the teacher and have him/her like me. I do not hide the little kid in me. I can take criticism without feeling hurt or upset. I love to sing or whistle. As a child I spent a lot Of time with my parents. PeOple like to come to me with their problems. I am neither too fat nor too thin. NO one will ever love me as much as members of my family do. I am an independent person. Someone in my family has ulcers. (Answer 1 or 4) 119. 120. 121. 122. 123. 124. 125. 126. 127. 202 When I was very young I was afraid of strangers. I rarely let people get close enough emotionally to hurt me. I don't feel responsible for my mother's happiness. When I was growing up, my mother didn't mind if I was not home for dinner. My family is not exceptionally close. My mother's children are her life. During some period in my life I was convinced I would die young. I sometimes go on eating binges. As a youngster, I did not take trips away from home by myself very Often, such as going away to camp. APPENDIX B BACKGROUND QUESTIONNAIRE FOR GENERAL SAMPLE APPENDIX B BACKGROUND QUESTIONNAIRE FOR GENERAL SAMPLE 1. My age is . 2. My sex is male female. 3. My race is . 4. I am married divorced single separated. 5. My educational level is less than high school high school graduate some college work college degree graduate degree(s) 6. I have children. yes no 7. My occupation is . 8. In my opinion, I am above average average somewhat below average in intelligence. 9. I was an only child. yes no. If no, indicate how many: brothers older sisters older younger younger same age same age 10. I have lived alone. yes no. If yes, for how long? At what age(s)? I liked living alone. not very much at all Okay very much 11. I am currently more than 15 lbs. overweight. yes no 12. I am currently more than 5 lbs. underweight. yes no 203 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 204 I am in perfect health. yes no If no, what health problems do you have (if you are willing to say)? I am currently a smoker. Yes no My parents currently live in the same house as I do. in the same town as I do. (If your parents are in the same state as I do. deceased, please indi— in the same regional area cate their location of the country as I do. to you when they were in another part of the living.) _ country. overseas. I call or visit my mother (if living) every day at least twice a week about once a week once every two weeks about once a month every couple of months a few times a year never As a child, I missed school less Often than most children. an average amount. more Often than most children. As a child I was hospitalized or sick at home for a period of time. yes no If yes, how many times? Averaging how many days? With what illness or illnesses? I was afraid to go to school as a child hardly ever sometimes often My father died when I was a child. yes no If yes, how old were you? My mother died when I was a child. Yes no If yes, how Old were you? My parents were divorced when I was a child. yes no If yes, how Old were you? 23. 24. 25. 205 If you lost one of your parents through death or divorce as a child, was that parental role taken over by a step—parent? es no by a relative or friend of the family? yes no During my childhood I was separated from my mother for a long period Of time. yes no don't know I was adOpted. Yes no APPENDIX C INSTRUCTIONS TO RESEARCH ASSISTANTS APPENDIX C INSTRUCTIONS TO RESEARCH ASSISTANTS Ask persons if they would be willing to participate in a research project for a Ph.D. thesis. Explain that approximately 20 minutes of their time is needed to fill out a personity test and background questionnaire and that you do not want their names or ask anything that would identify them. If they say no, do not attempt to persuade them to participate. If they ask what the test is trying to measure tell them that you cannot say until after they have taken it. Afterwards, you may tell them that "separation anxiety" was being mea— sured and explain it as best you can. Tell participants to answer the test items with one Of the four alternatives at the tOp of the page and record the number on the computer sheet in pencil. (Also remind them that the sheet is numbered across rather than down.) Ask them to fill out the questionnaire only after they have completed the test so that they are not cued to the trait being measured. If you leave materials with peOple and plan to collect them later: (a) make sure you designate a day and time for collection. If this is not done, peOple tend to forget about doing it. If it is done, they are usually very reliable. (b) make sure you tell them not to discuss the test items with a spouse or anyone else while answer- ing. Their response must be independent. Keep approximate track Of the number of peOple who turn you down and their reason(s) for refusing. (This number should be small.) Try and get a total sample that is about 3/4 female and 1/4 male. Also, get as many blacks as you conveniently can. 206 207 Make sure you hand out questionnaire-answer sheet sets that bear identical numbers. They have already been sorted in this fashion. But it doesn't hurt to check when you hand them out. APPENDIX D NEW TEST ITEMS ADDED TO ORIGINAL 127 ITEMS APPENDIX D NEW TEST ITEMS ADDED TO ORIGINAL 127 ITEMS I have Often felt that my father needed me as much as I needed him. I used to wish I had a different father. I am afraid of my father. I am very close to my father. When peOple I am close to go away on a trip, I worry that something might happen to them. My father likes to tell me what to do. My father's children are his life. My father is not overprotective toward me. My father listens and doesn't get upset when I tell him my problems. When I was growing up, my father didn't mind if I wasn't home for dinner. I feel guilty when I think negative things about my father. I used to regularly have big fights with my father. I have had older men friends or relatives who were like fathers to me. I have not dated as often as I would have liked to. I have said terrible things to my father. My mother has said that my father's relatives sometimes seem to mean more to him than we do. 208 APPENDIX E SEPARATION ANXIETY TEST, FORM B APPENDIX E SEPARATION ANXIETY TEST, FORM B Please respond to the following statements as quickly and as accurately as you can. Please do not omit any items. Read each statement, then select one of the four responses listed below, according to whether you think the statement is com— pletely false, mostly false, mostly true, or completely true. If you were raised primarily by someone other than your mother, for example, by a stepmother, grandmother, Older sister, or aunt, please answer the "mother" items with that person in mind. Do the same with the "father" items. Completely Mostly Mostly Completely True True False False 1 2 3 4 1. In most situations there is no clear right or wrong. 2. When I was growing up I didn't know if my parents had sexual relations or not. 3. I am not a jealous person. 4. When I was a child I felt nervous about returning to school after a vacation, even when I really wanted to go. 5. I am not very concerned with whether people like me or not. 6. If someone I care about very much is upset or hurt, I seem to feel their pain almost as if it were my own. 7. When I was a child I didn't have to check everything with my parents before I did it. 8. I have Often felt that my father needed me as much as I needed him. 9. My brother(s) or sister(s) and I fought quite often when we were kids. 209 '210 Completely Mostly Mostly Completely True True False False l 2 3 4 10. I have had older women friends or relatives who were like mothers to me. 11. When I was a child I sucked my thumb. 12. When someone I like doesn't like me or want to be around me I try and figure out why. 13. I used to wish I had a different father. 14. I can't bear the thought of failing at anything. 15. I think I was tied to my mother's apron strings. 16. My parents were so busy working and raising a family that they didn't go out much on weekends. 17. My mother or father is my best friend. 18. I usually know when to quit. 19. I am neither too tall nor too short. 20. I feel guilty when I leave someone. 21. I don't hide very much about myself from others. 22. Neither I nor anyone in my family has asthma. (Answer either 1 or 4) 23. I am not sure of what I want to do with my life. 24. I hate to make mistakes, because I am so hard on myself when I do. 25. I don't feel guilty as often as most peOple do. 26. I don't think I should take care of other peOple. 27. I do not catch colds easily. 28. have Often felt that my mother needed me as much as needed her. HH 29. I'm slow to let peOple know how much I like them because I'm afraid I'll come on too strong. 30. I am not a stubborn person. 211 Completely Mostly Mostly Completely True True False False l 2 3 4 31. I don't mind taking orders from my superiors. 32. I am afraid of my father. 33. My mother listens and doesn't get upset when I tell her my problems. 34. As a child, I never felt uneasy about going to the movies. 35. PeOple think I am just like my mother or my father. 36. The thought of committing suicide over failing at some- thing does not occur to me. 37. From the very first day of kindergarten I enjoyed school and couldn't wait to go. 38. I usually express my anger or irritation when I feel it. 39. I am a capable person. 40. My parents encouraged me to bring along my brother or sister when I went out with my friends. (If only child, leave blank) 41. I get frightened when one of my parents gets sick. 42. I am very close to my father. 43. Most peOple will not let you down. 44. I like the way my face and body look. 45. I always felt that my parents would be very unhappy if I moved away from home when I grew up. 46. I have never been shy. 47. If I had to choose, I would rather be taken care of than take care of others. 48. Some people think I am aggressive. 49. When peOple I am close to go away on a trip, I worry that something might happen to them. 212 Completely Mostly Mostly Completely True True False False l 2 3 4 50. My father likes to tell me what to do. 51. I never really give up, even when I should. 52. I like less physical affection than most peOple. 53. My mother was less critical of me than my father. 54. I eat when I feel upset. 55. As a child I was told more about sex than most children my age. 56. I think I am lovable. 57. My parents think it is awful if I get into a fight with my brother(s) or sister(s). (If only child, leave blank) 58. I used to regularly have big fights with my mother. 59. My father's children are his life. 60. It doesn't really bother me when peOple refuse to admit they are wrong. 61. I enjoy spending time in the house alone. 62. I feel scared when someone I am close to goes away. 63. I feel foolish when I cry. 64. As a child I was not absent from school very often. 65. My father is not overprotective towards me. 66. I am a talkative person. 67. I have never had trouble with my weight. 68. I enjoy sex more than most people. 69. I get attached to peOple easily. 70. My mother likes to tell me what to do. 71. When I am involved in a romantic relationship, I do not usually think that my partner will cheat on me. 213 Completely Mostly Mostly Completely True True False False 1 2 3 4 72. My father listens and doesn't get upset when I tell him my problems. 73. I feel guilty when I think negative things about my mother. 74. I expect much more of peOple in authority than of peOple in general. 75. I am a self-conscious person. 76. Most people do not take advantage of others. 77. When my parents raised me, they hardly ever yelled at me. 78. I can take care of myself in any situation. 79. My expectations for myself are tremendous. 80. When I was growing up, my father didn't mind if I wasn't home for dinner. 81. You should take care of yourself more than other people. 82. When I was a child and got into arguments and fights, my parents didn't get involved unless someone might get physically hurt. 83. I am afraid of my mother. 84. I want more affection from peOple than I have a right to expect. 85. I get jealous when I see a close friend of mine having a really good time with someone else. 86. I am a calm and relaxed person. 87. I feel guilty when I think negative things about my father. 88. It is possible to feel both love and hate for the same person. 89. I threw tantrums when I was younger if I didn't get what I wanted. 214 Completely Mostly Mostly Completely True True False False l 2 3 4 90. I have sometimes been homesick. 91. As a teenager I picked my own clothes. 92. I am not very critical Of myself. 93. Eating usually makes me feel better. 94. I am never positive that I have done a good job until someone else tells me that I have. 95. I get very frightened when I think that someday my parents will die. 96. I used to regularly have big fights with my father. 97. I have never smoked a pipe, cigarettes, or cigar regularly. 98. I have an active social life. 99. I enjoy it when my friends and family are physically affectionate towards me. 100. I am very close to my mother. 101. When I was younger I carried around a teddy bear, blanket, or favorite toy. 102. I like my looks just the way they are. 103. Caring for others makes me feel good. 104. I am a quiet person. 105. I, or someone in my family, get migraine headaches. (Answer 1 or 4) 106. I am Often afraid that someone I care about very much will find another person who they will like better than me. 107. My parents encouraged me to make my own decisions. 108. I was a late bloomer. 109. Ending a relationship is not particularly difficult for me. 215 Completely Mostly Mostly Completely True True False False l 2 3 4 110. I have had Older men friends or relatives who were like fathers to me. 111. I have said terrible things to my mother. 112. My mother is not overprotective towards me. 113. Sometimes I think I need more love than anyone could possibly give me. 114. I used to wish I had a different mother. 115. I used to get an upset stomach before I left for school in the morning. 116. I have not dated as Often as I would have liked to. 117. When I was in elementary school one of my parents would want to take me to school. 118. I like the idea of living alone. 119. I have said terrible things to my father. 120. I like my friends to be like me. 121. When someone rejects me I get very angry inside. 122. It doesn't bother me if someone close to me disagrees with me on an important matter. 123. If I ever let out all the anger in me I might really damage someone. 124. My mother has said that my father's relatives sometimes seem to mean more to him than we do. 125. As a child in school, it was very important to me to please the teacher and have him/her like me. 126. I do not hide the little kid in me. 127. I can take criticism without feeling hurt or upset. 128. I love to sing or whistle. 129. As a child I spent a lot of time with my parents. 216 Completely Mostly Mostly Completely True True False False l 2 3 4 130. PeOple like to come to me with their problems. 131. I am neither too fat nor too thin. 132. No one will ever love me as much as members of my family do. 133. I am an independent person. 134. I, or someone in my family, have ulcers. (Answer either 1 or 4) 135. When I was very young I was afraid of strangers. 136. I rarely let peOple get close enough emotionally to hurt me. 137. I don't feel responsible for my mother's happiness. 138. When I was growing up, my mother didn't mind if I was not home for dinner. 139. My family is not exceptionally close. 140. My mother's children are her life. 141. During some period in my life I was convinced I would die young. 142. I sometimes go on eating binges. 143. As a youngster, I did not take trips away from home by myself very Often, such as going away to camp. APPENDIX F SIGNIFICANT TEST ITEMS FOR SCHOOL PHOBIA (GENERAL SAMPLE) APPENDIX F SIGNIFICANT TEST ITEMS FOR SCHOOL PHOBIA (GENERAL SAMPLE) School phobic group, N = 43 Not school phobic group, N = 356 (Item numbers refer to Form B) 4. 15. 20. *36. *37. *39. *44. *46. 62. *64. 70. 73. 83. When I was a child I felt nervous about returning to school after a vacation, even when I really wanted to go. I think I was tied to my mother's apron strings. I feel guilty when I leave someone. The thought of committing suicide over failing at something does not occur to me. From the very first day of kindergarten I enjoyed school and couldn't wait to go. I am a capable person. I like the way my face and body look. I have never been shy. I feel scared when someone I am close to goes away. As a child I was not absent from school very often. My mother likes to tell me what to do. I feel guilty when I think negative things about my mother. I am afraid of my mother. 217 85. *86. 88. 89. 94. 95. 101. 106. *112. 115. 117. *122. 123. *131. 135. *139. 141. 143. 218 I get jealous when I see a close friend of mine having a really good time with someone else. I am a calm and relaxed person. It is possible to feel both love and hate for the same person. I threw tantrums when I was younger if I didn't get what I wanted. I am never positive that I have done a good job until someone else tells me that I have. I get very frightened when I think that someday my parents will die. When I was younger I carried around a teddy bear, blanket, or favorite toy. I am often afraid that someone I care about very much will find another person who they will like better than me. My mother is not overprotective towards me. I used to get an upset stomach before I left for school in the morning. When I was in elementary school one of my parents would want to take me to school. It doesn't bother me if someone close to me dis- agrees with me on an important matter. If I ever let out all the anger in me I might really damage someone. I am neither too fat nor too thin. When I was very young I was afraid of strangers. My family is not exceptionally close. During some period in my life I was convinced I would die young. As a youngster, I did not take trips away from home by myself very often, such as going away to camp. *Item reverse scored. Total number of items: 31 APPENDIX G SUBSCALES AND CORRELATION MATRICES APPENDIX G SUBSCALES AND CORRELATION MATRICES 1. Basic Personality Types Subscale (Passive-dependent/ Aggressive-independent) Includes items: 23. I am not sure of what I want to do with my life. 48. I have never been shy.* 66. I am a talkative person.* 78. I can take care of myself in any situation.* 98. I have an active social life.* 99. I enjoy it when my friends and family are physically affectionate towards me.* 104. I am a quiet person. 130. PeOple like to come to me with their problems.* 135. When I was very young I was afraid of strangers. 136. I rarely let peOple get close enough emotionally to hurt me. Correlation matrix: standard score coefficient alpha = .67 48 66 78 98 99 104 130 135 136 23 10 15 ll 11 7 13 0 9 8 48 41 22 17 6 34 9 23 10 66 10 32 10 58 11 24 29 78 16 3 13 17 17 5 98 30 30 21 19 25 99 8 6 6 20 104 7 33 21 130 5 7 135 18 *Item is reflected (reverse scored). 219 220 2. Ovepprotection Subscale Includes items: 7. When I was a child I didn't have to check every- thing with my parents before I did it.* 15. I think I was tied to my mother's apron strings. 50. My father likes to tell me what to do. 65. My father is not overprotective towards me.* 80. When I was growing up, my father didn't mind if I wasn't home for dinner.* 83. I am afraid of my mother. 107. My parents encouraged me to make my own decisions.* 112. My mother is not overprotective towards me.* Correlation matrix: standard score coefficient alpha = .64 15 50 65 80 83 107 112 7 26 22 18 34 14 26 2O 15 8 1 8 l7 19 18 50 29 24 4 22 13 65 27 17 21 24 80 2 14 9 83 27 26 107 24 3. Father Subscale Includes items: 8. 13. 32. 42. 59. 72. 96. I have Often felt that my father needed me as much as I needed him.* I used to wish I had a different father. I am afraid of my father. I am very close to my father.* My father's children are his life.* My father listens and doesn't get upset when I tell him my problems.* I used to regularly have big fights with my father. *Item is reflected (reverse-scored). 221 Correlation matrix: standard score coefficient alpha = .78 13 32 42 59 72 96 8 36 21 61 42 43 19 13 28 49 37 38 38 32 29 14 28 22 42 49 57 25 59 34 15 72 33 4. General Separation Anxiety Subscale Includes items: 4. When I was a child I felt nervous about returning to school after a vacation, even when I really wanted to go. 20. I feel guilty when I leave someone. 49. When peOple I am close to go away on a trip, I worry that something might happen to them. 62. I feel scared when someone I am close to goes away. 69. I get attached to people easily. 73. I feel guilty when I think negative things about my mother. 75. I am a self-conscious person. 85. I get jealous when I see a close friend of mine having a really good time with someone else. 87. I feel guilty when I think negative things about my father. 94. I am never positive that I have done a good job until someone else tells me that I have. 95. I get very frightened when I think that someday my parents will die. 106. I am Often afraid that someone I care about very much will find another person who they will like better than me. 113. Sometimes I think I need more love than anyone could possibly give me. 121. When someone rejects me I get very angry inside. 129. As a child I spent a lot of time with my parents. 132. NO one will ever love me as much as members of my family do. 137. I don't feel responsible for my mother's happiness.* 139. My family is not exceptionally close.* *Item is reflected (reverse-scored). 222 mN hmH mN mH NMH Nw HN vH mNH OH mH m mH HNH h 5 NH h mm MHH v w mH NH mm ov wOH NN vN mN mH om 5N mm mm MI 5 H N vm mm Hv mN vm oN 0N m mH mH vH vH om NH hm H w! mH o Hm mm ow mN mN MH mm mH NH m HH 5N 0N vN mN Hm NH NN mm om om NH NH 0N HN mH ow wH mm NH mN Nb NH NH HH N m hH mH mH mN MH NH mH wH mo 0H mH vH mH mm mm vw ow hm mm HN NH hm mH Nm NN mH HH o mH mH mN mm HN hN mH mH Nm 0N mv mv MH mN HH HI mN mH 5N om MN mH NH mN 5N OH mm «N ON w m NI HH HN mH MN 5N ON NH MH mH mH v NN NH mN v mMH hMH NMH mNH HNH MHH mOH mm em mm mm mm m5 mm Ne mv 0N mm. H mamHm ucmHOHmmmoo OHOom oumncmum "OHmoonm mumecm coHumummmm Hmumcmo How xHHqu GOHDMHOHHOU 223 5. Self-acceptance Subscale Includes items: 18. I usually know when to quit.* 36. The thought of committing suicide over failing at something does not occur to me.* 37. From the very first day of kindergarten I enjoyed school and couldn't wait to go.* 39. I am a capable person.* 43. Most peOple will not let you down.* 44. I like the way my face and body look.* 46. I have never been shy.* 64. As a child I was not absent from school very often.* 86. I am a calm and relaxed person.* 122. It doesn't bother me if someone close to me dis- agrees with me on an important matter.* 131. I am neither too fat nor too thin.* 141. During some period in my life I was convinced I would die young. Correlation matrix: standard score coefficient alpha = .62 36 37 39 43 44 46 64 86 122 131 141 18 15 14 25 10 8 8 3 23 21 3 5 36 15 15 9 10 11 10 13 15 0 15 37 6 10 15 14 18 22 17 9 16 39 3 21 10 12 18 8 18 8 43 17 10 15 16 20 8 9 44 21 14 9 15 52 8 46 5 10 14 8 8 64 2 l4 2 4 86 9 13 4 122 3 11 131 5 6. Mother Hostility Subscale Includes items: 58. I used to regularly have big fights with my mother. 70. My mother likes to tell me what to do. 88. It is possible to feel both love and hate for the same person. 89. I threw tantrums when I was younger if I didn't get what I wanted. 111. I have said terrible things to my mother. *Item is reflected (reverse-scored). 224 115. I used to get an upset stomach before I left for school in the morning. 117. When I was in elementary school one Of my parents would want to take me to school. 123. If I ever let out all the anger in me I might really damage someone. 124. My mother has said that my father's relatives sometimes mean more to him than we do. Correlation matrix: standard score coefficient alpha = .64 70 88 89 111 115 117 123 124 58 42 12 28 53 21 18 32 12 70 -2 9 17 10 14 21 5 88 12 18 -3 11 11 4 89 33 10 16 18 12 111 21 11 29 5 115 20 12 19 117 11 21 123 13 7. Psychosomatic Items These are additional items (not clustered) which were included in the instrument. 27. I do not catch colds easily.* 105. I, or someone in my family, get migraine headaches. 134. I, or someone in my family, have ulcers. *Item is reflected (reverse-scored). APPENDIX H CORRELATION MATRIX FOR CLUSTERS APPENDIX H CORRELATION MATRIX FOR CLUSTERS The following is the correlation matrix for the clusters cited in Appendix G. To simplify presentation of this matrix, the following numbers were assigned to each sub- scale: Basic personality type subscale Individuation subscale Father subscale General separation anxiety subscale Self-acceptance subscale Mother hostility subscale mmanl—I 0 Correlation matrix: 2 3 4 5 6 1 19 11 21 48 17 2 22 31 24 33 3 -21 30 35 4 41 25 5 54 225 APPENDIX I BACKGROUND QUESTIONNAIRE FOR FRESHMEN SAMPLE APPENDIX I BACKGROUND QUESTIONNAIRE FOR FRESHMEN SAMPLE Please check the apprOpriate response to the following questions. (Numbers following some responses are for com- puter coding purposes only, so you can just ignore them.) 1. Age: 18(1) 19(2) 20(3) 21 or over(4) Religious background: Catholic(l) Protestant(2) Jewish(3) Other(4) Race: White(1) Black(2) Other(3) Inducate number of older, younger, or same age siblings. If only child, leave blank. Brothers: Sisters: Older(l) Older(4) younger(2) younger(5) same age(3) same age(6) Current college major (or most likely choice if undecided): Job you would most like to hold after college: 226 10. 11. 227 What are your main reasons for making this choice? (Check no more than three.) parental influence(01) money(02) status(03) enjoy the work(04) want to help others(05) interesting work setting(06) Opportunity to trave1(07) talented in this area(08) can't get into field I really wanted(09) Opportunity for advancement(10) job security(ll) I have trouble controlling my weight: yes(l) no(2) I am currently overweight: yes(1) no(2) If yes, indicate how much: less than 10 1bs.(l) 11-20 1bs.(2) 21-30 1bs.(3) more than 30 1bs.(4) I am currently underweight: yes(l) If yes, indicate how much: 5 lbs. or 1ess(1) 6-10 1bs.(2) more than 10 1bs.(3) I am currently a cigarette smoker: yes(l) no(2) If yes, how much? occasional cigarette(l) half-pack a day or 1ess(2) one to one-and-a-half packs a day(3) two packs a day(4) more than two packs a day(S) I am in excellent health: yes(l) no(2) If no, what health problems do you currently have? 12. 13. 14. 15. 228 As a child, I missed school on the average of: almost never(l) a few days a year(2) about a week a year(3) 1-2 weeks a year(4) more than 2 weeks a year(S) I was sick at home quite a bit or for a long period of time as a child: yes(l) no(2) If yes, number of times: once(1) 2—3 times(2) 4-6 times(3) six or more times(4) Number of days: a few days(1) a week(2) 8 days to two weeks(3) 15 days to one month(4) more than one month(S) Illness(es): I was hospitalized as a child: yes(l) no(2) If yes, number of times: once(1) 2-3 times(2) 4-6 times(3) six or more times(4) Number of days: a few days(1) a week(2) 8 days to two weeks(3) 15 days to one month(4) more than one month(S) Illness(es) or injuries: There were times when I was afraid to go to school as a child: es(l) no(2) 16. 17. 18. 19. 20. 21. 22. 229 My father died when I was a child: yes(1) no(2) If yes, how Old were you? birth to three years old(1) 4-6 years old(2) 7-12 years old(3) My mother died when I was a child: yes(1) no(2) If yes, how old were you? birth to three years old(1) 4—6 years old(2) 7-12 years old(3) My parents were divorced when I was a child: yes(1) no(2) If yes, how Old were you? birth to three years old(1) 4-6 years old(2) 7-12 years old(3) If you lost one of your parents through death or divorce as a child, was that parental role taken over by a step-parent? yes(1) no(2) By a relative or friend of the family? yes(1) no(2) During my childhood I was separated from my mother for a long period of time: yes(1) no(2) don't know(3) I was adopted: yes(1) no(2) If yes, at what age? birth to six months(l) 7-12 months(2) 13 months to two years(3) two to three years(4) more than three years(5) I have lived away from home before coming to college? yes(1) no(2) 23. 24. 25. 26. 230 If yes, for how long? a few weeks(l) a few months(2) 6 months to one year(3) more than one year(4) For what reason? I am usually close to (include both friendships and romances): only one person at a time(1) a couple of peOple at a time(Z) several different people at a time(3) many different peOple at a time(4) During my life most of my close relationships have been with: not really anyone(1) family members(2) same sex peers(3) Opposite sex peers(4) equally with both sexes in peer group(5) adults Older than myself(6) In the future I would like to deveIOp more close relationships with: not really anyone(1) family members(2) same sex peers(3) Opposite sex peers(4) equally with both sexes in peer group(5) adults older than myself(6) I would rate my anxiety over being separated from my home, family, and old friends as: nonexistent(1) low(2) moderate(3) high(4) very high(5) 27. 28. 29. 30. 31. 32. 231 I would describe myself as (check three to five adjectives): agitated(01) strong(09) happy(02) confused(10) fearful(03) excited(11) confident(04) isolated(12) unhappy(05) depressed(l3) relaxed(06) cheerfu1(l4) nervous(07) Optimistic(15) well-adjusted(08) desperate(16) How well does the idea of living alone appeal to you? not at a11(1) somewhat(2) very much(3) Since I have arrived on campus I have made the following total number of contacts with someone back home (include telephone calls, visits, and letters received as well as made): none as yet(l) 1-3(2) 4-6(3) 7-10(4) more than 10(5) I consider myself to be homesick to the following degree: not at all(l) slightly(2) noticeably so(3) very much(4) So far I would rate my overall adjustment to college as: unsatisfactory to me(l) acceptable(2) very satisfactory(3) Please check if you are troubled by any of the following at the present time: upset stomach(1) diarrhea(2) sweaty palms(3) headaches(4) dizziness(5) difficulty breathing(6) 232 sleeping difficulties (includes nightmares)(7) lack of appetite(8) increased appetite(9) other physical complaints(10) Please indicate the nature of these: 33. I think about home: almost never(l) occasionally(2) regularly(3) constantly(4) REFERENCES REFERENCES Abelin, E. The role of the father in the separation- individuation process. In J. B. McDevitt & C. F. Settlage (Eds.), Separation-Individuation: Essays of Margaret Mahler. New York: International Uni- versities Press, 1971. Ainsworth, M. D. S. Objection relations, dependency, and attachment: A theoretical review of the infant- mother relationship. Child Development, 1969, 39, 969-1025. Ainsworth, M. D. S., & Bell, S. M. Attachment, exploration, and separation: Illustrated by the behavior of one- year-olds in a strange situation. Child Development, 1970, 31, 49-67. Alloway, T., Pliner, P., & Krames, L. Attachment behavior. New York: Plenum Press, 1977. Amrith, M. V. Introjective and projeCtive identification. Samiksa, 1975, 22, 1-12. Angel, K. On symbiosis and pseudo-symbiosis. Journal of the American Psychoanalytic Association, 1967, 15, 294-316. Angel, K. The role of the internal object and external object in object relationships, separation anxiety, object constancy and symbiosis. International Journal of Psychoanalysis, 1972, 5;, 541-546. Bak, R. D. The schizophrenic defence against aggression. International Journal of Psychoanalysis, 1954, 35, 129-134. Bak, R. D. Object-relationships in schizophrenia and per- version. International Journal of Psychoanalysis, 1971, 53, 235-242. 233 234 Bechtoldt, H. P. Construct validity: A critique. In D. N. Jackson & S. Messick (Eds.), Problems in_human assessment. New York: McGraw-Hill, 1967. Benedek, T. The psychosomatic implications of the primary unit: Mother-child. American Journal of Orth0p§y- chiatry, 1949, lg, 642-654. Bischof, N. A systems approach toward the functional con- nections of attachment and fear. Child Development, 1975, 36, 801-817. Bloom, B. L. (Ed.). Psychological stress in the campus community. New York: Behavioral Publications, 1975. Bowlby, J. Separation anxiety. International Journal of Psychoanalysis, 1960, 51, 89-113. Bowlby, J. Separation anxiety: A critical review of the literature. Journal of Child Psychology and Psy- chiatry, 1961, 1, 251—269. Bowlby, J. Attachment and loss (Vol. 2, Separation: Anxiety and anger). New York: Basic Books, 1973. Bowlby, J., Ainsworth, M. Boston, M., & Rosenblugh, D. The effects of mother-child separation: A follow-up study. British Journal of Medical Psychology, 1956, 22, 211-247. Brody, S. Patterns of mothering. New York: International Universities Press, 1956. Brody, S., & Axelrad, 8. Anxiety and ego formation in infancy. New York: International Universities Press, 1970. Buxhaum, E. The problem of separation and the feeling of identity. Child Welfare, 1955, 35, 8-15. Choi, E. H. Father-daughter relationships in school phobia. Smith College Studies in Social Work, 1961, 31, 152-178. Clyne, M. B. Absent. London: Tavistock Publications, 1966. Coleman, R. W., Kris E., & Provence, S. The study of vari- ations of early parental attitudes: A preliminary report. Psychoanalytic Study of the Child, 1953, 2, 20-470 235 Cronbach, L. J., & Meehl, P. E. Construct validity in psy- chological tests. In D. N. Jackson & S. Messick (Eds.), Problems in human assessment. New York: McGraw-Hill, 1967. Cullen, J. H. Experimental behavior. Irish University Press, 1974. Edelston, H. Separation anxiety in young children: A study of hospital cases. Genetic Psychology Monographs, 1943, gg, 3-95. Edwards, A. Technigues of attitude scale construction. New York: Appleton-Century-Crofts, 1957. English, H., & English, A. Psychological and psychoanalytic terms. New York: David McKay and Company, 1958. Erikson, E. H. Childhood and society. New York: W. W. Norton, 1963. Estes, H. R., Haylett, C. H., & Johnson, A. M. Separation anxiety. American Journal of Psychotherapy, 1956, 12, 682-695. Ezriel, H. The scientific testing of psychoanalytic findings and theory. British Journal of Medical Psychology, 1951, 24, 30-34. Fairbairn, W. R. Psychoanalytic studies of_thepersonality. London: Tavistock Publications, 1952. Feldman, S. S., & Ingham, M. E. Attachment behavior: A validation study in two age groups. Child Develop- ment, 1975, 46, 319-330. Fenichel, O. Psychoanalytic theorypof neurosis. New York: Norton, 1945. Feshbach, S. Aggression. In P. H. Mussen (Ed.), Carmichael's manual of child psychology (3rd ed.). New York: John H. Wiley, 1970. Fliess, R. Ego and body ego. New York: International Uni- versities Press, 1961. Forrester, R. Why do we have to say good-bye. Canadian Psychiatric Association Journal, 1974, 12, 517-522. Foss, B. M. Determinants of infant behavior. New York: John H. Wiley, 1961. 236 Freud, A. Aggression in relation to emotional development: Normal and pathological. Psychoanalytic Study of the Child, 1949, 3/4, 37-42. Freud, A., & Burlingham, D. Infants without families. New York: International Universities Press, 1944. Freud, S. [Mourning and melancholia.] In J. Strachey (Ed. and trans.), Standard edition of the complete works of Sigmund Freud. London: Hogarth Press, 1937. (Originally published, 1917.) Freud, S. [A general introduction to psychoanalysis.] (J. Riviere, Ed. and trans.) New York: Liveright Publishing Corporation, 1935. (Originally published, 1920.) Freud, S. [Inhibitions, symptoms and anxiety.] In J. Strachey (Ed. and trans.), Standard edition of the complete works of Sigmund Freud. London: Hoqarth Press, 1959. (Originally published, 1926.) Freud, S. [The problem of anxiety.] (H. A. Bunker, Trans.) New York: W. W. Norton, 1963. (Originally published, 1936.) Furer, M. The development of a preschool symbiotic psycho- tic boy. Psychoanalytic Study of the Child, 1964, 12, 448-469. Glenn, M. L. Separation anxiety: When the therapist leaves the patient. American Journal of Psychotherapy, 1971, 25, 437-446. Gorsuch, R. L. Facppr analysis. Philadelphia: W. B. Saunders, 1974. Greenacre, P. Trauma, growth, and personality. New York: W. W. Norton, 1952. Greenacre, P. Early physical determinants in the develop- ment of the sense of identity. Journal of the American Psychoanalytic Association, 1958, 6, 612-627. Greenbaum, R. S. Treatment of school phobias: Theory and practice. American Journal of Psychotherapy, 1964, lg, 616-634. Guilford, J. P. Psychometric methods. New York: McGraw- Hill, 1954. 237 Guntrip, H. Schizoid phenomena, opject relations and the self. New York: International Universities Press, 1969. Hansburg, H. G. Adolescent separation anxiety: A method for the study of adolescent ssparation problems. Springfield, 111.: C. C. Thomas, 1972. Hartmann, H., Kris, E., & Loewenstein, R. M. Notes on the theory of aggression. Psychoanalytic Study of the Child, 1949, 3/4, 9-36. Heinicke, C. M., & Westheimer, I. J. Brief separations. New York: International Universities Press, 1965. Helmstadter, G. C. Principles of psychological measurement. New York: Appleton—Century-Crofts, 1964. Henryk-Gutt, R., & Rees, W. L. Psychological aspects of migraine. Journal of Psychosomatic Research, 1973, 31, 141-153. Hinsie, L., & Campbell, R. Psychiatric dictionary. New York: Oxford University Press, 1970. Hoffer, W. Development of the body ego. Psychoanalytic Study of the Child, 1950, 3, 18-24. (a) Hoffer, W. Oral aggressiveness and ego development. International Journal of Psychoanalysis, 1950, 33, 156.160. (b) Hollender, M. C. The need or wish to be held. Archives of General Psychiatry, 1970, 33, 445-453. Kafka, E. On the deve10pment of the experience of mental self, the bodily self, and self consciousness. Psychoanalytic Study of the Child, 1971, 33, 217-240. Kerlinger, F. N. Foundations of behavioral research. New York: Holt, Rinehart, and Winston, 1964. Kestenbergh, J. From organ-object imagery to self and object representations. In J. B. McDevitt & C. F. Settlage (Eds.), Separation-individuation: Essays in honor of Margaret Mahler. New York: International Uni- versities Press, 1971. Klagsbrun, M., & Bowlby, J. Responses to separation from parents: A clinical test for young children. Projective Psychology, 1976, 33, 7-27 238 Knott, P. D. A further methodological study of the measure- ment of interpersonal aggression. Psychological Reports, 1970, 33, 807-809. Krout, M. H., & Tabin, J. K. Measuring personality in developmental terms: The personal preference scale. Genetic Psychology Monographs, 1954, 33, 289-335. Loevinger, J. Objective tests as instruments of psychologi- cal theory. In D. N. Jackson & S. Messick (Eds.), Problemsy3n human assessment. New York: McGraw- Hill, 1967. Maccoby, E. E., & Masters, J. C. Attachment and dependency. Mahler, Mahler, Mahler, Mahler, Mahler, Mahler, Mahler, Mahler, In P. H. Mussen (Ed.), Carmichael's manual of child psychology (3rd ed.). New York: John H. Wiley, 1970. M. S. On child psychosis and schizophrenia: Autis- tic and symbiotic infantile psychosis. Psychoana3y- tic Studyyof the Child, 1952, 3, 286-305. M. 8. On sadness and grief in infancy and child— hood: Loss and restoration of the symbiotic love object. Psychoanalytic Studyyof the Child, 1961, 33, 332-351. M. S. Thoughts about development and individuation. Psychoanalytic Study of the Child, 1963, 33, 307-324. M. S. A study of the separation-individuation pro- cess and its possible application to borderline phenomena in the psychoanalytic situation. Psycho- analytic Study of the Child, 1971, 33, 403-424. M. S. Rapprochement subphase of the separation- individuation process. Psychoana3ytic Quarter3y, 1972, 53, 487-506. M. S. Symbiosis and individuation: The psychologi- cal birth of the human infant. Psychoanalytic Study of the Child, 1974, 33, 89-106. M. S., & Goslinger, B. J. On symbiotic child psy- chosis: Genetic, dynamic and restitutive aspects. Psychoanalytic Study of the Child, 1955, 33, 195-212. M. S., Pine, F., & Bergman, A. The psychological birth of the human infant. London: Hutchinson, 1975. 239 Marcus, I. M. The experience of separation-individuation in infancy and its reverberations through the course of life: 2. Adolescence and maturity. Journal of the American Psychoanalytic Association, 1973, 33, 155-167. Masterson, J. Maternal clinging, separation-individuation, and the borderline syndrome. International Journal of Child Psychotherapy, 1973, 3, 331-344. McDonald, J., & Sheperd, G. School phobia: An overview. Journal of School Psychology, 1976, 33, 291-307. Meehl, P. E. The dynamics of "structured" personality tests. In D. N. Jackson & S. Messick (Eds.), Problems_3n human assessment. New York: McGraw- Hill, 1967. Parens, H. A contribution of separation-individuation to development of psychic structure. In J. B. McDevitt & C. F. Settlage (Eds.), Separation-individuation: Essays in honor of Margaret Mahler. New York: International Universities Press, 1971. Rabin, A. I. Adapting and devising projective methods for special purposes. In A. I. Rabin (Ed.), Prpjective techniques ingpersonality assessment. New York: Springer, 1968. Robertson, J., & Robertson, J. Young children in brief separation: A fresh look. Psychoanalytic Stugy of the Child, 1971, 33, 264-315. Sarnoff, I. Testing Freudian concepts: An experimental social approach. New York: Springer, 1971. Shirley, M., & Poyntz, L. The influence of separation from the mother on children's emotional responses. Journal of Psychology, 1941, 33, 251-282. Sperling, M. School phobias: Classification, dynamics, and treatment. Psychoanalytic Study of the Child, 1967, 33, 375-401. Spitz, R. A. Anaclitic depression: An inquiry into the genesis of psychiatric conditions in early child- hood, II. Psychoanalytic Study of the Child, 1946, 3, 313-341. (a) Spitz, R. A. The smiling response: A contribution to the ontogenesis of social relations. Genetic Psychology Monographs, 1946, 33, 57-125. (b) 240 Spitz, R. A. Anxiety in infancy: A study of its manifesta- tions in the first year of life. International Journal of Psychoanalysis, 1950, 33, 138-143. Spitz, R. A. Aggression and object relations. In R. Lowenstein (Ed.), Drives, affectsy behavior (Vol. 1). New York: International Universities Press, 1953. Spitz, R. A. The first year of life: A psychoanalytic study of normal and deviant development of object relations. New York: International Universities Press, 1965. Spock, B. The striving for autonomy and regressive object- relationships. Psychoanalytic Study of the Child, 1963, 33, 361-364. Sternschein, I. The experience of separation-individuation in infancy and its reverberations through the course of life: Maturity, senescence, and sociological implications. Journal of the American Psychoana3y- tic Association, 1973, 33, 633-645. Suttenfield, V. School phobia: A study of five cases. American Journal of Orthopsychiatry, 1954, 33, Szyrysnki, V. School phobia, its treatment and prevention. The Psychiatric Journal of the University of Ottawa, 1976, 3, 165-170. Talbot, M. Panic in school phobia. American Journal of Orthopsychiatry, 1957, 33, 286-295. ‘— Tennes, K. H., & Lampl, E. E. Stranger and separation anxiety in infancy. 39urnal of Nervous and Mental Disease, 1964, 139, 247-254. Tennes, K. H., & Lampl, E. E. Some aspects of mother-child relationship pertaining to infantile separation anxiety. Journal_pf Nervous and Mental Disease, 1966,‘333, 426-437. Waldfogel, S., Coolidge, J. C., & Hahn, P. B. The deve10p- ment, meaning and management of school phobia. American Journal of Orthopsychiatry, 1957, 27, 754-780. _— Williams, J. S. Aspects of dependence-independence conflict in children with asthma. Journal of Child Psychology and Psychiatry and Allied Disciplines, 1975, 16, 199-218. "- 241 Winestine, M. C. The experience of separation-individuation in infancy and its reverberations through the course of life: 1. Infancy and childhood. Journal of the American Psychoanalytic Association, 1973, 33, 135-154. Winnicott, D. W. Transitional objects and transitional phenomena: A study of the first not-me possession. International Journal of Psychoanalysis, 1953, 33, 89-97. Winnicott, D. W. The capacity to be alone. International Journal of Psychoanalysis, 1958, 33, 416-420. Wold, P. Family structure in three cases of anorexia nervosa. American Journal of Psychiatry, 1973, 130, 1394-1397. Woodbury, M. A. Altered body-ego experiences: A contri- bution to the study of regression, perception, and early development. Journal of the American Psy- choanalytic Association, 1966, 33, 273-303. Yamazaki, M. [A study of school phobia: Family dynamics that obstruct socialization of chronic cases with main reference to the personality of the fathers.] (F. Masuzewa, Trans.) Journal of Mental Health, 1973, 33, 29-48. Young, L. D., Suomi, S. S., Harlow, H. F., & McKinney, W. T. Early stress and later response to separation in Rhesus monkeys. American Journal of Psychiatry, 1973, 333, 400-405.