”Ky-infiz : 4, VS“ My.‘ ‘th‘ 4,4 A . 4 ' 3".‘r ,u 4 . .‘L‘v 90 f V 5 42.: u , 4‘ Iifi'flfifl‘: ‘ “3'1 3,141.4 3 4 «V 4 x... 54 , 1 .1, y L . i't'l . film >3, I. 1‘ wt ¢ Ar-f's‘i .Lw cr" is: ‘ F}- ‘4" s L .x . .‘L. M fii“ VIA. " 4 , xx IIIIIIIII IIIIIIIIIII IIIIIIII IIIIIII IIIIIIIIIIII THESIS This is to certify that the dissertation entitled Relationship of Depression to Gender Based Perception of Self-in-relation to Others presented by Nayda M. Flores has been accepted towards fulfillment of the requirements for ph,D. Psychology degree in W Major professor Date October 20, 1994 MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY MIchlgan State University PLACE ll RETURN 80X to roman this checkout from yout record. TO AVOID FINES return on or before date duo. DATE DUE DATE DUE DATE DUE MRI 4 4‘ ' I I I I i I II I ] MSU I. An Nflrdevo Adlai/Equal Opportunlty InstItqun W. “E I. l m1 ‘_. RELATIONSHIP OF DEPRESSION TO GENDER BASED PERCEPTION OF SELF-IN-RELATION TO OTHERS by Nayda Maria Flores A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1995 ABSTRACT RELATIONSHIP OF DEPRESSION TO GENDER BASED PERCEPTION OF SELF-IN-RELATION TO OTHERS by Nayda M. Flores This work explored the existance of two gender based types of self, the Relational/ Connected Self (RC3) and the Separate/ Objective Self (SOS), in a large sample of 726 participants (437 women and 289 men), and the relationship between type of self and depression in smaller matched groups (29 depressed and 29 nonde- pressed women, 19 depressed and 19 nondepressed men). The Beck Depression Inventory and Form B of the Depression Adjective Checklist were used to determine level of depression. Several measures were used to identify type of self, including the Revised Relationship Self Inventory (RRSI), an adapted version of the Hassles and Uplifts Scales, and projective stories elicited with some of the Thematic Apperception Cards and two incomplete stories. The RRSI failed to distinguish among individuals in this study and no conclusions were drawn from it. Results from the different measures were consistent in providing support for the existance of the two types of gender based self. Women were characterized by a re- ii lational/ connected self (RCS), while men were characterized by separate/objective self (SOS). Strong support was also observed for the prediction derived from the Self-in-Relation theory that in depression women experience a loss in their sense of relational self not experienced by men. Depressed women demonstrated fewer RCS elements than their nondepressed counterparts. In addition, they were not charac- terized as RCS. Their characterization as predominantly SOS was due to decrease in RCS elements without an increase in SOS as compared to nondepressed women. The proportions of depressed and nondepressed men evidencing RCS or SOS did not differ statistically. As part of this study a scoring scheme for projective stories was developed to identify type of self. iii ACKNOWLEDGEMENTS I want to thank all the people that contributed to this research. I am grateful to Dr. Dozier W. Thornton for his encouragement and guidance. Special thanks go to my mother, Eva Gonzalez de Flores, and my husband, Carlos W. Salgado, for their assistance in entering large amounts of data into the computer and running statistical programs for the analysis. Several friends made this task much easier. I thank Dr. John Benites for much needed encouragement at times when it seemed that the work would never be finished. Thanks to Vicky Grunder for her thoughtful feedback on the scoring scheme. Thanks to my dear friends: Elida and Felix Marti for their hospitality. iv TABLE OF CONTENTS page List of Tables .................................................................. viii List of Figures .................................................................. xi Chapter I - Introduction Incidence of Depression ............................................... 1 Statement of Purpose ........................................... 2 History of the Concept of Depression .................................. 2 Multi-levels Approach to Depression .................................. 6 Self-in-Relation Theory .............................................. 14 Depression and the Self-in-Relation Theory ........................... 20 List of Hypotheses ................................................... 26 Chapter 11- Method Participants ......................................................... 29 Instruments ......................................................... 29 Procedure ........................................................... 35 Analyses of Data .................................................... 38 Chapter III- Results Thesis I ............................................................. 47 Hypothesis A1 ............................................ 47 Hypothesis A2 ............................................ 47 Hypothesis A3 ............................................ 47 Hypothesis B ............................................. 51 Hypothesis C ............................................. 53 Hypothesis D ............................................. 55 Thesis II ............................................................ 55 Hypothesis E ............................................. 55 Hypothesis F ............................................. 59 Hypothesis G1 ............................................ 61 Hypothesis G2 ............................................ 61 Chapter IV- Post-Hoe Findings Primacy-of-Other Scale .............................................. 67 Violent Content in the Story Completion Task ....................... 68 Chapter V- Discussion ...................................................... 71 Chapter VI- Conclusions .................................................... 84 Appendix A .................................................................. 86 Demographic Information Questionaire ............................... 87 Written Instructions for the Thematic Apperception Test ............. 89 Story Completion Task: Instructions and Incomplete Stories .......... 91 Consent Form ....................................................... 96 Handout for Participants ............................................ 98 Appendix B ................................................................. 100 vi Mean Scores and Standard Deviations on the Beck Depression Inventory and Form B of the Depression Adjective Check Lists for Data Collected in 1988 and 1992 ...................... 101 Distribution of Scores on the Beck Depression Inventory for the Whole Sample (n = 726), Women (n = 437) and Men (11 = 289) ................................................ 103 Distribution of Scores on the Depression Adjective Check Lists- Form E for the Whole Sample (n = 726), Women (n = 437) and Men (11 = 289) ................................ 105 Correlations between the Beck Depression Inventory and Form B of the Depression Adjective Checklist ........................ 107 Appendix C ................................................................. 109 Appendix D ................................................................. 122 List of References .......................................................... 124 vii Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 LIST OF TABLES Participants’ Demographic Information .......................... Means, Standard Deviations and Range of Scores on the Beck Depression Inventory and Form E of the Depression Adjective Check Lists ........................................... Mean Scores on the Beck Depression Inventory and Form E of the Depression Adjective Check Lists for Depressed and Matched Nondepressed Samples ................................. Demographic Information from the Depressed and Matched Nondepressed Samples .......................................... Correlations Between Scales of the Revised Relationship Self Inventory for 437 Women and 289 Men .......................... Means and Standard Deviations on Relevant Revised Relationship Self Inventory Scales for 437 Women and 289 Men ........................................................ Percentages of Participants Endorsing Revised Relationship Self Scales as Self-descriptive ........................................ T-tests of Mean Difference Scores on the Adapted Hassles and Uplifts Scale Against a Zero Difference Hypothesis ........... T-tests Between 29 Nondepressed Women and 19 Nondepressed viii page ....30 ....39 ....41 ....42 ....44 ....48 ....50 ...52 Men on Hassles and Uplifts Subscales of the AHUS .................. 52 Table 10 T-tests of Mean Difference in Length of Stories from the Story Completion Tas Against a Zero Difference Hypothesis ............... 54 Table 11 Paired t-tests Between Mean Difference Scores of Depressed and Matched Nondepressed Participants on the Adapted Hassles and Uplifts Scale ........................................................ 57 Table 12 Paired t-tests Between Mean Scores from 29 Depressed and 29 Matched Nondepressed Women on Hassles and Uplifts Subscales of the AHUS ............................................. 58 Table 13 Paired t-tests Between Mean Scores from 19 Depressed and 19 Matched Nondepressed Men on Hassles and Uplifts Subscales of the AHUS .......................................................... 60 Table 14 Paired t-tests Between Mean Differences from Depressed and Matched Nondepressed Participants on the Story Completion Task ................................................... 62 Table 15 Paired t-tests Between Corrected Mean Differences from Depressed and Matched Nondepressed Participants on the Story Completion Task ............................................. 63 Table 16 Percentage of Women and Men that Included Violent Content in the Story Completion Task ....................................... 70 Table 17 Murders and Suicides Included by Women and Men in a Story ix Table 18 Table 19 about a Difficulty in a Relationship and in a Story about a Difficulty in Achievement ......................................... 70 Mean Scores and Standard Deviations on the Beck Depression Inventory and form E of the Depression Adjective Check Lists for Data Collected in 1988 and 1992 ................................ 102 Correlations Between the Beck Depression Inventory and Form E of the Depression Adjective Check Lists .................... 108 Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 LIST OF FIGURES page Multi-levels Approach to Depression .................................. 7 Distributions of Scores on SOS and RS for 29 Depressed Women, 29 Matched Nondepressed Women, 19 Depressed Men and 19 Nondepressed Men ..................................... 65 Distributions of Scores on RS for 29 Depressed Women, 29 Matched Nondepressed Women, 19 Depressed Men and 19 Nondepressed Men ............................................... 66 Distribution of Scores on the Beck Depression Inventory for the Whole Sample (n = 726), Women (11 = 437) and Men (n = 289) .................................................... 104 Distribution of Scores on the Depression Adjective Check Lists- Form E for the Whole Sample (11 = 726), Women (n = 437) and Men (11 = 289) ................................................ 106 Distributions of Scores on the Area of Help on Both Self-Types for the Four Small Groups Combined (n = 96) ..................... 123 xi Chapter I INTRODUCTION Incidence of Depression Marsella, Sartorius, Jablensky, and Fenton (1985) estimated that 100 mil- lion people worldwide suffer from a depressive disorder. For the United States, it has been estimated that one in five persons will experience a depressive disorder at some point in their lives (Weissman & Myers, 1978). More people were hospitalized in 1970 for depression (24%) than for schizophrenia (22%) and alcoholism (16%), becoming the leading cause of hospitalization in the United States (Schuyler, 1974). The incidence of depression in women is higher than in men. Weissman and Klerman (1977) reviewed prevalence studies for the United States and other countries between 1936 and 1973. They concluded that a 2:1 ratio was consistently found in the United States. Most other countries also showed a comparable higher incidence in women. As will be discussed later, cultural factors (i.e., presence of par- ticular symptoms vary among cultures) and methodological problems (i.e., different diagnostic criteria) have obscured cross-cultural research on depression (Marsella et al., 1985; Weissman & Klerman, 1977). Nevertheless, a sex difference is usually ob- served, where women are twice as likely as men to experience a depressive disorder during their lifetime. This observation has lead many to conclude that depression is mainly a woman’s disorder. Although one in five women in the United States will be depressed at some point in their lifetime, a significant number of men (one in ten) will also experience it (Schuyler, 1974; Weissman, 1981). Any theory attempting to explain the etiology of depression must be able to account for its existance in most cultures, the difference in manifestations across cultures, and the higher incidence in women. Self-in-relation theory has re- cently attempted to address these issues (Kaplan, 1986). The present study will test the validity of the two concepts of self advanced by this theory, relational self and separate self, and test their relevance in the experience of depression in women and men by exploring the hypothesis that depressed women experience a loss of sense of relational self that is not part of the experience of depressed men. History of the Concept of Depression Depression is one of the oldest known psychological disorders (Jackson, 1985; Marsella et al., 1985). Jackson (1985) traced the history of the concept of de- pression, finding evidence of its identification as a major disease in Hippocrates’ (460 - 377 BC.) writings. It was referred to as melancholia and defined as a prolonged fear or depression, associated with lack of appetite, insomnia, irritability and restlessness (Jackson, 1985). From the second century, when Galen described it, through the sev- enteenth century, melancholia was seen as a chronic form of madness which caused the afflicted to experience fear, sadness, misanthropy, tiredness of life, a particular type of delusion, and could include the other symptoms already mentioned by Hippocrates (Jackson, 1985). It was not until the sixteenth century that the experience of guilt became one of its symptoms (Jackson, 1985). Jackson (1985) explored the two ways of interpreting the behaviors associ- ated with depression that co-existed during the middle ages: melancholia and acedia. Melancholia was a disease for which the individual was not held responsible, and which required medical attention. Acedia was a sin that referred to behaviors that neglected the individual’s religious and communal duties. Acedia eventually became the sin of sloth, emphasizing laziness and losing the element of dejection. Interestingly, it is after identifying acedia as a sin that guilt appears as a feature of melancholia. The presence of a delusion as a necessary symptom, was discarded during the seventeenth century (Jackson, 1985). The term melancholia continued to evolve, gradually becoming known as depression (Jackson, 1985). The concept of depression has been used academically to cover a wide range of experiences: a temporary state of mind, a symptom, a mood, a syndrome, a character style, and a disease (Jackson, 1985; Lutz, 1985; Schuyler, 1974; Weissman & Klerman, 1977). Schuyler (1974) suggests that depression could be understood as a spec- trum, ranging from a normal temporary emotional state to a psychotic disorder. Others have used this approach (Bemporad, 1983; Coyne, 1985). This paper assumes that depressive experiences are part of the human repertoire of emotions, and occur on a continuum ranging from a passing mood to a chronic disorder. Depression is usually defined by the experience of feelings of dysphoria (Diagnostic and Statistical Manual of Mental Disorders [DSM] -III, 1981) and de- jection (Jackson, 1985). The DSM-IIIR (1987) introduced the concept of a mood disorder, where there is ”a prolonged emotion that colors the whole psychic life” (DSM-IIIR, 1987, p. 213). Mood disorders include Bipolar disorders where periods of elation are experienced at times, and Depressive disorders where manic states have never been present. The basic features of depressive disorders are: ”...depressed mood...or loss of interest or pleasure in all, or almost all, activities...feeling depressed, sad, hopeless, discouraged, down in the dumps...” (DSM-IIIR, 1987, pp. 218-219). The disorders are divided depending on duration and degree of impairment experi- enced. The DSM-IV (1994) did not introduce changes in the concept of depression. Although, a composite of symptoms resembling current ideas about de- pressive syndromes was shown to exist and cause concern early in western culture (Jackson, 1985; Marsella et al., 1985), consistency as to which symptoms occur to- gether has not been characteristic. These differences in patterns of symptoms not only occurred across time, but have also been found across cultures (Beeman, 1985; El- Islam, 1981; Marsella, 1980). Psychological symptoms such as guilt and self-reproach are not experienced in depressive states in some cultures (El-Islam, 1981). Even the patterns of physical symptoms experienced differ between cultures (Beeman, 1985). This diversity in symptomatology noticed in different cultures has led to the ques- tioning of the universality of depression (Beeman, 1985). But the diversity in symptomatology can be understood if culture is con- sidered a variable that affects the perception of depression. One way in which culture has confounded the study of depression has occurred through measurement. For ex- ample, some cultures lack a word for emotions associated with depression, yet such emotions might be experienced by people of those cultures (Marsella et al., 1985; Shweder, 1985). Also, people convey emotional states through metaphors, stylistic shifts, and tone of voice (Beeman, 1985), all difficult to capture in most measuring devices. But culture affects the perception of depression in a more profound way. Marsella (1980) suggested that depression will have different meanings and conse- quenses depending on the culture studied. By culture defining the experience of self (Hsu, 1985; Marsella, 1985), it influences the manifestation of depression which is a disorder of the self. Another way in which culture has an influence in how depression is experienced is through emotions. Besides its physiological determinants, emotions are the result of a person’s assessment of the relationship of the self to the world it is in (Keyes, 1985). These assessments are made through interpersonal interactions within the context of cultural rules limiting which emotions are considered appro- priate for the particular person to display. In turn, these assessments will influence relationships (Keyes, 1985; Lutz, 1985; Schieffelin, 1985; Shweder, 1985). Thus, in different cultures the experience of depression is organized differently. Multi-levels Approach to Depression In order to explain the diversity within the experience of depression and the higher rate in women, researchers have explored an array of variables. Even though significant associations have been found between many of these variables, none of the theories tested so far has been able to account satisfactorily for the sex difference in prevalence (Nolen-Hoeksema, 1987). Recently, Marsella et a1. (1985) proposed that a multi-level approach to depression might account better for the experience of depres- sion in different cultures. A multi-level model would integrate: environmental variables (death of significant other, family composition, early parental loss), cultural variables (sex roles, accepted symptoms as a function of status, age and sex of the afflicted person, interpretation of life events, rites related to loss), interpersonal interactions (within the family, with the community), personal coping skills, individuals person- ality factors (cognitions, previous learnings, psychological dynamics),and biological variables (hormones, genetic predispositons). Figure 1 presents a simplified version of a multi- levels model that orga- nizes the variety of variables that have been shown to be related to depression, while demonstrating the conclusion that a single level would not te able to explain depres- sion completely. It also offers a framework for the major theories about the etiology of depression. The environment provides the physical and historical context, including .l | ENVIRONMENT |.. ....... I Anthmpobglcal | CULTURE | I Sociological | sex nous | I | msm mum-m PATTERNS | I Psychological [ oven-r BEHAVIOR | ........ .l l a l a f l IBEHAVIOML H (:06wa 1...] PSYCHODYNAMIC | I J* l Biological | soar svsreus I Flour. 1: Hum-Levels Approach to Damion. specific geographical area, period in history, and composition of family. Examples of variables at this level that have been related to onset of depression are death of a sig- nificant other and unemployment (Brown & Harris, 1978). Another example is early parental loss through abandonment which has been correlated with later episodes of depression, with the variables of sex of parent lost, sex of the subject, and subject’s age at the time of loss, contributing to degree of vulnerability (Drill, 1987; Roy, 1985; Tennant, Hurry &. Bebbington, 1982). Sex roles seem to be associated with depression also. Women in west— ern cultures experience depression at a higher rate than men (Weissman & Klerman, 1977). It has been consistently found that married women account for this higher rate (Briscoe, 1982; Gove, 1972; Weissman & Klerman, 1977). Attributes considered feminine are more often used as signs of mental illness in general (Boverman, Bover- man, Clarkson, Rosenkrantz & Vogel, 1970) and of depression specifically (Hammen & Peters, 1978) than attributes considered masculine. Evidence is accumulating which suggests that lack of masculine role attributes might be more influencial in develop- ing depression than the presence of feminine traits (Heilbrun & Han, 1986; Funabiki, Bologna, Pepping & Fitzgerald, 1980; Roy, 1985; Welkowitz, Lish & Bond, 1985). Gen- der related coping skills are also beginning to be associated with depression (Funabiki et al., 1980; Parker, Brown & Blignault, 1986). Some studies have compared patterns of interpersonal interaction of de- pressed women and non-depressed women. Weissman and Paykel (1974) observed that depressed women experienced more communication problems with their spouses than non-depressed women, even after remission. A combination of life stress and lack of intimacy seems to contribute to the development of depression in women (Ten- nant, 1985). In a community survey, Brown, Bhrolchain and Harris (1975) found that emotional support from someone other than the spouse or boyfriend did not shield a woman from depression. Hopps, Biglan, Sherman, Arthur, Friedman and Osteen (1987) identified a dysfunctional family pattern where the mother’s dysphoric affect suppressed the family’s aggressive affect. In another study, inhibited verbal expression and deficient initiation of social interactions were found to be positively correlated with depression (Culkin & Perrotto, 1985). In this multi-levels model, overt behavior refers to the observable actions of a person. What the person does includes: the way of responding to others and cop- ing behaviors. There is some evidence that responses of depressed people to others elicit hostility and rejection from others (Hammen & Peters, 1978). As mentioned above, coping behaviors, or a person’s attempts to control or reduce stressful circum- stances (Billings et al., 1983), appear to be related to depression. Traditionally, psychology has approached the experience of depression at the level of personality. All the proposed theories assume that personal past experi- ences affect the perception and interpretation of new experiences. The cognitive approach assumes that how and what people think will de- termine their feelings and actions (Kovacs & Beck, 1985). According to Beck’s model (Kovacs & Beck, 1985), the process of thinking is governed by schemata, collections of related previous experiences. Depressed people generally have schemata composed of a negative view of self, the world, and the future, which might lead to depressive affect. These schemata are activated by stress, such as a loss. The probability of these schematas being activated is determined by degree of similarities between the present event and the content of the cognitive-affective structure. The aim of therapy from this perspective is to uncover the systematic error in the thinking process of the in- dividual. Beck’s model leaves unanswered why would women tend to have the kind of schemata that leads to depression or why their depressive schemata would be ac- tivated more often than occurs for men. Abramson, Seligman and Teasdale’s (1985) reformulated learned helpless- ness model proposes that the existance of depression and its extent will depend on the individual’s attribution of helplessness to three personality characteristics: globality versus specificity, internality versus externality, stability versus unstability. Once a person has learned that an outcome is uncontrollable, a cognitive deficit occurs, mak- ing it difficult for that person to learn later that responses produce outcomes. The aim of therapy from this perspective would be to identify the type of self—defeating attribution and challenge it to provide opportunities for new learning. The reformulated learned helplessness model does not address directly why incidence rates would differ between men and women. Findings have been incon- 10 sistent on whether women would tend to give negative global, internal and/or stable attributions more often than men. Early psychoanalytic theory suggested that depression was caused by the real loss of an object that had a established identification with the ego (Freud, 1959). The purpose of the depression was to punish the original object without openly ex- pressing hostility toward it. Klein explained that depressives were unable to express hostility directly to the object because of fear of destroying the object with their rage (Greenberg & Mitchell, 1983). As children, depressed people had parents that were emotionally unavailable to them, and concluded that their own destructiveness had made the parents unavailable. Fairbairn developed this issue further by suggesting that during childhood, depressives had fantasized that chaotic, unavailable parents were good, and would provide love if the child was different (Greenberg & Mitchell, 1983). Acknowledging the badness of the parents left them unable to make an im- pact. Thus by owning the badness, they hoped to control it. Later in life, depressives either select for or make new significant others into unavailable objects, just as the original ones, thus maintaining the initial ties (Greenberg & Mitchell, 1983). The aim of therapy based on psychoanalytic theories is to uncover the unconscious beliefs through the understanding of transference experiences. The difference in incidence of depression between women and men is not addressed. All the psychological theories discussed above assume that the process which leads to depression is the same for women and men. The psychoanalytic theories 11 provide an explanation of how certain affective experiences can predispose an individ- ual to depression. Beck’s model and the revised learned helplessness model propose learning mechanisms by which depressive thought and behaviors can be maintained. But, such psychological processes are embedded within cultural norms, that evolved through time. Being valued and loved by others has gained importance through evo- lution because being in relationships enhances ones’chances of survival. One of the advantages has been the division of tasks, allowing for varied supplies and more ef- ficient provision of needs. Because women bore children and breastfed them, women tended in early human history to do first gathering and then farming (Slocum, 1975). Men, being freer to be away for longer periods, tended to do the hunting and fishing (Leibowitz, 1975). Both sexes would have been equally valued, as each contributed necessary life maintaining supplies (Sacks, 1975). Women and men developed through history more and more efficient ways of getting the needed supplies (e.g., farming techniques and weapons), creating sur- plus and free time. Men, not being expected to care for the children, could use the free time to organize politically and pursue other tasks (Gough, 1975). Women, already assigned the duty of raising the children, had less free time, and during their free time pursued tasks that would not take them away from the children. In this way, labor was further divided. Men came to be valued for their power over others and tasks mainly 12 performed outside of relationships. As men organized politically and appropriated surplus, women were appropriated too (Sacks, 1975). As property, their contributions were expected but valued less than men’s. Being valued and loved by her mate be- came more important, raising the stakes in relationships for women. Although being in relationship with others was advantageous for both sexes, the development of sex roles and devaluation of women, created circumstances and demands for the growing child that depended on her/ his gender. Childrearing practices contribute to make girls and boys experieces unequal (Brooks-Gunn & Math- ews, 1979; Chodorow, 1974). These practices further encourage women through their development to understand their self-worth through relationships. Not only being in a relationship, but having an impact on their significant others. For men, Childrearing practices encourage them to determine their self-worth through achievements outside relationships, being guaranteed love from others if they succeed. The development of gender cognitive schemas provides cultures with a way of perpetuating sex roles (Bern, 1983). Children are taught to organize infor- mation about people according to their gender, encouraging their self-concepts and behaviors to be organized on the basis of their gender rather than on some other dimension. Oatley and Bolton (1985) hypothesized that when circumstances interfere with the roles by which people define their self-worth and there are no other sources of definition for them, depression is experienced. This would seem to explain why the 13 incidence of depression in college women and college men would be similar (Ham— men & Padesky, 1977). Female students probably have access to the same self—worth sources as male students (e.g., academics). Self-in-Relation Theory Self-in-Relation theory integrates cultural, individual and biological lev- els to explain psychological development. By looking at gender, the Self-in-Relation theory acknowledges that biological and cultural factors are interwined in creating sex differences in the development of the self. In this paper, the focus will be on how gender culture prescribes different developmental experiences based on a person’s sex, leading to the formation of different self-definitions in men and women. The role of biological processes will not be discussed further but is assumed to be an important factor also. The concept of self refers to a definition or consciousness of what one is about. It is linked to the individual’s perception of personal worth in the social world, and thus it is usually tied to the acknowledged value of that individual in the social structure. The self includes, besides the individual’s personality structure, the social role as well (DeVos, Marsella & Hsu, 1985). In Psychology the self is conceptualized as an internal image that devel- ops in each person. Most theories about how this image develops in us emphasize concepts of separation and individuation, and propose a series of stages leading to a 14 final understanding of ourselves as separate from others (Erickson, 1950; Kohlberg, 1981). But, this approach seems to describe mainly the experience of western men. In some eastern societies, the self includes others such as family and community (DeVos et al., 1985). Women from western cultures do not seem to follow a route toward a self understood as separate from others, either. Their sense of self develops within relationship, including interconnections to others as part of the definition of self. Gilli- gan (1982) does not see this as a deficiency in the development of women, but as a different type of development that strives to maintain a sense of self inseparable from interactions with others. Beginning at birth, we all have a self that is attuned to others and hope- fully experiences others (caregivers) attuned to us, each responding to each others emotions (Miller, 1984). The self is then experienced as an ”interacting” self, where effective emotional connections between the infant and the caregiver lead to healthy development in the infant (Miller, 1984). In understanding how men and women come to develop different types of selves, Chodorow (1979) adjusted psychoanalytic theory and proposed a process of differentiation/separateness that includes a capacity in the infant to perceive the emotional state of the care-giver as separate from the infant’s state and to respond and adapt to it. As the infant is adequately responded to by the care-giver, the infant develops a self through this relationship with the care-giver that is able to see the other as having different needs from those of the self. ”Concomitant with the earliest development of its sense of separate self, the infant constructs an in- 15 ternal set of unconscious, affectively loaded representations of others in relation to its self, and an internal representation of self in relationship emerges.” (Chodorow, 1979, p. 225) Although, Chodorow is still refering to the ego structure in psychoanalytic terms, where the relational self is an internalization of an earlier experience, not the on-going fluid relational self, defining itself in each interaction as the one espoused by Self-in-Relation Theory, her ideas about influence of gender in the psychological experiences of boys and girls can contribute to the model. Differences in the psychological experience of girls and boys has been observed very early in development. Hoffman (1977) found that infants and young children already differed in empathy. Both, girls and boys, were equally able to iden- tify someone else’s feelings, but girls’ experienced a vicarious affective response more often than boys. For Chodorow, gender differences in psychological experience would be encountered as the child begins to establish a sexual identity. Boys would be defin- ing themselves as not-like mother and in so doing, rejecting femenine patterns in order to develop their male identity. Girls would not confront such a situation when defin- ing themselves as female, since what is required is an identification with the mother, being like-mother and retaining feminine qualities. This would then enable girls to maintain the emotional connection with the mother, rather than to reject it. Here we will depart from individual psychology and look at the influ- ence of culture that the child faces. In western culture how we think about someone depends on which sex they happen to be. Thus, boys and girls have different develop- 16 mental experiences, where girls are encouraged to remain emotionally responsive to others, while boys are encouraged to become objective in their interactions with oth- ers. Miller (1984) suggests that boys are discouraged from their natural tendency to be emotionally connected to others and encouraged to direct their attention to their own development ignoring that of others. Because of this, boys develop a sense of effectiveness or power from individual action ”against others or over others” (Miller, 1984, p. 5), while girls self-esteem comes from emotional connections with others and taking care of the relationships. Gilligan (1982) describes a boy’s sense of self as separate / ob jective, rooted in individual agency and achievement. Boys do not perceive relationships as self- defin- ing elements, but in terms of objective reciprocity. Girls’ sense of self is described as relational/ connected, meaning that their sense of self is rooted in interdependence. Relationships are self-defining elements for girls, in which the concept of care becomes important as it leads to being emotionally connected with others. Agency in girls is directed at establishing and maintaining relationships. Miller (1984) also addresses this issue, pointing out that a sense of competence for girls derives from emotional connections. To be emotionally connected means not only to be in a relationship with someone but to take care of the relationship, contribute to the growth of the relation- ship. Miller (1984) contends that ”picking up the feelings of the other and attending to the interaction between” (p. 5) self and other becomes a motivation for girls. Thus, healthy development in girls leads to more complex ways of relating to others, not to 17 more individuation or separation. A self-in-relationship develops where ”both or all people involved are encouraged and challenged to maintain connection and to foster, adapt and change with the growth of the other” (Surrey, 1985, p. 8). Surrey (1985) describes the elements of women’s core self as: 1) an interest in, and attention to, the other person(s) which form the base for the emotional connection and the ability to empathize with the other(s), 2) the expectation of a mutual empathic process where the sharing of experience leads to a heightened development of self and other; and 3) the expectation of interaction and relationship as a process of mutual sensitivity and mutual responsability which provides the stimulus for the growth of empowerment and self-knowledge. (p. 7) Boggiano and Barrett (1991) provided evidence that interpersonal rela- tionships are more important for women in their self-evaluations than to men, whereas men report intelligence as more important in their self-evaluations than women. They also found that women had greater discrepancies between ideal self and actual self than the men. Another study by O’Brien (1991) offers more support, finding that men obtained higher scores on competence and women on likeability and lovability in a measure of sources of self-esteem. O’Brien (1991) cautioned that the differences were small, though. Several studies using thematic apperception techniques, have found that while both women and men evidence intimacy needs, these needs arouse more negative imagery for men than for women (Mazur & Olver, 1987; Pollak & Gilligan, 18 1982). Although Maccoby (1990) questions the role of culture and rearing prac— tices in producing differences in developmental experiences, she provides evidence that from very early gender differences are observed in the way that children interact among themselves. Preschool children already show a tendency to segregate by sex. Boys’ interactions are characterized by direct demands, girls use suggestions. Girls’ style is effective among the girls, but when applied in their interactions with boys it is not. Boys ignore girls’ suggestions, but do respond to demands from other boys. This is not a result of boys innability to follow suggestions, as reseach has shown that when an adult is present boys interact with girls in a more equitable way. Maccoby (1990) suggests that girls find interacting with boys aversive because they are un- able to influence them. Interactions among boys seem directed at establishing power, whereas interactions among girls seem directed at mutual influence. In a study designed to explore if the gender differences in type of moral development were strongly rooted in a person or were dependent on social expecta— tions, Bar-Yam (1991) found no significant differences according to sex. Her subjects were military men studying careers in the helping professions and military women, in an attempt to equalize the roles women and men were performing. She concluded that an inclination toward separateness or connection may be more related to indi- vidual differences and environment than to whether a person is female or male. Both the women and men in this sample are not representative of most people. Bar-Yam 19 intended it to be that way because it would help control for the influence of social ex- pectations derived from a person’s role. But by selecting a sample where the subjects had chosen to follow opposite roles than most people of their sex, it is questionable whether the results may be due to a preselection variable not included in the study instead of the influence of roles. Another problem with this study is that the subject were in their 40’ and 50’s. In a meta-analysis, Cohn (1991) established that gender differences in moral reasoning disappear in adulthood, where at higher levels of ego development people tend to use both criteria of care and justice in their judgements. Some variables that may be contributing to the disappearance of gender differences in moral reasoning for this age group are marriage and having children. The results from the above studies, with the exception of the last one, support the Self-in-Relation theory. Still, there is a need for validation of the concepts of separate/ objective self and relational/ connected self. As part of this study, several measures will be used to test the validity of these concepts, and explore whether these types of selves are distributed differently between men and women. Depression and the Self-in-Relation Theory Kaplan (1986) applied the Self-in-Relation Theory to the understanding of depression and the difference in the rates of depression consistently found for men and women. For Kaplan, the higher rate of depressive experience in women in comparison to men indicates that depression may be an exaggeration or distortion of the female 20 role and status in western culture. She points to common elements identified with depression and with normal psychological development in women, such as emotional loss, inhibition of action or assertiveness, inhibition of anger and aggression, and low self-esteem. But she goes further, suggesting that depression is triggered by different kinds of psychological loss in women and men. Most theories addressing depression accept that depressive affect is a response to a loss in the self. Freud (1959) pointed out that in mourning a loss is experienced in the outside world, while in melancholia it is experienced in the self. Kaplan observes that the loss of self men experience in depression appears to be related to what she calls a self-in-advancement. Depression is experienced by men when they fail to achieve a goal that they had come to believe would deliver to them the admiration and love of others. An object loss is experienced as they have failed to prove themselves worthy of it. This is supported by the finding that men from a college sample attributed their feelings of depression to academic concerns rather than to interpersonal relationships (Robbins & Tanck, 1991). Different from a man’s, a woman’s sense-of-self is ”grounded in the moti- vation to make and enhance relatednes to others” (Miller, 1986, p. 1). The loss women experience in depression is the loss of the self-in-relation. The loss is not merely that of an object and what this object would be providing, such as love and admiration, but the ”confirmation of their core self-structure as one which can facilitate reci- procity and affective connection in relationships” (Kaplan, 1986, p. 236). Difficulty in ability to communicate with the spouse has been linked to depression in women 21 (Weissman & Paykel, 1974). For women even succeding in their goal of providing the understanding and love to another, does not necessarily mean that it will be acknowl- edged (validated) and returned by the other. Thus, being involved in an unbalanced significant relationship would trigger depression in women. The probability of women experiencing an intimate relationship where their attempts at influencing their mates will be frustrated is higher than that of men, given the documented early tendency of men to disregard women’s attempts at influence, and women’s early tendency to accomodate men’s demands (Maccoby, 1990). The finding that married women are more likely to be depressed (Briscoe, 1982; Gove, 1972; Weissman & Klerman, 1977) could be explained by this, too. McLeod, Kessler and Landis (1992) found that speed of recovery from depression is associated with reaction of the spouse to the depressive episode. Nolen-Hoeksema and Girgus (1994) created a classification system for theories addressing the difference between women and men in rates of depression. In their system, the Self-in-Relationship theory would fall in the second model as a risk factor stemming from a characteristic of the personality. The studies cited there as contradictory evidence address the loss of particular relationships rather than a loss of the capacity to be-in- relationship. Actually, women would be less likely to develop a depression due to the loss of a particular relationship if the Self-in-Relationship theory is correct, as women would be more likely to obtain support from other rela- tionships through their already existing skills of emotional openess, while for men this 22 task would be more difficult because of lack of the same skills. Nolen— Hoeksema and Gingus (1994) did acknowledge evidence of gender differences in styles of interaction and in the report of distress resulting from oppossite sex interactions where women experience distress while men do not. Women who become depressed, according to Kaplan, feel that they do more work in the relationship (e.g., yield, provide understanding for the other person) than the significant other does. Depressed women believe that they are accomplishing their goal, giving the empathic understanding to the other and accepting the influ- ence of the other, but are not receiving the prize, being understood by the other and succeding in influencing the other. This then is interpreted as something being inherently wrong in them that makes them unlovable. Men’s dilemma would seem easier to resolve by selecting another goal that could be within reach. Women’s dilemma could be temporarily resolved in the same way, selecting another significant relationship. But given the nature of human relationships, with periods of closeness and detachment, and subject to stress, women would be constantly vulnerable to depression. Labile self-esteem has been found to be a good predictor of depression under stressful circumstances (Roberts & Monroe, 1992). This could explain the finding that depressed women are more likely than de- pressed men to have experienced more than one incident of depression in their lives (Amenson & Lewinsohn, 1981). Surrey (1985) expands on the need in women to understand others, not 23 just to be understood by others. Through this understanding of others, comes em- powerment or strengh for taking action. In applying this idea to women’s experiences in general and depressed women in particular, the following clinical observations are relevant. It is not whether women in general and depressed women consider their own needs (wants) because usually they do, but that they consider the other’s needs as well. When the other fails to engage in the expected empathic interchange and compromising expected, the women (depressed or not) are unable to take action and spend their energy trying to repair the relationship. This explains the passivity, am- bivalence and even the resentment toward the unyielding other observed in depressed women. Many women are unable to act without this type of understanding of others needs because their self- worth is bound to taking care of others. In therapy, depressed women often appear stuck, insisting on figuring out what they could provide for the other that would produce a validating response from the other. Their characteristic ambivalence is expressed in terms of wanting to act, but only with ”permission” from the other. For example, a common type of argument is: ”I want to leave him, but I want him to understand that it is the best for both of us”. Not hurting the other is, in their mind, of prime importance. The client is often perplexed about what would satisfy the other, what she needs to do to help the other become a satisfied person, i.e., happy, relaxed, sober, non-abusive, etc. Of course, resentment towards the other who is not providing empathy is also evident. These women usually benefit from dis- covering in therapy that the other’s psychological state makes it impossible for the 24 other to accept anyone’s influence and to respond adequately. That what the client or anybody else does for this person is irrelevant to the person’s satisfaction, in the same way that significant others’ in the client’s childhood could have never been satisfied by the client as a child. This explanation gives coherence to the depressed woman’s experience allowing her to base self-worth on her efforts at providing care, regardless of whether it is accepted or not. She is then relieved from the responsibility of taking care of the emotional life of this particular type of significant other, while continu- ing to define self-worth on capacity to be-in-relation. She can create a self-definition based on other relationships where she is more succesful at influence, and/or based on other types of achievements. At this point action and learning are possible. Robbins and Tanck (1991) found, using a college sample, that women do attribute their feelings of depression to problems in interpersonal relationships and men to academic concerns. Their study used a self report measure where students were asked to indicate what caused the depressed feelings. As the authors point out, these were perceived causes, not necessarily real causes. In the current study, we used both self-report and projective techniques to see if depression co—varied with differ- ences in self-definitions determined by gender. Self-report is easier to administer but questions about accuracy and influ- ence of social desirability in responses are always concerns. Projectives were included to provide verification, where participants did not know what was being measured with the projectives. The Thematic Apperception Test and a Story Completion task 25 were judged suitable for this purpose yielding the important themes in interpersonal relationships for each individual and forcing participants to address failures in rela- tionships and in academics. Self-in-relation theory offers a new explanation of depression that inte- grates cultural, individual and biological levels. It also offers an explanation of inci- dence patterns. The purpose of this dissertation is to test some of the assumptions made by this model: a) Women and men develop different self-definitions; women’s based on a perception of self-within-relationships, and men’s based on a perception of self-as-separate of relationships, b) Depressive experience will be associated with a self- definition limited by gender. In women depression will be associated with a loss of a relational sense-of-self and in men it will not. A detailed list of the theses explored and hypotheses tested follows. List of Hypotheses: Thesis I: People develop self-definitions limited by gender, where women have connected-relational selves and men have separate- objec- tive selves. Hypothesis A1: The mean score for women on the Relational/ Connected Self (R/ CS) Scale of the Revised Relationship Self Inventory (RRSI) will be 3.5 or higher (scale is self-descriptive), and the mean score for men will be lower than 3.5 (scale is not self-descriptive). 26 Hypothesis A2: The mean score for women on the Separate / Objective Self Scale (S/ OS) of the RRSI will be lower than 3.5, and the mean score for men will be 3.5 or higher. Hypothesis A3: The mean score for women on the Primacy-of-Other Self Scale (POC) of the RRSI will be higher than the mean score for men. Hypothesis B: In the Adapted Hassles and Uplifts Scale (AHUS), the mean difference between the Relationship subscale and the Out-of-Relationship subscale will be higher than zero for women and lower than zero. for men. Hypothesis C: The mean difference between the number of words used to com- plete a story about a difficulty in a relationship and the number used in completing a story about a difficulty in achievement on the Story Completion Task will be higher than zero for women and lower than zero for men. Hypothesis D: The proportion of women classified as relational-connected self on the stories elicited with the Thematic Apperception Test and the Story Comple- tion Task will be higher than the proportion of men classified in this way. Thesis II: Depressive experience in women is associated with a loss of a relational sense-of-self. Depressive experience in men is not associated with a loss of relational sense-of-self or of a sense-of-self separate from others. Hypothesis E: In the Adapted Hassles and Uplifts Scale (AHUS), mean dif- 27 ference scores between the subscales: Relationship and Out-of-Relationship will not differ between depressed and matched nondepressed participants within each sex. Hypothesis F: The mean difference between the number of words used to com- plete a story about a difficulty in a relationship and the number used in completing a story about a difficulty in achievement on the Story Completion Task will not differ between depressed and matched nondepressed participants within each sex. Hypothesis G1: The proportion of depressed women classified as having a relational-connected self (RCS)according to the stories elicited with the TAT cards and the Story Completion Task, will be smaller than the proportion of matched non- depressed women classified in this manner. Hypothesis G2: The proportion of men classified as RCS will not differ between depressed and matched nondepressed samples. 28 Chapter II METHOD Participants Undergraduate students were recruited at Michigan State University, from students at introductory Psychology classes. After eliminating a few incomplete pro- tocols, a total of 726 students remained as subjects, of which 437 were women and 289 were men. All the students who participated in the study received course credit in their introductory Psychology course as stipulated by the Psychology Department. Participants ages ranged from 17 to 47 years of age, with a mean age of 19 years and a mode of 18. The majority of the students were white (88%), born in the United States of America (98%), Christian (79%) and raised in an urban area (88%), in an average family size of four. Table 1 contains a detailed description of the subjects. Instruments A questionnaire asked for demographic information, country where person grew up, primary care-giver, and five sentence completion items (see Appendix A-l). The Depression Adjectives Check Lists (Lubin, 1967) was originally de- veloped to measure transient depressive mood or feeling. It consists of seven lists of adjectives found to discriminate between depressed and non- depressed individuals 29 Table 1 Participants’ Demographic Information . Percentage Demographic Variables Women (n=437) Hen (n=289) Age 17-19 75.7 63.7 20-25 23.6 33.2 Over 25 .5 2.1 Not Aswered .2 1.0 Race White 85.4 93.1 Black 9.2 2.4 Asian 3.9 3.1 Hispanic .9 .3 Native-American .5 .3 Not Answered .2 .7 Type of Religion Christian 81.5 76.5 Judaism 4.3 6.9 Eastern 1.4 1.0 Agnostic and Atheism 12.4 13.1 Not Answered 5 2.4 Country where Raised U.S.A. 97.5 97.6 Other 2.3 2.4 Not Answered 2 . Type of Area where Raised Urban 66.6 67.5 Rural 10.3 12.5 Not Answered .9 .0 Size of Family of Origin 2 1.1 1.7 3 11.9 10.4 4 79.2 84.1 5 6.9 2.8 6 5 1.0 Not Answered 5 .0 30 (Lubin, 1965). Split-half reliabilities range from .82 to .93 (Lubin, 1965). Form E was used in the present study. The Thematic Apperception Test (TAT) developed by Murray (1971) consists of 20 cards with pictures of ambiguous human situations. Subjects are in- structed by the tester to tell stories about the pictures. People tend to relate stories to TAT pictures where significant others are involved with each other (Murstein, 1972). Neugarten and Gutmann (1958) used a thematic card approach successfully to ex- plore people’s perceptions of sex roles and perceptions of typical interactions as seen by each sex. Depressive preocupations have also been observed in self-report as well as in fantasy material (Kovacs & Beck, 1985). The TAT was included in this study for two reasons. First, as a projective measure of the sense-of-self, the responses obtained with it were expected to be less censored by the participant and a better representation of the participants’ values (Neugarten & Gutmann, 1958). Second, the TAT specifically elicits the content and dynamics of the participants’ interpersonal relationships (Bellak, 1975). In the present study six pictures from the TAT were used for each sex. Women were asked to respond to cards: (3GF) - Woman standing in doorway, face covered with right hand. (4) - A man and a woman, she is clutching the man’s shoulders and his body is turning away from her. (6GF) - A woman seated looks at a man with a pipe standing behind her. 31 (7GF) - A woman and a girl seated on a sofa. (10) - A man and a woman embracing. (16) - Blank. Men were asked to respond to cards: (3BM) - A person on the floor, huddled against a couch. (4) - Same as 4 for women. (6BM) - An older woman standing with her back toward a young man. (7BM) - An older man looking at a young man. (10) - Same as 10 for women. ( 16) - Blank. The cards were photocopied in their original size, one per page, centered. Booklets were prepared for each student to use. The first page of the booklet in- structed the student to write a story for each picture (see Appendix A-2). Pictures appeared in the original sequence: 3,4,6,7,10 and 16. Individual booklets allowed for administration in large groups that included women and men. This arrangement was judged to be similar to normal classroom experience. The Story Completion Task had two short descriptions of situations where there was either an in-relationship conflict or an out-of-relationship conflict (see Ap- pendix A-3). Participants were intructed to complete the following stories: 1) Bob and Ann have been dating for six months. Today they had an argument - . . 32 2) Fred/Susan has worked very hard to get a scholarship. Today a letter came stating that his/her work was not good enough to merit the scholarship This exercise was included as a supplement of the TAT, to make sure that participants would address each type of situations. The Revised Relationship Self Inventory (Reinhart, Pearson, Strommen, Donelson, Barnes, Blank, Cebollero, Cornwell & Kamptner, 1987) was developed orig- inally to measure styles of relational self, based on the developmental stages proposed by Gilligan (1982). It consists of four scales: Separate/ Objective Self (S/ OS), Rela- tional/ Connected Self (R/ CS), Primacy- of-Other Care (POC), and Self and Other Care Chosen Freely (SOCCF). These scales were found to be reliable (alpha coeffi- cients ranged from .67 to .85) on a sample of 930 women and 228 men (E.A. Strommen, personal communication, 1987). It has 60 items which subjects rate on a five point scale ranging from ”Not like me at all” to ”Very much like me”. The following list gives two examples of items from each of the RRSI’s scales. From S/ OS: 3. I cannot choose to help someone else if it will hinder my self- development. 6. I find it hard to sympathize with people whose misfortunes I believe are due mainly to their shortcomings. 33 From R/ CS: 20. Being unselfish with others is a way I make myself happy. 27. Relationships are a central part of my identity. From POC: 1. I often try to act on the belief that self—interest is one of the worst problems facing society. 7. I try to curb my anger for fear of hurting others. From SOCCF: 4. I want to be responsible for myself. 24. In order to continue a relationship it has to let both of us grow. The long form of the Beck Depression Inventory (Beck, 1972) consists of 21 items, each with four response alternatives. Subjects are asked to ”pick out the statement in each group which best describes the way you have been feeling the past week, including today” (Beck, 1972). In a meta-analysis of research on the properties of the Beck Depression Inventory (BDI), Beck et a1. (1988) found mean coefficients alpha of .81 (ranging from .76 to .95) for nonpatients and .86 (ranging from .73 to .92) for psychiatric patients. An adapted version of the Hassles and Uplifts Scale (Delongis, Folkman 81. Lazarus, 1988) was used. The original scale was developed to measure stressful events happening in a day. It had 53 items, and asked respondents to rate how much 34 of a hassle and how much of an uplift each item was for them that day. The adapted version (AHUS) that was used in the present study asked respondents to rate how much of a hassle and an uplift each item was during the past week. Many items were eliminated because they were not relevant to the present study, and some items were adapted to make them relevant for undergraduate students. This version contained 18 items. Nine of the items belonged to the Relationship scale (items: 1, 2, 3, 4, 5, 6, 7, 17 and 18). Three examples of the type of items in this scale are: 1. Your relatives. 3. Time spent with your family. 5. Intimacy. The other nine items belonged to the Out-of-Relationship scale (items: 8, 9, 10, 11, 12, 13, 14, 15, and 16). Three examples from this scale are: 10. Your work load or academic load. 12. Enough money for necessities. 14. Exercise. Procedure Data collection was conducted during Spring, Summer and Fall terms of 1988, and Fall semester of 1992. No differences in means scores on the BDI or the DACL were found for the different years (see Appendix B-l). Instruments were administered in groups of varying sizes, ranging approx- 35 imately from 20 to 150 students. All sessions were conducted in the same way and lasted about two hours. Each student was handed a packet containing all the instru- ments (including a booklet with the TAT pictures). Students were instructed by the experimenter to fill out the questionnaires in the sequence in which they appeared. The experimenter told students when to begin and when to stop working on the first half of the session. First students were asked to fill out demographic information, a sentence completion exercise, and the DACL. This section took about ten minutues to complete. The experimenter then asked students to turn to the TAT section of their materials and ”make up a story for each picture... Write what has led up to the event shown in the picture, what is happening at the moment, what the characters are feel- ing and thinking; and then give the outcome. You can make up any kind of story... You will have five minutes for each story...” (see Appendix A for complete instruc- tions). After five minutes, the experimenter instructed students to stop writing and turn to the next page. Five minutes has been found to be the optimal time to obtain written stories for TAT cards similar to those that are obtained verbally (Murstein, 1963) The instructions were repeated for the other pictures in the following manner ”Now make up a story for the next picture. Remember to write what led up to the event shown, what is happening then and how it ends. Also, what the charac- ters are feeling and thinking.” 36 The experimenter gave the following instructions for the last story of this section: ” As you can see, there is no picture on this page. I want you to imagine a picture on this page. Then write a description of the picture you imagined, and then make up a story to go with it. Write it down as you go making it up. You will have five minutes to do this. You may begin now.” Then the students were asked to turn to the story completion task. The experimenter explained to the students that ”now you will write two more stories. In- stead of pictures, you will find a couple of sentences describing new situations. Please complete the stories, you will have five minutes per story... Remember to include how the characters are feeling and what they are thinking. Also include how the story ends.” (see Appendices A for complete instructions). A 10 minute break was offered at this point in every session, but all groups turned it down, electing to continue. Then students were instructed to fill out the rest of the questionnaires at their own pace. The questionnaires appeared in the follow- ing order: the RRSI, the BDI and the AHUS. These tasks took about 30 minutes to complete. To provide the students with a learning experience a short explanation about the symptoms of depression was given by the experimenter, and a handout was distributed with information of where students could get help for depressive symp- toms, as well as intructions about how to get a summary of the results of this study (see Appendix A-7). Students also received credit for their participation as stipulated 37 by the Department of Psychology. Analysis of Data Beck Depression Inventory and Depression Adjective Check List Form E. For each student, scores were obtained for the BDI and Form E of the DACL following the standard way of scoring each inventory. The distributions ob- tained with each instrument (see Appendices B-2 and B-3), as well as the correlation between the two instruments (Product-Moment correlation = .45, p < .001), were consistent with previous results in the literature (Appendix B-4 details correlations obtained). Table 2 shows the means and range of scores on the BDI and the DACL- Form E. Selection of Participants for the Depressed and Nondepressed Samples Four subgroups were selected based on sex of subject and scores on these two measures. Participants were judged to be depressed if they scored 16 or higher on both inventories, to eliminate false positives from each inventory. A score of 16 on the BDI is considered to reflect a mild-to— moderate depression in the ”guidelines for use” that accompanies the test. In Form E of the DACL, a score of 16 is more than one standard deviation from the mean in non-psychiatric populations (Lubin, 1981). In the current study, a score of 16 was one standard deviation above the mean on each of the instruments. About 7% of those tested (6.64% of the women and 6.57% of 38 Thbke2 Means, Standard Deviations and Range of Scores on the Beck Depression Inventory and Form E of the Depression Adjective Check Lists. Inventory/ Sample a u. an. Range 601 Whole Sample 726 8.94 .64 to 38 Women 437 9.65 .72 to 38 Men 289 7.86 .38 to 30 DACL-Form E Whole Sample 726 10.12 .01 to 30 Women 437 10.14 .96 to 30 Men 289 10.10 .08 to 27 39 the men) had elevated scores on both tests. These participants were the matched by sex, age, religion, country in which they were raised and type of area in which they grew-up with participants whose BDI scores were lower than 12 and DACL-Form E scores lower than 11. Always, the first person to match a depressed participant was selected for the control group. Table 3 shows the four groups established in this manner (Depressed Women, Depressed Men, Matched Nondepressed Women, Matched Nondepressed Men) with the means on the BDI and Form E of the DACL. Table 4 provides demographic information from these groups. Adapted Hassles and Uplifts Scale On the Adapted Hassles and Uplifts Scale (AHUS), three scores were ob— tained for each subject on the whole sample. Endorsements on items from 1 through 7, 17 and 18 were added to determine the Relationship score. Endorsements on items from 8 to 16 were added to obtain the Out-of— Relationship score. Scores on both variables had a possible range of 0 to 54. The difference between these two scores was also obtained. More detailed scores were obtained by specifying how much of each score came from the Hassles Column and how much came from the Uplift Column, for the depressed and matched nondepressed samples. 40 Table 3 Mean Scores on the Beck Depression Inventory and Form E of the Depression Adjective Check Lists for Depressed and Matched N ondepressed Samples . BDI DACL Sample n u. an. H. 3D Depressed Women 29 23.7 5.9 20.8 3.5 Men 19 21.8 4.6 19.7 3.2 Women and Men 48 22.9 5.4 20.4 3.4 Hatched Nondepressed Women 29 5 8 3 0 5.2 2 9 Men 19 4 3 2.9 5 2 2 3 Women and Men 48 5 2 3 0 5.2 2 6 41 Table 4 Demographic Information from the Depressed and Matched N ondepressed SamPles . Janen— __11en_ Demographic Variables D N D N Age 17-19 13 13 15 15 20-25 6 6 3 3 Over 25 0 0 1 1 Race White 25 25 19 19 Black 1 1 0 0 Asian 3 3 0 0 Type of Religion Christian 20 20 12 12 Judaism 1 1 2 2 Agnostic and Atheism 8 8 5 5 Country where Raised U.S.A. 29 29 18 18 Other 0 0 1 1 Type of Area where Raised Urban 26 27 17 17 Rural 2 2 2 2 Not Answered 1 0 0 0 Size of Family of Origin 2 0 0 1 0 3 4 5 1 3 4 19 23 15 16 5 or more 6 1 2 0 Note: For each sample of women, n:29. For each sample of men, n;19. 42 Revised Relationship Self Inventory On the Revised Relationship Self Inventory (RRSI), average item scale scores were obtained for the four scales. Responses to each item ranged from 1-” Not like me at all” to 5-”Very much like me”. Table 5 shows the correlations among the four scales of the RRSI for women and men. Because the Self/Other-Care-Chosen- Freely (S/OCCF) scale was positively and significantly correlated to both Sepa- rate/ Objective Self and Relational/ Connected Self scales, it was excluded from the analysis. S/OCCF was expected to correlate with these two scales as it attempts to identify people whose self definitions include characteristics of both types of selves. An average score of 3.5 or above on a scale defined a scale as self- de- scriptive for a participant. An average score of less than 3.5 on a scale was considered to indicate that the participant did not find the scale self— descriptive. This cut-off point has been used before with the RRSI (Pearson et al., 1985; Reinhart et al., 1985). Stories Elicited with Thematic Apperception Test and Story Completion Task The stories were used in two different ways. First, the responses of the four smaller samples to the Thematic Apperception Test Cards (TAT) and the Story Completion Task were scored for the presence or absence of characteristics that would describe a separate-objective self (SOS) and a relational self (RS). The scoring scheme was developed for this research. It was based on the theoretical descriptions in the literature and the items of the RRSI. The Scoring Scheme as well as the Scoring Sheet 43 Table 5 Correlations Between Scales of the Revised Relationship Self Inventory for 437 Women (above diagonal) and 289 Men (below diagonal). Scalesa S/OS SOCCF R/CS POC 8/08 .21* -.28* -.12* SOCCF .27* .43* .00 R/CS -.26* .42* .39s POC -.16* .07 .42* a Note: S/OS= Separate/Objective Self, SOCCF= Self/Other Care Chosen Freely, R/CS= Relational/Connected Self, and POC: Primacy of Other Care. * n < .05, based on two tailed tests. 44 can be found in Appendix C. Two coders were trained by the researcher on the scoring of responses using protocols collected from the larger sample of students but that were not se- lected for any of the smaller groups. Both coders were psychology graduate students, who were paid for their time. High levels of agreements were obtained between each coder and the researcher (87% and 92%) and between the coders (85% agreements for elements, 88% for areas) before proceeding to score the selected responses. Both coders were blind as to the purpose of the study and how the protocols were classi- fied. Instructions for scoring the protocols can be found in Appendix C. The level of inter—coder agreement for actual protocols was very good (81% for elements, and 84% for areas). After coders scored each protocol for particular elements, each major area was scored a 1 if there was at least one element present for that area in a story, and 0 if there were none present. These points were then added to obtain two scores per protocol, a score for SOS and a score for RS. The area of Help did not differentiate among subjects and thus was eliminated (see Appendix D). Scores on SOS and RS had a possible range of 0 to 24. For example: participant number 4 (P4) had a score of 11 on SOS and a score of 9 on RS, while participant number 16 (P16) had a score of 22 on SOS and a score of 4 on RS. To equate the means of both distributions and allow a better comparison of scores for each subject, the score on RS was adjusted as follows: 45 RS + 6.69 = RCS, where 6.69 was the difference between the means of SOS and DS from the matched nondepressed sample. In the previous example, the score of 9 obtained by P4 on RS was changed by adding 6.69 and became 15.69, the RCS score. The SOS score for P4 remained 11. For P16, RCS was 4 -+- 6.69 = 10.69, while the score on SOS remained 22. The next step was to compare the scores on SOS and RCS for each par- ticipant. The higher score of the two was said to represent that participant’s present state of relational connection, where SOS indicated a self not connected to relation- ships and RCS represented a self connected to relationships. Continuing the previous examples, P4 was said to have a RCS because the score of 15.69 on RCS was higher than the score of 11 on SOS, while P16 was classified as having a SOS because the score on SOS was higher than the score on RCS (22 > 10.69). The second way in which stories were used in this study was to count the number of words used by the participants to complete stories 7 and 8 (Story Com- pletion Task) only. Two scores were generated in this way, length of story 7 (L7) and length of story 8 (L8). The difference between L7 and L8 was also calculated. These scores were obtained for the depressed and matched nondepressed samples, and for the first 120 men and 120 women that did not belong to the samples just mentioned. 46 Chapter III RESULTS Thesis I: People develop self-definitions limited by gender, where women have connected / relational selves and men have separate/ objective selves. Hypothesis A1: The mean score for women on the Relational/ Connected Self (R/ CS) Scale of the Revised Relationship Self Inventory (RRSI) will be 3.5 or higher (scale is self-descriptive), and the mean score for men will be lower than 3.5 (scale is not self- descriptive). Hypothesis A2: The mean score for women on the Separate/ Objective Self Scale (S/ OS) of the RRSI will be lower than 3.5, and the mean score for men will be 3.5 or higher. Hypothesis A3: The mean score for women on the Primacy of Other Care Scale (POC) of the RRSI will be higher than the mean score for men. The data did not support these hypotheses. Table 6 shows the means and standard deviations obtained. For Hypothesis Al, the means of both, women (t = 31.69, p < .001, one-tailed test) and men (t = 18.29, p < .001, one-tailed test), were significantly higher than 3.5 indicating that both groups reported R/ CS as self- descriptive. For Hypothesis A2, both means, women’s (t = -36.41, p < .001, one-tailed 47 Table 6 e s d Dev' ' s ev t vised Relationship Self Inventory Scales for 437 Women and 289 Aden. Scale and Sample Kean SD Relational/ Connected Self Women 4.22 0.5 Men 4.05 0.5 Separate/ Objective Self Women 2.64 0.5 Hen 2.83 0.5 Primacy of Other Care Women 3.26 0.4 Men 3.24 0.5 48 test) and men’s (t = -22.09, p < .001, one-tailed test), were significantly smaller than 3.5 indicating that neither group reported S/OS as self-descriptive. For Hypothesis A3, the means for women and men did not differ sgnificantly (t = .51, p > .05, one- tailed test) on endorsing POC as self-descriptive. Although there were statistically significant differences between the means of women and men in the R/ CS (women’s mean = 4.22 with sd = .47; men’s mean = 4.05 with sd = .51; t = 4.46, p < .001, two-tailed test) and SOS (women’s mean = 2.64 with a sd = .49; men’s mean = 2.83 with a sd = .51; t = -5.02, p < .001, two-tailed test) scales of the RRSI, these differences were small and did not indicate true differences in self-definition as shown through the tests of hypotheses A1 and A2. The statistical significance stemmed from the large number of subjects. A look at the percentage of participants that endorsed each scale as self-descriptive (see Table 7) helps clarify this issue. Ninety-one percent of the whole sample reported R/ CS as self- descriptive and 92% reported S / OS as not self-descriptive. We can speculate that perhaps participants were responding to how socially desirable were the items. The percentages of men and women who endorsed POC did not differ, but this scale appears to be measuring something other than what would be socially desirable. A closer examination at the items of POC suggests that it may be measur- ing a denial of one’s own needs rather than a primacy of other’s needs (which implies a connected relational self) as proposed. This will be discussed in more detail in the discussion of results. 49 Table 7 Percentages of Participants Endorsing Revised Relationship Self Scales as Self-descriptive. Percentage Scale Total Women Men Separate/ Objective Self 8 5 11 Self/Other Care Chosen Freely 92 94 89 Relational/ Connected Self 91 94 85 Primacy of Other Care 32 31 33 Note: An average score of 3.5 or higher was considered self-descriptive. 50 These observations suggest that participants’ RRSI responses should not be used to infer sense of self. Thus, the RRSI will not be used to draw conclusions as to sense of self in this study. Hypothesis B: In the Adapted Hassles and Uplifts Scale (AHUS), the mean dif- ference between the Relationship subscale and the Out-of-relationship subscale will be higher than zero for women and lower than zero for men. Only half of the hypothesis is supported by the data. Table 8 shows that the mean difference for women was significantly higher than zero (t = 5.72, p < .001), indicating that women endorsed items about relationships as affecting their emotional state more than the out-of-relationship items. But for men the mean difference was not significantly different from zero (t = -0.56, p > .05), indicating that they reported being affected to the same degree by items related to relationships as by out-of-relationship items. In addition, the means differed significantly from each other (t = 4.20, p < .01, two-tailed test), showing that women displayed greater dif- ferences than men in how each subscale affected their emotional state. Examining a smaller sample of nondepressed students (who had been matched with the depressed students) in more detail (see Table 9), it becomes clear that women and men only differed in that women reported more emotional uplift from relationships than men did (t = 1.76, p < .05, one-tailed test). But these groups did not differ significantly in reported levels of hassles derived from relationships (t = 0.01), and hassles and uplifts stemming from out-of-relationship items (for hassles t Table 8 T-tests of Mean Difference Scores on the Adapted Hassles and Uplifts Scale Against a Zero Difference Hypothesis . Sample n. 1 SD. t 2 Women 437 1.78 6.52 5.72 .00 Men 289 -0.20 6.05 -0.56‘ .57 Note: Probabilities are based on two-tailed tests. Table 9 T-tests Between 29 Nondepressed Women and 19 Nondepressed Men on Hassles and Uplifts Subscales of the AHUS. __.aneL_ ..Jen__ Subscale 11 SD ll SD 1'. D. Relationship Hassles 4.59 4.19 4.58 3.72 0.01 .99 Uplifts 15.45 5.68 13.05 3.76 1.76 .08 Out-of- Relationship Hassles 10.62 4.08 8.89 7.25 0.94 .35 Uplifts 6.24 4.34 7.05 4.10 -0.65 .52 Note: Probabilities are based on two-tailed tests. 52 = 0.94, for uplifts t = -0.65). In general, students reported deriving more uplifts from relationships than from the out-of-relationship items (for women: t = 9.94; for men: t = 4.37; p < .001 with two-tailed tests), and more hassles from the out-of- relationship items than from the relationship items (for women: t = 6.69, p < .001; for men: t = 3.38, p < .01; two-tailed tests). Hypothesis C: The mean difference between the number of words used to com- plete a story about a difficulty in a relationship and the number used in completing a story about a difficulty in achievement on the Story Completion Task will be higher than zero for women and lower than zero for men. Again, the hypothesis was supported only partially. Women wrote longer stories when the issue addressed was that of a relationship rather than achievement (see Table 10), while men wrote equally long stories for both issues. In addition, the mean difference for women was significantly higher (t = 4.01, p < .01, one-tailed test) than the one for men. This difference in length was not an artifact of women writing on average longer stories than men. When using the relative difference: (L7- L8) / (L7+L8), where L7 is the number of words used in the relationship story and L8 is the number of words used in the achievement story, to eliminate the influence of the length of the story on the size of the difference, the means of the difference remained significantly different between women and men (Mean of difference and standard deviation for women = .09(.18), mean of difference and standard deviation for men = .02(.16), Student’t = 53 Table 10 T-tests of Mean Difference in Length of Stories from the Story Completion Tas Against a Zero Difference Hypothesis . Sample :1 ll SD 11 9. Women 120 17.70 31.69 6.12 .01 Men 120 3.00 24.63 1.33 . 10 Note: All probabilities based on one-tailed tests. 54 3.33, p < .001 based on a two-tailed test). Hypothesis D: The proportion of women classified as relational-connected self on the stories elicited with the TAT and the Story Completion Task will be higher than the proportion of men classified in this way. This hypothesis was strongly supported by the data. Based on the quan- tity of elements of relational versus separate self present in the stories, significantly more women were classified as exhibiting a relational- connected self than men (pro- portion of women = .62, proportion of men = .26, Binomial Nonparametric Test p < .001, based on a two-tailed test). Individually both of these proportions were significantly different from a chance proportion of .50 (p < .05, based on a two-tailed test). Thesis II Depressive experience in women is associated with a loss of a rela- tional sense-of-self. Depressive experience in men is not associated with a loss of relational sense-of-self or of a sense-of-self separate from others. Hypothesis E: In the Adapted Hassles and Uplifts Scale (AHUS), mean dif- ference scores between the subscales: Relationship and Out-of-Relationship will not differ between depressed and matched nondepressed participants within each sex. Table 11 shows that in the case of women, the data support the rejection of this null hypothesis. Depressed women did differ from matched nondepressed women 55 on how they endorsed items from the Relationship and the Out-of-Relationship sub- scales (t = -3.60, p < .001, two—tailed test). To clarify this finding, Table 12 presents the same data by showing endorsements of type of effect for each subscale. Although the overall impact of re- lationships in the women’s emotional state was the same for depressed and matched nondepressed women, the quality of the impact was different. Depressed women re— ported relationships as being more of a hassle (t = 3.62, p < .001, two-tailed test) and less of an uplift (t = -2.58, p < .05, two—tailed test) than matched nondepressed women did. For out-of- relationship items, the only difference was that depressed women reported these items as much more of a hassle than the matched nonde- pressed women (t = 4.87, p < .001, two—tailed test). The level of uplifts derived from these items did not differed between depressed and nondepressed women (t = 0.51, p > .05, two-tailed test). To summarize, depression had an effect in the quality of impact that relationships have on women’s emotional state, increasing in negative and decreasing in positive impact. But depression had an effect only on the negative impact of out-of-relationship issues, which increased. For men, the data did not support the rejection of Hypothesis E. There was no overall significant difference between depressed men and non-depressed men on how much they were affected by items from the Relationship and the Out-of- Relationship subscales (see Table 11), but a more detailed look at the data indicated that there were differences between the two groups of men. Table 13 shows how each 56 Table 11 Paired t-tests Between Mean Difference Scores of Dgaressed and Matched Nondepressed Participants on the Adapted Hassles and Uplifts Scale. Sample I SD. d1 L. 2 Women Depressed -1.93 6.15 28 -3.60 .01 Nondepressed 3.55 5.73 Hen Depressed -2.90 6.60 18 -1.76 .10 Nondepressed 1.53 8.53 ~ Note: Probabilities are based on two-tailed tests. 57 Table 12 Paired t-tests Between Mean Scores from 29 Depressed and 29 Matched Nondepressed Women on Hassles and Uplifts b esof be H Subscales 11 SD ll SD 1'. D. Relationship Hassles 8.07 3.61 4.41 4.10 3.62 .01 Uplifts 12.45 5.58 18.03 5.54 -2.58 .01 Out-of— Relationship Hassles 15.86 4.49 10.55 4.08 4.87 .00 Uplifts 6.86 5.05 6.21 4.33 0.51 .62 Note: a: = 28 for all tests. All probabilities are based on two-tailed tests. 58 group reported the kind of effect the items had on them. In the Relationship sub- scale, depressed men differed significantly from matched nondepressed men only in that depressed men reported experiencing less uplift (t = -2.21, p < .05, two-tailed test) than the matched nondepressed men. For the Out-of-Relationship subscale, the quality of impact was different between the two groups of men. Depressed men re- ported less uplift (t = -2.33, p < .05, two-tailed test) and more hassle (t = 1.66, p > .05, two-tailed test) than matched nondepressed men. Although the last differ- ence did not reach statistical significance, it is very close to being significant with a one-tailed test (p < .06). These results indicate that depression may have affected the quality of the impact from Out-of-Relationship items (most of which are related to performance in school or work), where depressed men derived less uplift from these items and showed a tendency to experience the items as more of a hassle than the matched nondepressed men. For the Relationship items, the only difference between the two groups of men was that depressed men derived less uplift from Relationship than nondepressed men. Hypothesis F: The mean difference between the number of words used to com- plete a story about a difficulty in a relationship and the number used in completing a story about a difficulty in achievement on the Story Completion Task will not differ between depressed and matched nondepressed participants within each sex. As Table 14 shows, the data did not provide evidence to reject this null hypothesis, as predicted by the theory. There was no difference on length of sto- 59 Table 13 Paired t-tests Between Mean Scores from 19 Depressed and 19 Matched Nondepressed Men on Hassles and Uplifts Subscales of M- Subscales I. SD. H. SD. 1. a Relationship Hassles 5.79 3.19 4.58 3.72 1.19 .25 Uplifts 8.84 6.29 13.05 3.76 -2.21 .04 Out-of- Relationship Hassles 12.79 6.68 8.89 7.25 1.66 .11 Uplifts 4.53 3.40 7.05 4.10 —2.33 .03 Note: d1 = 18 for all tests. All probabilities are based on two-tailed tests. 60 ries between depressed and matched nondepressed participants within each sex. De- pressed women wrote longer stories when addressing relationships than when address- ing achievement, just as nondepressed women did (t = -0.72, p > .05, two-tailed test). Men wrote equally long stories to both issues, regardless of whether they belonged to the depressed or nondepressed group (t = 0.17, p > .05, two-tailed test). Again, using the relative difference: (L7-L8)/(L7+L8), where L7 stands for number of words used to complete the story about a relationship and L8 stands for number of words used to complete the story about achievement, did not alter the results (see Table 15), demonstrating that the results obtained were not an artifact of women’s longer stories. Hypothesis G1: The proportion of depressed women classified as having a relational-connected self (RCS) according to the stories elicited with the TAT cards and the Story Completion Task, will be smaller than the proportion of matched non- depressed women classified in this manner. Hypothesis G2: The proportion of men classified as RCS will not differ between depressed and matched nondepressed samples. The data strongly supported these hypotheses. Using the Binomial Non- parametric Test to compare the proportion of depressed women evidencing RCS (.31) with the proportion of matched nondepressed women evidencing RCS (.62), the pro- portion of depressed women was found to be significantly lower (Z-approximation, p < .001, one-tailed test) than the proportion of matched nondepressed women. The 61 Table 14 Paired t-tests Between Mean Differences from Depressed and Matched Nondepressed Participants on the Story Completion Task . Sample I. SD. at. L. 2 Women Depressed 14.45 23.41 28 —0.72 .48 Nondepressed 18.79 36.23 Hen Depressed 3.95 23.16 18 0.17 .88 Nondepressed 2.63 28.20 Note: Probabilities are based on two-tailed tests. 62 Table 15 'r d -tests etw cte ' e ces rom Depressed and Matched Nondepressed Participants on the Story Completion Task . Sample Ii SD d1 1 9. Women Depressed .08 .14 28 -0.87 .39 Nondepressed .11 .21 Hen Depressed .02 .15 18 0.08 .94 Nondepressed .02 .16 Note: Probabilities are based on two-tailed tests. 63 NU‘U‘O’VQ 8 . éDepressed 7 éDepressed - 4 — 15— — L, o 5 10 15 20 25 o 5 10 15 20 25 sos WOMEN sos MEN . 8 . éMotched 7 éMotched Z— 6 - :— 4E— =_ ,2. E. rilruiliiiilii iliii 0 1111114 llgllllIl l-ll-li A o 5 10 15 20 25 o 5 10 15 20 25 505 WOMEN sos MEN Figure 2 Distribution of Scores on Separate-Objective Self (SOS) for 29 Depressed Women, 29 Matched Nondepressed Women, 19 Depressed Men, and 19 Matched Nondepressed Men. 65 proportion of men did not differ (Exact Binomial, p > .05, one-tailed test) for de- pressed (.22) and matched nondepressed (.26) men. Thus in depression, a loss of relational self is only relevant for women. This interpretation of the results is supported by the observation that depressed women fell into SOS category more often than nondepressed women, not because their scores on SOS were higher (t = 1.47, p > .05), but because their scores on RCS were lower (t = ~2.12, p < .05). The distribution of depressed women on these two variables is similar to that of nondepressed men (see Figures 2 and 3). Depressed men did not differ from nondepressed men in number of RCS elements included in the stories (t = - 1.29, p > .05), but they did have more SOS elements than the nondepressed men (t = 2.97, p < .05). 64 NU‘U‘O’QW NUDU‘O’VGJ 8 . iDepressed 7 éDepressed E- 5 E- :— 4 :— :— 3 E- r 2 g" :— 1 :- H 7 El 1 AJ 1 I —] ILL! 11 -] l [1 I] l C) :1 1 ll 1 11,1 I 1 ll 1 11,17 1 11 O 5 10 15 20 25 O 5 1O 15 20 25 SOS WOMEN SOS MEN Matched Motched NUDU‘O’IVQ 'UI'UIIT‘U'IIIU‘UIWI'I'T'I'I'YITTUI C) IVIII'VIY‘U'II'YITTTVV'UIVU'TTTUI'IVUIU ’ ’ [III 11 1J,l 11 ll lllLl lJ,l 1 11 [A C) :11 IL I] l 11.11 ll [1 J 1 A 5 10 15 20 25 O 5 10 15 20 25 SOS WOMEN SOS MEN Figure 2 Distribution of Scores on Separate-Objective Self (SOS) for 29 Depressed Women, 29 Matched Nondepressed Women, 19 Depressed Men, and 19 Matched Nondepressed Men. 65 NUJIU'GVQ a—l éDepressed I: » 7141111111111 0 5 10 15 20 25 RSWOMEN iMotched llllllll 11 “1111511 0 5 10 15 20 25 RSWOMEN NUbU‘O’Vm A Depressed .1 I VIII U‘II IUU'IVUVI'VU'UIVII‘III'" ll ll Hillllllllllll Figure 3 Distribution of Scores on Relational Self (RS) for 29 Depressed Women, 29 Matched Nondepressed Women, 19 Depressed Men, and 19 Matched Nondepressed Men. 0 5 1O 15 20 25 RS MEN E_Motched 7111111 I-lrlillllilllL O 5 1O 15 20 25 RS MEN' Chapter IV POST-HOC FINDINGS Primacy-of-Other Care Scale One RRSI scale provides some more information on the dynamics involved in depression for women and men. The Primacy-of-Other Care (POC) scale of the RRSI was originally designed and used to identify people who place the needs of others above their own to the point that self-gratification is not considered necessary. About 30% of the women and the men in this study endorsed this scale as representative of themselves. POC was found to correlate with depression for women (n = 437, POC with BDI: r = .24, with DACL: r = .12, each p < .001, one-tailed tests). but not for men (n = 289, POC with BDI: r = .09, with DACL: r = .10, each p > .05, one-tailed tests). The group of depressed women was the only one with a large proportion of high POCs (depressed men had .37, matched nondepressed men had .21, and matched nondepressed women had .24, while the depressed women had .66). If POC had been measuring a consideration for the feelings of others the following correlations with the variables from the projective stories would have been obtained: a positive correla- tion with the the sum of relational- connected self ’8 elements (r = .05, p > .05), and a negative correlation with the sum of separate-objective self ’s elements (r = -.11, p > .05). But POC did not correlate with these variables. A closer examination of 67 POC items suggests that it may be measuring a denial of one’s own needs more than a primacy of the needs of others. POC did have a moderate negative correlation (r = -.23, p < .01) with one element from the projective stories SOS which measured making decisions based on the advancement of own needs or moral rules. The denial of own needs may be more a reflection of low self- esteem (believing one does not deserve) than of a relational-self in this case. Of the 19 depressed women that had endorsed POC as self-descriptive, 13 were characterized as SOS based on their projective stories. POC items were phrased in terms of placing others needs first, but there was no evidence of this in the projec- tive stories. Violent Content in the Story Completion Task While analyzing the data for this study, an interesting difference was noted on the propensity to include violent acts in the projective stories. Since the use of violence might indicate a lack of empathy (to others when the violence is directed at others, or to the self when it is directed to the self as in suicide), the number of blatant violent acts were counted for the two incomplete stories for the following groups: depressed men, depressed women, matched nondepressed men, matched non- depressed women, and the first 120 men and 120 women not included in the previous groups (the later two groups are the same used to calculate length of stories earlier). Blatant violent acts were defined as unequivocal descriptions of purposeful physical 68 aggression, attempted and completed murders and suicides, purposeful killing of an- imals (included as other violence), bombings, etc. Table 16 shows the percentage of women and men that included violent acts in their stories. The data clearly shows that men are more likely than women to include such acts in their stories (21% of men and 6% of women). Further, most of the violence described by women were suicides (78%), while men were as likely to describe incidents of violence directed at others as to themselves. It is interesting to note on Table 17 that men described suicides more in response to the achievement story than to the relationship story, but were equally likely to describe murders on both types of stories. Also men were the only ones who described combinations of murder-suicide, or being killed as a consequence of going on a murder rampage in the stories. 69 Table 16 Percentage of Women and Men that Included Violent Content in the Story Completion Task . __!nmen___. .Hen Sample a l n. l Nondepressed 120 6 120 21 Hatched Nondepressed 29 3 19 16 Depressed 29 3 19 37 Table 17 Murders and Suicides Included by Women and Men inpa Story about a Difficulty in a Relationship and in a Story about a Difficulty in Achievement . __Relatinnshin__ ..Anhiaxament. Sample Murder Suicide Murder Suicide Women Nondepressed 2 1 0 6 (n:120) Hatched Nondepressed 0 0 0 0 (n:29) Depressed 1 0 0 0 (n=29) Hen Nondepressed 6 2 7 11 (n=120) Hatched Nondepressed 1 0 3 1 .05), while the women reported more impact from the relationship items than from the out-of-relationship items (t = 5.72, p < .001). This pattern of results agreed with the findings obtained with a direct pro- jective measure. Women used more words to complete a story about a fight between a couple than to another incomplete story about failure to earn a scholarship for college 72 (t = 6.12, p < .01). But men gave the same amount of attention to both stories (t = 1.33, p > .05). Again there was a gender difference (t = 4.01, p < .01), where women reacted differently when a relationship was addressed than when another potential source of esteem was addressed, in this case academics, and men reacted equally to the two issues. Previously it was speculated that for men relationships represent a prize received for achievement, meaning that men define themselves in terms of a self-in- advancement, while for women relationships represent not just a prize but the way in which they define themselves (Kaplan, 1986). The results from this study support this conceptualization. Although men report relationships as important contributors to their emotional state, they do not use it as the primary source of self-definition as women do. The results were also very consistent in providing support for the thesis that a loss of relational-self is experienced by depressed women but not by depressed men. Depressed women displayed a different pattern of how relationships impact their emotional lives as compared to nondepressed women, where depressed women derived more hassle (t = -3.39, p < .001, two tailed test) and less uplift (t = 2.03, p < .05) from relationships than nondepressed women. While the impact of Out-of—relationship issues on depressed and nondepressed women, differed only in that depressed women experienced more hassle (t = -4.65, p < .001) from these issues than nondepressed women. These results suggest that depressed women perceive an inability to handle 73 relationships successfully. For depressed men a different pattern of responses was observed. Out-of- relationship items were experienced as a higher source of hassles (t =-1.72, p < .05, one tailed test) and of less uplifts (t =2.70, p < .05, one tailed test) by depressed men as compared to nondepressed men. The impact of relationship items differed between the depressed and nondepressed men in being considered a lower source of derived pleasure by the depressed men (t = 2.50, p < .01, one tailed test). Here the results suggest that the presence of depression in men creates in them a perceived inability to handle out-of-relationship issues. By the direct projective measure, depressed women showed the same dif- ference in attention to relationships and achievement as nondepressed women (t = - 54, p > .05). Whether they were depressed or not, women were more interested in relationships than in achievement. Depressed men were equally interested in both, as nondepressed men had been (t = -.26, p > .05). Results from the TAT and incomplete stories provided strong support for a. gender difference in how depression affects the sense-of-self. Depressed women were mostly characterized by a separate-objective self (Proportion of SOS =.69), not the relational-connected self found earlier for nondepressed women (Proportion of non- depressed women with RCS =.62, proportion of depressed women with RCS = .31, Binomial Nonparametric Test p < .001). Depressed women fell into the SOS category more often than the nondepressed women because depressed women demonstrated 74 fewer elements of RCS (t = -2.12, p < .05), while maintaining the level of SOS el- ements that nondepressed women had (t = 1.47, p > .05). In fact, the distribution of depressed women on the RCS and SOS variables was similar to the one found for nondepressed men (see Figure 2 in the Results’ section). Thus, women experience a loss of relational self in depression, that was not there to begin with in men. Depressed men did not demonstrate a difference from their matched non- depressed counterparts in type of self exhibited, being characterized by a separate- objective self (Proportion of depressed men/ SOS = .78, proportion of nondepressed Jnen/ SOS = .74, Binomial Nonparametric Test p > .05). But depressed men demon- strated an exaggerated SOS as compared to nondepressed men (t = 2.97, p < .01), Vvhile their RCS remained stable (t = -1.29, p > .05). That depressed women experience a loss of relational self is supported by other studies that had used self-report measures. Using a sample of older women, Essex, Klein, Lohr and Benjamin (1985) found that the more depressed a woman was, t he more she experienced her intimate relationships as less friendly, less consistent and predictable, and as ones in which her feelings were not reciprocated. Earlier, Weissman and Paykel (1974) had found that depressed, as compared to nondepressed, women reported more difficulty in their ability to communicate with their spouses. Also, Brown et al. (1975) had found in a community survey that level of intimacy in the relationship with their mate, and the mate’s emotional support of the woman’s role played a significant part in preventing the development of depression during stressful 75 periods. Another study found that psychological well-being in housewives was related to having close social contacts and having husbands that valued their contributions (Briscoe, 1982). In a college sample, Chevron, Quinlan and Blatt (1978) had found an association between depressed affect and women describing themselves as less warm and less expressive; and men describing themselves as less competent. Because all of these studies were based on self-report, the question had remained as to whether the afflicted person’s perception was being influenced by social expectations, leading to appropriate parsimonious explanations such as Chevron et al. (1978), who suggested ”persons more prone to depression express their lowered self-esteem by describing themselves as deviating from the prescribed sex role” (p. 683). By using thematic apperception techniques, this study was able to establish that for depressed women there is a loss of sense of relational-self, not only in awareness, but also as projected in the stories, that does not occur for men. Some of the results of this study can be used to begin addressing the question of whether depression in women represents an exaggeration of the condition of being female in this culture, as Kaplan (1986) suggested, or perhaps a reversal of the norm in the culture. Having a relational- connected self, as demonstrated by the inclusion of empathy, attempts at maintaining relationships and attention to the needs of others in the projective stories, was not a characteristic of depressed women 0r men. A lesser relational capacity was observed for depressed women as compared to nondepressed women, which was not observed for men. As explained earlier, results 76 from this study would point to a reversal of the norm in depression for women and an exaggeration for men. Now that it has been established that type of self and gender are signif- icant variables in the understanding of depression, the next issue to address would be whether the differences between depressed and nondepressed participants within each sex are a predisposing factor to depression, or a consequence of the depression. To answer such a question would require a longitudinal study of depressed and non- depressed persons. Unfortunately, the RRSI failed to identify type of self in this study, as :most participants endorsed two of the scales as self-descriptive and rejected another of the scales. The development of a self-report measure would simplify testing consid- erably and thus would be worthwhile to pursue. One possible way to use self-report would be to present specific situations where the participant has to choose a response most like 3 / he would act or decide an issue. The Primacy-of-Other Care (POC) scale of the RRSI was originally de- signed and used to identify people who place others needs above their own to the point that self-gratification is not considered necessary. About 30% of the women and the men in this study endorsed this scale as representative of themselves. POC was found to correlate with depression for women (with BDI: r = .24, with DACL: r = .12, both P < .001, one-tailed test). but not for men (with BDI: r = .09, with DACL: r = .10, both p > .05, one-tailed test). The group of depressed women was the only one with a 77 significant proportion of high POCs (depressed men had .37, matched nondepressed men had .21, and matched nondepressed women had .24, while the depressed women had .66). If POC had been measuring a consideration for the feelings of others, a pos- itive correlation with the the sum of relational-self elements from the pro jectives, and a negative correlation with the sum of separate-objective self elements would have been obtained. But POC did not correlate with these variables. A closer examination of items from the POC scale suggests that it may be measuring a denial of one’s own needs more than a primacy of the needs of others. POC did have a moderate negative correlation with one of the elements of SOS from the projective stories which measured making decisions based on the advancement of own needs or moral rules. 'Two previous studies support the above interpretation of results. Culkin and Perroto (1985) found inhibited verbal self-expression in emotional situations to be associated with depression in a sample of college women. Also, Jack and Dill (1992) found a strong correlation between depression and a scale measuring issues similar to those measured by POC. Of the 19 depressed women that had endorsed POC as self-descriptive, 13 were characterized as SOS. POC items were phrased in terms of placing others needs first, but there was no evidence of this in the projective stories. It could be that depressed women’s stories were mainly descriptions of events that did not ex- Plain feelings or reasons for actions, thus yielding lower RCS score than nondepressed Women without producing a difference in SOS scores. This, together with the finding 78 that depressed women were as interested in relationships as nondepressed women, would point to an inability to understand others most likely related to a denial of their own feelings and motivations. POC did not provide information about the dy- namics involved in depression for men as the proportion of depressed men with high POC did not differ from the proportion of nondepressed men with high POC. Oatley and Bolton (1985) have suggested that depression is experienced when a person can not continue to perform the role by which personal worth was determined and the person has no other available source of self-worth. Warren and McEachren (1985) reported that women whose main source of self- definition was interpersonal relationships reported more symptoms of depression than women who had other sources as well. Thus, in populations where women are limited to only one source of self-worth (such as homemakers), while men have access to more than one, the usual 2:1 ratio of depressed women versus depressed men would be observed. In a college population, where women and men would have access to the same two possible sources of self-definition: relationships and academic achievement, even though one may be preferred over the other, a 1:1 ratio was expected. About seven percent of the Sample was found to be depressed in this study, with almost the same rate for women as for men (6.64% of the women and 6.57% of the men). Previous studies with college Samples have reported similar findings (Hammen 81. Padesky, 1977). It can be argued that using a college sample controls for the possible cOnfounding factor of unequal access to sources of self-esteem. Thus, finding a gender difference in self-definiton in 79 this sample would not be due to a limitation in access to the sources of self-esteem. Such a difference was not considered to stem from lack of interest to one of the sources determined by gender either, as the results from this sample show that women and men report similar levels of emotional impact from both types of sources. Student-t values ranging from .01 to 1.76, were obtained on how much of a hassle and of an uplift relationship and out-of-relationship items, such as academic and job concerns were for both genders, all p > .05. Even when a 1:1 ratio is found by equating the access to sources of self-esteem, a gender difference emerges in how depression affects sense-of-self. By choosing a college sample another potential complication was hope- :fully avoided, possible biologically determined differences in attachment to children that could weight self-definition in women more strongly toward relationships than in men. College students usually do not have children yet, so it was not expected to surface in this age group. But given that Oatley and Bolton (1985) cite some evidence that having more than three children may be a vulnerability factor for depression in Women, not allowing time to develop other self-definitions, future studies of relational self and depression should control for this variable directly. This can be easily done by asking if the participant has children and how many. The approach to depression in this study was based on the assumption that depressed affect is a continuum ranging from a passing mood to an affective disorder; an approach used often in the study of depression. It does not explore pos- 80 sible etiological differences in what is considered clinical depression and less severe depressions. Weissman (1981) did not find differences in the incidence of minor de- pression and depressive personality between women and men in a longitudinal study, but did find a higher rate of women than men with Affective Disorders. The age of onset for some affective disorders is later in life than the range encompased by an undergraduate college sample. This could also account for the 1:1 ratio usually obtained in college samples. Still, a college sample can be used to draw conclusions about depressive affect and some depressive conditions. Studies including an older sample would be needed to control for the factors just mentioned. In this study two instruments were used in an attempt to eliminate false positives. To further select participants that were more likely to be experiencing a. depressive syndrome, the cut-off point was placed at one standard deviation from the mean for both instruments. While analyzing the data for this study, an interesting difference was noted on the propensity to include violent acts in the projective stories. Since the use of violence might indicate a lack of empathy (to others when the violence is directed at others, or to the self when it is directed to the self as in suicide), the number of blatant violent acts were counted for the two incomplete stories. As shown on Table 16, men were more likely than women to include such acts in their stories (21% of men and 6% of women). Further, most of the violence described by women were suicides ( 78%), while men were as likely to describe incidents of violence directed at others 81 as to themselves. It is interesting to note on Table 17 that men described suicides more in response to the achievement story than to the relationship story, but were equally likely to described murders on both types of stories. Also men were the only ones who described combinations of murder-suicide, or being killed as a consequence of going on a murder rampage in the stories. These results mirror what happens in real life where men commit most of the extreme violent acts. Of course the fact that someone describes violence in projective stories does not mean that they will actually engage in such acts. But the closer the character in aprojective story is in age, sex, occupation, etc. to the person writing the story, the more likely it is that the story will be more relevant to the writer’s real life (Karon, 1981). Thus further research on use of violence and the inclusion of violence in these stories might prove to be useful. Nevertheless, the results indicate that extreme lack of empathy was projected more often by men than by women, and that men portrayed violence for both types of stories. The number of depressed and nondepressed participants for which this variable was relevant were too small to draw any conclusion as to effect of depression on violence, but a trend toward more violence in depressed versus nondepressed men should be explored further. It would also be useful to see if the suicides portrayed predominantly to the achievement story were altruistic, in order to relieve the family of financial burden. Pollak and Gilligan (1982) had observed a higher frequency of violent imagery by men in response to pictures with a heterosexual couple than to other 82 pictures. They interpreted this to reflect a fear of intimacy, as they noted that the introduction of structure and other characters, decreased the amount of violent im- agery. In the present study, there was no structure provided, but it is interesting to note that a substantial portion of the violent imagery was directed at the story where there was only one character mentioned and where intimacy was not an issue. The use of thematic apperception techniques as a screening device for identifying potential violence (due to lack of skills at handling frustration, interpersonally or individualy) appears promising. 83 Chapter VI CONCLUSIONS The present findings support the existance of two gender based types of self, relational/ connected self and separate/objective self, as proposed by the Self-in- Relation Theory. The results also supported the notion that type of self is relevant to depression. As predicted by the Self-in- Relation Theory, depressed women demon- strated less elements of relational self than nondepressed women, without showing more elements of separate self. Also, depressed and nondepressed men differed mainly in the number of elements of separate self, where depressed men had more of this elements than nondepressed men. There was no difference in number of elements of relational self for men. Because projective techniques were used with self- report mea- sures, these patterns are not believed to be just an artifact of socially accepted ways of reporting the experience of depression. This study did not find validation for the Revised Relationship Self Inventory (RRSI) as a measure of sense-of-self, as it failed to distinguish among individuals. An expected ratio of 1:1 was obtained for depressed women vs. depressed men in this college sample. Post-hoe observations about the re— lationship between the Primacy-of-Other Care Scale from the RRSI and depression, and about the inclusion of violent content in the projective stories were discussed. 84 Li APPENDICES 85 APPENDICES A 86 Demographic Information Questionaire 87 3 1 . E |' Age: Race: Religion: Country where you grew up: __ U.S.A. __ Other: Before age 15, you lived: __ City __ Small town __ Rural area Circle those that lived in the same household as you when you were growing up: Mother Father Siblings Grandmother Grandfather Other: Which of the above spent more time nurturing you before the age of 15? (If more than one person, indicate them and estimate what percentage of nurturing they did, e.g.: mother 50% and sibling 50%) Waitress: I feel good about myself when... Something I feel proud of is... What I most enjoy is... When a friend and I disagree, I... I know someone loves me when... 88 Written Instructions for the Thematic Apperception Test 89 PICTURE STORIES Each of the following six pages has a picture on it. Make up a story for each picture and write it on the corresponding numbered sheet. Write what has led up to the event shown in the picture, what is happening at the moment, what the characters are feeling and thinking; and then give the outcome. You can make up any kind Of story. Write the story as you go making it up. If you want to change something, do not erase it, just draw a line over it. QQ_nQL_ h§g1n_until_instructed_tn. You will have 5_minutes for each story. You will be instructed when to turn the pages and begin your stories, and when it is time to stop each story. Do not worry about not having time to finish a story. Just go to the next story when asked to. Remember to include what led up to the event, what is happening then. what the characters are feeling and thinking, and what is the outcome. Story Completion Task: Instructions and Incomplete Stories 91 STORY COMPLETION TASK You will write two more stories. Instead of pictures you will find a couple of sentences describing new situations. Please complete the stories, you will have 5 minutes per i H story. You will be instructed when to begin, and when to stop. Again do not worry about not having time to finish a story, just go on to the next story when asked to do so. Remember to include how the characters are feeling and what they are thinking. Also include how the story ends. 92 1. Bob and Ann have been dating for six months. they had an argument... 93 Today 2. Fred has worked very hard to get a scholarship. Today a letter came stating that his work was not good enough to merit the scholarship... 94 b I 2. Susan has worked very hard to get a scholarship. Today a letter came stating that her work was not good enough to merit the scholarship... 95 Consent Form 96 CONSENT FORM Thank you for your interest in participating in this research project. If you choose to participate you will be asked to fill out several questionnaires. You will be asked also to create several stories. Your participation will take about two hours. Your signature at the bottom of this page indicates your willingness to participate, and an aknowledgement that you have read and understood the following: 1. Your participation in this study has been explained to your satisfaction and is understood. 2. You freely consent to participate. 3. You understand that you may stop taking part in the study at any time without penalty. 4. All information you supply is confidential and anonymous. You will not be required to provide any identifying information. 5. There are judged to be no risks of any kind associated with participation in this study. 8. Your participation in this study does not guarantee any beneficial results to you. 7. At your request, you will receive additional information about this study after your participation is completed. 8. At your request, you will receive a summary of the results of the study. Date Signature 97 Handout for Participants 98 STUDENTS’ STORIES RESEARCH Thank you for participating in this study. Perhaps you learned something about your- self while completing the questionnaires and creating the stories. The purpose of this study was to explore whether there would be differences in the way relationships were portrayed in stories told by students who were feeling de- pressed and those who were not. Normally people experience a wide range of feelings, including temporary depressive feelings. But, about one in five adults in the U.S.A. will experience a pro- longed, serious episode of depression once in their lifetimes. A serious episode of de- pression is one that has lasted more than two weeks and has been present nearly every day for that time period. It is characterized by feeling: depressed, sad, blue, hopeless, low, and/or irritable. Other symptoms experienced are: poor appetite or increased appetite, insomnia or hypersomnia, psychmotor agitation or retardation, loss of in- terest or pleasure in usual activities, fatigue, feelings of worthlessness, self-reproach, excessive guilt, inability to think or concentrate, and recurrent suicidal thoughts. Treatment for depression has been very succesful. Some resources for stu- dents who are experiencing depression or other kinds of psychological problems are: MSU Counseling Center Phone # 355-2310 MSU Psychological Clinic Phone # 355-9564 Listening Ear (Confidential Crisis Line) Phone # 337-1717 RESULTS OF STUDY If you are interests in the results of this study, please write your name and address on a separate piece of paper. You can leave it in my box: Nayda M. Flores, M.A. Graduate Student Department of Psychology 135 Snyder I will mail you a summary of the results when the analysis is finished. Again, I thank you for your participation! 99 APPENDICES B 100 Mean Scores and Standard Deviations on the Beck Depression Inventory and Form E of the Depression Adjective Check Lists for Data Collected in 1988 and 1992. 101 Table 18 Mean Scores and Standard Deviations on the Beck'Depression Inventory and form E of the Depression Adjective Check Lists or Data ollected ' an 1988 1992 Inventory/ Sample a 21 SD. a ll SD. BDI Women 271 10 7 172 11 7 Men 200 8 6 89 9 7 DACL-Form E Women 271 10 6 172 11 6 Hen 200 10 6 89 11 6 102 i Distribution of Scores on the Beck Depression Inventory for the Whole Sample (n = 726), Women (11 = 437) and Men (11 = 289). 103 4o 30 20 10 l l l l l l l l l l l rnl 10 20 3O O TTITIIIIIIIIITITTII BDI 4O 30 20 10 l l l l l l l l l 1 hi 10 20 30 O llllllllTITllllllll BDlWOMEN 4O 30 20 10 O lllllllll-I—l—at-‘rl-J—lr4—11 O 10 20 30 ll TlllTllTlIllll BD'MEN Figure 4. Distribution of Scores on the Beck Depression Inventory for the Whole Sample (n = 726), Women (n = 437) and Men (n = 289). MM Distribution of Scores on the Depression Adjective Checklist Form E for the Whole Sample (n = 726), Women (n = 437) and Men (n = 289). 105 4O 3O 20 10 O lllJlLllllLlpjl'fill 10 20 3o lIT TTVTIIITIIIIII O DACL 4O SO 20 10 FITTITIIIIIITIIIIIT O 1111].ng llll h! l—ll 1 10 20 30 O DACL WOMEN 4O SO 20 10 O IWIIT 1411 ll 1 ll lhl—Ll l [1 10 3O O N C) DACL MEN Figure 5. Distribution of Scores on the Depression Adjective Checklist Form B for the Whole Sample (n = 726), Women (n = 437) and Men (n = 289). 106 Correlations Between the Beck Depression Inventory and Form E of the Depression Adjective Checklist. 107 Table 19 Correlations Between the Beck Depression Inventory and Form B of the Demession Adjective Check Lists. Sample 11 r. 9. Whole Sample 726 .45 .001 Women 437 .45 .001 Men 289 .47 .001 Note: All probabilities are based on two- tailed tests. 108 APPENDIX C 109 INSTRUCTIONS FOR IDENTIFYING TYPE OF SELF FROH PROJECTIVB STORIES When reading the TAT stories you will be trying to determine how the person who wrote the stories defines her/his self in relationship with other people. The self is a person's internal image that tells the person her/his worth in the world. Two types of selves will be identified: Separate Objective Self,and Connected Self. Neither of the categories will be considered healthy, unhealthy, more developed or less. This study assumes that there are different types of fully developed selves. You will be rating each TAT story for the presence of elements from both types, which means that each story will be rated twice. For each type you will be determining the presence or absence of statements related to four major areas: Reason for action or decision, Empathy, Relationship portrayed, and Help. Elements related to each area will be identified by letters. There can be more than one element in each area. If no elements from a particular area are present then you will code a D for that area. inx_indinate WWW ' Do not infer intentions, for each element scored you must be able to find the specific clause that corresponds with it. The reason for this rule is that people can describe the same action for different reasons. For example, in a story about a child who apologizes after being asked to do so by a parent, the child says: "I am sorry for having done this". So far there would be no way of determining type of self, since the child's action could be equally due to trying to avoid a punishment as to mend a relationship. Further statements in the story may tell you intention or result and then it could be scored. But if there is no further information, Reason for action or decision would get a 0 in both scorings for type of self. When proceeding to score the stories: a)Always review the scoring instructions and examples before beginning. b)Read the story, then read each element listed on all areas and score for presence of each element. If no element of an area is present in the story, give a D to that area. Scoring for SOS and CS can be done independently: some stories will have elements of both types of selves, some stories will not. 110 c)If the story is difficult to score go on to the next story and score the difficult one last for that subject. 111 W The self in this category is understood to be completely separate from others. People who fall in this category do not include others as part of their understanding of themselves. Thus, priority is given to self-development and achievements in areas outside of relationships, such as careers, sports, sexual conquests, etc. SOS take responsibility for their lives and their actions because they see each person as an individual responsible for him/herself only. Empathy for others is not relevant in their view, since each should care for him/herself. Help from others is not expected nor is help provided for others. Relationships are approached as contracts; being ended if others make demands or do not live up to expectations. Decisions are made based on concepts of right vs. wrong, principles and objective justice (i.e. "an eye for an eye...“). The fact that people in this category are not sympathetic to others' problems, does an; imply being destructive toward other people. The root is in their understanding of people as individuals, as completely separate entities responsible for themselves. Examples of benign SOS could be the stereotype of a judge or a pastor making judgements about themselves and other based on an objective law of right or wrong. A less benign example could be a person looking for vengeance. Remember to score for all the elements present in one area. 1. Reason for Action or Decision: a)Own advancement or satisfaction of own needs niihnnt concern of the effect on others: **There must be an explicit statement that the person does not care for adverse consequence of action on others; nQL.an omission category. Guided by the principle of "every man for himself". Examples: "Luke fell for it, now Bob knew nothing could stop him from becoming rich". ”He got what he wanted from her and left her to deal with the consequences". 112 **One character kills or attemps to kill another to cover up an illegal act committed: or the story describes an unexplained killing (or another act of violence) where the killer and the victim are not intimately related. Examples: Robber shooting victim so that he will not be recognized or a stalker killing a victim with no indication as why. b)A moral rule is used to determine what is to be done. Principles supercede others' feelings. Examples: "She was scared, but he knew what the right thing to do was". "His reasons did not matter, what he did was evil, she would not forgive him". c)Vengeance. Guided by the principle of "an eye for an eye and a tooth for a tooth". Examples: "He cheated on her, so she cheated on him.“ "Jake was going to ake him pay for what he had done.“ d)Fear of consequence (other than end of a relationship) such as being scolded, berated, receiving a punishment, being harmed by another: or simply to avoid conflict. Examples: "She never told anyone, lest she be disgraced." "He did not do it because he knew he would be caught.“ "He complied. That was the only way to stop the nagging." e)Action would be futile, no escape possible. Example: "She knew she could not get away with it." 2. Empathy: a)There is no statement or action that explicitly states empathy for another's feelings. If empathy is not an issue or is not called for in the story, you will still score it as a in this area. Do not score if any element of CS Area 2 (Empathy) is present. 113 b)Lack of sympathy for others' misfortunes. People's misfortunes are due to their own shortcomings. Example: "Hike deserved to lose his job, he was not hardworking like Joe (main character)". . c)Lack of understanding for others. Example: "She could not understand why he was so mean to her." 3.Relationship portrayed: a)Characters ties to each other are commercial or professional in nature, and no love is involved. Also score as a interactions between acquaintances or strangers where no love is involved. Examples: Employer and employee; co-workers: business partners. b)Characters involved in a sexual relationship where no love is involved or in an exploitative sexual relationship. Examples: Porno actors, prostitute and client, brief sexual encounter, rape. c)Characters intimately related (by blood, by marriage, etc.) and love of some kind can be assumed but: **At least one of the characters feels a need to escape the relationship. Example: "But the daughter spent her time planning how to escape from her mother's attentions." **One of the characters feels that the demands of the relationship are damaging. Example: “The son felt he could not stand one more minute in that house." **Relationship where one person is clearly manipulating another (look for things like: pretended to agree, could get her to do what he wanted, etc.) or where one person believes s/he owns the other (Example: "George made it clear, she was his and no one elses“). 114 **The relationship is contract style. It is ended if demands are not met. No compromise is reached when there is a disagreement. Example: “There was no use in trying to stay together, they would never agree on this." **Relationship where love has ended but that is maintained for negative reasons (such as to make the other suffer, or there is nothing better ”out there" anyway.), for reasons of individual gains even if both gain individually from it (such as financial gains, or social standing). or for religious reasons. Also score this element as present if an reason showing interconnection between the couple (such as companionship, loyalty, friendship) is explicitly given. Sex will not be considered as a reason showing interconnectedness. Staying together for the childrens'sake does not show interconnectedness either, so it would be scored for this element. **Characters do not interact at all. d)At least the main character has a strong negative feeling towards another (such as: hate, resentment, anger) which is not resolved. The feelings may be mutual. e)No relationship portrayed, single character concerned with issues not related to relationships, such as achievement. 4. Help: a)No help is expected from others, or provided from others. One can only rely on oneself. Score a even if this area does not appear to be relevant for the story. Do not score as a if emotional support is provided by other ("Her parents could not pay for her education but gave her the encouragement she needed to try again."- would get a score of 0 for this element). An example of a score of a would be:. "He knew it was up to him to get out of this one." 115 b)Does not help others, and others deserve what they get. Example: "Jean had brought this onto herself, Amanda would not bail her out." c)Doing for others only when repayment is anticipated. Example: "She knew that if she was nice to him, he would buy her a new dress.“ d)Help offered by others is rejected in favor of handling situation by self. 116 W There is inclusion of others as part of the self. People in this category see the self as interconnected to others. The self is defined within relationships. Attempts at uderstanding the other and at explaining oneself are observed. The feelings of others are acknowledged. This person is aware of the effect his/her actions will have on others. Maintaining relationships is a goal, so usually harmony and cooperation are emphasized. People influence and adapt to each other. Help is expected and provided for and from others. Satisfying everyone involved is a goal. 1.Reason for action or decision: a)Character feels responsible for another person, has to take care of another. This must be explicitly given as the reason. Example: “She had to take care of her elderly mother.” ”He could not leave her alone now, she needed him." b)Effect of one's actions on others are important considerations. Usually a compromise solution is reached so that others are not adversely affected. Example: "They worked out a plan: she would be going back to school and he would be working the extra time he needed in his career. That way she would not feel lonely and he would not give up his dreams either." "Having his mother move closer to him made both happy." “The mother felt torn by her obligations and her daughter's needs. The daughter was able to see that the mother cared.“ c)Lack of any self-interest. Sacrifice own needs for someone's elses. Example: “She would go with him, even though it meant losing those dear to her." "She did it for him, his well being was worth the humiliation.“ d)Own values are ignored for the sake of keeping a relationship. Example: ”He could not bear being without her, even if she would never be faithful to him." “She will do anything to keep him.“ 117 c)Desire to do s.t. depends on not hurting others' feelings. Example: “She would never become a showgirl, it would hurt her parents too much." f)Guilt as a reason for action. Must be explicitly mentioned as the reason. Example: "Leaving her was out of the question, he could never live with the guilt of causing his mother suffering." g)Character is responsible for s.o. else‘s actions (where guilt is not mentioned). Example: “She drove him to suicide." 2.Empathy: a)Explicitly states feelings of others. Examples: “Finally he understood she needed time with her friends." "She could see that he was hurting.“ b)Action that clearly shows understanding of how the other feels by responding to the other's needs, prayidnd that a reason of a different nature is nah Biven. Examples: "He consoled her." "Suzy was so bored, her mother looked at Suzy's face, closed the book and asked: What would you like to do?" "Don was in such pain, Greg stayed there until the crisis passed." c)Explicitly states that the character is hiding own feelings so as not to hurt someone else‘s feelings. Example: "He would never confess his love for Sonia to Maria. He did not want to hurt Maria." d)Two characters sharing the same emotion, excluding sexual satisfaction. Examples: falling in love, mourning, reminiscing positives of a long relationship like in an anniversary. nn51_hs_nnsitiya, negative emotions such as hate or indifference would not 118 be scored in this element. Two parents mourning death of son without becoming hurtful to each other would be scored for this element. 3. Relationship portrayed: a)Hain character must be in love or care (again must be stated explicitly) for another. b)Ending of relationship ,separation or threat of separation is followed by misfortune for a character: **A relationship ends or one character threatens to end the relationship (such as arguments, or one character expressing need for time without the other) and one character or both characters experience misfortune. Examples of misfortunes: suicide, threat of suicide, accidental death or loss of wealth. **A temporary or permanent separation (such as: being in college, or a trip) from significant others is followed by a threat to life (such as surviving an accident) or some other misfortune (significant other getting a terminal illness, see examples of other misfortunes above). c)Hain character is altruistic in the relationship. 4. Help: a)Help is expected from others or provided for others. Examples: "His parents reassured him that they would find a way to pay for his education." Also score 1 if help is only emotional support: "They could not help with money, but they listened and gave him good ideas on what to do next.“ b)Relying on another to care for self. Example: "Thank God, there was his good friend Jake to raise his spirits back." 119 c)Cooperation, mutual help or support are major themes. Example: "They had always counted on each other for help (support, etc.)." d)Helping and being generous makes one happy. Example: "Taking care of her elderly mother, made her feel inner peace." 120 3.0.— iv emanate.- seas-scum assess baa o: man van new Nan 2H¢< WEHSNJH §~0h DOPUOIEOU s... -v . H. sea-ate.- ..e=o.ocnn aeoasus h: m: v: n: N; w<fl¢< WEE—£84m cause "a .825; =8 ease-sauon-a: 121 APPENDIX D 122 50 45 4O 35 30 25 20 15 10 U' O O "W 'IIII'IIII[IIIIIIIII'IIIIIIIII'IIII'IIII'IIII b p- p p i- P n.— r- O N O o IIIIIIII'IIIIIIIII'IIII'IIII'IIIIIIIII'IIII IIII RS HELP SOS HELP ‘ Figure 6 Distributions of Scores on the Area of Help on both Self-Types for the Four Small Groups Combined (n = 96). 123 List of References Abramson, L.Y., Seligman, M.E.P., & Teasdale, JD. (1985). Learned helplessness in humans: Critique and reformulation. In J .C. Coyne (Ed.), Essential papers on depression (pp. 259-310). New York: New York University Press. Amenson, C.S., & Lewinsohn, PM. (1981). An investigation into the observed sex difference in prevalence of unipolar depression. Journal of Abnormal Psychology, 9Q, 1-13. American Psychiatric Association (1981). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: Author. American Psychiatric Association (1987). Diagnostic and statistical manual of mental disorders (3rd ed. rev., pp. 213-233). Washington, DC: Author. American Psychiatric Association (1994). Diagnostic and statistical manual of 124 mental disorders (4th ed.). Washington, DC: Author. Bar-Yam, M. (1991). Do women and men speak in different voices? A comparative study of self-evolvement. International Journal of Aging and Human Development, 12, 247-259. Beck, A.T. (1972). Beck Depression Inventory. Philadelphia; Center for Cognitive Therapy. Beck, A.T., Steer, R.A., & Garbin, M.G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, §, 77-100. Beeman, W.O. (1985). Dimensions of dysphoria: The view from linguistic anthropology. In A. Kleinman 85 B. Good (Eds), Culture and depression: Studies in anthropology and cross-cultural psychiatry of affect and disorder (pp. 216-243). Berkeley: University of California Press. Bellak, L. (1975). The TAT, CAT and SAT in clinical use (3rd ed.). New York: Grune and Stratton. 125 Bern, S.L. (1983). Gender Schema Theory and its implications for child development: Raising gender-aschematic children in a gender- schematic society. Signs: Journal of Women in Culture and Society, S, 598-616. Bemporad, J. (1983). Psychotherapy of the depressive character. In M.B. Cantor & M.L. Glucksman (Eds), Affect: Psychoanalytic theory and practice (pp. 185-209). New York: John Wiley and Sons. Billings, A.G., Cronkite, R.C., & Moos, RM. (1983). Social-environmental factors in unipolar depression: Comparisons of depressed patients and nondepressed controls. Journal of Abnormal Psychology, Q, 119-133. Boggiano, A.K., & Barrett, M. (1991). Gender differences in depression in college students. Sex Bples, 25, 595-605. Boverman, I.K., Boverman, D.M., Clarkson, F.E., Rosenkrantz, P.S., & Vogel, SR. (1970). Sex-role stereotypes and clinical judgements of mental health. Journal of Consulting Psychology, 14, 1-7. 126 Briscoe, M. (1982). Sex differences in psychological well-being (Monograph). Psychological Medicine, (Suppl. 1). Brooks-Gunn, J ., & Matthews, W.S. (1979). He and she: How children develop their sex-role identity. New Jersey: Prentice Hall. Brown, G., Bhrolchain, M., & Harris, T. (1975). Social class and psychiatric disturbance among women in an urban population. Sociolog , 9, 225-254. Brown, G.W., & Harris, T. (1978). The social origins of depression: A study of psychiatric disorder in women. New York: Free Press. Chevron, E.S., Quinlan, D.M., & Blatt, SJ. (1978). Sex roles and gender differences in the experience of depression. :Journal of Abnormal Psychology, _8_Z, 680-683. Chodorow, N. (1974). Family structure and feminine personality. In M.Z. Rosaldo & L. Lamphere (Eds.), Women, culture and society (pp. 43-66). Stanford, CA: Stanford University Press. 127 Chodorow, N. (1979). Feminism and difference: Gender relation, and difference in psychoanalytic perspective. Socialist Review, 4_6_, 42-64. Cohn, L.D. (1991). Sex differences in the course of personality development: A meta-analysis. Pflchological Bulletin, L09, 252-266. Coyne, J .C. (1985). Ambiguity and controversy: An introduction. In J .C. Coyne (Ed.), Essential papers on depression (pp. 1-22). New York: New York University Press. Culkin, J ., & Perroto, RS. (1985). Assertiveness factors and depression in a sample of college women. Psychological Reports, 51, 1015-1020. DeLongis, A., Folkman, S., & Lazarus, R. (1988). The impact of daily stress on health and mood: Psychological and social resources as mediators. Journal of Personality and Social Psychology, fl, 486-495. DeVos, G., Marsella, A.J., & Hsu, F.L.K. (1985). Introduction: Approaches to culture and self. In A.J. Marsella, G. DeVos & F.L.K. Hsu 128 (Eds.), Culture and self: Asian and western perspectives (pp. 2-23). New York: Tavistock Publications. Drill, R.L. (1987). Young adult children of divorced parents: Depression and perception of loss. ournal of Divorce 10 169-187. ,——, El-Islam, M.F. (1981). Epidemiologic research on depressive disorders. In T.A. Ban, R. Gonzalez, A.S. Jablensky, N.A. Sartorius & F.E. Vartanian (Eds.), Prevention and treatment of depression (pp. 75-85). Baltimore: University Park Press. Erikson, E. (1950). Childhood and society. New York: W. W. Norton. Essex, M.J., Klein, M.H., Lohr, M.J., & Benjamin, LS. (1985). Intimacy and depression in older women. Psychiatry, fl, 159-177. Freud, S. (1959). Mourning and melancholia. In S. Freud, Collected Papers (Vol. 4, pp. 152-170). New York: Basic Books. Funabiki, D., Bologna, N.C., Pepping M., & Fitzgerald, KC. (1980). Revisiting sex differences in the expression of depression. 129 Journal of Abnormal Psychology, _8_9, 194-202. Gilligan, C. (1982). In a different voice: Psychological theory and women’s development. Cambridge, MA: Harvard University Press. Gough, K. (1975). The origin of the family. In R.R. Reiter (Ed.), Toward an anthropology of women (pp. 51-76). New York: Monthly Review Press. Gove, W.R. (1972). The relationship between sex roles, marital status and mental illness. Social Forces, fl, 34-44. Greenberg, J .R., & Mitchell, SA. (1983). Object relations in psychoanalytic theory. Cambridge, MA: Harvard University Press. Hammen, C.L., & Padesky, GA. (1977). Sex differences in the expression of depressive responses on the Beck Depression Inventory. Journal of Abnormal Psychology, _8_6, 609-614. Hammen, C.L., & Peters, SD. (1978). Interpersonal consequence of 130 depression: Responses to men and women enacting a depressed role. Journal of Abnormal Psychology, 8_7, 322-332. Heilbrun, A.B., Jr., & Han, Y.L. (1986). Sex differences in the adaptative value of androgyny. Psychological Reports, 59, 1023—1026. Hoffman, M. (1977). Sex differences in empathy and related behaviors. Psychological Bulletin, 8_4, 712-722. Hops, H., Biglan, A., Sherman, L, Arthur, J., Friedman, L., & Osteen, V. (1987). Home observations of family interactions of depressed women. Journal of Consulting and Clinical Psychology, _5_5_, 341-346. Hsu, F.L.K. (1985). The self in cross-cultural perspective. In A.J. Marsella, G. DeVos & F.L.K. Hsu (Eds.), Culture and self: Asian and western perspectives (pp. 24-55). New York: Tavistock Publications. Jack, D.C., & Dill, D. (1992). The Silencing the Self Scale: Schemas of intimacy associated with depression in women. Psychology of 131 Women Quarterly, 16, 97-106. Jackson, SW. (1985). Acedia the sin and its relationship to sorrow and melancholia. In A. Kleinman & B. Good (Eds.), Culture and depression: Studies in the anthropology and cross-cultural psychiatry of affect and disorder, (pp. 43-62). Berkeley: University of California Press. Kaplan, A. (1986). The ”Self-in-Relation”: Implications for depression in women. Psychotherapy, 2_3, 234-242. Karon, BR (1981). The Thematic Apperception Test (TAT). In A.I. Rabin (Ed.), Assessment with profictive techniques: A concise introduction (pp. 85-120). New York: Springer Publishing Company. Keyes, CF. (1985). The interpretive basis of depression. In A. Kleinman & B. Good (Eds.), Culture and depression: Studies in the anthropology and cross-cultural psychiatry of affect and disorder (pp. 153-174). Berkeley: University of California Press. 132 Kohlberg, L. (1981). The philosophy of moral development: Moral stages and the idea of justice. San Francisco: Harper and Row. Kovacs, M., & Beck, A.T. (1985). Maladaptive cognitive structures in depression. In J .C. Coyne (Ed.), Essential papers on depression (pp. 240-258). New York: New York University Press. Leibowitz, L. (1975). Perspectives on the evolution of sex differences. In R.R. Reiter (Ed.), Toward an anthropology of women (pp. 20-35). New York: Monthly Review Press. Lubin, B. (1965). Adjective checklists for measurement of depression. Archives of General Psychiatry, 12 57-62. —’ Lubin, B. (1967). Depression Adjective Check Lists. San Diego, CA: Educational and Industrial Testing Service. Lubin, B. (1981). Manual for the Depression Adjective Check Lists. San Diego, CA: Educational and Industrial Testing Service. Lutz, C. (1985). Depression and the translation of emotional worlds. In A. 133 Kleinman & B. Good (Eds.), Culture and depression: Studies in the anthropology and cross-cultural psychiatry of affect and disorder (pp. 63-100). Berkeley: University of California Press. Maccoby, EB. (1990). Gender and relationships: a developmental account. American Psychologist,45, 513-520. Marsella, A.J. (1980). Depressive experience and disorder across cultures. In H.C. Triandis & J .G. Draguns (Eds.), flandbofi of cross- cultural psychology (pp.237-290). Boston: Allyn and Bacon. Marsella, A.J. (1985). Culture, self, and mental disorder. In A.J. Marsella, G. DeVos, & F.L.K. Hsu (Eds.), Culture and self: Asian and western perspectives (pp. 281-307). New York: Tavistock Publications. Marsella, A.J., Sartorius, N., Jablensky, A., & Fenton, ER. (1985). Cross-cultural studies of depressive disorders: An overview. In A. Kleinman & B. Good (Eds.), Culture and depression: Studies in the anthropology and cross-cultural psychiatry of 134 affect and disorder (pp. 299-324). Berkeley: University of California Press. Mazur, E., & Olver, RR. (1987). Intimacy and structure: Sex differences in imagery of same-sex relationships. Sex files, 16, 539-558. McLeod, J .D., Kessler, R.C., & Landis, KR. (1992). Speed of recovery from Major Depressive Episodes in a community sample of married men and women. Journal of Abnormal Psychology, 19;, 277-286. Miller, J.B. (1984). The development of women’s sense of self. Work in Progress. Wellesley, MA: Stone Center for Developmental Services and Studies, Wellesley College. Miller, J .B. (1986). What do we mean by relationships? Work in Prcgress. Wellesley, MA: Stone Center for Developmental Services and Studies, Wellesley College. Murray, HA. (1971). Thematic Apperception Test. Cambridge, MA: Harvard University Press. 135 Murstein, BL (1963). Theory and research in projective techniques (Emphasizing the TAT). New York: John Wiley and Sons. Murstein, BL (1972). Normative written TAT responses for a college sample. Journal of Personality Assessment, 36, 109-147. Neugarten, B.L., & Gutmann, D.L. (1958). Age—sex roles and personality in middle age: A thematic apperception study. Psychological Monographs: General and Applied, 12, 1-33. Nolen-Hoeksema, S. (1987). Sex differences in unipolar depression: Evidence and theory. Psychological Bulletin, _1_0_ZL, 259-282. Nolen-Hoeksema, S., & Girgus, J .S. (1994). The emergence of gender differences in depression during adolescense. Psychological Bulletin, IQ, 424-443. Oatley, K., & Bolton, W. (1985). A social-cognitive theor of depression in reaction to life events. Psychological Review, 92, 372-388. O’Brien, E.J. (1991). Sex differences in components of self-esteem. 136 Psychological Reports, 68, 241-242. Parker, G., Brown, L., & Blignault, I. (1986). Coping behaviors as predictors of the course of clinical depression. Archives of General Psychiatry, 43, 561-565. Pearson, J.L., Reinhart, M.A., Donelson, E., Strommen, E.A., & Barnes, CA. (1985). Marital status and Gilligan’s developmental theory of care of self and other. Paper presented at the 198 Biennial Meeting of the Society for Research in Child Development, Toronto, Canada. Pollak, S., & Gilligan, C. (1982). Images of violences in Thematic Apperception Test stories. Journal of Personality and Social Psycholog ,fl, 159-167. Reinhart, M.A., Pearson, J.L., Strommen, E.A., Donelson, E., Barnes, C.L., Blank, L, Cebollero, A.M., Cornwell, K., & Kamptner, N.L. (1987). Revised Relationship Self Inventory. East Lansing, MI: Michigan State University. 137 Reinhart, M.A., Pearson, J.L., Kamptner, N.L., Cornwell, K., Barnes, C.L., Strommen, E.A., & Donelson, E. (1985). Assessment of Gilligan’s model: development of the Relationship Self Inventory. Paper presented at the 1985 Biennial Meeting of the Society for Research in Child Development, Toronto, Canada. Robbins, P.R., & Tanck, RH. (1991). Gender differences in the attribution of causes for depressed feelings. Psychological Reports, 68, 1209-1210. Roberts, J .E., & Monroe, SM. (1992). Vulnerable self-esteem and depressive symptoms: Prospective findings comparing three alternative conceptualizations. Journal of Personality and Social Psycholog, 2, 804-812. Roy, A. (1985). Early parental separation and adult depression. Archives of —, General Psychiatry, 42 987-991. Sacks, K. (1975). Engels revisited: Women, the organization of production, and private property. In R.R. Reiter (Ed.), Toward an anthropology of women (pp. 211-234). New York: Monthly Review 138 Schieffelin, EL. (1985). The cultural analysis of depressive affect: An example from New Guinea. In A. Kleinman & B. Good (Eds.), Culture and depression: Studies in the anthropology and cross-cultural psychiatry of affect and disorder (pp. 101-133). Berkeley: University of California Press. Schuyler, D. (1974). The Depressive spectrum. New York: Jason Aronson. Shweder, RA. (1985). Menstrual pollution, soul loss, and the comparative study of emotions. In A. Kleinman & B. Good (Eds.), Culture and depression: Studies in the anthropology and cross-cultural psychiatry of affect and disorder (pp. 182-215). Berkeley: University of California Press. Slocum, S. (1975). Woman the gatherer: Male bias in anthropology. In R.R. Reiter (Ed.), Toward an anthropology of women (pp. 36-50). New York: Monthly Review Press. Surrey, J .L. (1985). Self-in—Relation: a theory of women’s development. Work in Progress. Wellesley, MA: Stone Center for Developmental 139 Services and Studies, Wellesley College. Tennant, C. (1985). Female vulnerability to depression. Psychological Medicine, _16, 733-737. Tennant, C., Hurry, J ., & Bebbington, P. (1982). The relation of different types of childhood separation experience to adult psychiatric disorder. Brittish Journal of Psychiatry, Lg, 475-482. Warren, L.W., 8:: McEachren, L. (1985). Derived identity and depressive symptomatology in women differing in marital and employment 9 133-144. —’ status. Psychology of Women Quarterly, Weissman, M.M. (1981). The epidemiology of affective disorders in the US. In T.A. Ban, R. Gonzalez, A.S. Jablensky, N.A. Sartorius & F.E. Vartanian (Eds.), Prevention and treatment of depression (pp. 105-118). Baltimore: University Park Press. Weissman, M.M., & Klerman, CL. (1977). Sex differences and the epidemiology of depression. Archives of General Psychiatry, 14, 98-111. 140 Weissman, M.M., & Myers, J.K. (1978). Affective disorders in a U.S. urban community. Archives of General Psychiatry, 3_5, 1304-1311. Weissman, M.M., & Paykel, ES. (1974). The depressed woman: A study of social relationships. Chicago: University of Chicago Press. Welkowitz, J ., Lish, J .D., & Bond, RN. (1985). The depressive experiences questionnaire: Revision and validation. Journal of Personality Assessment, fl, 89-94. 141 "I111111111llllllf