lllllllllllllllllllllllllllllllllllllllllllllllllllllllllll THESIS 3 1293 103825851 This is to certify that the thesis entitled Maternal Separation in Late Adolescent Women presented by Patricia Ann Ponto has been accepted towards fulfillment of the requirements for Masters Psychology degree in [/45:;;_ (35A,a,ég:a—\_~ Major professor Date 5/1u/81 0-7639 MSU LIBRARIES n. RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. mum WTR 1983 Juwjzjfi‘issg MATERNAL SEPARATION IN LATE ADOLESCENT WOMEN By Patricia Ann Ponto A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1981 ABSTRACT MATERNAL SEPARATION IN LATE ADOLESCENT WOMEN By Patricia Ann Ponto The purpose of this research is to investigate the process of separation that takes place between mothers and daughters when the daughters leave for college. The sample consists of 101 mother-daughter pairs. Five variables are studied: autonomy, identification, accuracy of perception, mothers' employment, and daughters' choice of spending free time. The first three are measured through a 180-item questionnaire, the Structural Analysis of Social Behavior. The last two involve answers to questions in a behavioral inventory. Analyses of variance and trtests are employed in testing the hypotheses.' The major conclusion of the study is that distancing from mother is an important part of the process of late adolescent separation for women. Specific distancing mechanisms include low identification with mother, low accuracy of perception, and free time spent with friends. In addition, many questions are raised about internalized autonomy. The meaning of the concept, particularly its role in late adolescent separation, is questioned, as is the accuracy of its measurement through Series C of the SASB. ACKNOWLEDGMENTS Writing this thesis has been a complex, painful, and intriguing process--large1y because it focused in many ways on my own developmental struggle of separating from my deceased mother and, thereby, gaining autonomy. Not only was the content appropriate to my struggle, but also the process. The process (in this case, my style) was to work in a very autonomous fashion--being even more independent at times than I really wanted to be or found useful. In retrospect, I sigh a bit and realize that such situations arise when one embarks on such an over- determined piece of work. I'm grateful to . Elaine Donelson, my chairwoman, for her very careful and thorough editing of my work and for her acknowledgment and support of my competence and autonomy; John Schneider, committee member, for his thoughtful comments about separation as loss; John Hurley, committee member, for his reasonable perspective on masters' theses; Wendy Fielder, for her sympathetic acceptance of my stuckness and procrastination; Nancy Egan, for giving me some necessary pushes; Lynn Wendyger, for listening to my craziness, for being there, and for finishing and, thus, showing me that it could be done; ii Peggy Walsh, for acknowledging and helping me understand the complexity of the work as it related to my own life. It's good to be finished. iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . vii INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Parental Separation in Late Adolescence . . . . . . . . . . . . . 2 Theoretical Contributions . . . . . ... . . . . . . . . . . . . 3 Normative Research . . . . . . . . . . . . . . . . . . . . . . . 6 Clinical Literature Regarding Problematic Separation . . . . . . 9 The Mother-Daughter Attachment and Separation . . . . . . . . . . 10 Theoretical Contributions . . . . . . . . . . . . . . . . . . . 10 Normative Research . . . . . . . . . . . . . . . . . . . . . . . 17 STATEMENT OF HYPOTHESES . . . . . . . . . . . . . . . . . . . . . . 20 METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Subjects . . . . . . . . . . . . . . . . . . . . .g. . . . . . . . 27 Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Measurement of the Variables . . . . . . . . . . . . . . . . . . . 34 Tests of the Hypotheses . . . . . . . . . . . . . . . . . . . . . 36 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Questionnaire Return . . . . . . . . . . . . . . . . . . . . . . . 39 The Autonomy Scores . . . . . . . . . . . . . . . . . . . . . . . 41 Basic Data Outline . . . . . . . . . . . . . . . . . . . . . . . . 44 Hypothesis 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Hypothesis 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Hypothesis 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 iv TABLE OF CONTENTS (Cont'd.) Hypothesis 4 . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Hypothesis 5 . . . . . . . . . . . . . . . . . . . . . . . . . . 74 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Introjected Autonomy, Autonomy in Relation to Mother, and the Relationship between the Two . . . . . . . . . . . . . 30 Autonomy and Identification as Perceived Similarity . . . . . . 33 Autonomy and Accuracy of Perception . . . . . . . . . . . . . . 91 The Effects of Mothers' Autonomy . . . . . . . . . . . . . . . . 96 The Effects of Mohters' Employment Outside the Home . . . . . . 101 The Effects of Ways of Spending Free Time . . . . . . . . . . . 104 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 APPENDIX A--SASB Questionnaire . . . . . . . . . . . . . . . . . 128 APPENDIX B--Marlowe-Crowne Social Desirability Scale . . . . . . 138 APPENDIX C--Cover Letter to Mothers . . . . . . . . . . . . . . 140 APPENDIX D--Behaviora1 Inventory for Mothers . . . . . . . . . . 142 APPENDIX E--Behaviora1 Inventory for Daughters . . . . . . . . . 144 APPENDIX F--Diagrams of Hypotheses . . . . . . . . . . . . . . . 146 10. 11. 12. 13. 14. 15. LIST OF TABLES Basic Data Outline The Relationship between Daughters' Introjected Autonomy and the Identification Correlation . The Relationship between Daughters' Autonomy in Relation to Mother and the Identification Correlation The Relationship between Daughters' Introjected Autonomy and Self-report Identification The Relationship between Daughters' Autonomy in Relation to Mother and Self-report Identification The Relationship between Daughters' Correlated Accuracy of Perception and Daughters' Introjected Autonomy . The Relationship between Mothers' Correlated Accuracy of Perception and Daughters' Introjected Autonomy . The Relationship between Daughters' Correlated Accuracy of Perception and Autonomy in Relation to Mother The Relationship between Mothers' Correlated Accuracy of Perception and Autonomy in Relation to Mother The Relationship between Reported Accuracy of Perception and Daughters' Introjected Autonomy . The Relationship between Reported Accuracy of Perception and Daughters' Autonomy in Relation to Mother . The Relationship between Daughters' Estimate of the Accuracy of the Mothers' Perceptions of Daughters and and Daughters' Autonomy . The Relationship between Mothers' Estimate of the Accuracy of the Daughters' Perception of Mothers and Daughters' Autonomy . The Relationship between Mothers' Introjected Autonomy, Their Autonomy Giving Behaviors, and Their Daughters' Sense of Autonomy . Iftests Regarding the Relationship between Mothers' High Introjected Autonomy Group and All Others . . vi 45 48 49 51 52 54 55 55 S6 58 6O 61 63 65 66 16. 17. 18. 19. 20. 21. LIST OF TABLES (Cont'd.) The Relationship between Mothers' Outside Employment and Autonomy Giving Behaviors . A Further Delineation of the Relationship between Employment Outside the Home and Mothers' Autonomy Giving Behaviors . . . The Relationship of Free Time to Introjected Autonomy . The Relationship of Free Time to Autonomy in Relation to Mother . One-way Analysis of Variance on the Relationship of Free Time to Autonomy . The Means of the Analyses of Variance Tests on the Relationship between Identification and Autonomy vii 68 7O 74 75 77 8S LIST OF FIGURES 1. SASBModeI....................‘.....125 2. Pole and Quadrant Titles . . . . . . . . . . . . . . . . . . 126 3. Affiliation and Autonomy Weights . . . . . . . . . . . . . . 127 viii INTRODUCTION The relationship between mothers and daughters is a tOpic that has recently gained prominence in literature concerning women. Current writers are convinced that this relationship is very different from those of other dyads and that its quality has far-reaching effects for women (Friday, 1977; Hammer, 1975). The relationship is seen to be changing as, through the women's movement, more women commit themselves to self-development. When women who are mothers begin to define themselves more sharply as individuals rather than being sub- sumed by the caretaker role, they will be able to encourage their daughters to achieve greater autonomy and clearer senses of themselves as individuals. It is the daughter's autonomy or separateness that is of particular interest to me. The issue of autonomy is involved throughout the life span of the mother-daughter relationship, but it seems to me that there are several points in typical development at which it is particularly significant. Those are: in late infancy, when autonomy is first experienced; at puberty, when biology dictates that sexuality be addressed; at the end of adolescence, when daughters separate physically or psychologically from home to create their own lives; at marriage, when the daughter also becomes a wife; and, perhaps most significantly, at the birth of a child (especially a female child), when the daughter joins the ranks of mothers. Of these critical points, one that has not been addressed sufficiently is the separation at late adolescence. The importance of this separation has, I think, been underestimated. As women move more and more to establishing their own identities rather than subscribing to the earlier "automatic" commitments of marriage and family, it is this separation that will gain in importance for increasing numbers of women because it will be a critical factor in the development of a sense of self. Though there is very little literature concerned specifically with maternal separation in late adolescent women, there are several related areas that require investigation in a study with this focus. The first of these is the theoretical and empirical work that has been done on late adolescent separation from parents in general. This includes both normative and clinical theory and research. Second is the information about the mother-daughter relationship and the implica- tions it holds for late adolescent separation. Parental Separation in Late Adolescence Much like its more specific counterpart of maternal separation in late adolescent women, this area has not received nearly as much attention as seems its due. The process of separation from parents is implicitly contained in many theories of adolescence as a crucial aspect of the transition from childhood to adulthood. It is also referred to in its negative form as differentiation failure (which resultsimiadolescent psychopathology). The normative research done in this area is, however, very sparse. Theoretical Contributions Among those theorists who have specifically addressed the issue of separation from parents in adolescence are Peter Blos, Edith Jacobson, Helm Stierlin, and Irene Josselyn. Cutting the "psychological navel cord" (1967, p. 176) is Peter 8105' colloquial definition of the process of adolescent separation. According to 8105, one of the foremost psychoanalytic theorists of adolescence, the shedding of family dependencies is critical to the adolescent's becoming a member of the adult world. In this pursuit, he/she is aided by peers who permit experimentation with the newfound independence and alleviate guilt feelings that accompany the separation. The actual process of separation involves, basically, the disengage- ment from the parents as primary love objects. This is accomplished by a return to the early phases of development and a reactivation of infantile emotional involvements. In 8105' view, the conflicts surrounding those involvements, particularly the ones relating to Oedipal wishes and fantasies, are more sharply focused at this point in time. After they are worked through at this stage of greater maturity, they are,hopefully, better resolved. The result of the process, then, is disengagement from the internalized infantile objects. When it is accomplished, a new equilibrium is established. The new equilibrium is expressed by the adolescent as a personal and autonomous life style. Following this expression of autonomy, rapprochement or reidentification with the primary object representations at an adult level takes place and completes the adolescent separation process (8105, 1962). Ego psychologist Edith Jacobson asserts that "affectionate ties [to parents] must be loosened to guarantee future freedom of object choice and permit sound readjustment to one's own generation, and a normal adjustment to adult social reality" (1964, p. 170). She views the loosening of the symbiotic bond as essential to the processes of ego development, identity formation, and differentiation. If the process of separation fails, chronic pathology results. Such failures often occur, in Jacobson's opinion, because of parental interference with the process. (The interference may take several forms: the parent living through the child, the parent dominating or overprotecting the child, or the parent treating the child as a narcissistic extension of him/herself.) The major task of parents during this period, then, is to encourage the instinctual and emotional freedom of the adolescent. The major task confronting the adolescent, on the other hand, is to alter his/her identification of parents as parents to parents as sexually active persons, who will grant him/her the right to engage in sexual and adult activities. Jacobson describes this task as the ”most incisive and difficult step" (1964, p. 175) in the process of separation. Its result is increased freedom for the adolescent from both external and old superego influences. Like Jacobson, Helm Stierlin is concerned with parent-child interactions that discourage adolescent separation. He believes that parents may discourage their adolescents from separating because they themselves are in the midst of a developmental crisis that they would prefer to ignore. Stierlin pr0poses that parents interfere with separation by "binding" or "delegating" their adolescents (1974, p. 10). The "binding" mode involves attempting to tie children to them by means of affect, cognition, or the superego. Affective binding results from overgratification of the regressive wishes of the child by the parents. The result is that the child is so spoiled that only his/her parents will tolerate him/her. Cognitive binding involves parental interference with the child's ability to perceive and articulate his/her own feelings, needs, motives, and goals. The result is an adolescent who cannot trust himself/herself and, consequently, depends on parental judgment. In superego binding, the adolescent is taught to believe that his/her parents cannot survive without him/her. The adolescent, then, views separation as murder. Finally, parents block separation through the "delegating" mode of interaction by putting pressure on the adolescent to fulfill some "mission" in the world that is designed to meet the needs of the parents or to compensate for their own inadequacies. Irene Josselyn describes separation from parents as an ambi- valent process, quite similar to the one that occurs in infancy. She believes that the adolescent experiences pulls toward both dependence and independence (1971). As in infancy, the adolescent asserts his/her separateness through negativism to parents and other authorities. A large part of the motivation for rejection of parents in Josselyn's view relates to the adolescent's budding sexuality and the possibility that the accompanying, newly experienced emotions will be turned toward the parents. The adolescent needs, then, to reduce or deny the intensity of his/her tie to the parents in order to avoid the sexual arousal closeness might initiate. Josselyn notes in addition that the adolescent's growing ego ideal requires that he/she be a separate entity. As the ego ideal assumes more power in an individual's life, the childhood superego must recede. The rebellion against parents in adolescence may actually be, in this theorist's view, a projected battle with the adolescent's own childhood superego. Following these battles for independence in the earlier stages of adolescence, the late adolescent realizes that emotional maturity does not mean total independence from others. When the late adolescent has achieved "independence in the area of individual competence" (1971, p. 185) and no longer fears dependence as re-engulfment in the childhood relationship, he/she is free to engage in mature inter- dependence. This state is based on acceptance of the idea that others need the individual and that the individual has needs that others can gratify. In Josselyn's words, it is at this point that "childhood dependence has grown up" (1971, p. 185) and adult life is achieved. Normative Research As mentioned earlier, normative research concerning the separation of adolescents from the family is sparse. The notion that an adolescent must become independent of parents' control to become an adult was reported as early as 1904 by G. Stanley Hall. The first reserach related to that construct, however, came thirty years later. In the 19305, McDill developed a lZO-item questionnaire, the Emancipa- tion from Parents Scale. His purpose was to study emancipation with regard to the variables of religion, socioeconomic status, intelligence, height, and weight. The correlations he obtained were quite low and, apparently, interest in the instrument subsequently waned. A similar instrument for use with university students was, however, constructed by Sherman in 1946. His concern, much like McDill's, was to discover the relationship between emancipation scale scores and various subject characteristics (sex, intelligence, social class, religion, age, time at college). He compared two groups scoring high and low on his emancipation measure. His results also showed only slight relationships between emancipation and the various characteris- tics. Sherman concluded that future research should focus on the relationship between the respondents and their parents. It was about twenty years later that the next research relating specifically to adolescent separation was done. In the early sixties, Murphey, Silber, Coelho, Hamburg, and Greenberg studied the development of autonomy in college freshmen and its relationship to various patterns of parent-child interactions as part of an NIMH project focusing on the development of competent adults. The research team interviewed nineteen students and their parents at the end of the students' senior year in high school and four times during the first year in college. The students were grouped with regard to their relative positions on two dimensions: autonomy and relatedness. Autonomy was defined as the ability to make "separate, responsible choices” (1963, p. 645). Related- ness was defined as the student's expressed satisfaction "in a pre- dominantly positive relationship with his parents" (1963, p. 645). When the nineteen subjects were grouped according to these two dimensions, only one person was found to be low in both autonomy and relatedness. Consequently, the researchers studied only these three groups: high-autonomy, high-related; high-autonomy, low-related; and low-autonomy, high-related. Their findings revealed clearcut differences with respect to separation among the parents of the sUbjects in the three groups. Parents of the high-autonomy, high— related subjects were supportive of their children's development of autonomy and were confident of their children's ability to handle situations. They did not usually offer help unless they were asked for it and they favored their children's separation as necessary for growth. The parents of the low-autonomy, high-related subjects were unclear about the boundaries between themselves and their children. They often wanted their children to fulfill their frustrated aspirations and did not clearly perceive their children's interests and abilities. They also lacked confidence in their children's ability to achieve autonomy. The parents of the third group of subjects (high-autonomy, low-relatedness) viewed their children as separate and provided them with opportunities to develop their own abilities and interests, both important conditions for the development of autonomy. However, these parents could not later accept the subjects' independence and changing values because they required modification of the roles of the subjects in the families. The parents considered these roles unChangeable. In a study of adolescents in 1966, Douvan and Adelson found that autonomy is characteristic of adolescents whose parents have encouraged gradual detachment from the family. They were very success- ful at differentiating autonomous from non-autonomous subjects on the basis of whether the adolescent spent most of his/her time with peers or family members. In a very recent study (1979) of home-leaving in late adolescents, Hotch attempted to provide a framework for separation by adopting a general personality construct, engagement style. There are three forms of engagement style in the framework by McKinney (1978) that she employed: agent (in which the individual views himself/herself as a "doer"), patient (the self is viewed as acted upon) and communal (the self both acts and is acted upon). Hotch hypothesized that the interaction of self-sufficiency and relatedness would be the best predictor of style of leaving home. Her results indicated that style of perceived home-leaving could be reliably assessed and that related- ness, not self-sufficiency or the interaction of the two, was the best predictor of style. Clinical Literature Regarding Problematic Separation The importance of separation in the development of a healthy family system is supported by the literature on lack of differentiation and the subsequent psychopathology. Morris and Wynne (1965) have suggested that pathological families frequently involve symbiosis or lack of differentiation. Olson adds that the presenting patient often has few ego boundaries and is "defined" in terms of the family system. Bowen (1960) describes how symbiosis is involved in the development of schizophrenia and how families with a schizophrenic child experience a crisis when the child approaches adolescence and the expected separation from parents. In addition to this rather back-handed affirmation of the importance of the process of separation, Singer and Wynne also affirm 10 its significance in a more straightforward manner. They report (1966) that one of the two major elements of a healthy family is a separate and potent identity fer each family member. This idea is further supported by the results of a survey conducted by the Group for the Advancement of Psychiatry in 1970, which revealed that one of the three primary therapeutic goals listed most frequently by its members was improved autonomy and individuation. . In a clinical study directly related to separation, Stierlin (1971) investigated families of adolescents who were in family therapy. His interest was to discover the way in which parental perceptions of adolescent children influenced the child's ability to separate. This study involved family sessions, individual inter- views with family members, family Rorschachs, and family art evaluations. The findings revealed that there are three areas of parental perception that seem most influential in determining the individual's capacity to separate: (1) whether or not the adolescent is seen by the parents as potentially able to be autonomous; (2) whether or not the adolescent is seen by the parents as potentially able to form relationships outside of the family; and (3) whether the adolescent's separation is seen by the parents as good and normal, or bad and destructive. The Mbther-Daughter Attachment and Separation Theoretical Contributions The special attachment between mothers and daughters and the particular difficulty involved in mother-daughter separation has been 11 discussed by such theorists as Sigmund Freud, Helene Deutsch, Nancy Chodorow, Lois Hoffman, and J.H. Block. Freud was well aware of the importance of separation from parents. He wrote: From the time of puberty onward the human individual must devote himself to the great task of freeing himself from the parents; and only after this detachment is accomplished can he cease to be a child and so become a member of the social community (1935, p. 295; emphasis mine). He was also aware of the particular difficulty daughters have in freeing themselves from their mothers--largely as a result of their childhood attachment and identifications. In his essay "Female Sexuality," Freud pointed to the infantile "phase of exclusive attachment to the mother" and concluded that it "is far more important to women than it can claim to be for men" (1953, p. 258). As proof of this, Freud asserted that many women choose husbands that will enable them to duplicate their childhood relationship to their mother and that this often results in a difficult relationship with the husband. Freud believed that "development of womanhood consists mainly in transferring affective ties from the mother to the father-object" (1953, p. 258). He theorized that this occurred at the phallic stage, when the daughter, angry with the mother because they both lack penises, turns from her to the father as primary love object. Freud saw this turning away (a separation) as a very important step in female development. After the important turning away occurs, however, the daughter becomes concerned that the mother might retaliate against her because of the daughter's rivalry for the father and withdraw her love from the daughter. Therefore, the daughter 12 represses her love for the father and identifies with the mother. The identification with the mother, which was always strong because of her initial position as primary love object, intensifies at this point because of the threatened loss of love. The consequent attachment to the mother is very close. Indeed, Freud speculated that it is so close that many women never abandon it later (1953, p. 253). Helene Deutsch, a widely-recognized psychoanalytic theorist on the psychology of women, fills in many more details of the special- ness of the mother-daughter relationship and the difficulty involved in the separation process. Deutsch summarizes the process of motherhood: ”Women's two greatest tasks as a mother are to shape her unity with the child in a harmonious manner and later to dissolve it harmoniously" (1944-5, p. 294). The attachment between mothers and daughters and its subse- quent dissolution is more difficult than that between mothers and sons for several reasons. First, because of the identification process, mothers can more easily justify close attachment to their daughters in the interest of encouraging greater femininity. On the other hand, mothers are urged to separate from their sons in Deutsch's view because of fears of incest and making the boy a sissy. The attachment between mothers and daughters is, therefore, more acceptable and stronger, and the separation that much more difficult. The relationship between mothers and daughters is also special because it is often based on a wish by the mother that she will "be reborn in her daughter, endowed with all the charm of the new being" (1944-5, 325). This attachment, based on the daughter providing the mother with fulfillment, makes 13 separation more difficult because it appears to some extent to the mother as abandonment of a part of herself. Finally, Deutsch believes that the presence of a daughter reactivates for the mother many of the problems she had with her own mother. This complicates the relationship between the mother and daughter by adding to it aspects of the mother/ grandmother interaction. This complication, of course, makes separation between mother and daughter less clearcut and, therefore, more difficult. Nancy Chodorow (1974), a contemporary theorist interested in family structure and personality development in women, is very concerned about ego boundaries and the mother-daughter relationship. On the basis of clinical evidence, she asserts that separation from mother and the development of a consistently individuated sense of self are particularly difficult psychological issues for Western middle-class women. She summarizes Slater's findings (1961) that "though most forms of personal parental identification correlate with psychological adjustment, personal identification of a daughter with her mother does not" (Chodorow, 1974, p. 59). This is because the identification is often excessive. Chodorow believes that the over- identification leads to blurred ego boundaries. As a result of the boundary confusion, many mothers do not provide experiences of differentiating ego development for their daughters or encourage the breaking of the daughter's dependence. In comparing Western women to those in East London, Indonesia and Java, Chodorow concludes that the major difference is that the others experience a mature dependence based on full differentiation of 14 ego and object, whereas women in the Western middle-class are caught up in issues of infantile independence. Lois Hoffman (1972), in "Childhood Experiences and Achievement," discusses the difficulty of separation from a same sex primary care- taker. She renews Freud's argument that the increased identification with mothers makes it less likely for a girl to establish an early and independent sense of herself. She believes further that establishing the self as separate from mother is easier for boys because they have more conflict with the mother and these conflicted encounters facilitate a sense of separateness. Hoffmann concludes that many girls experience too much maternal nurturance and rapport in the early years and, therefore, find themselves having difficulty facing stress and motivat- ing themselves for autonomous achievement as adults. Block (1973) writes of separation as a sex-role phenomenon. Her theory is that parent-son relationships are based on control, while parent-daughter relationships emphasize relatedness, protection, and support (all of which focus on dependence rather than autonomy). She maintains, furthermore, that characteristics that are essential for individuation and self-expression are defined in our culture as "masculine” and that women are called upon to submerge themselves in communal roles, thereby renouncing both achievement and autonomy. She sees the socialization of women, then, as detrimental to their develop- ment of independence. Donelson (1977) agrees that the socialization of girls and women emphasizes communal at-one-ness with other people at the expense of a sense of separateness and autonomy. Pinches summarizes the effects 15 of socialization nicely: ”. . . women are trained to strive for harmony and fusion whereas men are trained to strive for autonomy" (1978, p. 34). The notion of separation between mothers and daughters is most clearly the domain of Signe Hammer's book, Daughters and Mothers, Mothers and Daughters. In fact, Hammer explicitly devotes two chapters to this topic. The first, "Culture and the Separate Self," focuses on the revolution occurring in motherhood. Hammer notes that, previously, being a good mother seemed to be based on acting as though there was no difference between self and other. The mother's power was, at that time, expressed through her control of the family. Because it would diminish their power base, mothers worked hard to inhibit separation. These old ideas of motherhood are, however, undergoing a revolution, in Hammer's opinion. Instead of encouraging their daughters to assume identities similar to their own, mothers are now working to alter mother-daughter relationships and the ways in which their daughters define themselves as girls and women. Like Freud and Hoffmann, Hammer also emphasizes that the process of separation (so important to self definition) is more difficult for daughters because of the same sex identification, which runs very deeply through the relationship. In "Body and the Separate Self," Hammer focuses on the mother- daughter relationship's effect on sexual identity. She writes of the "mystification' (1975, p. 48) of the female body that the daughter feels because of the mother's ambivalence about her own sexuality. This ambivalence is communicated to the daughter through toilet training procedures, early efforts to inhibit expressions of infantile sexuality, 16 and, sometimes, an overidentification with the daughter's Oedipal conflict. The last, particularly, causes difficulty in a major aspect of the early process of separation, choosing the father as the love object. At another point in the book, Hammer notes that daughters typically turn away from their mothers in early adolescence. The intense rebellion that is often part of the relationship between mothers and their adolescent daughters results, in her opinion, from the closeness caused by the identification and from the nature of the relationship, which is usually confined to the emotional and personal sphere. Finally, Hammer speaks directly about the separation in late adolescence of going away to college. She notes that it is often the first big step toward separation for young women and that, at this point, the separation between mothers and daughters becomes a physical, if not immediately an emotional, reality. She believes that the physical separation is often undertaken in the hope that emotional separation will follow. In her very popular book, My Mother, My Self, Nancy Friday also addresses the issue of separation openly. She defines separation as the freedom to let another be herself (1977, p. 68). ThOugh separation is feared by some because it sounds very final and suggests an end to love, Friday maintains that, in fact, it enables love to develop further. Like Hammer, Friday is convinced that separation is a necessary precondition for true sexuality. She writes: ”To grow into 17 a sexual woman we must fight the person closest to us" (1977, p. 77). Friday believes this process is different for men, who are not raised with similar fears. She concludes: "Sex does not present them with the idea of losing mother” (1977, p. 78). In other important points, similar to those of Hammer's, Friday notes that: 1. Separation is easier for mothers who have a life of their own (in addition to the role of mother). 2. Adolescent girls go through a process of identifying with other women and rejecting their mothers as a part of separation. 3. Undefined boundaries between mothers and daughters are the greatest impasse to separation. 4. And, finally, that "Seeing mother plain, seeing her whole, a mixture of good and bad, is in itself an enormous step toward separation" (1977, p. 402). Normative Research Theorists, then, seem to agree that separation is an important step toward womanhood for adolescent daughters. The results of several normative studies, however, bring that conclusion into question. One of the implications of a study by Schenkel (1973) on field-independence was that women seem better able than men to use the support of their families in forming an identity. In reporting this study, Marcia notes that "they [women] seem to feel less need to separate themselves from their backgrounds and, in general, consider autonomy less of an issue than men do" (1975, p. 105). Marcia further comments that this study 18 is consistent with Douvan and Adelson's (1966) findings that girls tended to have relatively comfortable relationships with their parents and felt little need to change the relationships. Hotch (1979), in her study on home-leaving, also found that self-sufficiency (another expression of separation) was significant as a predictor of style in perceiving home-leaving for males but not for females. In contrast, Jill Allen (1976), in her unpublished dissertation, Identity Formation in Late Adolescent Women, found that a critical, distancing stance toward the mother marked the moratorium phase (Marcia, 1970) of identity achievement. Separation in the form of distancing, then, seemed to her to be an essential step in the formation of identity as a woman. Allen also studied separation-impeding interactions and mothers' perceptions of daughters in her dissertation. She hypothesized that the identity crisis (operationalized through Marcia's statuses) would be most successfully resolved in those mother-daughter relationships that had the fewest separation-impeding interactions. This hypothesis was tested by her subjects' responses to a subset of items from the Interpersonal Perception Method (IPM). The items selected for study involved attitudes and behaviors on the part of both mothers and daughters that would, according to theory, impede separation. Allen's second hypothesis was that the most successful resolutions of the identity crisis would be made by the daughters who were most accurately perceived by their mothers. The mothers' predictions of their daughters' responses to the IPM were used to test the second hypothesis. The results of the study failed to confirm either of these hypotheses, but 19 some conclusions were drawn about a third hypothesis that focused on the degree of mother-daughter identification and the four Marcia identity statuses. Allen defined identification as mother/daughter similarity and measured it as profile similarity on the Interpersonal Adjective Checklist. She found that daughters in the moratorium identity status needed to be critical of and distant to their mothers, and demonstrated low identification with them. In contrast, identity achievement daughters appeared to have stopped the criticizing and distancing processes and to have re-established identification with their mothers. Foreclosure daughters showed high identification and could not, apparently, risk criticizing their mothers. Diffusion daughters experienced low identification with their mothers and a distant attitude toward them. Thus, Allen's findings suggest that there is a relationship between the process of identity formation and the use of a critical and distant attitude as a defense. STATEMENT OF HYPOTHESES In preparing the hypotheses fer this paper, the author focused on the theoretical notions of Hammer and Friday and the empirical approach of Allen. The reason for this choice of focus was the extensive work of these authors with the mother/daughter dyad specifically. One of the central themes running through the literature on adolescent separation in general, and even more strongly through that part of it concerned with mothers and daughters, is that identification plays an important part in the process of separation. The identifications of childhood must be transformed if one is to become an autonomous being, an adult. This transformation seems to be particularly difficult for women because of the intensity of their identifications with their mothers. Identification is, of course, an extremely broad and complex concept. Ferguson (1970) defines it as an ongoing developmental process during which an individual takes on the characteristics or behavior patterns of another person until he/she has actually incorporated aspects of that pattern into his/her own personality. In working with such a large concept, it is essential to narrow it to a manageable size, often by investigating only one aspect of the concept. In this case, the aSpect of identification under study is perceived similarity. The choice receives some support from Donelson, who writes "Perceived 20 21 similarity to a model is a likely manifestation of identification . . ." (1973, p. 457). Donelson describes the function of perceived similarity in identification: "One identifies with those who are salient in one's perceptual field. The salience may be due to the perception of a model as either similar (underlining mine) or powerful to the learner . . ." (1973, p. 483). In measuring identification, then, this study focuses on the perceived similarity of the daughter (the learner) to the mother (the model). Some insight to the transformation of identification (as perceived similarity) in the late adolescent separation process is offered by Allen. Her study revealed that foreclosure daughters, who are by definition not separated from their mothers, remained closely identified with them without ever going through a distancing period. As noted above, however, Allen also found thatidentity achievement daughters, after passing through a distancing phase with their mothers, renewed close identification with them. (As noted earlier, 8105 [1967] refers to this process as "rapprochement.") Furthermore, Allen's investigation of moratorium daughters, those in the process of separating, revealed that they felt little identification with their mothers. Building on Allen's conclusions, it seems that daughters who have not separated (or show low autonomy) and those who have separated (or show high autonomy) would identify more with their mothers than those between the two extremes (who would be assumed to be in the process of separating). This idea is stated as the first hypothesis of this study: Hypothesis 1: Daughters demonstrating low autonomy and those 22 demonstrating high autonomy will show greater identification with their mothers than those in the mid-range. Implicit in the definition of identification as perceived similarity is the important role of perception in the overall process of identification. Donelson writes, "The perception of similarity captures attention and motivates the search for additional similarity," which results in a kind of snowball effect (1973, p. 458). This effect would, it seems, be diminished by perceptions that reveal differences. Since accurate perception of another person should reveal both similarities and differences and requires some distance between the perceiver and the perceived, it should be an important step in defining individuality or separateness. Autonomy would, then, involve accurate perception, "seeing it like it is." Allen comments: it would seem that inaccurate perception would be a necessary result of inadequate separation from that person, because the existence of blurred boundaries between oneself and another person results in a tendency to project one's own feelings and attitudes onto that person (1976, p. 28). Allen chose to focus on the accuracy of the mother's perceptions of the daughter as the prerequisite to separation. It seems to me, however, that the relationship is reciprocal and that it is both the mother and the daughter who must be clear about their boundaries to perceive each other accurately enough for separation. Friday (1977) lends some tangential support to this conclusion in writing that a major step in separation for daughters is perceiving their mothers as whole human beings. Jacobson Speaks of the adolescent's need to tone down the "idealized . . . parental images" in favor of "realistic concepts" as a part of the transformation in identifications that leads to separation (1964, p. 176). Lidz addresses this issue as well when he speaks of the adolescent's task of overcoming "his childhood image of his parents as omniscient and perfect” (1969, p. 110) in order to relinquish his/her dependency on parents. The relationship between perception and separation is expressed as the second hypothesis: Hypothesis 2: The daughters who perceive their mothers most accurately and who are perceived most accurately by their mothers will show higher autonomy than the others. The remaining three hypotheses of this study focus in various ways on the relationship between autonomy and the roles of motherhood and daughterhood. The basic assumption of these hypotheses is that women who largely define themselves in terms of their roles as mothers and daughters will have particular difficulty in separating or expressing autonomy. Donelson comments on this situation in a general way when she notes that people who define themselves in terms of their relationships to others view rejection as "a threat to existence itself“ (1973, p. 482). Since separation appears to involve a process of distancing that could easily be perceived as rejection, it seems reasonable to conclude that individuals whose self-definitions depend on their relationships or roles would experience difficulty separating. This issue has been discussed much more often from the point of view of mothers than from that of daughters. The difficulties that a mother submerged in her role has in adjusting to her children leaving home have been described extensively as the "empty nest syndrome." Both Deutsch and Friday, as cited above, have considered this aspect of the 24 problem. Further discussion is found in Donelson's chapter, "Social Responsiveness and a Sense of Separateness" (1977). To a lesser extent, however, daughters face the same issue. The Douvan and Adelson study (1966) cited earlier showed that autonomy of adolescents could be predicted on a comparison of the amount of free time spent with peers and family. Since time spent with family is a behavioral indicator of the tightness of the family unit, it seems reasonable to conclude that those adolescents spending more time with family would experience greater identification and dependence and, therefore, would define themselves more in terms of the familial role and would have more difficulty separating. Women who have a sense of identity beyond their roles as mothers would, it seems, experience more autonomy than women who do not. Because of this, they would be in a better position to foster the development of autonomy in their daughters and to serve as role models of autonomous women. Consequently, their daughters would also show high autonomy. These ideas will be tested as the third hypothesis. Hypothesis 3a: Mothers who demonstrate high autonomy will more readily foster the development of autonomy in their daughters than other mothers. Hypothesis 3b: Their daughters will demonstrate higher autonomy than the other daughters. A more specific variation of this question relates the autonomy of the mother to a behavioral dimension, employment outside the home. Such employment should, it seems, be one means of broadening a woman's definition of herself, thereby making the separation from children somewhat less threatening to her. The assumption of the effect of women's employment on their self-definition receives some support from a study by Birnbaum (1975). In comparing a group of mothers who were faculty members at a large university with a comparably aged group of mothers who had graduated from college with honors but had not pursued further education or a career, she found that the unemployed mothers had lower self-esteem and a lower sense of competence. Furthermore, Gullahorn notes that the employed mother "tends to have more favorable attitudes about relationships with her children than do nonemployed mothers" (1977, p. 269). She believes that this greater satisfaction with interpersonal relationships results from greater satisfaction with the self that is a consequence of employment. This possible relationship between outside employment and autonomy will be tested as the fourth hypothesis. Hypothesis 4: Mothers employed outside the home will foster the growth of autonomy in their daughters more than those who are not. Finally, a behavioral dimension similar to the one employed in the 1966 Douvan and Adelson study will be used to test the daughters' ability to separate. The theoretical notions, again, are that adolescents who spend more time with their peers identify with them more closely and are less likely to view themselves in terms of familial roles. They should, therefore, separate more easily from parents than those whose primary free activities center on the family. The two parts of the fifth hypothesis are the test for this prediction. Hypothesis 5a: Daughters who spent the largest amount of free time during high school with their peers will show greater autonomy than the others. Hypothesis 5b: Daughters who spent the largest amount of free time with their families will show less autonomy than the others. In addition to studying the relationship of time spent with family or peers and autonomy, the author is also curious about the autonomy of those adolescents who spent the largest part of their free time alone. This group seems to be more difficult to make a prediction for--a case could be made that their time alone indicates autonomy, withdrawal, or a variety of other attributes. Because of the unpredictability of this group, the issue will be stated as an exploratory question rather than a hypothesis. Exploratory Question: What is the relationship (if any) between time spent alone during adolescence and autonomy? METHOD Subjects The aim was to have a sample of lOO'freshman women at Michigan State University and their mothers participate in the study. As discussed later, 101 mother/daughter pairs were actually studied. The qualifications for the participation of the daughters in the study were that the subject: (1) was l7, 18, or 19 years of age; (2) was living in a residence hall on campus; (3) had been living with both of her natural parents prior to leaving for school; and, (4) had not lived away from her mother previously for a significant time period (defined as three months or longer). The first two qualifications were an attempt to restrict the level of maturity and to ensure the normative nature of the subjects. The last two qualifications were viewed as necessary to control for earlier separations from the mother that would very likely influence the current experience. The sign-up sheet for the daughters labelled the study "Mother/Daughter Relationships." This title probably restricted the sample, though it is difficult to evaluate precisely how. It seems likely that subjects who felt they had good relationships with their mothers would be drawn by such a title. However, it also seems possible that daughters with problematic relationships might have been interested in the opportunity to "state their side of the story" or, 27 28 even, to learn something that might be helpful. It seems least likely that freshmen who were indifferent about their relationships to their mothers participated in the study. The study was conducted at the end of the term and the timing may also have produced a biased sample. Many students (particularly the more conscientious, forward-thinking ones) might have completed their experiment participation requirements by that time. This sample may, then, be biased to favor a less active, conscientious student group. Instruments The major instrument was Lorna Smith Benjamin's Structural Analysis of Social Behavior Questionnaire. The questionnaire is based on a model (see Figure 1) that describes the structure of social behavior in a way similar to the models of Leary (1957) and Schaefer (1965). The basic dimensions of the Benjamin model are affiliation (love-hate) and interdependence (emancipation-control). The model can be used to measure the interaction between any dyads. It has been used to study differentiation failure, developmental changes in parent-child interactions, and the interaction between client and therapist. The results of the questionnaire appear as "maps," charts that indicate which items were most salient for the test—taker. An average score is also available for affiliation and autonomy. Finally, the measure provides a means for quantifying identification (in the form of perceived similarity of behavior) and complementarity of behavior (if X does action 1, does Y respond with its complement?). The 29 autonomy and identification scores were employed in this study. In addition, a correlation that Benjamin does not employ, but that is believed by the author of this study to be a measure of accuracy of perception, was studied. The model is presented as three diamond-shaped surfaces. The diamond shape is chosen (in preference to Leary's and Schaefer's cir- cumplexes) because Benjamin feels that it is conceptually "more parsimonious" and because it "allows the poles of the axes to be more salient than they would be if the surfaces were circles" (1974, p. 397). Each of the three surfaces represents a focus of the interaction. The top surface is named "focus on other" and represents parentlike behaviors--what is done to or for the other person. The second surface is "focus on self" and includes behaviors prototypically regarded as childlike--what is done to or for the self. The third surface, called the "introject," describes intrapersonal behaviors and attitudes (in contrast to the interpersonal nature of the two above). Points on this third surface were named by deducing what would happen if parentlike behaviors charted on the first surface were directed toward the self. In other words, the introject is assumed to result from the taking in of the experience of others (especially parents) to one's self. The interpersonal experience with others is, then, trans- formed into this intrapsychic way of being with one's self. The surface is indicated in the code number as the hundreds' digit: focus on other behaviors are the 1005, focus on self are the 2005, and focus on introject are the 3005. 30 In addition, each surface is divided into four quadrants by the intersection of the affiliation and interdependence axes. The quadrants are titled according to the stance they represent. For example, if the focus is on the other and the behavior falls on the love end of affiliation and the independence side of the inter- dependence dimension,the action would be one that "encourages friendly autonomy." The quadrant and pole titles of the model are presented in Figure 2. Please note that for the interdependence dimension, maximum interdependence is at the bottom of each diamond and maximum independence at the top. The bottom half of the first two diamonds represents behaviors saturated with control, either in the sense of dominance (focus on other, controlling the other) or submission (focus on self, being controlled by the other). The four quadrants are labelled by Roman numerals and denoted as the tens' digit in the code numbers. There are also nine topics, which represent subdivisions of the quadrants and are named tracks. The tracks or topics are: primitive basics; approach-avoidance; need fulfillment, contact, nurturance; attachment; logic and communication; attention to self-development; balance in relationship; intimacy-distance; and identity. They are represented in the code numbers as the ones' digits. Any items with the same ones' digit, then, represent the same track or subdivision. The questionnaire consists of items describing each of the chart points (see Appendix A). The subject rates the applicability of each item on a scale from 0 to 100 (O = never, not at all; 50 = sometimes, moderately; 100 = always, perfectly). There are four series. Series A describes the interpersonal behavior of another in 31 terms of the first two surfaces (focus on other and focus on self). An example from the daughters' questionnaire of Series A with the focus on other is "My mother manages, controls, oversees every aspect of my existence." Focus on self is illustrated by "My mother yields, submits, gives in to me." In Series B, the roles are reversed; instead of "he/she," the subject of the items is "I." An example of Series B with the focus on other (again, from the daughters' questionnaire) is "I manage, control, oversee every aspect of my mother's existencei'and an example of focus on self is "I yield, submit, give in to my mother.” Series C is a measure of how the subject treats him/herself rather than of how he/she relates to others. ("I control, manage myself according to my carefully thought-out goals for myself.") It can be rated by the subject him/herself (and is then called the "introject"), or by someone else judging the subject's self-concept. Series D asks that a third party rate the relationship between the members of the dyad. For any dyad, Series D is given twice, reversing the subject and object the second time. This model allows the researcher to do both quantitative and qualitative analyses of the affiliation, interdependence, and identification (in the form of perceived similarity) of dyadic members. The model also provides weighted autonomy and affiliation scores that present a summary statement of the basic thrust or orientation. The weights used in computing the affiliation and autonomy scores are given in Figure 3. Note that items approaching the affiliation and autonomy poles are given progressively greater positive weights while those approaching the disaffiliative and interdependence poles are given 32 progressively greater negative weights. The affiliation or autonomy score is obtained by multiplying the endorsement for each item by its weight as assigned in Figure 3. In this study, data on both affiliation and autonomy were gathered. Because the two relate to each other in a highly complex way, however, it seemed necessary to focus on one or the other for the sake of relative simplicity. The autonomy score was chosen for two reasons. First and foremost, it seemed to be more relevant than affiliation because the issue under consideration is separation. Second, the autonomy score is affected less by a social desirability set than is affiliation in a normal population (Benjamin, 1974, p. 423). With regard to the negative autonomy scores, please note that they indicate dominance if generated from the focus on other surface (for daughters, "My mother manages me" or "I manage my mother") and submission if generated from the focus on self surface (for daughters, "My mother yields to me” or "I yield to my mother").. Negative autonomy scores from the introject surface indicate introjected control ("I manage myself"); while positive autonomy scores indicate intro- jected autonomy ("I listen to and follow what I find deep within myself"). The second instrument was a brief questionnaire designed especially for this study to measure the separation in behavioral terms and to get some impressions about the process from mothers and daughters (Appendices D and E). The Marlowe-Crowne Social Desirability Scale (SDS) (see Appendix B) was also given, but it was not analyzed. This measure is a 33-item 33 true-false scale that was developed as an alternative to the Edwards Social Desirability Scale because of an objection to the pathological content of many of Edwards' items. The authors of the SDS chose items that reflect culturally-sanctioned behaviors that are highly improbable in occurrence. Therefore, the test is more appropriate for use with a normal population than the Edwards. The authors report reliability coefficients of .88 for internal consistency and .89 for test-retest reliability. Procedure The freshman women were contacted through the human subjects pool. They were requested to meet with the researcher in a large room on campus. The researcher first explained that the goal of the study was to investigate the mother/daughter relationship and its effects on women's development. Next, each participant was handed a packet containing a consent form, a copy of Forms A, B, and C of the SASB questionnaire, a copy of the behavioral questionnaire for daughters, and a copy of the Social Desirability Scale. All of these items had been coded to allow for matching with the mothers' questionnaires without involving identification by name. After completing the questionnaires, the subjects were instructed to address an envelope to their mothers. The researcher later placed a cover letter (Appendix C), consent form, a second set of the questionnaires, and a pre-addressed return envelope in the packet to be sent to the mothers. The packets were mailed to the mothers in the envelopes their daughters had addressed. The mothers were asked to return the material in the pre- 34 addressed envelope without a return address. The only possible source of identification of subjects (once the consent forms were separated from the mothers' questionnaires) was, then, the code numbers. A list ofthe participants and their code numbers has been maintained because follow-up research on these subjects is planned. The list is held in confidence in the Department of Psychology. The procedure was approved by the Human Subjects Committee. Measurement of the Variables As indicated above, the SASB provides a weighted autonomy score for each member of the dyad for every surface. Autonomy was, then, operationalized in this study as the weighted autonomy scores for the appropriate surfaces (other, self, introject). High Autonomy was defined as the third of the subjects having the highest weighted autonomy score as specified in the hypothesis. Low autonomy included the third of the subjects having the lowest weighted autonomy scores. Identification was assessed in the form of pgrceived similarity by correlating each daughter's Series A (her view of her mother in relation to herself) and her Series B (her view of herself in relation to her mother) of the SASB (Benjamin, 1979, p. 10). This measure involves a comparison of the similarity the daughter perceives between her own behaviors and her mother's. For example, one item in the correlation compares the daughter's rating of "My mother manages me" (daughter's Series A) with "I manage my mother" (daughter's Series B). If those two items were rated equivalently by the daughter, the correlation would be high. The conclusion would, then, be that the 35 daughter perceives herself and her mother to be similar with respect to this behavior. The overall identification correlation used in this study is based on the daughters' opinions of a series of 72 such paired behaviors. Accuracy of perception, the third variable, is a difficult concept to employ. The difficulty is inherent in the term accuracy. What perception is accurate if all are affected to an extent by the subjectivity of the perceiver? In this study, accuracy of perception was defined as agreemeent of perception--that is, to what extent mother and daughter view the situation in the same way, though from their separate points of view. For the daughter, the accuracy of perception was defined as the degree of agreement between her Series A questionnaire (her view of her mother in relation to herself) and her mother's Series B questionnaire (her mother's view of herself in relation to her daughter). For example, one item in the correlation that represented the daughter's accuracy of perception was the comparison of her rating of "My mother manages me" (daughter's Series A) to her mother's rating of "I manage my daughter" (mother's Series B). A similar accuracy of perception index was calculated for the mother (comparison of her Series A with her daughter's Series B questionnaire). The difference between this variable and identification is that this measure of accuracy of perception involves the views of the individual's performance of a particular behavior by bg£h_members of the dyad. In contrast, the identification correlation involves only the daughters' perception of their own and their mothers' behaviors. 36 Implicit in this design is the use of self-perception as the standard by which the accuracy of the other's perception is measured. That is, the daughter's perception of her mother is compared to the mother's self-perception and the reverse. Although there are clear complications involved because self-perception can as easily be distorted as perception of another and because both mother and daughter can misperceive in the same way, this procedure seemed to be the most appropriate means of approaching the concept of accuracy of perception. Tests of the Hypotheses (See Appendix F for diagrams) For Hypothesis 1, concerning autonomy and identification, the freshmen women were divided into thirds on the basis of their weighted introjected autonomy scores (Daughters' Series C). The identification correlations (obtained from the matching of Series A and Series B for each daughter) of these three groups were compared through an analysis of variance and Eftests. The first relationship between accuracy of perception and autonomy predicted in Hypothesis 2 (that daughters who most accurately perceive their mothers will show greater autonomy than those who least accurately perceive them) was tested by dividing the daughters into three groups--high, medium, and low accuracy of perception. The accuracy of perception was measured as the correlation between the daughters' Series A (my mother in relation to me) and their mothers' Series B (I in relation to my daughter) questionnaires. The introjected autonomy 37 scores of the three groups of daughters were compared through an analysis of variance and £;tests. The second relationship between accuracy of perception and autonomy predicted in Hypothesis 2 (that daughters who are most accurately perceived by their mothers will show greater autonomy than those who are least accurately perceived by them) was similarly tested by dividing the mothers into three groups on the basis of their accuracy of perception correlations. Then, the introjected autonomy scores of the daughters of these three groups of mothers were compared through an analysis of variance and petests. Hypothesis 3a relates the mothers' autonomy to the fostering of autonomy in her daughter ("give autonomy" in Benjamin's terms). It was tested by dividing the mothers into thirds on the basis of the weighted introjected autonomy scores and comparing their weighted autonomy scores on the Series B focus on other level (you in relation to your daughter) and on the daughters' Series A focus on other level (my mother in relation to me) through analyses of variance and Eftests. (Note: This hypothesis involves two separate comparisons based on the mothers' perceptions of themselves and the daughters' perceptions of their mothers.) Hypothesis 3b relates mothers' autonomy to that of their daughters. It was tested by selecting the same groups of mothers as in 3a and comparing the averages of their daughters' weighted intro- jected autonomy scores through an analysis of variance and a Eftest. For Hypothesis 4, mothers were divided into two groups, worked at home only and worked outside the home. These two groups were 38 compared on their weighted autonomy scores for their Series B focus on other level and their daughters' Series A focus on other level, as in the test for Hypothesis 3a. Iftests were employed. The relationship between the behavioral dimension of "time spent" and autonomy that is described in Hypothesis 5 was measured by three tests. First, the group of daughters who indicated that they spent the largest portion of their free time with peers were compared to all others with respect to the weighted introjected autonomy scores. It was anticipated that this group would have higher autonomy scores than the others. Second, the group that indicated that they spent the most time with family was compared to all others with the prediction that their autonomy scores would be lower. Finally, the group that spent the most time alone was compared to the others without a specific prediction. All three comparisons involved Eftests. RESULTS Questionnaire Return The questionnaires were completed by 175 freshman women. Twenty-five of those reSpondents were found to have not fulfilled one of the criteria (see below in Basic Description of Sample) or not completed the questionnaire properly; therefore, 150 questionnaires were mailed to the mothers. Of these 150, 121 were returned, for a response rate of 80.7%. In 20 of these cases, the mothers did not follow the directions or left large blocks of questions (in some cases, entire pages) unanswered. Those questionnaires were eliminated from the study. Questionnaires on which only occasional items were skipped were included in the sample. Benjamin's score for missing data (-1) was assigned to the blank items. The result, then, was an N of 101 mother-daughter pairs. Investigation of the questionaires that were eliminated revealed that they did not differ in any obvious ways from those included in the study. Basic Description of Sample The 101 daughters that comprise this sample met the basic criteria for inclusion in the study. They were all 18 or 19 years of age and freshmen at M.S.U. They lived in a residence hall on campus. The subjects had been living with both of their natural parents before coming to school. None of the subjects had lived away from her mother for more than three months in the past. 39 40 The mean number of siblings for these women is 2.75. Seventy-one of the subjects have at least one older sibling and forty-four have at least one older sister. Two of the subjects are only children. In terms of distance of their parents' home from M.S.U., these freshmen can be grouped as follows: 1 0- 20 miles 10 20- 50 miles 51 50-100 miles 24 100-200 miles 9 200-500 miles 6 more than 500 miles The 101 mothers of these daughters range in age from 38 to 60. Seventy-three of these women have worked outside their homes at some point during their marriage and 91 worked before the marriage. Their occupations varied extensively. According to the Hollingshead occupational classification system, the jobs of the 73 women who have worked after marriage can be grouped as follows: 1 Higher executives, proprietors of large concerns, major professionals 17 Business managers, proprietors of medium-sized concerns, lesser professionals 13 Administrative personnel, small independent businesses, minor professionals 33 Clerical and sales workers, technicians, owners of little businesses 2 Skilled manual employees 6 Machine operators and semi-skilled employees 1 Unskilled employees 41 The Autonomy Scores Some comments about the variable on which much of this study is built, the introjected autonomy score (based on Series C), need to be made before the results of the study are reported. For daughters, the range of the introjected autonomy score was from -103.00 to 37.00 with a mean of -22.88. For mothers, the range was from -105.00 to 74.00 and the mean was -16.57.. The scale theoretically ranges from -250.00 (highest interdependence) to 250.00 (highest autonomy). The negative means indicate that for both daughters and mothers the true state of affairs is one of introjected interdependence, rather than introjected autonomy. In fact, only 24 mothers and 15 daughters reported positive introjected autonomy scores. These findings are difficult to evaluate. There is, of course, an initial question about the validity of the measurement technique. The question is, does Series C actually measure internalized autonomy? At this point, there is no ready answer, but the question remains important and will be further addressed in the Discussion section of this paper. Another possibility, discussed further in the section on norms that follows, is that the measure is valid and that this sample is typical of the population--the conclusion being that most people have introjected interdependence, rather than autonomy. A third possibility, suggested earlier in the Methods section, is that this particular sample is biased to favor interdependence for some reason. As noted above, the method of recruiting subjects may be a source of bias in that the study may have mainly attracted students who felt they had good relationships with their mothers. Many people 42 seem to define good mother/daughter relationships as close ones. Comments like, "That mother and daughter have such a good relationship-- they tell each other everything/do everything together,U are frequently made and illustrate that perspective. Such relationships clearly involve a significant amount of interdependence. If subjects were attracted to this study on the basis of that type of "good” relationship, the reason for the bias toward interdependence is clear. A second possible source of bias for this sample is that the timing of the study at the end of the term may have involved getting subjects who are not "go-getters" (assuming that the "go-getters" would have already fulfilled their experiment participation requirements) and that the sample was, therefore, biased away from autonomy. That does not explain the similar findings of interdependence for the mothers. However, as this study suggests, it seems likely that there is a relationship between the mothers' and daughters' experiences of autonomy and, perhaps, that relationship accounts for the similar findings for the mothers. Testing hypotheses and understanding the results of the tests in a situation where the measure of autonomy is, actually, that of less interdependence could easily result in problems. These potential problems are accentuated in this study because there is no guarantee of the validity of the assumption that autonomy is the opposite of interdependence or that more of one necessarily implies less of the other. In an attempt to reduce the problems that could result from such a situation, two procedures were employed. The originally proposed procedure of testing for introjected autonomy was implemented 43 with the assumption that those scores do reflect relative, if not absolute, autonomy (in the form of less interdependence). The second procedure involved testing another autonomy score, the one that represents the daughter's view of herself in relation to her mother (Daughters' Series B, focus on self). That score was chosen for two reasons. First, it represents the daughter's view of the autonomy/interdependence that she possesses with respect to her mother. For this study on maternal separation, it seems to be a very appropriate choice. It is, perhaps, even more to the point than the originally proposed introjected autonomy score in that the latter is a measure of a general internalized sense of autonomy and the former is directed to the specific object of the separation under study, the mother. This score was not originally chosen as the target because the author's initial interest was in the internalized sense of autonomy and because the daughters' Series B (focus on self ratings) is also involved in other variables in this study (see Methods). Further discussion of the possible relationship between these two forms of autonomy and the theoretical implications of each will be found in the Discussion section. The second reason for the choice of the autonomy in relation to mother score is that it appears to be more clearly a measure of autonomy than interdependence in terms of the scores the subjects obtained. The mean of the daughters' autonomy in relation to mother scores is located on the autonomy side of the axis (11.55). The range of scores also covers a larger part of the autonomy-interdependence dimension (from -57.00 to 143.00 out of a possible -250.00 to 250.00). 44 Basic Data Outline Table 1 presents an overall picture of the findings. It consists of the means and minimum and maximum ratings obtained for the five autonomy scores involved in this study. The possible range for each score is from -250.00 (greatest interdependence to 250.00) (greatest autonomy). As Table 1 indicates and as was noted earlier in the section on the autonomy scores, the mean introjected autonomy scores for both mothers and daughters (-l6.57, -22.88, respectively) fall on the interdependence side of the dimension. The other three mean autonomy scores are located on the autonomy side of the axis. The highest of these mean scores (13.41) is the mothers' view of their autonomy giving behaviors. Next highest (11.55) is the daughters' view of their own autonomy behaviors in relationship to their mothers. The lowest of these scores is the daughters' mean rating of their mothers' autonomy giving behaviors (4.89). Since Benjamin is just in the process of collecting data for norms at this time, there is not much available to set a context around these scores. An early study involving medical students and psychiatric patients was the basis for this statement by Benjamin in the Manual for UsingTSASB: . norms . . . suggest that mothers exerted enough control to average in Quadrant IV (interdependence, positive affiliation) for some groups, and all groups reported a Quadrant IV average level of submissiveness in relation to mother (1979, p. 21). The introject results (Series C) for this particular study are also 45 oo.on oo.nm . ow.e nuocuo co mzoom .< mowpom .muouzwsmov now>mzoo wcw>ww xsocouzm .muonuoe Hmong mo zow> .mnouzwomc oo.ew co.vm . ~v.m~ Apozuo :o msoow .m mowwom .mpocuozv muow>mzon mcw>mm xaocousm Agog» mo zow> .mwozuoz oo.mv~ oo.nm n mm.- Amfiom :o wouom .m mowhom .mpouzwsmov owoom segues ou :owumaou cm >50:o~:m .mwounmsma oo.am oo.moH- ww.-- Au mmfleom .mpooemsaov ouoom xsocousm wouooHOhucw .mwoucwsmo oo.ah oo.mos- em.os- Au mowuom .meoeuozv ououm xeosousm couoonoa»:w .mwozuoz answxmz Ezswcwz :moz mouoom xsocoun< mouoom xaocoua< mo menswxmz paw .waaswcwz .mcmoz oocwmuno H ofinmh 46 found as graphs in the manual and indicate the average scores for this series also fell in Quadrant IV (interdependence, positive affiliation). Thus, the findings of the current study on mothers and daughters, which place the daughters on the autonomy side of the axis for their experience of autonomy in relation to mother and which also place the mothers' autonomy giving behaviors on that side of the axis, are different from these early norms. On the other hand, the findings for introjected autonomy in this study of mothers and daughters are consistent with these early norms. It is difficult to know what to make of these findings since the norms are so sketchy and specific to certain groups (medical students, psychiatric patients). Perhaps the most important suggestion of the norms for this study on mothers and daughters is the possibility that interdependence rather than autonomy is far more likely to be introjected on the average by any group of individuals; the implications being that the women in this study are not unusual in this regard and that such an introject is not limited to women only. The tests of the hypotheses that follow are based on various combinations and sub-groupings of the five mean autonomy scores presented in Table l. Hypothesis 1: The Relationship Between Autonomy and Identification as Perceived Similarity The first hypothesis stated that daughters demonstrating low autonomy and those demonstrating high autonomy would show greater identification with their mothers than those in the mid-range. The 47 classification of autonomy (low, mid-range, high) was made first on the basis of the daughters' weighted introjected autonomy scores (e.g., the score derived from the ratings of such Series C statements as "I listen to and follow what I find deep within myself.”). Three groups of 33, 34, and 34 freshman women were formed. Identification was assessed in the form of perceived similarity by comparing the raw scores of the daughters' Series A (my mother in relation to me--e.g., "My mother manages me") and the daughters' Series B (I in relation to my mother--e.g., "I manage my mother") questionnaires through a Pearson product-moment correlation. The Pearson 335 were then transformed to 5-scores using Fisher's formula. This standardization method was employed to meet the assumptions of normality and homo— geneity for the analysis of variance procedure. Next, a one-way analysis of variance was performed on the transformed 3 scores. The results of that procedure are found in Table 2. As shown in Table 2, the correlations are quite high (.66 to .79) and the difference among the means is highly significant with a probability of .008. Two E-tests were computed next to test the specific predictions of the hypothesis. The first compared the group that had the highest weighted introjected autonomy (Group 1) with those in the mid-range (Group 2). The results of that comparison were not significant. Thus, there was no evidence for the prediction that the group with high autonomy would show greater identification than the mid-range group. In fact, the mean of the second group is slightly = 1.0837 higher than that of the first (2 = 1.0618; Group 2 ’ 2Group 1 .795, r = .786). rGroup 2 = Group 1 48 Table 2 The Relationship between Daughters' Introjected Autonomy and the Identification Correlation ANOVA Mean SD F P Group 1 1.0618 .381 5.074 .008 (High autonomy) (r=.786) Group 2 1.0837 .452- (Mid-range) (r=.795) Group 3 .7965 .402 (Low autonomy) (r=.66l) T-tests Mean t DF P Group 1 1.0618 -2.77 65.11 .0035 Group 2 1.0837 Group 3 .7965 - .22 63.77 NS Group 2 1.0837 The second t-test compared the standardized correlations of the low autonomy group (Group 3) and the mid-range group (Group 2). The results of that procedure were t = -2.77, p = .0035. This comparison of means shows highly statistically significant differences, but in the reversed direction. The mean transformed £_score for Group 2 is significantly larger than that for Group 3 (z = 1.0837, Group 2 .7965; r = .795, .661). 2Group 3 = Group 2 rGroup 3 = The same testing procedure was next employed for the alternate autonomy score, the daughter's view of herself in relationship to her mother (Series B, I in relation to my mother, e.g., "I manage my 49 mother"). The results for both the one-way analysis of variance and the t-tests are summarized in Table 3. Table 3 The Relationship between Daughters' Autonomy in Relation to Mother and the Identification Correlation ANOVA Mean SD F P Group 1 .8963 .5192 .930 NS (High autonomy) (r=.715) Group 2 1.0255 .3878 (Mid-range) (r=.772) Group 3 1.0153 .3700 (Low autonomy) (r=.768) T-tests Mean T DF P Group 1 .8963 -.1292 '59.2 NS Group 2 1.0255 Group 3 1.0153 -.0102 65.9 NS Group 2 1.0255 The hypothesis was not supported in any of the three tests. Inspection of the means reveals that the group of daughters who reported highest autonomy in relation to mother had the lowest mean identification correlation (z = .8963, r = .715). The average correla- tions for the groups reporting low and mid-range autonomy are very close (2 = 1.0153, 2 7 = 1.0255; r Group 2 Group 3 = '768’ r = Group 3 Group 2 .772). U) 5 5., 50 Another approach to the measurement of identification in the form of perceived similarity was taken in the supplement to the SASB. Daughters were asked directly to rate their similarity to their mothers on a six-point scale: l-very dissimilar, 2-moderately dissimilar, 3-slightly dissimilar, 4-slightly similar, 5-moderately similar, 6-very similar. The mean rating was 4.6134. This assessment technique was compared to the SASB-derived identification correlation through a Pearson product-moment correlation. The correlation between the two techniques is .4954, which is significant at the .001 level. The testing procedures that were employed with the SASB measure of identification were also carried out on this self-report measure. The results are summarized in Table 4. With respect to the analyses based on the groups that were defined by the introjected autonomy score, the hypothesis was not maintained in any of the three tests. Though the means are close in size (4.70, 4.68, 4.47), it is the low autonomy group that shows the least identification. The analyses based on the daughters' view of her autonomy with respect to her mother are presented in Table 5 and do, in contrast, show some significant results. First of all, the one-way analysis of variance indicates that there is a highly significant difference among the means (p = .0058). The E-tests reveal that the difference between the means for these high and mid-range groups is significant at the .02 level, but that the difference is in the reversed direction from the prediction--that is, the mid-range group has higher identification than the high autonomy group (4.71, 4.03, respectively). The second test The Relationship between Daughters' Introjected Autonomy and Self-report Identification 51 Table 4 ANOVA Mean SD F P Group 1 4.70 .2866 .272 NS (High autonomy) Group 2 4.68 5120 (Mid-range) Group 3 4.47 .3759 (Low autonomy) T-tests Mean T DF P Group 1 4.70 .0205 98.0 NS Group 2 4.68 Group 3 4.47 .2059 98.0 NS Group 2 shows a trend (p = .107) in the direction predicted for the contrast between the low and mid-range autonomy groups (the low autonomy group shows greater identification than that of the mid-range--5.09, 4.71, respectively). 52 Table 5 The Relationship between Daughters' Autonomy in Relation to Mother and Self-report Identification ANOVA Mean SD F P Group 1 4.03 1.5907 5.439 .0058 (High autonomy) Group 2 4.71 . 1.1423 (Mid-range) Group 3 5.09 1.2152 (Low autonomy) T-tests Mean T DF P Group 1 4.03 - .6756 98.0 .02 Group 2 4.71 Group 3 5.09 .3824 98.0 .107 Group 2 4.71 Hypothesis 2: The Relationship between Accuracy of Perception and Autonomy The predictions of the second hypothesis were: (3) daughters who most accurately perceive their mothers will show higher autonomy than those who least accurately perceive them; and (b) daughters whose mothers most accurately perceive them will show higher autonomy than those whose mothers least accurately perceive them. Three groups (high, medium, low accuracy) were formed for each of the two sets (mothers, daughters) by dividing the 101 subjects into thirds on the basis of the accuracy of perception score. For mothers, this score is a 53 correlation between the mother's view of her daughter (Series A--my daughter in relation to me) and her daughter's self-view (Series B-- I in relation to my mother). The daughter's self-view is, then, the standard of reference and the correlation measures how close the mother's view of her daughter approximates the daughter's view of herself. The correlation is called "accuracy of mothers' perceptions of daughters." For the accuracy of the daughters' perception, the correlations are, similarly, between their own Series A (the daughter's view of her mother) scores and their mothers' Series B (the mother's self-view). As was done for Hypothesis 1, Pearson product-moment correlations were computed and transformed to g_scores. One-way analyses of variance comparing the three groups of each set on the daughters' weighted introjected autonomy scores were computed first. Iftests comparing only the groups of each set that were high and low on accuracy of perception were computed second. The results follow in Table 6. Table 6 indicates that the results of the analysis of variance show no significant differences among the means of the three groups (-l7.15, -22.91, -28.41) based on the accuracy of the daughters' perception of their mothers. However, the E-test comparing the means of the high and low accuracy of daughters' perception groups (-l7.15, -28.41, respectively) does yield significant results (t = 1.68, p = .0485). Therefore, the part of the hypothesis that states that the daughters who most accurately perceive their mothers will show higher autonomy than those who least accurately perceive them was supported. The results of the analysis of variance of the accuracy of the mothers' perception of their daughters (presented in Table 7, below) 54 Table 6 The Relationship between Daughters' Correlated Accuracy of Perception and Daughters' Introjected Autonomy ANOVA Mean SD F P Group 1 -l7.15 19.1 1.432 NS (High accuracy) Group 2 -22.91 ° 26.99 (Medium) Group 3 -28.41 33.46 (Low accuracy) T-test Mean T DF P Group 1 -17.15 1.68 65 .0485 Group 3 -28.41 revealed highly significant differences (p = .0003) among the means (-l4.88, -16.21, -38.00). The_£-tests comparing the groups that were high and low (-l4.88, -38.00) on this measure of accuracy of mothers' perception were also found to be highly significant (t = 3.52, p = .0005). The second part of the hypothesis--that daughters whose mothers most accurately perceive them will show higher autonomy than those whose mothers least accurately perceive them--was also supported. The same set of tests were performed on the alternate autonomy score, the daughters' view of their autonomy with respect to their mothers. The results of the tests are condensed in Table 8 (Daughters' Accuracy) and Table 9 (Mothers' Accuracy). 55 Table 7 The Relationship between Mothers' Correlated Accuracy of Perception and Daughters' Introjected Autonomy ANOVA Mean SD F P Group 1 -14.88 22.33 8.660 .0003 (High accuracy) Group 2 -l6.21 _ 22.79 (Medium) Group 3 -38.00 30.60 (Low accuracy) T-test Mean t DF P Group 1 -14.88 3.52 58.47 .0005 Group 3 -38.00 Table 8 The Relationship between Daughters' Correlated Accuracy of Perception and Autonomy in Relation to Mother ANOVA Mean SD F P Group 1 10.35 29.55 2.369 .0989 (High accuracy) Group 2 3.35 30.66 (Medium) Group 3 20.91 39.26 (Low) T-test Mean t DF P Group 1 10.35 - 1.24 65 NS Group 3 20.91 56 Table 9 The Relationship between Mothers' Correlated Accuracy of Perception and Autonomy in Relation to Mother ANOVA Mean SD F P Group 1 9.29 30.48 .971 NS (High accuracy) Group 2 7.35 . 37.28 (Medium) Group 3 18.21 33.72 (Low accuracy) T-test Mean t DF P Group 1 9.29 - 1.13 63.91 NS Group 3 18.21 Though none of the results is statistically significant, inspection of the means reveals a pattern with this autonomy score that is very different from that for introjected autonomy. As just stated, for the latter, high accuracy of both mothers and daughters was paired with greater autonomy for daughters and low accuracy with less autonomy. In this case, daughters who have the least accurate perception and the daughters of mothers who have the least accurate perception show the greatest autonomy (means of 20.91, 18.21, respectively). Those who rank in the high accuracy groups have the next largest autonomy means (10.35, 9.29, respectively). The lowest autonomy is reported by those in the mid-range (3.35, 7.35, reSpectively). 57 The second assessment of accuracy of perception, similar to the second measure of identification in Hypothesis 1, was done to supplement the SASB. Mothers were simply asked to evaluate the accuracy of their daughters' perception of them by the question, ”How accurate is your daughter's view of you?" They were asked to reply using another six-point scale: l-very accurate, 2-moderately accurate, 3-slightly accurate, 4-slightly inaccurate, 5-moderately inaccurate, 6-very inaccurate. The reply is, then, the mother's estimate of the accuracy of her daughter's perception of the mother. That is, it is the mother's estimate of her daughter's accuragy, not the daughter's estimate of her own accuracy that is being measured. Daughters were asked in a simlar fashion to evaluate the accuracy of their mothers' perceptions of them. Their replies are described below as the daughters' estimate of the accuracy of the mothers' perceptions of the daughters. The two measures of accuracy of perception (SASB correlation and simple report) were compared through a Pearson product-moment correlation. The correlation between the mothers' SASB scores that rated the accuracy of their perceptions of their daughters and their daughters' simple report of the accuracy of their mothers' perception was .3097. For the daughter's accuracy, the correlation was .3896. Both of these correlations are significant at the .001 level. One-way analyses of variance dividing the subjects on the basis of the six simple report accuracy scores were computed for the daughters' introjected autonomy score and the daughters' autonomy in relationship to mother score. The results of these analyses for introjected autonomy follow in Table 10. 58 Table 10 and Daughters' Introjected Autonomy The Relationship between Reported Accuracy of Perception Mothers' Estimate of the Accuracy of the Daughters' Perception of Mother ANOVA Mean N SD F P Group ' -25.71 28 ° 31.33 1.003 NS Group -20.21 62 25.28 Group — 7.00 3 19.08 Group -39.67 3 11.15 Group -50.50 2 62.93 Group -l9.00 l 0 Daughters' Estimate of the Accuracy of the Mothers' Perception of Daughters ANOVA Mean N SD F P Group -l9.00 31 22.36 .803 NS Group -22.78 49 27.05 Group -25.40 10 41.70 Group -47.50 4 38.38 Group -27.00 3 18.36 Group -20.25 4 17.23 As shown in Table 10, neither the comparison based on the daughters' estimate of the mothers' accuracy of perception nor that based on the mothers' estimate of the daughters' accuracy yielded significant results with respect to introjected autonomy. Inspection 59 of the means reveals that there is no clearly consistent pattern, but that higher introjected autonomy scores tend in general to be associated with higher accuracy of perception ratings while low scores are found with the lower accuracy ratings. For instance, for the tests based on the mothers' estimate of the daughters' accuracy, the introjected autonomy scores for the three high accuracy groups are -25.74, -20.21, and -7.00. Those scores for the low accuracy groups are -39.67, -50.50 and -19.00. In contrast, as presented in Table 11 below, the results of the analyses of variance that compared the mean scores of daughters' autonomy in relationship to mother are significant for both the groupings based on the mothers' estimate of their daughters' accuracy of perception and the daughters' estimate of their mothers' accuracy (p = .0018, .0013, respectively). Inspection of these means also reveals an inconsistent pattern, but one in which high autonomy scores tend to be associated with the "very inaccurate" and "slightly accurate” groups, while low scores are found in the ”very accurate" and "moderately inaccurate" categories. For example, for the tests based on the daughters' estimate of the mothers' accuracy, the autonomy in relation to mother scores for the "very inaccurate" and "slightly inaccurate" groups are 77.75 and 11.25, respectively. Those scores for the ”very accurate" and "moderately accurate" groups are 2.65 and 5.67. Tftests were computed next to test the actual differences between the high and low accuracy groups. Because there were so few subjects in each of the three categories representing inaccuracy, the 60 Table 11 Relation to Mother The Relationship between Reported Accuracy of Perception and Daughters' Autonomy in Mothers' Estimate of the Accuracy of the Daughters' Perception of Mother ANOVA Mean 'N SD F P Group 5.18 28 33.24 4.167 .0018 Group 12.16 62 30.72 Group 36.67 3 21.73 Group - 3.67 3 39.83 Group 11.00 2 48.08 Group 143.00 1 0 Daughters' Estimate of the Accuracyyof the Mothers' Perception of Daughters ANOVA Mean SD F P Group 2.65 31 35.43 4.368 .0013 Group 9.86 49 26.68 Group 22.90 10 33.71 Group 11.25 4 20.60 Group 5.67 3 47.82 Group 77.75 4 43.66 61 researcher decided to condense those three groups into one that represented low accuracy. High accuracy was defined as the "very accurate" classification. The results of the t-tests for the daughters' estimate of the mothers' accuracy are presented in Table 12. Table 12 The Relationship between Daughters! Estimate of the Accuracy of the Mothers' Perceptions of Daughters and Daughters' Autonomy Tftest Mean N SD t DF P Introjected Autonomy Group 1 -19.00 31 22.36 1.41 14.96 .0895 (High accuracy) Group 2 -32.00 11 27.58 (Low accuracy) Autonomy in Relation to Mother Group 1 2.65 31 35.43 --l.95 13.95 .0355 (High accuracy) Group 2 33.91 11 48.68 (Low accuracy) As the table indicates, the daughters' estimate of the mothers' accuracy of perception proved to be near significant (p = .0895) in the predicted direction for introjected autonomy. Thus, this form of mothers' high accuracy is associated with greater internalized autonomy among daughters than is low accuracy (means of -19.00 and '32-00, respectively). Furthermore, the daughters' estimate of the mothers' accuracy proved to be significant (p = .0355) for the autonomy 62 in relationship to mother score. It was significant, however, in the reversed direction from the prediction-~those who estimated that their mothers were inaccurate in their perceptions showed greater autonomy in relation to mother than those who estimated that their mothers were very accurate (33.91, 2.65, respectively). As shown in Table 13 below, mothers' estimate of the daughters' accuracy of perception did not have a statistically significant impact, but inspection of the means reveals that these scores are consistent with the pattern for the daughters' estimate of the mothers' accuracy: introjected autonomy is affected as predicted with this form of high accuracy being associated with greater autonomy than low accuracy (-25.71, -39.83, respectively). The opposite is true for autonomy in relation to mother: the daughters whose mothers estimate that they view their mothers inaccurately show greater autonomy than the daughters whose mothers say their perception of their mothers is very accurate (25.67, 5.18, respectively). 63 Table 13 The Relationship between Mothers' Estimate of the Accuracy of the Daughters' Perception of Mother and Daughters' Autonomy Iftest Mean N SD T DF P Introjected Autonomy Group 1 -25.71 28 31.33 1.00 7.33 NS (High accuracy) Group 2 -39.83 6 31.21 (Low accuracy) Autonomy in Relation to Mother Group 1 5.18 28 33.24 - .73 5.54 NS (High accuracy) Group 2 25.67 6 66.73 (Low accuracy) Hypothesis 3: The Relationship Between Mothers' Own Autonomy and Their Abiligy to Give Autonomy Hypothesis 3 is intended to test the effect of the mother's sense of her own autonomy on her ability to "give" autonomy to her daughter and on her daughter's sense of autonomy. The first part of the hypothesis states that mothers possessing high autonomy will more readily foster the development of autonomy in their daughters; the second, that their daughters will demonstrate higher autonomy. For the tests of this hypothesis, the 101 mothers were divided into three groups (high, mid—range, low autonomy) based on their introjected autonomy scores. Four analyses of variance were, then, computed. First, the mothers' senses of their own behaviors of "giving" autonomy were 64 compared in the form of the mothers' weighted score of their Series B focus on other level (you in relation to your daughter). Second, the daughters' views of their mothers' behaviors of "giving" autonomy were compared through the daughters' weighted autonomy score of the Series A focus on other level (my mother in relation to me). Third, the three groups were compared on the basis of the daughters' introjected autonomy scores (Daughters' Series C). Finally, the groups were compared on the alternate autonomy score for daughters, the daughters' view of their autonomy in relation to mother (Daughters' Series B, focus on self). The results of these analyses follow in Table 14. As Table 14 indicates, the results of three of the four analyses of variance were not significant. No significant differences were found in the tests of the mothers' view of their ability to give autonomy, the daughters' view of their mothers' ability to give autonomy, and the daughters' experience of autonomy in relationship to their mothers. The single significant result (p = .003) was found for the test based on the daughters' weighted introjected autonomy scores. The result was in the predicted direction--the daughters whose mothers reported their own greater internalized autonomy experienced greater internalized autonomy. The means of the daughters' introjected autonomy score for the three groups (based on the mothers' high, mid- range, and low introjected autonomy scores) were, respectively, -9.91, -28.l7, and -30.24. Because the original prediction involved comparisons of the group of high autonomy mothers with all others, Ertests using two groups (high autonomy and all others) were also performed on these variables. The results of these tests are found in Table 15. 65 Table 14 The Relationship between Mothers' Introjected Autonomy, Their Autonomy Giving Behaviors and Their Daughters' Sense of Autonomy ANOVA Mean SD F P Mothers' View of Their Autonomy Giving Behavior Group 1 13.09 22.41 .030 NS (High autonomy) Group 2 14.29 22.05 (Mid-range) Group 3 12.79 33.89 (Low autonomy) Dagghters' View of Their Mothers' Autonomy Givipg Behavior Group 1 5.00 27.74 1.950 NS Group 2 - 1.60 29.19 Group 3 11.67 25.95 Daughters' Introjected Autonomy Group 1 - 9.91 23.38 6.134 .003 Group 2 -28.17 26.63 Group 3 -30.24 27.87 Daughters' Autonomy in Relation to Mother Group 1 15.12 32.95 .892 NS Group 2 14.29 35.13 Group 3 5.09 33.72 T-tests Regarding the Relationship between Mothers' High Introjected Autonomy Group and All Others 66 Table 15 T-test Mean SD t DF P Mothers' View of Their Autonomy Giving Behaviors Group 1 13.09 22.41 .09 78.08 NS (High autonomy) Group 2 13.56 28.56 (All others) Daughters View of Their Mothers' Autonomy Giving Behaviors Group 1 5.00 27.74 .03 64.54 NS Group 2 4.84 28.26 Daughters' Introjected Autonomy Group 1 - 9.91 23.38 3.69 72.48 .0005 Group 2 -29.18 27.06 Daughters' Autonomy in Relation to Mother Group 1 15.12 32.95 3.61 .0005 Group 2 9.69 These results did not show significant differences for the mothers' or the daughters' views of the mothers' behaviors of giving autonomy, but did show highly significant differences (p = .0005 in both cases) with respect to the daughters' weighted introjected autonomy 67 score (means of -9.91 for those whose mothers scored high and ~29.18 for all others) and the daughter's view of her autonomy in relation to mother (15.12 and 9.99, respectively). Thus, the first prediction of this hypothesis, that mothers who demonstrate high autonomy will more readily foster the development of autonomy in their daughters than other mothers, was not upheld. The second prediction, that the daughters of these mothers will demon- strate higher autonomy than other daughters, was supported. Further inspection of the results reveals that there is very little difference among the average scores of the mothers' view of their behaviors of giving autonomy (13.09, 14.29, 12.79 for mothers' high, medium and low introjected autonomy groups, respectively). In the daughters' view, the mothers with the least introjected autonomy give the most (11.67) followed by those mothers with the most autonomy (5.00). The daughters' introjected autonomy scores parallel the pattern of the mothers--high for both mothers and daughters, mid-range for both, low for both. The means for the daughters are -9.91, -28.17, -30.24, respectively. The daughters of mothers who report high and medium internalized autonomy have quite high autonomy in relation to mother scores, while 1 those whose mothers report low introjected autonomy tend to have a lower autonomy in relation to mother score (15.12, 14.29, 5.09, respectively). l rm (7) 68 Hypothesis 4: The Relationship between Methers' Employment and their Abilipy to Give Autonomy This hypothesis states that mothers employed outside the home would foster the growth of autonomy in their daughters more than those who are not. For the tests of this hypothesis, the mothers were divided into two groups, those employed outside the home at any time since marriage (Group 1) and those not employed outside the home since marriage (Group 2). These groups were compared on the same Benjamin "give" autonomy measures as in Hypothesis 3a--the mothers' view of their autonomy giving behaviors (mothers' Series B focus on other level) and the daughters' view of their mothers' autonomy giving behaviors (daughters' Series A focus on other). The results of the Eftests on these two groups for these two scores are presented in Table 16. Table 16 The Relationship between Outside Employment and Autonomy Giving Behaviors I:test Mean SD t DF p Mothers' View of Their Autonomnyivimg Behaviors Group 1 13.05 25.10 -.2 42.49 NS (Outside employment) Group 2 14.32 29.82 (Home only) Daughters' View of Their Mothers' Autonomy Giving Behaviors Group 1 4.48 27.64 -.23 46.60 NS Group 2 5.96 29.25 69 The hypothesis was not supported in either test. In both cases, the means of the autonomy giving behaviors of the two groups were quite close (13.05 and 14.32 in the mothers' view, 4.48 and 5.96 in the daughters' view). In each case, inspection of the means reveals that it is the mothers who have not been employed outside the home who view themselves and whose daughters view them as giving more autonomy, which is the reverse of the original prediction. This hypothesis was tested further by a breakdown of mother's employment to ”before marriage" and the following categories after marriage: 1: worked when daughter was aged 0 to 2 2: worked when daughter was aged 2 to 5 3: worked when daughter was aged 6 to 12 4: worked when daughter was aged 13 to 18 5: worked when daughter was aged 18 onwards (Please note that each of these classifications has been treated as a discrete unit and, consequently, a mother who worked only when her daughter was aged 2 to 5 would be found in Group 1 for the test for that age category and in Group 2 for all of the others. Also, please note that the autonomy giving behaviors are measured with respect to the current situation, not the time period(s) during which the mother worked.) Iftests were run on each of these groups and their counterparts of mothers who were not employed outside the home at the time. The results of these tests are presented in Table 17. Members of Group 1 were employed outside the home for all or part of the stated time 7O vo.wm om mH.o N nacho OH. mm.m~ mm.Ht wm.©m HH n~.m u H aono zo~> .mHouzwamc mm.hm om mm.m~ N QZOHO mz wk.AH NH. - om.oH HH oo.4H H ozouu 36H> .mnonuoz N ea c mm: Houswsmo con: poxnoz mm.mm Hm ow.mH N macho «no. NH.NH mm.H- Am.wm oH mo.m H ozouo 36H> .mpoemwsmo Acho oeozv mm.m~ Hm oH.mm N ooouo AucosonQEo oommuzov moo. mo.HH mm.~- m4.m~ oH No.HH H ozouo owmmwpmz opomom 6&6: ecu oewmuao eoonmEm a as H em 2 :66: ouooum 6561 can oufimuso mH6w>msom wcw>wo xsocouz< .muozuoz use ha o~nm& acmeonmsm :oozuon chmcoHumHoz onu mo :oHumocwHoo Hoguasm < 71 Nm.wN Ne NN.NH N ooouo m2 mm.Hw Nm. oo.mN om 4H.4H H 62666 on> .muozooz NH OH NH mm: Houmwsmo cog: poxpoz oo.oN Ho oe.w N Hoouo mNeo. HH.mo mo.H- Ho.mN oe oo. H ooouo 36H> .muounwsmo oo.oN Ho mo.mH N ooouc m2 oo.on NH. ow.mN o4 mo.mH H ozouo 36w> .mhosuoz NH on 0 mm: Houammmo cog: poxaoz HN.NN Hm No.N N ooouo Nmo. ow.mm em.H- oo.NN ON mm.m H ozouo zow> .maounwsmo Nw.NN Hm mN.eH N ozouo m2 em.Hv SN. - NH.oH ON mo.oH H ooouo 36w> .muocuoz m ou N we: Houmwwma :65: poxhoz a no p cm 2 :66: umoutfl H.o.o=oov NH oHHae 72 Nc.om no mw.w N @3090 Nmo. wo.mw Hm.H- 4H.mN em 64.H H ozooo 36H> .mHouawmoo No.NN N4 NN.oH N groom mz HN.oN oH.H- 6N.NN 4m oe.oH H ooouo 36w> .muonuoz mphmzco NH mm: soumwsma con: poxwoz NN.NN N4 No.m N ooouo m2 He.ow NN. - oN.NN om NN.4 H oooou zow> .mHommwwmo o an N am 2 coo: omoonm a.w.u=ouv NH oHnMH 73 period, members of Group 2 were not employed outside the home at that time. As Table 17 indicates, there are only three significant differences. Those are found in the mothers' view of their autonomy giving behaviors in the worked before marriage grouping and in the daughters' perception of their mothers' autonomy giving for the categories of ages 2-5 and 6-12. The one-tailed probabilities equal 1008, .037, and .0475, respectively. Inspection of the means reveals, however, that the results are all in reverse of the predicted direction: those mothers who worked before marriage see themselves as giving less autonomy now than the mothers who did not (11.02, 35.10, respectively). In the daughters' view, the mothers who worked when they were 2-5 years of age and 6-12 years of age also give less autonomy now than those employed at home only (-3.95, 7.07 for the 2-5 group, -.60 and 8.49 for the 6-12 group). The same tendency for mothers employed outside the home to give less autonomy now is noted in the three cases of near-significance: the daughters' views of their mothers' autonomy giving behaviors in the before marriage (p = .074), 0 to 2 years (p = .10) and 18 on (p = .097) categories. Hypothesis 4, then, received no support. Indeed, inspection of the means for the significant findings indicate that, contrary to the original prediction, mothers' employment outside the home during the periods specified (before marriage, daughters' ages 2-5 and 6-12) is associated with less fostering of autonomy in their daughters at present than is employment at home only during those years. 74 Hypothesis 5: The Relationship of Ways of Spending Free Time to Autonomy_ The predictions of Hypothesis 5 were: first, that daughters who spent the largest amount of their free time during high school with peers would show greater autonomy than the others and, second, that daughters who spent the largest amount of free time with families would show less autonomy than others. No prediction was made for the group that spent the most time alone. Two sets of three Eftests were employed to test Hypothesis 5. One set was based on the daughters' introjected autonomy score and the other on the daughters' view of their autonomy in relation to mother. The first test in each set compared those fresh- men who spent the largest amount of free time with friends to all others the second compared those who spent that time with family to all others, and the third compared those who spent that time alone with all others. The results for introjected autonomy are presented in Table 18. Table 18 , The Relationship of Free Time to Introjected Autonomy I-test Mean N SD t DF P Most with Friends -21.26 54 25.71 .63 92.37 NS All Others -24.74 47 29.28 Most with Family —20.37 35 23.21 .72 84.46 NS All Others -24.21 66 29.39 Most Alone -32.27 11 38.59 -.88 11.11 NS All Others -21.73 90 25.70 75 As the table indicates, the predictions of this hypothesis were not supported for any of the tests of introjected autonomy. An inspection of the means reveals that the average weighted introjected autonomy scores for daughters who spent the most free time with friends was highest (-20.37), followed by those who spent that time with family (-21.26), and, finally, those who spent it alone {-32.27). For the autonomy in relation to mother score (presented below in Table 19), the contrast between the group that spent the most time with family and all others (means of 1.26 and 17.02, respectively) was significant in the predicted direction (p = .02). The contrast for the group that spent the most time with friends and all others (means of 15.83 and 6.64, respectively) nearly reached the .05 level of significance (p = .089). The result for the test comparing those who spent the most time alone to all others (means of 20.27 and 10.49, respectively) was not significant, but that group did show the highest autonomy in relation to mother score (20.27). Table 19 The Relationship of Free Time to Autonomy in Relation to Mother lftest Mean N SD t DF P Most with Friends 15.83 54 32.63 1.36 94.75 .089 All Others 6.64 47 35.08 Most with Family 1.26 35 29.98 -2.37 78.97 .02 All Others 17.02 66 34.84 Most Alone 20.27 11 46.39 .68 11.21 NS All Others 10.49 90 32.27 76 Because the first set of groups was based on only partial use of the information available (the rank of 1), a second approach was developed that more specifically defined the groups using all of the rank orderings. These six groups were defined as: Free Time Spent First Second Third Group 1 family friends alone Group 2 family , alone friends Group 3 friends family alone Group 4 friends alone family Group 5 alone family friends Group 6 alone friends family Twelve of the daughters did not rank order the list; indicating, instead, with one "x" how they spent most of their free time. They were not included in this grouping. Analyses of variance were done on these groups with the dependent variables of daughters' introjected autonomy and daughters' view of autonomy in relation to mother. The results of these procedures are found in Table 20. Again, the predictions of the hypothesis were not supported for the tests of introjected autonomy. Inspection of the means reveals very little difference among the average scores for Groups 1, 2, 3, 4 and 6 (-20.24, -23.80, -23.03, —24.38, -24.11, respectively). The mean of Group 5 (those who spent their time in the order alone, family, friends) is, however, much lower (-69.00). The results of the one-way analysis of variance on the autonomy in relation to mother score approached the .05 level of significance (p = .0804). The means were rank-ordered as follows: 77 Table 20 One-way Analysis of Variance on the Relationship of Free Time to Autonomy ANOVA Mean N SD F P Introjected Autonomy Group 1 -20.24 29 22.19 1.230 NS Group 2 -23.80 5 32.95 Group 3 -23.03 36 25.92 Group 4 -24.38 8 31.78 Group 5 -69.00 2 48.08 Group 6 -24.11 9 34.06 Autonomy in Relation to Mother Group I 2.79 29 28.37 2.047 .0804 Group 2 -12.40 5 40.14 Group 3 -13.00 36 29.73 Group 4 -32.13 8 44.88 Group 5 - 9.50 2 7.78 Group 6 26.89 9 49.11 Groups Ranking M332 friends, alone, family .13 alone, friends, family .89 friends, family, alone .00 family, friends, alone .79 alone, family, friends .50 family, alone, friends .40 78 Inspection of the means reveals that the greatest autonomy in relation to mother was expressed by those groups who spent their free time during high school mainly with friends and alone and, thus, ranked family third. The lowest autonomy was associated with those who spent that time with family and alone, ranking friends in the third position. In conclusion, this hypothesis was not supported with respect to daughters' introjected autonomy, but it did receive some support in the area of daughters' autonomy in relation to mother. In the latter, the prediction that those who spent the most time with family would experience less autonomy than the others was upheld at the .02 level of significance. Further, there was a trend that indicated that those daughters who spent most of their free time with friends would show greater autonomy than the others (p = .089). The results of the tests based on all three rankings lend support to these findings (at non-significant levels) from a different angle. The difference is viewing the results in terms of the ways the daughters spent the least amount of their free time rather than the most. Those findings indicate that it was those daughters who ranked family last that reported the most autonomy and those who rated friends third that reported the least. DISCUSSION The purposecdfthis study is to investigate the process of separation from mothers that takes place when late adolescent daughters leave home for college. The basic question of the study is: What aspects of the mother-daughter relationship affect the amount of autonomy the daughter experiences at this important transitional time? More specific points of interest involve the relationship between the daughters' autonomy and their perceived similarity to their mothers, the accuracy of their perception of their mothers and their mothers' perceptions of them, and the amount of internalized autonomy their mothers experience. Also of interest is the relationship of the mothers' ability to foster autonomy in their daughters to the mothers' sense of internalized autonomy and to their employment experiences. The final focus of the study is the relationship between the daughters' autonomy and the ways they chose to spend their free time during high school. This study, then, focuses entirely on autonomy, excluding any analysis of affiliation. As such, it represents only one slice (though a major one) of the process of separation. In the Methods section, it was noted that the decision to focus on autonomy alone was viewed as necessary to simplify the project. It was also stated that autonomy was selected because of its presumed importance in the process of separation and its lesser susceptibility to social desirability factors. In looking at the major, global processes of separation, identification, and accuracy 79 80 of perception, then, this study focuses on the behavioral manifestations of autonomy as measured by the SASB and the simple report questions. Introjected Autonomy, Autonomy in Relation to Mother, and the Relationship between the Two As noted specifically earlier and evident throughout the Results section, the tests of the hypotheses of this study were conducted on two quite different autonomy scores, the introjected autonomy score and the autonomy in relation to mother score. As a prelude to the discussion of the results of those tests, some effort must be made to understand the concepts the two scores represent and their relationship. The introjected autonomy score is intended to be a measure of an internalized sense of autonomy--it is the person's own view of the autonomous behaviors that he/she currently experiences with respect to himself/herself, e.g., "I feel solid, integrated, together, acceptant of my inner core." In Figure 2, Benjamin notes that it is the "introject of other to self." As described earlier, the introject, results from the taking in of the experience of others to one's self. In other words, the interpersonal experience with others is transformed into an intrapsychic way of being with one's self. Thus, introjected or internalized autonomy (the two terms are used interchangeably in this paper) is quite a complex phenomenon. It is complex because it involves the internalization of many experiences and relationships, of which the relationship to mother, however important, is only one. The initial assumption of this study, that the daughters' introjected autonomy would have some direct or clearcut connection to their relationship to their mothers, appears to 81 be too simple. Furthermore, it is not possible on the basis of the data in this study to define the various contributors to the daughters' sense of internalized autonomy or the size of their contributions. Under- standing this, however, may be of use in explaining the differences in results obtained from using the two autonomy scores. In approaching these differences, it is also important to keep in mind that, for most of the women in this study, interdependence in the form of introjected control, rather than autonomy, has been internalized. That is, the mean scores for both mothers and daughters in this sample fall on the interdependence side of the dimension. As indicated in the Results section, the method of recruitment may have biased these findings. One possibility suggested was that students were attracted to the study because they felt they had good, close relationships with their mothers and that the closeness involved inter- dependence. A second possibility involved the timing of the study and the hypothesis that the more autonomous students may have already filled their participation requirements and, thus, may not have been interested in taking part in this study. In addition to the possible relationship of this finding to subject recruitment, the author wondered initially if the introjected control (interdependence) was related to the traditional Social roles assigned to women. The only norms available at the time of this writing reveal that, in the one study reported, both male and female medical students and psychiatric patients reported behaviors that averaged on the interdependence side of the axis. Thus, the available norms do not point to such a difference. It seems to the author, however, that it would be 82 both interesting and important to investigate any sex-related differences in this form of autonomy. The impact of the findings that interdependence rather than autonomy has been introjected by most subjects in this study and that internalized autonomy is more complex than originally thought is to cast some doubt on the meaning and/or validity of the concept of internalized autonomy (in contrast to introjected interdependence or control), at least as measured in this study by this score. Further questions about the issue will be raised as necessary in the discussion that follows. In contrast to the introjected autonomy score, the autonomy in relation to mother score seems to be simpler and more direct. As noted earlier, it is a measure of the behaviors that the daughter states that she currently engages in with respect to her mother. It is, therefore, a measure of one specific relationship where the object is clearly identified. As such, it seems unlikely that the score that represents this concept would be as influenced by other relationships as the introjected autonomy score is. Autonomy in relation to mother is also more direct than introjected autonomy in that it does not involve the additional step of internalization of the experience of the other. The relationship between these two forms of autonomy is not specifically defined by Benjamin and can, thus, only be speculated about. Looking at the concepts from a very general and long-range perspective, it seems likely that introjected autonomy ultimately encompasses autonomy in relation to mother and all other significant persons. At any 83 particular point in time, however, it appears that the relationship between the two forms of autonomy is highly interactive. In late adolescence, for instance, the number of autonomous behaviors a daughter engages in with respect to her mother would certainly be influenced by the strength of her sense of internalized autonomy. Similarly, the autonomy that she feels with respect to herself would be affected by the autonomy that she has expressed to her mother. It seems, then, that the two forms of autonomy are interactive, but that it is not possible to specify the exact nature of the relationship at any particular point in time--at least not on the basis of the data in this study. Further work in the area of the development of internalized autonomy and the relationship between this internal sense and the autonomy experienced in relating to significant others needs to be done. Specific questions about this issue will be addressed as appropriate in the discussion of the individual hypotheses. Autonomy and Identification as Perceived Similarity As indicated earlier, identification is a complex process involving many phenomena, only one of which (perceived similarity) is under investigation in this study. Perceived similarity is defined in this study as the correlation between the daughter's ratings of her own behaviors and the daughter's ratings of her mother's behavior. As such, it is the daughter's perception of the similarity between herself and her mother. The results of this study do not support the predictions of Hypothesis 1 that the daughters who scored high and low on autonomy 84 would show greater identification (perceived similarity) with their mothers than those in the mid-range. Inspection of the means for both the SASB and self-report measures (summarized in Table 21) reveals some interesting patterns, however. For introjected autonomy, the pattern on both measures is that those in the high and mid-range autonomy groups show greater perceived similarity than those in the low autonomy groups. For autonomy in relation to mother, the pattern is that those high on autonomy in relation to mother show less perceived similarity than those in the mid- range or low autonomy groups. Before commenting on the implications of the results, the issue of the two measures of identification must be addressed. As noted in the Results section, the correlation between the two identification measures (SASB and self-report) is .495 (p = .001). This indicates that about 25% of the variability in the SASB measure is shared with the self- report measure. The two techniques for assessing perceived similarity are, then, tapping the same material to that significant, but quite limited extent. The differences between the measures may be related to the more global and clearly self-report nature of the second. Inferences from the test results need to be made with an awareness of these differences. The predictions of this hypothesis were based, largely, on the results of Jill Allen's dissertation. The theoretical notions behind them were: those daughters who rated high on autonomy would have already separated and would, therefore, be in a "rapprochement" phase with their mothers and would show greater identification (like Allen's identity 85 Nwwo.m . amon.v momo.v Magoo: on :oHumHoz :N xeo:6u:< saw: :oNumuHmwuconH unomoaumHom conv.v mono.o choo.v x56:6p:< wouoonouucH now: :owumowwHucopH uHomoHumHom mmHo.H mmNo.H meow. Hosea: op :oHumHox cH xEocoHa< Law: :oHumHoHHou :oHumonNucopH mooN. ANNO.H NHoo.H Nsoeoo=< oooooHouooH Hon :oHumHoHHou comumomeucoeH azoav m macho nomcmutpwzv N macho Axeocous< :MHIV H moose oHomem> xsocous< ecu :omumuflmwucopu :oozuon mwsmcoHumHom ecu :o mpmob oucmwum> mo mommec< ecu mo mcmoz 6:? Hm manmh 86 achievement group);those low on autonomy would not have separated yet and would, therefore, maintain the higher identification of the pre-separation period (as Allen's foreclosure daughters did); and those in the mid- range would be in the process of separating themselves from their mothers and would, therefore, show less identification with them (Allen's moratorium group). The discrepancy between the results of this study and Allen's may have several causes. First of all, the measurements of identifica- tion are quite different although they are all based on the notion of similarity. Allen used the Interpersonal Adjective Checklist and measured identification as mother-daughter profile similarity. In this study, the identification was assessed entirely from the daughters' perspective. As described above, the identification correlation was the match between the daughter's view of herself and her view of her mother. The self-report assessment of identification was also based on the daughters' perspective only. It was their general View of the similarity between them and their mothers. In addition to the difference in perspective, the measurement techniques also differ in what they measured. Allen used adjectives or attributes that describe personality traits, the SASB identification correlations are based on statements about behaviors, and the self-report is one global rating. A second area that may account for the differences is the subjects' ages. Allen's daughter subjects ranged in age from 19 to 23 and represented all four years of college, while the subjects in this study were all 18 or 19 and college freshmen. The differences in age 87 present several possible reasons for the differences in results. One possibility is that the subjects in this study, who had all arrived at the University nine weeks prior to participating in the study, were still so involved in the separation process that individual differences were masked by the intensity of their common experience or, perhaps, the demands and excitement of adjustment to University life had left them little time to think about the process of separation. The brief nine-week time period may also indicate that these daughters were still in the process of grieving their separation and that autonomy could not develop until a later date when the grieving was completed. Using Schneider's model (1981) of the grieving process, these daughters could have been involved in the "holding on/letting go" stage--where their early reaction to the separation was either to experience the depth of their interdependence (holding on) or to deny the importance of the relationship and focus entirely on their independence (letting go). In either case, the daughter would be unable to experience any true or lasting autonomy. The daughters' paragraphs at the end of the behavioral inventory, which focus on their reactions to the process of leaving home, lend some support to this possibility. Many daughters spontaneously commented that leaving home made them realize how much their families meant to them and how much they missed them. Another large group indicated that they were enjoying the separation because of the freedom and independence they experienced. In summary, the age differences between the subjects in the two studies may account for the different resultsirrthat the daughters in 88 this study were all in the very early stage of physical separation with its attendant excitement and grief. Developing the issue of the age difference further leads to a second, related possibility. The younger age of the daughters in this study may invalidate the assumption that was made about the relationship between the groups of the two studies: that the high autonomy group of this study would be comparable to Allen's identity achievement group, the mid-range to the moratorium group, and the low autonomy group to her foreclosure group. Because of the differences in ages between the two sets of subjects, the groups may be in different stages of the process of separation and, therefore, the findings about autonomy may not be comparable. For instance, the daughters high on autonomy in this study may, because of their younger ages, be in the moratorium phase or the process of separating from mother rather than the identity achievement or rapprochement stage. If that is the case, a second look at the reSults of this study may show more support for Allen's results than originally thought. As noted above, inspection of the means reveals that both tests based on the daughter's view of her autonomy in relationship to her mother follow the same pattern: the medium and low autonomy groups show greater identification scores and the high autonomy groups show much smaller identification scores. (This pattern is observed for both the SASB and simple report scores with respect to autonomy in relationship to mother.) Since the basis for this autonomy rating is the daughter's view of her behavior to her mother, the daughters who rate high on this score may well be the ones who are in the process of separating or 89 distancing themselves from their mothers. This view is based on an idea that the scores depicting the process of separating from mother might be best represented by a curve in which the greatest amount of autonomy in relation to mother occurs at a point when distance from mother is most crucial to the development of a separate sense of self. Those daughters who have already achieved their own sense of identity might show lower autonomy in relation to mother scores because they would no longer require as much distance from their mothers and, therefore, could acknowledge the behavioral aspects of their relation- ships that involve interdependence. The low identification (perceived similarity) scores of the daughters experiencing the greatest autonomy in relation to mother would, then, reflect the process of distancing, the need to see one's self as being separate or different from mother. This result parallels what Allen discovered about the moratorium daughters in her study. They were also the ones involved in the process of distancing themselves and they also showed less identifica- tion with mothers than the other groups. For introjected autonomy, the identification correlation was found to be high and quite close for those in the high and mid-range groups and lower for those in the low group. The three self-report means followed the same pattern, although they were very close in size. Introjected autonomy, then, also relates to perceived similarity in a way that is very different from the original prediction. The discrepancy may be based on the concept of internalized autonomy and its possible lack of meaning for the women in this study. As pointed out earlier, that lack of meaning is suggested by the finding that 9O interdependence is more likely to be internalized than autonomy. The actual relationship between interdependence and perceived similarity may, then, be very different from the one predicted for internalized autonomy. That difference may account for the discrepancy between the predicted and the actual findings. A second possibility is that this measure of internalized autonomy (as is) may only become relevant with respect to this form of identification when it is studied at a more advanced stage, perhaps at the end of the separation process. In that case, the discrepancy between the predicted and the actual results would be due to the point in the separation process that the study focused on. A third possible reason for the discrepancy assumes that the pattern identified does have meaning for the separation process as measured in this study. If so, the implication is that relative autonomy or decreased interdependence in its internalized form depends to an extent on identification at this stage of development. Perhaps identification is an important "bottom line" on which internalized autonomy is built. Those freshman women who report high interdependence may do so because they do not identify with their mothers and are, therefore, lacking a basis on which to build a sense of internal autonomy during the very transitional time of leaving home. Take, for instance, those women who are beginning college and planning profes- sional careers, but whose mothers have not gone to college or had careers. The identification of the daughter with the mother in these cases might be quite low. At the stage of transition in which these women were involved at the time of the study, this lack of identification could be experienced as a real lack in grounding that leaves them 91 feeling quite insecure and, therefore, experiencing themselves as highly interdependent. In summary, the results of this study indicate that strong identification with mother (in the form of high perceived similarity) is associated with high and mid-range introjected autonomy and low autonomy in relation to mother. The most likely reason for the introjected autonomy finding is that identification is an important base line on which this form of autonomy is built. The second finding, regarding autonomy in relation to mother, is best explained by the stage in the process of separation that the daughters in this sample were involved in at the time of the data collection. It seems likely that those experiencing high autonomy were in the process of distancing themselves from their mothers and, thus, perceiving themselves as different or separate from the mothers (resulting in low perceived similarity scores). Autonomy and Accuracy of Perception For the accuracy of perception measure based on the SASB, the results of this study supported both parts of Hypothesis 2 for the daughters' introjected autonomy. The findings were that (a) daughters who most accurately perceive their mothers do show higher introjected autonomy than those who least accurately perceive them, and (b) the daughters of mothers who most accurately perceive them do show higher introjected autonomy than those whose mothers least accurately perceive them. With respect to the daughters' autonomy in relationship to mother score, there were no statistically significant findings, but inspection 92 of the means revealed the opposite effect--1east accuracy by both mothers and daughters was associated with highest autonomy for the daughters. Before commenting on the results of the tests based on the simple report accuracy of perception measure, the ways in which this technique differs from that derived from the SASB need to be considered. First, the SASB accuracy of perception measure compared the daughter's view of her mother to her mother's self-view, while the simple report is the mother's estimate of the accuracy of her daughter's perception (that is, the mother evaluates the accuracy of her daughter's view). In both cases, the mother is used as the standard of reference for accuracy but, in the simple report measure, this use is only implicit. It is implicit in that it is based on the assumption that the mother evaluates the accuracy of her daughter's perception by comparing it to her own view of herself. The simple report measure is also different from the SASB in that it is based on the mother's estimate rather than a comparison of data from both mother and daughter. As such, it is the mother's Opinion of the daughter's accuracy and nothing more. This opinion is, of course, a global impression and that is the final way in which this measure differs from the SASB method. The simple report is one general estimate, whereas the correlation is based on 72 pairs of items. The differences between these two measures of accuracy of perception show up clearly in the correlations between them. The mothers' accuracy ratings show a correlation of .3097, while the correlation for the daughters' accuracy ratings is .3896. Although 93 both are significant at the .001 level, the correlations indicate that only about 9% and 16% (respectively) of the variability of the SASB measure is shared with the simple report measure. Like the instruments used to measure identification, these techniques are, to a large extent, measuring the concept of accuracy of perception in different ways and this must be acknowledged in drawing inferences. The results of the tests based on the simple report technique are, then, as follows. First, the prediction of the hypothesis concerning the daughters' estimate of their mothers' accuracy was nearly supported for daughters' introjected autonomy; that is, there was a trend (p = .0895) that indicated that the daughters who view their mothers to be most accurate in perceiving them experience the highest introjected autonomy. The comparison of the mothers' high and low accuracy groups also was significant (p = .0355) for the daughter's view of her autonomy in relationship to mother, but in the reverse direction; that is, the daughters whose mothers are seen by daughters to be most accurate experience the least autonomy in relation to mother. No significant results were found for the daughters' accuracy, but the patterns were the same: high accuracy was associated with high introjected autonomy and low autonomy in relation to mother. The results of this study with respect to Hypothesis 2 can, then, be summarized as follows: For introjected autonomy, the predictions are upheld as anticipated: high accuracy of perception is associated with high autonomy. (In the four tests that compared high and low accuracy groups, two showed significant results and one showed near-significance.) 94 For the daughters' view of autonomy in relationship to mother, the pattern appears to be the reverse: least accurate perception is associated with high autonomy. (Significance is reached in only one of the four tests.) The differences between the results for the two autonomy scores is intriguing. The results of the tests on daughters' autonomy in relation to mother coincide with Allen's findings about maternal accuracy. She concluded that maternal accuracy of perception was important in the early childhood separation-individuation, but not in late adolescent separation. Furthermore, she hypothesized that late adolescent daughters might "not allow their mothers to know them very well during this period in order to facilitate separation" (1976, p. 92). If this is the case, it appears that maternal inaccuracy of perception plays an important role in the late adolescent separation process. In addition to supporting Allen's hypothesis (albeit, partly at non-statistically significant levels) about maternal accuracy, this study also suggests that the daughters who are least accurate in their perceptions of their mothers are most autonomous in their relationships to them. It seems very possible that this difference is, similarly, a part of separation. For some daughters, distancing from mother may also be achieved by inaccurately perceiving her, for at least a time. Perhaps, during this stage of development, then, daughters both hide themselves from their mothers and hide their mothers, as they really are, from themselves in order to facilitate the necessary separation. 95 The support that the predictions received with respect to introjected autonomy underscores, again, the difference between these two forms of autonomy. As mentioned earlier, there is some question about the introjected autonomy score in general and its meaning for this sample. The introject is mainly of interdependence and that may change what should be expected from the relationship of this score to accuracy of perception. If the introjected autonomy score is meaning- ful, this study shows that greater accuracy of perception by both mothers and daughters is associated with the development of higher internalized autonomy in the daughters. The suggestion in such a case is that internalized autonomy involves accuracy of perception, seeing things/people as they really are. One explanation of this finding is viewing introjected autonomy as the end of the process of separation and, thus, at a more developmentally advanced stage than the part of the process that is associated with high scores for daughters' autonomy in relation to mother. High accuracy of perception would, then, become important after the separation from mother is completed because the inaccurate perception would have served its purpose and could be put aside. This possibility seems consistent with the notion of rapprochement, the coming together after distancing. Perhaps, the process of putting the inaccurate perception aside is, even, part of the rapprochement. The viability of this argument for this study is, however, limited in that the daughters in this sample are more likely to be involved in (than to have completed) the process of separating from their mothers. 96 Perhaps, then, the accuracy of perception is involved as predicted in the process\10\ U1:- r-i—lr-JHH—i "l'l‘Tl'fl'Tl'Ti’Tl Flt-3 ‘r‘l'fl 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Before voting I thoroughly investigate the qualifications of all candidates. I never hesitate to go out of my way to help someone in trouble. It is sometimes hard for me to go on with my work if I am not encouraged. I have never intensely disliked anyone. On occasion I have had doubts about my ability to succeeed in life. I sometimes feel resentful when I don't get my way. I am always careful about my manner of dress. My table manners at home are as good as when I eat out in a restaurant. If I could get into a movie without paying and be sure I was not seen I would probably do it. On a few occasions, I have given up doing something because I thought too little of my ability. I like to gossip at times. There have been times when I felt like rebelling against people in authority even though I knew they were right. No matter who I'm talking to, I'm always a good listener. I can remember "playing sick" to get out of something. There have been occasions when I took advantage of someone. I'm always willing to admit it when I make a mistake. I always try to practice what I preach. I don't find it particularly difficult to get along with loud-mouthed, obnoxious people. I sometimes try to get even rather than forgive and forget. When I don't know something I don't at all mind admitting it. I am always courteous, even to people who are disagreeable. At times I have really insisted on having things my own way. There have been occasions when I felt like smashing things. I would never think of letting someone else be punished for my wrong-doings. I never resent being asked to return a favor. I have never been irked when people expressed ideas very different from my own. I never make a long trip without checking the safety of my car. There have been times when I was quite jealous of the good fortune of others. r-ir-lr-ir-i r-i F 29. F 30. F 31. F 32. F 33. 139 I have almost never felt the urge to tell someone off. I am sometimes irritated by people who ask favors of me. I have never felt that I was punished without cause. I sometimes think when people have a misfortune they only got what they deserved. I have never deliberately said something that hurt someone's feelings. 140 APPENDIX C COVER LETTER TO MOTHERS November 10, 1979 Dear I am a graduate student in psychology at Michigan State University. Currently, I'm working on a research project that focuses on the relationship between mothers and daughters. Your daughter has agreed to participate in this research and has filled out the same forms that I'm sending to you. I'm wondering if you would please take the time (one hour or less) in the next few days to complete the enclosed questionnaires and return them to me. Your daughter received extra credit in her introductory psychology course as a result of her participation in my study. Unfortunately, I have no similar concrete reward to offer you! I'm hoping that you'll agree to participate because of your good will, the knowledge that I'll only be able to use the information that I have if both the mother and daughter respond, and your curiousity about the fascinating relationship you and your daughter are involved in. (By the way, if you are curious about it and would like to have a copy of the general results of my study, please check the appropriate box on the enclosed consent form.) Now, let me describe specifically what I'd like you to do. The first thing is to read and sign the enclosed consent form. Then, it's time for the questionnaires themselves. These can be filled out in pen or pencil. The first three pages take the most time and require a little explanation here. They are basically the same questionnaire, but you're being asked to complete it from different points of view. The first page (both sides) is your view of your daughter in relationship to you. The second page (both sides) is your view of yourself in relationship to your daughter. The first side of the third page asks for your view of yourself. The second side of the third page focuses on what you believe to be your daughter's view of herself. It's a little complicated, but 1 think it will make sense when you start. Some additional details: l. A few times in the questionnaire, you'll see "mother/daughter." For you, the choice is always "daughter." 2. If you have more than one daughter, fill out these question- naires on the basis of the relationship between yourself and your daughter who is a freshman at MSU only. 3. Please do not talk with your daughter about the question- naires until you've completed them. 4. If you are like your daughters, you can expect to get tired of the questions about halfway through. Please continue anyway! After page 3, it goes very quickly. 141 5. Don't spend too much time wondering about any one question. Your first response is usually the most accurate. 6. Be sure you answer every question, even if the answer is only your best guess. 7. Return the consent form and questionnaire to me in the enclosed envelope. One last thing--the results of this study will only have merit in helping us learn about this very important relationship between mothers and daughters if the questionnaires are filled out very honestly. I would like to encourage you, then, to give as accurate a picture of yourself and your relationship to your daughter as possible. I want to assure you that the confidentiality of your responses will be strictly maintained. (Your daughter will not, then, every know anything about your responses.) My procedure for collecting the questionnaires has, by the way, been approved by the University committee concerned with ethical issues in research. Thank you very much in advance for your cooperation. If you have any questions at this time or after you receive the general results, please feel free to contact me at this address and phone. Department of Psychology Snyder Hall Michigan State University East Lansing, MI 48824 1 - 517 - 355 - 9564 Sincerely, /) Eoitfi l i / , . a . Patricia Ponto Faculty Supervision by Elaine Donelson, Ph.D. Important P.S. If at all possible, would you please complete the fbrms before your daughter comes home for Thanksgiving. Thank you!!! 142 APPENDIX D BEHAVIORAL INVENTORY FOR MOTHERS How many children do you have? sons daughters How many of your children are older than your daughter who is a freshman at M.S.U.? sons daughters Did you work before marriage? yes no If yes, for how long? At what occupation(s)? Have you worked outside your home since marriage? yes no If yes, for which years of your daughter's life have you worked? (for example--age 7 to present) What is your age? How similar are you and your daughter? very dissimilar moderately dissimilar slightly dissimilar slightly similar moderately similar very similar How accurate is your daughter's view of you? very accurate moderately accurate slightly accurate slightly inaccurate moderately inaccurate very inaccurate How would you describe the relationship between you and your daughter? very interdependent (both pe0ple are dependent or one is dominant and one submissive) moderately interdependent slightly interdependent slightly autonomous moderately autonomous very autonomous (both people are very independent) How would you rate your relationship to your daughter in comparison to other mother-daughter relationships you've seen? much more loving than most others somewhat more loving than most others slightly more loving than most others about the same as most others slightly less loving than most others somewhat less loving than most others much less loving than most others 4. ON THE REVERSE, PLEASE WRITE A FEW SENTENCES DESCRIBING WHAT YOUR DAUGHTER'S LEAVING HOME TO GO TO MSU HAS BEEN LIKE FOR YOU. 144 APPENDIX E BEHAVIORAL INVENTORY FOR DAUGHTERS How far is your parents' home from M.S.U.? (Check one.) 0-20 miles 20-50 miles 50-100 miles 100-200 miles 200-500 miles more than 500 miles Have you visited home since college started? yes no If yes, how many times? If 29, when do you plan to make your first visit home? Have you spoken with your mother on the phone since you left home for college? yes no If yes, how often? every day more than once a week, but not every day once a week every two weeks once a month less than once a month Did your mother work before she was married? yes no If yes, for how long? At what occupation(s)? Has your mother worked outside the home since marriage? yes no If yes, for which years of your life has she worked? (for example-- age 7 to present) At what occupation(s)? What is your mother's age? How many siblings do you have? brothers sisters How many of your siblings are older than you are? brothers sisters How similar are you and your mother? very dissimilar moderately dissimilar slightly dissimilar slightly similar moderately similar very similar How accurate is your mother's view of you? very accurate moderately accurate slightly accurate 145 slightly inaccurate moderately inaccurate very inaccurate How would you describe the realtionship between you and your mother? very interdependent (both people are dependent or one is dominant and one submissive) moderately interdependent slightly interdependent slightly autonomous moderately autonomous very autonomous (both people are very indepen- dent) How would you rate your relationship to your mother in comparison to other mother-daughter relationships you've seen? much more loving than most others somewhat more loving than most others slightly more loving than most others about the same as most others slightly less loving than most others somewhat less loving than most others much less loving than most others During your high school years, would you say you spent the largest amount of your free time with your family, with your friends, or alone? (Please rank 1, 2, 3 with 1 indicating the largest part of free time.) with family with friends alone What has leaving home to come to college been like for you? (Please write a few sentences.) APPENDIX F DIAGRAMS OF HYPOTHESES It is rather difficult to picture how the SASB is used because of the complexity involved with the three series of the questionnaire and the three foci of the model. For clarification, each of the five hypotheses is presented in a diagram on the following pages. The general diagram to be employed is shown as Figure 4 on the next page. It is important to note that each diamond in Figure 4 results in one weighted autonomy score. For each mother and daughter, then, five measures of autonomy were made. Basically, the four diamonds that go across the top of the diagram (focus on other) represent ability to give or encourage autonomy with reference to mother or daughter. The middle four diamonds (focus on self) measure the ability to take or receive autonomy from mother or daughter. The fifth diamond is an introjected autonomy score, which represents a sense of one's own autonomy without a specific reference to another person. For each hypothesis, both the means used to group the subjects and the comparison made are identified. The score used to group the subjects is indicated by vertical lines through the diamond. Horizontal lines through the diamonds indicate the scores used in comparisons. The correlations involved are represented by segmented lines. KEY l l _. ..._.- Score used to group @ Score used in comparison Correlation .mzucm some so“: mo_uum ago me some cm assuage am an Sena see we o_:zaxo :: me =o>_u on: mezcswac 9:» nose: mommegs cab .eoozuzes 1:: Lagoon Lou Loom omega use ace momaom cacao ego ouez .xczum mmsu cm com: me am .uousuzms x5 .05 ca mc_o_x cu c—omz _ assoc! x: Ga =_o_x _ Louzmzae x: 147 \\ \.\ /. .« —0m :6 \\\\ msucg .uuzoca x2 .02 mouacfi= .Loozuzee xi .oa moasees execs: — Loewe! x: oases: _ Leagues: x2 / \.\/ Sfio :o \\\\ \\ .\\ msucm w m w u m o. 32% Lugswmmm hmzmmz 2moueoz ma cowaau_u.ucap— .m macaw; .w_um>l .m_omxe goon cancel _ ou=:=a _ iota:— .mo;ooa as .03 co ms.o_x .Loasusev As .93 cu mc_o_x . ea c.0«x _ nuance >2 0» p.9mx _ eeusuzep x: u_om /\/ .5 I. / m:uOm I 143 .uosuoa x5 .95 nuances .Leuzuzac >5 .oe meaazee ounces _ assoc: x: ounces _ aeozu:mc x: lli /// Lesa: . - - llll; . :c ./.\\\\ / mzoo; w m « .ac m a .323 mmmmwmmm Logos: >&—:<;_2~m ::>_:U=:; m< ZO—F232€92< zzzzbza ;_:mzo_%<4:z :2? o. _ m_m::boa>: 149 .maa:.o 0:“ eagu Leeccusa no;a_= zosm .__3 wooeuoe umozd x: x_ousa:uue one! cs>_oees; one c:3 1cm x_oueezuoe «mos meozuoa mass. o>.ouao: as: ma0.:a:=c ac:o can m=c_ou_usa: as? .mouoom comuzcuuo; we xumazuum Lassa ac m_ms; 3;» :o m::¢eu sous» cuem cosm>na .o>coe ceuaa_1 0;» =_ amazemoaeou can sue: .az mo_uomv mo>.omeo:u mo 39_> .mauzoce 9:» cu someaflou mac A< moweomv .m:o_um_osecu :uzoL:u 1o>_c>:m :cquzoupo: mo xuauzuoa 0:9 m:= mueaaoz u=__a_m < Lo mouaawumo LLegm .om:msos Lonesome .muogsmzee use no; meseuoa soc: Lou avasmaos mm m_mo;uoax: m_;u =_ .u_omxz ems:=a — Aw \ .Losuos x5 so o_s_x _ Pzii. .mozuce x5 ounces _ . -'|| 1 m >2CZC+:< :z< .05 o» m=_e_x assoc! x: I; xv .oa moaszms Logos: x: 2%“:me owl . my“. 34233.; .uzuu< .maouzmza: .w—umxa amazes _ lily ‘l‘ .aauzuzec x2 as a_a_A _ mascu< .maeeuoz .Lsu;a=e1 x5 sweets _ l xusezuu< .meosuoz 0 III II . II n 'I zc_+;;u::; I- In! ‘1' It .111 O... L. Q .. a z m cm .naosucs you 0155 use mesugazac m_ cem_am:aoo meozuoa m_e:u muouzuaae 1:: .o ae=a_a “yum iced:— .95 o» ma_mmx you:2:e1 x: u_am :c msoo; .95 moumzcs Louzuzec >2 eszuc :o mzuo; moqeom <1 ;: >3<==uz< zzzzéaa ;.:mz:_+<;;x 2:? ”N m—m::&3;>: .xeo:0n:s no;un: onanumzoEop .ancaonosn .~__3 mnouzazat n.9gu usgu .12: anousazau n.o;. en >1o=o~== once s>_u on 0.:e on —__3 nascenze gun: oo:o_nogxo cs: mnseoos ween one m:c_nu_1sna 0:? .monOem xecccnze conson0nn=m .mnongazea nnuzu no moaano>c sen an. ace .monoum nosuc cc mzucm < mo_nem .mnou:a:aa n_o;n guy .moncum no:ua cc azucm m mo_nom =20 n_o:n A_~ .monoum xio=cnao ponuoqcnucm n_s:u "on ouozmon :u_3 conaalso eozu one mzzonu eon:u och as m_m=: as» :6 mL30nm eon;n cuen =01M>nc one anognOE oz» .mnmasncax; m_:u no; .m_amxs umecza _ .no:uoe LE on 1_ann _ \\/ Q .nazuol >6 ounces _ m >1323§2< m>_u Ob >+_a_:< =_u:% :z< >Zozc+z< 233 .mzxzéoz 23:39:: ;_:mZO—é4<:x .0! on ms_owx nogaos x: / .se nuances 353 n2 l oua=as _ azcna so can: uncum .nenzuaoz xi on c_o_» _ .nonzazas x: smegma _ al .5 on:a_m noon ion»:— .95 an ms_e_> nougazee x2 \/ :e :0 ////.\\x mzuom \ .oa moaa=aa nonsusac x: nose: :o mzuc; « mo_nom M=Fr ”m m-m::+OL>: .mnoeno 9:» seen one! aneugazae nne;n z. >5o=cuze mo enronu use nouns» _—_3 use: ego eanmnzo eo>c.:59 mnozuot emcee once a. :cnnu_1ena 0:5 .onoum noznc cc mzucm < mownsm .nnougasse ego 1:5 encum noeno :0 mass» m monnsm .n.9;~c5 a:n "seen .caeu .ona monoum >lczonze we meomnnazaou oak .olc: sen ocnmnso neoe>oqaeo no :o.nno:: men on normee ego >s pocnemnonep one unmoguoz>z m_:n no» uncenu 0:9 .a snzuna .m_om>l .m—um>a oaecmfl — om=:ea _ \ uuoq -onu:_ ./\ .noznos >5 .95 on m1.o_> censuses >5 .95 cu m3_e_> on e_o.> _ no;>05 >2 cu cdoq> _ nsuzuzmc >2 < < O / J .noznos >5 .05 newscae neugazaa >5 .05 mommcee omaeez _ nocnoa >2 amazes _ noozmzms >2 \. ///x \M.- nosuo . , =0 \\\\ azuo; -- .l m a 82% =| < L 5529...? xiv C... >._._.:=< 2.2:... =z< ._.z”..2>3._._2..._ .mzm=_._.02 zm..._z._.u.=_ L:_mzc_._.<._m_z 3:... ”v m—mn__._b._>: .9:o_= olnn eon» Lo ozzasa amounmn 9:u uc9zn 6:3 emceo new 9155 an :o_u9_son: oz .mnosuc use ems» >5occnzs mas. seen ___3 >__55& :n_3 sen» umoe ecu u:9:m oer omega was» 5:: mnegno on. cezn >eo:cn:e nencsna seem ___z mn99: :nnz 95mg umc5 osn neoam 9;: 9ncgn neg» one mzcmn9naon: 0:5 .9>oee 59.9num:___ we maneum >50:¢o:= =9u99HOnn=_ n_9:n :o sensasco seen one >925 .o5_o eons acnccozm new :99_c;u nmesn mo mmmsa can :c aozzsnu one mnonzmaac .m_mo;no;>: .ec_u 925 no; .a on:u_u .2_0m>5 .u_oa>5 ouncea _ saezea _ \Rnlr \//// uuum \illll'l rHH/ \ -Onn=_ \ \ _ .n9;nc5 >5 .95 o“ m3_9_> duneuzmv >5 .95 on ma_9_> on 1_9_> _ n9;noe >2 on 2_om> _ nonsuzma >2 .\/ . . .. < . 0 ”mm .n9:ac5 >5 .95 nowezs5 .nonngec >5 .95 mome=a5 oum5=5 _ nognoa >2 ouecc5 _ nonsmcac >2 x\ nezuc co \\\\ mzuog m .5. a m m a 82% >2620+=< Ch 22.? mzz; cz—czaam 2c w>: