INTER—GENERATIONAL FAMILY RELATIONS ASSOCIATED WITH THE PSYCHOLOGICAL ADJUSTMENT OF COLLEGE STUDENTS A Disser’taflon for the Degre‘e of DI). D. MICHIGAN STATE UNIVERSITY Edward C. Teyber I977 -. v.‘ V“ .1" I. J ;_.-.-' Ill;umlzglgululllluwll"will: ~ e y . v v‘1‘3.’ '3"!— 5 1-51“; I” (”Mn“- ." I‘Vllbikgsfill WMc: l O A {m University This is to certify that the thesis entitled Inter-Generational Family Relations Associated with the Psychological Adjustment of College Students I presentedrby Edward CharlesiTeyber has been accepted towards fulfillment of the requirements for Ph.D. degree in PS [CATO-log! ,— AN.M (M R ~LL’LC; M'K ajor professor Dategwfi “‘2“? T 7 0-7639 ABSTRACT INTERFGENERATIONAL FAMILY RELATIONS ASSOCIATED WITH THE PSYCHOLOGICAL ADJUSTMENT OF COLLEGE STUDENTS By Edward C. Teyber The purpose of the present study was to explore if different types of sub-group formations and relationship patterns in the family were associated with the psychological adjustment of college-aged offspring. This approach to studying family functioning is called structural family relations. Family theory postulates that the marital bond is the pri- mary two-person relationship in the healthy family. A body of clinical research and theory, particularly with schizophrenic-member families, suggests that problematic families are characterized by dysfunctional cross-generational relationships and alliances that supersede the mari- tal bond. Theory further suggests that parental failure to establish the marital relationship as the primary dyad in the family reflects unresolved loyalties and bonds to their own parents in their family of origin. Students in two large introductory psychology courses were asked to select the strongest two-person relationship in their family as they were a child growing up, based on the bonds of emotional closeness and involvement. Two groups of male and female subjects from.two, three, or four-child intact families were selected: (1) those who chose the Edward C. Teyber marital dyad as the primary relationship and (2) a group who selected another dyad (grandparent-parent, grandparent-child, parent-child, or child-child) as primary. One hundred and two subjects who met the criteria were interviewed in order to gain their perceptions of the primary relationships and sub-group formations in their families of origin. The Eysenck Personality Inventory (1956) which measures Neuroticism and Extraversion, and Constantinoples' (1969) test of Ericksons' late adolescent-young adult psychosocial crises (Industry, Identity, and Intimacy) was also administered to subjects. Results revealed conditional support for the main hypothesis that offspring who reported the marital dyad as primary would be better adjusted than subjects who reported another non-marital dyad as primary. No significant overall findings were obtained from the MmNOVA; but individual comparisons were performed and revealed the prediction to be supported for females but not males. The strongest findings obtained from the interview concerned the role of the father in the family. In support of the family theory cited, when father was not the primary object for mother, she seemed to turn to one or more children for her primary relationships. This mother-child bond did not expand over time to allow a strong relationship and identification to develop between the children and father, as it did in the marital-relationship-primary group. Finally, sex differences in perception of primary dyads were also obtained, with males more frequently reporting the mother-father and females the mother-child dyad. Edward C. Teyber The results were discussed as centering on the Oedipal triangle. Reasons postulated for the lack of three-generational findings were the normal sample and the self—report method used. The adjustment differ- ences obtained between the two experimental groups for females, but not males, was suggested to be a reflection of the different stage in the identity crisis for males and females at 18 and 19 years of age. Suggestions for improving the present methodology and directions for further research in structural family relations were proposed. INTER-GENERATIONAL FAMILY RELATIONS ASSOCIATED WITH THE PSYCHOLOGICAL ADJUSTMENT OF COLLEGE STUDENTS BY Edward C. Teyber A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1977 ACKNOWLEDGEMENTS I would first like to thank.my chairman and supervisor Lucy Ferguson. She gave me the freedom to do what I wanted to do, and was able to respond so well when I needed help. I feel a great deal of respect and appreciation for her-I have thoroughly enjoyed the time we had together. I would also like to acknowledge my very fine committee. My friends Gary Stollak and Larry Messé have responded to me with affection and respect for five years. I have enjoyed each of them and their families very much. They have also influenced me to become a more scholarly clinical psychologist, something I did not anticipate when I came here. I would also like to thank Don Grummon for his helpful participation on this dissertation. My time here has been very rich. I am taking a lot away with me. Through my relationship with Bill Mueller I became a clinical psychol- ogist, and especially a therapist. I think John Powell is a full man and a christian soul. He has given me a lot mostly through being who he is. And Sam. Sam Plyler taught me a "bunch", but it is best to say we just had such a wonderful time together. Right now it feels very sad to end so much that I love. I have shared my life this past year with two very good friends. See you later, Jan Schreiber and Mike Hollander. Well, California, here I come! ii TABLE OF CONTENTS CHAPTER ' Page I. INTRODUCTION.OCOOOOOOOOOOOOO000......OOOOOOOIOOOOOOOOOOOO 1 Literature Review..................................... 2 The Incest Taboo: A Framework..................... 2 Early Studies of Alliance Patterns in the Family: Schizophrenia................................... 3 A.Mu1ti-generational Framework of Family Relations. 7 The Triad.......................................... 11 Alliances and Therapeutic Interventions............ 15 Social Schemas..................................... 16 Summary and Hypotheses............................. 19 II. WODOOOO0.0.00.00....OOOOOOOOCOOOOOOCOOO0.00.00.0000... 22 subjects 0 O O C O C O O O C O O O O O O O O O O I O O C C O O O O C O O O O O O O O O O O O O O O O 22 measures 0 O 0 O O O O O O O O O O I O O O O O O C O C O O O O I O O O O O O O O O O O O O O O C O O 23 Family Intefliew. O O O O O C O I O O O C O C O O O O O O O O O O O O O O O O I O O O O O O 24 III. RESULTSOOOOOOOO0.0......O...COOOOOCCOOOOOOOOOOOOOOOOOO0.. 26 Description of the Primary Relationships.............. 26 Family Size........................................... 29 Sample Composition for Dependent Measures and Interview.......................................... 30 Adjustment Measures................................... 32 A Non-predicted Finding............................... 36 Family Interview...................................... 37 Iv. DISCUSSIONOOOCOOCOC000......OOOOIOOOOOOOOOOOOO0.0.0.0.... 63 The Mother-Child Dyad................................. 64 Child Adjustment as a Function of Marital Primacy..... 67 'Family Structure...................................... 71 The Father's Role..................................... 74 Conclusions and Directions for Future Research........ 78 BIBLIWWROOOOOOOOOOOOOOOO0.00.00.00.00...00.0.00...00.0.00... 82 iii TABLE OF CONTENTS--continued Page APPENDICES A. QUESTIONNAIREOOOOOOOOOOOOOO0.0.0...OOOOOOOOOOOOOOOOOOOIOO 85 B. PMT I: DESCRIPTION OF SCORING CATEGORIESOC OOOICOIOOOOOO 86 PART II: FAMILY INTHVIEWOOOOOOOOOOOOOCOOOCOOIOOIOOOOOOOO 91 C. CHI SQUARE TABLES FOR MOTHER-CHILD AND FATHER-CHILD WINGS.IO...OOOOOOOOOOOOOOOOOOOO0.00....OOOOOIOOIOOO... 95 D. NON-SIGNIFICANT FAMILY INTERVIEW CHI SQUARE TABLES . . . . . . . 97 / E. SMLE CORHIATIONMTRHOOOOOOOCOOOOOOOOOOOOOOOOOOOOOOOO 114 iv LIST OF TABLES TABLE 1. 2. 3. 5. 6. 7. 8. 10. 11. 12. 13. 14. 15. Dyads Reported Primary by Sex............................... MRNP Female Respondents: Observed Frequency of Primary Dyads by Family $138.0...0..OOOOOOOOOOOOOOOOOOOOIOO0.0.0.... Actual Sample Obtained in Approximation of Original Respondents.OOOOOOOOOOOO0.000000000000000000000.0.0.0.000... Mean, Standard Deviation, and Cell Size for Measures of Mjmmnt.0.0.00.00.00.00...0.00.0000...OOOOOOOOOOOOOOOOOOO Univariate Interaction Effects of Sex x Group for Marginally Simificant Dependent MeasuresOOOOO0.00000000000000000000000 Individual Comparisons of Interaction Effects for Female Subjects.................................................... Univariate Main Effects for Sex for Dependent Measures...... PGF—F Relationship as Reported by Females................... PGMPF Relationship as Reported by Females................... Summary of Chi Square Analysis for Each Grandparent, by Sex Of subjeCtooooo00000000000000.0000000000000000...oooooone... Mother-Third Child Relationship as Reported by Females...... Same and Cross-Sexed Identification Patterns for Oldest Child as Reported by Females................................ Same and Cross-Sexed Identification Patterns for Second ChildaaReported by.m1e8..OOOOOOOCOOOOOOOOOOOO...0.0.0.... Same and Cross-Sexed Identification for Second Children as Reported by FWleaOCOOOOO0.0000000000000000IOOOO0.00.0.0... Same and Cross-Sexed Identification for Fourth Child as Reported by FMIEBOIOOOOOOOOOIOOOOOOOOO0.000....0.0.0.0.... Page 26 3O 32 33 35 35 36 38 39 42 44 46 46 47 48 LIST OF TABLES--continued TABLE 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Presence Reported of Cross-Sexed Identification for Any Child by Fem1e80000OOOOOOOOCOOOOOOO...OCOOOOOOOOC...O... All Children Reported Closer to Mother by Males............. All Children Reported Closer to Mother by Females........... Nuclear Family Seen as a Sub-group Excluding Grandparents as Reported by MaleBOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.0.... One Nuclear Family Member Excluded as Reported by Females... Description of Father by Males.............................. Description of Father by Females............................ Recent Involvement with Family for Males.................... Recent Involvement with Family for Females.................. Changes in Bad Child Role Position for Females.............. Same and Reported Same and Reported Same and Reported Same and Repoyted Cross-Sexed Identification for Oldest Child by m1e800000000O...O0.0....OOOOOOOCOCOOOOCOOOO0.0. Cross-Sexed Identification for Oldest Child by Paul-$00....0....OOOOOOOOOOOOOOOOOOOOOOOO0.0... Cross-Sexed Identification for Second Child by MleBOOO00.0.0000...OOOOCOOOOOOOOOOOOOOOO...O... Cross-Sexed Identification for Second Child by FeMIeBOOOOOOOOOOOOOOOOOOOOCOOOOOOOOOCOOOO0.0... vi Page 49 50 50 53 54 56 56 57 58 59 60 60 61 61 CHAPTER I INTRODUCTION During the last twenty years there has been a shift in psychology and psychiatry whereby important determinants of an individual's behavior are thought to reside in the social context in which s/he lives. There has been a movement away from.studying individuals in the family to viewing the family as a whole social system with group proper- ties superseding individual and dyadic elements. The primary influence of the interpersonal and social environment of the family on individual development has also become apparent. The family is the primary social unit in society, the major source of security for the developing child, and the most powerful socializing agency. Many psychologists believe the family is essential for the child's survival, for his/her learning of adaptive techniques, and for his/her development of personality characteristics through identification and other methods of internaliza- tion. In the study proposed here the social system.of the family is investigated through structural relations within the family. The ques- tion is whether different types of sub-group alignments and inter- generational relationship patterns between family members will be associated with differences in the young adults' level of psychological adjustment. Below, a theoretical framework is introduced and then followed by a review of research on the concepts of alliance, coalition, and split which represent structural family relations. Literature Review The Incest Taboo: A Framework The concept of structural family relations can best be understood when viewed in the context of the two basic axes and functions of famdly life. The generation boundary divides those members of the family who are sexually active with each other and those for whom this is inter- dicted. It divides the family into parents who perform tasks of nurtur- ing, controlling, and decision-making, and children who are dependent and socialized. As the marital dyad broadens to become a parental dyad, a generational division of responsibility, role, and affectional rela- tionships must exist between parent and child. The second axis isthe division of sexes in the famdly. The same sexed parent serves as a roledmodel for identification while the opposite sexed parent provides the basic love object. In the family the son first learns that he is masculine and how to become a boy and a man, and a girl that she is feminine and how to become a girl and a woman. Effecting and maintain- ing the generation and sex boundaries is considered the most important task of the family. In effect, this paradigm is the establishment of the incest taboo (Fleck, 1966; Click and Kessler, 1974; and Lidz et al., 1957a,b). It is implied in this paradigm that the marital bond is the pri- mary relationship in the well-functioning family. This is the main thesis of this study and will be elaborated further. The paradigm of sex and generation boundaries also provides a theoretical framework onto which the various alliance and coalition groupings that are to be dis- cussed can be fitted. Early Studies of Alliance Patterns in the Family: Schizophrenia In the late 1950's investigators were studying the family of the schizophrenic. Several papers were written which used alliances and coalitions to describe the interpersonal processes they observed in these families. In one of these early theoretical/clinical articles, Hynne (1961) noted that alliances and alignments, splits and alienations, were phenomena observable in all persisting groups, including families. He thought of alliances and splits as "structural points of reference" which represented a new focus on the social System aspects of families. Alliances were defined as "the perception or experience of two or more persons that they are joined together in a common endeavor, interest, attitude, or set of values and that in this sector of their experience they have positive feelings toward one another. A split is here de- fined as a comparable perception or experience of opposition, differ- ence, or estrangement, with negative feelings. The alignments and splits within a social system define, to a considerable degree, the emotional organization of the system" (p. 96). The thesis of the paper was when an alliance has developed within a given family therapy group, look for an emerging split at another level or in another part of the group. In this regard, alignments and splits are seen as functional in the hemoeostatic maintenance of fami- lies as social systems. In his schizophrenic-member families Wynne reported that the structure of alignments and splits seemed to shift in a bewilderingly rapid fashion such that the meaning of any particular alignment or split remained confusing. wynne related this unintelligi- ble shifting of alliances within the schizophrenic family to the concepts of pseudo~mutuality and pseudo-hostility; differences between family members were being blended. Haley (1959) was also studying the family of the schizophrenic and described coalition processes in similar terms to wynne (1961). In this paper he hypothesized that no alliances were permitted within the family nor were family members allowed to establish an intimate coalition with someone outside the family. He also suggested a family system operating so that if the patient breaks the family rule about forming relationships outside the family, others in the family would break down. In the schizophrenic families he was describing Haley proposed that the individuals were unable to form and maintain an alli- ance of two against one. If mother and father were in acknowledged agreement against son, one parent would have to dissolve it. Mbther and son may then begin complaining about father; as soon as this brief coalition is labeled by one member, the dyad breaks up and shifts again. It is as if an alliance between two of them is inevitably a betrayal of the third person. Haley suggests this is because "they have difficulty functioning in a two-person relationship, and as a result the separa- tion of any one of the three from the others is a particular threat." . ' ‘1 I ‘1- ‘ l . This formulation can be seen as a failure to establish the marital bond as the primary relationship in the family, resulting in a breach of generations; it also suggests why the marital bond fails to form. The major contribution to the conceptualization of alliance patterns comes from Lidz and Fleck's study of the family with a schizo- phrenic member (Lidz, Fleck, Cornelison, 1966; Fleck, Lidz, Cornelison, Schafer, and Terry, 1959). This group keyed on the parental personality patterns that determine the character of the parental coalition. One of their main findings was the chronic marital disharmony among parents of young schizophrenics. These authors discuss the concept that in order to direct properly the personality development of the child, the spouses 2‘ ’ must form a parental coalition, maintain boundaries between generations, ,9; and a here to their respective sex-linked roles. They believe the achievement of a cohesive identity depends upon a reasonably harmonious integration of identification with two parents. Parents who are irrecon- cilable in reality create intrapsychic conflict and can even split the child's personality with alternative "ego" and "superego" formations for each parent. The authors describeweeveralfldefective coalition patterns they have observed. They propose a "schismatic" marital relationship of openly warring spouses that results in a "schismatic family" because the children are also forced to join exclusively with one or the other I parent. One of the major problems for offspring in such a family is that the spouses devalue each other, making it difficult for the child to want to be like the same-sexed parent or to appreciate the opposite— sexed parent as the prototype of a desirable love object. This is hypothesized to interfere with the development of a clear sexual "1 identity, homosexual proclivities are fostered, and incestuous problemsj persist. The child may accurately perceive s/he is more valued by one parent than the spouse. When this happens, both axes of family life are violated. A second distortion of the parental coalition was termed "skewed". The authors described this pattern as rooted in the severe psychopath- ology of one spouse, coupled with a passive appeasing partner who gives in to the other. This lopsided coalition is unsuited to the completion of basic family tasks. The Lidz group brought special attention to the father-child dyad,' noting the coalition pattern of mother and child frequently aligning against father. They also presented another form of defective coalition occurring when one or both spouses have not emotionally fully separated from their families of origin. In this case parental role tasks of decision~making and responsibility are often abrogated to grandparents and a weaker commitment to the marital dyad is found. These alliance patterns are viewed as a deformation in structural family relationships that are presumed to be either a significant causal factor in relation to a schizophrenic offspring, or to reflect family disturbance accomr panying the presence of a schizophrenic child. Click and Kessler (1974, pp. 16-26 and 77-81) summarized the information provided largely by the investigators cited above and offered a schematization of some family coalition patterns. They state that "the core of the family is the marital coalition", which means that the spouses have been able to loosen their ties appropriately from i their families of origin and have been able to develop a sense of their own individuality and self-worth. Their major thesis is that the mari- tal coalition is the strongest pathway in the healthy family with open channels between all other family members. In contrast are various dysfunctional families: schismatic families with a weak marital coali- tion and strong alliances across sex and generation, such as mother to son, and father to daughter; skewed families which have one relativelyuj isolated member and other members aligned; an excessive "generation gap" with strong marital bonds and offspring bonds but little inter-genera- tional interaction; "pseudodemocratic" in which marital coalition and parental roles are not well defined and there is equal importance of relationships; and the "disengaged family" where members are cut-off from each other with little positive interaction. Again, the main theme is the criticalness of the marital bond as the primary coalition in the family which attenuates exclusive and dysfunctional cross- generational and cross-sexed coalitions. These authors designed this schematization for the parent and child generations; a significant addition to their schema is offered by Boszormeny-Nagy and Spark (1973) with a similar framework for explaining the entire continuum of healthy- functional to pathological-dysfunctional alignment patterns in a three- generational analysis of the family. A Multi-generational Framework of Family Relations Boszormeny-Nagy and Spark (1973) provide important elaborations and extensions of the alliance concepts presented so far. They intro- duce the term "loyalty" which is "reciprocal obligations and merits existentially owed between family members" and which is expressed in the form of concrete services to family members or may be combined with ‘ emotional attachments and involvement. The authors state that "the issue of loyalty fabrics in families is closely connected with those of alignments, splits, alliances, and sub-group formations, often discussed in the literature of family therapy". Boszormeny-Nagy believes alliances as Wynne (1961, see above) defines them are functional and important in family life but "more significant relational dimensions of family align— ment are based on guilt-laden loyalty issues as they are affected by the balance of reciprocal obligations and merits" (p. 38). Thus, these authors present structural family relations but believe they are defined by loyalty on a "deeper" level than the "emotionaleexperiential" levels which are used by wynne and others who discuss alliance. A question of the depth of different alliance bonds is raised here, but again patterns of family relations or alignments are proposed as fundamental to under— standing family processes. Vertical loyalty commitments owed*to a previous or subsequent generation keep conflicting with horizontal commitments owed to one's \ mate, siblings, or peers in general. Boszormeny-Nagy and Spark (1973) state "A very important deep-seated paradox lies in the antithetical relationship between individuation and family loyalty" (p. 51). A.wish for "everlasting symbiotic togetherness" is balanced in every family against the developmental task of separation. New loyalty commitments are necessitated with marriage and then again with birth. The question becomes: "Whom do you choose?" The authors discuss many clinical phenomena as representing a conflict between each spouse's unresolved loyalty to the family of origin and loyalty to the nuclear family. The: ./ 7 _J marital relations constitutes a loyalty commitment to the family of TX failure of a child to develop autonomy and devote himself to peer and origin--one which is overtly deplored in the family but covertly valued. In this regard they conceptualize a child‘s symptoms as a mask for loyalty conflicts between generations. The goal of their therapy is to try and re-balance the ledger of obligations that has accrued between family members and obtain a new sense of fairness or justice in the family relationship system that allows for relatedness with separateness. Thus, these authors use the term loyalty which they believe con- veys a deeper subjective emotional bond than does alliance and coalition as others have defined them. Through loyalty bonds they discuss struc- tural family relations and develop more fully a three-generational analysis of family patterns. Rather than focusing on schizophrenic families, they look at broader psychopathological phenomena as well as emphasizing the task of separation for all families. Like others, they propose that the marital bond must be the primary alliance or "loyalty bond" in the family, and explicate the vicissitudes of the transition of loyalty commitments from family of origin to spouse and children. Others have recognized the inter-generational framework and separation issues elucidated by Boszormeny-Nagy and Spark. Summarizing an early observational/clinical study, Bell (1962) stated "disturbed families have been unable to resolve conflicts with the extended kin outside the nuclear family" and that well families have achieved resolu- tion of the problem of ties to extended kin. One of the things many family theorists are stressing, as well as the quality of the separation 10 from family or origin, is viewing the importance of the whole family network process. The sociologist Sussman (1959), for example, concluded in his study of American family relationships that the idea of the iso- lated nuclear family is a fiction. Although he found many changes since 1890, the 1960 family is "closely integrated within a framework of mutual assistance and activity which can be described as an interdepend- ent family system". Similarly, Spark (1975) argues that whether ties with maternal and paternal grandparents are sources of conflict and competition or resources of positive support and constructive influence for the nuclear family, these relationships do exist and are not to be minimized or denied. In a study of normal families, Sweetser (1964) provided evidence supporting the idea that the mother-daughter relationship is the axis around which extended family relations revolve. Although there is pre- sumed relative independence of the nuclear family and general bilater- ality of obligations to kin, she reported cross-cultural data in the literature syntonic with her own finding of stronger ties between married daughters and parents than between married sons and parents. This was shown through exchanges of help during illness and other services, frequency of visits, transmission of family possessions and traditions, and the reported quality of the relationship. Much of this was thought to be based on child-rearing concerns and responsibilities which are carried out by women. 11 The Triad Eight years later, Haley (1967) wrote a second paper further describing coalition and alliance phenomena in families. In attempting to describe a pathological social system he addressed the "Eternal Triangle", which he calls the gnly social relationship ever named in folk speech. Haley says that if we examine the past literature and activities of(§;ngwe can find a triangle taken for granted and never made explicit that describes a pathological social system. Its char- acteristics are: "l. The people responding to each other in the tri- angle are not peers, but one of them is of a different generation from the other two. 2. In their interaction together the person of one generation forms a coalition with the person of the other generation against his peer (a coalition is a joint action of two people against the third person; an alliance is two people together in a common inter- est independent of the third person). 3. The coalition between the two persons is denied" (p. 17). The perverse triangle is one in which the separation between generations is breached in a covert way. The system is pathological if this is a repetitive pattern. Haley notes that in a nuclear family with grandparents, two par- ents, and two children there are 21 triangles in which parents and children can form coalitions across generations. If the triangular groups consist of amiable members the situation is not complex. If the child is at the nexus of conflicting triangles, however, difficulties arise. For example, if a child is at the nexus of two triangles or groups which are in conflict, he may please his mother and maternal grandmother but may concurrently displease his father and paternal 12 grandmother. Haley postulates that one can explain the symptoms of schizophrenia as adaptive to this kind of conflicting set of groups. Like Lidz and Fleck cited above, Haley describes Freud's Oedipal conflict as the prototype of the breach of generations. He notes the pattern of the grandparent-parent cross-generational coalition which is usually phrased in terms of an excessively dependentiparent. In the next generation, the inability of the parents of a disturbed child to maintain a common front to enforce discipline is a reflection of their inability to maintain a separation between generations. A similar breach of generation appears in the case of a disturbed child who associates with his parents but avoids his peers. Haley goes on to make a further hypothesis about cross-generational coalitions. He says the existence of a coalition between a disturbed child and a parent occurs so often in conjunction with a coalition of one of his parents with a grandparent that one might suggest they are inseparable. A breaching of generations with the child will coincide with a breaching at the next generational level. "If such a triangle at one generation always accompanies a similar one at the next generational level, we can suspect a regularity in networks of family relationships where the patterns in any one part of the family are formally the same as those in some other part" (p. 19). This hypothesis is one of the areas investigated in the study proposed here. Bowen (1966) is a family therapy theorist who proposed "differenziw tiation of self" in the marital, parental, and extended family relation- ships in order to get out of the amorphous "we-ness" of the "undifferen- tiated family ego mass". He construes all families and individuals as 13 located along a single adjustment continuum of psychological differen- tiation. Describing family interaCtion, Bowen asserts that "within the family emotional system, the emotional tensions shift about in an order- ly series of emotional alliances and rejections. The basic building block of any emotional system is the triangle" (p. 354). He believes that during calm periods two members of the triangle have a comfortable emotional alliance and the third person is in the "unfavored outsider position". In tension situations, the outsider is in a favored position and both of the emotionally overinvolved ones will make efforts to in- volve the third in the conflict. Bowen calls this "triangling", when tension in a two-person system exceeds a certain level and the conflict is permitted to expand and shift about within a triad. The therapeutic stance is for the therapist to remain emotionally detached; this pre- vents triangling and facilitates the two-person tension resolving. "Triangling" is consonant with the formulations presented so far; an unsatisfactory marital alliance facilitates an over-involved dysfunc- tional coalition with a third member from.another generation. Bowen (1960) also offers a description that epitomizes the role of\\ \ l the marital alliance in the family system that is being discussed. ‘ i From his study of the hospitalized families of schizophrenics he states I "The striking observation was that when the parents were emotionally close, more invested in each other than either was in the patient, the I patient improved. When either parent became more emotionally invested in the patient than in the other parent, the patient immediately and automatically regressed. When the parents were emotionally close they could do no wrong in their "management" of the patient. The patient . 14 responded well to firmness, permissiveness, punishment, "talking it out",* or any other management approach. When the parents were "emotionally divorced", any and all "management approaches" were equally unsuccessful". Thus again, this succinct statement claims it is the structure of family relationships-~especially the primacy of the marital coalition--that influences children the most. The issue of psychological separation from family of origin n25 been entwined with many of the formulations of structural family rela- tions and Bowen (1966) also has some interesting ideas in this regard. He described two different dysfunctional family separation styles. He termed "exploding" a style characterized by little contact beyond for- malities after the child leaves home and a "cohesive" style with very close and continued contact with parents. He has observed that a person who has dealt with the family separation task by "exploding" can often marry a "cohesive" spouse and move into the emotional orbit of his/her family. Two "exploders" can also marry and become very dependent upon each other. In either case, there is no resolution of their emotional attachments to their families of origin. Bowen considers a successful separation to have occurred if individuals can invest themselves more fully in work and social situations and not just with their spouse. The study proposed here will explore whether the manner in which the parents have resolved their relationship with their own parents is repeated with their own children, and if this resolution is associated with the psychological adjustment of the young-adult offspring. 15 Alliances and Therapeutic Interventions In the previous section, alliance or relationship patterns were presented that attempt to describe some aspects of family functioning. Alliances have also been utilized as a target for therapeutic inter- ventions designed to change dysfunctional family systems. For example, Minuchin (1965) attempts to break up pathogenic coalitions by physicaIIyfl\ removing a family member from a therapy session and having them observe 4 other members through a onedway mirror in the presence of a co-therapistaj Satir (1964) uses illustrations of pathogenic coalitions in teaching } QI' family therapy to students. Haley (1967) says: "The various methods of "O 1 family therapy that have appeared would seem to have one factor in I common: a focus on the problem of coalition both within the family and //: between therapist and family members" (p. 25). Haley notes that it is generally assumed to be unwise for a therapist to join one member of a family against another, yet members of disturbed families are "exasperatingly skillful" at provoking a therapist to side with them and at antagonizing him to side against them. He suggests the art of family therapy is to develop ways of taking sides with all members at once, remaining ambiguous, or clearly acknowledging taking sides when this seems appropriate. Camp (1973) has stated as therapeutic objec- tives: a) weaken the existing cross-generational alliance; b) develop 2 a stronger parental coalition; c) assist parents and children to extend and strengthen relationships with peers at their own generation level; and d) strengthen the relationship between the isolated member in a triangulated situation and one of the cross-generational participants (e.g., develop a relationship between isolated husband and son). 16 He believes the "structure of a family is defined by the patterning of the coalitions and splits among family members", and the therapeutic goals he has outlined virtually summarize the material presented by the various investigators above. Zuk (1968, 1971) has written most about the therapeutic use of coalitions and alliances in family therapy. Zuk believes that most family therapy is still conceptualized in dyadic terms reflecting the ties to traditional psychoanalytic beginnings. He stresses the con- ceptual leap to triads with resultant group processes of coalition and mediation and notes that the triadic concept "par excellence" is the coalition. Zuk introduced the "go—between process" as a therapeutic strategy. This characterizes the family therapist as taking and trading the roles essentially of the mediator and the side-taker. He views negotiations occurring between therapist and family in.which parties vie for control. By selectively supporting, enhancing, or interrupting different coalitions, Zuk believes the therapist can obtain the leverage to breakrup pathogenic relating. Although therpeutic styles and orien- tations may vary, there is a consensus among family therapists that effecting changes in patterns of family alignments is the most funda- mental goal. Social Schemes A.well-defined framework of alliance and alignment structures in families has been developed by clinically-oriented researchers. Clinical investigators have presented their observational reports of structural family relations. There has been no direct empirical testing 17 of these clinical and theoretical formulations, however. In this sec- tion two empirical studies are presented that do not directly address the topic of this study-intergenerational and alliance patterns in the family-but which empirically study structural family relations in other contexts. Hamid (1970) explored differences in the way first and later-born children organize and group their own family members. If first-born children feel displaced (dethroned) from their normal maternal attach- ment by their younger siblings, then one would expect to find idiosyn- cratic arrangements when they are asked to place themselves in a spatial relation with the other members of the family. He hypothesized that a) first-barns displace themselves significantly further away from mother than second- and later-horns, b) that first-barns displace them- selves significantly further away from their younger siblings than second- and later-barns, and c) first-borne displace themselves signifi- cantly closer to father than second- and later-barns. One hundred and fifty normal children equally divided by sex (50 first-barns, 50 second- borns, and 50 third- or later-barns, no only children) aged 5-11 years (mean age - 8.6 years) were included. Kuethe's (1962) method for the representation of social groupings by free-figure placement was used. The method requires the subject to place a series of felt figures (mother, father, child/self and a series of other children--brothers and sisters) on a large felt board. Subjects placed a figure labelled self on the board, then figures the experimenter labelled Mother, Father, Brothers and Sisters. Physical distance between figures was recorded. 18 Results supported the hypothesis concerning clear birth order differences in perceived family relations. More first-barns placed theme selves significantly further from mother and closer to father than second- and third-horns. The sibling rivalry hypothesis is also sup- ported in that deposed first—borne also place themselves further away from their younger siblings than second— and third-horns. This study offers a successful method for revealing systematic differences in per- ceived family relations. Gerber and Kaswan (1971) used a family doll placement technique as an index of the separateness and connectedness within the family as this varies over positive and negative emotional themes. Additional indices used were the orientation of dolls in relation to one another and a family grouping schemata category system which depicted the groupings and subgroups of dolls. Physical distance and directness of the doll interaction were assumed to be related to psychological dis- tance in the family. All members of ten intact families (mother, father, two children) which had one child with learning difficulties partici- pated. Subjects placed dolls representing their family so as to repre- sent specified affective states (a loving, happy, worried, sad, and angry family). Unfortunately, there was no control group. . Results for linear distance showed themes with negative emotional connotations elicited significantly more distant placements of doll family members than themes with positive connotations. The doll place- ments reflected more than a psychological distance continuum, however. In loving and happy emotions, the doll family is typically grouped in one unit facing each other. In worried and sad themes there are a l9 considerable number of three-member groupings with one member separate. In angry themes, dolls are typically grouped into two dyads or dolls are isolated from one another. The patterning of interpersonal involvement for each emotion could be described. Dyads in happy and loving themes are focused on each other or in the same direction. Dyads in the worried and angry themes are facing one another or withdrawn. Sad dyads are not actively involved with one another, but withdrawn. Parents were found to represent the family as a more close-knit group than the chil- dren. Finally, the child with learning difficulties had the greatest number of dolls oriented away from one another in the sad and angry themes, possibly reflecting a tendency for children with learning dif- ficulties to withdraw from painful feelings. Again, a method is pre- sented that successfully differentiates family relations. Both of these studies demonstrate differences in the structure of family rela- tions, although they address birth order influences and situational affect rather than enduring alliance and alignment patterns within the family. Summary and Hypotheses Alliances, coalitions, loyalties, and alignments, although variably defined, all represent relatively enduring patterns of rela- tions within the family. They have been viewed by most investigators as forming the most basic structure of family organization. As these ideas about alliance patterns have been developed in the literature, they have necessarily become drawn together with the two inter-related family tasks of boundary maintenance (sex and generation) and 20 psychological separation from family of origin. The literature reveals a coherent set of formulations about structural family relations which have been widely adopted as therapeutic guidelines in family therapy practice (see Camp, 1973; above). However, despite this general con- sensus, there has been no direct empirical testing of these hypotheses. Further, virtually all the work done with these concepts, which I am calling structural family relations, has been done in the area of severe : ‘mental disturbance--schizophrenia. There is very little study of alliance patterns in relation to more prevalent neurotic or other less severe forms of emotional disturbance, or of changes in structural family relations that occur over the course of the family life cycle. The present study addresses these topics. The following predictions were tested: (1) When the marital relationship is perceived as the primary relationship in the family, this will be associated with healthy psycho- logical adjustment of the offspring. When the marital relationship has not been reported as the strongest emotional bond in the family system, the development of the child will be problematic. An inter-related family structure is also hypothesized to occur in conjunction with this central proposition. (2) That is, when the marital relationship is not reported as the primary relationship in the family, there will also be reported both a cross-generational grandparent to parent alliance and a parent to child alliance. Whereas these were the main hypotheses of the study, several other topic areas were explored which were expected to occur in con- junction with hypothesis (1) and to differentiate families in which the 21 marital bond was reported as primary or not primary. The following were proposed to describe the marital-relationship-not-primary group. (3) The manner in which the grandparents and parents have resolved their relationship (a) will be problematic and (b) the particular mode of resolution will be replicated in the parent to child relationship. (4) There will be a pattern of a polarized "all-good" child to whom the other children should look up and an "all-bad" child who will be considered the "black sheep" or problem of the family. (5) There will be more children perceived as identifying with or being emotionally closer to the cross-sexed parent. (6) There will be more cross-generational and extended family alliance and sub-group formations in the family. (7) There will be more problems in living reported for other family members. It will also be explored if there are differences between the sample groups in: (8) The fathers' reported role in the family, (9) Changes in family relationship patterns as offspring have grown up and begun to move away from the family, and (10) Types of extrafamilial parental relationships. These areas are expanded in the description of the interview in the next chapter. CHAPTER II METHOD subjects College-aged students hold a special vantage point for viewing family functioning. Their major developmental tasks are separating psychologically from their family of origin, establishing a more autono- mous identity, and exploring commitments in heterosexual relations that will lead to new family formation. In this regard, college-aged persons may be seen in a transition that places them between families. At this inrand-out phase they are able at least partially to observe and describe the family operations they have experienced. For these reasons a college-aged sample was selected for this study. Subjects for the study were obtained in the following manner. The experimenter asked all students in two large introductory psychology courses (total enrollment was 730 students) to fill out a form asking for their name, sex, phone number, age, marital status, year in school and the age and sex of any siblings. The form continues by noting there are three generations in the family: grandparents, parents and children, and that there are relationships between different generations (e.g., grandmother-mother or mother-child) and within generations (e.g., mother- father or child-child). Each person is asked to report the most basic 22 23 and primary two person relationship in their family as they were a child growing up. Which relationship had the strongest bonds of emotional closeness and involvement? If they are uncertain which relationship was primary, they are instructed they can report "uncertain". Finally, the form asks if the parents have died or been divorced and if so, what age was the respondent at the time of separation (see Appendix A). The investigater planned to collect the responses and select 30 males and. 30 females who reported the marital relationship as primary, were single, had one, two or three siblings (thus no only children or families with more than four children), were a 18 or 19 year old freshman or sophomore, and whose parents had not died or divorced before they were 12 years old. A second group was to be selected having the same characteristics, except that the marital relationship was not reported as the primary relationship in the family. Thus, 120 students who fit these criteria were to be selected, contacted by phone, and asked to participate in a study of family relationships that occur across generations. For their participation they would complete two short personality questionnaires, respond to a family interview, and receive 3 volunteer credits toward their course requirements. A total of 262 questionnaires were returned that met the criteria. Measures Two personality tests were administered (following the Family Interview described below) in order to determine personality differences between the two groups reporting the marital relationship as primary and not primary. Constantinople (1969) has developed an Ericksonian 24 measure of personality development in college students. The test in- cludes 5 items reflecting successful and 5 items unsuccessful resolu- tions of Erikson's first six stages of psychosocial development. The items consist of short phrases or single words and the subject uses a 7-point scale to describe how characteristic or uncharacteristic that item is of him/her. Test-retest r's after 6 weeks ranged from .45 to .81 with a median reliability of .70 (n . 150). For the purposes of this study only half the test was administered so that only the three psychosocial stages of late adolesence were measured: successful and unsuccessful resolution of industry vs. inferiority, identity vs. M i. identity diffusion, and intimacy vs. isolation. It was hypothesized that subjects reporting the marital relationship as the primary rela- M’uw-fl tionship in the family would score higher on successful resolution and ; lower on unsuccessful resolution of these three psychosocial stages than: would the group reporting the marital relationship not primary. As a measure of psychological adjustment, the Eysenck Personality_Inventgry (1956), was also administered to all subjects. This test has college student norms based on a small college sample. Test-retest reliability ranges from .84 to .94 after one year elapsing between tests. The Neuroticism scale provides a general index of anxiety or adjustment that has been correlated with other anxiety measures. Family Interview The experimenter trained four assistants (blind to which of the two selection groups the subject belonged) to interview subjects about 25 the patterns of relationships they have observed in their families. Haley (1967a) argues that family coalitions are covert processes occur- ring simultaneously on many different levels and thus cannot be validly assessed through self-report. On the other hand, in his review of family research methods, Radkin (1965) recommends ". . . what occurs within the confines of the family can be fully studied only from the viewpoint of the individual member's perception of these events and not in terms of some a priori categories of researchers." He cites Ausubel (1964) who has shown with children that it is not what happens to the child, but his perception of acceptance, rejection, etc., that is pre- dictive. Accordingly, this investigater used an interview to obtain self-reports of family alignment and relationship patterns. Thus, this study will only address subjecta' conscious and reported perception of family relationship patterns. Unconscious or covert alignment patterns that may be present were not explored. The interview procedure is presented in Appendix B (Part II) along with a description of the 7 categories (Part I) utilized in the interview. CHAPTER III RESULTS Description of the Primary Relationships As discussed in the previous chapter, 262 questionnaires were returned which met the criteria for the sample. Table 1 below divides those 262 responses into the frequency with which the various possible dyads are reported as primary, by sex. Table 1. Dyads Reported Primary by Sex Percent Dyad Male Female Mother-Father 46.4 35.5 Grandparent-Parent 5.0 1.8 Grandparent-Child 6.0 3.0 Parent-Child 31.3 46.6 Child-Child 11.1 12.8 Total 100.0(N-99) 100.0(N-163) Total N - 262 Mother-father and parent-child are perceived as the most important dyads in the family. These two relationships combined comprise 78% of the male and 822 of the female responses. Dyads which include a 26 27 grandparent were infrequently reported as primary. A sex difference is noted here; males more frequently report a dyad which involves a grand- parent than females (112 males; 4.8% females). The child-child dyad was more apt to be perceived as primary than the grandparent-parent or grandparent-child dyads, but it was also infrequently reported by both males and females. These normative results emphasize that nuclear family dyads involving one or both parents are perceived as the most central relationships in the 3-generational family system. Inspection of Table 1 also reveals a very important sex difference in perception of marital bond. Males most often perceive the mother- father dyad and females the parent-child dyad as the primary relation- ship in the family. In reciprocal fashion, parent-child is the second most frequently reported dyad for males and mother-father for females. 7 ‘This pattern of males reporting the mother-father and females the h-I-ufl—H’fi \ parent-child dyad as primary reflects a fundamental sex difference in \\\/ the perception of family relationships. {A chi square analysis of these ‘patterns of relationships and sex differences in Table l proves nearly significant (X2(4)- 8.5, pfi.10). A further chi square analysis for only the mother-father and parent-child dyads within Table l was then performed and found to be significant (X2(1)- 5'29’.B‘:'°5)' The data in Table 1 show the high frequency with which the parent- child relationship is perceived as primary, especially by females. More detailed questioning during the actual interview revealed that within this parent-child dyad it was the mother-child relationship in particu- lar that was being reported (see Table 1, Appendix C). Thirty-nine percent of the MRNP males and 502 of the MRNP females reported the ”m lam-n.4— M 28 mother-child relationship as primary; and none of the males and 231 of the females reported father-child as primary. For the MRP group, 482 of the males and 482 of the females perceived mother-child as second to the marriage; while only 131 of the males and 72 of the females reported father-child as second to the marriage. It is noted that father is most important for the MRNP females and least important for the MRNP males. Chi square analyses were performed for each of these comparisons and significant differences were obtained only between the MRNP males and females for the father-child dyad, X2 - 4.22, pf=.05. Overall, these (1) results point out the central role of the mother in the family system. In summary, important differences are noted in these perceptions of primary family relationships. Males tend to perceive mother-father as the primary dyad in the family whereas females report parent-child most frequently. The data suggest that when mother and father do not form the basic coalition in the family mother aligns herself with a child to form the primary dyad. The data obtained did not differentiate whether this mother-child dyad referred to the mother-son, mother- daughter, or both relationships, but there was a tendency for females to report the mother-child bond as primary more often than did males. Some differential effects hypothesized to be associated with the relative ranking of these dyads will be investigated below. First, however, the original normative responses will be further studied for differences between the two—, three- and four-child families. 29 Family Size As noted above, one criterion for participation was that the sub- ject came from a two-, three-, or four-child family, since it was thought that only-child and very large families would contain different patterns of family relationships. The breakdown for family size was: 301 of the males and 291 of the females were from a two-child family; 381 of the males and 442 of the females were from.a three—child family; and 321 of the males and 302 of the females were from a four-child family. Chi square analyses were performed in order to determine if there were differences in the choice of primary dyad across the three family sizes, by sex. No significant differences were obtained between males and females in the MRP group for the three family sizes (X2(2)- 1.75, 2):.20). Similarly, no significant differences were found between males and females in the MRNP group for the three family sizes by four possible dyads, X2(6)- 6.16, 2):.20. Further, looking within just male MRNP subjects, no significant differences were obtained for the three family sizes by four possible dyads, X2 - 4.14,.p}’.20. A signifi- (6) cant difference for female MRNP subjects was obtained, however, and is presented in Table 2 on the following page. The significant chi square value observed was largely attributable to two differences. Within the two-child family, there were more grand- parent-child dyads and fewer child-child dyads reported than expected. When there are only two children in a family, female MRNP subjects report that they are not as apt to form a primary bond as children in a three- or four-child family. The results suggest that these two 30 Table 2. MRNP Female Respondents: Observed Frequency of Primary Dyads by Family Size Primary Dyad 2 child 3 child 4 child Total Grandparent-Parent Grandparent-Child 4 l 0 5 Parent-Child 24 34 18 76 Child-Child 0 11 10 21 Total 29 46 30 105 18(6)- 18.8, p_< .01. siblings are more apt to turn to a grandparent than each other in order to form a primary coalition. Although less influential, the reverse pattern can be seen in the four-child families. More child-child pri- mary dyads are found than expected and no grandparent-child dyads are reported as primary. Perhaps with more children coalitions are formed amongst them and they do not need to turn to the grandparents for a pri- mary relationship. The overall results, however, reveal that the perception of primary dyads is relatively constant across the three family sizes. A Sample Composition for Dependent Measures and Interview 4A mmaller group of 120 subjects was selected from the original 262 respondents to be interviewed for the study. Two contrast groups were obtained by selecting 60 subjects (30 males and 30 females) who reported the marital relationship primary and 60 subjects (30 males and 30 females) who reported another (non-marital) dyad primary. In order 31 to generalize from the results back to the original population the 120 subjects selected for the two groups were representative of the distri- bution of family size among the larger sample of original respondents. This was done by selecting 30 males and 30 females who reported the marital relationship as primary in proportion to the number of two-, three- or four-child families obtained in the total sample. Thus the 30 male and 30 female subjects in each of the two experimental groups (marital relationship reported as primary and MRNP) contained approxi- mately equal numbers from two-, three- and four-child families. However, an unexpectedly large number of subjects (231) changed their choice of primary relationship when they were asked to rank the dyads in their family a second time during the actual interview two to seven weeks later. One hundred and thirty-three subjects were interviewed in total; 12 subjects switched away from the marital relationship as primary and 19 subjects switched to the marital relationship as primary during the interview. (A.chi square analysis revealed that 31 was not a sig- nificant number of overall switches, x2 - 1.05,‘p)’.20). These 31 (1) subjects who changed their choice of primary relationship were dropped from.the analysis. Thirteen extra subjects were selected in order to maintain as close an approximation of the original sample as possible. The final sample of 53 females (27 MRP; 26 MRNP) and 49 males (23 MRP; 26 MRNP) is presented in Table 3 on the following page. 32 Table 3. Actual Sample Obtained in Approximation of Original Respondents Family Size 2-child 3-child 4-child Dyad Reported Primary Male Female Male Female Male Female Mother-Father 5 5 12 13 6 7 Grandparent-Parent 3 1 2 l Grandparent-Child 0 2 l 0 Parent-Child 5 4 3 5 Child-Child 0 0 1 2 Total 13 12 23 24 13 15 Adjustment Measures Hypothesis 1 predicted that children whose parents had the most basic and primary 2-person relationship in the family would be better adjusted as young adults than those from families having another (non- marital) dyad as primary. A 2 (Sex) X 2 (Group) multivariate analysis of variance for the nine measures of personality presented above was performed. The multivariate analysis was performed in order to insure that the nine dependent measures were independent of each other. The means and standard deviations for each variable for each of the four groups of subjects are presented in Table 4 on the following page. Constantinoples' (1969) measures 1 to 3 theoretically reflect successful resolution of three of Erickson's late adolescent/young adult stages; measures 4 to 6 represent unsuccessful resolution of these psychosocial crises; and measures 7 to 9 are Eysenck‘s (1956) 2-factor test of Neuroticism (anxiety) and Extraversion, and a Lie scale. aNm.m o~w.e NH».H .n.m 33 co.ma MN.OH mn.~ new: mzmz m 0N e ema.n mso.e mmn.H .n.m mm.NH HH.m oo.~ nods. mm: m um m emu.e nun.¢ coo.H .n.m om.ma em.m on.~ and: mzmz 2 on N mmH.m con.e «we.a .n.m mo.ma on.oa HN.N new: mm: .8 ma H nowmuo>muuxm xuoomhm .amwuwuounoz Monomhu sag susomhm .>sn .vum nacho Mom 2 Hana m m b one: eoaaauuou a «Hana mum. mnw. «aw. nmo. mmo.a ooh. .n.m Hm.N 9N.m co.m NN.m h.¢ o.n dam: mam: h ON a non. now. can. now. has. Neh. .n.m NM.N wm.n an.N cm.n H.n hn.n nods mm: m 5N .m Han. mom. cma. «Ho. son. «an. .n.m Nm.N om.m NN.m m¢.m no.m mm.n anus. mzmz Z on N com. com. Hun. Huh. Nae. Hue. .D.m mh.N nn.m m~.n Hm.n No.n cm.m anuz. mm: 3 MN H soaumaonu seamsmmav mufiuouuouoa homawuou huaunoem unannosH .>oa .eum moose Mom 2 Hana o muauooeH e m N H can: n unmfiumsno< we mourned: now seem dado one .nouumd>mn pudendum .osz .e manna 34 No significant overall effects attributable to Group were obtained from.the MANOVA: Group main effect 2(9,90)- .6554; Sex X Group Inter- oaction F(9,90)- 1.19. Although no significant effects were found to warrant further analyses, the data were explored further. Winer (1972, p. 392) notes that further comparisons should be performed to test hypotheses which are a part of the design. The conclusions made from these further analyses should only be regarded as highly tentative, however. Within this context, the Efratios and p.values for effects attributable to Group (that were in the predicted direction) are pre- sented below. Inspection of the cell means--and to some extent the univariate §_ratios for the Sex X Group interactions--suggested that findings for female subjects (but not males) tended to be in the pre- dicted direction. Although there were no significant Sex X Group interaction effects obtained, two very strong trends are observed in the univariate results for successful resolution of Industry and Intimacy; and three lesser trends are also noted for Identity, Inferiority, and Neuroticism (see Table 5 on the following page). A consistent pattern of results is obtained in all five of these trends. Female subjects tended to support the hypothesis: the MRP group scored better adjusted than the MRNP group on each of the five measures revealing trends. Individual comparisons (l tailed t tests) were performed and found to be significant for four of the five trends. Between group differences were obtained for Industry, Identity, Intimacy, and 35 Table 5. Univariate Interaction Effects of Sex X Group for Marginally Significant Dependent Measures Dependent Measure Z ratio 3 value 1. Industry 3.739 .0561 2. Identity 1.809 .1817 3. Intimacy 3.147 .0792 4. Inferiority 2.574 .1119 5. Neuroticism 2.580 .1115 df - 9/98 Inferiority (see Table 6 below). As noted above, these results should be considered tentatively. The correlation matrix for the MANOVA re- vealed these measures to be in fact correlated (up to .49) and not independent which could account for the cluster of significant results obtained (see Appendix E for sample correlation matrix).- Table 6. Individual Comparisons of Interaction Effects for Female Subjects Dependent Measure t obs. p.value 1. Industry 2.95 p< .005 2. Identity 1.83 p< .05 3. Intimacy 1.89 p_< .05 4. Inferiority 2.27 p< .025 5. Neuroticism 1.34 p< .10 36 In sumary, a pattern of results tended to support the prediction for females but not males. These results will be explored in the dis- " ' W -—4<-—‘e-... cussion section in conjunction with the findings obtained from the family interview. A Nonjredicted Finding Although main effects for Sex were not predicted, two significant main effects were found for Inferiority and Isolation. Males reported the five items comprising the Inferiority scale to be more character- istic of them than did females (X- males - 3.229; 3 females - 2.789). Similarly, males responded to the items comprising Isolation as more characteristic of them than females (3 males - 2.804; 3? females - 2.445). Finally, the trend for Identity Diffusion yielded the same pattern of results (X- male - 3.563; R female - 3.325). Thus, the three measures demonstrated a poorer resolution of these developmental tasks for males than females. Table 7 below presents the g ratios and p values for the main effects due to Sex. Table 7. Univariate Main Effects for Sex for Dependent Measures Dependent Measure 1 ratio 1 value Inf eriority 6 . 3119 . 0137 Identity Diffusion 2 . 4542 . 1205 Isolation 5 . 1101 . 0260 df - 9/98 37 Family Interview The second part of the study explored various differences within the families of the two MRP and MRNP groups as revealed in an interview. The distribution of responses to each of the interview questions was analyzed using a chi square. Each family structure interview question is represented below for group (MRP,MRNP), controlling for sex (M,F), by response choice. Overall patterns of response obtained for both groups which would not be discovered by between group chi square differ— ences will also be presented. Normative responses to these types of questions are of interest as well as the between group differences pre- dicted. Because of the large number of analyses computed only the strong trend and significant chi square tables will be presented in the results section and non-significant or no trend results will be pre- sented in Appendix D. Few significant results were obtained but a larger number of strong and consistent trends were found; these are presented below. gguestion 1 Subjects were asked to describe the type of relationship each of their parents had with their own parents (the subjects' grandparents). Seven discrete categories of relationship types were provided for the subject to choose from (see interview procedure in Appendix A for a complete description of the seven relationship categories). A.1. Paternal Grandfather-Father relationship as reported by male subjects. No significant pattern of results for this relationship pair differentiated the two groups (MRP snd MRNP). The chi square obtained 38 - 4.0 with 5 df; probability - .548 (see chi square table Appendix D). The results for both groups combined loaded heavily on the very positive (1:241) and adequate (2:471) categories. The male subjects from both groups perceived the PGF-F relationship positively using few (28%) of the more problematic categories 3-7. A.2. Paternal Grandfather-Father relationship as reported by female subjects. Three relationship categories demonstrated a pattern of group dif- ferences that closely approximated a significant result (p - .0548). The major difference between the two groups (female MRP and MRNP) is that the MRNP group tended to load more heavily on category 2 (a good, adequate, satisfactory relationship) while the MRP group spread out with greater usage of both the highly positive category 1 (an exceptionally satisfying and rewarding relationship) and category 5 (emotional dis- tance, little involvement). The frequency counts are presented in Table 8 below. Table 8. PGF-F Relationship as Reported by Females Category MRP MRNP Total 1 6 2 8 2 7 16 23 3 3 3 6 4 1 3 4 5 7 l 8 6 2 l 3 7 1 0 1 Total 27 26 53 Chi square - 12.36 with 6 df; probability - .0548. 39 B.1. Paternal Grandmother-Father relationship as reported by male subjects. No significant pattern of results was obtained for this relation- ship differentiating the two groups. (Chi square - 5.75 with 5 df, probability 8 .3304, see chi square table in Appendix D.) Most of the male subjects from both groups combined described this relationship in either of the positive categories 1 or 2 (67%). B.2. Paternal Grandmother-Father relationship as reported by female subjects. As in Table 8 above, females closely approximated a significant result (p - .0505). The pattern of results is very similar to that obtained for the PGF-F relationship above but with one difference. The female MRNP group loaded more heavily on category 2 (adequate, normal) while female MRP subjects spread out with greater usage of the very positive category 1 relationship and the highly involved category 4 relationship (see Table 9). The same pattern of results is obtained for the PFC-F and PGMrF relationship except females in the MRP group described the PGF-F relationship more often as distant and uninvolved (category 5) and the PGMrF relationship more often as overinvolved (category 4). Table 9. PGM-F Relationship as Reported by Females Category MRP MRNP Total 1 4 0 4 2 13 19 32 3 2 0 2 4 7 3 '10 5 1 2 3 6 0 0 0 7 0 2 2 Total 27 26 53 Chi square - 11.04 with 5 df; probability - .0505. 40 C.l. Maternal GrandfatheréMother as reported by male subjects. There was not a significant difference or strong trend in the pattern of results for this relationship pair differentiating the two groups. (Chi square obtained - 7.39 with 6 df; probability - .2855. The chi square table is located in Appendix D.) As observed earlier, the type of relationship reported for the MRP and MRNP groups combined were the positive categories 1 and 2 for most subjects (862). C.2. Maternal GrandfatheréMother relationship as reported by female subjects. As with the male subjects, no significant pattern or trend of results was obtained between the two groups. (Chi square - 3.29 with 6 df; probability 8 .7707. The reader is referred to Appendix D for this chi square table.) Again the majority of the subjects for both groups combined reported category 1 (312) and 2 (452). D.l. Maternal GrandmotheréMother relationship as reported by male subjects. No significant results were found differentiating the two groups. (Chi square obtained - 2.13 with 5 df; probability - .8295. The chi square table is represented in Appendix D.) Although no differences were found between the two groups here, both groups of subjects tended frequently to report category 1 (351) and the overinvolved category 4 (312). D.2. Maternal GrandmotheréMother relationship as reported by female subjects. No significant results or strong trends were found between the two groups. (Chi square obtained - 5.15 with 6 df; probability - .5235. The chi square table is presented in Appendix D.) Again, describing 41 the maternal relationship for both groups combined, female subjects reported high category 1 and 2 usage, but also frequent (211) usage of category 4 relative to the other categories. Summary of Question 1 Results Although the results only yielded two strong trends for female subjects, a consistent overall pattern of results can be observed. These results can best be summarized by saying that there are particu- larly strong emotional bonds across generations between mother and child, and especially with daughter. For example, the grandfather's relationship with his sons and daughters was most frequently described with the positive categories 1 and 2, and to a much lesser extent, the distant category 5. The grandmother's relationship with her daughters more frequently contained the highly involved category 4 as well as category 1 and 2; and this same result of higher category 4 usage was also found for the female MRP group with the grandmother-son relation- ship. Overall these results support the findings in Table 1 above of stronger mother-child ties. Further, the between group differences of more very positive category 1 and highly involved category 4 usage for female MRP than MRNP subjects is also noted. The MRNP subjects tend to report less intense relationships as evidenced by their frequent usage of bland category 2. Question 2 Question 2 asked if each grandparent was able to (1) easily and readily accept the parents' marriage or (2) accepted it slowly, with difficulty, or not at all. A chi square analysis for each grandparent, 42 by sex of subject, was computed. There were no significant results differentiating the MRP and MRNP groups for this question. Table 10 below summarized the chi square probability value for each test; the reader is referred to Appendix D for the complete chi square tables. Table 10. Summary of Chi Square Analysis for Each Grandparent, by Sex of Subject Sex of Chi Square Subject Grandparent Probability M PGF .9531 F PGF .2327 M PGM .9510 F PGM .1809 M MGF .8354 F MGF .4361 M MGM .4215 F MGM .8945 df - 1. Although it is only a weak trend, the two results of female PGF (.2327) and female PGM (.1809) are noteworthy in Table 14. In both cases the.MRP females reported the fathers' parents (both mother and father) had more difficulty accepting their son's marriage than the female MRNP group (frequency counts are 5 vs. 2 and 9 vs. 3, respec- tively). This weak trend goes against the prediction that the MRP group grandparents would more easily accept their children‘s marriage. The two female MRP groups that reported the most parental diffi- culty in accepting their son‘s marriage still had 752 and 832 of the subjects from both groups combined respond (1) "easily and readily 43 accepted". Looking at the other results for both groups combined, 75% to 902 of all subjects were reported as having responded (l) "easily and readily accepted". In general, the grandparents of both groups are described as having readily accepted their offspring's marriages. The reader's attention is called to an emerging pattern of sig- nificant results being obtained between the MRP and MRNP groups only for the paternal grandparents. These results for questions 1 and 2 suggest a thread of "normal" Oedipal relationships across the three generations being reported by the MRP females. This differential role of the father and the father‘s parents in the family system for the two groups is noted here and will be explored in the discussion section. Question 3 Question 3 asked the subjects to describe the type of relation- ship each child in their family had with his/her parents. {Again, seven discrete categories of relationship types were provided (see Appendix D for interview procedure). No significant results or strong trends were found between the two groups for father and any of the four pos- sible children as reported by males or females. (The chi square tables for these analyses are presented in Appendix D.) Only one significant difference was found between groups for mothers and any of the four possible children as reported by males and females. (The non-significant results are presented in Appendix D.) Table 11, on the following page, presents the significant result found for mothers and the third child as reported by female subjects. This maternal involvement with the third child suggests high investment in the maternal role for MRP mothers. 44 Table 11. Mother-Third Child Relationship as Reported by Females Category MRP MRNP Total 1 6 l 7 2 6 14 20 3 2 0 2 4 6 3 9 Total 20 18 38 Chi square - 9.59 with 3 df; probability - .0214. The results obtained in Table 11 parallel the results obtained above in Tables 8 and 9. Females report greater category 2 usage (adequate) for MRNP subjects; and greater category 1 (very positive) and category 4 (very involved) for MRP subjects. Again, more intense parent-child relationships are reported by the MRP than MRNP subjects. The results obtained from both groups combined did not yield an overall pattern of results in themselves but can be compared to the grandparent-parent relationships. Two contrasts can be made with the grandparent relationship ratings from question 1: There was more conflictual category 3 usage (for most relationship pairs approximately 201) than reported for the grandparent-parent relationship; and the feuding/embittered category 6 which was low for the previous generation is virtually non-existent between parent and child. There seems to be somewhat more conflict overtly expressed between the parents and chil- dren, which is no longer present between grandparent and parent. Perhaps this more common parent-adolescent conflict subsided in the 45 previous generation for most, or moved into an unresolved long-term stand-off (category 6) for a few. Question 4 Question 4 asks which parent, if either, each child in the family was "emotionally closer to or more identified with" as they were growing up. A chi square analysis was computed for each of the four possible children in the family by sex, and by group, for the five response choices of: boy-mother, girl-father, girl-mother, boy-father, and no difference between the parents. A.1. Oldest child's identification as reported by male subjects. A very weak trend was obtained differentiating groups. There is a tendency for the MRP males to report the father-son and father-daughter relationships more frequently than the MRNP group (7 vs. 3); and for the MRNP males to report greater usage of the mother-son and mother-daughter relationship (14 vs. 8). (Chi square . 5.83 with 4 df; probability - .2117, the chi square table is presented in Appendix D.) A.2. Oldest child‘s identification reported by female subjects. A very strong trend was found differentiating for female subjects (p . .0598; see Table 12 on the following page). The MRP subjects reported the girl—mother dyad frequently (30%) but most heavily reported no difference (48%). The MRNP females used this last category infre- quently (152). A cross-generational coalition is being reported by the ‘ MRNP females between mother and daughter. 0n the other hand, the MRP females form a triad and identify with the parental coalition. 46 Table 12. Same and Cross-Sexed Identification Patterns for Oldest Child as Reported by Females Relationship MRP MRNP Total Boy-mother l 3 4 Girl-father l 3 4 Girl-mother 8 14 22 Boy-father 4 2 6 No difference 13 4 17 Total 27 26 53 Chi square - 9.05 with 4 df; probability - .0598. 3.1. Second child as reported by male subjects. A clear pattern of results differentiating the two groups ap- proached significance for the second child (p_- .0598; see Table 13 below). Table 13. Same and Cross-Sexed Identification Patterns for Second Child as Reported by Males Relationship MRP MRNP Total Boydmother 6 11 17 Girl-father 0 3 3 Girl—mother 2 3 5 Boy-father 9 2 11 No difference 6 7 13 Total 23 26 49 Chi square - 9.05 with 4 df; probability - .0598. There was more boy—father same-sexed identification for the MRP than the MRNP group and more boy-mother cross-sexed identification for the MRNP 47 group. This pattern for male second children of cross-sexed maternal identification reported by the MRNP group and same-sexed paternal identification by the MRP group is consistent with the original predic- tions. 3.2. Second child as reported by female subjects. A significant pattern of differences between the two groups was obtained on this analysis (see Table 14 below). The MRNP group had somewhat higher same-sexed responses for both "girl—mother" and "boy- father", but the largest contrast was that the MRP group used "no dif- ference" thirteen times as compared to once for the MRNP group. This high usage of "no difference" for the MRP group was also found above in Table 12 for female subjects with the oldest child. The MRP females do not report a parent-child coalition but share in a triad or identify with a father—involved parental dyad. Table 14. Same and Cross-Sexed Identification for Second Children as Reported by Females Relationship MRP ‘ MRNP Total Boy—mother 5 3 8 Girl-father 3 5 8 Girl-mother 3 10 13 Boy-father 3 7 10 No difference 13 l 14 Total 27 26 53 Chi square - 16.64 with 4 df; significance - .0023. Differences between males and females for the second child are the high "no difference" for the female MRP group again; and the high 48 same-sexed "boy-father" relationship for MRP males, which was not found for females. C.1 and 0.2. Third child as reported by both males and females. Because subjects for the study were chosen from two-, three-, and four-child families there are fewer subjects in the third and fourth child analyses. No significant differences between the two groups were found for males or females. (Chi square for males (N - 37) - 3.49 with 4 df; probability - .4799; and for females (N - 38) - 3.64 with 4 df; probability - .4588, see Appendix D for chi square tables.) D.l. Fourth child as reported by males. No differences between groups were found for this small sample (see Appendix D for chi square table; chi square (N - 12) - 3.08 with 3 df; probability - .3786). D.2. Fourth child as reported by females. A strong trend was found between groups for this small sample (N - 15) that parallels the results found above (see Table 15). Again, there was a tendency for higher same-sexed "girldmother" usage for the MRNP group and high "no difference" usage for the MRP group. Table 15. Same and Cross-Sexed Identification for Fourth Child as Reported by Females Relationship MRP MRNP Total Boy-mother 2 1 3 Girl-father 0 1 l Girl—mother 1 4 5 Boy-father 0 1 1 No difference 5 0 5 Total 8 7 15 Chi square ‘ 9.10 with 4 df; probability - .0585. 49 The responses to the above questions were collapsed or summed in order to look at three overall patterns which are presented below. First, a chi square was computed to explore differences in the presence of cross-sex identification for the MRP and MRNP groups summing over the four-child relationships above. No significant differences were obtained for males between groups. (Chi square - .8834 with 1 df; probability - .3582; see Appendix D for chi square table.) A significant difference between the two groups was found for47 females, however. Table 16 below shows that the MRP group split about evenly in reporting cross-sexed identification patterns in their family; while the MRNP group reported this type of relationship more frequently, as predicted. Table 16. Presence of Cross-Sexed Identification for Any Child Reported by Females MRP MRNP Total YES 13 21 34 NO 14 5 19 Total 27 26 53 Chi square - 4.79 with 1 df; probability - .0286. A similar tendency was evident for both males and females of the MRNP group to report cross-sexed identification patterns more frequently than the MRP gréfi, although it only reached a significant level among female subjects. From this combined analysis it is not possible to determine if this difference is composed of "boy-mother", "girl-father", or both relationships. 50 The second question asked whether all children in the family were reported as being emotionally closer to the mother. Again, the same pattern of results is found among both males and females; more MRNP subjects responded "yes" to this question than did MRP subjects, although significance was reached only for females. This pattern can be observed in Tables 17 and 18 below. Table 17. All Children Reported Closer to Mother by Males MRP MRNP Total YES 3 8 11 NO 20 18 38 Total 23 26 49 Chi square - 1.30 with 1 df; probability - .2538. Table 18. All Children Reported Closer to Mother by Females MRP MRNP Total YES 0 7 7 N0 27 19 46 Total 27 26 53 Chi square - 6.19 with 1 df; probability = .0128. The third question asked if any or all of the children in the family were emotionally closer to the father. No significant results were obtained between groups here and, interestingly, only three 51 subjects responded "yes" to this question. This supports an earlier finding of the primacy of the mother-child rather than father-child relationship. (The chi square tables are presented in Appendix D.) Summaryuof Question 4 Results Question 4 explored the prediction that there would be more cross- sexed identification (boy-mother and girl-father) in the MRNP than MRP group. Although this was found to be generally true, examination of the overall results yields an unexpected finding. First, several trends were obtained for the male MRP subjects to report both boys and girls closer to fgghgg, The female MRP subjects reported "no difference"- both mother and father. Second, the MRNP males repeatedly reported boys closer to mother; and the MRNP females consistently reported girls closer and more identified with mother. The meaningful difference in these results is that in the MRNP group both males and females reported the greater relevance of mother, In the MRP group, males reported the greater relevance of the fathgg_and females saw the mother and father as equally important. The trend for MRNP males and significant tendency for MRNP females to report all children closer to mother also supports this pattern. These results suggest that if the parents do not form a strong horizontal bond, mother turns to one or more children for her primary relationship. If mother and father form the primary bond, female MRP subjects are able to form a triad and identify with their relationship and not form a parent-child coalition. Although the saliency of the father is stressed as highly influential in these results, the primacy of the mother-child relationship discussed earlier 52 is noted here again also. Virtually no subjects reported any or all children in the family being closer to father. (Question 5 Question 5 was made up of several questions concerning cross— generation alliances and sub-groups existing within three generations of the MRP and MRNP families. A, The first question asked if there were cross-generation alli- ances reported in the family system. The hypothesis that there would be more cross-generational alliances in the families of the MRNP group was not confirmed. There were more of these reported for the MRNP than the MRP group, but this was not even a strong trend (chi square prob- abilities-males - .3284; females - .6382; the chi square tables are presented in Appendix D). For both groups combined, 502 of the males and 642 of the females reported cross-generational alliances occurring in the family system. Q, The second question asked if cross-generational alliances were repeated across generations. For example, if there was a GP-P relation- ship reported, was there also a P-C or GP-C relationship reported. This hypothesis was not confirmed by any significant results or strong trends between the two groups. The chi square probabilities were .5752 for males and .3704 for females (see Appendix D for chi square table). Almost all subjects (combined for the two groups) said this repetition did not occur (822 males; 902 females). 9, The third question asked if the nuclear family (mother, father and children) was seen as a sub-group excluding the grandparents. 53 It was hypothesized that this pattern would reflect a better separation and individuation between generations and be characteristic of the MRP group. A trend (p - .0898) was found in this direction for male MRP subjects (see Table 19). No differences were found between the two groups for females. (Chi square probability - .9732; see Appendix D for chi square table.) It should be noted that this is also a rare response for all subjects (only 7 of 102). Table 19. Nuclear Family Seen as a Sub-group Excluding Grandparents as Reported by Males MRP MRNP Total YES 4 0 4 N0 19 26 45 Total 23 26 49 Chi square - 2.87 with 1 df; probability - .0898. 2, The fourth question asked if there was an alliance from which only one member of the nuclear family is excluded. It was hypothesized that this would be more characteristic of the MRNP group. This was not found to be true for males (probability . .6693) but was true for females (see Table 20). It should be noted that this is also a rare response for all subjects (only 10 of 102). Overall, subjects very infrequently reported one member of the family being excluded. 54 Table 20. One Nuclear Family Member Excluded as Reported by Females MRP MRNP Total YES 0 6 6 NO 27 20 47 Total 27 26 53 Chi square - 4.91 with 1 df; probability - .0266. Question 6 Question 6 asked if there were polarized "all-good" and "all-bad" role positions in the family system. The prediction that the MRP group would have fewer of these extreme family positions was not confirmed. Although there were more "all good" children reported for the MRNP group than the MRP group, there was not a significant difference (chi square probabilities - .5375 for males; .4946 for females). Similarly, the greater frequency of "all bad" role positions hypothesized for the MRNP group was not found. (Chi square probabilities - .7730 for males; .9182 for females; the chi square tables are presented in Appendix D.) An inspection of the results for the two groups combined shows that about one-third of all subjects reported the "good child" role in their family; and about one-third of all subjects reported the "bad child" role (322 vs. 362). Data were not collected determining if subjects reported both roles occurring simultaneously in the family. Question 7 Question 7 asks if any family members have experienced problems in living such as alcoholism, drug arrest, police records, school 55 expulsions, etc. The prediction that the MRNP group would report more problems in living was not confirmed. (Chi square probabilities I .8058 and .9182 for males and females respectively; the chi square tables are presented in Appendix D.) The results for both groups combined reveal that 392 of the male subjects and 362 of the female subjects responded yes to this question. Question 8 Question 8 had subjects describe their fathers using four sets of adjectives which are diagrammed on the two axes below. confident, assertive 4 1 aggressive, critical supportive, sympathetic shy, wary A significant difference in attribution of these adjectives was found between the two groups for Bali‘subjects (see Table 21 on the following page). The MRP group described their futhers most often as quadrant 1-confident and loving; while the MRNP group most frequently described their fathers as confident but aggressive-~quadrant 4. Quadrants 2 and 3 were not used--virtua11y no fathers from either group were described as "shy, wary, obedient". The same pattern of results was evident for females. However, the relationship was weaker so that only a strong trend was obtained (p_ - .0681). 56 Table 21. Description of Father by Males Adjectives MRP MRNP Total 1. Confident and supportive l8 8 26 2. Confident and aggressive 4 14 18 3. Shy and supportive 1 1 2 4. Shy and aggressive 0 3 3 Total 23 26 49 Chi square I 12.26 with 3 df; probability I .0065. Table 22 below shows that the MRP group most frequently described their fathers in quadrant l and the MRNP group in quadrant 4. Similar to the males, very few fathers were described as "shy, wary, obedient". This finding again highlights the differential role of the father in the families of the MRP and MRNP subject; and also the cultural stereo- type of fathers in general as relatively assertive. Table 22. Description of Father by Females Adjectives MRP MRNP Total 1. Confident and supportive 21 13 34 2. Confident and aggressive 4 9 13 3. Shy and wary 2 1 3 4. Shy and aggressive 0 3 3 Total _ 27 26 53 Chi square I 7.12 with 3 df; probability I .0681. Question 9 Question 9 is made up of several different questions that concern changes in the patterns of family relationships that have occurred in 57 recent years. A, The first question asked about parental divorce in recent years. There were no differences in parental divorce in recent years for the two groups. (Chi square probabilities I .6931 and .5033 for males and females respectively. The chi square tables are presented in Appendix D.) There were very few divorces reported (9 subjects out of 102). g, The second question asked if subjects were becoming 1) more interested and involved in family affairs, 2) less concerned, or 3) no change. No significant differences were found between the two groups. A.meaningful pattern of results was observed, however, and is presented in Table 23 below and Table 24 on the following page. There was a trend for the male MRP group to be more interested and the MRNP males to be less interested in family affairs. In contrast to this trend for males, more females of both groups reported "more" interested in family affairs and only 2 female subjects reported "less" interested in family affairs. An overall sex difference for interest in continuing family relationships is evident here with females demonstrating more active continuing involvement with their families than males. Table 23. Recent Involvement with Family for Males MRP MRNP Total More 15 11 26 Less 2 8 10 No change 6 7 13 Total 23 26 49 Chi square I 4.12 with 2 df; probability I .1272. 58 Table 24. Recent Involvement with Family for Females MRP MRNP Total More l7 17 34 Less 0 2 2 No change 10 7 17 Total 27 26 53 Chi square I 2.51 with 2 df; probability I .2842. _C_. This question asks about changes in the "good" child role posi- tion that have occurred in recent years. No significant differences were found between groups for males (chi square probability I .8914). The majority of male subjects reported "no change". A trend was ob- served between groups for female subjects (p_I .1056), however. Whereas the preponderance of the MRNP females reported "no change" (922), only 752 of the MRP females reported "no change" and nine subjects spread out with small usage of the other three categories (gone away, newly arose, new person in old role). For the "bad child" role, a significant difference was obtained between groups for females only (see Table 25 on the following page). The female MRP group reported an increase in the "bad role" position in recent years, contrary to the prediction. This change may be a reflec- tion of the separation process currently being negotiated between parents and young adult. (The non-significant chi square table for males is presented in Appendix D.) The overall results reveal that very few subjects actually reported a change in these "all good” and "all bad" role positions in the family--they remain relatively constant. 59 Table 25. Changes in Bad Child Role Position for Females Category MRP MRNP Total 1. Gone away 5 3 8 2. Newly arose 6 0 6 3. Remained but with new people 0 0 0 4. No change 16 23 39 Total ‘ 27 26 53 Chi square I 7.74 with 2 df; probability I .0209. D, The last question asked about changes in the patterns of same and cross-sexed identification in recent years. A significant pattern of results was obtained for both first and second children and is pre- sented below. No significant results were obtained for third and fourth children, where there was a smaller N. A significant group difference was found for males reporting on the oldest child. Table 26, on the following page, reveals that there was more boy-mother and girl-mother identification for the MRNP males, as was found in question 4 above. The salient effect and change is the report of "no difference" (mother and father) for bg£h_males as well as females in recent years for the MRP group. This table also demonstrates the greater salience of the father for the MRP subjects through higher girl-father usage. A similar significant pattern was obtained among females for first child (see Table 27). The MRNP females reported more of all the same and cross-sexed categories and the MRP females loaded most heavily on "no difference" again, although the girldmother relationship was also utilized by the MRP group similarly to the MRNP group. 60 Table 26. Same and Cross-Sexed Identification for Oldest Child Reported by Males Relationship MRP MRNP Total Boy-mother l 7 8 AGirl-father 6 2 8 Girldmother 2 8 10 Boy-father 4 7 11 No difference 10 2 12 Total 23 26 49 Chi square I 16.12 with 4 df; probability I .0029. Table 27. Same and Cross-Sexed Identification for Oldest Child Reported by Females Relationship MRP ‘MRNP Total Boyhmother 0 3 3 Girl-father l 5 6 Girl-mother 10 12 22 Boy-father 2 3 5 No difference 14 3 l7 Chi square I 13.15 with 4 df; probability I .0106. A significant result was not found among males for second child (see Table 28, on the following page) as it was with.males for first child. There was not the high "no difference" for the MRP group as there was above. There was a trend towards more "boydmother" usage for the MRNP group again; and higher MRP "boy-father" usage continuing the MRP father and MRNP mother pattern found throughout. A significant difference between groups was obtained for females (see Table 29). The MRP group reported "no difference" while the MRNP group used more of both the same and cross-sexed categories again. 61 Table 28. Same and Cross-Sexed Identification for Second Child Reported by Males ' MRP MRNP Total Boy-mother 3 9 12 Girl-father 10 3 13 Girl-mother 4 3 7 Boy-father 0 5 5 No difference 6 6 12 Total 23 26 49 Chi square I 7.65 with 4 df; probability I .1051. Table 29. Same and Cross-Sexed Identification for Second Child Reported by Females MRP MRNP Total Boy-mother 1 5 6 Girl-father l 5 6 Girldmother 3 9 12 Boy-father 4 5 9 No difference 18 2 20 Total 27 26 53 Chi square I 21.23 with 4 df; probability I .0003. The results obtained here repeated those found earlier in question 4. For males, the groups again differed in that the MRP subjects chose father or father and mother ("no difference") and the MRNP subjects chose mother. Also the females differed in that the MRP subjects chose both parents ("no difference") in every case, unlike the MRNP group, who showed more frequent differential choice of one parent, usually mother. The differing role of the father is again highlighted for the two groups. 62 One interesting difference from question 4 here is that the MRP males reporting on the oldest child selected both parents ("no differ- ence") rather than just "father" as was done for both the second child here and for all the MRP males when they were younger (question 4). This suggests a developmental trend, with the MRP males moving toward integrating both parents as they get older, which the MRP females have reported all through their lives, and the MRNP subjects have not done at all. (Question 10 Question 10 asked if the mother and father had significant friend- ships with people outside of the family. No significant results or trends were found for males (chi square probability I .7755; .9205) or females (chi square probability I .6005; .6932) for mother and father respectively. (The chi square tables are presented in Appendix D.) Looking at the responses to this question for both groups combined re- veals that fathers have more relationships outside the family than mothers. Subjects reported approximately 292 of the mothers had sig- nificant relationships outside the family and about 442 of the fathers. Here again, the greater maternal involvement in the family or the greater centrality of the family for women is suggested. Finally a chi square was computed to examine differences in the sample for race or parental education. No significant differences were found for either factor. All but 7 of the 102 subjects were white; females (chi square probability I .1199) had a slight tendency to have better educated parents than males (chi square probability I .9235). The chi square tables are located in Appendix D. CHAPTER IV DISCUSSION The data in Table l reflected an important sex difference in sub- jects' perceptions of family relationships. Males perceived the marital bond as the.most basic relationship in the family, followed by the parent-child dyad. Females reversed this order and ranked the parent- child dyad as primary followed by the marital bond. This inverse pat— tern suggests that a parent-child dyad is more salient for females such that it overrides the primacy of the marital relationship. A different experience of the basic family structure is being reported here by young men and women. Beyond this reciprocal sex difference, the data in Table 1 also reveal that young adults preponderantly view the mother-father and parent-child dyads as the most important relationships in the family. Although one or the other dyad was perceived as primary for males and females, 812 of the subjects reported one of these two relationships as primary. In contrast, relationships between children and with grand- parents were infrequently perceived as the primary dyads in the family. Further questioning during the interview revealed that within parent- child it was the mother-child dyad, and not father-child, that was being reported. Thus, the fundamental dyad is in the nuclear family, between mother and child or mother and father. 63 64 The Mother-Child Dyad It was noted that when the parents were perceived as the primary two-person relationship in the family (MRP group), the mother-child dyad was then reported as the second strongest bond in the family. Male (482) and female (482) MRP subjects alike ranked the mother-child relationship as second to the marital dyad most frequently. The mother-child rela- tionship was also the dyad most frequently reported as primary in the MRNP group. Here, there was a non-significant tendency for females to rank the mother-child relationship as primary more often than males. The interviews did not reveal whether it was differentially the mother- daughter, mother-son, or both relationships that were being ranked. Theory stressing the importance of a primary marital bond was presented in the introduction. These results suggest that this is perceived more by males than females, and that the mother-child relationship is a fundamental bond for virtually all subjects. Psychoanalysts have always asserted that the child‘s relationship to his/her mother is the foundation stone of personality. Theorists are in agreement about the far-reaching influence of this first human relationship on all later personality development, although the nature of this powerful tie is disputed. Bowlby (1969), for example, offers a biologically based theory involving instinctive attachment behaviors that have the function of maintaining proximity to the mother. A secondary drive learning theory is also widely accepted wherein the mother becomes important as the source of gratification for the child's needs. From whatever viewpoint, the maternal bond found with young th In t1 re] rec 65 adults in the present study.has been observed with offspring throughout animal and human life. In the animal world, washburn, Jay, and .Lancaster (1965) assert "this pattern of enduring social relations be- tween mother and her offspring will be found in other species of primates". Bowlby (1969) notes that attachment behavior in humans decreases with maturity but continues as a normal aspect of personality into adult life. He goes on to suggest that in many societies the attachment of mother and daughter continues more strongly than that of son to mother. Young and Willmont (1957), for example, found the bond between mother and daughter plays a great part in western urbanized social life (see also Sweetser, 1964, above). This effect can be seen in the sex.differences obtained, with males more frequently perceiving mother-father and females perceiving mother-child as the primary rela- tionship in the family. Several results from the interview also exemplify strong and con- tinuing maternal bonds to children, and especially to daughters. Questions 1 and 3 both showed greater usage of the highly involved (category 4) relationship for grandmothers and for mothers with their male and female offspring in both groups. Although there were some significant results showing greater category 4 usage for the MRP than the MRNP group, grandmothers and mothers overall were reported to have this type of very close and involved relationship with their children. In a related vein, questions 9C and 10 brought out important female ties within the family unit. A sex difference revealed that females reported more continuing and active involvement with their families in recent years than did males. Looking at the relationship boundaries of 66 the family unit, mothers were reported as having fewer (292) signifi- cant friendships outside of the family than fathers (442). The results for question 4, which are presented more fully later, point up an important difference in mother's relative salience for the two groups, however. Subjects in the MRP group did not report being "closer and more identified with" mother while the MRNP subjects (both males and females) did report being closer and more identified with mother. Although these various results all emphasize the mother-child relation- ship, it is clear that this relationship is significantly affected and ’5 qualified by the father‘s differential saliency for both mother and/’ L/// children in the two groups. If father is perceived as primary for \ mother, the children also tend to identify with father. In the MRNP \\ group, where father is not primary for mother, the mother-child bond \ remains strong and the father does not become as significant for the 1 children. These results may be seen as supporting the family theory 3 which postulates that in the "healthy" family the mother-father bond ) must be more important than the parent-child or any other bond. -“X The results of the present study confirm the important and con- tinuing role of the mother-child relationship within the network of family relationships, although they do not as clearly elucidate mother- daughter and mother-son differences. The results obtained do suggest the possibility that there is less difference in importance between the mother-daughter and mother-son relationship when the marital bond is primary than when it is not. The overall findings, however, do emphasize the importance of both the mother-child and mother-father relationships. The discussion will go on to elucidate some of the 67 different effects of this fundamental maternal relationship on the off- spring as a function of the varying role or salience of the father and husband. Child Adjustment as a Function.of Marital Primacy A sex difference was obtained in the results for adjustment dif- ferences between the MRP and MRNP groups. The results showed a strong tendency for females who perceived their parent's marriage as the pri- mary relationship to be better adjusted than females who did not. The MRP females scored higher on successful resolution of the psychosocial crises of Industry, Identity, and Intimacy and lower on unsuccessful resolution of Inferiority and Neuroticism, as predicted. This cluster of results lends conditional support to the main hypothesis of the study that the children from a marital relationship primary family will be better adjusted at college-age than those from.a family with a non- marital dyad as primary. Although further analyses of the results for males were not statistically warranted, the patterns of means obtained will be dis- cussed and extrapolated to theory in a speculative or clinical fashion. The results showed that there were virtually no differences between the MRP and MRNP males on any of the six Ericksonian measures. The overall mean difference for each measure between groups was very small, although on all six measures the MRP males scored less well adjusted than the MRNP males, contrary to the prediction. A more substantial mean differ- ence between the two male groups was revealed in the trend found for 68 Neuroticism. Again, the MRNP males were (non-significantly) less anxious than the MRP males, against the predictions. One explanation for these results is simply that the prediction tends to hold true for females but not males. This may occur because females characteristically function in a social-emotional and males in a career-achievement modality. The identity of a woman has been tradi- tionally tied to marriage and other intimate relationships--having a parental model of a successful emotional bond may make the MRP females more secure in their own capacity to establish a significant marital relationship than the MRNP females. Because much of a male's identity is established through work and success, the model of a primary marital bond may be less relevant to the identity challenges he faces as a beginning college student. Although parsimonious, this explanation does not address the contradictory tendency towards poorer adjustment for the MRP males. A more encompassing interpretation of the results is constructed below, which speculates that the converse pattern for males is indeed consistent with clinical theory. The freshman and sophomore subjects used in the present study have fairly recently separated from their family of origin and are making an adjustment to college life. The developmental task addressed at this time is the identity crisis. Research evidence in recent years, however, has shown that this is a very different process for males and females. As briefly sketched above, our society offers differential roles to men and women. The sociologists Parsons and Bales (1955) have suggested an instrumental/expressive sex role distinction. More specifi- cally, Douvan and Kaye (1962) have noted the dominant motivation of 69 males going to college is vocational while that of females tends to be more social. The major component of the male's identity is an occupa- tional one while the female is more involved with her sex role as wife and mother. For males, then, the identity crisis is met in the early years of college when there are independence demands, more separation from.family ties, and critical career choices to be made. The identity process is more protracted for females who are reinforced by our society for maintaining closer family ties and relationship emphases (the "culture bearers"). Because of this, females are not yet in crisis dur- ing this early college period. In support of this, Constantinople (1969) found that females scored as more mature (higher on the success- ful and lower on the unsuccessful resolutions of the Ericksonian measures used in the present study) than males during the freshman year. Stark and Traxler (1974) found female underclassmen to be less anxious than men. This difference was also found in the present study where signifi- cant main effects for sex reflected that females were less Isolated, less Inferior, and had a trend towards less Identity Diffusion than males. Marcia (1964) has done a series of studies on identity formation. He has developed an interview procedure that assigns one of four identity statuses to men and women in their late adolescence who are in different stages in their achievement of an identity. From Erickson's theory, he abstracted out the two variables of "commitment" (personal commitment to a course of action or ideas) and "crisis" (a decision period when the individual seems to be actively involved in choosing among alternatives). Those who had passed a crisis and were firmly 70 committed to an occupation and an ideology were assigned the status of Identity Achievement (plus crisis, plus commitment). Those who did not experience a crisis, yet were firmly committed, were scored as Foreclosure (minus crisis, plus commitment). Those who were presently in crisis and were either uncommitted or only vaguely committed were given the status of Moratorium.(plus crisis, minus commitment). Those who did not experience a crisis period and lacked a commitment were assigned the status Identity Diffusion (minus crisis, minus commitment). The Identity Achievement and Moratorium groups were found to be more similar while the Foreclosure and Identity Diffusion groups were more alike. Marcia and Friedman (1970) have shown that persons in the former two statuses are more anxious than the latter two groups because they are in crisis (Moratorium) or alienated from a peer group (Identity Achievement). Although they scored higher on the welsh anxiety test (the Moratorium subjects highest, Foreclosure subjects lowest), they are still considered at a higher stage of ego development because they were also less authoritarian, less vulnerable to self esteem manipula- tion, and had better cognitive performance under stress than the Fore- closure and Identity Diffusion subjects. The critical point here is that it is necessary and adaptive for males, but not females, to experience a "crisis" with heightened anxiety during the early college years. Avoidance of this period of turmoil for males yields a Foreclosure status, that does not develop an indi- viduated identity, or simply continued Identity Diffusion. One explana- tion for the results of the present study is that the MRP males are in fact better adjusted, as predicted, and are developmentally in step by 71 experiencing the higher anxiety inherent in the identity crisis. These males would be in.Marcia's Moratorium status. Consonant with this specu- lation, the MRP males also scored relatively lower on the Eysenck Lie scale and higher on Extraversion than the MRNP males. The MRNP males who were found to be less anxious may have settled for a Foreclosure status and avoided the period of crisis. Framed another way, one specu- lation is that the MRP males are currently experiencing the same separa- tion conflicts their fathers may have experienced as reported in Question 2, where the MRP grandparents are described as having more trouble letting go of their sons, which was not true for the MRNP males or any females. As expected, it was found that females in the MRP group are better adjusted than the.MBNP females on the measures used because they are not yet in "crisis". Constantinople (1969) found that females will become more stressed during the later college years. Within this theoretical framework, the findings of the present study are speculated to be supportive of the original predictions. Family Structure One purpose of the present study was to explore patterns of cross— generational relationships and fixed role positions in the family. Several questions were asked during the interview addressing different aspects of these two topics. Concerning fixed role positions in the family, responses to question 5C suggested that one member of the nuclear family is rarely excluded or held away as an outcast. Although there was a significant difference obtained in the expected direction of greater occurrence in the MRNP group, only 102 of the subjects 72 overall reported this. Similarly, question 6 asked if there were fixed "all-good" and "all-bad" role positions in the family. About one-third of all subject reported each of these roles in their family, although differences between the MRP and MRNP groups were not found as predicted. Question 90 followed up this same question by asking about changes in these "all-good" and "all-bad" role positions in recent years. Although some changes were seen that may reflect the separation crisis occurring at this time (more MRP females in "bad-child" role), these role posi- tions seemed to remain relatively constant for family members. Finally, question 7 explored if there would be more problems in living for other family members (alcoholism, school expulsion, police records, etc.) in the MRNP than the MRP groups. Again, about one-third of all subjects reported these phenomena but differences between the two groups were not found. These attempts to find differences in the role positions and broader family system for the two groups were not supported. Another group of questions focused on three-generational relation- ship patterns. In general, the results from questions 1 and 3 showed that children reported their relationship with their parents and the relationship between their parents and grandparents in a positive manner. The normative results from Table 1 revealed that the nuclear family relationships of mother-father and mother-child were perceived as the most fundamental relationships. Although dyads involving a grandparent were rarely the most important, they did seem to be important to the nuclear family. Virtually no subjects in question SC reported their nuclear family as a sub-group excluding the grandparents. Relationship teams or alliances were reported by over half of all subjects across I 73 generations, although the hypothesis that cross—generational alliances would be repeated between grandparent-parent and parent-child relation- ships was not found to be true. Throughout these questions it was pre— dicted that there would be differences between the two MRP and MRNP groups. The hypothesis that the parents who do not establish the pri— mary bond in the family have unresolved primary bonds to their own parents, which then repeat in cross-generational alliances with their own children, was not confirmed. The observed means were in the pre- dicted direction but did not approach significance. One explanation for this lack of findings is the retrospective self-report method utilized. Several theorists cited earlier stressed the unconscious and covert nature of these cross-generational ties (e.g., Haley, 1967). Perhaps this information cannot accurately be obtained from a retrospective and conscious self—report method. Another possibility is that the three- generational processes hypothesized to occur are only found in a particu- lar sub-group of the sample. The 72 of the subjects who reported "grandparent-parent" or "grandparent-child" or the 122 that reported the "child-child" relationship as primary contain an extended genera- tional member and may be families where these patterns occur. Because the majority of the MRNP sample selected parent-child as primary, dif- ferences in the three-generational family system of the other primary dyads may have been diluted. This is discussed as a possibility for future research below. One other explanation that should be considered for all these results is the normal population selected. The family theory which describes structural family relations is based on clinical samples and these same three-generational processes may not occur in a 74 normal sample. The most powerful discriminator between the two groups did not lie in the three-generational relationships or fixed role posi- tions as expected, but rather in the differential role of the father. The Father's Role The strongest results differentiating the two groups in this study concerned the father. The only two near significant results (p<=.06) obtained between the MRP and MRNP groups in question 1 involved the father and his parents. The MRNP group described their paternal- grandparent relationships most frequently as falling in the "adequate category (2) while the MRP subjects had more of the very positive cate- gory (l) for both grandmother and grandfather, more of the "highly involved" category (4) with the grandmother, and for one case, in the other extreme, more of the "distant" category (5) for grandfather and father. It seems that the MRP subjects' fathers usually had an especial- ly good or occasionally a very distant relationship with their fathers, and a very good or highly involved relationship with their mothers. These more extreme grandfather-father relationships coupled with the intense grandmother-father relationship preceded the father establishing the primary bond in his own family with his spouse. It was stated earlier that these results suggest a more "typical" or normal Oedipal configuration in the father's background for the MRP than MRNP subjects, or more simply a history of closer family ties for these fathers. Interestingly, again it was the grandfather-father and grandmother- father relationships that approached significance for question 2. In both cases the MRP group reported the father's parents (both grandmother ti id 75 and grandfather) had more trouble accepting their son's marriage than the MRNP group. This trend suggests stronger ties between father and his parents for the MRP group. This tie could reflect a greater conflict with separating and approving of the son's decisions and independence or simply stronger affection. The results from question 1 were both very positive and more conflictual and either or both may be contributing to this trend. The important effect, though, is that it is the father's relationship with his own parents and not the mother's background that differentiates the ’1 two groups on both questions. In this regard, Barry (1970) summarizes ,1 E ,3 k. the marital research literature by saying that the husband's past history, } f happiness of his parent's marriage, and his relationship to his father I are all more strongly related to his present marriage than similar I factors for the wife. \\\\’)/ Questions 4 and 9D asked subjects who they were "closer and more identified with" as they were growing up and again in recent years. These two questions yielded a consistent set of highly significant find- ings that demonstrated the father's relative saliency for the two groups. The MRP males differed significantly from the MRNP males in reporting strong father-son and father-daughter relationships, while the MRNP males reported only the cross-sexed "boy-mother" relationship. Similar- ly, the MRNP females also reported only the "girl~mother" relationship. In contrast, the MRP females reported "no difference" each time-both '/// mother and father. In short, the MRP subjects reported an identifica- tion that included the father, and the MRNP subjects reported a maternal f E identification. This same pattern of results was found to remain in j I \l 76 recent years, except the MRP males were more apt to report "no differ- ence" for the oldest child. The MRP and MRNP groups, in one sense, were originally selected on the basis of the emotional relevance of the father to the mother. These results are showing this same effect across generations. When the father was the primary relationship for the mother he is also the primary identification figure for his sons and for his daughters, as perceived by females. If father is not primary to mother as in the MRP group, he is also not the primary identification figure for any of his children, all of whom are more identified with mother. It was stated earlier that without father, mother seems to turn to one or more children for her primary relationship. This bond con- tinues through development without expanding to allow for a strong father-child attachment. This is probably the most important finding demonstrating the effects of having the marital relationship as the pri- mary bond in the family. a} It is also interesting to note in these results that the MRP females all along, and the MRP males in recent years, report both par- 4-!”‘AV’- ents equally as objects of identification. Perhaps not only a single parent but also a relationship is internalized or identified with by a child. In object relations terms, this would be the establishment of a ’ l I I" constant internalized representation of the relationship between the two parents separate from.their individual images. The Lidz group (1957) cited earlier suggested the possibility that the children of a schisma- tic couple could not internalize and identify with a parental relation- ship that was irreconcilable in reality. In one study, Westley and Epstein (1970) studied high and low adjusted college freshmen through 77 extensive testing and family interviewing. They argued that the child's emotional health was more related to the quality of the emotional rela- tionship between the parents than the psychological status of the individual parents. Processes other than single same-sexed identifica- tions may be occurring. These results suggest that females move from their primary attachment with mother to incorporate a relational pair when father is an emotionally relevant figure. This would be syntonic with their sex role orientation towards interpersonal relationships. If males have an available father, they may move from the initial attach- ment to mother, to an identification with father, and then to a final integration of that relationship during young adulthOod. It is clear, though, that the father is salient to the MRP subjects such that the initial bond to mother is relinquished or at least allows for a close relationship and identification with father. Children from.the MRNP group do not seem to leave the early bond to mother and move toward father. The results from question 8 provide further information and sup— port this framework. A highly significant difference was obtained be- tween the two groups'descriptions of their fathers. The MRP subjects described their fathers as "confident and supportive" (warm, sympathe- tic, affectionate, understanding) while the MRNP subjects described their fathers as "confident and aggressive" (critical, irritable, frank). One of the most definitive findings in the literature on fathering is the many positive effects on child development of a nurturant or emo- tionally available father for sons and daughters (Biller, 1974, pp. 33- 34). The results demonstrate that this affective bond was not perceived 78 between father and child for MRNP subjects. In this regard, Aberlin (1971) offers a theory derived from his research with young children in an ego developmental framework. He postulates that it is the father's role to draw and attract the very young child away from mother and into the "real world of things and people"—-to support and encourage the child's first steps into the world of external reality and be there for the mother to turn to. Aberlin believes it is the father's role to help the mother and child to separate from their symbiosis and begin to individuate. He suggests that this task,"might be impossible for either of them to master without their having the father to turn to"(p. 248). Although he is concerned with very early childhood experience, this may be an important prototype for further personality development which is relevant to the findings obtained here. Conclusions and Directions for Future Research \\// In summary, the results of the study can be interpreted as center— ing upon the Oedipal triangle rather than finding strong patterns of relationships in the extended three-generational family system. The important role of the mother-child bond was clearly established through- out the results. The pivotal or differentiating factor was the salience of the father. If he was primary in relationship to mother, the mother- child bond expanded to allow the children more of a relationship with father. Likewise, if father was not primary for the mother, he did not become as significant for the children who then retained the original bond to mother. The difference between the two groups of fathers was along a supportive/loving vs. critical/aggressive dimension. 79 These results suggest that the mother-child relationship may be a more fixed or constant entity which is affected by a more widely varying factor of an emotionally present and supportive or more distant and critical father. The more influential role of the father's background than that of the mother's, cited above, supports this. That differen- tial effects are associated with the relative ranking of the mother- child or mother-father relationship as primary is suggested in the marginally significant findings of higher adjustment for MRP over MRNP females. Other than some small trends concerning the father's parents, which suggested a more typical Oedipal configuration for the parents and grandparents of the MRP than MRNP subjects, few differences were found between the two groups concerning other family members, role posi- tions, or three-generational relationship patterns. Several issues are raised in the present study that could be addressed in future research. One specific direction would be to select subjects according to Marcia's (1964) four identity statuses described above. This would offer a more precise evaluation of the subject's level of ego development and the lack of finding or trend towards reversed findings for males may be better explicated. The important differences between the two groups concerning maternal and paternal identification could then be explored more thoroughly with these more differentiated subject groups. Some other changes in the present methodology could also be made. There was little variance for each of the dependent personality measures (and some inter-correlation). A.more heterogeneous sample could be selected or tests better designed to control for social desirability 80 could be administered which may yield a greater variability of scores. A better possibility may be to select an initial group of high and low adjusted subjects and study their family relationships and primary bonds. Another factor that entered into the present study was the loss of 232 of the sample because of changes in relationship selected as primary. It is possible that with a larger sample the many trends and marginally significant findings obtained would be strengthened. Although this large percentage of subjects changed their choice of primary rela- tionship, this perception may actually be more reliable and stable if obtained under different conditions. The subjects were asked to make their relationship rankings in a large lecture hall where it is possible that many were afraid to ask questions to clarify the procedure or put little time or thought into the anonymous task. During the one-to-one interview situation when they could respond more slowly and with more thought, subjects were confident and specific about making the relation- ship rankings. Those that switched often said they were in a hurry the first time and did not think of "that one". The lack of three-generational findings has also been noted. From the initial normative data it was found that only 192 of the rela- tionships reported as primary involved other than mother-father or mother—child relationships. Perhaps a sufficient sample of subjects reporting "grandparent-parent", "grandparent-child", and "child-child" relationships primary could also be selected and compared with a mother- father and mother-child primary group, for each of Marcia's identity statuses. Differences may have been blended in the present study which R _ 81 would reveal more of the three-generational influences hypothesized if samples of these minority pairings were obtained and compared with each other. It would be an important direction for future research to see if different constellations of family relationships are found to char- acterize each of the various primary dyads. Finally, as well as looking at characteristic adjustment and rela- tionships differences between each of the possible primary dyads, dif- ferences should also be explored across family size. In the present study two—, three- and four-child families were chosen because they were thought to be more homogeneous than single or many-child families. Although this was found to be generally true, a trend towards the two- child-family siblings not allying with each other but turning to a grandparent more often than in three- and four-child families was ob- served. This pattern may be reversed in very large families where siblings become more important to each other as parents are less avail- able. Data obtained from single-child families not analyzed for the present study revealed a very different pattern of primary dyads. Thirteen of the 17 single—child—family respondents perceived the parent- child dyad as primary-a much greater proportion than found for the two—, three- and four-child subjects. Future research could control for family size and make comparisons between the one-, two- and many- child families as well. The family is the medium in which personality develops. Marked differences in individuals' perceptions of the basic structural relationships in the family as a function of sex, family size, or age may yield important information about personality development and functioning. BIBLIOGRAPHY BIBLIOGRAPHY Aberlin, E. L. The role of the father in the Separation-Individuation process. In J. B. McDevitt and C. F. Settlage (Eds.), Separation- Individuation: Essuys in Honor of Maugaret S. Mahler. New York: International Universities Press, Inc., 1971, 222-253. Barry, W. A. Marriage research and conflict: An integrative review. Psycholggical Bulletin, 1970,.13, 41-55. Bell, N. W. Extended family relations of disturbed and well families. Family Process, 1962, ;, 75-93. N I, / Biller, H. B. Paternal Deprivation. Lexington, MA: D. C. Heath and [ //’ Co., 1974. Boszormeny-Nagy, 1., and Sparks, G. Invisible Lgyalties: Reciprociuy in Intergenerational Family_Therupy. New York: Harper and Row, 1973. Bowen, M. A family concept of schizophrenia. In D. D. Jackson (Ed.). The Etiology of Schizophrenia. New York: Basic Books, 1960. Bowen, M. The use of family theory in clinical practice. Comprehensive ngchiatry, 1966,.1, 345-376. Bowlby, J. Attachment and Loss. Vol. I. New York: Basic Books, 1969. Camp, H. Structural family therapy: An outsiders perspective. Family Process, 1974, gg, 269-277. Constantinople, A. An Ericksonian measure of personality development \ in college students. Developmental Psychology, 1969, 1, 357-372. Douvan, E. and Kaye, C. Motivational factors in college entrance. In N. Sanford (Ed.), The American College, New York: Wiley, 1962. Eysenck, H. J. The questionnaire measurement of Neuroticism.and Extraversion. Revista oi Psicolo 1a, 1956, 59, 113-140. Fleck, S. An approach to family pathology. Comprehensive Psychiatry, 1966, 1, 307-320. 82 \\\J 83 Fleck, S., Lidz, T., Cornelison, A., Schafer, S., and Terry, D. The intrafamilial environment of the schizophrenic patient: Incestu- ous and homosexual problems. In J. Masserman (Ed.), Individual and Familial Dynamics. New York: Grune and Stratton, 1959. Gerber, G., and Daswan, J. Expression of emotion through family group- ,,/ ing schemata, distance, and interpersonal focus. Journal of /// Consulting and Clinical Psychology, 1971, 36(3), 370-377. /' / Glick, I., and Kessler, D. Marital and Family Therapy. New York: // Grune and Stratton, 1974. V Haley, J. The family of the schizophrenic: A model system. The Jour- nal of Nervous and Mental Disease, 1959, 129, 357-374. Haley, J. Toward a theory of pathological systems. In G. H. Zuk, and I. Boszormeny-Nagy (Eds.), FamilyATherapy and Disturbed Families. Palo Alto: Science and Behavior Books, 1967a, 11-27. Hamid, P. Birth order and family schemata. Perceptual and Motor Skills, 1970, 31, 807-810. Keuthe, J. Social schemes. Journal of Abnormal Social Psychology, 1962, 64, 31-38. Lidz, T., Cornelison, A., Fleck, S., and Terry, D. Intrafamilial en- vironment of the schizophrenic patient, 1: The father. Psychiatry, 1957a,,29, 329-342. Lidz, T., Cornelison, A., Fleck, S., and Terry D. Intrafamilial environ- ment of schizophrenic patients, II: Marital schism and skew. American Journal of Psychiatry, 1957b, 114, 241-248. Lidz, T., Fleck, 5., and Cornelison, A. Schizophrenia and the Family. New York: International Universities Press, 1966. .\ Marcia, J. E. Development and Validation of Ego-Identity Status. \ Journal of Personality and Social Psychology, 1966, 3, 551-558. \/ Marcia, J. E. and Friedman. Ego identity statuses in college women. Journal of Personality, 1970, 3g, Minuchin, S. Conflict-resolution family therapy. Psychiatgy, 1965, 28, 278-286. Parsons, T. and Bales, R. F. Family socialization, and interaction process. Glencoe, Illinois: Free Press, 1955. Radkin, L. Y. The patient's family: research methods. Family Process, 1965, i, 85-105e 84 Satir, V. Conjoint Family Therapy. Palo Alto, Calif.: Science and Behavior Books Inc., 1964. Spark, G. Grandparents and intergenerational family therapy. Family Process, 1975, 13, 225-236. Stark, P. A. and Traxler, A. J. Empirical validation of Erickson's theory of identity crisis in late adolescence, Journal of Psychol- ogy, 1974, 86, 25-33. Sussman, M. The isolated nuclear family: fact or fiction. Social Problems, 1959, 6, 334-340. Sweetser, D. A. Mother-daughter ties between generations in industrial societies. Family Process, 1964, 332-343. Washburn, J. 8., Jay, P. C., and Lancaster, J. B. "Field studies of old world monkeys and apes". Science, 1965, 150, 1541-1547. ' Westley, W. A. and Epstein, N. B. Silent Majority. San Francisco, /\\Q/ Calif.: Jossey-Bass, 1970. Winer, B. J. Statistical Principles of Experimental Desigu. New York: McGraw-Hill, 1971. Wynne, L. C. The study of intrafamilial alignments and splits in ex- ploratory family therapy. In N. W. Ackerman, F. Beatman and S. N. Sherman (Eds.), Exploriuthhe base for family therapy. New York: Famdly Service Association of America, 1961, 95-115. Young, M. and Willmont, P. "Family and kinship in East London." London: Routledge 8 Kagan Paul; 1957, New York, The Free Press. Zuk, G. When the family therapist takes sides: A case report. Psychotherapy: Theory, Research, anngractice, 1968, 5, (1), 24-28. Zuk, G. Family therapy: A triadic based approach. Behavioral Publi- cations, Inc., 1971. APPENDICES APPENDIX A QUESTIONNAIRE Please print clearly Ask for assistance Fill out completely if directions are not clear 1. Name Circle: .Male Female 2. Telephone number 3. Class circle: Frosh.; Soph.; Junior; Senior; Other 4. Marital Status circle: single. married other 5. Thinking of the bonds of emotional closeness and involvement in your family as you.were a child growing up, what was the most basic and primary 2-person relationship in your family? Circle one: 1. A grandparent and parent 2. A grandparent and grandchild 3. Mother and Father 4. A parent and a child 5. Child and child 6. How many children are there in your family? 7. Was there a parental death or divorce in your family before you were 12 years old? Circle: Yes No 85 APPENDIX B: PART I DESCRIPTION OF SCORING CATEGORIES It was stated in questions 1 and 3 above that the Interviewer and Sub- ject reached agreement on one of seven possible categories describing the relationship. These seven categories are: l. 3. An exceptionally satisfying and rewarding relationship with much respect, closeness, and independence. A good, positive, satisfactory, or adequate relationship. A conflicted relationship described either as critical, domi- nating or controlling; or an angry, rebellious, and fighting relationship. The relationship is characterized by overt conflict and struggle. A highly involved relationship where the people involved share an unusual amount of each other's personal and intimate affairs. In this relationship it is difficult for one to live in a way that would be disapproving or upsetting to the other, it may be hard for these people to move away from each other, or to have their own private, separate and independent activi- ties. An emotional distance, disinterest, lack of concern, and little involvement characterizes the relationship; or the person has an emotionally uninvolved, absent, alienated relationship where he/she is an "outsider" to the family. 86 87 6. An emotional wall of bitterness and rejection exists-one is actively held away, there is virtually no interaction. 7. If one person in the relationship described died before the other was aged 12 it is scored 7. If the person died after age 12 a description of the relationship that was had up to that time is obtained. The following two questions are asked to help determine the difference between categories 1 and 4. a) "Would you describe this relationship as either over-protected, babied, spoiled, very dependent, or too close?" b) "Would you describe the grandparent and parent or the parent and child in this relationship as being equals, friends, companions and intimates rather than as having a close parental relationship?" For category 1 the 8 must describe the relationship with a grandparent and nuclear family member (parents and children) as an especially good, emotionally close and satisfying relationship. Although Ss may report the family member may spend more or less time with the grandparent, the quality of this time is high and very much valued by the participants. 83 say things like: "My sister and I have always been very close to my dad's mother, we often share things. She's really neat." "My grand- mother and my mother have always been very good to each other, they seem to have a real respect for each other." "She is a marvelous woman." When the S describes this type of relationship, it is always determined if the participants described are able to also have their 88 own private affairs and independent activities. That is, whenever a category 1 relationship is described, the interviewer must be careful to distinguish it from a category 4 relationship. The S answering "yes" to either of questions (a) or (b) above decides the relationship as category 4. Category 2 describes a more neutral relationship that may stretch from generally positive to not particularly close or involved. Category 2 covers modal family relationships. It is not a highly emotionally in- volved relationship and does not form a real integral relationship for the person. 88 may say things like "All of us kids have always gotten along fine with them." "When we were kids it was exciting to visit them." "My mother has a good relationship with her mother." "They live in Arizona now and we haven't had much to do with them." Whereas categories 1 and 2 reflect very positive and generally neutral relationships, categories 3 through 6 are made up of different types of problematic relationships. Category 3 is for family relationships characterized as conflicted. These relationships are described as either critical, dominating, or controlling; or rebelling, fighting and arguing. 88 say things like "My grandmother doesn't think my mother is good enough for her son." "My grandfather was always telling us we would never make it if we didn't shape up." My grandmother lived with us and would try and tell my father what to do." Or, Ss say things like "My brother is always fighting with her." "My parents have been having trouble making him obey them." 89 Category 4 describes grandparent-parent and parent-child relationships that are intrusive, like in 3 above, but the parents or children seem to accept or need this involvement. These relationships are over- involved or too close such that a parent may be dependent upon a grand- parent and feel insecure without their support. 88 say things like "My mother is very close to her, she would never move away from her home town." "My father works for his dad and always does what he says." S's description may initially sound very warm, close and integral as in category 1, but in 4 the offspring are unable to leave the parent psychologically and be secure without them. Categories 5 and 6 are scored when there is an emotional "wall" between members of the two generations. Category 5 expresses a total dis- interest or lack of concern with the person. Ss may describe a rela- tionship blandly that may initially sound like category 2, but further discussion reveals there is no cathexis, involvement, or loyalty to this grandparent, parent or child. For example, "My dad doesn't have anything to do with his parents, I never hear him talk about them or anything." "They live out West and we never hear from them or call them either." "There's nothing to say really, they've never even been part of the family." Or, 83 may say things like "My brother is sort of out of the picture, he doesn't have much to do with the rest of us." "She is alienated from the rest of us, she got married young and was always away from hom." In category 5 there is an alienation or absen- tion from the family. 90 Category 6 is also scored when there is an emotional boundary that holds the two people apart such that there is no real interaction or flow between them. A grandparent, parent, or child may be actively held away but with feelings of bitterness and rejection described. 88 say things like "My mother thinks her father was mean and she doesn't want anything to do with him." "My dad's mother was always trying to tell him.what to do and he just ignores her now." Category 7 is used when a family member is dead and has not had a rela- tionship with certain family members. If a person has more recently died the relationship is described that was had up to the time of death. If the person died before age 12, the relationship is scored as 7. APPENDIX B: PART II FAMILY INTERVIEW The interviewer says to the subject: "I am going to ask you a series of questions about relationships in your family as you characteristically remember them during your childhood. Current patterns of relationships in your family will be considered separately near the end of the inter- view. '1). First, I would like you to draw a diagram of your family which in- cludes maternal and paternal grandparents, your mother and father, all offspring and their ages. Now I would like you to describe the type and quality of relationship that occurred between each grandparent and your mother and father. Again, I would like you to describe the rela- tionship as you characteristically saw it and heard it described, and not just how it has been in recent years." The interviewer should discuss these four relationships with the S and then hand the subject the 7 category sheet. After reading the 7 categories the interviewer should help the subject reach a decision. The subjects answers are scored as follows: 7. PGF-F 1 2 3 4 5 6 7 8. PGMPF 1 2 3 4 5 6 7 (1_7) 9. mores 1 2 3 4 5 6L__];___ 10. names 1 2 3 4 s 6 7 2). "How did each of these four grandparents tend to accept their son or daughter-in-law and their marriage. was it easily and readily accepted, or with difficulty or not at all?" scored: 1) easily and readily accepted 2) accepted slowly, with difficulty, or not at all 3) grandparent deceased or absent at time of marriage 11. PGF___ 12. PGM___ 13. MGF____ 14. nou____ (1‘3) 3)_ "I would like you to describe the quality and type of relationship that generally occurred between each parent and child. Again, describe the relationship that you characteristically remember as you were a child growing up; current relationships will be discussed later." 91 92 The interviewer again writes down a description of each of these rela- tionships and discusses them.with the subject. The interviewer then hands the subject the 7 category sheet and helps him reach a decision. The subjects answers are scored as follows: 15. F-Cl 1-7, 16. F-C2 1-7, 17. F-C3 1-7, 18. F-C4 1-7, 19. MéCl 1-7, 20. M-CZ 1-7, 21. M-C3 1-7, 22. MeC4 1-7 4)_ The interviewer asks: "Were any of the children in your family perceived as being emotionally closer or more identified with one parent than another? Was one parent characteristically more important or closer to a particular child or was there no real difference over the years? Please answer for each child which, if either, parent they were most identified with while they were growing up. Again, current rela- tionships or changes in recent years will be discussed later on." 1. boy-mother 2. girl-father 3. girl-mother 4. boy-father 5. no difference C1 1-5 C2 1-5 C3 1-5 or 8 C4 1-5 or 8 Interviewer sums: ____(1,2) Presence of cross-sexed identification: Yes/No ___ (1,2) All children reported closer to mother: Yes/No ____(1,2)1 All children reported closer to father: Yes/No §)_ "1 would like you to describe the characteristic sub-groups of rela- tionships that existed in your family. Nearly all families contain several smaller sub-groups or teams where there may be a particular closeness or involvement between certain family members. What is the most characteristic 2, 3, 4 or more person sub-group within your family?" .___ (1,2) are there cross-generational alliances Yes/No ____ (1,2) if there is a GP-P alliance is there also a P-C or GP-C alliance also reported Yes/No ____(1,2) is the nuclear family seen as a sub-group excluding grandparents Yes/No ____ (1,2) within generation alliance: ‘ 1. Parental 2. Sibling 3. None ____(l,2) is there an alliance where only one member of the nuclear family (mother, father, child) is excluded Yes/No 93 62_ "Have members of your family been in polarized all-good or all-bad roles? That is, has anyone been considered to be the problem in the family, a bad guy, a black sheep? On the other extreme, is there a person in the family who is considered the good child, a person others are supposed to be like or who is seen as doing everything right? Or is there no one in these extreme roles?" all good: Yes/No (1,2) ordinal position (1,2,3,4) or 8 sex (1,2) or 8 all bad: Yes/No (1,2) ordinal position .(1,2,3,4) or 8 sex (1,2) or 8 no-one in these roles Yes/No (1,2) 12_ "I would like you to look over this list of two person relationships and rank order them according to the strength of the bonds of emotional closeness and involvement. Rank.the strongest and most primary rela- tionships first, then the second, and so on down through the list. Again, think of the relationships as you characteristically remember them when you were a child growing up." ::2: (either grandparent) M-F M-C (any child) (1-7) F-C - GP-C C-C is the marital relationship the primary relationship Yes/No (1,2) is the nuclear family (F-F, MEG, F-C) ranked above the extended family (M-GP, F—GP) Yes/No (1,2) (1,2) Yes/No agreement with intake sheet: M-F ranked first or not first in both cases §)_ "Have members of your family experienced problems in living such as alcoholism, drug arrest, therapy, police records, school expulsion, running away from home or unwed pregnancies?" This includes children, parents, and grandparents. Yes/No (1,2) 2) "Which of these two sets of descriptions comes the closest to fitting the role your father played in the famdly as you were growing up?" Select "a" or "b" and then "c" or "d" a) confident, assertive, direct, leads, guide, teach l. or b) shy, wary, obedient, conforms 94 c) supports, sympathizes, affectionate, agreeable, understanding, warm 2. or d) aggressive, firm, forthright, critical, frank, irritable The interviewer summarizes this such that: l. a a and c 2. I a and d 3. I b and c (1-41’ 4. I b and d lQ)_ "I have been asking questions in terms of your past history. I would now like to ask you about any changes in patterns of relation- ships that have occurred in recent years (since leaving home for school, or age 18)." .___ (1,2) "Has there been parental divorce in recent years?" Yes/No ___ (1-3) "Are you becoming more interested, involved with, and concerned about your relationships with your family now or are you becoming less concerned, less involved and less interested in family affairs or is there no real change?" Scored: 1. More 2. Less 3. No change ____ (1-4) Good child 1. gone away 2. newly arose 3. remained but with new people 4. no change ____ (1-4) Bad child 1. gone away 2. newlyarose 3. remained but with new people 4. no change (1,2) Male child moved into cross-sex identification Yes/No (1,2) Female child moved into cross-sex identification Yes/No (1,2) Male child moved out of cross-sex identification Yes/No (1,2) Female child moved out of cross-sex identifica- tion Yes/No ll) "Extrafamilieal friends for parents: significant friendships out- side the family offering sources of support, meaning and companionship?" Mother 1. none or very few 2. some or many (1,2) Father 1. none or very few 2. some or many (1,2) 12) Race Parents' education (for most educated parent) ___ (1-5) 1. less than high school graduate 2. high school graduate 3. some college 4. college graduate 5. advanced professional degree. APPENDIX C CHI SQUARE TABLES FOR.MDTHERPCHILD AND FATHERPCHILD RANKINGS Four chi square analyses are presented below concerning the rela- tive ranking of the mother-child and father-child relationships. There were seven possible rating positions for each relationship. An asterisk. (*) is placed by the relevant finding discussed in the results section for each table. Table l. Ranking of Mother-Child Dyad by Males RANK. MRP MRNP 1. O 392 N I 10* 2. 482 N I 11* 82 N I 2 3. 262 N I 6 272 N I 7 4. 132 N I 3 232 N I 6 5. 92 N I 2 42 N I 1 6. 42 N I 1 O2 7. 92 N I 0 02 TOTAL 1002 N I 23 1002 N I 26 49 X2(5) I 18.52 Probability I .0024 Table 2. Ranking of Mother-Child Dyad by Females RANK .MRP MRNP l. 0 502 N I 13* 2. 482 N I 13* 82 N I 2 3. 332 N I 9 152 N I 4 4. 72 N I 2 112 N I 3 5. O2 112 N I 3 6. 72 N I 2 42 N I 1 7. 42 N I 1 02 TOTAL 1002 N I 27 1002 N I 26 53 x2(6) - 27.51 Probability - .0001 95 96 Table 3. Ranking of Father-Child Dyad by Males RANK MRP MRNP l. 0 0* 2. 132 N I 3* 192 N I 5 352 N I 8 192 N I 5 172 N I 4 232 N I 6 5. 262 N I 6 152 N I 4 6. 42 N I l 232 N I 6 7. 42 N I 1 0 TOTAL 1002 N I 23 100 N I 26 49 X2(5) I 6.40 Probability I .2689 Table 4. Ranking of Father-Child Dyad by Females RANK. MRP MRNP l. O 232 N I 6* 2. 72 N I 2* 272 N I 7 3. 332 N I 9 192 N I 5 4. 482 N I 13 152 N I 4 5. O2 82 N I 2 6 82 N I 82 N I 2 7. 42 N I 0 TOTAL 1002 N I 27 1002 N I 26 53 X2(6) - 17.67 Probability - .0071 APPENDIX D NON-SIGNIFICANT FAMILY INTERVIEW CHI SQUARE TABLES Appendix D presents the non-significant chi squares performed for the family interview. The tables are presented for each of the 10 ques- tions by sex. Significant or strong trend results were presented in the results section. Question 1 A.l Grandfather-Father Relationship for Males CATEGORY MRP MRNP 1 8 4 2 11 12 3 1 2 4 1 2 5 l 2 6 0 0 7 1 4 TOTAL 23 26 49 x2(5) . 4.00 Probability - .5482 B.1 Grandmother-Father Relationship.for Males _QATEGORY MRP MRNP l 5 4 2 10 14 3 0 3 4 4 3 5 2 0 6 0 0 7 2 2 TOTAL 23 26 49 x2(5) = 5.75 Probability - .3304 97 98 C.l Grandfather-Mother Relationship for Males CATEGORY MRP MRNP 1 2 7 2 ll 12 3 3 0 4 3 2 5 l 2 6 l 0 7 2 3 TOTAL 23 26 49 X2(6) - 7.39 Probability - .2855 C.2 Grandfather-Mother Relationship for Females CATEGORY MRP guuui l 9 6 2 10 12 3 1 2 4 3 3 5 3 2 6 l 0 7 0 1 TOTAL 27 26 53 X2(6) I 3.29 Probability I .7707 99 D.1 Grandmother-Mother Relationship for Males CATEGORY MRP MRNP 1 6 ll 2 5 4 3 l l 4 7 8 5 2 l 6 2 l 7 0 0 TOTAL 23 26 49 x2(5) - 2.13 Probability - .8295 D.2 GrandmotheréMother Relationship for Females CATEGORY MRP MRNP l 8 6 2 10 10 3 1 4 4 6 5 5 0 1 6 1 0 7 1 0 TOTAL 27 26 53 X2(6) - 5.15 Probability - .5235 Question 2 1. 3. 4. Paternal Grandfather for Males MRP MRNP EASILY 14 14 DIFFICULTY 1 2 TOTAL 15 16 31 x2(1) - .000 Paternal Grandfather for Females Probability I .9531 MRP MRNP EASILY 13 21 DIFFICULTY 5 2 TOTAL 18 23 41 x2(1) - 1.42 Paternal Grandmother for Males Probability I .2327 MRP MRNP EASILY 18 16 DIFFICULTY 2 3 TOTAL 20 19 39 x2(1) - .000 Paternal Grandmother for Females Probability I .9510 MRP MRNP EASILY 17 19 DIFFICULTY 9 3 TOTAL 26 22 48 X2(1) - 1.79 Probability I .1809 an... 8. Maternal Grandfather for Males 101 MRP MRNP EASILY 16 14 DIFFICULTY 3 2 TOTAL 19 16 35 x2(1) - .043 Maternal Grandfather for Females Probability I .8354 MRP MRNP EASILY 15 19 DIFFICULTY 7 4 TOTAL 22 23 45 X2(l) I .606 Probability I .4361 Maternal Grandmother for Males MRP MRNP EASILY 16 17 DIFFICULTY 2 6 TOTAL 18 23 41 x2(1) - .646 Maternal Grandmother for Females Probability I .4215 MRP MRNP EASILY 20 21 DIFFICULTY 4 5 TOTAL 24 26 50 x2(1) - .017 Probability I .8945 (Question 3 3. 102 Father-Child 1 Relationship for Males CATEGORY MRP MRNP l 3 O 2 ll 15 3 6 4 4 3 4 5 O 3 6 O O 7 O 0 TOTAL 23 26 49 x2(3) - 7 Probability - .1358 Father-Child 1 Relationship for Females CATEGORY MRP MRNP 1 9 4 2 12 14 3 2 3 4 3 2 5 l 3 6 O O 7 O 0 TOTAL 27 26 53 x2(4) - 3.45 Probability - .4841 Father-Child 2 Relationship for Males CATEGORY MRP H CON-FLQJ-‘Ug NO‘UIJ-‘UNII‘ OOOO\-§‘O# TOTAL 23 26 49 x2(4) - 3.60 Probability - .4625 103 4. Father-Child 2 Relationship for Females CATEGORY MRP MRNP l 6 4 2 14 12 3 4 6 4 2 2 5 l 1 6 0 0 7 0 0 TOTAL 27 26 53 x2 (4) - 1.26 Probability - .8867 5. Father-Child 3 Relationship for Males CATEGORY MRP MRNP 1 0 0 2 16 11 3 l l 4 1 4 5 1 l 6 0 l 7 0 0 TOTAL 19 18 37 x2(4) - 3.70 Probability - .4479 6. Father-Child 3 Relationship for Females CATEGORY MRP MRNP 1 5 1 2 8 13 3 2 l 4 4 2 5 1 l 6 0 0 7 0 0 TOTAL 20 18 38 x2(4) - 4.76 Probability - .3123 104 7. Father-Child 4 Relationship for Males CATEGORY MRP ‘MRNP 1 0 2 2 1 3 3 l 0 4 3 l 5 0 l 6 O O 7 O 0 TOTAL 5 7 12 22(4) - 5.82 Probability - .2132 8. Father-Child 4 Relationship for Females CATEGORY MRP MRNP 1 5 2 2 1 4 3 0 1 4 2 0 5 0 0 6 0 0 7 0 0 TOTAL 8 7 15 x2(3) - 6.04 Probability - .1094 9. Mother-Child 1 Relationship for Males CATEGORY MRP MRNP l 3 4 2 14 13 3 2 2 4 3 6 5 1 1 6 0 O 7 0 0 TOTAL 23 26 49 X2(4) I .999 Probability I .9098 105 10. Mother-Child 1 Relationship for Females CATEGORY MRP MRNP l 5 3 2 11 16 3 2 3 4 8 4 5 1 0 6 O 0 7 0 0 TOTAL 27 26 53 x2(4) - 3.94 Probability - .4139 ll. Mother-Child 2 Relationship for Males CATEGORY MRP MRNP l 4 7 2 9 12 3 5 3 4 5 4 5 0 0 6 0 0 7 0 0 TOTAL 23 23 49 x2(3) - 1.68 Probability - .6413 12. Mother-Child 2 Relationship for Females CATEGORY MRP MRNP 1 6 1 2 10 15 3 6 5 4 4 2 5 0 2 6 1 l 7 0 0 TOTAL 27 26 53 x2(5) - 7.31 Probability - .1984 106 Mother-Child 3 Relationship for Males CATEGORY MRP MRNP 1 2 2 2 10 9 3 O 2 4 7 5 5 O O 6 O O 7 O 0 TOTAL 19 18 xzca) - 2.36 Probability - .5010 Mother-Child 4 Relationship for Males cartoon! MRP MRNP 1 0 1 2 1 2 3 1 2 4 2 2 5 1 0 6 0 0 7 o 0 TOTAL 5 7 x2(4) - 2.40 Probability I .6626 _QATEGORY MRP MRNP 1 2 1 2 2 5 3 0 0 4 4 l 5 0 0 6 0 0 7 0 0 TOTAL 8 7 x2<2) - 3.36 Probability I .1857 37 12 Mother-Child 4 Relationship for Females 15 Question 4 A.1. C01. C02. 107 Identification Patterns for Oldest Child as Reported by Males Relationship MRP MRNP Total oy-mother 4 9 13 irl-father 6 l 7 irl-mother 4 5 9 y-father 1 2 3 0 difference 8 9 17 Total 23 26 49 x2 (4) - 5.83 Probability I .2117 Identification Patterns for Child 3 as Reported by Males Relationship MRP MRNP Total oy-mother 6 9 15 irl-father 2 2 4 irl-mother 4 1 5 Boy-father 3 1 4 0 difference 4 5 9 Total l9 18 37 x2(4) - 3.48 Probability I .4799 Identification Patterns for Child 3 as Reported by Females Belationship MRP MRNP Total y—mother 3 6 9 irl-father 3 4 7 irl-mother 4 4 8 Boy-father 3 1 4 0 difference 7 3 10 otal 20 18 38 22(4) - 3.64 Probability I .4538 D.2. Question 5 108 Identification Patterns for Child 4 as Reported by Males ‘ elationship 'MRP MRNP Tota1__ Boy-mother 2 2 4 Girl-father 0 0 0 Girlemother l l 2 Boy-father 0 3 3 mo difference 2 1 3 Total 5 7 12 x2(3) - 3.08 Probability - .3786 Presence of Cross-sexed Identification for any Child Reported by Males MRP MRNP Total es 10 16 26 BEL 13 10 23 [Total $1 26 49 X2(l) I .8834 Probability I .3582 Presence of Cross-Generational Alliance in the Family for Males MRP MRNP TOTAL has 10 16 26 mo 13 10 23 ”TOTAL 23 26 49 X2(l) - .955 Probability - .3284 Presence of Cross-Generational Alliances in the Family for Females \ MRP MRNP TOTAL ES 16 18 34 O 11 8 19 OTAL 27 26 53 X2(1) I .221 Probability I .6382 109 B. Cross-Generational Alliances Repeating Across 2 Generational Teams for Males MRP MRNP TOTAL YES 2 5 7 NO 20 20 40 TOTAL 22 25 47 X2(1) - 1.10 Probability - .5752 C. Cross-Generational Alliances Repeating Across 2 Generational Teams for Females D. Nuclear Family as a Subgroup for Females MRP MRNP TOTAL! YES 2 1 3 NO 25 25 50 TOTAL 27’ 26 53 X2(1) - .0011 Probability - .9732 E. One Nuclear Family Member Excluded for Males IMRP “MRNP TOTAL YES 1 3 4 NO 22 23 45 TOTAL 23 26 49 X2(1) - .155 Probability - .6931 MRP MRNP TOTAL YES 4 1 5 NO 23 25 48 TOTAL 27 26 53 X2(l) - .802 Probability I .3704 Question 6 Presence of All-Good Role for Males 110 MRP MRNP TOTAL YES 6 10 16 NO 17 16 33 TOTAL 23 26 49 X (1) I .380 Probability I .5375 Presence of All-Good Role for Females MRP MRNP TOTAL YES 7 10 17 NO 20 16 36 TOTAL 27 26 53 X2(l) - .466 Probability I .4946 Presence of All-Bad Role for Males MRP MRNP TOTAL YES 7 10 17 NO 16 16 32 TOTAL 23 26 49 x2(1) - .083 Presence of All-Bad Role for Females Probability I .7730 _ MRP ‘MRNP TOTAL YES 9 10 19 NO 18 16 34 LToTAL 27 26 53 22(1) - .010 Probability I .9182 Question 7 Question 9 A. 111 Problems in Living for Males MRP MRNP TOTAL YES 8 11 19 NO 15 15 30 TOTAL 23 26 49 X2(l) . .060 Probability I .8050 Problems in Living for Females MRP MRNP TOTAL YES 10 9 19 NO 17 17 34 TOTAL 27 26 53 X2(l) - .010 Probability I .9182 Parental Divorce for Males MRP MRNP TOTAL YES 1 3 4 NO 22 23 45 TOTAL 23 26 49 x2(1) - .155 Probability I .6931 Parental Divorce for Females MRP MRNP TOTAL YES 2 3 5 NO 25 23 48 TOTAL 27 26 53 x2<1) - .481 Probability I .5033 112 Changes in Good-Child Role for Males CATEGORY MRP MRNP TOTAL Gone away Newly arose Remained but new people No change 18 20 38 TOTAL 23 26 49 X2(3) I .925 Probability I .8194 Changes in Bad-Child Role for Males CATEGORY MRP MRNP TOTAL Gone away 6 13 Newly arose 5 Remained but new people 0 1 No change 13 14 27 TOTAL 23 26 40 x2(3) - 1.43 Probability - .6972 113 Question 10 Significant Friendships Beyond Family for Mather (MALES) MRP MRNP TOTAL YES 7 8 15 NO 16 18 34 TOTAL 23 26 49 X2(1) - .081 Probability I .7755 Significant Friendships Beyond Family for Mather (FEMALES) MRP MRNP TOTAL YES 9 6 15 NO 18 20 38 TOTAL 27 26 3;; x2(1) - .2742 Probability I .6005 Significant Friendships Beyond Family for Father (MmLES) MRP IMRNP TOTAL YES 10 12 22 NO 13 14 27 TOTAL 23 26 49 X2(1) - .0099 Probability I .9205 Significant Friendships Beyond Family for Father (FEMALES) MRP MRNP TOTAL YES 10 12 22 NO 17 14 31 TOTAL 27 26 53 X2(l) - .155 Probability - .6932 114 oo.a ouo.u ono.u Hum.n oaH.u «no.1 Hmm. mom. moo. aoaoaoaououm oo.H mao.u mom. ooa. and. mr~.u ann.- ooa.u araoaoouaoz oo.a «Ho.u awa.u ao~.u Nae. Hue. cad. oea co.H aha. emu. oca.u nNm.u mua.u aoaooaoaH 8:333 oo.a moo. oaa.- aom.n aoa.u hoaoaooH co.” aoo.u ~nm.i omo.u aoeaoaoomoH oo.a oao. use. mooaaoaa oo.H coo. moaoooou oo.H aboaaaau .oooue .oooaoz man .ooaooH .man.oH .oaH .oaH .oaooH .oaH NHmaflz ZOHH