II. '7. WV, . I I I‘ I. 3-2.“ " x“. Y I'IIIIIII o‘flrIH'll'III ”k: ”HUI. “II”. II‘JI‘ \WJ‘II‘TtI' I I “I“ I .IIIIII.1II"‘ 'II'III‘ IIIIIIIIM}M II IIIIJTM" ... I 'I‘ I I I' II III' “‘ ' IIIIII II’III 1" IIIIII “I I. 'I" I . I III II I I I I III II III,II:II‘IIAII|I|III'II ’.II'IIII. 'TI :II0I‘3"I.:“II". “VIIIIII r“. I I II “MIII'W I‘I‘I'III II II It] I ' III’ I III III'I‘IIIIIIIIII' |.'I' I‘I IIIIIIIIMII‘I I-III III‘I IIIIIIIIIIII IIIIIIIIIIr‘II' I I 'I IIIII IIIIIIKIII‘III I IIIIIIIIII . III II III” “III.“ II' IIII‘IIIIIII III IIIIIIII" . IIIII ‘ IIIII'I III mlI'IlI III. . ' . I" III“ I III . II" " . IIIIIIL I'I, I' ." “I ”II” "' 'III‘ IIIIIIIIIIIIIIIIII.I’II:II' I ”III" II' .II. I‘III. ' IIII‘IIIIII ‘IIII‘IIII I II III- “II II”In”: ”II ' WW II' '.II' 'I.;I"II" . ‘II' ‘ .II'IIIIII' "I II ' 'I'If‘II III III IIIIIII' II 1..“ IIII.I‘ I IIIIIII‘II IIII' II‘IMIII . .IIIII'I'III IIIII IIIIIII IIIII III . ‘I II‘III II'ILIIIVII' "“I III III “'I' " ‘1 ' I ‘HI'IIH" "' " I 'I- H ,IMI ' I. IIII'W'IIIII IIIWIIIfiI’II I II I'IIIIII'I IQ. III 'IIIII‘:,I II'IIIIIIII'II‘I'IIIIIIII “II, I' IIIIIIIIIIII'IIIIII‘ ‘.I‘ 'I" IIIIIIIII' "II-III .IIIIIIIIIIII,‘ III , .‘I I. IIH III I I III” I N I III I'N nu" IJIII I,’ ".f “I. . “III I III:III'I":‘ “ II'II‘I'I‘IIIIIIIIII I II‘I I “I.“IIIIII ‘I IIIIIgII‘IIIIIII III'I 'qu v‘IllI w‘ttgql 'I'IIII:;~£VI’IIIII’ yEII'1III IIIII'II ‘ “III-”10"?"JI‘JI‘I‘IIIII,III' 'III'rIIIIII‘H‘ I " ?. .' ‘1‘.“ ”I IIII LIX” I“ I‘IIII‘II'QII‘ I‘I-I'I‘I IIMfiI" I‘“ I III-IIIIIII .I IIII'I'I. HIIIIBI 'I I" “II 'I'II; III-III I I. “I l . I I II IIIIIlIII‘I III III’IIMi IIb .I'IIII'IE‘TNII I I III'II I III! 'III II‘IIII IIIII “III ‘I‘ Ifi'III 'IIIfi‘I IIHIIIIIIIIIIII‘III . rIII IIIII' I I‘I‘ a II I IIIIJIII I'III' LEM IIINI‘III I IL'thI IL, . . I‘II IQ)” I III' qr . {Ifirfiidw ~ . I :I I". J”) I I “III IIIIIIII IIIUIIIIIIIIIIIII ‘I' .IIIIIJI . >~ . .v ..l‘ :1" wufr» ‘Iquu. II N . ‘II‘IIL’F IIII) I I I III-III; . ' II III; I :.I..II I. I I _ 1'. .‘I III. «I ‘III III-(“III III" III.- I’ 'I' "‘I IIIII'II _II‘II‘.. ~ .I II, ' I. WW“ II IIIII!’ III II; N" I I I I". I IIIIIIII I II I I I. I. I I I ‘ .I‘I'UII IIIH‘ III"... [II . . ‘ I. I. - I .‘lI . . I . '. IIII'I‘. II'I I I ”I" I: .‘ "I: 0'... I I a; . ‘I II‘II IIIII'IIIII '5' ..I I Dug—VI I- . . I H 'I_'.I.I:': .‘II'WII'II'I E with . '..' I . . -; .3: ”If. . ‘ . I I. . I . I I . I . I‘. . I ‘ I» "I I :« IIoI .' I. I“ I I I . II III I I . I I I I . II. II I . "I. -I .‘II. . . "I-III‘iII -(III I I IIIIIII’. III] Inf. \IJ III (I ")II’A II: II I'.’ I.‘ I'I I -."-‘ . .I‘II uI‘ IIII' II‘I. III.IIII . ”IW;II'I- I‘I‘I'IIZIIf’I I " , I ‘ ' . I II ‘ I I'.'III " “I “f . - | I I I I I ' III. I ,‘ I I II II I I III I I 'I'III I. '-I r... H "JI-I II wr‘Iv‘" ' #6: ‘ f‘IHTI—{II' I 'I I I I . I ' ' II ." I“ I 'I ' I I I I . I III] . I ~ . ‘,. . I III '. .- I I 'I I II II .II I IIIIIIIII'III . IIII II II III IIII 'Il'I ‘III’IIII/IVI.I~\IIIII;I "IIII‘I I ' . .I’ I ., I. . ., . ' '. ‘ -. III I- I, I I . .." . IIIIII . I .I. v I . III. I.‘ II‘. I ‘. II ‘ IIIII I .I' “II I“‘H‘ IIIIII.‘ I. ‘II I' I III.” I ‘l'IvJ‘ ‘IIIIO .‘I I. III.I'4.'I-‘.,I I. we! RH" I” I I 0‘?! I I... ff r I. . I‘ bI'IIf IIII I II. I . II III .: I“‘.’I.I' I.I.I.I"'I..§IIII‘ I I I. ,"I ' . Iv.. ' “HI" I .' II I III I I . I '. I I,” II IIII'. I,IIIIIIIII I II III .I III I .IIIII’ (H ,I .I'IIIIIIII .' I I I'I .I :1. I a 'I III] "II? ijw \l 1 '{I II ”I ' 'II II-‘IIfI‘I 'I I‘ ‘ I I I ‘ ‘ ‘I'I“I' .II'I I‘I ' I 4“"'III'D(';._ I I I II II II III I I I ‘ 'I‘ I'IIIII I I II/I'I‘I III II I II . I .II)!“ II'::}III|‘ flIII‘L‘hI‘L II“. ‘ L III. I II I '4 ‘ I “ IIIII‘ I .“ IOII ‘ III.” .II IIIII‘I'III I.IIII.I. I' I.I.. I.II . III 'I‘ I' Fl I “I" I II II 'II t I .I, II. I “ ‘..'I"I "':""I 'Iv‘II'I IIIII‘VII .‘~J""' .I'f'lI-I IIII’I. .III ‘ "III I‘ I " ' I' 1‘“ I I""" ‘II III.‘ III-I I ”I" IIIIII "I “III III 'I. H III. I I . I . I‘III I“I:' IU 'IIIIIII I'II' i” “In“, III‘ IIIIIII I I I I. I . . . , l‘ I I ~ I I' I“ II 1. ‘ . ‘nII I_ III. "I'M" I-I I-I'IIIIIII “I IIIIIIIII ‘I " ‘ ‘ ‘ . ' I}. II‘II III\',"' ..,II . ~ ‘.‘II II‘II ,I. "I . I'~I‘I I.I II I I II :1?“va II .I I II‘jII‘I‘III II” II LII I .II I! IIIII-IIII' ‘I' IIII,.,III . I It’. I I. III ”JIM . ....I ‘I .. r‘ f I I III I I)” I II I" III‘I‘ 2” III .I ..I.'..'.&‘fImI ILIIn‘. “I'm ..I".II”|' IIII .MHII .‘ If“; H ”I‘lfll II IHIJI IIIIIIfiIiIImI . ' LIBRARY Michigan State University rugsvs 3 1293 10062 8670 This is to certify that the th’essis entitled ABUSIVE VS. COMPETENT MOTHERS: PREDICTING PARENTING BEHAVIORS FROM SELF-REPORTED LIFE HISTORY VARIABL‘ES presented by Diane Elizabeth Johnson has been accepted towards fulfillment of the requirements for Ph.D. degree in Psychology Major professor Date May 30, 1979 0-7639 OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. DH; 01 2008 mg 120408 1 .4? . { MW\TTTJLT\\\\\1\\M\\2WT§T§T\M rig“? 31 Ear 5&3;th This is to certify that the theSis entitled ABUSIVE VS. COMPETENT MOTHERS: PREDICTING PARENTING BEHAVIORS FROM SELF-REPORTED LIFE HISTORY VARIABLES presented by Diane Elizabeth Johnson has been accepted towards fulfillment of the requirements for Ph.D. 4169.66 in Psychology Major professor / Date May 30, 1979 0-7 639 OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. r\ I MCU “0 £3. :2 ' 3:“. ‘f‘ _- A ELL: 012003 mfgg 1204 08 L (‘9 Copyright by DIANE ELIZABETH JOHNSON 1979 ABUSIVE VS. COMPETENT MOTHERS: PREDICTING PARENTING BEHAVIORS FROM SELF-REPORTED LIFE HISTORY VARIABLES By Diane Elizabeth Johnson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1979 ABSTRACT ABUSIVE VS. COMPETENT MOTHERS: PREDICTING PARENTING BEHAVIORS FROM SELF-REPORTED LIFE HISTORY VARIABLES by Diane Elizabeth Johnson Determinants of mothering behavior were investigated by applying a variety of measures of developmental processes, personality attributes, and demographic characteristics suggested by the contemporary literature to samples of child-abusing and unusually competent mothers. More than direct preparation for their eventual role as mothers, culturally prescribed myths appear to guide most girls toward this goal. Perhaps contributing to the dissatisfaction of mothers in our culture is evidence that the mothering role may be becoming more stereotyped and narrow. The widely held assumption of repetitious parenting behavior across generations suggests that women may be especially vulnerable to unresolved symbiotic attachment to their mothers. Theoretically, failure to resolve automony issues during the second stage of ego-development (1% to 3 years) is related to the formation of destructive life-scripts in early childhood, which in turn heavily influence such subsequent behaviors as childrearing. To the extent that the role of motherhood is transmitted as traditional Diane Elizabeth Johnson and institutionalized, rather than as an autonomously experienced process, the individual is discounted. In extreme cases, child- abuse may be one result of such a process. Seventeen abusive and twenty competent mothers were recruited from rural and urban locales. Each mother responded to interview questions about early childhood parenting experiences, the meanings she attached to those experiences, the developmental history of her expectations about parenting, and how these compared to her actual experiences. Each subject also completed five instruments, including demographic information, life-stress (Schedule of Recent Experiences), Erikson's developmental stages (E-Scales), empirically-derived predictors of parent-child difficul- ties (Michigan Screening Profile for Parenting--MSPP), and two checklists derived from Transactional Analysis (TA) theory about parents' problem-solving behavior and time structuring. All interviews were audio-taped and independently assessed by teams of two or eight raters for the presence of items represent- ing ten variables drawn from TA "life-script" theory. Raters also judged the quality of scripting for parenthood, the age that the basic life decisions were make, and the ego-state responsible for decisions about parenting. Six of the raters were Clinical Member trainees (intermediate level), while two were at more advanced levels in the International Transactional Analysis Association. The variables from each method were initially screened using stepwise discriminant function analysis to select only those Diane Elizabeth Johnson which significantly related to the criterion of abuse versus competence. Comparison with direct solutions showed minimal losses of predictive efficiency. Rated TA variables emerged as the best predictors of group membership (89.2% correct classifica— tion), closely followed by self-reported TA and demographic measures (both 83.8%), and trailed by the Ericksonian (78.4%), MSPP (73.0%), and life-stress (58.6%) variables. Suggesting that TA variables were especially useful for accurately assigning these mothers to the abuse and competence groups was the fact that six TA measures (3 rated and 3 self-reported) played important roles in the final discriminant function solution. This final solution, which also included MSPP Emotional Needs Met, ESZ (Automony), ES7 (Generativity), Presence of Telephone, and Number of Children, yielded 100% correct classification into the abusive versus competent groups. Thus, these groups differed notably in some aspects of their upbringing and their reactions to it. A pg§t_hgg analysis of linkages among the eleven most discriminating variables supported a theoretical bridge between Eriksonian and TA theories, along an Autonomy (E52) versus passivity (TA Discounts) dimension. Examination of the relationships among TA variables unexpectedly highlighted the importance of fathers' expectations in these women's life scripts. Time structuring also appeared to play an important role in competence, as constructive time use (one manifestation of life-script decisions) was strongly related to the criterion of group membership. The patterns among these TA variables generally appeared consistent with TA theory. Diane Elizabeth Johnson The Erikson and SRE findings, which differed from previous studies, were discussed. Based upon the major results, strategies for assessment, prevention, and therapeutic intervention with abusive mothers were proposed. To Bob and Andrea, whose love means so much . ii ACKNOWLEDGMENTS Many people have contributed to the successful execution of this study. First were my committee members, John Hurley, Denise Coburn, Lucy Ferguson, and Gary Stollak. They helped me throughout to clarify my ideas and methodology. I found them to be encouraging, and to offer many helpful suggestions. Judith Taylor and Bob Wilson gave highly qualified assistance in establishing the methodology and statistical format of the study. Bob also aided in formulating the TA interview questionnaire, rating scales, and rating procedure. Dr. Ray Helfer of MSU's Medical School provided the forms for the Michigan Screening Profile for Parenting. Both Helfer's office and Dr. James Hoffmeister's office at the University of Colorado provided scoring services for that instrument. Many agencies gave permission to contact their clients, and I especially want to thank here the many workers who took the time to make personal contact and encourage their clients to participate: Roberta Rodgers, Kay Williams, Joan Haug, Rex Hedge, Jim Brown, Ethel Comer, Terry Zelenka, Vonnie Haas, Dave Campbell, LaVonne Bennett, Janie Rodriguez, Pat Zelenka, Claire von Buchwald, Sharon Reinhart, Ann Hallock, Maureen Cleary, Carmen Gomez, Catherine Cole, Lou Krash, and many other staff in their agencies. The Ionia Cooperative Extension Service, Lakewood Public Schools, Belding iii Public Schools, and the Family Growth Center all provided interview- ing space, and the latter provided child care. Beth Waldron, Cynthia Gutierrez, and Helen Paine contributed volunteer time to help give continuity to the experience when several mothers were scheduled simultaneously. The Lansing Council on Child Abuse provided free transportation for Lansing area mothers when needed. The TA raters comprised another special group who spent about forty hours of their busy lives for little reward other than their interest in the project. I am grateful to Denise Coburn, Sandy and Garry Schwartz, Mark Nestrate, Rob Flanders, Kathy Koch, Virginia Martinat, and Larry Sarbaugh for their clinical contribu- tions as well as their sustained interest and thoughtful feedback on the interviewing and rating process. iv TABLE OF CONTENTS Page LIST OF TABLES ............................................... vii LIST OF FIGURES .............................................. viii LIST OF APPENDICES ........................................... ix INTRODUCTION ................................................. l Assumptions About Parental Background .................... 2 The Sociological Model ................................... 3 The Transactional Analysis Model ......................... 4 Origins of Parental Scripting ............................ 5 The Consequences of Motherhood Scripting ................. 7 The Alternatives to Scripted Motherhood .................. 8 Summary of the Problem ................................... 9 METHOD ....................................................... ll Instruments .............................................. ll Demographic Assessment ............................... ll Transactional Analysis Instruments ................... l2 Erikson Instrument ................................... l4 Michigan Screening Profile for Parenting ............. l4 Subjects ................................................. 16 RESULTS ...................................................... 20 Final Discriminant Function Analysis ..................... 2l Post Hoc McQuitty Analysis of Important Predictor VarTEEles .............................................. 23 Post figg_Review of TA Variables .......................... 24 DISCUSSION ................................................... 27 Findings ................................................. 27 Differences in Upbringing ............................ 27 Prediction Potential ................................. 31 Contributions to Theory of Parental Abuse/Competence . 32 Contributions to Personality Theory .................. 37 Implications for Future Research ......................... 40 Implications for Treatment and Prevention ................ 41 V Page APPENDICES .................................................. 48 BIBLIOGRAPHY ................................................ 129 vi Table LIST OF TABLES Page Summary of Discriminant Function Analyses of Six Sets of Data for Predicting Memberships in Abusive and Competent Groups .................................. 44 Discriminant Function Analysis of Best Predictors from Initial Screening Analyses for Predicting Membership in Abusive and Competent Groups ....................... 45 vii LIST OF FIGURES Figure Page l. Graph Comparing Percent Correct Classification into Abusive and Competent Groups for Direct and Step-wise Discriminant Function Analysis of Six Variable Sets .... 46 2. McQuitty Elementary Cluster Linkages between the Ten Best Predictors of Abusive and Competent Group Membership Selected by the Discriminant Function Analysis ............................................... 47 viii LIST OF APPENDICES Appendix A. Structured Interview: Scripting for Parenthood ....... 8. Training Manual for Rating Script Elements from Audio-Taped Script Interviews ......................... C. TA Script Variable Rating Forms: Decisions About Parenting and Early Programming (Script) Checklist .... D. Rater Reliabilities and Homogeneity Estimates for Clinician-Rated and Self-Reported TA Variables from Sample Data ........................................... E. Self-Report TA Checklists A and B ..................... F. Estimates of Homogeneity of the Erikson Developmental Scales (ES) from Sample Data .......................... G. Assessment of Mothering Abilities and Practices ....... H. Sample Recruitment Letter ............................. I. Variable Individual Scores, Means, and Standard Deviations by Criterion Group ......................... J. Measures of Association between Demographic Variables and Membership in Geographic Groups and Abusive/ Competent Groups ...................................... K. Scoring Outcomes for Six Variable Sets by Direct and Step-wise Discriminant Function Analysis .............. L. Pooled Within Group Correlation Matrix for Eleven Best Predictors of Group Membership .................... M. Pooled Within Group Correlation Matrix for All Clinician-Rated and Self-Reported TA Variables .... ix Page 49 54 83 86 93 95 98 TOT 103 116 118 125 127 INTRODUCTION Adler (1939) stated that the whole of mankind is endangered if mothers fail. Skillful mothers, he noted, have been educated for motherhood from an early age and in such a way that they "like the prospect of being a mother, consider it a creative activity, and are not disappointed by their role when they face it in later life." Though much attention has been paid to the short-term effects of parenting behaviors on the learning abilities and social competencies of young children (Baumrind, 1967; Bell & Ainsworth, 1972; Clarke- Stewart, 1973; Hess & Shipman, 1965), little has been learned about the precise nature of education and experience required in childhood to produce skillful mothers. The purpose of this study was to investigate, albeit retrospectively, the childhood experiences of women who became mothers. In particular, we wanted to demonstrate that the "education" of women who became skillful parents, satisfied with their role, was different from that of women who grew dis- satisfied with their role and who functioned poorly as parents. 0f the group who fail as parents, mothers who physically abuse their children are clearly at one extreme. With the growing emphasis in our culture and in its professional literature on the problem of child abuse, it was expected that information about the upbringing of abusive mothers would be useful in strengthening both treatment and prevention efforts. It was with this goal in mind that abusive women were chosen as the contrast group for this study. Assumptions About Parental Bagkground There is a pervasive assumption throughout the literature on child-rearing and abuse that parents raise their children as they were raised (Fraiberg, 1967; Helfer & Kempe, 1976; Steele, 1974). Specifically, it is believed that parents who abuse their children were themselves abused in childhood (Bakan, 1971; Burland, et al., 1973; Disbrow, Doerr, & Caulfield, 1977; Gil, 1971; Helfer, 1973; Kempe, et al., 1962; Komisaruk, 1966; Morris & Gould, 1963; Reiner & Kaufman, 1959; Steele 8 Pollack, 1968; Tuteur & Glotzer, 1966; and Wasserman, 1967). There is some support for this view from observa- tion. Thus, Helfer,et al. (1977) found that a high percentage of abusive parents had failed to have their own emotional needs met as little children (80-95%), where as 60-70% of those designated as "good" parents reported "happy childhoods." There is also evidence, however, to refute this assumption. Gil (1970) found that only 14.1% of mothers and 7% of fathers in a nation-wide abusive sample had been childhood victims of abuse. Jayaratne (1977) and Gelles (1973) both pointed out the lack of normative comparison) groups in these early studies. Kadushin's (1974) review concluded, "there is little valid evidence to support the theory that abusive parents were themselves abused as children." The Sociological Model If the relationship is not a direct one between early experi- ence and later performance, where can we look for explanations? Gelles (1973) proposed a sociological model formulating patterns of demographic variables associated with child abuse. Empirical studies of abusive parents have confirmed that they tend to belong to working or lower socioeconomic classes (Bennie & Sklare, 1969; Galdston, 1965; Gil, 1971), where intrafamilial violence is more common (Blumberg, 1964; Steinmetz & Straus, 1971), and female (Bennie & Sklare, 1969; Gil, 1971; Resnick, 1969; Steele & Pollack, 1968; Zalba, 1967). The abused child is usually quite young, often under three and a half years (Bennie & Sklare, 1969; Galdston, 1965; Kempe, et al., 1962; Resnick, 1969). The social context typically includes unemployment, especially of the father (Gil, 1971; O'Brien, 1971), an unwanted pregnancy (Bennie & Sklare, 1969; Kempe, et a1, 1962; Resnick, 1969; Wasserman, 1967; Zalba, 1967) in an already large family, i.e., four or more children (Gil, 1971), religious differences of parents (Bennie & Sklare, 1969) and marital conflict or disruption (Bennie & Sklare, 1969; Zalba, 1967). While this model usefully demonstrates the wide variety of stressors which may accompany child abuse, and has been supported by life event research with abusive parents (Justice & Justice, 1976), it fails to direct practitioners to why these parents fail to cope, or to specific strategies for change. The Transactional Analysis Model The concept of scripting proposed by Transactional Analysis (TA) offers a promising alternative. TA theory posits that the out- comes of all "important behaviors“ are predictable if we know a per- son's early decisions about himself, others, and what it takes to survive in the world. These early decisions are thought to affect length of life span, choice of being well or sick, choice of marital status and partner, occupation, and child-rearing positions (Berne, 1972). If the behaviors chosen in any one of these areas show the impact of early parental influence, compliance with that influence through decisions made in the first three to five years of life, and if the behaviors are organized as part of an overall program or plan with specific expectable outcomes, then they are part of a "life- script," according to Berne (1972). Steiner (1974) added that the scripting process entails giving up autonomy of choice over develop- ment of potential under severe threats to survival in the family. The greater the degree of threat, the earlier the decisions will be made, and the more self-destructive or self-limiting they will be. The Schiffs' (1971) work with passivity clarifies what happens when autonomy is relinquished. The individual is continually in the position of manipulating others to take care of her as she has given up responsibility for herself. In order to maintain such a dependent, symbiotic relationship, the person must engage in dys- functional thinking (discounting) and dysfunctional behavior (passivity). The implicit contract in the symbiotic relationship is that only one person's needs can be met. As Erikson (1950) pointed out, this mode of adaptation is generally accompanied by considerable insecurity, in the form of shame and self-doubt. He added that the abult who still despairs of autonomy experiences a revival of "oral rages and hallucinations" when the conflicts surface. According to the Schiffs (1971), a mother would be "at high risk" for the appearance of violence (one form of passivity) if she competes with her child for the dependent position in the relation- ship. If she feels trapped and without choice vis-a-vis the infant's demands for nurturance, she also appears to be at risk according to extrapolations from Erikson (1950). Justice & Justice (1975, 1976) found this to be true in their study of abusive parents. These parents exhibited unstable symbiotic patterns of relating with both spouse and child, with no one consistently getting needs met. The Justices postulated that other symbiotic parents who are not abusive may have some kind of internal mechanism that prevents them from assuming the dependent role vis-a-vis their own children. The lack of a comparison group in the Justice's (1975, 1976) studies precluded a test of this possibility. Origins of Parental Scripting Although an individual's life script is formed out of the interaction between her and her parents, the role of the parents is to act as transmitters of the culture as they understand it. A cultural mandate for women to give priority to family roles and obligations, to bear several children and to care for them alone has been well documented (Griffith, 1973; Laws, 1971; Russo, 1976; Sherman, 1976; Steinman & Fox, 1966; Wilson, Bolt, & Larsen, 1975; Zellman, 1976). As early as 1934, Horney had documented through case studies that "an overvaluation of relationships with men and parent- ing" was becoming common. Friedan (1963) observed that after 1949, the goal of nearly all American women became to seek fulfillment as wives and mothers. According to a mid-1950 study (Douvan & Adelson, 1966) of nearly 2000 girls 12-18 years old, 95% of the girls expected to marry and have children. Russo (1976) cited statistics that 80 to 85% of young women were married and mothers. Babcock and Keepers (1976) stated that "many couples still [marry and] have babies because they assume that they're supposed to have babies, and that something is wrong if they don't want to." Friedan (1963) said, "our culture does not permit women to accept or gratify their basic need to grow and fulfill their potential as human beings." Wyckoff (1974) found that "women are trained to accept the mystification that they are incomplete, inadequate, and dependent . . . . They are trained to be adaptable." Rich (1976) decried the victimization of women which has them teach their daughters self-hatred and low expectations of life, as well. These views were pre-dated by Horney's (1934) belief that there was a "social narrowing of the woman's sphere of work." About the same time, Adler (1939) observed cultural forces which devalued motherhood and homemaking, and a little later, Erikson (1950) attributed these processes to the changing outcomes expected of child-rearing. Adler foresaw that girls would not prepare themselves with enthusiasm for what was presented to them as an inferior task, and sounded an alarm which only recently has been heeded. Specific myths which influence parental scripts have been discovered, also. Women feel they must choose between a "successful professional career" and "marriage, children, and resentment" (Elder, 1977). The latter choice may lead her to see her children as depriving her from her own share of nurturing due to the infant's excessive demands (Boulton, 1977). Women expect themselves to know instinctively how to mother (Boston Women's Health Book Collective, 1976; Jongeward 8 Scott, 1976), and to be infinitely loving and patient 24 hours a day--the "Madonna and Child" myth (Justice 8 Justice, 1976). Babies are commonly depicted as clean, cuddly little creatures who smile lovingly at their mothers and who sleep peace- fully in their arms--the "bundle of joy" myth (Justice 8 Justice, 1976). The Consequences of Motherhood Scripting Friedan (1963) claimed that women who buy into these myths are remarkably incompetent and infantile as mothers, living through their children in an unhealthy symbiosis. Erikson (1950) said that for this type of mother "remnants of infantility join advanced senility to crowd out the middle range of mature womanhood, which thus becomes self-absorbed and stagnant." Empirically, well-adjusted women tended to be identified with low-feminine mothers (Heilbrun 8 Fromme, 1965), while high-feminine women more frequently experienced spontaneous abortions than less feminine women (Sherman, 1971), raised maladjusted children, reflected in child guidance referrals (Marks, 1961), and were judged more inadequate in all their assumed roles (Cohen, 1966; Sherman, 1976). To the consternation of investigators, independent raters found "masculine" women more attractive, maternal, and competent in their sex roles (Peskin, 1968). The passive-dependent life style of high-feminine women also appeared to impede full intellectual development (Maccoby, 1966; Maccoby 8 Jacklin, 1974; Sherman, 1974; Sontag, Baker 8 Nelson, 1953). The Alternatives to Scripted Motherhood TA has the added advantage of clearly specifying qualities of good and bad parenting. In particular, good parenting involves caring; cuddling, holding, feeding, and comforting infants; setting rational limits for safety and health; investing time in listening and in showing children how to do things; eXamining one's own beliefs and values to decide what to pass on to another generation; modeling after others who are doing a good job; and recognizing one's own needs and working out mutually satisfying ways of meeting needs (Babcock 8 Keepers, 1976; Jongeward 8 Scott, 1976). The development of one's autonomy as the central issue to good mothering emerged from Friday's (1977) biographical research on contemporary women. The mother must be able to appropriately end the symbiotic infancy stage and support her daughter's separation and growing autonomy in early childhood. Later, during adolescence, identification and role modeling play an important part, opening the door to choices. Jongeward and Scott (1976) emphasized that the more models a child has, the more opportunity she has to expand her repertoire of parenting behaviors, and the more stable will be her personality. Friday quoted Mio Fredland as saying, "for a woman to be maternal . . . somebody has been very maternal to them," even if "somebody" was "the father, or an uncle" (Friday, 1977, p. 221). Rich (1977), echoing Friday, said that women need a sense of their acceptability to their own mothers in their unique (autonomous) choices vis-a-vis womanhood. Rich (1977) proposed the following resolution to the cultural dilemma as follows: "To destroy the institution [of motherhood] is not to abolish motherhood. It is to release the creation and sustenance of life into the same realm of decision, struggle, sur- prise, imagination, and conscious intelligence as any other difficult but freely chosen work.” Summary of the Problem The goal of this study was to explore the early antecedents of abusive and good parenting outcomes. Theory suggests that the early interaction between the individual and her social environment is crucial to the establishment of trust and autonomy (Erikson, 1950; Berne, 1972). Steiner's (1974) contention that environmental oppression from parents creates pressure to give up autonomy clearly points to Erikson's Stage 2: Autonomy vs. Shame and Doubt (ages 1 1/2 to 3 years) as very important to scripting and the development of essentially symbiotic, manipulative, self-limiting relationships. 10 Erikson (1950) would seem to imply that particular scripting for adult roles, including parenthood, is absorbed in the next stage, Initiative vs. Guilt (ages 3 to 5 years). These ages overlap exactly with the period in which scripts are supposed to be formed according to TA. METHOD To compare the explanatory power of the models discussed earlier, instruments were chosen to measure demographic, TA, and Eriksonian variables. An additional well-known empirically derived predictive measure was included to assess its comparative efficacy for identifying abusive women. Instruments Demographic Assessment Data was obtained on current and past social milieu and life events thought to be significant in the etiology of abuse from an information sheet completed by the mothers. The Schedule of Recent Experiences (SRE) developed by Holmes and Rahe (1967) was administered for more specific data on the stressfulness of recent life changes (items and scoring criteria available in Holmes 8 Rahe, 1967). The SRE is a list of 42 events which might normally occur in a person's life. This technique assumes that such events require an individual to expend energy toward readjustment, hence cause stress. Consensus on the stressfulness of the items has been demonstrated across a variety of rater groups (all rfs > .90, except for racial group, where r_= .82; Masuda 8 Holmes, 1967). The level of life stress experienced has been shown to be related to onset of illness (Rahe, et al., 1964) and the beginning 11 12 of pregnancy (Knittel 8 Holmes, in Dohrenwend 8 Dohrenwend, 1974). Justice and Justice (1976) used the SRE to study abusive and non- abusive parents and found that the life change scores of the abusive (A) parents were significantly higher than those of non-abusive (NA) parents (EA = 234, XNA = 124, ;_= 4.28, p_<:.001). While not pre- suming that life stress caused abuse, the Justices pointed out that exhaustion associated with on-going crises may lead to lower defenses and weaker controls against acting out. Transactional Analysis Instruments Because scripting for parenthood has not previously received systematic research attention, an interview format was constructed to elicit the necessary clinical information, and a rating system was created for summarizing these data. The interview included 21 content areas tapping early rearing experiences, the person's understanding of those early circumstances, and the development of ideas and self-concepts related to eventual parenthood. This inter- view required from 60 to 90 minutes (see Appendix A). General models for written script questionnaires were available (Corsover, 1977; McCormick, 1971; Steiner, 1974) and three clinical studies have been reported (delinquents: McCormick, 1971; reading disabled children: McCormick, 1977; and prison inmates: Corsover, 1977). While no reliability or validity statistics were reported, these researchers claimed clinical validity, i.e., that experienced clinicians got both similar and meaningful results in inferring life- script elements from the information derived from these questionnaires. 13 Script information from the tape-recorded interviews in the present study was quantified through clinical judgment along pre- scribed script element dimensions (see Appendix B for the Training Manual and definitions, and Appendix C for the rating forms used). These ratings were performed by eight TA clinicians. Six were candidates for Clinical and Special Fields memberships in the Inter- national Transactional Analysis Association, one was a Special Fields member, and one was a Provisional Teaching member. They ranged in age from late 205 to early 505 and included four men and four women. All were involved professionally in teaching or mental health- related activities. All had completed 50 hours of advanced didactic and experiential training, 50 hours of supervised practice, and 150 hours of clinical experience using TA methods, and two had passed comprehensive written and oral exams on knowledge of TA theory. All raters additionally completed four hours of training specific to this script rating task. Training followed Wilson's (1976, 1979) model for teaching cognitive discrimination tasks, which allows for successive approximations to the actual task. This method has been shown to produce high interrater reliabilities quickly (Wilson, 1976; see Appendix D for the reliabilities obtained in this study). The subjects also filled out two brief TA-derived checklists enumerating different responses which they recalled their parents making to parent-child problems when they were little (Checklist A, Appendix E), and the average amount of time spent daily in the six modes of time-structUring (Checklist 8, Appendix E) described by TA theory. These self-reported data comprised an additional source of l4 script information (see Appendix D for reliabilities of self-rated variables). Erikson Instrument Since the TA model closely resembles the earlier Eriksonian model of personality development, especially on issues of autonomy and symbiosis, Evans' (1976) E-Scale (E5) was chosen to measure the subjects' mastery of the seven stages described by Erikson (1950). Evans (1976) used the ES to assess developmental mastery in a very similar sample of abusive and non-abusive mothers, and found that stages 1 (Trust vs. Mistrust), 5 (Identity vs. Role Diffusion), and 6 (Intimacy vs. Isolation) were included in the discriminant function solution for good predictors of group membership. Evans' (1976) ES was a revision of an earlier measure by Constantinople (1969), and included re-phrasing the items to enhance comprehension by less-educated, lower-income populations; he also added ten items to cover Stage 7 (Generativity vs. Stagnation). The ES consists of 70 items, ten for each stage. Evans (1976) did not report reliabilities or validities, but does list the items and scoring criteria (see Appendix F for reliabilities obtained in this study). Michigan Screening Profile for Parenting_ The development of the Michigan Screening Profile for Parenting (MSPP) by Helfer, Schneider, and Hoffmeister (1977) was the culmination of a series of efforts to devise and validate 15 an instrument for predicting the potential of parents to abuse their children. The original items were based on clinical experience which suggested that abusive parents had more anxiety about dealing with child problems, more concern about criticism and isolation, higher expectations of children's performance, and more severe physical punishment in their own childhoods than had non-abusive parents. The version of this instrument used inthe present study consisted of 50 items. It was made available by one of its authors (Helfer, MSU; currently available from James K. Hoffmeister, Test Analysis and Development Corporation, 2400 Park Lake Drive, Boulder, CO 80301). Initial test-retest stabilities (Helfen,et al., 1978) for 92 mothers appeared adequate for two scales (Relationship with Parents, 69% of the scores remained stable; Emotional Needs Met, 85%) and marginal for two (Expectations of Children, 62%; Coping, 65%). Emotional Needs Met has been identified as the best of the four scales for predicting membership in abusive versus non-abusive groups (Helfer, et al., 1977). Validity studies to date indicate reasonable sensitivity (85.7% of 98 mothers with problems correctly classified) and speci- ficity (79.8% of 138 mothers with no apparent problems correctly classified; Helfer, et al., 1977). Spinetta (1978) found that low- income abusers scored at the highest risk level for all but one of the abuse potential categories. However, his samples were extremely small (only 7 abusers and 13 neglectful parents, for example), and the middle-income comparison group produced the most significant differences. This was a doubtful base for comparisons with a 16 low-income abusive group, since socialization, values, and oppor- tunities for relief are so different between low- and middle-income groups. Subjects The subjects for this study were selected to represent extreme manifestations of parenting behavior, physical abuse versus competent parenting. Physical abuse was defined as: harm to a child's health by a person responsible for the child's health which occurs through non-accidental physical injury or maltreatment (from the 1975 Child Protection Law, State of Michigan). Competence was defined by the six following criteria selected from the child-rearing and child-development literature: 1. Does not view children or child care as a burden, or complain inappropriately about their demands being excessive. 2. Responds to interruptions or demands by child in an interested, concerned, and patient way, yet can be appropriately firm. 3. Encourages child to express questions, ideas and feelings, and responds with interest. 4. Feels good about role as mother, likes what she's doing, and who she is as a mother. 5. Seems able to solve problems with the child, and to constructively use worker's advice in relation- ship to problems with the child. 6. Seeks and enjoys recreation with and for her children. The sample consisted entirely of mothers, as there is much evidence that men's and women's socialization for parenthood is quite different in this culture (Douvan 8 Adelson, 1966). Further, race has been 17 found to be related to some of the personality variables to be studied (Evans, 1976), so race was also controlled (one black in each group, both from the same geographic area). Seventeen abusive mothers (AM) were recruited in all, ten from the Department of Social Services (DSS) Protective Service (PS) and Parent Aide programs in the rural county, and seven from a nearby urban setting's service agencies, to which they had been referred subsequent to PS investigation and verification of abuse. Many of these abusive mothers (AMs) were or had been receiving treatment, but previous experience indicates that this has no effect on the variables of interest here (Evans, 1976). The choice of highly competent mothers (CMs) for the compari- son group reflects the desire to understand how, in similar life circumstances (age, income, marital status, size and age of family, and life stress), people come to be different in the quality of parenting they offer to their children. Previously demonstrated relationships between abusive behavior and levels of stress (Justice 8 Justice, 1976) and socioeconomic level (Bennie 8 Sklare, 1969; Galdston, 1965; Gil, 1971) have been questionned severely on the grounds that they may well be spurious effects determined by avail- ability of subjects (PS referrals tend to be lower-class, multi- problem families). Therefore, every attempt was made to equalize the groups on these variables. Twenty competent mothers were recruited from nominees by caseworkers of 055 Family Services, Extension Services, low-income housing, and preschool programs. There was equal urban and rural 18 representation. Statements documenting the observed high quality of parenting and the absence of known abuse were obtained from the nominating sources (see Appendix G). Subjects were told that a group of people at Michigan State University (MSU) were studying mothering, and wanted to assess the common idea that parents raise their children the way they were raised. Parents were encouraged to contribute their own unique experiences to our knowledge of parenting (see Appendix H for sample recruitment information).‘ In accord with ethical standards of research with human subjects, mothers were promised confidentiality and anonymity in final reports, and advised that the overall results would be made available to anyone wishing them. The project was reviewed and approved by the State of Michigan 055 Bureau of Finance, in order to protect the rights of DSS recipients, and to guarantee voluntary participation. The two groups were similar in most respects on the demo- graphic variables (see Appendix I for a complete data summary). The median age of AMs was 29 years, while the median age for CMs was 28 years. The median income claimed by both groups was in the $4000- $6000 per year range. AMs tended to be married (47%), while the largest group of CMs were divorced (45%). Slightly more AMs were separated (11.8%) or never married (17.6%) than CMs (0 and 10%, respectively). The CMs were better educated (55% had at least some college) than AMs (only 17.6% had some college). AMs had somewhat more children (median = 4) than did CMs (median = 2) and AMs' children were slightly older (median age of youngest child was 31-47 19 months for both groups, but median age oldest child was 9-12 years for AMs, but 4-6 years for CMs). Both groups appeared to be experi- encing substantial life stress, as the average life stress score of each group exceeded the criterion of 300 points (Holmes 8 Rahe, 1967). Only one of these differences, education, was statistically significant using x2 and t_tests (see Appendix J). There were no important differences between the urban and rural subgroups. The two groups required quite different management tech- niques. Nearly twice as many AMs as CMs refused to participate in the study when initially approached (45% versus 26%), and more dropped out after initially agreeing to participate (4 versus 3). It was necessary to schedule AMs more times to get a completed interview (means of 2.2 versus 1.5 appointments) despite offers of transportation and child-care, and reminder phone calls. The AM drop-outs were also more indecisive. The AMs scheduled an average of three missed sessions before they quit (mode = 3), while the CMs averaged only two sessions (mode = 1) to decide. The AMs' apparent ambivalence about their participation was both exasperating and expensive for the researcher. RESULTS Because of the unwieldly and unstable correlation matrix created by the large number of variables (48) at the initial stage of data analysis and the smaller number of subjects (37), the number of variables was reduced by regressing them against the abusive versus competence criterion by groups, according to instru- ment "package." The six packages were: (a) demographic variables (already reduced to the five measures that correlated significantly with group membership); (b) SRE variables (3); (c) clinician-rated TA variables (17); (d) self-reported TA variables (12); (e) Eriksonian scales (7); and (f) the MSPP scales (4). Discriminant function analysis with stepwise inclusion (Nie, et al., 1970) reduced the total number of variables to seventeen (see Appendix K for the complete summary of these computations). The direct solutions for each package were also computed and compared to the stepwise solutions to check the loss in prediction efficiency due to eliminating variables. As can be seen in Table 1 and Figure l, the direct solutions discriminated between AM and CM groups slightly better than did the stepwise solutions. However, by selecting the most powerful predictors, the stepwise solutions lost very little power in predicting group membership, and often yielded dramatic increases in statistical significance level. On the basis of these results, it appears appropriate to use only the 20 21 most powerful predictors from each package for the final analysis. The surviving variables were: demographic-—Telephone, Number of Children, and Education; 5RE--none; clinician-rated TA--"Litt1e Professor" Program, Original Script Level, Discounts, and "Adult" Program; self-reported TA--Activities, Withdrawal, Paternal Allowers, Pastimes, Games, and Paternal Drivers; Eriksonian-- Generativity vs. Stagnation (E57) and Autonomy vs. Shame and Doubt (E52). The variable packages are presented in Table 1 in order of best to worst predictive efficiency. It is noted that the rated TA script variables were the best predictors, followed by self-reported TA variables, demographic variables, Eriksonian scales, MSPP measures, and the SRE. Percentage of correct classifi- cation using the stepwise solutions ranged from 56.8% to 89.2%, and all stepwise solutions except the life-stress (SRE) solutions were significantly (p_< .05) related to the criterion. Final Discriminant Function Analysis This study's purpose was to explore the nature and origins of abusive and competent parenting in women. To the extent that variance between abusive and competent groups can be attributed to the variables in the discriminant function solution, we may conclude that the study has identified meaningful differences between these groups of mothers. A stepwise discriminant function analysis (Nie, et al., 1970) was conducted on the group of 17 variables selected in the previous step to meet this objective. 22 Eleven variables were selected by the analysis for the final solution (Eigenvalue = 7.32; cannonical §_= .94; x2 = 62.49; g:_= ll; p_< .001). Table 2 illustrates these results. Using the discrimi- nant function equation composed of these 11 variables to predict membership in the criterion groups results in perfect classification of all 37 cases, an unusually clear and powerful outcome. From these results, we can conclude that there is a meaning- ful linear relationship between the competence of mothers and a variety of antecedent and concurrent parental conditions. Since five of the 11 variables were related to presumed childhood antecedents (TA Original Script, Discounts, and "Little Professor," E52, and MSPP's ENM), we can also conclude that maternal abuse of children is linearly related to women's perceptions and responses to their own early parenting. Further, since three time-use vari- ables are included (Pastime, Withdrawal, and Activity), we can conclude that abuse is linearly related to the way in which these women currently choose, perhaps according to their life scripts, to structure their time and relationships. The absence of telephones among abusive mothers may be similar to TA Withdrawal as an indicator of their use of isolation in structuring relationships although these two measurers did not correlate (§_= -.13) signifi- cantly. Family size (Number of Children) may be indicative of the actual difficulty of the parenting task. The presence of variable E57 (Generativity vs. Stagnation) was congruent in the sense that this stage is theoretically the time during which an adult is 23 developmentally ready to assume that mature parenting role of nurturing and educating a new generation. Post Hoc McQuitty Analysis of Important Predictor Variables To further understand the relationships among those best predictor variables singled out by final discriminant function analysis, the eleven surviving measures were subjected to a McQuitty elementary linkage cluster analysis (McQuitty, 1957) based upon a reduced pooled within-groups correlation matrix (see Appendix L). Three clusters were identified, as illustrated in Figure 2. The first cluster, positively anchored by Autonomy (E52) and nega- tively anchored by TA Discounts, labeled "Autonomy vs. Discounts," embraced six variables, including Eriksonian Autonomy and Generativity, self-reported TA Activities and Pastimes, and rated TA Discounts and "Little Professor." This large primary cluster also related to the smaller tertiary cluster through the inverse linkage of TA Original Script Level to TA Discounts. The relationships within the first cluster indicate that the mother who describes herself as having resolved issues of autonomy and control (E52) is more likely to be judged as taking responsibility in problem-solving (rated Discounts), to report spending time in social conversation (Pastimes), to show a high interest in the care and development of children (E57), and be judged to parent on the basis of information and experience ("Adult") rather than from prejudice or attempts to please others (“Little Professor"). 24 The secondary three-measure cluster appears to represent "social isolation." Across the whole sample, the number of children was negatively correlated to the amount of time these mothers reported spending alone, but positively correlated with the probabil- ity that she had a listed telephone number. Despite this general pattern, the AM subsample had mpgg_children (p_< .05), were 1§§§_ likely to have a telephone (p.< .006), and said they spent mppg time alone (p < .01) than did the CM subsample. The tertiary cluster appears to represent "early parental deprivation" and contained two variables, MSPP ENM and Original Script Level. The more positive the woman's memories of nurturing received in early childhood were, the less destructive the life- script level she chose. Post Hoc Review of TA Variables Because the present data concerning TA variables is unique, in the sense that no previous systematic quantitative research attention appears to have been given to these TA constructs, an exploratory examination of the relationships among these variables seemed timely. This discussion is based on the correlation matrix in Appendix M. On the self-reported TA variables, the women tended to describe their mothers and fathers as similar in both Passivity (g_= .37) and Allowers (§_= .43). However, Maternal Drivers were strongly associated with Maternal Passivity (§_= .58), while Paternal Drivers were strongly inversely associated with Activities 25 (§_= -.45). Thus, high perceived maternal demands were accompanied by high levels of perceived maternal withdrawal and/or manipulation, while high perceived paternal demands were accompanied by low levels of current reported use of Activities (productive work and play). The time structuring variables were associated roughly as might be expected according to TA theory. Women high in productive Activities appeared to be low in their use of Rituals, high in amount of Intimacy, and low in Withdrawal or isolation. Mothers who structure thier interpersonal relationships with Games (discomfort) appeared less able to relate around superficial, but benign, common interests (Pastimes). Among the clinician-rated TA variables, the salient relation- ships were between Original Script Level and Injunctions (p = -.76), Original and Current Script Levels (§_= .64), and between Current Script Level and Discounts (§_= -.47). These findings imply a high correlation between the number of injunctions given by parents and the destructiveness of the early script decisions. Original Script was a fairly good predictor of Current Script, which indicates that scripts may be quite stable over time. For this sample, the fewer the number of discounts reported from their own parents, the more benign was the outcome of the woman's life script. Examination of the comparable variables based on rated versus self-reported sources showed Passive Behaviors to have the strongest relationships (EMF/PB = .29 and EPP/PB = .51). The correlations between rated Allowers and self-reported parental Allowers were 26 positive but nonsignificant. Rated Drivers fared poorly, correlat- ing near zero with self-reported maternal and paternal Drivers. Three variables accounted for 55% (11 out of 20) of all the statistically significant correlations between self-reported and rated TA measures. They were Rituals (5), "Little Professor" (3), and Activities (3). Reported high use of Rituals was accompanied by more rated Passive Behaviors, more destructive Payoffs, the presence of "Gallows" Transactions, and more destructive early and current life-scripts. "Little Professor" influences on parenting were positively correlated with number of perceived Paternal Drivers, and greater use of Games and less use of Pastimes in structuring relationships with others. High use of Activities was associated with more rated Permissions, less destructive Payoffs, and low levels of rated Racket: Shame. DISCUSSION Findings The purpose of the study was to find out if abusive mothers were notably different in their upbringing than competent mothers. Differences in Upbringing While these retrospective interview and questionnaire data may not accurately represent actual circumstances, the two groups were found to differ substantially in their perceptions of their upbringing on many variables. The survival rate of the TA life- script measures was gratifying. In fact, six of the final eleven best predictors of membership in Abusive versus Compenent groups were related to scripting and script decisions. Life script analysis appears to have considerable utility for exploring the origins of parenting behaviors. The main Eriksonian Autonomy-stage conflict is that of control, which seems closely related to the responsibility conflicts characterized in TA theory by passivity. Thus, the inclusion of E52 in the discriminant equation points to an important bridge between TA and Eriksonian theories, and emphasizes the importance of early developmental mastery in the etiology of parenting behavior. The presence of Emotional Needs Met, the best MSPP predictor, also supports the importance of early nurturing 27 28 in developing good mothers. Eriksonian Generativity concerns the individual willingly assuming the responsibility for both self Egg. others. The presence of Telephone and Number of Children supports the idea that AMs are socially isolated and may be subject to increased pressures of child-rearing, and raise another aspect of control, that over reproductive decisions. The following vignettes were drawn from two actual inter- views to illustrate the differences between the groups in early parenting and current functioning. The CM with the highest discrimi- nant function score and the AM with the lowest score were chosen for contrast. The two cases thus selected had annual family incomes notably greater ($10-15,000 versus $4-6,000) than the typical woman in the competent and abusive groups, suggesting that gross level of income is not an important determinant of parenting skill. 11: Vignette 1. Lucy, a CM, was raised by her natural parents as the middle child in a family of five. Many relatives lived nearby. Lucy described her mother as "very emotional," adding, "she could not make the kids do what she wanted." Her mother said she was "always lonely," and "had worked for everything she had." Her father, on the other hand, was "very domineering . . . egotistical . . . intelligent . . . the strong one of the two," who always "had everything given to him." "He came first," Lucy said, "and I've never agreed with it." In Lucy's family, "Mpm_gave the love," while "Dad was not outwardly affectionate. . . ." When Lucy or a sib- ling was hurt, mom would "hold us on her lap . . . she re- members one spanking. Usually, dad would "sit us down and give us a really hard tongue-lashing." *Not her real name. 29 When she was five, her mother let her help with her new sister--what was special was "letting me be close." Her nicest experience with her father was during a trip she and her parents took to celebrate 8th grade graduation. One evening just she and her father went out to a movie, came back to the hotel, and he bought her coffee, just like an adult. The worst things that happened were around trust--"I get feelings from my father, strong bad vibes . . . they just didn't trust their kids, you know." On one occasion, dad banned her girlfriend. "I told him I felt he was wrong. He said he didn't care, it was his house and that's the way it would be. I almost ran away. . . ." Another time she overheard her mother tell friends that she didn't trust Lucy and her boyfriend. Lucy told her mother, "If that's really what you feel then I_really don't care anymore." Lucy's mother showed her how to do things most, explain- ing how and why it was done that way. Mistakes were not exaggerated, just accepted. However, Lucy did not feel her parents understood her problems, and "gave up talking to them" fairly early. Lucy decided that, "my kids were going to be able to . . . talk to me about problems, because I always felt strongly about not being able to . . . I wouldn't expect them to be excellent in everything, but I expect them to try. . . . I would praise them when good and come down on them when they didn't do things right, but not so hard they'd feel really guilty. . . ." She liked the way her mother gave affection, and wishes she had her dad's talent for speaking, making things understood. Lucy is now 28, married for the second time. Her first husband was not the baby's father--she got pregnant "by accident." She had felt that the child came out of love, and she should accept it, and be responsible for her mistake. She finished college, and the family has an adequate income ($10-15,000 annually). She describes her pregnancy as "exciting," and her delivery as "very easy." She says her daughter "was like a miniature human being . . . so much to learn, and I want it to learn the right things in the right way. . . ." Lucy finds being a mother "rewarding." She conSiders herself patient and easy going, but did lose her temper twice when her daughter was small, hurting her mildly the first time. The second time she did not act on her feelings, but "I told myself I wasn't going to ever have it happen again." 30 At the other extreme, we have Sara, an AM. Vignette 2. Sara* was adopted by her grandparents and raised as youngest in a family of three. Her sister, 16 years older, was "really my mother," and was "made to help raise me." Sara describes her mother as "always trying to help people, and me--a good cook, sewer." Dad was "quiet--never punished me at all, spanked me." Her sister "made up for it." She said wistfully that all she knew about her real father was "he was dark-haired and real good-looking . . . and he didn't want me." Sara thought her sister might have been happy about the pregnancy at first, but very hurt and rejected when the father wouldn't marry her. She said, "when my adopted parents did find out, they sent her to [city] to have it." She imagined that her sister might have reacted to her at first by saying, "'take it away,’ or something like that." She did feel wanted by her adopted parents, though, and wishes "I'd never been told about being adopted." When angry, Sara's parents would yell at her, "Why can't you ever do things right?" She only remembers four spankings, and considers her adoptive dad's spanking the worst thing that ever happened to her. Her adoptive mother belittled her abilities in comparison to her bio- logical mother's other daughter. The worst thing her biological mother ever yelled was, "I don't see how m-m-m-ma ever stood you." Affection was expressed through hugs, and they'd tell her if they were proud, or "take me out to dinner." She remembers when they were disappointed "they wouldn't say anything at all, really . . . I kept asking them about it . . ." They would finally say, "Why couldn't you act your age?" Sara "got hit by a car and almost died when I was 12, so after that they were kind of protective . . . I felt li-li-li I was loved more . . . when I was older." When Sara made a mistake doing chores, her mother would “show me how to do it and then make me do it . . . made me do it over again until I got it right." Sara decided she was "not going to spoil my kids at all. Gotta make sure they mind you real good . . . none of that talking back the way I used to do . . . they shoulda spanked me right then and there." Sara couldn't think of anything her parents did that she wanted to repeat. *Not her real name. 31 Sara is now 29, married, with two young children. She and her husband wanted to "have the first years all to ourselves"--they waited two years to have children. She knew her husband "wanted kids, one or two," and "everybody else was having them but me," so she finally got pregnant. Sara finished high school, and her family has a comfortable income ($10-15,000 annually). Sara says she "felt real good during my pregnancy," but "I didn't think delivery would hurt so much . . ." As a new mother, "I was scared . . . afraid I was going to drop them and hurt them all the time." Now, she says of motherhood, "It's an all day job! I thought it was going to be pretty easy, but it's ppp." She did have a friend who helped her out, as her husband now does, for which she sounds grateful. Sara says she "likes hearing my children call me 'Mom'." She "gets upset when they don't mind," and threatens to tell their dad. She also spanks them quite a bit. Last fall, in a fit of anger, she told her husband she "just had them for him, so he took them and left." They are back together now, and when she gets mad now, she leaves the room. Sometimes he helps by warning the kids to leave her alone. Things appear to be improv- ing for Sara. Prediction Potential In screening the variable packages, it was found that the clinician-rated TA variables were the best classifiers of abuse/competence, closely followed by the self-reported TA, demographic, and Erikson variables. The MSPP package was distinctly less effective, and least effective were the life stress (SRE) measures. Predictive efficiency ranged from 89.19% to 56.76% correct classification, and all step-wise solutions except life stress (SRE) were significantly related to the criterion. Several other recent studies have attempted to develop effective sets of measures for discriminating child abusers from other parents with varying degrees of success (Evans, 1976: 97.5% correct classifica- tion; Disbrow, et. al., 1977: 88% correct classification; 32 Helfer, et al., 1977: 83% correct classification of mothers having actual child-rearing difficulties). The present study's correct classification rate of 100% appears unique. The feasibility of an effective set of discriminators for abusive mothers now seems well established. There are several reasons for caution in this approach to abuse. The two studies with the highest correct classification rate (Evans and this study) need replication. Some shrinkage of predictive validity is to be expected when the measures are applied to a different sample, and considerably more if the study is a longitudinal one. The TA methodology developed here and the Eriksonian scales both have rather limited internal or inter-rater consistency (Appendices D and F, respectively), which contribute to sampling error. None of these four studies have examined in depth either the family system or the marital relationship. The latter two variables have been shown to be highly relevant to the maintenance of abuse (Justice 8 Justice, 1976). And finally, not all the variables used in each study were measured in ways that were equally meaningful or easy. The predictors which are most meaningful, accurate, and easy to administer without loss of predictive power need to be further identified. Contributions to Theory of Parental Abuse/Competence TA ratings of interview data and the self-reported TA data were both powerful in predicting differences between the groups. Abusive mothers were rated as subjected to more 33 destructive early parental scripting and as showing more passive (symbiotic) behaviors and a more negative, survival-oriented parenting program than were CMs. The father's influence on scripting of women emerged as unexpectedly potent in the present data. Fathers of CMs appeared to have high standards for their daughters, but also to allow considerable leeway for individuality in personality and style. Anecdotal support for this specific finding came directly from the interviews. At least five CMs mentioned specific events, typically occuring near puberty, where they were either given special attention by their fathers or given an "adult" privilege for the first time. Not one AM mentioned such an event. Competent mothers generally evidenced less symbiotic relating and greater productivity than did their AM counterparts, including being involved in more activities, fewer games and pastimes, and less withdrawal. V The results of this study suggest that a combination of self-report TA variables, selected demographic variables, and selected Erikson scales would make up a quick, meaningful, and very economical instrument package for discriminating abusive from non-abusive mothers. Specifically, the variables Activity, Withdrawal, Pastimes, Games, Paternal Allowers and Paternal Drivers from TA; E52 (Autonomy vs. Shame and Doubt) and E57 (Generativity vs. Stagnation) from the E-Scale, and Education, Telephone, Number of Children, and Number of Significant Adults 34 in childhood appear to be the best discriminators from the screening analyses. Discounts, a rated TA variable, would be fairly easy to learn to rate from ordinary interview material. Discounts would be a valuable addition to the total package, as it is a prominent variable in the TA correlation matrix and appears as the single most powerful variable in the final discriminant function analysis. The raters identified eight women, two AMs and six CMs, who appeared to have chosen ppt_to comply with their parents' scripting messages. Both AMs were still rated as choosing a tragic life script, whereas all six CMs were rated as choosing banal scripts. CMs also reported interacting during their child- hood with an average of 2.5 significant adults besides their parents, while these two AMs reported none. These observations seem pertinent to the work of Anthony (1975) and Garmezy (1976) on "invulnerables" or "superkids" and suggest that perhaps the child with sufficient alternative parenting ppp make autonomous, construc- tive decisions in childhood, even in the face of heavy destructive life scripting by parents. Demographic variables were the second most efficient predictors. Lack of telephone suggests social isolation and control issues around help-seeking, an issue mentioned frequently in the child abuse literature (Davoren, 1974; Helfer, 1977; Steele, 1974). The AMs' higher number of children may also reflect a passive and inadequate stance toward controlling their own reproductivity and/or 35 sexual behavior. Educational level, lower for AMs, may be related to premature parenthood, and to general devaluation of a rational approach toward problem-solving. The Eriksonian findings were expected in the sense that Autonomy's role was congruent with TA theory, while the presence of Generativity was predictable directly from Eriksonian theory (Erickson, 1950). From the psychodynamic viewpoint, symbiosis is the opposite of autonomy and its manifestation is passivity. The MCQuitty cluster analysis revealed Discounts as strongly inversely linked to Autonomy in Cluster 1, and Discounts are theoretically one of the two general manifestations of passivity (the other is Passive Behavior). These findings indicate an empirical bridge between the Eriksonian and TA theories, and suggest that scripting occurs during the resolution of the Autonomy stage (1% to 3 years). Anecdotally, CMs frequently cited the patience of their parents, who seemed able to accept mistakes as an inevitable accompaniment of their immaturity and learning, and had rewarded their cooperation. In sharp contrast, the AMs were more likely to give examples of their parents trying to control them through coercion or "shaming." The Eriksonian findings in this study differ substantially from those of Evans (1976)* who reported that Trust (E51), Identity (E55), and Intimacy (E56) were the best predictors, using the same *The present E-Scale scores are a linear transformation of those which would have been obtained using Evans' (1976) scoring ES + 40 - E 0 rules [ESJ - Number of Scorable Items]. These difference do not affect the statistical analyses in any way. 36 instrument with somewhat similar samples. On closer examination, two differences between these central Michigan samples may have been important: (a) mothering ability, and (b) dependency (passivity). The present CMs were selected from a low income population, but primarily for their excellence in parenting. Evans' control mothers were randomly selected cases, presumably free from abuse. All were recipients of Aid to Dependent Children (ADC), while the present CMs were about equally divided between women who were low-income but never on ADC, "graduated" ADC clients, and women temporarily on ADC while receiving additional training. Thus, divergent samples may have importantly contributed to these discre- pant findings. The MSPP produced two variables, Emotional Needs Met and Expectations of Children, which discriminated moderately well between groups. While this package yielded weaker discrimination than did the TA, Eriksonian, and demographic approaches, the results seem consistent with the work of Helfer, et a1. (1977). These two MSPP variables may be peripherally related to symbiosis, in that failure to have needs met in childhood was related to AMs' subsequent inability to tolerate much dependency in their children. The SRE's failure is sharply contrary to the report of significant differences between abusive and non-abusive groups by Justice and Justice (1976). Their abusive mothers were experiencing moderate life crisis (Y'= 234), while their non-abusive counterparts were only in mild life crisis (Y'= 124; t_= 4.28, p_< .001). In 37 this study, the differences were not significant, and both groups were considered to be in severe life crisis (7A = 411; Yk = 347; 3.: .83, p_< .10). The Justices' groups (N's = 35) were matched quite carefully on age, income, size and age of family, and education. Since the current groups were similary equated, the different findings may well be accounted for by divergent sampling. Contributions to Personality Theory The pp§t_ppp_analysis of best predictors identified three clusters of variables. The first, characterized by "Autonomy (E52) vs. Discounts," appears to confirm the notion that the autonomy stage (1% to 3 years) is an important one for development of a healthy, capable parent. It also supports the idea that the autonomy-symbiosis dimension is the bridge between Eriksonian and TA theories. The second "social isolation" cluster indicates that the more children one has, the greater the degree of one's social contact generally, and the less privacy is available. Children thus may represent a socially-involving force for most parents. The third cluster, involving only Original Scrip Level and ENM, supports previous observations on the importance of the nurturance of the early environment in future parent role performance. As a whole, these findings lend strong support to the original hypothesis that early upbringing strongly influences adult parenting, including the occurrence of child abuse. The mechanisms of that influence appear consistent with the notion of "life-scripts," the limiting and/or giving up of autonomy which starts in early childhood. 38 The pp;t,ppg_examination of the TA variables also revealed that the women tended to see their parents as responding similarly to their children on the Passivity and Allower dimensions. The similarity on Passivity supports the Justices' (1976) finding that couples tended to have interlocking scripts that perpetuated passivity and symbiosis in each partner, and contributed to the maintenance of abuse in the family. Parental Driver behavior was experienced differently from mothers and fathers. These women may have seen their mothers' controlling demands as part of the manipulation to meet mother's needs, whereas fathers' demands had a strong inhibiting effect on the women's later ability to be productive people. The former finding seems consistent with Friday's (1977) observations that the main symbiotic tie for adult women is with their mothers. The latter finding appears consistent with TA theory in that if a person spends time trying to please another, she will spend less time in autonomous pursuits such as productivity and intimacy. These were in fact the relationships observed among the time-structure variables in this study. Together, these findings lend some support for TA theory's postula- tion that girls get their instructions on what to strive for from their fathers, but turn to their mothers for information about how to do it. The relationships within the clinician-rated variables generally fell into patterns consistent with TA theory. Specifically, Injunctions were highly inversely correlated with Original Script 39 Level, and have been proposed as the primary source of parental influence on a person's life-script (Berne, 1972; Steiner, 1974). The negative correlation between Discounts and Current Script Level is also important theoretically, as Discounts are one of the two passivity mechanisms through which a person maintains destructive scripted positions and roles. It is difficult to know how much bias was introducted into these correlations by the fact that the ratings on the variables were not independent, i.e., each rated by a different rater. Furthermore, the raters were highly committed to the TA model. Therefore, the opportunity to examine cross—source relationships is even more important than the separate findings. The raters and subjects showed most agreement on the Passivity measures, less on Allowers, and none on Drivers. It is particularly gratifying to find agreement on the Passivity measures, since passivity plays such an important role theoretically as an apparent precursor of child abuse and poor parenting. The total lack of relationship between Driver variables may be due to the difficulty that was encountered in translating the TA variables into specific behavior which might be understood by the women on the self-report checklist. Most other items (e.g., Allowers, Passive Behaviors, etc.) were more readily describable and were often presented in the same language to both the subjects and the TA raters. The patterns of correlations of the three most prominent variables in the cross-source matrix appeared to be consistent 40 with TA theory. Two other observations from these data are of special interest. The first is that women who appeared heavily "counterscripted" by their fathers (Drivers) tended to operate out of "Little Professor" and to relate to others primiarly through manipulation (Games). Second, Permissions given in childhood, appeared positively linked with self-confidence and productivity in adult women. Thus, the counterscript element Drivers may be as important to script outcome as the more traditional script element, Injunctions. It also appears useful to look for such positive elements as Permissions and Allowers in assessing the extent of constructive maternal scripting. Implications for Future Research One logical extension of the present study would be to administer the best predictor instruments to several hundred randomly selected low income women. After a reasonably long period, say two years, each participant would be followed up with the referring agency staff and the State Child Abuse Registry to find out which, if any, of the women had been found to have physically abused their children in the interim period. The predictions from the measures would be compared with the actual group membership at follow-up. This would be a complex and difficult study to execute, and it would be necessary to keep rather close contact with the referring agencies throughout. One way of doing that might be to request interim observational data from the caseworkers quarterly. 41 Such a strategy might also allow for better follow-up accessibility to the participants than is usually possible. It would be expected that there would be shrinkage in the accuracy of group membership prediction over such an extended period, with a new sample, and with a true predictive model. However, the ability to assess intervening variables might give valuable clues to effective prevention strategies. Implications for Treatment and Prevention Although the present study's correlational character does not permit causal statements about the relationships among the variables examined, it seems important to speak to the implicit issues of treatment and prevention. To counteract destructive early scripting in treatment, the therapist needs to strongly' encourage the abusive mother to think and feel, to express her needs and intentions clearly, to observe others carefully and to trust her own senses, and to develop her ability to think through interpersonal problems and find solutions. The therapist needs to set firm limits on destructive maternal behaviors, but to allow the mother a considerable degree of latitude in meeting those limits (Erikson, 1950). According to the findings of this study, she may be embroiled in an Autonomy (E52) conflict and be very fearful of external control. She will also probably be experiencing severe self-doubt and shame reactions accruing through both real and imagined inadequacies. Since curiosity appears to be an opposite of confusion, which is a common consequence of the shame 42 response (Lewis, 1971) the therapist can encourage curiosity. The mother must get permission from the therapist to attend to her own legitimate needs, but she must also learn that healthy relationships require a balance between attending to needs of the self and the needs of others. The interview developed for this study appeared potent in helping these women focus on the origins of their ideas about mothering in an unthreatening way. Several spontaneously resolved during this interview to find out more from their own mothers, and several expressed relief after discussing their own experiences and feelings. The sense of continuity of parenting behavior between generations seemed to help them get perspective and some empathy for all three generations. The importance of a wide range of services, including family planning, options to upgrade education, and the provision of a tele- phone, to the possible prevention of child abuse are suggested by the present findings. However, these findings need verification in subsequent studies. The availability of a telephone seems sufficiently important and inexpensive that granting one (perhaps without a long distance option) to every welfare recipient with small children could be a truly significant prevention strategy. Preventatively, the abused or "at risk" child needs adequate support for a positive resolution of the Autonomy stage. Parent training for parents of toddlers and preschoolers focused on issues of autonomy and control and creative problem-solving might help parents to manage their own power and control conflicts 43 during this period. Preparatory guidance in teaching methods could help the parents to be clear in their directives and supportive of their children's mistakes and efforts to please. Another form of support for the child appears to be the provision of a home visitor. This person, herself a good mother who had good parenting, acts as a friend of the family who visits regularly, discusses problems sympathetically with the parent, and acts as a positive parenting model for the child (Kempe, 1976). Such a visitor may introduce the child and family to a perspective of reason and sanity into the family which is unique in their experience. As the findings of this study suggest, the presence of other significant adults can be tremendously important in off- setting negative effects of the parents. In projects to date using such persons, reported incidents of abuse have been dramatically reduced or eliminated (Barry, undated; Gray, Cutler, Dean 8 Kempe, 1977; Hallock, 1976). 44 .mmmu men» go» mumcwsgmamoco rr o~.mm .. .. r. .. .. .. amazaapm maucaaaaaxm ou.~o one. m mn.~ mm. mm. oo. uumgmo acmomm mo mpzumsum .o ocwpcmgwo No.mm ooo. m om.~ Po. ow. em. mmwznmum mo opweoco no.- moo. o oo.u mu. we. AN. pumcwo ocwcmmcum :aowsowz .m om.m~ moo. N mo.~p . oo. om. ow. mmwzomam ucmsoo~m>oo eo mo.Fm ooo. A mo.ep No. Po. oo. pumcwo mmoaum m.:omxvgm .q on.mo ooo. m oo.m~ mm. mo. mo. mmwzampm o¢.oo Foo. o mm.P~ Pm. on. om. pumcwo owgoacooEmo .m on.mo moo. o NF.oN mm. oo. oo. mmwzomum mmpnapsa> o~.oo ooo. NP Po.op om. Fm. mo. uumgwo <5 topaz .F cowuauweommapo m. on mcazcm moose; :owumpmgcoo m=Fa> venom: pom muao aumccoo ego .mxpwz Pau_coccmo -cwowm co_pucou ocmucmo peacpswcumwo .moooco pomumosoo new m>wm=n< cw owcmcmnsmz ocwpuwomco toe aumo we mumm xwm eo mmmapac< cowuucso acaewswgumwo mo zgmssam--.~ m4o zuv>wuacmcmo .N «N. _e. mo. om.- cotaa_omo .m> suaeeacc .o um. mm. NP. n~.- :owmzeewo upom .m> zupucmoo .m Fo. om. mp. mN.- apwcowamecH .m> xcumzoco .e om. me. NF. m~.- upwso .m> m>wuawuwco .m cm. oe. oo. nm.- macaw .m> xsocop=< .N Pm. FA. mm. _P.- pmaapmaz .m> aware .P cowumpmccoo Empoucmpeo .mmpaum FaucmEQopm>mo comchu to» Aesop666 63.2: 66¢ 216666 66.6666 66.6.6 66. uc—Louno 66.66.66» .7...» .6 66:. .... 25:8 6.63.66 «concoaon co museum .muuonaam acouoalou v:- osvmanww¢0 ¢>uw¢0 020000 200000 09:50.80 .—.~ 305 50g 108 92:0 99.9 09.. ——.~ mo.— oo.~ 99.9 n... —..9 09.9 .9.9 .9999 99.» on. no.~ mm.. 99.n n~.n m~.9 09.9 nu.n n~.0 999: 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.. 99.. 99.9 99 9..» 9.9 99.9 99.9 99.. 99.9 n... 99.9 99.. 99.. on 99.9 99.9 9.9 99.9 9.9 99.. 99.9 99.. 99.9 99.. on 99.. 99.9 99.9 99.9 99.. 99.. 99.9 99.9 99.~ 99.. .9 99.9 99.. 99.9 99.9 99.9 99.9 99.9 99.. 99.9 “9.. an 99.9 99.9 99.. 99.9 99.~ 99.9 99.. 99.9 99.9 99.9 «n a. 99.9 9.9 99.9 99.9 99.. 9~.~ 99.9 99.9 99.. 99.. .9 99.9 99.. ...9 99.. 99.9 99.. 99.9 99.9 99.9 99.. 99 .m 99.» 9.9 99.. 99.9 99.9 99.... 99.... 99.9 99.. 99.. 9... m 99.9 9.9 99.. 99.. 99.~ 99.~ 99.9 99.~ 99.. 99.. 99 .. 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 99.9 - 9. 99.. 99.9 99.9 99.. 99.. 9..~ 99.9 99.. 99.9 99.9 99 m 99.9 9.9 9.9 99.9 99.9 99.9 99.9 99.9 99.9 99.. 3 m 99.9 9.9 99.9 99.. 99.9 99.. 99.. 99.. 99.9 9... .w 99.9 9.9 99.9 . 9.9 99.. 9~.. 99.. 99.9 99.. 99.. aw 99.. 99.9 99.9 99.. 99.~ 99.9 99.. 99.9 99.9 99.9 - 99.. 99.9 9... 99.9 9..~ 9..~ 99.9 99.9 99.~ 99.. .9 99.9 99.. ...~ 99.. 99.9 99.. 99.9 99.9 99.9 9..» 9~ 99.9 9.9 9.9 99.. 99.9 99.~ 99.~ 99.9 99.. 99.. 9. 9..» 99.9 99.. 9..» 9..~ 99.9 99.9 99.9 99.9 99.9 9. 99.9 99.. 9~.p 99.~ ~_.~ n9.m no.9 ~n.n m—.m o~.m :99: 99.. 99.9 9.9 99.. 99.9 99.9 99.~ 99.9 99.. 99.9 9. 99.9 99.9 99.. 9.9 99.9 99.9 99.9 99.. 99.9 99.. o. 9... 9.9 99.9 9... .9... 99.9 99.. 99.. 99.9 .... m. 99.9 99.. 99.. 99.. 99.9 99.9 99.9 99.. 99.9 99.9 c. 99.9 99.. 99.. 99.9 99.9 99.9 99.9 99.9 99.9 99.9 9. 99.9 9.9 9... 99.9 9~.~ 9..» 99.9 99.9 99.9 99.. ~. 99.. 9.9 9.9 99.9 on... 99.9 99.9 99.. 99.9 99.9 .. M 9..9 9... 99.9 99.9 9... 9~.~ 99.9 99.~ 99.9 99.. 9. n. 99.9 9.9 99.. 99.. 99.. 99.. 99.9 99.. 99.9 99.9 9 M 99.9 9.9 9.9 9..~ 9..~ 9~.n 99.. 99.. 99.~ 99.. 9 9. 9... 9.9 9.9 99.9 9.9 99.9 99.9 99.9 99.9 99.9 9 m 99.. 9... 9.9 99.9 99.~ 99.9 99.. 99.. 99.. 99.9 9 _m 99.. 99.9 99.9 ou.~ 99.9 9..n 99.~ 99.9 99.9 99.. n 99.9 9.9 99.. 99.. 99.9 99.9 99.. 99.. 99.9 99.9 9 99.9 99.9 99.. 99.9 9.9 9m.. 99.. 99.9 99.. 99.. 9 9... 9... 9... 99.9 9... 9... 99.~ 99.9 99.9 99.. ~ 9... 99.9 9..» 9... 9..~ 9..~ 99.9 99.9 99.. 99.9 . >99.9z. 999.9 >99..u¢ us..m 18 years Never Married Widowed No No Physical Both $10-15,000 $15,20,000 > $20,000 No 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 113 Educational Level 1.00 - 8th 2.00 - 8th 3.00 - 10th Close Friend/Relative Death 1.00 - Yes Presence of Telephone 1.00 - None or Unlisted Had Premature Baby 1.00 - Yes Had C-Section 1.00 - Yes Own Birth Difficult 1.00 - Yes Breast Fed as Infant 1.00 - Yes Breast Fed Own Children 1.00 - Yes Number of Children Expected (total) Number of Siblings (total) Position in Birth Order 1.00 - Oldest 2.00 - Youngest Who Raised Subject 1.00 - Natural parents Number Mother's Siblings (total) Number Father's Siblings (total) 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 12th Some College College Completed No Yes No No No No No Middle Only Other 25. 26. 27. 28. 29. 114 Other Important Adults in Childhood (total) Other Important Females in Childhood (total) Parental Drug and Alcohol Use 1.00 - Yes 2.00 - No Own current Alcohol Use 1.00 - Never 4.00 - Several times a week 2.00 - Once a month 5.00 - Every day 3.00 - Once a week Own Current Drug Use Same as Number 28, above. (VARIABLES 30-69 ARE CONTINUOUS VARIABLES.) 30. 31. 32. Schedule of Recent Experiences - 6 months - 1 year - 2 years Erikson Scale 1, Trust vs. Mistrust 2, Autonomy vs. Shame 3, Initiative vs. Guilt 4, Industry vs. Inferiority 5, Identity vs. Role Diffusion 6, Intimacy vs. Isolation 7, Generativity vs. Stagnation Michigan Screening Profile of Parenting: Emotional Needs Met Relationships with Parents Expectations of Children Coping. TA Self-Report: Withdrawal Rituals Pastimes Activities Games Intimacy TA Rated: Racket l, Distress Racket 2, Fear Racket 3, Anger Racket 4, Shame 71. 72. 115 TA Self Report: Maternal Allowers TA Rated: Paternal Allowers Maternal Drivers Paternal Drivers Maternal Passivity Paternal Passivity Injunctions Permissions Drivers Allowers "Gallows" Transactions Passive Behaviors Discounts Payoffs Original Script Level Current Script Level Age of Compliance 1.00 - 1-3 years 4.00 13-17 years 2.00 - 4-6 years 5.00 - 18-30 years 3.00 - 7-12 years 6.00 - 30 years "Adult" Influence 1.00 - Yes 0.00 - No "Little Professor" Influence 1.00 - Yes 0.00 - No APPENDIX 0 MEASURES OF ASSOCIATION BETWEEN DEMOGRAPHIC VARIABLES AND MEMBERSHIP IN GEOGRAPHIC GROUPS AND ABUSIVE/COMPETENT GROUPS 116 APPENDIX TABLE J.l.--Measures of Association between Demographic Variables and Membership in Geographic Groups and Abusive/Competent Groups. 117 Urban vs. Rural > Abusive vs. Competent Demographic Variable I? E 2 2 I .E :2 E B 2 I: .2 Age 21.57 .31 .16 .35 15.53 .69 .15 .38 Income 4.20 .52 .23 .17 5.07 .41 .10 .58 Marital Status 1.91 .75 4.91 .30 Age Youngest Child 3.57 .61 -.12 .48 .89 .97 .06 .74 Age Oldest Child 2.79 .84 .22 .19 6.24 .40 -.08 .62 Number of Children 12.96 .02* .15 .38 7.78 .17 -.33 .05* Education 2.75 .60 -.03 .88 9.49 .05* .49 .00* Birth Order 1.47 .69 2.76 .43 Presence of Telephone .53 .47 7.67 .01* Alcohol Use 3.57 .31 -.25 .13 2.15 .54 .22 .19 Drug Use 1.89 .39 .03 .84 1.77 .41 .21 .21 Who Raised Subject .08 .78 1.10 .29 Childhood Punishment 2.26 .52 -.19 .26 3.83 .28 .01 .97 Own Birth Difficult .02 .89 .44 .51 Separated from Mother .10 .75 .10 .75 Childhood Pets .04 .85 .04 .85 Serious Illness .07 .79 .07 .79 Relative Died .01 .93 .01 .93 Child by Caesarian .01 .93 .01 .93 Child Premature .44 .51 2.43 .12 Breast Fed by Mother .01 .95 .12 .73 Breast Fed Own Kids .04 .84 .21 .65 Parent Drug/Alcohol Use .04 .85 .04 .85 Number of Children Expected 6.28 .39 .06 .73 5.61 .47 -.33 .06 Number of Siblings 4.92 .84 .12 .80 1.70 .23 -.22 .19 Maternal Sibship 12.17 .20 -.15 .42 5.20 .82 .02 .91 Paternal Sibship 10.55 .31 .11 .57 2.34 .17 :11 .55 No. Significant Adults .69 .22 -.07 .67 8.80 .07 .41 .01* No. Signif. Female Adults .09 .99 -.02 .91 5.29 .15 .35 .03* .Significant at p: .05 APPENDIX K SCREENING OUTCOMES FOR SIX VARIABLE SETS BY DIRECT AND STEP-WISE DISCRIMINANT FUNCTION ANALYSIS 118 119 cowpmuzum .m $58 ooo. m 8.- mm. a. 8. =22...5 .3 aces: .N % mcosngm» .P mupzu< upmsmm acme -Pc_=m.m co Lungs: m»_=u< u:mu.c.=m.m ....... Rae.om _oo. m om.p~ mm. 0“. ea. cc 259232 Sauces cacup_;u co guess: mcognmpmh cowumuzum .. .... ...“... 4% ”Wrasse“. E ...... .3395 23358 can 35:3 5 3:23sz mcwpuwumca com mmpam_cm> uwcamcmosmo umuumpmm mo mwmapmc< cowuuczu ucmcwewcumwo--.F.x m4m=mmvm umuzpucH mmpnmrcm> waspmz acmucwa vsu .mxp_3 pmuwcocmo .mazocw ucmgmaeou ucm m>wman< c? awsmcwasmz mcwuupvmca cow Ammm "mmucmwgoaxm acoumm yo mpzumgumv mmgzmmmz mmmcum mew; mmcgh do mpmapmc< cowuuczu acmcwewcumwo--.m.x m4mm4 Row.pm mocozpccH =SP=u<= .e mucaoumwo .m wmwzmmpm pn_aum Pacwaeao .N mucm:_w:H .commmCOLa apoges. ._ macaw gmmc< cmmu mmmcumwo ”umxumm qusou op :opmwumo pm>m4 pawcum newcczu Fm>m4 pawgum szmmwco mucmzpmcu :Lommw$OLm mPHHwA: emp. NP m¢.mm Fe. nu. m¢.— mucus—ucH =ppzu<= aumcwa mcowuummcmcp mzoFme pm>m4 mmoxmm mucaoumwo mcow>mzmm m>wmmwa mcmzop_< mcm>wca meowmm_sgma meowuuczncfi cowpmu_cwmmm_u yumccou ucmucmq .m .1. mcmzcm munEoA cowumpmccou mapm> cu ego .mxpwz _au_=o=mu -cmmwm umuapucfi mmFaa_am> venom: .mazoco ucmpmasou use m>wm=n< cw apgmcmnswz m=_puenmga 20c ma_amwam> pgeaum wco chgmumm .m mmsmw .m mmspumma .e &wm.mm moo. m n~.om mm. mm. mm. mgmzopp< pwcgwuwm .m N F szaaucp_z . mm.u_>Ppu< . xums_u=H magma mm_uw>wuu< mmswpmma mpmappm szmcucuwz Nm~.mw moo. NP Fm.m_ om. FR. mm. xuw>wmmm¢ Pacemamq uumcwa xuw>wmmma chgmumz mgm>wgo pmccmuma mcm>PLo Puccmumz mcmzopp< pacemama mcmzop~< chcmpmz cowomuwcwmmm_u poocgou ucmocma .. .I. mcmsam muaqu :owumpmggoo wzpm> ecu .mxpwz Pa0w=ocmu -cmm_u umu=_u=H mapnmwam> cocoa: Q. u. 'U .mazoco pcmumasou can m>mm=n< cw nwsmcmnsmz mcwpuwumca cow mmpamwcm> wumcmcmc .p scrumcmmum-»uw>wumcmcmw cowumpomH-aUms_u:H cowmaccvo apex-»gwp=mua xmo.~m wmo. N mm.¢P No.. Po. om. xuwcopgwecfi-zgumaucH mmwmwm “Frau-m>wumpuch msmsm-»socop=< umacumpz-pm=ch cowumu_$_mmmpu .. ... mcmaam cease; coppm . Faggou wzpm> wwwwumw a mu Psu .mxppz Pauwcoccmu -cmmpm umuapucH mwpnmvcm> uocumz .mazoco ucmgmqaou ucm m>wmaa< cw apnmcmnsmz mcwuurumca soc Ammv mmpmum Faucmsaopm>mo :mpcomxpcm mo mpmzpm=< cowuuzzm ucmcpswgumpo--.m.x m4m umczpuca mmpnmvcm> vogue: ucmucma mco .mxpwz quwcoccmu -cmmmm .quocu pcmumaaou ucm m>wm=n< cw nwzmcmnsmz mcmuuwumca com mucoum mpmum Aammzv m:_u=mcma com mp_$ogg mcwcmmcum cmmwgopz we mwmzpmc< cappuczu ucmcwewcumwo--.o.x m4m’uu< .uucueuw "umu o~. we. . mp.- mo. so. oo. cc. po. ~—. oc._ m~.- ocosau—o» Pm. sn.- mo.- up. eo.- p..- o..- em.- mp.- m~.- oo.p pa>u4 un—cum poceopco opu Low x*suux cowuupossou unsouw :.:u¢3 topooa--.—.A man