ovcnouz FINES: 25¢ perm per item RETURNING LIBRARY MATERIALS. J {l --\\\“\‘ L:' .\ gwu” Place in book return to remove I," charge from circulation records A COMPARISON OF THE CHILDBEARING AND INITIAL CHILDREARING EXPERIENCES OF TEENAGE AND OLDER MOTHERS BY Mark W. Roosa A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family Ecology 1980 ABSTRACT A COMPARISON OF THE CHILDBEARING AND INITIAL CHILDREARING EXPERIENCES OF TEENAGE AND OLDER MOTHERS BY Mark W. Roosa The purpose of the study was to determine the pos- sible causal mechanisms of the developmental deficits that longitudinal studies have attributed to the children of teenage mothers. Sixty-two primiparous mothers from 15—32 years old and their non-twin infants were the subjects of the study. The mothers were contacted and interviewed dur- ing the last trimester of their pregnancies and the mothers and infants were studied until the infants were three months old. Demographic data were gathered in home interviews, medical data for the birth were recorded, and the Brazelton Neonatal Behavioral Assessment Scales were administered shortly after birth. At one, two, and three months post- partum the mother and child were systematically observed while interacting in their home, the home environment was assessed, and the mothers completed scales rating maternal attitudes and infant temperament. The results indicated that teenage mothers (n = 14) were more similar to older mothers than generally has been Mark W. Roosa reported. However, the teenage mothers had significantly lower SES levels, began prenatal care later in pregnancy, and lived in more crowded homes than the older mothers. The teenage mothers also spent significantly less time talk- ing to their infants, less time talking to their infants during mutual gazing, and were significantly less respon- sive to their infants' distress signals than the older mothers. On the other hand, the teenage mothers had easier births than the older mothers. Only one of seven a priori Brazelton scales differentiated between the two groups of infants, with the infants of the older mothers scoring higher. Systemic analysis of the data, using LISREL, a maxi- mum likelihood method of estimating linear structural rela- tions, indicated that SES was the most influential variable relating various predictors of a child's developmental status. Maternal age, with SES controlled for, was nega- tively related to the same predictors; i.e., given equal SES, young mothers of their infants did better than the older mothers or their infants on each of the predictors of developmental status. The results of the systemic analysis need to be treated cautiously since the model is not sufficiently precise in its present state of develop- ment, at least for the small sample that was studied. Refinement of the research model and its application to other data sets (especially longitudinal data sets) were recommendations for future research. To Mr. & Mrs. Fred Rose and Mary Kaufman ii ACKNOWLEDGMENTS In any undertaking as complex as the present study, any attempt to thank all those who contributed to the com- pletion of the task is doomed to failure. However, I must thank the William T. Grant Foundation who provided the funds that made the project possible; the manufacturer who donated diapers that were useful incentives for partici- pation in the study; the numerous physicians, agencies, and school officials who distributed study brochures to potential participants; Dr. Stack, Dr. Gruca, Bette Chap- man, and the staff of Gratiot Community Hospital for all their help and cooperation; and to the most important people of all, the participants who gave up part of their privacy and large blocks of time to make the study possible. I also must thank Kris Busk, Gale Grant, Olga Hernandez, Betty Abedor, Sally Boerner, and Martha Brewster, the research assistants who worked so hard gathering and recording data, and to Cheryl Spiekout and Lisa Rath, the data coders who did a remarkable job handling the immense amount of data collected. Colleen Esch, Diane Osburn, Jean Grof, Mary Grace and other members of the staff of the iii Institute for Family and Child Study were extremely helpful throughout the study. Each member of my doctoral committee, Dr. Beatrice Paolucci, Dr. Hiram Fitzgerald, Dr. Nancy Carlson, Dr. Edward Fink, and Dr. Margaret Bubolz, has played a role in the development of the final product of the study. However, special thanks go to Dr. Fitzgerald and Dr. Carlson, who have been extremely helpful and supportive throughout the study, from the shaping of the initial research idea to the preparation of the final report. iv TABLE OF CONTENTS Page LIST OF TABLES O O O O O O O O O O C O O Viii LIST OF FIGURES. . . . . . . . . . . . . x Chapter I. INTRODUCTION . . . . . . . . . . . 1 Statement of the Problem. . . . . . . 1 Objective of the Study . . . . . . . 7 II. REVIEW OF THE LITERATURE . . . . . . . 9 The Continuum of Reproductive Risk . . . 10 'Experiences of Pregnancy and Delivery. . 10 DeveloPmental Status at Birth . . . . 16 Summary. . . . . . . . . . . . 25 The Continuum of Caretaking Casualty. . . 28 rTeenage Parenting . . . . . 28 Maternal Attitudes and Child DeveloPment. 33 Home Environment and Child Development . 34 Socioeconomic Status and Child DeveloPment. . . . . . 37 Parenting Style and Child Development. . 39 Mother-Infant Interaction. . . . . . 46 Mother-Infant Attachment . . . . . . 49 smach O O O O O O O I O O O 59 A Systems Approach to Child Development. . 61 Model for Examining the Effects of Maternal Age on Child Development . . 64 Summary . . . . . . . . . . . . 68 Chapter Page III. DESIGN 0 O O O O O O O O O 0 O O 70 Subject Recruitment and Selection . . . 70 Informed Consent Procedures . . . . . 76 Research Staff. . . . . . . . . . 78 Method . . . . . . . . . . . . 79 Initial Interview . . . . . . . . 79 Obstetric Data . . . . . . 80 Neonatal Behavioral Assessment. . . . 80 Home Observations . . . . . . . . 81 Maternal Attitudes. . . . . . . . 82 Home Environment . . . . . . . . 82 Infant Temperament. . . . . . . . 82 Developmental Assessment. . . . . . 83 Instruments . . . . . . . . . . 83 Data Coding . . . . . . . . . . 99 Analysis . . . . . . . . . . . 100 IV. RESULTS AND DISCUSSION. . . . . . . . lll Characteristics of the Two Groups of Mothers and Infants . . . . . . . 111 Socioeconomic Status . . . . . . . 111 Other Demoqraphic Characteristics. . . 116 Pregnancy and Delivery Experiences . . 118 Neonatal Status. . . . . . . 121 Maternal Attitudes and Infant Temperament . . . . . . . . . 123 Home Environment . . . . . . . . 126 The Caretaking Environment . . . . . 129 Discussion of the Descriptive Analysis. . 137 The Continuum of Reproductive Casualty . 137 The Continuum of Caretaking Casualty. . 138 Summary . . . . . . . . . . . 145 The Estimates of the Parameters of the Research Model . . . . . . . . . 146 The Measurement Models . . . . . . 146 The Theoretical Model. . . . . . . 157 Evaluation of the Research Model . . . 163 Discussion of the Estimated Model . . . 165 Evaluation of the Model . . 165 The Relationship of the Estimated Model to Theory . . . . . . . . . . 172 vi Chapter Summary . . . . . . . V. SUMMARY AND CONCLUSIONS . . Summary . . . . . . . Conclusions. . . . . . Limitations . . . . . Strengths. . . . . . Implications. . . . . REFERENCES 0 O O O O O O O O APPENDICES Appendix A. Subject Recruitment Letters . . Informed Consent Procedure Materials . Scoring and Instruction Manuals. B C. Instruments . . . . . . D E . Correlation Tables for the Research Model and Glossary of Terms . . F. LISREL Analysis of Research Model . vii Page 179 181 181 186 186 187 188 191 215 218 222 254 272 277 Table l. 10. 11. 12. 13. LIST OF TABLES Complications of Pregnancy and Delivery that are Reported to be More Common for Teenagers than for Women in Their Twenties, and Supporting Studies . . . . . . . . . Studies Reporting Differences Between the Gestational Ages and/or Birth Weights of Infants Born to Teenagers and Older Mothers. Characteristics of Teenage Mothers Reported in the Literature . . . . . . . . . . Age-Specific Live First Births for Clinton, Eaton, Gratiot, and Ingham Counties, Michigan Residents, 1978 . . . . . . . Response Description . . . . . . . . . Age Distribution of Subjects . . . . . . Interrater Reliabilities for Observation Itenls I O O I O O O O O I O O 0 Definition of Symbols Used in Figure 3 . . . Structural Equations for the Theoretical Model in Figure 3 . . . . . . . . . . . Structural Equations for the Measurement Models in Figure 3 . . . . . . . . . Socioeconomic Characteristics of Study Families by Maternal Age . . . . . . . Descriptive Characteristics of Subjects by Maternal Age . . . . . . . . . . . Characteristics of Pregnancy and Delivery by Maternal Age . . . . . . . . . . . viii Page 11 21 29 72 73 75 97 104 107 108 113 114 119 Table V Page 14. Characteristics of Newborns by Maternal Age . . 122 15. Brazelton Summary Scores of Newborns by Maternal Age. . . . . . . . . . . . 124 16. Attitudes Toward Child by Maternal Age As Measured By the Know Your Baby Scale. . . . 125 17. Infant Temperament Characteristics by Maternal Age--Resu1ts of the Michigan Infant Tempera- ment Scale . . . . . . . . . . . . 127 18. Characteristics of the Home Environment at Three Months by Maternal Age . . . . . . 128 19. Characteristics of the Caretaking Environment at Three Months by Maternal Age . . . . . 130 20. Characteristics of Mother-Infant Interaction by Maternal Age: Maternal Behaviors as a Preportion of Infant's Awake Time. . . . . 132 21. Contingent Re5ponsivity by Maternal Age . . . 134 22. Characteristics of Mother-Infant Interaction by Maternal Age--Infant Behaviors as a Ratio of the Infant's Awake Time . . . . . . . 135 23. Selected Correlations Between Maternal Variables and Variables Related to Child Deve10pment. . 142 24. Standard Errors for the Estimates of the Parameters of the Measurement Models. . . . 156 25. Standard Errors for the Estimates of the Parameters of the Theoretical Models. . . . 164 ix LIST OF FIGURES Figure Page 1. Model for Examining the Effect of Maternal Age upon Child DevelOpment . . . . . . . . 65 2. Research Protocol for Comparing Teenage Mothers and Their Infants with Older Mothers and Their Infants . . . . . . . . . . . 84 3. Research Model for Examining the Effect of Maternal Age Upon Child Development. . . . 103 4. LISREL Estimates for the Measurement Models for the Socioeconomic Status Variable £2 and the Prenatal Care Variable n1 (Reliabilities in Parentheses) . . . . . . . . . . . 147 5. LISREL Estimates for the Measurement Models for the Neonatal Status Variable n2 (Reliabilities in Parentheses) . . . . . . . . . . 148 6. LISREL Estimates for the Measurement Model for the Neonatal Status Variable n3 (Reliabilities in Parentheses) . . . . . . . . . . 149 7. LISREL Estimates for the Measurement Model for the Maternal Attitude Variable flu (Reliabili- ties in Parentheses). . . . . . . . . 150 8. LISREL Estimate for the Measurement Model for Infant Temperament Variable n5 (Reliabilities in Parentheses) . . . . . . . . . . 151 9. LISREL Estimate for the Measurement Model for Home Environment Variable n5 (Reliabilities in Parentheses) . . . . . . . . . . 152 10. LISREL Estimate for the Measurement Model for the Mother-Infant Interaction Variable n7 (Reliabilities in Parentheses) . . . . . 153 Figure Page 11. LISREL Estimates for the Theoretical Model for Comparing the Childbearing and Childrearing Experiences of Teenage and Older Mothers . . 158 12. Standardized Solution--LISREL Estimates for the Theoretical Model for Comparing the Childbearing and Childrearing Experiences of Teenage and Older Mothers . . . . . . 159 xi CHAPTER I INTRODUCTION Statement of the Problem Each year in the United States, approximately one million women 15-19 years old, or about one-tenth of the women in this age group, become pregnant. An additional 30,000 girls younger than 15 become pregnant annually. These teenage pregnancies result in about 600,000 live births each year and over 90% of young mothers keep their babies, a reversal of the situation little more than a decade ago (Alan Guttmacher Institute, 1976; Center for Disease Control, 1980; National Center for Health Statis- tics, 1977). The fertility rate for teenage women declined by almost 45% from 1960 to 1977. However, the fertility rate for older women dropped by about 50% during the same period. Since the fertility decline of teenagers did not equal that of older women, and since the overall size of the teenage population increased until 1977 due to the baby boom, teen- agers accounted for an increasing percentage of all births during most of the period from 1960 to the present. The percentage of all births accounted for by teenagers rose 1 from 14% in 1960 to a high of about 19% in 1974 and declined to about 17% in 1977 (Baldwin, 1976; Center for Disease Con- trol, 1980). Interest in teenage pregnancy, childbearing and Childrearing grew as the percentage of all women giving birth who are teenagers increased. Some of the interest in teenage pregnancy has centered upon its causes (Gottschalk, Titchener, Piker & Stewart, 1965; Klein, 1978; Meyerowitz & Malev, 1973; Moore & Caldwell, 1977; Russ-Eft, Spreenger & Beever, 1979; Vincent, Haney & Cochrane, 1969). A far greater number of studies have been focused upon the consequences of teenage pregnancy for the pregnant woman (and occasionally the male involved) in terms of education, relative economic status, family size and marital stability (Alan Guttmacher, 1976; Bacon, 1974; Burchinal, 1965; Butman & Kamm, 1965; Cannon-Bonventre & Kahn, 1979; Card & Wise, 1978; Coombs, Freedman, Friedman & Pratt, 1970; David, 1972; Freedman & Thornton, 1979; Freedman & Coombs, 1966; Fursten- berg, 1976, 1979; Moore, 1978; Moore, Hofferth, Caldwell & Waite, 1979; Moore, Hofferth & Wertheimer, 1979; Nye, 1976; Osofsky, 1968; Presser, 1980; Rolfe & Roosa, 1979; Roosa, 1977; Stickle & Ma, 1975; Trussel & Menken, 1978; Trussell, 1976; Waite & Moore, 1978). The medical aspects of adoles- cent pregnancy, including higher rates of maternal and infant mortality, prenatal and postnatal complications, and birth defects, also have received considerable attention (Alan Guttmacher Institute, 1976; Ballard & Gold, 1971; Coates, 1970; Dott & Fort, 1976; Duenhoelter, Jimieniz & Baumann, 1975; Dwyer, 1974; Hollingsworth, Moser, Carlson & Thompson, 1976; Hulka & Schaaf, 1964; Jones & Placek, 1979; McCarthy, Abbott & Terry, 1979; Mednick, Baker & Sutton- Smith, 1979; Mellor & Wright, 1975; Menken, 1972; Nye, 1976; Osofsky, 1968; Ryan & Schneider, 1978; Stickle & Ma, 1975; Walters, 1975). However, very few researchers have examined the impact of teenage childbearing upon the child. Several medical studies mentioned above did discuss the infant's physical status immediately after birth. Nevertheless, until recently, few studies have examined the long term consequences of teenage childbearing and Childrearing for the children involved. Two groups of researchers who did examine the long term social and psychological consequences for children of teenage parents found that by age 6 these children score lower on intelligence tests, are more likely to be dependent and distractable, to have behavioral problems, and to be deficient in reading-grade level than are children born to nonteenage parents (Oppel & Royston, 1971; Hardy, Welcher, Stanley & Dallas, 1978). In the first of these longitudinal studies, Oppel and Royston (1971) matched 86 mothers under 18 with mothers who were 18 or over at the delivery of their child. Matching was on the basis of socioeconomic status, birth weight of the child, maternal parity and maternal race. When the children were 6-8 years old and again when they were 8-10 years old, social and psycho- logical data were collected. The children were individually administered the Stanford-Binet and Weschler IQ tests and the Wide-Range Reading Achievement Test. Psychological ratings of emotional adjustment and personality traits were obtained as were data concerning mother-child relationships and family socioeconomic status. Oppel and Royston (1971) found that the children of the younger mothers: (a) were more likely to be reared by persons other than their bio- logical parents; (b) generally had more siblings and lived in larger households; (c) had a lower mean height; (d) scored lower on the Stanford-Binet IQ test but not on the Weschler; (e) scored lower in reading-grade level; and (f) were more often rated as dependent and distractable than children born to older mothers. The Johns Hopkins Child Development Study (Hardy £5 21., 1978; Hardy, 1971) had a more extensive and complex data collection protocol and a larger sample (4,557 mothers, 706 of whom were 17 years of age or less at the time of delivery) than the Oppel and Royston study. The children in this study were born from 1959 to 1965 to predominantly black (77%) lower-middle or lower socioeconomic families from the geographic region near Johns Hopkins University in Baltimore. All children were followed for 8 years; a representative sample of 466 children (77 with mothers 17 years of age or less) were followed for 12 years. At the 8 and 12 year follow-ups, 88% of the sample were examined. When children born to adolescents (17 years of age or less) were compared to children born to women 20-24 years of age, children born to adolescents: (a) scored lower on the Bayley mental test at 8 months; (b) scored lower on the Stanford-Binet IQ test at 4 years; and (c) scored lower on the Weschler Intelligence Scale for Children (verbal, per- formance and overall) and on the Wide-Range Achievement Test (spelling, reading, and arithmetic) at 7 years. The results at the 12-year examination were similar to those at 7 years (Hardy £5 31., 1978). Though Hardy gt El- did not take birth weight, parity, race, or socioeconomic status into account, the results of the study are quite similar to those reported by Oppel and Royston (1971). In neither study did the chil- dren born to teenagers do better than children born to older mothers on any of the developmental measures used. Vandenberg (1976, reported in Chilman, 1980) reports find- ing similar results in a retrospective study with a large, all white sample. Several correlational studies that link a child's IQ score with maternal age at delivery also sup- port the results reported above (Illsley, 1967a, 1967b; Broman, Nichols & Kennedy, 1975; Record, McKeown & Edwards, 1969; Lobl, Welcher & Mellits, 1971; Dryfoos & Belmont, 1979). Recently, Baldwin and Cain (1980) reviewed the results of several National Institute of Child Health and Human Development (NICHD) studies that looked at the long term impact of teenage parents upon their children's devel- opment. With few exceptions, the studies reviewed were, like the Oppel and Royston, Hardy gt 31. and Vandenberg studies, reanalyses of longitudinal data originally col- lected for other purposes. Though quite diverse in their design and methodology, the authors of the NICHD studies reached basically the same conclusions as the previous studies: While excellent prenatal care of the teenager may result in the birth of a healthy infant, the subse- quent health of her child may be severely jeopardized by early parenthood. All analyses show deficits in the cognitive development of children (especially male children) born to teenagers; much, but not all, of the effect results from social and economic consequences of early childbearing. Less consistent effects are found for the children's social and emotional devel- opment and school adjustment (p. 37). None of the studies cited above nor any of those reviewed by Baldwin and Cain (1980) offers an adequate explanation for the developmental deficits that were found. Though the educational and economic disadvantages associated with early parenthood and the greater likelihood of marital breakup or of social isolation are reported to be related to the various deficits, we are given no hint as to the mechanisms through which these rather massive status variables operate. Since about one of every six children born in the U.S. is born to a teenage mother, it is important to discover the etiology of the developmental differences that were cited above. Objective of the Study The objective of this study was to determine the possible causal mechanisms of the developmental differences cited by Oppel and Royston, Hardy 35 31., and Baldwin and Cain. To accomplish this objective, the researcher examined and compared the childbearing and early Childrearing experi- ences of teenage mothers and older mothers. The following research questions were posed: 1. Are the background characteristics of teenage mothers different from those of older mothers? If differences are found, can they be related to the cited developmental differences? Are the childbearing experiences of these two groups of mothers different and, if so, can this difference be related to later developmental differences? Do the children of these two groups of mothers differ from one another at birth? Could such a difference be related to later developmental dif- ferences? Do teenage mothers differ from older mothers in their attitudes toward their children and could such a difference be related to later differences. in the children's behavior and abilities? Do the home environments provided by teenage mothers differ from those provided by older mothers and what would be the relationship of any differences to later development? 6. Are the children of teenage mothers temperamentally different from those of older mothers even in the first few months of life? 7. Are the childcare practices of teenage mothers dif- ferent from those of older mothers? Are such dif- ferences related to differential patterns of devel- opment? A short-term study such as the present one cannot possibly provide final answers to questions about the eti- ology of the developmental deficits cited above. Instead, by focusing upon the late pregnancy-early postpartum period, I hope to be able to describe the processes that may be the early precursors of developmental deficits. There- fore, the results of the present study should be useful in generating hypotheses about the developmental problems associated with teenage parenting. Only detailed longi- tudinal studies can begin to determine the value of hypoth- eses so generated. CHAPTER II REVIEW OF THE LITERATURE There exists an extensive body of literature con- cerned with developmental risks associated with pregnancy and the immediate postpartum period and there is a growing body of literature concerned with postnatal effects upon development. Factors which appear to have a negative effect upon a child's development can be conceptualized as lying on two dimensions or continua: the continuum of reproductive casualty, incorporating the biological events from the time of conception to the end of the perinatal period, and the continuum of caretaking casualty, incorporating the social and environmental events that occur after the time of birth (Pasamanick & Knobloch, 1966; Sameroff, 1975; Sameroff & Chandler, 1975). In general, researchers involved with developmental risk have tended to focus upon one of these two continua to the exclusion of the other. Sameroff (1975; Sameroff & Chandler, 1975) was among the first to argue that the major weakness in risk research has been the assumption that the continua were mutually exclusive, when, in fact, the relationships among the factors along each continuum are transactional. 9 10 In the review that follows, the focus is upon fac- tors located on each of these continua, especially as each is affected by the age of the mother. Evidence for the transactions between the factors on the continua will also be examined. Finally, a model will be developed, based upon the evidence presented in the literature review, for examining the effects of teenage childbearing and child- rearing upon child develOpment. The Continuum of Reproductive Risk There are several aspects of childbearing in which the medical literature reports differences according to maternal age. For purposes of this review these aspects of childbearing will be divided into obstetric and gynecologic factors (experiences of pregnancy, labor and delivery) and neonatal factors (the state of the neonate at birth). Experiences of Pregnancy and Delivery It has been widely reported that teenagers tend to experience complications of pregnancy and delivery more often than women in their twenties (Nye, 1976; Jones & Placek, 1979). Several researchers have reported that pregnant teenagers have higher rates of toxemia or hyper- tension, anemia, and prolonged labor than women who are in the normative childbearing age range of 20-30. Table 1 contains a summary of the research in this area. 11 Table l Complications of Pregnancy and Delivery that are Reported to be More Common for Teenagers than for Women in Their Twenties, and Supporting Studies Complication Supporting Studies Toxemia or hypertension Anemia Prolonged labor Baldwin (1976); Fielding (1978); Hollingsworth g£_al. (1976); Alan Guttmacher Insti- tute (1976); Stickle and Ma (1975); Dott & Fort (1976); McAnarney (1975); Menken (1972); Coates (1970) Fielding (1978); Alan Gutt- macher Institute (1976); McAnarney (1975); Menken (1972); Baldwin (1976); Stickle and Ma (1975); Menken (1972) Baldwin (1976); Stickle and Ma (1975); Menken (1972) 12 It is important to note that these researchers are not unanimous in indicating that teenagers experience each of these difficulties more than other mothers. In fact, the results of several studies with very young adolescent mothers (: 15 years of age), for whom the incidence of these complications is generally reported to be high (Nortman, 1974; Dott & Fort, 1976), provide evidence which appears to contradict these generalizations or at least suggest that the differences may well be less than is gene- rally reported. In three such studies the pregnancy experi- ences of large samples (N = 137 to 471) of 12-15 year old primigravidas (women who are pregnant for the first time) were compared with older primigravidas. No differences were found between the groups in terms of anemia and pro- longed labor (Hulka & Schaaf, 1964; Coates, 1970; Duenhoelter 95 21., 1975). However, the young mothers were reported to experience a somewhat higher incidence of toxemia or hyper- tension in two of these three studies (Coates, 1970; Duen- hoelter, 1975). In a related study focusing upon early adolescents, Dwyer (1974) attempted to determine whether the provision of an adequate prenatal program would eliminate the diffi- culties that have been reported for teenage pregnancies. After providing such a program for 231 pregnant 12-16 year olds, Dwyer reported finding no significant differences between the pregnancy and delivery experiences of young 13 mothers and those of older women (cf.: Berg, Taylor, Edward & Hakanson, 1979). As Dwyer (1974) suggests, teenagers may be at risk during pregnancy because few of them get good prenatal care. Other researchers have reported that teenage mothers begin prenatal care later than older women and generally have fewer prenatal visits (Dott & Port, 1976; Hulka & Schaaf, 1964; Coates, 1970). Furthermore, Ryan and Schneider (1978), in a study of 220 teenagers, reported finding a strong rela- tionship between both the timing of the first prenatal care received and the pregnancy and delivery experiences of adolescents. In taking a closer look at the relationship of pre- natal care to the pregnancy experiences of teenagers, Mednick 33 31. (1979; see also Sutton-Smith, 1979) examined two large studies, one American and one Danish, which showed a linear relationship between maternal age and the number of complications during pregnancy; that is, teenagers experienced fewer complications than older mothers. There was no evidence of physiological immaturity of the teenage mothers. In both studies, all the participants were enrolled in free, high quality prenatal care programs early in their pregnancies. The authors attribute this factor alone with being responsible for a finding that is contrary to that usually reported. Mednick 33 31. (1979) argue that teenage pregnancy occurs most frequently in low SES groups. Since social 14 status has been shown to be related to the quality of obstetrical care available to a person, it seems reasonable to assume that teenage mothers generally would not have access to high quality programs of prenatal care. More- over, since teenagers often hide the fact of pregnancy as long as possible, they are less likely to enroll in any type of prenatal care in the first trimester (Mednick 33 31., 1979). Dott and Fort (1976) offer evidence to support this line of reasoning while agreeing that, with adequate prenatal care, there should be few differences between the obstetric performance of teenage and older mothers. In a major prospective study, the research team of the Collaborative Perinatal Project followed over 26,000 children from birth (Broman _E _1., 1975). The authors reported that the four-year IQ scores of this group were related to both the number of prenatal visits and the length of gestation at prenatal registration. However, both maternal education level and SES were more strongly related to the four-year IQ score. Thus, the relationship between maternal age and the experiences of pregnancy and delivery may not be as direct as is commonly thought. The critical variable in teenage childbearing does not appear to be physiological immaturity. Instead, the critical variable for predicting pregnancy experiences appears to be the timing of onset, quality and quantity of the mother's prenatal care. In turn, a mother's 15 pattern of prenatal care is influenced by socioeconomic factors and maternal age, which are probably covariates (Illsley, 1967b). Complications of pregnancy, labor and delivery such as those discussed above are thought to have a negative impact upon fetal development and therefore on later devel- opment. For instance, Pasamanick and Knoblock (1966), in a review of several retrospective studies, reported that several developmental disorders, including mental defici- ency, behavior disorders, and reading disabilities, were related to complications of pregnancy and delivery, especi- ally toxemia and maternal bleeding. Similarly, Stott (1957) found a strong relationship between later development and the incidence of complications of labor and delivery using a sample of 200 mentally retarded children. However, it should be noted that the relationship between obstetric and gynecoloqic factors and later devel- opment is probably not a strong one. Studies that have used more representative samples than that used by Stott (1957) have reported weaker relationships between pregnancy and delivery factors and the development of very young chil- dren (Goldstein 33 31., 1976; Kopp & Parmelee, 1979). In fact, the relationship between prenatal factors and post- natal development becomes even weaker with increasing age, while the influences of environmental factors become greater (Drillien, 1964; Harper & Weiner, 1965; Nortman, 1974; McDonald, 1964; Sameroff & Chandler, 1975; Sameroff, 16 1975). Thus, one would expect any direct relationship among maternal age, experiences of pregnancy and delivery, and the developmental status of school age children to be quite small. Developmental Status at Birth The two most commonly used indicators of neonatal well-being are Apgar scores and a general evaluation of intrauterine development based upon either gestational age or birth weight or both. A third indicator, the Brazelton Neonatal Behavioral Scale, has been used in several research studies but is not commonly used in hospital delivery rooms or nurseries. In this section, each of these indicators of neonatal viability will be reviewed for evidence of their relationship to maternal age and child development. Apgar scores. Originally devised by an anesthesi- ologist to evaluate the deleterious effects of obstetrical medication upon the newborn (Apgar, 1953), Apgar scores taken at one and five minutes postpartum have become part of normal hospital routine. Five vital signs (heart rate, respiratory effort, muscle tone, reflex irritability, and color) are evaluated and each is scored on a scale of 0-2. The Apgar score is the sum of these five subscores. It should be stressed that these scores are limited to vital functions and that only depressed functions are measured (i.e., a low pulse rate results in a diminished score while both a normal or elevated heart rate receive an optimum 17 score). It is also probable that the evaluation of color at birth (from blue or pale to completely pink) is of little value for nonwhite babies. Despite its imperfections as a sound psychometric instrument, the Apgar scores at one and five minutes have become widely accepted as indicators of neonatal viability. In general, a score of 7-10 indicates a vigorous infant; 4-6, a depressed infant; 0-3, a markedly depressed infant in need of immediate attention (Apgar, 1953). Only a few studies have compared the Apgar scores of infants born to teenagers and infants born to older women. Using 220 teenagers, Ryan and Schneider (1978) found a posi- tive relationship between maternal age and infant Apgar scores. Jones and Placek (1979) reported national data which support such a conclusion. However, Sandler (1979; in Baldwin & Cain, 1980) failed to find any relationship between maternal age and Apgar scores. How do Apgar scores relate to later development? In one study a randomly selected sample of over 200 newborns were rated by a trained observer who had no other delivery room responsibilities (Edwards, 1968). The one- and five- minute Apgar scores were significantly correlated with IQ scores on a concept formation task, and fine and gross motor skills at four years of age, with the highest corre- lations obtained for the motor skills. The five-minute Apgar score accounted for more of the variance in four year 18 mental and motor development scores than either the one- minute Apgar score or the infant's birth weight. Other researchers who have compared Apgar scores with later developmental status have reported much weaker relationships than those reported by Edwards. For instance, in a study of 233 infants, no significant relationships were found between newborn Apgar scores and infant developmental indexes (Gesell & Cattell) at 12 months of age (Caputo, Taub, Goldstein, Smith, Dalack, Pursner & Silberstein, 1974). The authors suggest that the failure to find significant corre- 1ations may be attributable to: (a) limited variability of the scores due to extreme negative skewness of the distribu- tion; (b) biased ratings (the reason for the extreme skew- ness) due to the fact that the attending physician adminis- tered the examination; and (c) use of the one-minute Apgar score rather than the five-minute score, which, at least according to Edwards, is the stronger predictor. - In summary, it appears that the newborns of teenage mothers are likely to receive lower Apgar scores than are other newborns. Secondly, research reports indicate that Apgar scores may be related to later development and there- fore might be useful predictors for the developmental dif- ferences reported for children of teenagers relative to other children. However, it is too Optimistic to assume that the predictive power of the Apgars will be as strong as Edwards' data suggest. First, as noted earlier, the predictive value 19 of perinatal events diminishes with age (Drillien, 1964; Harper & Weiner, 1965; Kopp & Parmelee, 1979; McDonald, 1964; Sameroff & Chandler, 1975; Weiner 31 31., 1968). Secondly, Apgar scores are extremely biased when obtained from the attending physician (Caputo 33 31., 1974; Drage & Berendes, 1966). Finally, it is difficult to imagine how an instrument with both a skewed distribution (by design) and such a limited range (0-10) could prove to be a major predictor of later development over such a wide range of outcome variables. Gestational age and birth weight. Two commonly used indicators of the newborn's status are gestational age and birth weight. Although gestational age technically refers to the time in utero from conception to birth, measuring true gestational age is difficult. Gestational age is routinely recorded in most hospitals and infants with a gestational age of less than 37 weeks are usually considered to be premature and at risk. Generally, gestational age is estimated based upon either the pregnant woman's knowl- edge of the date of conception, the date of her last men- strual cycle, or the date of quickening (Illsley, 1967b). A rarely used clinical method for assessing gestational age, the Dubowitz Scale (Dubowitz, Dubowitz & Goldberg, 1970), has been shown to provide highly accurate estimates. At least partially due to the difficulty of accu- rately estimating gestational ages, birth weight generally 20 is used alone, or in conjunction with gestational age as an indicator of neonatal status. In general, low birth weight infants (< 2,500 gms) are considered to be at risk relative to older infants: "the risk of death in the first year of life is 17 times the risk among infants weighing 2,501 gms or more" (Chase, quoted in Menken, 1972). A much higher degree of risk is associated with that subset of low birth weight infants who are classified as "very low birth weight" (< 1,500 gms). It is widely reported that children of teenagers are more likely than children born to older mothers to be of short gestation, low birth weight, or both (see Table 2). In only two published studies did researchers fail to find any differences in the gestational ages or birth weights of the infants of teenagers and older mothers (Coates, 1970; Duenhoelter 33 31., 1975). However, the relationship of gestational age and/or birth weight to later development is not totally clear. For instance, Edwards (1968) failed to find any relationship between birth weight and any four- year developmental measures. On the other hand, in the Caputo 31 31. (1974) study, both gestational age and birth weight (which are highly intercorrelated) were significantly related to measures of infant development at one year of age. A relationship between gestational age and intellec- tual performance at age nine has also been reported (Muller, Campbell, Graham, Britain, Fitzgerald, Hogan, Muller & Ritterhouse, 1971). 21 Table 2 Studies Reporting Differences Between the Gestational Ages and/or Birth Weights of Infants Born to Teenagers and Older Mothers Gestational age Dott & Port (1976); Illsley (1967b); Nye (1976); McAnarney (1975); Vandenberg (1976, reported in Chilman, 1980) Birth weight Alan Guttmacher Institute (1976); Baldwin (1976); Bro- man 33 31. (1975); Fielding (1978); Hulka & Schaaf (1964); Jones & Placek (1979); Mellor & Wright (1975); Menken (1972); Osofsky & Osofsky (1970); Ryan & Schneider (1978); Stickle & Ma (1975); Lobl 33 31. (1971); Zachau- Christiansen (1975) In one of the few prospective studies dealing with prematurity, Hunt and Rhodes (1977) followed 56 infants, divided into four groups based upon gestational age: 27-31 weeks; 32-34 weeks; 35-37 weeks; and 38-44 weeks. Testing these children almost every month for the first year, Hunt and Rhodes found consistent differences between the groups on the Bayley mental scales. However, when the infants were compared at equal conceptional ages, no differences were found. Thus when compensation is made for unequal gestational ages, the results for all groups are quite similar. From these results, one would expect prematurity p33 33 to have little effect upon development at school age. 22 Sigman (1976) compared the exploratory behavior of premature and full term infants, defining prematurity as a combination of gestational age (< 37 weeks) and birth weight (< 2,500 gms). When preterm and full term infants were compared at eight months conceptional age, the preterm infants explored a familiar object longer and showed less preference for a novel object. Citing other studies which found that exploratory behavior of novel objects at six months is significantly correlated with three-and-a-half year Stanford-Binet performance, Sigman concluded that pre- maturity may be related to lower performances on cognitive abilities tests in the preschool years. When considering studies that use birth weight alone as the major independent variable, one again finds great confusion and disagreement in the literature regarding its relationship to later development (usually later IQ scores). However, there is a chronological pattern to the results as the following reports reveal. For instance, researchers studying children born prior to 1960 regularly report find— ing a relationship between birth weight and later IQ scores (Churchill, 1965; Harper & Weiner, 1964; McDonald, 1964; Weiner, Rider, Oppel & Harper, 1968; Willerman & Churchill, 1967). In studies of children born since 1960, researchers have not found a correlation between birth weight and later IQ, even when infants who were less than 1,500 gms at birth are included (Davies & Stewart, 1975; Francis-Williams & Davies, 1974; Hack, Fanaroff & Merkatz, 1979). However, 23 even recent studies concur that small—for—date children (children whose birth weight is less than expected for their gestational age) perform significantly poorer on IQ tests than do average-size-for-date children (Als, Tronick, Adamson & Brazelton, 1976; Davies & Stewart, 1975; Francis- Williams & Davies, 1974; Kopp & Parmelee, 1979). The explanation for the contradictory results cited above lies in changes that were made in the methods used in neonatal intensive care units, methods that were related to neurological and sensory handicaps experienced by very low birth weight infants (Davies & Stewart, 1975). Due to radical changes in medical procedures, data from earlier eras cannot be applied to the prognosis for today's low birth weight children. By the early 705 the 80-90% of surviving very low birth weight infants were free from serious mental or physical handicap (Hack, Samaroff & Merkatz, 1979). However, though school performance has not been completely analyzed for these later cohorts, there seems to be an increased incidence of learning, reading and behavioral dis- orders despite normal intellectual development (Davies & Stewart, 1975; Francis-Williams & Davies, 1974; Hack t al., 1979). In summary, it appears that for at least some groups, there is a tendency for children of teenagers to be prema- ture and/or of low birth weight (LBW). Though the develop- mental effects of prematurity alone may be negligible, low birth weight may be related to later learning 24 difficulties. The develOpmental outcome for LBW children is also similar to that reported above for children born to teenagers. However, these relationships appear to be con- founded by SES since early parenthood, prematurity, low birth weight and poor developmental outcome all covary with SES (Broman 33 31., 1975; Davies & Stewart, 1975; Dott & Fort, 1976; Francis-Williams & Davies, 1974; Hack 33 31., 1979; Illsley, 1967a, 1967b; Mednick 33 31., 1979; Sameroff & Chandler, 1975; Thompson, 1976). Brazelton Neonatal Behavioral Scale indicators. The Brazelton Scale consists of 20 indicators of neurological development and 27 indicators of behavioral abilities (Brazelton, 1973). It is designed to be used on infants less than one month old and has been used primarily as a research tool with neonates. (A more detailed explanation of the Brazelton Scale is included in Chapter III.) In two studies that have used the Brazelton Scale to compare the newborns of teenagers and older mothers, the results are somewhat different. In a Nashville study, in which "the quality of medical care was maintained," no differences were reported between the two groups of infants on day two, even though the teenage mothers experienced more stress during pregnancy than did the older women (Sandler, 1979, reported in Baldwin & Cain, 1980). A second study was conducted with samples from Florida and Puerto Rico (the quality of medical care is not mentioned). When these 25 infants were tested at two days postpartum, in both samples the children of teenage mothers, though their scores fell within the normal range, were significantly more likely to be underaroused or overaroused than babies of older mothers (Lester, 1978, reported in Baldwin & Cain, 1980). Again, the difference in the results of these two studies may be related to the prenatal care received by the mothers. The relationship between scores on the Brazelton Scale and later development is not well established and more studies in this area are needed. To date, only one long term follow-up study of the Brazelton Scale has been published. Low scores on the Brazelton exam were shown to be predictive of abnormal neurological development at age seven. In comparison with alternative newborn examinations, the Brazelton Scale correctly detected 80% of the abnormal children while achieving a much smaller false alarm rate, i.e., incorrectly labeling normal children as abnormal (Tronick & Brazelton, 1975). Summary The above review of factors on the continuum of reproductive risk has indicated several variables that apparently discriminate between the childbearing experiences of teenagers and older mothers. Unfortunately, the rela- tionship between these factors and later development is not as clear as one would like. For instance, even though many teenagers apparently experience more difficult pregnancies 26 and deliveries than older mothers, the relationship of pregnancy and delivery complications to later development of the child is small. Furthermore, both the experiences of pregnancy and delivery and the child's developmental status are strongly related to the mother's SES level and to the type of prenatal care she received. Similarly, the relationship of perinatal factors to later development is unclear. Apgar scores, which may discriminate between the children of teenagers and older mothers, are reported to be related to later development, but only in a controlled setting. In a normal setting where the attending physician, who obviously has a vested interest in the results of the pregnancy, is the one who assigns the Apgar scores, the reliability, and therefore the useful- ness of the scores, may be in doubt. The relationships of maternal age to prematurity and of prematurity to later development also are unclear. In part, this lack of clarity probably is due to the diffi- culties inherent in any determination of gestational age. Thus, prematurity p33_33 may not be an appropriate variable for use in predicting later outcome. On the other hand, birth weight does seem to be related to developmental outcome and to maternal age, at least for some samples. The relationship between birth weight and later development is strongest for children of very low birth weight (< 1,500 gms). Small-for-date infants also are at risk. In addition, the developmental 27 consequences of low birth weight are similar to those that have been described for the children of teenage mothers. Therefore, birth weight, probably in combination with gestational age, may be a useful perinatal indicator of the later development of children born to teenagers. However, it is important to take into account the relationship between SES and maternal age, prenatal care, birth weight, and developmental outcome that continually reappears in the literature (Davies & Stewart, 1975; Francis-Williams & Davies, 1974; Hack 33 31., 1979; Illsley, 1967; Kopp & Parmelee, 1979). Low income mothers are more likely to become pregnant as teenagers, to receive little or poor prenatal care, to have low birth weight infants and to have children with developmental problems. When children of very low birth weight are divided into groups based on SES, higher scores are characteristic of the higher SES group (Francis-Williams & Davies, 1974). These relationships must be taken into account in any study of the impact of teenage childbearing on child development. Finally, it is important to remember that studies reported above showed that the predictive ability of any pregnancy/delivery variable diminishes with time, at least when IQ is the dependent variable. Unfortunately, similar studies have not been done with the other developmental problems such as those associated with learning, reading, and behavior. Thus one would not expect that the experi- ences of pregnancy, delivery, and the immediate postpartum 28 period, by themselves, to account for the developmental deficits that would have been reported for the children of teenagers. The Continuum of Caretaking Casualty In their review, Sameroff and Chandler (1975) indi- cate that parents influence their child's development through their behavior, their attitudes and the psychological and socioeconomic environments they provide their children. In this section, information relating to the quality of teen- age parenting and its potential impact upon later develop- ment will be reviewed. The effects of parenting style, parental attitudes, and home environment upon child develop- ment also will be discussed. Teenage Parenting Several researchers have examined teenage parenting knowledge and practices. Reviewing the studies on this topic, one finds that a wide variety of approaches have been used. For instance, the samples studied have varied from 12 to 300 subjects, from all white to all black, from all rural to all urban, with some consisting of institutional- ized subjects. The methodologies used have varied from the exclusive use of questionnaires to the exclusive use of unstructured interviews. Nevertheless, the results of these diverse studies are relatively consistent (see Table 3). In general, researchers have concluded that teenage parents may lack 29 Table 3 Characteristics of Teenage Mothers Reported in the Literature Characteristic Citation (a) (b) (C) (d) (e) (f) (g) Unaware of developmental milestones Prone to use physical punishment Insensitive to infant's signals or needs Unaware of how to stimu— late a child's develop- ment Ambivalent toward motherhood Not inclined toward spontaneous play Likely to spend less time looking at and talking to their babies than older mothers deLissovoy (1973, 1975) Epstein (1979) Walters (1975) deLissovoy (1973, 1975) Furstenberg (1976) deLissovoy Mercer (1980) Williams (1974) Epstein (1979) Crumidy (1966) Mercer (1980) deLissovoy (1973, 1975) (1973, 1975) contradicted by Williams, (1974) Sandler (1979, Cain, 1980) in Baldwin & 30 both the knowledge and abilities necessary for adequate parenting, especially when the teenager is the sole care- giver. Teenagers are reported to be insensitive to their children's needs and signals, inconsistent in their care- giving, and apparently more concerned with their own wel- fare than with that of their child. Unfortunately, the studies from which these conclu- sions are drawn suffer from a number of weaknesses. Fore- most among the weaknesses are small samples, samples of mixed age groups, samples of institutionalized subjects, a general lack of comparison groups, either a complete reli— ance upon survey methods or upon unstructured observations, and the rare use of any form of statistical analysis of the differences reported. Only the weight of repeated reports of inadequate parenting among adolescents from studies using varied methods and samples leaves one with any faith in the findings presented (see Nelson, 1973). To date, no study in which teenage parenting has been studied systematically, especially in comparison to the parenting practices of older mothers, has been published. Until that occurs, one must accept the results reported above. Theories of development during the adolescent period provide support for the descriptions of teenage parenting reported above. During adolescence, children are making the transition from childhood to adulthood, establishing their identity or self-concept and establishing their inde- pendence from their parents (Chilman, 1980; Erikson, 1959). 31 At this time, the change to the formal operations stage in cognitive development usually occurs resulting in a period of strong egocentricism (Elkind, 1967). Thus psychologists have suggested that teenage mothers are more likely than older mothers to provide inconsistent care to their children, to experience shifts in mood and behavior, and to be ambiv- alent toward motherhood (Bemis, Diers & Sharpe, 1976; Chilman, 1980; Walters, 1975). Williams (1974) reports that the discussions among the participants of a national workshop for persons working with adolescent parents pro- vide support for the conclusion reached by Bemis 33 31. (1976). In a detailed study of the characteristics of mothers of colicky infants, one finds a description of mothers that is quite similar to that of teenage mothers given above (Lakin, 1957). Mothers of colicky infants are described as: (a) being ambivalent toward their role and role function; (b) having poor self concepts with respect to role function and feelings of inadequacy; (c) experienc- ing poor marital adjustments (see Roosa, 1977, for a review of the marital adjustment of teenage marriages); (d) being tentative and insecure; and (c) lacking facility in carry- ing out mothering activities. Lakin suggests that the feelings associated with these attributes greatly affect the quality of mother-infant interaction. Such feelings are reflected in less adequate responses to the infant, 32 in less effective infant need reduction behavior, and, therefore, in greater mutual tension and discomfort. Finally, descriptions of abusive parents indicate that there may be striking similarities between them and teenage parents. For instance, abusive mothers are described~ as: (a) having inappropriate expectations of the child's development; (b) being unaware of the child's needs (Bavolek, Kline, McLaughlin & Publicover, 1978; Delsordo, 1974; John- son & Morse, 1974); and (c) having a strong belief in the value of physical punishment (Johnson & Morse, 1974). Abusive mothers usually have their first child before they are 20 years of age (Smith, Mumford, Goldfarb & Kaufman, 1975), are from low SES groups, and often are experiencing severe marital conflict (Johnson & Morse, 1974). In a recent review, Bolton (1980) also notes the numerous simil- arities between adolescent parents and abusive parents. He concludes that, while there is no empirical base to sup- port a connection between adolescent parents and abusive parents, there are too many similarities in demographic and behavioral variables to ignore (cf.: David, 1972). To summarize, all the available evidence indicates that teenagers may be less than adequate as parents. Only reports that teenagers' interest in playing with their chil- dren (Williams, 1974) and their energy, enthusiasm and will- ingness to learn parenting skills (Epstein, 1979; Mercer, 1980) provide an encouraging note to the review. However, the insufficient quality of the studies reviewed and the 33 circumstantial nature of the other evidence presented do not generate much confidence that a true picture of teenage parenting in the U.S. has been presented. Maternal Attitudes and Child Development Though not much work has been done directly with maternal attitudes and child development, the studies that do exist suggest that this relationship could be an important one. In one of the recent studies in this area, Broussard and Hartner (1970, 1971) evaluated the attitudes of 120 mothers of full-term, normal, firstborn infants at both day 2 and at one month postpartum. Maternal attitudes were found to be quite unstable during the first month. However, the authors found that children who were rated as average or below by their mothers at one month were much more likely to need psychological intervention by four-and- a-half years of age than were children who were rated posi- tively. This need for intervention was not related to SES or to changes in SE8, to prenatal or postpartum complica- tions, type of delivery, religious preference of the mother, or sex of the child. It is reported that disturb- ances in a mother's early attitudes toward her infant may lead to a chronic disturbance in the mother-child relation- ship (Bibring, 1961), a point made much earlier by Fries (1944). Also, a mother who is insensitive to her infant's cues, or who will not or cannot respond to these cues, can severely harm the developing relationship with her child 34 (Korner, 1974). Finally, self-confidence appears to be an important factor in the way a mother cares for her child (Benedek, 1949). Summarizing the pre-l970 literature, Walters and Stinnett (1971) concluded that parental acceptance, warmth and support are positively related to favorable emotional, social and intellectual development of children. On the other hand, extreme restrictiveness, authoritarianism, and punitiveness, without acceptance, warmth and love were negatively related to a child's emotional and social devel- opment. Recent reviews support the conclusions advanced by Walters and Stinnett (e.g., see Osofsky & Connors, 1979). Obviously maternal attitudes affect the mother- child relationship, although succinct demonstrations of these effects are not readily available in the literature. One would expect maternal attitudes to shape all aspects of a mother's relationship with, and behavior toward, her infant. Especially with the reports of negative or ambivalent attitudes toward parenting for teenage mothers, and their expressed lack of self-confidence as mothers, maternal attitudes may be a factor that distinguishes teenage mothers from older mothers. Home Environment and Child Development One of the more neglected, although potentially significant, influences on child development is the type of home environment in which the child is reared. There is 35 little in the literature on teenage parents to suggest that they provide a qualitatively different environment for their children than do older mothers. However, since teenage pregnancy is associated with lower SES both before and after pregnancy, and since teenage childbearing is often associ- ated with a rather rapid transition to "independent" living arrangements, one might hypothesize that the environments provided by teenage mothers are less Optimal, in general, than those provided by older mothers. An early study of the effects of home environment on child development examined language and intellectual development longitudinally. Seventy-six children born in London were followed from the time they were six months old until they were eight years old. At two-and-a-half years, the child's toys, books and experiences with these items, the parents' use of example and encouragement with the child, and the level of acceptance, warmth and sensitivity of the parents to the child were rated on five—point scales by observers. At two-and-a half, none of these variables showed a significant correlation with concurrently given developmental tests, but each was strongly correlated with IQ, vocabulary, comprehension, and reading quotient scores at age eight (most correlations were about .60). These correlations generally remained significant, but smaller, when social class was statistically controlled (Moore, 1968). Thus, it appears that at least global measures of the home 36 environment during the early years are significant predic- tors of later development and language acquisition. Using more detailed measurements of the environment, Elardo, Bradley and Caldwell (1975; Bradley & Caldwell, 1976) found similar results when assessing: (a) emotional and verbal responsivity of mother, (b) avoidance of restric- tion and punishment, (c) organization of physical and tem- poral environment, (d) provision of appropriate play mate- rials, (3) maternal involvement with child, and (f) oppor- tunities for variety in daily stimulation. Using multiple regression, the multiple correlation of these six variables measured at six months with three-year Binet scores was .54. Measured at twelve months, this correlation was .59; at twenty-four months, it was .72. The importance of the home environment appears to increase with the age of the child, probably in direct relationship with the increased use of language (cf.: Golden & Birns, 1976). Research in this area suggests the mechanisms for some of the subtle influences of SES on child development. For instance, Wachs, Uzgiris and Hunt (1971), using a version of the Bradley and Caldwell Home Scale, documented differences between homes rated as "disadvantaged" and "average." Contrary to common belief, overstimulation of infants in disadvantaged homes accounted for the main dif- ference between home environments. Tulkin and Kagan (1972) obtained similar results in their study comparing working class homes to middle class homes. Excessive noise, too 37 many people providing stimulation, and, in general, levels of stimulation too high for the child's current level of functioning were negatively correlated with psychological development. It seems that environmental stimulation has a curvilinear relationship with development with both high and low levels detrimentally influencing child development (Wachs 33 31., 1971; Tulkin & Kagan, 1972; Bradley & Cald- well, 1976; Elardo 33 _1., 1975; McCall, 1979). The studies reported above clearly suggest that measurements of the home environment (including aspects of the parent-child relationship) can be useful as predictors of later developmental status. The unanimity of the find- ings is impressive considering the variety of variables measured, methods used, and samples tested. The relation- ship reported between the home environment variables and measures of development are also generally larger than the relationships between other variables and development. Socioeconomic Status and Child Development Socioeconomic status is an abstraction generally based upon education level, income and/or occupational pres- tige. The primary use of SES is to divide populations into groups whose members are assumed to share similar character- istics, behaviors, values and social histories. When SES is used as a variable in research, it is used to represent the qualities that the members of these groups share and not social class per 33 (cf-z Illsley, 1967b). 38 Throughout the previous sections of this review, references to the interaction of SES with various predictors of developmental status have been made. SES has been reported to be negatively related to the probability of becoming a teenage mother, and to the number of complica- tions of pregnancy, labor and delivery. It is positively related to the status of the child at birth and the quality of the physical and psychosocial environment in which the child will develop. The influence of this rather gross variable appears to be pervasive. Therefore, it is easy to understand why the positive relationship between SES and development is one of the strongest and most widely reported relationships in developmental research (Ainsworth & Bell, 1973; Bayley, 1965; Broman 33 31., 1975; Carew, 1977; Golden & Birns, 1976; Illsley, 1967; Kopp & Parmelee, 1979; McCall, 1979; Moore, 1968; Ricciuti, 1977; Sameroff, 1975; Sameroff & Chandler, 1975; Wachs, Uzgiris & Hunt, 1971; Werner 33 .31., 1968). Exactly how does SES impact upon child development? The answer to this question is extremely complex. First, as mentioned above, at birth the children of low SES mothers are more likely than other children to be defined as "at risk." Such children are then reared in an environment which may not provide adequate nutrition or medical care, which may be unclean and unhealthy, which may be over- crowded and loud, and which, because of the parents' educa- tional level and value system, may not be responsive to the 39 child's needs and may not provide adequate stimulation to the child. As it was expressed by Lewis and Freedle (1977), the environment of the poor seems to be constructed in such a manner that the children of the poor seem predisposed to remain in the same social environment as their parents. In this sense, children of the poor are born to parents who feel (are) powerless in their environment, an experience which shapes their attitudes and feelings toward life. In turn, the parents socialize their children in a way that prepares the children to feel powerless, a lesson that is constantly reinforced by the parents' lack of responsivity and by the environment in general (Tulkin, 1973). In such an environment, the lack of parental respon- sivity trains the child in "learned helplessness" (Lewis & Goldberg, 1969), thus helping to maintain the poverty cycle. It is the fact that so much of a child's environment is influenced by parental SES that makes SES such a potent influence on develOpment. Parenting Style and Child Development Only one study has been found which compared the parenting styles of older mothers and teenagers (Sandler, 1979, in Baldwin & Cain, 1980). Sandler reported that the older mothers spent more time than the teenage mothers talking to and looking at their babies. Furthermore, these social interactions were associated with higher Bayley scores at nine months of age. 40 There is a growing body of knowledge regarding the relationship between parents' (usually the mother's) behav- ior and the children's later developmental status. A wide variety of methods have been used and numerous questions remain to be answered. Nevertheless, it is quite clear that parenting styles do influence child development. For instance, Epstein and Evans' (1979) longitu- dinal study of children from age two to seven showed that the maternal-infant interactional style was more signifi- cantly related to later intelligence test performance than SES. Furthermore, interaction styles at age two were pre- dictive of academic success at age seven, though not of interaction style at age seven. (This last result is probably due, at least in part, to a radical change in measures from age two to seven. Though the authors refer to their measures of interactional style, it is clear that their measures focused primarily upon maternal verbal behav- ior toward the child at age two. At seven, the focus appears to have changed from the mother to both the mother and child.) In this study, a positive interactional style was one in which the mother was verbally supportive, expanded upon what the child was saying, and asked questions as a teaching device. Negative interactional styles con- sisted of the use of positive and negative commands in response to the child's efforts. Clarke-Stewart (1973) made repeated observations of 36 mothers and their nine—month-old firstborns over a nine 41 month period. A complex maternal factor composed of the appropriateness of maternal behavior for the child's age and ability, expressions of affection, social stimulation, contingent responsiveness, acceptance of the child's behav- ior, and stimulation and effectiveness with materials was significantly correlated with measures of the child's com- petence. Maternal effectiveness and infant attachment to the mother, as well as maternal effectiveness and infant irritability, were significantly related. Maternal restric- tiveness was related to the child's object orientation score on the Hunt-Uzgiris scale. In other words, mothers who pro- vided more stimulation to their children (especially verbal stimulation) were positively influencing their child's devel- 0pment, especially intellectual develOpment. Contingent responsiveness also was related to future intellectual development. In another study, 24 infants who had been adopted within ten days of birth were observed and tested once when they were seven months old and again at eleven months. Though the infant's developmental test scores were found to be correlated with the natural mother's SES, parenting styles of the adoptive mothers were found to have a significant influence as well. Infants who were spoken to less, touched less or given less opportunity to explore their home scored lower on the developmental examinations (Beckwith, 1971). Yarrow, Rubenstein, Pederson and Jankowski (1972) examined the relationship between various types of 42 environmental stimulation (physical and social) and infant abilities among 41 black five-month-old infants from low SES families. Both level and variety of social stimulation were related to performance on the Bayley Mental Development Scale (MDI score). Furthermore, the caregiver's contingent response to infant distress was positively correlated with MDI scores. Yarrow 33 31. suggested that ”relatively early intervention in distress states may have a general facilitating effect, either by freeing the infant to respond to external stimuli or by making him aware of a contingent relationship with the caretaker" (1972, p. 212). Bradley and Caldwell (1976) observed 77 normal infants and their mothers in another study of multiple environmental influences. They reported that children whose performances increased significantly from six months to three years had mothers who were rated more sociable and affectionate toward the child than the other mothers in the study. At 12 months the correlation between this "maternal involvement" variable and the three year Stanford- Binet score was .47 and at 24 months, .55. Tulkin and Kagan (1972) examined the differences in parenting style between middle class and working class white mothers and their firstborn daughters. At ten months, there were few social class differences in the areas of social contact, prohibitions, and nonverbal interactions. However, every verbal behavior observed was more frequent among the middle class mothers, though this difference may "be 43 attributable to a subgroup of middle class mothers who were highly verbal with their infants" (Tulkin & Kagan, 1972, p. 38). There was a tendency for working class mothers to believe that their infants did not have the ability to express emotions and to communicate with others. Therefore, they saw little reason to interact with them. Tulkin and Kagan suggest that working class mothers care for their children as intensively as other mothers do, with the excep- tion of verbal stimulation of cognitive development. Several studies have shown the differential impact of various types of caregiving upon the developmental status of premature or low birthweight infants. For example, one group of researchers followed 51 premature infants for nine months (Beckwith, Cohen, Kopp, Parmelee & Marcy, 1976). After naturalistic observations at one, three and eight months, the Gesell developmental schedules and a sensori— motor scale were administered at nine months. Infants with high Gesell D.Q.s spent less Of their awake time at one and three months receiving physical care, had experienced more postural stress when being held (i.e., they were made to provide some support for themselves), and were given more Opportunities for floor freedom and exploration at eight months. Infants who received higher sensorimotor scores experienced more mutual gazing with the caregiver, more interchanges Of smiling during mutual gazing, more con- tingent response to distress and greater levels of social interaction including more responsiveness to nondistress 44 vocalizations. Beckwith 33 31. (1976) contend that the most significant aspects of enhanced development are the reciprocal social transactions, that is, transactions that occur contingently to the infant's signals, either simul- taneously as in mutual gazing or successively as in con- tingency to distress or nondistress vocalizations. Even though these reciprocal social transactions occupy only a small part of the infant's daily activities, they appear to be the underlying factors in the development of compe- tence. Three other studies with premature infants illus- trate the relative efficacy of even minimal amounts of infant-caretaker interaction. Each study provided extra stimulation to small groups of low birth weight infants during the newborn period. White and Labarba (1976) found an increase in weight gain, compared to six controls, in six infants who were given only 15 minutes additional daily stimulation for a ten day period. Solkoff and Matuszak (1975) found that 11 infants who received 7% minutes of extra handling each hour for 16 hours a day over a 10 day period showed considerably more improvement on the Brazel- ton Scales than did controls. Powell (1974) also demon- strated the benefits of extra stimulation on the develop- ment Of low birth weight infants, though this study was weakened by a 50% attrition rate that was especially pro- nounced in the control group. 45 These studies with premature infants seem to demon- strate that even minimal additional stimulation or inter- action with caregivers can lead to enhanced development for biologically disadvantaged infants, even if the mechanisms for the improvements are not fully understood (cf.: Cornell & Gottfried, 1976). Additionally, certain patterns of care- giving are more likely than others to help infants overcome deficits that may be due to biological disadvantages at birth. Researchers consistently have found a relationship between mothering styles and a child's developmental prog- nosis. Apparently, the critical variable is the amount Of social stimulation, as contrasted to the amount of time spent providing physical care. Though social class differ- ences have been reported for the amount of stimulation pro- vided, parenting style has been reported to be more strongly related to development by school age than is SES, in con- trast to all the previously reviewed influences on develop- ment. However, it is probably more logical to assume that parenting style is simply a mechanism through which SES influences are transmitted. Possibly the most encouraging aspect of research in this area has been pointed out by Moore (1977). One suspects from these observations that the mothers of competent children do not have mothering skills that other mothers lack; none of the behaviors they display would appear to require a high degree of social skill or expertise. But they do require a "mind set" about mothering in which the mother is conscious of her part 46 in nurturing her child's budding intellectual and social competences as well as in caring for her child's physical needs (Moore, 1977, p. 68). Mother-Infant Interaction Up to this point, mother-infant interaction has been seen as a relationship in which only the mother (or caregiver) has influence upon the child's development. The infant appears to be a passive acceptor of environmental stimulation. This is, in fact, the way that much of the literature views mothers and infants. However, there is considerable evidence that the infant plays a significant role in shaping his/her own environment and in determining, to some extent, how caretakers respond to his/her needs. The infant's role in mother-infant interaction is readily apparent in extreme cases. For instance, Shaw (1977) found that mothers of babies labeled "chronic cryers" interacted less with their infants and were less responsive to the infant's cues than mothers of other babies. Similarly, abused children often have abnormal character- istics that appear to predispose the parents to be abusive. These children often are Of low birth weight, may be mentally or physically ill, or may be temperamentally diffi- cult or trying (Sameroff & Chandler, 1975). Another extreme example is provided by work done in a village in which undernourishment is prevalent (Chavez, Martinez & Yaschine, 1975). One group of mothers were provided vitamin supplements early in their pregnancies 47 and their children received food supplements from the twelfth to the sixteenth week of life. By the twenty-fourth week, the supplemented infants were more active and demand- ing than unsupplemented infants. As a consequence, the supplemented infants had higher levels of interaction with their mothers and even their fathers, though it was unusual for fathers in this village to be involved in child care for very young children. These supplemented children also were moved more and spent more time outside the home than unsup- plemented children and thus were exposed to a greater vari- ety of environmental stimuli. However, the child's first influences upon his/her own development are more subtle and occur earlier than the examples given above. Bennett (1976) noted that caretakers, whether professionals or parents, tended to label various aspects of the baby's personality from the first moments. These labels, combined with the infant's particular active- sleep cycle, were related highly to the type and amount of mother-infant interaction by as early as the second week. Even before the infant is able to develop expressive mech- anisms, he/she is exerting an influence on his/her social environment. Mother-infant interaction is more complex than either of the unidirectional models presented thus far. It is best described as a fitting together of two complex orga- nisms and personalities, a bidirectional organization Of behavior. Each infant is born with a unique rhythm and 48 organization of the arousal and inhibitory systems (Bennet, 1976; Brazelton, 1976; Lozoff, Brittenham, Trause, Kennell & Klaus, 1977; Schaffer, 1974) and a characteristic tempera— ment (Bonem, 1978; Thomas, Chess & Birch, 1963). One of the first tasks of the mother is to adapt to the infant's basal characteristics, to learn his/her rhythms and to learn to identify and interpret his/her cues (Sander, 1976, 1977; Thoman, 1975). Her mode of adaptation to these character- istics alters them and thus alters the stimuli to which she must respond. It appears that the infant is in charge at this stage, with the mother adapting to the infant's needs and cues (Brazelton, 1976; Lozoff 33 31., 1977). However, the mother is simultaneously shaping the baby's behavior by altering its sleep-wake cycle (Sander, 1970, 1972) and by presenting her particular style of caretaking and handling to the infant (Call & Marshak, 1976; Thoman, 1974). In optimal situations, mother and infant learn each other's characteristics and styles and learn to identify the signals each uses. In this case, behavioral synchrony has been established and mother-infant interactions generally are characterized by sequences of give-and-take (Brazelton, 1976; Sander, 1970, 1976; Schaffer, 1974). Asychrony in these early interactions is usually related to neurologi- cal Or severe physiological handicaps in the infant (Brazelton, 1976; Condon, 1975; Lozoff 33 31., 1977) or may be predictive Of developmental traumas such as failure to thrive (Thomas, 1975). Since adults are more capable 49 of conscious adaptation, it is extremely important for the infant to experience a sensitive and responsive mother or other caregiver during the first months of life (Dunn, 1976). Mother-Infant Attachment In order to understand the full significance of mother-infant interaction one must understand the concept of attachment. Attachment is one Of the major adaptive pro- cesses for both mother and infant and, because of the devel- opmental level and characteristics of teenage mothers, may be an extremely difficult process for teenage mothers and their infants. Unfortunately, theoretical perspectives Of attachment vary greatly depending upon one's theoretical back- ground. In this section, the varied theoretical perspectives of attachment will be reviewed and the relevance of attach- ment to adolescent parenting will be discussed. Instinct theorists were the first to use the term attachment as they attempted to apply to humans the concept of imprinting that ethologists discovered in animals (Bowlby, 1958). Though the newborn infant did not actually follow its mother, it did behave in a manner that kept the mother close to it. According to this theory, the infant has an innate drive to maintain proximity to its mother (and only its mother) because she is the only source of food (the breast), comfort and safety. Likewise, the mother is innately attached to the infant, for the good of the 50 species. The primary distinction between humans and ani- mals in attachment behavior is that the infant, instead of remaining close to the adult, produces behavior such as crying or smiling that keep the parent close to it. There are a number of problems with Bowlby's theory of attachment. First, Scrimshaw (1978) argues that infanti— cide has been the most widely used method of p0pulation control throughout human history. If attachment is well organized at, or very shortly after, birth, infanticide would interfere with the mother's desire to reproduce. The fact that cultures which practiced infanticide survived seems to preclude early, innate attachment (Freedman, 1974). Second, researchers have shown that fathers are ade- quate caregivers.and fathers and infants do form attachments, even when fathers are not involved in routine caretaking (Parke, 1979). In general, fathers and infants interact differently than do mothers and infants. However, these differences probably result from the socialization of the father and are not necessarily innate. In contrast to Bowlby (1958), Gewirtz and Boyd (1977) and social learning theorists in general, View attachment as a process of mutual conditioning. As the caregiver's contingent responses condition her infant's behavior, the infant's contingent responses condition the caregiver's behavior. Attachment then is seen as a process Of mutual learning and reciprocity. According to this viewpoint, facial features and expressions develop discrimination 51 functions as the mother responds to the infant's needs (cf.: Cairns, 1972). Originally the mother is reinforced by the infant's signals that it's needs have been met and the infant is reinforced for crying by having it's needs met. Through mutual conditioning, eventually facial features, smells, body movements, and more specific sounds may replace the original reinforcers. Others who speak of attachment tend to emphasize the role of psychological attitudes with the behavioral aspects receiving second priority (Klaus & Kennell, 1976; Kennel, Voos & Klaus, 1979; Leifer, 1977; Lozoff 33 31., 1977). Leifer (1977) found that emotional attachment begins as early as the second trimester of pregnancy, after fears of miscarriage have passed. After quickening, women tend to become increasingly emotionally invested in the fetus as the fetus becomes personified and preparations for the birth take place. Leifer considered this linear increase in affect toward the fetus during pregnancy as a significant developmental task of pregnancy and as an intricate;part of the postpartum bonding process. Women who did not develop these affectional ties during pregnancy tended to have difficulty in the transition to parenthood and in establish- ing attachments with the infant. Klaus, Kennell and their associates (1976, 1977, 1979) tend to focus on the immediate postpartum period for the development of mother-infant attachment. This formu- lation suggests that there is a sensitive period for the 52 formation of parental bonds during the first hour or so after birth (see also Ainsworth, 1973; Brazelton, 1976). During this period Klaus and Kennel (1976) propose that there is a species-specific response tO the human infant that lays the foundation for maternal-infant bonding. Stresses that occur during pregnancy, or even during pre- vious pregnancies, which leave the mother feeling unsup- ported or ambivalent about the pregnancy or which precipi- tate concern for the health or survival of either the mother or infant, can delay preparation for the infant and retard bond formation (Cohen, 1966; Leifer, 1976). Therefore, if the mother has positive psychological attitudes toward the child (i.e., the child is wanted and the mother is ready to love and care for the child), the mother will be much more accepting of the child and ready to adapt to the child's needs than if she had negative atti- tudes. Then, if the mother is psychologically attached to the infant at birth, the infant will become attached to the parent much more positively and rapidly than otherwise. However, complications of pregnancy, labor, and delivery, deformities or morbidity of the infant at birth, or intensive care for the infant, can delay or retard the development of mother-to-infant bonds and therefore, infant-to-mother bonds (Goldberg, 1979; Klaus & Kennell, 1976). In this approach, attachment is seen as a two phase unilateral process with the parent doing all the adapting before birth and during 53 the first few days afterwards and the infant doing most of the adapting thereafter. In a more holistic and systemic approach, attachment is seen as the constructed, cognitive-social relationship between a caregiver and an infant. It is a continuously negotiated and achieved adaptive fit between these two individuals as they adapt to one another and become a unit during the early months Of infancy. This complex inter- action between genetic and environmental elements is linked to further socio-emotional organization and development of the child (Ainsworth, 1969; Ainsworth & Bell, 1973; Ains- worth 33 31., 1974; Freedman, 1974; Sroufe, 1979). One of the critical features of this view is the importance of behaviors or events that are consistent and contingent upon the infant's signals (Ainsworth, 1969, 1973; Ainsworth & Bell, 1973; Ainsworth, Bell & Stayton, 1974; Blehar, Lieberman & Ainsworth, 1977; Brazelton, 1976; Lewis & Goldberg, 1969; Lozoff 33 31., 1977; Robson, 1967; Sander, 1977; Sroufe, 1979). When a caretaker responds, or fails to recognize or respond, to the infant's cries or movements, the caretaker is providing feedback to the infant about the efficiency or usefulness of the infant's signal- ing. In turn, the infant's responses to the caregiver's ministrations indicate to the caregiver the acceptability or utility of those efforts. Through this process of sig— naling and feeding back, the caregiver and infant shape each other's behavior (Ainsworth, 1969). The resulting behavior 54 may not be as either would have desired, but it will evolve from this mutual adaptation and reciprocity. However, this process of shaping behavior is funda- mentally different from the mutual conditioning process of Gewirtz and Boyd (1977). As Gewirtz suggests, shaping of another's behavior requires responses that are contingent upon that behavior. The term "contingency of reinforcement" implies that both the mother and infant are passive recep- tors of stimulation. The term feedback, on the other hand, suggests that each of these individuals is actively involved in their interactions, that each learns from the informa- tion received from the other's response, that each uses this information to reorganize behavior in a trial-and-error fashion, and that each response, rather than being the end of a stimulus-response chain Of behavior, is a link in a complex chain of behavior (Ainsworth, 1969). In this view, infant development, rather than being a passive process, is seen as one in which there is a continuous trend on the infant's part toward more active participation in producing effective stimuli (Sroufe, 1969). Attachment is not solely based upon contingencies and responsivity. The mother's ability to provide adequate caregiving and to set an affective mood, her acceptance of the infant, and her sensitivity to the child's affective signals are an important part of the process as well (Ains- worth 33 31., 1974; Sroufe, 1979). Sensitive caregivers provide the proper affective climate, help the infant achieve 55 and maintain an Optimal level of tension, and help their infants organize the behavior to which she will respond con- tingently. However, with sufficient interactions, attach- ments can develop even in harsh, punishing environments (Ainsworth, 1973). This early interaction is the infant's first attempt at the negotiation of an interpersonal relationship and the progress of this relationship is thought to affect future interpersonal relationships (Ainsworth, 1969, 1973; Sroufe, 1979). After all, this is where the child begins to learn the rules that govern such relationships. Furthermore, the infant's ability to control or shape the caregiver's behav- ior provides the infant's first experiences in controlling aspects of its environment. The relative success or failure of these early efforts is thought to affect future cognitive development (Ainsworth, 1969, 1973; Ainsworth & Bell, 1973; Lewis & Goldberg, 1969; Schaffer, 1971; Sroufe, 1979; Sugarman, 1977). In general, infants who are relatively successful at environmental control learn from their experiences, refine their signaling behaviors, and apply their efforts to other parts of the environment. Such securely attached infants are also much more likely to feel free to leave the mother, when they become mobile, and use her as a base from which to explore their environment, a behavior that has been linked to cognitive development (Ainsworth, 1969; Kennell & Klaus, 1976; Sroufe, 1979). On the other hand, infants 56 whose efforts continually meet with failure and frustration, experience learned helplessness (Lewis & Goldberg, 1969). Over time, such infants can be expected to become passive and listless and may even experience traumas such as the failure-to-thrive syndrome (Thoman, 1974). Thus, experi- ences gained by the infant during the first few months of life are seen as critical to much of the child's later development. Unlike the behaviorists, psychobiologists do not see attachment as purely learned behavior. Because the human newborn is so helpless at birth and therefore completely Adependent upon adult caretakers for sustenance and protec- tion, the mutual attachment of adults and infants is obviously a necessity for species survival. Therefore, one would expect inherited thresholds or pathways Of learning to exist to facilitate the attachment process (Ainsworth, 1969; Freedman, 1974; Klaus & Kennell, 1976). The accept- ance of genetic support for the attachment process is one of the major differences between the psychobiological view of attachment and the social learning view. For instance, most theorists agree that social smiling plays an important part in the attachment process. However, social smiling does not occur until the infant is two to three months old (McFarlane, 1974). During the period before social smiling begins, involuntary or non- elicited smiling does occur occasionally. During this time the infant begins making mutual eye contact with the 57 caregiver. Soon, smiling occurs during mutual eye contact and the infant takes on "human" qualities for the parents. Mutual eye contact and social smiling are the infant's first adult-like positive reinforcers. This develOpmental pro- cess, probably genetically influenced, is hypothesized to have the evolutionary utility of allowing mothers eventually to accept early neonatal death without threatening further reproduction. Secondly, it allows another caretaker to rear the child, and form attachments with the child, in the event of maternal death during birth (Freedman, 1974). Another apparently innate develOpmental factor is the infant's ability to adapt to the communication rhythm of those around him/her (Condon & Sander, 1974a, 1974b; Kennell 33 31., 1979). Speech communication is a uniquely human characteristic. In one child development laboratory, microanalysis of videotapes has supposedly shown that hours old neonates who are moving begin to coordinate their move- ments with the rhythm of the human speech around them (Condon & Sander, 1974a, 1974b). Such an ability no doubt would help the infant in its initial primitive communications with its caregivers. Furthermore, reciprocal, contingent behavioral patterns which mothers and infants develOp would be similar in form to human speech; that is, they intercollate their behaviors whereas later they will intercollate their speaking turns. The infant's ability to adapt to the rhythm of adult speech would be helpful both in the devel- Opment of reciprocal, contingent patterns as well as the 58 development of speech. If other researchers can replicate Condon and Sander's findings, this ability will be another example of the interaction of what is probably a genetic pathway and social learning processes that is crucial to overall human development. Thus, in the psychobiological approach, attachment is a critical element in human evolutionary adaptation and as such has meaning only as a long term ongoing process. It is an intimate relationship that is constantly negoti- ated and changing between a caregiver and an infant and which is shaped by the infant's biological status and the mother's developmental history. It is not a singular, easily defined or recognized process but it is the process through which mother-infant interaction style is developed and expressed. Furthermore, because attachment is a process, one can only infer its presence when one sees attachment behaviors (Schaffer, 1971; Ainsworth, 1969, 1973; Klaus & Kennell, 1976; Sroufe, 1979). These behaviors are generally described as proximity-maintaining behaviors, baby's reacting differently toward parents and strangers, harsh reactions from infants separated from their mothers for long periods of time, exploratory behaviors, and reciprocal, contingent behaviors between caregiver and infant. As the above description has shown, attachment is thought to have a significant effect upon a child's 59 socio-emotional and cognitive development. Because of its complexity and its role in all of a mother and infant's interactions, it is a central concept in child development. The significance of attachment for a study of adolescent parenting behavior lies in the question, "DO adolescent parents and their children generally have more difficulty in the attachment process than do other parents and chil- dren?" Since teenagers are known to have more unplanned pregnancies, to be ambivalent toward pregnancy, to experi- ence more stresses during pregnancy, to have more complica- tions during pregnancy, and to have more low birth weight and/or premature infants than Older women, there is reason to believe that teenage mothers would be less prepared to form attachments with their children. Add the fact that teenagers are more likely than older mothers to be single parents or, when married, to have distressed marital rela- tions, and the likelihood of experiencing difficulties dur- ing the attachment process appears to increase. For these reasons, attachment is seen as potentially the most important element that differentiates teenage mothers and their children from Older mothers and their infants. Summary Though the quality of research on teenage parenting practices and attitudes is less than desirable, the results of these studies have been consistent. In all cases, the 60 quality Of teenage parenting has been found wanting. Teen- agers have been reported to lack self confidence in their parenting abilities, to be ambivalent toward the parenting role, and to be unaware Of, and/or unresponsive to, their infant's signals. It has even been suggested that teenage parents are more likely to be child abusers than other parents (Bolton, 1980). The characteristics of persons in the developmental stage in which most teenage parents find themselves also were shown to be in conflict with the demands of being a parent. The elements on the continuum of caretaking casualty have consistently been shown to be related to child devel- Opment. The mothers' attitudes toward parenting in general or her child in particular have been shown to be related to the way the mother interacts with the child and the child's later developmental status. The quality of the home environ- ment and the quantity and type of stimuli that are available to the child have also been shown to be related to later development. Parental social class has been shown to have a strong effect upon many aspects of a child's development and to be strongly related to later developmental status. Finally, the mother-child interactional unit, which actually cannot be separated from the other elements dis- cussed above, appears tO have a strong role as a mediator Of other influences on child development. Because of the contrast between what is known or suspected about teenage parenting behavior and what is necessary for normal or 61 enhanced development for children, the elements of the con- tinuum of caretaking casualty may help explain the devel- opmental deficits reported for children of teenage mothers. A Systems Approach to Child Development The review of attachment above illustrates the need for approaching complex phenomena with methods more com- plex than the bivariate or multivariate unidirectional models (also known as mechanistic or medical models) that have characterized much of the child development research to date. For instance, throughout the review it has been noted that SE8 interacts with many of the independent- dependent variable relationships described. There is a need for multivariate transactional models that take several variables and their interactions into account to explain complex phenomena. The use of transactional models is referred to as the systems approach. A system is a group Of interacting, interdependent elements. The interaction and interdepend- ence of the elements create a unique functioning whole. Any change in the parts produces systemic changes in the whole. The whole can be understood only by considering all the parts and their interactions simultaneously (Buckley, 1967; Kantor & Lehr, 1975). Complex systems can themselves be constructed of interdependent, interacting systems. Thus when the systems approach is applied to an organism the focus is upon 62 organism (a complex system)-environment systems (several complex systems which may include organisms). In such a complex system, one studies the transactions among the ele- ments; i.e., one looks at the ways that the organism alters the environment and is, in turn, altered by that environ- ment. No longer should one be content with simple stimulus- response networks. One must consider an organism that learns from its interactions with the environment and makes conscious efforts to alter that environment. This approach assumes constant change throughout the system in response to the feedback received from the changes occurring else- where in the system. Thus, behavior loses the essence of its meaning outside the context (environment) in which it occurs (Overton & Reese, 1977; Willems, 1977; Hook & Paolucci, 1970; Paolucci, Hall & Axinn, 1977; McGurk, 1977). Willems (1977) outlines the basic assumptions of applying such a transactional approach to the study of humans. These assumptions are: (a) that human behavior must be viewed at levels of complexity that are quite atypical in behavioral science; (b) that the complexity lies in systems of relationships linking person, behavior, social envi- ronment, and physical environment; (c) that such sys- tems cannot be understood piecemeal; (d) that suCh behavior-environment systems have properties that change and unfold over long periods of time; (e) that tampering with any part of such a system will probably affect the other parts and alter the whole, which in turn means we must develop an ecological awareness of the many ways in which simple intrusions can produce unintended effects and the many ways in which long- term harm may follow from short-term good; and (f) that the focal challenge is to achieve enough understanding of such systems so that the effects of interventions 63 and planned changes can be anticipated in a compre- hensive fashion (Willems, 1977, p. 22). The need for such holistic and complex models in the study of the effects of teenage parents on their chil- dren's development should be apparent by now. The review of the literature above has shown continually that univari- ate Or multivariate predictors of development are generally only weakly related to most measures of developmental out- come. Furthermore, most of these relationships are con- founded by other relationships (often unmeasured) with the independent and/or dependent variables, e.g., confounding often occurs with SES, race, marital status, or age of mother (cf: Blank, 1976; Sameroff, 1975; Sameroff & Chandler, 1975; Social Research Group, 1977). Thus it has been shown that early disorders in development, by themselves, have very little relationship to later developmental status. It is only when the con- tinuum of reproductive casualty and of caretaking casualty are examined together that complex transactions are found which reduce or amplify early problems in behavior. Devel- opmental outcomes are, therefore, the products of a child's characteristics and environments, and the transactions between the two (Sameroff, 1975). Inborn deficits cannot be regarded as a static characteristic if the child is thought of as actively and constantly engaged in attempts to organize and structure his/her world. According to this approach, the constants Of development are not some set of 64 traits but rather the processes by which these traits are maintained in the transactions between the organism and environment (Lester, 1979; Lewis & Starr, 1979; Sameroff & Chandler, 1975). Furthermore, the child's perpetual state of active reorganization is thought of as a "self-righting ability," means Of correcting inborn deficits. One explanation for failure of this self-righting mechanism could be an insult to the organism's integrative mechanism which prevents the self-righting ability from functioning. This implies a major insult such as severe neurological damage. It is also possible that environmental forces present throughout development could prevent the normal integrations that would occur in a more modal environment (Sameroff & Chandler, 1975). Model for Examining the Effects of Maternal Age on Child Development By combining this type of holistic, integrative and systemic thinking with the above review of the liter- ature, the model in Figure 1 emerges. This model includes all the causal variables mentioned in the literature. How- ever, the pathways indicate the mechanisms through which each variable's influence is transmitted, rather than link- ing each variable directly to the final outcome variable. By doing so, the emphasis is upon the processes of the influences, thus reducing the proposed importance Of any 65 .ucmEmOHw>OQ paflno coma mad Hmcnoumz mo uomwwm on» msflcflfimxm How Hmco=||.H madman mom Hoonom um msumum Hmucwfimoao>wo m.@Hfl£o u a: maumum Hmumcomc mo muoumoflvcfl Hmowmhgm n m: sofiuomuwucfl unmwcflnuonuos mo huHHmsv u A: Apnea um mcumum Hmuofi>mnon cam HMOfimOHousw: n «C coflumassaum ucwficouw>cw wEoz n w: :ofiumummoum Hmumcwum H H: acmemummaou ucmucfl u m: , mmm n Nw mopsuwuum Hmcuwums u :C wmm Hmcuoume N am mC“ mcl)‘ um mCllTllllnc or cofiuomumDCH HCMMCHIHOSMOZ UCOECOHfiNVCW GEO: ucmfimummflwfi/ NC I‘VII//\ Haw \mwosuflue E)\)\\\.\.\\l\| DGMMCHInguoz guuwm um :ofiumummoum mowumfluouomumno wsumum ucmwcH Hapmcwum Hmcuoumz 66 single variable. This conforms with suggestions that there is a need to shift the emphasis from assessing changes in outcome measures to identifying the processes underlying the changes (Social Research Group, 1977; Sameroff & Chandler, 1975). As an example of this type of emphasis, look at the position of prenatal preparation in the model. The litera- ture suggests that both maternal age and SES are related to the month in which prenatal care begins and the number of prenatal visits. In turn, prenatal care is related to the state of the infant at birth which should be evident both in the medical indicators of birth status and in the results of the Brazelton examination. Both the status of the infant at birth and the mother's degree of preparation for the birth are related to the mother's attitudes toward her child. These same variables are related to the child's temperament. The mother's attitudes will affect her behav- ior with her child. It is only through this complex net- work that prenatal care is seen to have an effect on child development. It is hard to imagine prenatal care having a direct effect upon child development as the results of correlational studies imply. Unlike prenatal care, both maternal age (£1) and SES (52) are expected to have a broader range Of effects upon child development. Whether the effects of these vari- ables in the model are both direct and indirect as shown is an empirical question. 67 It is hypothesized that all the relations noted should contribute to the child's developmental status six to twelve years later. Obviously, there would be numerous environmental and even physiological influences on devel- Opment in the intervening years. For this reason, the model is limited in the amount of variance in intellectual development that it could account for. However, because the model focuses upon the critical mechanisms of develop- ment that will be operating during these intervening years (i.e., the mother-child interactional system), it is reason- able to expect a stronger relationship than is usually found in research that predicts later developmental status (Lewis & Starr, 1979). Notice also that the model does not include any contemporary measures of developmental status at the one to three month level. Because of the widely recognized diffi- culty of using developmental status prior to about three years of age (when language development becomes very important) to predict developmental status beyond age three, it would not be advantageous to include such a measure in the model. The effects of SES and caretaking environment would be expected to be much more potent shapers of development in the intervening years than would develop— mental status at three months, a primarily biological phenomenon (Honzik, 1976; Kagan, 1979; Korner, 1976; Lewis, 1976; McCall, 1973, 1979; Sameroff, 1975). 68 Summary In this chapter, research related to the child- bearing and Childrearing experiences of teenagers relative to older mothers and research related to child development has been reviewed. The factors on the continua of repro- ductive casualty and caretaking casualty have been examined with regard to their ability to differentiate between teen- age mothers and older mothers and their association with the later developmental status of the child. Experiences of the prenatal and perinatal period as well as the physio- logical and neurological status of the newborn were shown to be related weakly, if at all, to later developmental status, except in extreme cases. On the other hand, factors on the continuum of caretaking casualty, including parental attitudes, SES, parenting style, and parent-infant inter- action were shown to be related to a child's developmental‘ status. Caretaking factors were shown to be capable of diminishing or amplifying the effects of genetic or con- genitally acquired deficits. Finally, it was argued that the continua of repro- ductive and caretaking casualty were not mutually exclusive. In fact, in order to gain a better understanding of the complexity of human development, one must examine both continua and the transactions that occur among their ele- ments. Furthermore, rather than focus upon any set of ele- ments, one should examine the processes that maintain or enhance child development if one is interested in predicting 69 later developmental status. A multivariate model was devel- oped which attempted to define the causal pathways and mech- anisms of those "independent" variables in the child devel- opment literature which also appear capable of differenti- ating between teenage mothers and older mothers. CHAPTER III DESIGN The research design for this study was a natural- istic observation, causal-comparative design which uses two groups of mother-infant pairs. Data were collected by inter- viewing mothers prenatally, by using medical records of the pregnancy and delivery, by examining newborn children, by Observing mother-infant interactions in the home, by evalu- ating home environments, by measuring maternal attitudes toward the infant, and by assessing behavioral capabilities of infants at 12 months conceptional age. The remainder of this chapter describes the sample, the consent proce— dures, the details of the methodology, the instruments used, the measurements of variables, and the method of analysis. Subject Recruitment and Selection The subjects for this study were English speaking, primiparous mothers and their well, non-twin infants who were without any known defects. Subjects were excluded if their infants could not be maintained in an open crib after birth, or if the child was given up for adoption. One group of mothers was less than 20 years of age at the time 70 71 of their delivery. The second group met the same criteria as the first except that they were at least 20, and not more than 35 years of age at the time of the birth. All subjects were self-referred. A letter describ- ing the project (Appendix A) was distributed by organiza- tions teaching prenatal classes. Copies of the letter were left in the offices of cooperating physicians, prenatal clinics, family planning agencies, and school counselors which prospective subjects could be expected to frequent. The letter: 1. Explained the nature of the study, 2. Explained the type of subjects needed, 3. Explained the confidential nature of the data collected, 4. Explained the time obligations of volunteers, 5. Explained the remuneration available to partici- pants (12 boxes of disposable diapers), 6. Contained a telephone number that mothers could call if they had additional questions or wanted to volunteer, and 7. Had a business reply postcard attached which mothers could complete and send if they wished to volunteer. Using the letter and postcard, subjects were recruited from Eaton, Clinton, and Ingham counties in Michigan. In Gratiot County, cooperating physicians helped recruit subjects by urging their participation and indicating 72 that the information gathered would become a part Of their medical records. The letter to subjects specifically men- tioned this arrangement (Appendix A). The distribution of live births for these counties for 1978 is shown in Table 4. Table 4 Age-Specific Live First Births for Clinton, Eaton, Gratiot, and Ingham Counties, Michigan Residents, 1978 Age of Mother County 15-19 20-29 All Ages Clinton 68 200 292 Eaton 85 273 392 Gratiot 88 171 268 Ingham _333_ 1,185 1,707 Totals 621 1,829 2,659 aMichigan Department of Public Health, Office of Vital and Health Statistics, 1979. Seventy-eight mothers and their infants took part in all or part of‘the study. The present study is limited to 62 mothers and infants who met all the criteria for participation and for whom complete data are available. Table 5 shows the disposition of all 81 subjects. Table 6 shOws the age distribution of the sample. 73 Table 5 Response Description Subject Category Reason for Losses Number Completed initial 78 interview Healthy child 1 teenager's child died 74 born shortly after birth 1 Older mother delivered in Midland; child could not maintain body temperature and therefore could not be tested 2 teenagers did not inform us of a birth Hospital and l teenage mother decided to 71 Brazelton data drop out after the inter- gathered view 1 teenage mother did not contact researcher when child was born; child was two months Old when researcher reestablished contact 1 older mother moved out of state between inter- view and birth First home visit 1 Older mother left area to 67 completed care for a sick relative; was gone over two months 1 older mother's child was placed in a foster home 1 older mother moved out of state after delivery 1 teenage mother was evicted at the time of the first visit; did participate in the second and third visits Second visit 68 completed 74 Table 5 (continued) Subject Category Reason for Losses Number Third visit 5 mothers (3 teens) gave 63 completed birth beyond the end of the March 31 deadline. As much data as were possible were gathered for future analysis. Cases for final 1 teenage mother partici— 62 analysis pated in only two of three home visits 75 Table 6 Age Distribution of Subjects Age Frequency (n = 62) 15 4 16 l 17 O 18 5 19 4 20 2 21 0 22 3 23 I 6 24 5 25 3 26 9 27 6 28 3 29 5 30 4 31 1 32 ' l 76 Informed Consent Procedures Because of the sensitive nature of studies of teen- age pregnancy, a great deal of effort was taken to be sure that free, informed and legal consent was obtained before any data were collected. For this purpose, the population was divided into three groups with a consent procedure for each group. These groups and procedures were: 1. Women 20 years old or older and women 16-19 years old who were self-sufficient and who were not wards of the Probate Court (emancipated minors): For these groups only the signature of the woman was required. However, the signatures of the husband, significant male, relatives or relevant others were obtained where applicable, to gain their support for the study and to eliminate a possible source of tension as the study progressed. 2. Women 16-19 years of age who were self-sufficient and who were wards of the Probate Court: For this group, after the consent of the woman was obtained, the signature of the Probate Court judge would be obtained before collect- ing any data. Again the signatures of any relevant others with whom the woman may have been living would have been Obtained also. (No women fitting this description volun- teered to be in the study.) 3. Women 16-19 years of age who were not self- sufficient and women 15 years old: For these groups, the signatures of the woman and at least one parent were 77 obtained before data were collected. Whenever the woman was married, her husband's signature was sought also. Immediately upon receipt of a postcard or telephone call from a pregnant woman, a researcher made an appoint- ment with the woman and her husband (or others as appropri- ate). During this meeting, the researcher explained the nature of the study in somewhat greater detail than was available in the letter and answered questions that the woman (or others) may have had. The researcher explained that the study concerned the many ways that different groups of people have of car- ing for children. The researcher was concerned with any— thing that may affect a family's method of Childrearing, including the family's background, the medical records of the pregnancy, delivery and birth, the abilities of the infant at birth, and the mother's style of caring for the infant. The researcher gave the family a copy of the Schedule of Activities (Appendix B) which outlined the study simply and graphically. The researcher also explained that, if for any reason the infant could not be maintained in an open crib by the end of two weeks after birth, the family would be dropped from the study. Furthermore, the time requirements of the study were outlined as: first inter— view, 15—30 minutes; newborn examination, 30 minutes; home visits, about 2% hours each. The informed consent forms (Appendix B) were explained in detail. 78 If the woman were interested in participating in the study after the explanation and after her questions were answered, the relevant signatures, as outlined above, were sought. Any of the signees could terminate the subject's participation. The letters to expectant mothers (Appendix A) and the consent form (Appendix B) used in Gratiot County men- tioned specifically that the data collected would become part of the woman's and baby's medical records. However, women from Gratiot County who wanted to participate in the study without this stipulation could sign the consent form that was used in the other counties. No one took this option. Research Staff The staff for the present study consisted of seven graduate students: six female, one male. Each was trained in the use of all of the instruments used in the study except the Brazelton Scale, which requires special training and certification. (Two staff members who received certi- fication administered all of the Brazelton examinations.) After the end of the initial training sessions, staff meetings were held approximately every two weeks for the remainder of the project. Additional training took place at these meetings and unusual data gathering or recording situations were discussed. 79 The assignment of subjects to staff members was more or less random with two exceptions. As subjects enrolled in the project they were assigned to the staff on a rotating basis, depending upon work loads. A single researcher was assigned to each family for the duration of their participation, including accompanying the Brazelton examiner during his/her visit with the family. However, subjects assigned to the male staff member who indicated during the initial interview that they would be breast feed- ing their baby and would be uncomfortable doing so with the male observer present were reassigned to another member of the staff for the duration of the study. Furthermore, all subjects from the Alma area were assigned to the author due to the time commitments required. Method Initial Interview When informed consent was granted, the Initial Interview Formwwas completed (Appendix C). The purpose of this interview was to obtain background economic and demo- graphic information on the subject and, where relevant, her husband (or other significant male) and parents. At this time the subject was asked to contact the research staff at the time of the birth. The expected date of birth was also obtained. Using this information, the subject was called a few weeks prior to the delivery date to remind her of the study's interest in her delivery. 80 During the initial interview, each mother was engaged in Open conversation about the pregnancy and impend- ing birth. Throughout the conversation the researcher indi- cated his/her genuine interest in the mother's experiences and in children in general. Establishing good rapport with the mother at this time reduced her anxiety about partici- pating in the study and began to desensitize the mother for future visits. Obstetric Data While the mother and baby were in the hospital, the medical charts of the mother and infant were examined. The Obstetric Information Form (Appendix C) was completed at this time. Information about the mother's weight gain dur- ing pregnancy, medications given during pregnancy, labor and delivery, any complications that occurred and informa- tion about the neonate at birth were recorded on this form. In hospitals in which the research staff were not allowed access to the subjects' medical records, the obstetric information was obtained from the mother after she got it from the nursing staff. This indirect system worked well with only a few exceptions. Neonatal Behavioral Assessment Approximately ten days postgpartum, a member of the research team who was certified to do so administered the Brazelton Neonatal Behavioral Assessment Scale (Brazelton, 1973; Appendix C). The Brazelton exam assessed 81 the infant's neurological intactness and behavioral capa- bilities. During this visit, attempts were made to involve the mother in a discussion of the delivery and the general results of the assessment. This interaction also helped to desensitize the mother to the presence of the observer during later visits. Home Observations All infants were Observed at equivalent conceptional ages, based upon the gestational age of the infant at birth, since coqnitive development is related to biological age and not extrauterine age (Amiel-Tison, 1968; Hunt & Rhodes, 1977). Observations took place at l, 2, and 3 months from the expected dates of birth, or at 44, 48, and 52 weeks conceptional age. At those times, naturalistic observations were made in each of the subjects' homes as the family members proceeded with their usual everyday activities. During each of these visits, the babies were Observed through a cycle that consisted of waking from sleep, being fed, and all other activities that occurred until they fell asleep again (or for a maximum of two hours after waking from sleep, whichever came first). The behaviors of the infant and the mother were recorded every 15 seconds using a checklist (Appendix C). The Observers used an automatic timer which emitted a click every 15 seconds through an earphone. 82 Maternal Attitudes During the three home visits the Know Your Baby Scale (Appendix C), a 13 item scale based upon the Neonatal Perception Inventory (Broussard & Hartner, 1971), was administered. During one of the asleep periods before or after the observation, each item was read to the mother and her responses were recorded. Home Environment The home environment was evaluated during visits one and three by using the Home Environment Assessment Instrument (Appendix C), an instrument that was developed from the Purdue Home Stimulation Inventory (Wachs, Francis & McQuiston, 1978). Information for the home environment evaluation was obtained by interviewing the mother, by Observations made during the visit, and by inspecting the child's toys and room. This evaluation took place during one of the infant's sleep periods. Infant Temperament The Michigan Infant Temperament Scale (Bonem, 1978; Appendix C) was given to the mother to complete during a sleep episode in visits one and three. This scale consists of 164 true-false items and took 15-20 minutes to complete. The researcher, at the mother's discretion, cared for the infant's needs during this period, should the infant wake up, to facilitate the mother's uninterrupted completion of the instrument. 83 Developmental Assessment During the third home visit, or within five days of the third home visit, the Bayley Scales of Infant Develop- ment were administered (Bayley, 1969). Each member of the research team was trained to administer and score the Bayley Instrument. The complete research protocol is described in Figure 2. Instruments Initial Interview Form. The Initial Interview Form was designed to gather background data on the mothers in the study, the significant males in their lives (husbands, boyfriends, fathers of the infants who are neither husbands nor boyfriends), and one of their parents. This instrument also included some exploratory attitudinal items which allowed the mother and researcher to begin a dialogue about her pregnancy and the impending birth. Like all of the instruments, this form was designed to be self-coding to facilitate data handling and keypunching. Obstetric Information Form. The Obstetric Informa- tion Form was designed to gather those medical data about the mother's pregnancy, labor and delivery, and about the neonate which are suspected of being most relevant to the infant's later development. After showing copies of the hospital permission forms (Appendix B) to the appropriate 844 .musuucn nudge can muozuoz anode soar nusuucm wanna use cheapo: encodes unuuoosou new accououm noumwmoznl.~ uneven .ossno 0:» do mucusuuan on» can: ocoaoo uoc new: new cofiuamocoo uouuo canoe :ucou on» an ”as: H gucoz.. .meu .aaued concusu coscwusoo can meme .anso canon muosuos nu“: muomucoo unuau. .usoamoam>oa acouCH uo «macaw Sodium on» mean: unwammommfl HancosmoHo>oo .onom acosmuomema accucu savanna: any means usosnuom lama undue“ uo noduusac>m .OOMoumwcwaon ucosauumcm ucoaumommc nauseoua>cn use: .oamom Soon use» noon oz» can: consumes acoucw on» ouu30u nouauduuu Housman: .auom coduo>uomno mac: on» new»: vouoon :OAuoououcn unawcnuuunuo: .oHcOm mama use» 3ocx Ono saw: consumes ucmwcw on» cacao» nou=uwuuc Hacuouoz .Euom coflua>uomno OED: gnu mafia: vouoom :Oau nucuoucn unmucnuuozuoz .oHoom acoEwuoman unnucn cameras: may mean: acoeuuomeou acumen no coauuanu>n .uouwu amends—cu anon—swung" acme unwound usoazouq>cu OED: .oHcom anon use» 30:2 ecu saw: couscous accuse on» ochOu mousuwuuu Housman: .EhOb acauu>uomno 030: via mean: couoom coda nouuoucn accwcnnumsuo: .nommouun Houocauom and cone uOu vocwsflxo ocuooou Housed: .oauom ucosmmonud m.:oumououn moans ucoa lounmmo: Hau0w>o£om can HoodooHouaoz douccooz .uumv owzmmumosoo can uafiOCOOOIOAOOn Monaco ou Hugues 30w>uou=H .ncoauoo unn> noHou Has no nousuccmwn uconcoo uuauo~c« :«uuno .:o«u nonesseuu can wowuwufin (accommou .huwuuwucwo Iwucoo .u00noum sandmxm n sucox N suco: cad cane! axon cos segue: weaned Havocoum mcoHuc>hwmno acumen no sauna .uouucoo amass 85 person and filing a copy in the charts of the mother and infant, information about the medications given the mother during pregnancy, labor and delivery (using a list compiled by Brackbill, 1979), the length of labor, and the number and types of complications that might have occurred were recorded. Data also were recorded regarding the time of delivery and the weight, length and head circumference of the infant. Brazelton Neonatal Behavioral Assessment Scale. The Brazelton Scale was designed to measure the behavioral capa- bilities of the infant before the extrauterine environment has had much of an opportunity to begin to shape the infant's behavior (Brazelton, 1973). It measures those behaviors and responses that the infant uses to negotiate with its physi- cal and social environments. The Brazelton Scale measures the infant's ability to control its state; its reactions to various stimuli; its ability to shut out aversive, inter- fering stimuli; its ability to attend to the signals of the caregiver; and the muscular development of the infant. As part of this evaluation, the examiner also performed a neurological examination to determine the adequacy of the infant's developing neurological functions at birth (Prechtl & Beintema, 1964; Prechtl, 1977). This part of the examination included testing various basic reflexes and motor responses. 86 The Brazelton Scale consists of 46 items: 26 behav— ioral responses which are scored on nine-point scales and 20 elicited responses (neurological items) scored on a three- point scale. For the elicited responses, 80% of all infants will score two; scores of one and three indicate abnormal responses and each is typical of only 10% of all infants. Each of the behavioral items is scored on a unique scale that is defined in the Summary of Brazelton Scale Scoring Definitions (Brazelton, 1973). The Brazelton examination was administered by trained and certified examiners. Training consisted of observing training films, Observing others giving the exami- nation and performing the examination on 30-40 infants. Certification required traveling to Boston where the exami- nation was performed and scored with a minimum level of 90% agreement with a trainer at the Children's Hospital Medical Center. Agreement was defined as scoring each behavior item within 11 of the score assigned by the trainer and disagreeing with the trainer on no more than 1 of the elicited response items. Two members of the research team were trained and certified in this manner. Previous research with the Brazelton has shown that testers trained to a .90 criterion of reliability will remain at that level of reliability for a year or more (Brazelton & Tryphonopoulii, 1972). Test-retest stability over a period of four weeks is .592 at the 11 criterion level and .783 at the 12 criterion level (Horowitz 33_31., 87 1971). These stability data were gathered before the present form of the scale was developed and the least reli- able items had been deleted (Brazelton, 1973). In the only long-term follow-up study to date with the Brazelton exam, the predictive validity of the earlier version of the instrument is compared to that of the stan- dard neurological exam developed by the nationwide Collabo- rative Study sponsored by the National Institute for Nervous Disease and Stroke (Tronick & Brazelton, 1975). The Brazelton exam was comparable to the neuro- logical exam in detecting children who later were judged abnormal. However, the Brazelton achieved this level of prediction without including as many normal newborns in the abnormal category as did the other exam. The relative efficiency of the Brazelton was probably because it elicited higher order functioning of the CNS to predict a recovery process (Als, Tronick, Lester & Brazelton, 1977). The Brazelton Scale also has been shown to be correlated with measures of infant temperament intensity at ten weeks (Sostek & Anders, 1977; Brazelton et al., 1979). The present study uses a seven-cluster scoring sys- tem for the Brazelton scale (Lester gt al., 1978). This §_p£i2£i scoring system groups the 46 Brazelton items into the following clusters: 1. Habituation: How well does the infant maintain state control in the presence of nonharmful environmental 88 irritants? How quickly does the infant habituate to those irritants? 2. Orientation: What is the infant's capacity to attend to and process simple and complex environ- mental events? How does the infant respond to audi- tory and/or visual stimuli? 3. Motor performance: How well is the infant able to maintain adequate muscle tone and to control motor behavior? 4. Range of state: How easily is the infant irritated? How stable or changeable is the infant's behavioral state? 5. Regulation of state: How capable is the infant of controlling his/her own state, or how responsive to a caretaker's ministrations is he/she, once irritated? 6. Autonomic regulation: How integrated are the child's basic control mechanisms with regard to skin color and nonelicited muscular reflexes? 7. Reflexes: What are the signs of possible neuro- logic defects or insults? The decision rules for assigning the appropriate scores for each category are outlined in Appendix D. This system of scoring has several advantages: 1. Several of the Brazelton scales are U-shaped functions. The scoring system makes these scales unidirectional and monotonic. 89 2. By collapsing several related items into a single variable, the composite score may be more reliable than any single score. 3. The 46 items of the Brazelton scale are reduced to a more manageable number of items. A clustering approach with the Brazelton scale has been used successfully in a number of studies to discrimi- nate between groups of infants. A four-cluster approach has been used to discriminate between small-for-gestational- age infants and appropriate-for-gestational-age infants (Als, Tronick, Adamson & Brazelton, 1976). Sepkoski, Coll and Lester (1976) were able to use this same approach to classify successfully 58-80% of a group of neonates as being at risk due to obstetric factors. The seven-cluster approach was recently developed to be a more sensitive measure of infant abilities. Know Your Baby Scale. The Know Your Baby Scale was designed to measure the mother's attitudes toward her young infant. This scale is based upon the Neonatal Perception Inventory or the average baby scale of Broussard and Hartner (1971). In the original version, this instrument consisted of six items referring to negative or unpleasant aspects of caregiving (crying, difficulty with caregiving, difficulty with bowel movements). The mother was given a card contain- ing these six items and asked to describe her impression of the average baby by checking a five-point scale, from "a 90 great deal" to "none" for each item. Then the mother was handed a second card and asked to describe her own baby on the same scales. To get a score for a mother, the scores for the two cards were subtracted. There are a number of statistical difficulties with the Neonatal Perception Inventory (NPI). Such category scales "do not produce estimates of the perceptual magni- tudes of the stimuli used, but rather produce estimates of the relative discriminability of the stimuli" (Shinn, 1974). That is, the relationship of the points on the scale used in the NPI assumes a different meaning for each stimulus. Secondly, category scales are based upon the assump- tion that the "equal appearing intervals" between the response categories are indeed equal. This assumption rarely holds and the distances between categories usually becomes greater at the extremes (Shinn, 1974). The data from such a scale will be nonlinear and probably ordinal and therefore incompatible with most common statistical procedures (Shinn, 1974). Thirdly, instruments like the NPI have severely truncated ranges: ceiling and floor effects must be expected. Once again, this has the effect of invalidating or at least weakening the more common statistical pro— cedures. Finally, using subtraction of two category scales to obtain a final score for a subject simply compounds the problems discussed. When there are already problems due to 91 nonlinearity, unequal intervals, and truncated ranges, it is difficult to assign a theoretical meaning to a result obtained by subtraction. With the Know Your Baby Scale, the technique of magnitude estimation scaling (MES) was used to avoid the problems presented by the original instrument. The pro- cess actually simplified the instrument since each pair of items became a single item. Additional, positive items were added to generate a more complete and, hopefully, more reliable measure of the parents' attitudes toward their child. When using MES, the subjects "attempt to match the magnitude of a number to the magnitude of the sensation produced by a stimulus . . ." (Hamblin, 1974). To facili- tate this process the experimenter should define a standard (in this case, the score of the average baby) and a zero point. The data generated by the MES procedure are gene- rally log normally distributed. In the bivariate case, one transforms the results to a linear form by the follow- ing formula: d = log (c + x) where d transformed score c integer constant determined by plotting the original data: the integer above the point at which the graph crosses the axis 92 x = original score After this transformation, the data will have all the prop- erties of a ratio scale and will be amenable for use with most common statistical procedures. However, in systemic models, in which a variable is both estimated and used as an estimator, one must search for an appropriate transformation. To accomplish this, one uses an iterative procedure to produce transformations similar to the log transformation above. In the present case, the author used (x + c)k where x = the original data, c = -l, =%, 0, 8, l and k = -l, -8, 0, k and l, which pro- duces transformations on either side of the one discussed above. (When the coefficient k = 0, the transformation log (x + c) was used.) After doing these iterations for all variables on the Know Your Baby Scale, the statistical qualities of each transformation were evaluated. The skew- ness (degree to which a distribution approximates a normal curve) and kurtosis (relative peakedness or flatness of a curve defined by a distribution) of each transformation was examined to determine which one produced the distribution that most closely approximated a normal curve, i.e., pro- duced values closest to zero for each statistic (Nie, Hall, Jenkins, Steinbrunner & Bent, 1975). As indicators of positive maternal attitudes, the ratings for the questions regarding smiling, cooing, play- ing with the infant, cuddling, and how much the infant enjoyed being fed and playing with the mother were used. 93 The transformation that was most effective with this group of variables was log (x + 1). After the transformation, these six items form a scale of positive maternal affect with an alpha coefficient of .81. Thus positive maternal affect appears to be a reliable scale as measured by the Know Your Baby Scale. A second transformation [(x + l)%] was the best choice for the negative maternal affect scale which included items such as crying, spitting up, and having trouble with feed- ing, sleeping, bowel movements, and "settling down to a regular pattern of eating and sleeping." After the trans- formation, the negative maternal affect scale had an alpha coefficient of .62 indicating that this scale is not as internally consistent as the scale of positive affect. Interestingly, the scale of negative affect is based directly upon the scale of Broussard and Hartner (1970). The original NPI has been used to predict the need for therapeutic intervention with the mother-infant pair. In a sample of 85 mother-infant pairs, when the primiparous mothers' attitudes toward the child at month one were nega- tive, two-thirds of the cases needed intervention by age 4%. When the mothers' attitudes at one month were positive, about one-fifth of the cases needed intervention. Further- more, when mothers' attitudes toward the infant were con- sistently negative over time, the need for intervention was much more likely than when the mothers' attitudes were 94 consistently positive. In that study, the need for inter— vention was associated with the mothers' attitudes but not with the "educational level of the mother or father, the father's occupation, changes in income since delivery, pre- natal or postpartum complications, type of delivery, age of mother at delivery, religious preference of the mother, moves, or sex of the child" (Broussard & Hartner, 1971). The Know Your Baby Scale, with its improved statistical qualities, should prove to be at least as predictive of the need for intervention as the NPI. Home Observation Form. The Child Home Observation Scale was developed by Parmelee and Beckwith (1972) to study mother-infant interaction in the home environment. Designed to be used during a sleep-awake-sleep cycle during the first three months, the instrument allows one to record what behavior is occurring, who is producing the behavior, and to whom the behavior may be contingent or directed. The observer wears an earphone that emits a click every 15 seconds. The observer then records all behaviors that occurred during that 15 second interval. Most measures, with the exception noted below, will be ratio scores, derived from the frequency counts of the number of 15 second units in which a given behavior occurs divided by the length of awake time during the observation. Contingent responses to baby's vocalizations will be expressed as a percentage of nondistress vocalization; 95 contingency to distress will be expressed as a percentage of distress signal episodes; mean length of a hold will be reported as a frequency of 15 second episodes. Beckwith assessed the observer reliability with the instrument during 30 minute observation samples of 10 babies. Reliability coefficients of individual categories on the instrument ranged from .80 to .98 with the majority greater than .90 (Beckwith 35 al., 1976). The instrument used for the present study (Appendix C) is a variation of one used by Beckwith and Parmelee. In a telephone conversation with Beckwith, the author deter- mined which items were least reliable or were least useful in differentiating between types of mothering behavior. The Home Observation Form and its instruction manual (Appendix B) were developed based upon the original items and the.sugges- tions of Beckwith. In a study of caregiver—infant interactions with 51 infants who were born prematurely, the items on the Child Home Observation Scale were shown to be related to the child's later deve10pment. Specifically, at three months, control, defined by the categories "commands," "criticisms," and "stress musculature," was significantly related to the infant's Gesell Development Quotient scores at nine months of age. Smiling at age three months and mutual gazing at one month were significantly related to sensorimotor skills at nine months of age. It was also discovered that infants who obtained higher Gesell Development Quotients at nine 96 months of age, Spent less of their waking time at one and three months in physical care, that is, in being burped, diapered, dressed or bathed. Finally, at three months, those infants who had experienced more contingent responses to their fuss cries did attain higher sensorimotor scores at nine months (Beckwith, 1976). For the present study, a rigorous training program was developed for the observers. Videotapes of mother- infant interaction were obtained from another research pro- ject. These tapes were scored numerous times by the research staff until concensus scores were obtained and the conceptual definitions of the behavioral categories were sharpened. Training sessions were also held with live mothers and babies both in artificial and home settings. These training sessions were concentrated in the first few weeks of the pro— ject but were also held every three to four weeks throughout the project. Data for calculating interrater reliabilities were collected approximately every 90 days. The interrater reliability coefficients for the observation items are shown in Table 7. Behavior in other categories did not occur frequently enough during the testing sessions for accurate reliabilities to be calculated. Home Environment Form. The Home Environment Assess- ment Form is based upon the Purdue Home Stimulation Inventory of Wachs, Francis and McQuiston (1978, 1979). As Wachs (1978) describes it: 97 Table 7 Interrater Reliabilities for Observation Items Variable Alpha Bottle Feed .994 Held/Pick Up .922 Stress Musculature .853 Affectionate Touch/Pat/Rock .982 Talk/Vocalize .973 Infant Fuss .980 Infant Nondistress Vocalization .863 Look 4821 Average .933 The Purdue Home Stimulation Inventories (PHSI) Sections I-III is [sic] an attempt to measure physical parameters of the child's home environment. The emphasis of this instrument is on the physical stage upon which social interpersonal interactions take place rather than upon the social interpersonal transactions themselves. The PHSI was designed as an improvement over the Caldwell Inventory of Home Stimulation (1966) which mixes social and physical stimuli, and the Yarrow Instrument (1975), which is quite limited in scope (Wachs, 1978). The interrater reliability of the individual items on the PHSI ranged from .46 to 1.00 according to Wachs (1978). For the present study, an interrater reliability of .90 was achieved during training. 98 There were a number of difficulties with the PHSI which prompted the development of the present instrument. The items in the PHSI were quite poor statistically. Scales for these items were variously multidimensional, categorical, nonlinear, had truncated ranges and were otherwise inadequate for most statistical uses. The present instrument was developed in an attempt to correct these difficulties by eliminating some items of question- able value, and by using magnitude estimation scaling (see page 91 for a detailed explanation). The data gathered by these procedures are more reliable and stable than that gathered by the PHSI. Additionally, the items are much more amenable to statistical analysis. Those categorical items that remain are used for descriptive purposes only. Michigan Infant Temperament Scale. The MITS was developed by Bonem (1978) as an alternative to the most widely used instrument for measuring temperament, the Carey Infant Temperament Scale (CITS) (Carey, 1970). Bonem has demonstrated that the scales of the CITS do not reliably measure the nine temperamental variables: activity, adap- tability, intensity, threshold, persistence, mood, approach, distractibility, and rhythmicity. The coefficient alphas ranged from .30 to .68, indicating a low level of internal consistency. The MITS on the other hand, has coefficient alphas ranging from .71 to .86 on eight of the nine scales (persistence was dropped as a variable since it appears to 99 be a multidimensional item with low reliability at its present level of deve10pment). In addition, the MITS scale was shown to have a high degree of short term stability (r = .62-.89 after two weeks) and a moderate level of temporal stability over a longer period (r = .4l-.64). These stability figures are similar to those achieved by other measures of infant temperament (Bonem, 1978; Persson-Blennow & McNeil, 1979). The key for the Michigan Infant Temperament Scale is in Appendix D. Data Coding With the exception of the Home Observation Form, all instruments used in this study were designed to be pre- coded to eliminate the need for a separate coding operation and to facilitate entering data into the CDC 750 computer. As each completed instrument was given to the data coding staff, it was reviewed for completeness and legibility. Any questions about the instrument were immediately referred to the staff member who had completed it. Information from complete and legible forms was typed directly into disk- stored computer files. Later, another staff member compared a print-out of each computer file with the completed instru- ment to detect errors in data entry. Any such errors were then corrected. The Initial Interview Form had one question that required coding. When subjects were asked about their 100 occupation, the researcher entered a complete description of the job on the form. Using tables from Reiss, Duncan, Hatt and North (1961) two trained members of the coding staff independently assigned a prestige score to each sub- ject based upon the job description. Whenever there were disagreements in the assignment of a score, the two staff members conferred until an agreement was reached. When an agreement could not be reached, the author was consulted to make the final decision. The Home Observation Form is a complex instrument requiring a great deal of effort to code. For the present analysis, the data on this form were coded in the form of frequency counts. In each behavior category, a count was made of the number of times that each code was entered. Because of the complexity and tedium of this task (there were 35 behavior categories, six behavior codes and up to 480 observation periods per form), each form was double checked by a second coder before being entered into a com- puter file. As with the other forms, the computer print-out was also checked for accuracy and any errors were corrected. Analysis The primary analysis for this study consisted of an estimation and evaluation of the model that was develOped in Chapter II (Figure l, p. 65). This model was evaluated by using a computer program called LISREL®, the analysis of linear structural relationships by the method of maximum 101 likelihood (Joreskog & Sorbom, 1978). LISREL is speci— fically designed to handle causal models similar to that in Figure 1, especially those with cases of mutual causa- lity. There are a number of advantages to the use of LISREL in the analysis of complex causal models. First, LISREL allows one to use models with two parts: a measure- ment model and a theoretical model. The measurement model allows the researcher to specify how the theoretical or unobserved variables will be measured using the observed variables. For each theoretical variable that uses more than one observed variable as an indicator, a factor analy- sis is performed to define the theoretical variable by means of the common underlying variance of the indicators. The reliabilities for the measurements of the observed variables (e.g., interrater reliabilities) may be supplied by the researcher or they may be calculated by the program. The theoretical model or the structural equation model indicates the causal effects (Joreskog & Sorbom, 1978). Second, LISREL allows the researcher to evaluate not just the specific causal linkages but also the entire causal model. LISREL performs a chi-square goodness-of-fit test to determine how well the estimated model explains the vari- ance of the data. For exploratory research, this goodness- of-fit test provides the researcher with important informa- tion about which causal linkages are most or least useful. 102 Third, LISREL is a full-information maximum likeli- hood statistical routine. This means that LISREL uses the information that is available for all variables in the model to make each estimate. It also means that LISREL's esti— mates of population parameters are those values of the population parameters that are most likely to have generated the observed sample data. For these two reasons, the LISREL estimates are both consistent and efficient (Fink, 1980). The model presented in Chapter II is a theoretical model. The complete research model for the present study is presented in Figure 3. A definition of the symbols used in the model is presented in Table 8. The structural equations for the theoretical model are presented in Table 9. The equations for the measurement model are presented in Table 10. The following decisions were made with regard to the measurement of certain variables used in the model in Figure 3 (see Table 8 also). 1. As an indicator of SES, total family income was used rather than the income of specific individuals. The objective was to measure the total financial resources that were available in the child's immediate environment. 2. As another indicator of SES, maternal education was used. It was often the case that the husband or, in the case of unwed teenagers, the parents had a higher education level than the mother. However, the mother in 103 .ucmeaoHo>oo uuanu can: 0m< Hocuouux uo vacuum mzu megawadxm new Honor nonconomnu.m muDOMm u can now ~nu «nu on» ....s. mm mm NN AN on on an AA an a U u U u U u u U u ...... ... mm» vs» n~> «a» nu» ow» was on» ”as on» a» Mag H~« was o~ ans mu» an» was L L .H L “Hz Ha« Ha« wax nag mg «g a“ ham mam mam cam ma «ufl um afi saw one mag oau man «an Nu a» 104 Table 8 Definition of Symbols Used in Figure 3 51 E2 01 maternal age x1 = mother's age at the time of delivery socioeconomic status x2 = the highest occupational prestige rating among the household members x3 = maternal education level x, = family income, total income from all sources of all family members in the household (divided by 1,000) prenatal preparation y1 = did the mother participate in a birth preparation - class (1 = yes, 2 = no) y; = month of pregnancy in which prenatal care began neurological and behavioral status at birth y3 = Brazelton Neonatal Behavioral Assessment Scale (BNBAS) orientation summary score y» = BNBAS motor performance summary score y5 = BNBAS range of state summary score ya = BNBAS regulation of state summary score y, = BNBAS autonomic regulation summary score ye = BNBAS reflexes summary score Note: the BNBAS habituation summary score was removed from the model due to an excessive amount of missing data physical indicators of neonatal status Y9 = birth weight of infant, in kilo grams Table 8 (continued) 105 Du 05 05 Y10= gestational age at birth, in weeks Y11= S-minute Apgar score maternal attitudes toward child Y12 Y13 Yiu Yis Yle Y17 positive negative positive negative positive negative attitude attitude attitude attitude attitude attitude index index index index index index from from from 'from from from month month month month month month infant temperament Y1e Y19 Yzo Y21 Y22 Y23 an st Michigan Infant Temperament Scale (MITS) activity score MITS mood score MITS intensity score MITS threshold score MITS distractibility score MITS rhythmicity score MITS approach score MITS adaptibility score home environment stimulation st Y2? Yze Y29 the number of rooms per person measured at 3 months number of caregivers reported at 3 months is there a quiet place where the child can be put when sleeping? l = yes 2 = no number of audio/visual response toys at 3 months 106 Table 8 (continued) n7 = quality of the mother-infant interactions (all 3 months) Y3o = percentage of time mother held child while he/ she was awake y31 = percentage of time mother talked to child Y32 = percentage of time child vocalized y33 = percentage of distress periods responded to by mother in 45 seconds or less ya. = percentage of time spent in mutual gaze Yss = percentage of time spent in feeding and care- taking activities, i.e., feeding, burping, bathing and diapering As (lambda) = regression coefficients relating unobserved or theoretical variables to observed vari- ables ¢ (phi) = covariance of the exogenous theoretical variables 51 and 52 Ys (gamma) = regression coefficients relating exogenous, theoretical variables (5's) to the endogenous variables (n's) 85 (Beta) = regression coefficients interrelating the endogenous theoretical variables Gs and as (delta and epsilon) = measurement errors in observed variables, x and y, respectively gs (zeta) = errors of prediction 107 Table 9 Structural Equations for the Theoretical Model in Figure 3 01 = Y1§1 + Y2€2 + C1 02 = Y3§1 + Yugz + ‘8101 + 'ana + C4 03 = Ys€1 + Y6€2 + ”8301 + C3 m. = Y7€1 + 7852 + -Ba.n1 + -Bsn2 + -Ben3 + -87ns + -Ben7 + Cu 05 = ‘8902 + ’81003 + C5 05 = Y9€2 + C6 07 = Y10€1 + Y11€2 + 'Biinu + ‘Bizns + “81306 + C7 every case was the primary caregiver. The mother's intel- lectual resourcefulness, vocabulary, knowledge, and moti- vation would constitute or shape the child's environment more than would those qualities of any other person and therefore have a greater impact upon the child's develop- ment (cf.: Leibowitz, 1974a, 1974b). 3. All home environment variables were measured during the third month. These measures were quite unstable during the first three months. Many mothers spent four to six weeks with the baby before returning to school or work, thus having greatly different numbers of caretakers from month 1 to month 3. Several mothers in both age groups moved during the child's first three months (one older mother moved three times), thus completely changing the measured physical environment during this time. The third 108 Table 10 Structural Equations for the Measurement Models in Figure 3 Y1 = 1101 + 81 Y21 = A2105 + £21 Y2 = 3201 + 62 Y22 = A2205 + €22 Y3 = A302 + E3 Y23 = A2305 + €23 Yu = 4402 + 8» qu = 32405 + €24 Y5 = 1502 + 85 st = 12505 + 825 Ya = 1602 + £6 st = 42506 + 626 Y7 = X702 + £7 Y27 = 82706 + £27 Ya = A802 + 68 Y2a = AZBNG + €28 Y9 = A903 + 89 Y29 = X2906 + £29 Yio = A1003 + 610 Yao = A3007 + €30 Y11 = A1103 + €11 Y31 = A3107 + 831 Y12 = A1204 + £12 Yaz = A3207 + £32 Y13 = A1304 + 813 Y33 = A3307 + €33 Y1u = A1404 + £14 Yau = X34n7 + €34 YlS = 41504 + €15 Yas = 13507 + £35 Y16 = Alena + 816 Y17 = A1704 + 817 X1 = A3651 + 51 Yie = A1605 + 818 X2 = 33752 + 52 Y19 = X1905 + 619 X3 = 43352 + 53 Y2o = A2005 + £20 X4 = X3952 + 5» A. 135 = l; 61 = 0 since 51 has a single indicator. B. A1, A3, A9, A12, A13, A25, A30, A37 were set equal to 1 so that the measurement model for each theoretical variable was identified. 109 month measures were assumed to be more stable than first month measures and to be more indicative of maternal prepa- rations or plans. 4. The measures of the quality of mother-infant interaction are based on data from all three observations. By combining the data from all the observations, these measures are more representative of the true state of the mother-infant dyad than are measures from any one observa- tion. Measures based upon a single observation could be quite misleading if the visit was extremely short (a common occurrence during the first observation) or if the mother or infant were experiencing an atypical day. All other measures were quite straightforward as defined in Table 8. In the few cases in which there were missing data, the mean values for the appropriate age group were substituted. One of the critical issues in the estimation of such a complex model is that of identification. Essentially this is the issue of whether or not there is enough informa— tion available to make the estimates requested. Models which are underidentified cannot be estimated. Models which are just identified can be estimated but cannot be tested. Overidentified models are constrained to have a unique solution and the degree of overidentification (the amount of excess information available upon which to base the estimates) is related to the reliability of the results. 110 The model in Figure 3 meets the necessary conditions for overidentification with 679 degrees of freedom (Joreskog & Sérbom, 1978). CHAPTER IV RESULTS AND DISCUSSION To facilitate the presentation of the analysis, this chapter has been divided into two parts. In the first part, the data describing the characteristics of the two groups of mothers and infants will be presented and discussed. In the second part, the estimation of the research model will be presented and discussed. Characteristics of the Two Groups of Mothers and Infants The literature review suggested that teenage mothers and their infants may differ from older mothers and their infants in some ways that are critical to the infants' development. The data analysis that follows indicates that, indeed, these two groups do differ in many ways that may affect infant development. However, not all of the differences favor the older mothers. Only differences for which E < .05 will be considered significant. Socioeconomic Status As might be expected, the two groups of mothers differed greatly in measures of socioeconomic status 111 112 (Table 11). The teenage mothers had significantly fewer years of education (though several were still in school; see Table 12), their families had significantly lower total incomes, and the workers in teenage families held signifi- cantly less prestigious jobs. The males involved in the teenage pregnancies in this sample also had significantly less education than their older counterparts. None of these socioeconomic results are surprising since education, income and occupational prestige are all functions of age, to a certain extent. As shown in Table 12, most of the teenage mothers were still in school and thus were not part of the labor force. Furthermore, almost half of the young families depended upon public assistance for their income, a significantly larger proportion than in the older families. In those cases in which a teenage mother or her male partner were working, the jobs they held were often entry-level or part-time. Two older mothers and one younger mother reported no income. In all three cases the mother was living alone and was being supported by her family. This family sup- port apparently varied as did the mothers' needs and was not regular. These mothers had applied for, and were expecting to receive, public assistance. Note also the differences in the standard devia- tions of the two groups for the variables in Table 11. One of the assumptions of the F-test is homoskedasticity: the variances of the two samples are assumed to be equal 113 Table 11 Socioeconomic Characteristics of Study Families by Maternal Age Teenage Older Mothers Mothers F n = 14 n = 48 Years of Education Completed Mother mean 10.8 14.83 37.6 (p<.01) SD 1.5 2.30 df=(l,60) range 9-14 12-21 Father* mean 10.7 15.20 19.8 (p<.01) SD 1.4 2.60 df=(1,50) range 9-12 12-20 Family Income mean $8,385 $21,553 13.9 (p<.01) SD 9,051 12,264 df=(l,60) range 0-34,400 0-54,000 Occupational Prestige mean 25.1 57.6 18.1 (p<.01) SD 24.2 25.3 df=(l,60) range 0-72 0-93 *For fathers, n = 7 for the teenagers and n = 45 for the older group. 114 Table 12 Descriptive Characteristics of Subjects by Maternal Age Teenage Mothers Older Mothers [1:14 11:48 Frequency (percent) Frequency (percent) Mother in School 10 (71%) 8 (17%) Father in School 1 (14%) 10 (22%) Mother Works 2 (14%) 33 (69%) Father Works 5 (71%) 41 (91%) Family receives public assistance 6 (43%) 4 ( 8%) Maintains relation- ship with infant's father 7 (50%) 45 (94%) Race - Mother White 11 (79%) 48 (100%) Black 3 (21%) 0 ( 0%) Father White 6 (86%) 40 (89%) Black, 1 (14%) 0 ( 0%) Hispanic 0 ( 0%) 3 ( 7%) Other 0 ( 0%) 2 ( 4%) Living Arrange- ments of Mother with male 2 (14%) 43 (90%) with her parents 6 (43%) 0 ( 0%) alone 3 (21%) 4 ( 8%) with male & other relative l ( 7%) 0 ( 0%) with other relative 1 ( 7%) 0 ( 0%) with female roommate l ( 7%) l ( 2%) 115 Table 12 (continued) Teenage Mothers n = 14 Frequency (percent) Older Mothers n = 48 Frequency (percent) Age - Mother mean range Father* mean range 17.3 15-19 22.1 17-31 26.0 20-32 27.0 22-37 *For fathers, n = 7 for the teenage families and n = 45 for the older families. 116 (SD = /V§EI§KEE). If the differences between the variances of the two groups are statistically significant, the F-test is inappropriate and the results in Figure 11 may be mis- leading (Glass & Stanley, 1970). These are numerous instances of possible heteroskedasticity (unequal variances) in the tables that follow and the interpretation of the results should reflect this potential source of bias. There are several possible explanations for the instances of possible heteroskedasticity mentioned above. First, one group may have given more accurate responses to the questions and scales than the other. Second, the small number of teenage mothers in the study suggests that the sample may not have been large enough for an accurate esti- mation of the means for this group. Third, heteroskedas- ticity can result from bounded ranges due to the homogeneity of the sample. In this case, the relationship of one vari- able to another may not be linear and transformations may need to be found to produce linear relationships before applying most statistical techniques (cf.: Hanushek & Jackson, 1977, pp. 142-144). Other Demographic Characteristics In this sample, teenage families had higher incomes and prestige scores if the mother continued to live with her parents. In some of these cases, the higher income was achieved simply by having more persons under one roof receiving public assistance. In both age groups, mothers 117 who lived alone were more likely to receive public assis- tance than were the other mothers. Half of the teenage mothers maintained a relation— ship with the father of the infant during and after the pregnancy, while almost all (94%) of the older mothers maintained such a relationship. All but one of the older mothers who maintained such a relationship with the infant's father were married to him. Among the younger mothers, fewer than half of those who maintained a relationship with the infant's father were married to him. (These comparisons may be misleading since staff field notes indicate that many of the unmarried couples appeared to be cohabiting by the third month postpartum.) Table 12 also contains the distribution of the study families by race, living arrangements and age. The samples were predominantly or all white. While 90% of the older mothers lived with the infant's father at the time of the initial interview, only 21% of the teenagers did so. Almost half of the teenagers continued to live at home. Several mothers in each group lived alone during and after the pregnancy. In the age distributions in Table 12, it is interesting to note the age difference between the teenage mothers and their mates. The teenage mothers apparently selected (or were selected by) mates who were several years older than themselves, though only two of the males were older than 21. 118 Pregnancy and Delivery Experiences In Table 13, information about the pregnancy and delivery experiences of the two groups of mothers are dis- played. The rate of participation in prenatal classes (Lamaze, Leboyer, LaLeche) among the older group of mothers was more than double that of the young mothers. The differ- ence between the participation rates of the men is even greater with a considerably larger proportion of the men in the older group participating. Though there was very little difference between the two groups of mothers in the type of prenatal care received, there was a significant difference in the timing of the onset of that care (Table 13). On the average, teenage mothers in this sample began prenatal care almost one month later than did older mothers. Even more important than the average is the fact that some teenage mothers did not begin prenatal care until the seventh month or third tri- mester of the pregnancy. None of the older mothers began prenatal care later than the fifth month or second tri- mester. Note the possible heteroskedasticity. There was very little difference between the two groups of mothers in the amount of weight gained during pregnancy (Table 13). Interestingly, the teenagers appear to have done a better job of weight management during preg- nancy and avoided the extremes experienced by some of the older women. Characteristics of Pregnancy and Delivery 119 Table 13 by Maternal Age Teenage Older Mothers Mothers F n = 14 n = 48 Mother participated in birth preparation classes 5 (36%) 40 (83%) Father participated in birth preparation classes* 1 (14%) 37 (82%) Type of prenatal care personal physician 10 (71%) 41 (85%) clinic 4 (29%) 7 (15%) Month of pregnancy that prenatal care began mean 3.10 2.30 5.89 p<.05 SD .40 .14 range 1-7 1-5 df=(1,60) Weight gain during pregnancy (1bs.) mean 27.60 30.50 .93 SD 8.40 10.57 range 18-51 15-73 df=(l,60) Expected a partner to attend delivery 12 (86%) 46 (96%) A partner attended delivery 12 (86%) 43 (90%) Partner was father 4 (29%) 41 (85%) Type of delivery normal vaginal 12 (86%) 29 (60%) forceps l ( 7%) 12 (25%) caesarian section 1 ( 7%) 7 (15%) 120 Table 13 (continued) Teenage Older Mothers Mothers F n = 14 n = 48 Sex of child female 9 (64%) 21 (44%) male 5 (36%) 27 (56%) Planned to breast— feed 9 (64%) 46 (96%) Actually did breastfeed 8 (57%) 41 (85%) *For fathers, n = 7 for the teenage sample and n = 45 for the older group. 121 The age groups had similar expectations with regard to having a partner with them during the delivery, and in most cases, their expectations were correct. However, if the mother was older, the father of the infant was much more likely to be the mother's partner than if the mother was young. On the other hand, of those teenagers who main- tained a relationship with the father, over half had the father with them during delivery. Significantly fewer teenage mothers than older mothers planned, or did, breast- feed the infant. The data in Table 13 also indicate that the teen- age mothers generally had easier deliveries than the older mothers. A larger percentage of teenage mothers had normal vaginal deliveries than did the older mothers. Data on the length of labor and the complications experienced during pregnancy, labor and delivery were incomplete and in many cases quite unreliable (due to varying definitions and means of reporting) and therefore are not reported here. Neonatal Status In Table 14, the indicators of neonatal status taken from medical records are shown. Overall, the two groups of infants were quite similar with respect to gestational age, birth weight, length, head circumference, and Apgar scores. Though the differences in the average gestational age and length of the two groups were statistically significant, the differences are small and the means for the infants of Characteristics of Newborns by Maternal Age 122 Table 14 Teenage Older Mothers Mothers F n = 14 n = 48 — Gestational Age (weeks) mean 38.86 40.04 5.44 p<.05 SD 1.96 1.58 — range 34-41 36-44 df=(l,60) Weight (grams) mean 3,201.71 3,523.04 3.53 SD 717.41 512.58 range 1,960-4,536 1,512-4,564 df=(l,60) Length (centimeters) mean 49.71 51.58 4.78 p<.05 SD 3.47 2.54 _ range 43-56 42-56 df=(l,57) Head circumference (centimeters) mean 33.31 34.02 1.12 (n=13) (n=4l) SD 2.56 1.98 range 29-38 30-37 df=(l,52) 1-minute Apgar mean 7.73 7.68 0.01 (n=11) (n=44) SD 1.73 1.25 range 4-10 3-9 df=(l,53) 5-minute Apgar mean 8.82 8.88 0.10 (n=45) SD .98 .57 range 6-10 7-10 df=(l,54) 123 teenagers are well within the normal range. However, the teenage mothers in this study did tend to deliver slightly smaller babies and to have shorter pregnancies than the older mothers. It is noteworthy that only two teenagers and one older mother gave birth to an infant weighing less than 2,500 gms., and that no infant weighed less than 1,500 gms. at birth. Note the possible heteroskedasticity. The Brazelton 7-Cluster Summary scores are shown in Table 15. Once again, the two groups of infants are quite similar with only one significant difference between them, that being their motor performance scores. The dif- ference in motor performance scores primarily was due to the more optimal scores received by the children of older mothers on two of the Brazelton Scales: activity level and defensive movements. However, it is interesting to note that the infants of teenagers consistently score lower than the infants of older mothers on each of the items. The sample sizes for the habituation item were reduced because this item could be scored only if the infant were asleep when testing began. Maternal Attitudes and Infant Temperament There were no significant differences between the attitudes toward their infants that the two groups of mothers expressed (see Table 16). Each group rated their infants high on the positive scales and low on the negative scales. However, the younger mothers had a broader range 124 Table 15 Brazelton Summary Scores of Newborns by Maternal Age Teenage Older Mothers Mothers ‘F n = 14 n = 48 Reflexes mean .79 1.06 .72 SD 1.05 1.08 range 1-4 0-5 df=(l,60) Habituation* mean 3.38 3.86 .21 SD 3 25 3.50 range 3.0-8.0 1.5-9. df=(1,34) Orientation mean 6.18 6.38 .24 SD 1.03 1.46 range 4.5-8.0 2.8-8. df=(l,60) Motor Performance mean 4.97 5.39 4.29 .p<.05 SD .82 .61 range 3 6-6.0 4.0-6.6 df=(l,60) Range of State mean 3.46 3.95 3.10 SD 1.03 .88 range 105-408 105-500 df=(l,60) Regulation of State mean 4.77 5.11 .94 SD .98 1.21 range 3.3-6.5 2.3-7. df=(l,60) Autonomic Regulation mean 5.93 6.65 3.51 SD 1.56 1.17 range 2.0-8.3 3.3-8. df=(l,60) *n = 8 and 28, respectively. 125 Table 16 Attitudes Toward Child by Maternal Age As Measured By the Know Your Baby Scale Teenage Older Mothers Mothers g n = 14 n = 48 Positive attitude summary scores Month 1 mean 31.389 31.327 0.00 SD 6.005 3.216 range 22.6-47.3 26.8-39.9 df=(l,60) Month 2 mean 34.162 33.244 .275 SD 8.977 4.490 range 26.1-60.3 22.7-44.6 df=(l,60) Month 3 mean 33.661 34.051 .051 SD 9.406 4.142 range 25.8-59.2 27.3-44.5 df=(l,60) Negative attitude summary scores Month 1 mean 15.220 15.083 .024 SD 2.291 3.037 range 10.9-17.7 8.6-20.5 df=(l,60) Month 2 mean 15.892 15.843 .003 SD 3.503 3.059 range 10.2-22.0 7.9-22.3 df=(l,50) Month 3 mean 16.277 16.036 .077 SD 3.031 2.811 range 11.0-19.1 9.0-20.2 df=(l,60) 126 (on the high end of the scale) and a larger standard devi- ation on the positive scales than the older mothers. On the negative scales, the teenage mothers gave a narrower range of ratings than did the older mothers. Thus it seems that some of the teenagers seemed to assign quite high posi- tive scores to their children while all teenagers seemed to consider their infants as relatively average on the negative items. This may indicate a social desirability bias in the response of the teenage mothers. The results of the Michigan Infant Temperament Scale are shown in Table 17. There were no significant differences or consistent patterns of differences between the two groups on the eight temperament variables for which a "1" indicates a high level of the attribute and a "0" indicates the com- plete absence of the attribute. Home Environment Though there were no significant differences in maternal attitudes or infant temperament, there were several differences in the environments the mothers provided for their infants. The teenage mothers in this study lived in more crowded conditions than did the older mothers (Table 18). The homes in which the teenage mothers lived had both more occupants (note the possible heteroskedasticity) and less space per occupant than the homes of the older mothers. These differences are due, in part, to the fact that several teenagers lived with their parents or other relatives while 127 Table 17 Infant Temperament Characteristics by Maternal Age--Results of the Michigan Infant Temperament Scale Teenage Older Mothers Mothers E n = 14 n = 48 Activity Score mean .40 .46 1.74 SD .14 .16 df=(l,60) Mood Score mean .67 .64 .20 SD .16 .16 df=(l,60) Intensity Score mean .40 .37 .24 SD .15 .17 df=(l,60) Threshold Score mean .66 .64 .09 SD .19 .22 df=(l,60) Distractibility Score mean .46 .50 .40 SD .23 .21 df=(l,60) Rhythmicity Score mean .64 .63 .03 SD .23 .17 df=(l,60) Approachability Score mean .85 .87 .22 SD .17 .14 df=(l,60) Adaptability Score mean .71 .67 .59 SD .19 .19 df=(l,60) 128 Table 18 Characteristics of the Home Environment at Three Months by Maternal Age Teenage Older Mothers Mothers E n = 14 n = 48 Number of adults living in house mean 2.60 1.96 13.174 ‘p<.01 SD 1.27 .20 range 1-5 1-2 df=(l,60) Density (number of rooms per person) mean 1.43 1.73 4.043 p<.05 SD .51 .49 "' range .89-2.33 1.00-3.00 df=(l,60) House has quiet place for baby 9 (64%) 47(98%) Baby's room is decorated with pictures that stand out from background 10 (71%) 41 (85%) Window shades are not drawn all the time 11 (79%) 43 (90%) Number of toys with audio-visual response mean 5.10 8.70 3.895 SD 3.63 6.54 range 0-12 2-35 df=(l,60) There are noncom- mercial items infant can use as toys 13 (93%) 37 (77%) There is a mobile over the crib 6 (43%) 44 (92%) 129 the older mothers lived with their spouse or alone (see Table 12). The socioeconomic differences cited earlier probably contributed to the differences in crowdedness (£_= .31-.45 between education, occupation, income, and density; p < .05; see correlation tables, Appendix E). Because of the relatively crowded conditions of the teenage mothers' homes, a significantly smaller proportion of their homes, compared to the older mothers' homes, had a quiet place where the infant could be put during naps. Also, the homes of the older mothers were slightly more likely to provide stimulation to the infant from the wall decorations and open windows than were the homes of younger mothers. In terms of other sources of stimulation, the older mothers provided their infants with more audio-visual response toys than the younger mothers provided their infants; one younger mother provided none of these toys by three months. However, the homes of younger mothers offered more noncommercial items that the infant could use as toys (small ashtrays, knick-knacks) than those of the older mothers. A significantly larger percentage of the older mothers than the younger mothers had a mobile over the infant's crib. The Caretaking Environment The caretaking environment of the two groups of mothers is described in Table 19. There was very little 130 Table 19 Characteristics of the Caretaking Environment at Three Months by Maternal Age Teenage Older Mothers Mothers F n = 14 n = 48 Number of caretakers mean 2.400 2.600 0.25 SD 0.940 .870 range 1-4 1-6 df=(l,60) Proportion of times that mother is primary care- taker mean .704 .723 0.08 SD .227 .201 range .333-l.000 .286-l.000 df=(l,60) Number of times mother and infant left neigh- borhood in week prior to the 3-month visit mean 4.100 5.600 2.03 SD 2.540 3.690 range 0-10 l-18 df=(l,60) Number of times mother and infant visited neighbors in week prior to the 3-month visit mean 2.400 1.700 0.59 SD 4.400 2.400 range 0-15 0-11 df=(l,60) Infant's sleep schedule regular 8 (57%) 12 (25%) demand 6 (43%) 16 (33%) variable 0 ( 0%) 20 (42%) Infant's eating schedule regular 5 (36%) ll (23%) demand 7 (50%) 20 (42%) variable 2 (14%) 17 (35%) 131 difference between the groups in either the number of caretakers or the proportion of time that the mother was the primary caretaker. It appears that the older mothers made trips beyond their neighborhood somewhat more often than younger mothers, probably due in part to the older mothers' greater access to cars. However, teenage mothers took their infants to visit with neighbors more often than did older mothers. The teenage mothers in this study also had more persons in the home during the observation period than did the older mothers, the result of living with more people and of having more visitors during this time period. In Table 19, the eating and sleeping schedules of the two groups of infants are compared. Significantly more of the teenagers than the older mothers described their infant as having a regular sleep schedule as opposed to sleeping only whenever he/she was tired or varying between these alternatives. While none of the teenage mothers described their infant's sleep schedule as variable, over 40% of the older mothers did so. The eating schedules of the two groups of infants were similar, though again fewer teenage mothers considered their infants' eating schedule to vary between "regular" and "demand." The pattern of the responses suggests that a social desirability bias could be operating here. The characteristics of the mother-infant interaction patterns of the two groups of mothers are described in Tables 20, 21 and 22. As is shown in Table 20, the two 132 Table 20 Characteristics of Mother-Infant Interaction by Maternal Age: Maternal Behaviors as a Proportion of Infant's Awake Time Teenage Older Mothers Mothers F n = 14 n = 48 Mother holds infant mean .686 .690 .005 SD .232 .185 df=(l,60) Mean length of hold (no. of 15-second periods) mean 38.891 35.882 .249 SD 24.261 18.451 df=(l,60) Mother affection- ately touched, patted or rocked infant mean .346 .437 1.842 SD .228 .216 df=(l,60) Mother used posi- tive commands with infant mean .013 .030 1.989 SD .014 .042 df=(l,60) Mother used nega- tive commands with infant mean .007 .006 .067 SD .008 .009 df=(l,60) Mother talked to infant mean .309 .478 8.070 p<.01 SD .218 .189 df=(l,60) Caretaking activities mean .485 .502 .137 SD .155 .157 df=(l,60) 133 Table 20 (continued) Teenage Older Mothers Mothers F n = 14 n = 48 Mother aroused, stimulated, soothed infant mean .033 .059 1.584 SD .084 .064 df=(l,60) Radio, television or loud appliance on mean .730 .617 .950 SD .352 .387 df=(l,60) 134 Table 21 Contingent Responsivity by Maternal Age Teenage Older Mothers Mothers F n = 14 n = 48 Proportion of awake time spent in mutual gaze with mother vocalizing mean .069 .114 5.187 p<.05 SD .053 .068 df=(l,60) Proportion of non- distress vocali- zations to which mother responded contingently mean .153 .202 1.652 SD .147 .119 df=(l,60) Proportion of dis- tress vocalizations to which mother responded con- tingently mean .360 .561 12.504 p<.01 SD 2.09 .180 df=(l,60) Table 22 135 Characteristics of Mother-Infant Interaction by Maternal Age--Infant Behaviors as a Ratio of the Infant's Awake Time Teenage Older Mothers Mothers .F n = 14 n = 48 Fussing mean .073 .072 .003 SD .053 .036 df=(l,60) Nondistress vocalizations mean .214 .263 1.527 SD .142 .127 df=(l,60) Mouthing mean .044 .055 .341 SD .053 .061 df=(l,60) Playing with objects mean .033 .028 .141 SD .061 .036 df=(l,60) Looking at people mean .244 .219 .825 SD .120 .083 df=(l,60) Smiling mean .031 .040 1.083 SD .027 .030 df=(l,60) Smiling during mutual gaze mean .015 .025 2.485 SD .014 .023 df=(l,60) 136 groups of mothers are similar in the proportion of time they held their infants, touched their infants affection- ately, or used positive or negative commands with their infants. The proportion of time spent in caretaking activ- ities or in stimulating the infant and the proportion of time that a radio, television, or loud appliance was on also were similar for the two groups. However, older mothers spent a significantly larger portion of the infant's awake time talking to the infant than did younger mothers, almost 50% of the infant's awake time compared to less than one-third. In Table 21, we see that on two of the three vari- ables the older mothers were contingently responsive to their infant's signals significantly more often than were the younger mothers. Older mothers spent a significantly larger proportion of time than the younger mothers vocal- izing to the infant while mother and infant were looking at one another's eyes. The older mothers also were signifi- cantly more responsive to the infant's distress signals than were the younger mothers. The older mothers appear to have been slightly more responsive to the infant's nondistress vocalizations as well. There were no significant differences between the activities of the two groups of infants (Table 22). The infants of teenage mothers spent slightly more time than the infants of older mothers looking at people, including their mothers. The infants of the older mothers spent 137 slightly more of their time vocalizing, mouthing, smiling, and smiling during mutual gaze. Discussion of the Descriptive Analysis The Continuum of Reproductive Casualty As suggested in the review of the literature, the teenage mothers in this study did differ from the older mothers on some variables on the continuum of reproductive casualty. The teenage mothers had shorter pregnancies than the older mothers, and their infants were slightly smaller than those of the older mothers, supporting the findings of numerous studies of teenage pregnancy (see Table 2, p. 21). However, the teenage mothers also had easier deliveries than the older mothers (cf.: Mednick 23 21., 1979), though the younger mothers started prenatal care later and were less likely than the older mothers to take birth preparation training courses. For this sample there was no relationship between prenatal care and delivery experiences (see the table of correlations, Appendix E). Note that in a total of 15 tests we could expect to find one significant at the p < .05 level just by chance. In fact, five tests were significant. Note also that these tests were not independent (see Appendix E). The differences that were found in the pregnancy and delivery experiences of the two groups of mothers do not appear to have had much effect upon the status of the infants after birth: there 138 was no difference in the Apgar scores of the two groups of infants and only one of seven Brazelton scores differenti- ated between the two groups. Thus, the continuum of reproductive risk does not seem to be significant, by itself, in differentiating between the teenage and older mothers. The differences that did exist between the two groups of mothers were few and relatively small. Unless the caretaking environments of these infants interacted directly with the reported dif- ferences, there is no reason to believe that the variables of the continuum of reproductive casualty should be related to any developmental differences that may be found to exist between the children of these two groups of mothers at a later date. The Continuum of Caretaking Casualty Maternal Attitudes. The review of the literature suggested that there might be differences between teen- agers and older mothers in their attitudes toward their infants. Since teenage pregnancy is usually unplanned and therefore causes major disruptions in the mother's life, it would not be surprising to find that adolescent mothers held less positive attitudes toward their infants than did older mothers. As measured by the Know Your Baby Scale, there were no differences between the maternal attitudes of the two groups of mothers in this study. There was no evidence 139 for the negative or ambivalent attitudes of teenage mothers reported elsewhere (Crumidy, 1966; Mercer, 1980). It should be pointed out that in both the Crumidy (1966) and Mercer (1980) studies, maternal attitudes were rated by an observer and neither study compared the attitudes of young mothers directly with a comparison group of older mothers. Home Environment. In the review of literature, it was reported that various elements of the home environment apparently had an effect upon a child's development. There was no direct evidence that the homes of teenage mothers differed from those of older mothers. However, since teen- age mothers and older first-time mothers generally differ in SE8 level (Card & Wise, 1978), and since home environ- ments generally differ by SES level (Wachs gt al., 1971; Tulkin & Kagan, 1972), one would suspect the home environ- ments of teenage mothers to differ from those of older mothers. The homes of the teenage mothers in this study did differ from those of older mothers in the types and amount of stimulation each provided to the infants. In the homes of the teenager there were more adults, there was less space per person, and there were more peOple in the home during the observation period than in the homes of older mothers. Thus, it appears that the infants of teenage mothers were exposed to considerably more social stimulation than the infants of older mothers. Furthermore, the teenagers' homes 140 were less likely than those of the older mothers to have a quiet place where the infant could be put to escape from stimulation when upset or tired. Since there are no clear definitions of under, optimal, or excessive levels of social stimulation, it is impossible to draw any firm conclusions about the implications of these data. It is at least a possibility that the infants of some of the teenage mothers in this study received excessive amounts of social stimu- lation from which they could not escape. On the other hand, the teenagers' homes offered fewer opportunities than the older mothers' homes for some sources of nonsocial stimulation that have been related to cognitive development. Compared to the homes of the older mothers, the teenagers' homes were less likely to have con- trasting decorations in the baby's room which would attract the infant's attention, less likely to have a mobile over the infant's crib, and there were fewer audio-visual response toys to stimulate the infant. Though the teenagers' homes were more likely than the older mothers' homes to have non- commercial items that the infant could use as toys, these items not provide the same stimulation as audio-visual response toys nor was there any effort to use these items as toys in the author's experienCes during observations. Thus, there were a number of significant differences between the environments provided for the infants of the two groups of mothers during the infants' first three months. The infants of teenagers received less stimulation from 141 their nonhuman environment than the infants of older mothers. The infants of teenage mothers lived in more crowded condi- tions than the infants of older mothers. Were the above cited environmental differences to persist, one would expect that the environments provided for the two groups of infants would be a significant contributor to the developmental differences cited by Oppel and Royston (1971) and Hardy gt 1. (1978). Socioeconomic Status. The differences between the two groups of mothers on the SES variables were quite large. The older mothers averaged four years more education, over two-and-a-half times as much family income, and had family job prestige levels that were double those of the younger mothers. Generally speaking, these comparisons indicate that the older mothers had many more material, informational and experiential resources than the younger mothers to help them in their role as parents. The pervasive influenc of SES on the various aspects of child development i apparent from a review of the correlation tables for the va 'ables in the research model (Table 23; see also Appendix E). Not only are maternal age and the three indicators of SES highly positively corre- lated, but each of the four variables are highly correlated, positively or negatively, with several of the variables from the continuum of reproductive casualty as well as the home environment, maternal attitudes, and mother-infant 142 Table 23 Selected Correlations Between Maternal Variables and Variables Related to Child Development* Age Oczzgailon Education giggig Status Birth weight .23 .17 .21 .17 Gestational age .23 .14 .29 .15 Home density .40 .31 .40 .45 Number of A/V toys .23 .09 .26 .17 Talks to infant .30 .22 .25 .42 Mutual gaze .19 .07 .10 .25 Response to distress .38 .21 .40 .39 *If r 3 .25, p < .05. 143 interaction variables. Thus, it appears that, as suggested in the review of the literature, both maternal age and SES are potentially strong influences upon child development. Whether the influences of maternal age and SES are inde- pendent or complementary remains to be seen. Mother-Infant Interaction. Measures of infant characteristics that influence mother-infant interaction indicated that there were no significant differences between the two groups of infants on eight temperament character- istics. There was no reason to suspect that any differ- ences in temperament should exist since temperament is an inherited characteristic which underlies personality and personal style (Thomas 23 al., 1963). However, one would expect that infant temperament would be related to maternal attitudes and mother-infant interaction. A review of the correlation tables (Appendix B) shows that this is not the case. Of the 98 correlations between the temperament vari- ables and the maternal attitudes or mother-infant inter- action variables, only three are statistically significant, approximately as many as one would expect to find by chance. Unfortunately, bivariate correlations hide any influence that temperament may have on attitudes or interaction since correlations are gross measures that reflect both direct and indirect influences. 144 Among the more direct measures of mother-infant interaction, only three significant differences were found between the two groups of mothers and infants. Older mothers talked to their infants more, spent more time vocalizing during mutual gaze, and were contingently responsive to the infants' distress signals more often than the younger mothers. These three interaction variables have been found to be significantly related to a child's cognitive develop- ment (cf.: Beckwith, 1971; Yarrow g; 31., 1972; Beckwith gt al., 1976). Furthermore, both vocalizing during mutual gazing and responding to distress signals are the types of contingent responses that are thought to play a significant role in the development of attachment as well as in the child's socioemotional and cognitive development (cf.: Ainsworth, 1969, 1973; Ainsworth & Bell, 1973; Ainsworth _e__t 31., 1974; Blehar gt a_l., 1977; Brazelton, 1976; Lewis & Goldberg, 1969; Lozoff, 1977; Robson, 1967; Sander, 1977; Sroufe, 1979). The difference in the type and quantity of mother-infant interaction between older mothers and their infants and between teenagers and their infants on these three critical variables is potentially related to the developmental differences reported by Hardy 33_21. (1978) and Oppel and Royston (1971). Another interesting relationship is found by review- ing the correlation tables (Table 23; Appendix E). The three SES variables are correlated with the three inter- action variables and each of these correlations approaches 145 significance or is significant at the p < .05 level. Thus, once again, the causal influences of maternal age and SES are confounded. Summary The descriptive analysis has shown that the results of this study generally support the findings in the research literature. Teenage mothers generally began prenatal care later in their pregnancies, had shorter pregnancies, and gave birth to slightly smaller infants; their homes provided more human and less nonhuman stimulation, they had fewer material or educational resources to apply to child rearing, and they generally exhibited less contingent interaction with their infants than the older mothers. Furthermore, SES was found to be as strongly related to most of the variables discussed as was maternal age. It was noted that some of the differ- ences between the two groups may be due to the violation of the homoskedasticity assumption of the F-test. However, the teenage mothers were found to have had easier deliveries than the older mothers despite beginning prenatal care later. No differences were found between the maternal attitudes of the two groups of mothers or between the temperament characteristics of the infants and only one difference was found between the two groups on the Brazelton summary scores. In general, the teenage mothers in this study exhibited much more positive characteristics than have been reported in the research literature. 146 The bivariate relationships that were examined have generally supported the research model (Figure 3, p. 103). However, because of the complex relationships that are hypothesized to occur among these variables, an examination of bivariate relationships cannot provide sufficient infor- mation about the value of the theoretical model in Figure 3. An evaluation of the research model will be presented in the sections that follow. The Estimates of the Parameters of the Research Model The analysis of the research model was done using LISREL, a full-information maximum-likelihood method of estimating and evaluating systems of linear structural equa- tions (Joreskog & Sorbom, 1978). In LISREL, information about all the relationships in a model are used to generate estimates for each parameter of that model. Though a full- information analysis technique was used, for ease of discus- sion the results will be presented separately for each sec- tion of the model. (The entire analysis also appears in Appendix F.) The Measurement Models Figures 4 through 10 contain the results of the estimation of the measurement model for each of the theor- etical variables. The parameter estimates of the measure- ment model for the exogenous variable g2, socioeconomic status, are shown in Figure 4. As in all the measurement 147 52 AK VAR 52 = 425.298 Socioeconomic Status \x\\\\\ \\\1 A38 = .441 ! Y X“ ) VAR a, = 89.525 (.472) (.544) (.455) Family Mother's Family Occupational Educational Income Status Level n1 Prenatal Care \*\ A1 = 1.000 A2 = 1.528 $ YI Y2 1 4 VAR £2 = 1.004 VAR €1-= .326 (.308) (.229) Prenatal Month prenatal preparation care began class Figure 4.--LISREL Estimates for the Measurement Models for the Socioeconomic Status Variable £2 and the Prenatal Care Variable n1 (Reliabilities in Parentheses). .1418 .Anumozucuudm as moauasaamaaom. ~c manuaum> mauoum auuucooz an» ace mauve: ucoamusnua: on» sou «macaquuu anumHa--.m magmas nouuuaawuu manna no «aqua oven-heuuom noxoauum unacc0u9< couuuusmux mo «and: none: noduuuauwuo Acco.v Anoo.v Amqo.v Aooo.v Amoo.v Aooc.v Noa.n I on mts. noc.a I an m<> sad." I no m<> man. I mu m¢> «we. I :u m¢> o~a.~ I nu m<> o» h» m» m» :h a» SHN.-- . as sms.n~ . as oen.en . as oo~.~a- . ma non.» . 3‘ ooo.~ - ms oauuuo unaccouc mo cucunuuvsq m msumum Hmumcomz man How Hmooz ucmfimusmmwz may now mmumEHumm qmmmHAII.m musmwm ouoow Hawae moccasin mwm Hmsoaumumou uanmB euufim Aooo.v Awoo.v Aqflo.v mas. u Haw ma> msa.~ u as. m<> Ham.~m n mu m¢> as» as» as a “GO.H- u 22‘ amm.H u 024 ooo.a n ma //////, \\\\\\ maumum anomaooa mo wMOumoHc:H Hmofivmz m: 1550 .Amomosucoumm ca moauafianmaswxv .c onnmaum> muauauu< Hocuoumz on» you Mano: acmemusmum: an» ace unassumm ammeAnn.n magmas m ugou> n uuau> N awnw> N uuww> H awnw> nousufiuu< novauguu< amusuauu< novauauu< novauauu< a usaa> o>wunwuz u>uuunom o>auuwuz o>uuwuom o>uuuwuz novauwuu< «>«uauom Ansa.o Asn~.c Aoua.v AHNN.V AenH.e Amen.v o~s.o . as. m4> one.» . as. m¢> «as.» n ma. ma> ma~.a . saw m¢> Hoa.s . m2. m<> "on.oa . N2. ¢¢> sum ofla mnh :Hm «H» -h son.) . has ons.~ . was con.u . was na~.~ . 32‘ sas.- . mas oco.H . «as ovauauua Huauouut :c 1551. .Amommcucoumm :« mofiuaafinwaawm. m: manmwum> ucwamuomama acumen uOu Hmuo: ucwfiouamom: ecu uOm mumaaumm qmmmHAuu.m ousmwm auuflaamunav< auuaupanuaounn< huauuanuhsx huaaanfiuouuuman peonmuush huumcoucn poo: xuu>auo< Com; 8%; 82¢ 33$ 33¢ 88; $23 283 HHS.H . mNu m<> nom.o . .Nu m¢>nna.a . ma. m¢>mas.s . Na. m<> kam.n . "Nu m¢> cso.~ . o~u m¢>sna.~ . mam m¢>ssn.~ . as. m<> mNh :Nh mwh NNh ~Nh omh w?» o~% nho.o . m~a mnn.n . :Na aka.m . mus knm.H . «as Hmo.n . Has nno.flu . °~2 «nm.~ . mas ooo.a . was usuauuomaaa ucmucu m: 152 .Ammmmnucmumm CH mwfluHHwQMflHmmv a: mabmwum> ucmecoufl>cm meow may MOM H0002 acmEmMSmmmz map How mumEflumm qmmqull.m musmflm whoa oncoammu Hmsmfi> mac: :H mumxmuoumu mufimcmp cavam mo umnasz woman umaao mo Honasz vaonmmsom Amso.c Awms.v ANHO.V Amsm.v ssm.mm n mam m¢> “so.o u mmw ma> aoh.o n h32> cho.cs u 8N8_m¢> mm> mm» nmh mm% n has ooo.a u was unmacoufi>cm $803 w: 1553 .Anomozusoucm :« naauaaaauasomc Bc cannauo> coauouuoucn uauucnuuoauoz can you “ova: ucoauusauuz on» you cassava» anxmaauu.oH ouamam nowua>uuuu ucwxnuouwu :« anon: weanum acouunuv uouwaauo> adage cu adage snag «o unsoa< snags: ou unaoauox adage use.» aunuox oases gonna: Aaoo.v AONH.V Anun.v Ao~c.v Aube.v Amoo.v mon.~ I mmu ¢<> osc.o I saw m<> nom.~ I mmu m¢> Hoo.u I umu z¢> can.~ I ”mu n¢> mnu.n I on» m<> mnh :mh mmh th umh on» sd~.~- . was coo.H . :na o~5.as . ”as mon.~ . «ma ~c~.oa . ans oco.H . ans couuuwuuucn ucaucunuonuo: 5: 154 models, one of the coefficients relating the theoretical variable 52 to its indicators, the observed variables x2 (family occupational status), x3 (mother's education level), and x4 (family income), was set equal to one. Arbitrarily the coefficient for the first indicator was chosen to be the "reference indicator" in each case. Without this restriction, each measurement model would be underidentified and an infinite set of consistent estimates would be pos- sible, with no means of choosing among them. As a result of this restriction, each measurement model has a unique solution and the estimated coefficients are calculated rela- tive to the restricted parameter (Fink, 1980). For the analysis presented here, the variance- covariance matrix for the 39 observed variables was anal- yzed. Therefore the socioeconomic status measurement model in Figure 4 has the following interpretation: Var x2 = 1.000 Var g, + 425.298 Var x3 = .104 Var £2 + 3.367 Var x4 = .441 Var £2 + 89.525 where Var = variance. That is, 52 is created by factor analyzing X2, x3, and x4. The coefficients indicate how the variance of £2 would be used to recreate the observed variables. The Gs, or error of measurement terms, indicate how much of the variance of the observed variables is not accounted for or recreated by the decomposition of £2. (The as for the 155 y-variables are interpreted similarly.) The size of the error terms depend on both the measurement scales that were used and upon the amount of common underlying variance (degree of intercorrelation) of the indicator variables. A large measurement error does not indicate that the measurement of a particular indicator variable was poorly done, only that the measurement of that indicator as it relates to the theoretical variable in this model is poor. Thus, if the indicator variables are not highly intercorre- lated, the measurement errors will be large. For com- parison purposes, the reliabilities of the measurements (again as related to the theoretical variables) were calculated by using the following formula and were written in parentheses under each measurement error term: Error variance Total variance Reliability = 1 - One way to evaluate the significance of an estimate is to compare it to its standard error. This comparison is provided in Table 24. If an estimate is at least twice as large as its standard error, it is significant at the p < .05 level (Hanushek & Jackson, 1977; Wheaton, Muthén, Alwin & Summers, 1977). Most of the ratios in Table 24 are considerably less than 2.0 indicating that few of the esti- mates for the measurement models are significant. Another important aspect of the measurement models is the interpretation of the directionality of the created 156 Table 24 Standard Errors for the Estimates of the Parameters of the Measurement Models a Standard Error/ Parameter Estimate Standard Error Estimateb 12 1.528 .408 3.75 A“ 8.363 519.975 0.02 15 -12.208 851.734 -0.01 15 36.549 2493.883 0.01 17 23.784 1625.900 0.01 13 -27.214 1811.687 -0.02 110 1.339 .590 2.27 111 -1.067 .501 -2.13 113 -.494 .169 -.72 11. 2.293 .407 5.63 115 -.505 .184 -2.74 115 '2.430 .413 5.88 117 -.509 .168 -3.03 119 2.854 2.318 1.23 120 -l.033 1.235 -.84 121 3.051 2.561 1.19 122 1.837 1.875 .98 123 3.979 3.162 1.26 12. 3.338 2.592 1.29 125 6.073 4.926 1.23 127 -.036 .040 -0.90 123 -.069 .016 4.31 129 .558 .293 1.90 131 10.702 220.419 0.05 132 1.305 26.161 0.05 133 11.726 237.815 0.05 13. 1.806 36.364 0.05 135 -1.217 17.391 -0.07 137 .104 .021 4.95 138 .441 .084 5.25 a These values for the standard errors were taken from an approximate solution since none were available for the final solution. be greater than 12.0, p < .050 157 or unobserved variable. By assigning a regression coeffi- cient of 1.0 to the first indicator, one forces the unobserved variable to vary in the same direction as the reference indicator. In Figure 4, increasing occupational status is indicative of increasing SES, thus the positive coefficient means that large values of 52 indicate high SES. On the other hand, in the lower half of Figure 4, prenatal class participation, as the variable is coded (1 = yes, 2 = no) is negatively related to quality prenatal care. Since the coefficient is positive, large values of n1 indicate poor quality prenatal care and small values indicate a higher quality of prenatal care. Similarly, the month that prenatal care began is negatively related to quality prenatal care. However, the positive coefficient supports the inverted interpretation of n1 given above. In all the other measurement models, except for n5 (Figure 8), the reference indicator is associated with posi- tive aspects of the unobserved or theoretical variable. For n5, infant temperament, high values of the reference indicator, activity, are indicative of a "difficult" infant, a possibly negative temperamental characteristic. Thus, high values of n5 are indicative of difficult characteristics and low values, of more flexible characteristics. The Theoretical Model In Figures 11 and 12, the estimates of the param- eters of the theoretical model are presented. The gammas (Y) and betas (B) are regression coefficients which 1558 .muunuo: uuuao can museums uo nousowuomxu acaunouvafico can mcwuuonvuagu «nu mcaummaoo Haw «one: Hoowumuouna an» haw nouoaquum Amzmnquu.aa unavam use. u an ace. I an oooo. I m abox N mOHv.OVHH I .I I.c~ I mp..ANMMIIIIIIIMI n: no.- u «an- . a . o: cum. r as 1?}b\\\\\‘ - . cg- 260' 2a.: . . ooe.- . was- “an. n 22a- ows.~ . as H n.c~- . ma- . 89.: . 5 2. Allis): «06.: I «u coo.o I ~u 159 .mumnuoz Hmpao 0cm mmmcmme mo mwocmwummxm mcwummuoawno can mcflumwncafino map mcfiummeou How ammo: Hmowuwnomne map How mmumEHumm AmmmHAIICOquHom omwflpumocmumnu.ma musmwm .Iuofl I m: ”mg” m m: use.) u «an- 38.3 n m; s A .4: S o “ mew. u e 160 indicate the relationships among the eXOgenous (g) and endogenous (n) theoretical variables. The size of a regression coefficient is sensitive both to the measure- ment scales of each of the two variables it relates, to the strength of the relationship between them, to the effects of other variables in that equation. and t0 the structure Of the whole model. For these reasons, one cannot directly compare the coefficients in Figure 11 to one another. In order to compare regression coefficients, the standardized solution to the theoretical model, which is an accurate description of the relationships in the model when the model is not scale-free (Wheaton g_ _1., 1977), is presented in Figure 12. (See also Appendix F.) In the standardized solution, the variance of each theoretical variable is 1.0. In Figure 11, one of the first points to notice is the opposing direction of the effects of the two exogenous variables, £1 (maternal age) and £2. £1 and 52 were highly positively correlated as indicated in Figure 11 where ¢ represents the covariance of £1 and £2 and in Figure 12 where ¢ represents the correlation of £1 and £2. The gammas (7) relating the exogenous variables to the endogenous vari- ables indicate the independent effects of maternal age and of SES; that is, the effect of each exogenous variable after the effect of the other has been partialled out. Thus, with the effect of SES accounted for, maternal age had a positive impact on prenatal care (ml) and a nega- tive impact on both neonatal status variables (n2 and n3), 161 on maternal attitudes (nu) and on mother-infant interaction (n7). Since high scores on the prenatal care variable (n1) are indicative of poor prenatal care, the positive relation- ship between E1 and n1 shows that given equal SES, younger mothers had better prenatal care than older mothers. In fact, all the coefficients for 51 indicate that young mothers or their offspring did better than older mothers or their offspring on each of the endogenous variables. With the effects of maternal age partialled out, SES had a negative impact upon prenatal care and a positive impact upon all the other endogenous variables it affected. In other words, high SES mothers or their infants did better than low SES mothers or infants on each of the predictors of child development incorporated in the model. The impact of SES was consistently larger than that of maternal age, or indeed, of any other variable (Figure 12). According to the estimates in Figure 11, prenatal care (n1) had a positive impact upon both neonatal status variables (n2 and mg) and upon maternal attitudes (04)- Because of the inverted interpretation of n1 these relation- ships mean that, when combined with the other predictors in the model, poor prenatal care (high values of n1) con- tributes to good neonatal status and to positive maternal attitudes. Brazelton neonatal status (n2) had a negative impact on maternal attitudes (nu) and a positive impact on infant temperament (n5). That is, in combination with the 162 other predictors, the high Brazelton status scores con- tributed to lower maternal attitude scores and to higher (more difficult) temperament scores. Neonatal physical status (n3) had negative impacts upon the Brazelton neonatal status variable (n2), maternal attitudes (n.) and infant temperament (n5). High levels of physical status thus contributed to lower levels of Brazel- ton neonatal (nz) status, lower or negative levels of maternal attitudes (nu), and to lower (more positive) infant temperament scores (n5), when used as a predictor with other variables as indicated in the model. In combination with other predictors, high levels of infant temperament (us), which are indicative of a diffi- cult child, had a negative impact upon both maternal atti- tudes and mother-infant interaction. Similarly, high levels of the home environment variable (ns--associated with more space and toys) made a negative contribution to the level of mother-infant interaction. Finally, the level of mother-infant interaction made a positive contribution to maternal attitudes while maternal attitudes made a positive contribution to mother-infant interaction. The :3 are the errors of prediction or the amount of the variance of the endogenous variables that are not linearly predicted by the variables associated with it. Comparing the values of the cs with the variances of the endogenous variables (see Eta-Eta matrix, Appendix F, P. 294), we see that two of the errors of prediction (c. 163 and t7) have values larger than the variance of the vari- able with which they are associated. This may indicate that the assumption that the error terms for the indicators of ng and n7 are uncorrelated is incorrect. To evaluate the significance of the parameters of the theoretical model, the standard errors of the estimates are presented in Table 25. As the ratio of the estimates to their standard errors shows, few of the parameters are sta- tistically significant (if ratio 3 2.0, p < .05). Evaluation of the Research Model To evaluate the significance of the complete research model, a x2 goodness-of-fit test was performed (Figure 11). This test is performed by comparing the original variance- covariance matrix of the 39 observed variables (Appendix F, pp. 279-280) with the estimated variance-covariance matrix (see Sigma matrix, Appendix F, pp. 291-292) which is based upon the paths or specifications of the research model. The goodness-of-fit test determines how well the original matrix is approximated by the estimated matrix (see S-Sigma matrix, Appendix F, pp. 292-294 for the matrix of differ- ences between the estimated and original matrices). In the goodness-of-fit test, the null hypothesis, H is that the model as specified is the true model of CI the relationships among the variables or that the model accurately reproduces the variance-covariance matrix that was analyzed. The alternative hypothesis, H1, is that 164 Table 25 Standard Errors for the Estimates of the Parameters of the Theoretical Model a Standard Error/ Parameter Estimate Standard Error Estimateb Y1 .029 .044 .65 Y2 -.024 .012 -2.00 Y3 -.287 101.071 -0.00 y. .408 148.309 0.00 Y5 -.097 22.947 -0.00 ya .114 15.020 0.00 Y7 -12.841 65,065.961 -0.00 Ye 21.852 97,667.165 0.00 Y9 .130 .037 3.51 Y1o -.316 6.441 -0.05 Y11 .151 3.117 0.05 '81 19.888 7,069.441 0.00 '82 -5.092 2,054.079 -0.00 -83 5.310 529.987 0.01 -8. 1,113.939 4,674,356.085 0.00 -85 -26.381 323,656.130 -0.00 -86 -333.135 1,379,750.219 -0.00 -87 -2.926 5.041 -0.58 '83 27,294 575.436 0.05 '89 5.484 393.271 0.01 -810 -.254 .085 -2.99 "811 .587 15.374 0.04 '812 -.033 19.720 -0.00 -813 -.400 11.259 -0.04 aThese values for the standard error were taken from an approximate solution since none were available from the final solution. be greater than or equal to $2.0, p < .05 (Duncan, 1975; Wheaton gt al., 1977). 165 the estimated matrix is any positive definite matrix (Joreskdg & Serbom, 1978). Contrary to most significance tests, the goal here is not to reject Ho' The probability level that is given in Figure 11 "is defined as the probability of getting a x2 value larger than the value actually obtained given that the hypothesized model is true" (Joreskég & Sorbom, 1978, p. 57). Thus large significance levels for the xz-test values approaching 1.0, under the assumptions of multinormality and in large samples, are indicative of models that fit the data well. In Figure 11, the results of the goodness-of-fit test were x279 = 1140.4103, p = .0000. If this test were the only means of evaluating the model, we would have to conclude that the model as Specified in Figure 3 (and Figures 4 through 12) does not fit the data well. Furthermore, an error message suggests that one parameter, and therefore perhaps the whole model, is under-identified. If this is the case, the test results are irrelevant. Discussion of the Estimated Model Evaluation of the Model According to goodness-of-fit test, the model does not fit the data well. However, the goodness-of-fit test is sen- sitive to sample size and.must be interpreted cautiously (Joreskag, 1974, p. 4; 1978, p. 14; Wheaton g5 31., 1977, p. 99). Instead of using the x2 value to accept or reject the model as a whole, the results of the goodness-of-fit test can be used to compare the fit of various models to the data: 166 . . . if a value of x2 is obtained, which is large compared to the number of degrees of freedom, the fit may be examined by an inspection of the residuals, i.e., the discrepancies between observed and repro- duced values. Often the results of an analysis, an inspection of residuals or other considerations will suggest ways to relax the model somewhat by intro- ducing more parameters. The new model usually yields a smaller x2. A large drop in x2, compared to the difference in the degrees of freedom, indicates that the changes made in the model represent a real improve- ment (JoreskEg, 1974, p. 4; see also Joreskog, 1978; Wheaton £3 31., 1977). An inspection of the residuals matrix (the S-Sigma matrix, Appendix F, pp. 292-294) suggests that the model does a fairly good job of reproducing the original variance- covariance matrix. Most of the residuals are quite small, especially in comparison to the size of the corresponding variance or covariance terms. Furthermore, the xz/df ratio of 1.679 suggests that the present model is a reasonable representation of the data though it might be improved by the addition of theoretical variables and/or paths between variables (cf.: Wheaton EE 31., 1977, p. 99). The present model is highly overidentified, that is, the number of the degrees of freedom is large. This degree of overidentification is the result of numerous assumptions that were made or restrictions that were placed upon the model (e.g., the error terms of the measurement models are uncorrelated and the error of prediction terms are uncorre- lated). Because of the status of theory in the social sciences, there were no theoretical guides to the placement of these restrictions (cf.: Land & Felson, 1978. p- 288). some of which are quite strong assumptions. Therefore, 167 after inspecting the residual matrix, there are numerous submodels that can be created by relaxing some of the over- identifying assumptions. These submodels can then be com- pared to the original model in the manner suggested by Joreskég above. Such a strategy should be applied to the present model according to the guidelines provided in the following section. However, as the inspection of the standard errors above indicated, most of the parameter estimates of the model are not significant. While the estimates are con- sistent (a large sample property), with the sample size utilized many standard errors are so large that often the hypothesis that the estimate is zero cannot be rejected. By this criteria one cannot have faith that the estimates obtained are the true values of the parameters, reducing the usefulness of the model for making predictions with any degree of precision. Suggestions for alternative models. A close inspection of the information in the LISREL analysis (Appendix F) provides several possible explanations for the large standard errors reported above and provides indi- cations of ways to improve the model. First, as reported above, £1 (maternal age) and £2 (SES) were highly corre- lated. Highly correlated exogenous variables result in multicollinearity, a situation in which there is only limited information to separate out the independent effects 168 of each variable. Multicollinearity reduces the precision (increases the standard errors) of the estimates. Further- more, because the estimates are based upon so little infor- mation (only the variance that is left after the joint effects are removed), they become very sensitive to slight modifications of the model specifications. Thus, in cases of multicollinearity, the estimates may prove to be arti- facts of the sample used and the model specification (Hanushek & Jackson, 1977). One possible solution to multi- collinearity is to create a single theoretical variable out of £1 and 52. Second, the values of the variance-covariance matrix for the theoretical endogenous variables (see Eta-Eta matrix, Appendix F) show that the variances for n2, n3, ns, and n, are quite small, with the variance of mg being ".000." This indicates that the sets of indicators for these vari- ables were not good choices in that the amount of common underlying variance is so small. The intercorrelations for each set of indicators supports this conclusion (see Appendix E). This situation may result from the lack of precision of the scales used (not precise enough to show variability) or from the use of orthogonal (perfectly dis- tinct) factors as indicators. It also may result from the use of small scaling metrics, e.g., scales with maximum values between 0 and l. The message about insufficient arithmetic precision on the LISREL results appears to sup- port the latter possibility (Appendix F). 169 For instance, each of the Brazelton summary scores is an indicator of neonatal well being. However, as the correlation matrix shows (Appendix E), the intercorrelations among the indicators are sometimes quite small suggesting that these variables are not indicators of a single con- struct. The extremely small measurement reliabilities for these indicators add strength to this contention. Similar conclusions can be reached for the indicators of n3, n., and n5. Two possible solutions to the problem of weak indi- cators are to (1) create additional unobserved variables or (2) reduce the number of indicators used for the unob- served variables (Fink & Mabee, 1978; Duncan, 1975; Costner & Schoenberg, 1973). The resultant sets of indicators should be characterized by relatively strong intercorrelations and by theoretical relevance. The choice of indicators for maternal attitudes (nu) poses another type of problem. Since these indicators are the results of repeated measurements, one would expect to find the measurement errors to not be truly random, but to reflect factors left out of the model which may be corre- lated, a possibility that was not reflected in the model. This possible misspecification is probably the cause of the inflated error of prediction for maternal attitudes (Cu). The model should be modified to allow the appr0priate corre- lations among the error terms. The standard error is also influenced by the quality of the measurement instruments. If the instruments do not 170 have a high level of validity, we cannot be certain that we are measuring the construct of interest; that is, the rela- tionship between that which is measured and the theoretical concept may be weak. If the instruments do not have a high level of reliability, a certain amount of the variation in the data will be due simply to the unreliability of the instruments that were used. Low reliability leads to larger measurement errors and to larger estimation errors. As mentioned above, the fact that the estimates are consistent is a large sample property and the x2 test is a large sample test. One would expect an improvement in the estimates of the model and in the results of the goodness- of-fit test simply by increasing the sample size. An increase in the sample size would also reduce the size of the standard errors. Thus one alternative is to estimate the current model with data from much larger samples. Multiple indicators were used throughout the model to improve the reliability of the measurements. However, in some cases (Brazelton items, MITS) the intercorrelations of the indicators were quite low, thus the validity of the resulting construct was probably quite low. Though the instruments used were the most reliable ones available, or were redesigned to improve reliability, none of the instru- ments was perfectly reliable. Obviously, some portion of the standard error is due to the reliability and validity of the measurements used. 171 If a model is "empirically underidentified," large standard errors will result (Fink, 1980). Though the number of degrees of freedom of the model (calculated by subtract- ing the number of estimates to be made from the number of elements in the data matrix) is large and nonnegative, this is only a necessary condition for being overidentified. However, "empirical" identification depends, for example, upon various covariances not being equal to zero (Fink, 1980). A review of the variance-covariance matrix (Appendix F, pp. 279-280) indicates that some covariances are quite small, possibly contributing to the size of certain standard error terms. Finally, the instances of possible heteroskedasticity mentioned above imply that the assumption of a multivariate normal distribution, used to estimate the parameter estimates and used in the x2 test for goodness of fit, was violated. If some of the relationships in the model are not linear (one source of heteroskedasticity), the assumption for the estimates is violated and the model is misspecified. By choosing appropriate transformations, the specification problem can be corrected (e.g., the relationships may be made linear) and such transformations often produce normal dis- tributions (cf.: Fink, 1980, pp. 134-135). Where hetero- skedasticity exists in the model above, appropriate trans- formations could be found by using residual analysis of ordinary least squares regressions. 172 In summary, the research model does not completely account for all the variance that appears in the data set. It is possible that adding theoretical variables or causal linkages between variables would strengthen the model. The large standard errors for the estimated parameters reduce one's confidence in the research model. These large standard errors may be reduced by using the observed exo- genous variables as indicators for a single unobserved exo- genous variable. Improved measurements would help to reduce the standard errors also. The Relationship of the Estimated Model to Theory Given the results of the goodness-of-fit test and the large standard errors, it is difficult to provide an interpretation of the results with any degree of confidence. However, the discussion that follows provides an interpre- tation as though the estimated model were the true descrip- tion of the theoretical linkages represented. Future research may provide evidence in support of, or in contrast to, that presented here. One of the more interesting aspects of the results presented above is the finding that maternal age and SES have opposing effects upon the endogenous variables in the model. The fact that maternal age and SES have causal effects in opposite directions challenges much of what has been reported about teenage parenting. Even an examination of the correlation tables (Appendix E) suggests that the 173 causal effects of these two variables would be similar, i.e., the size and direction of the correlation of each with the other variables are quite similar. It is the high positive correlation between maternal age and SES that creates this illusiOn and confounds the results of studies dealing with teenage childbearing. It appears that the negative effects of adolescent childbearing upon child development are not due to maternal age per 53. Instead, the negative effects are due to the correlates of early childbearing: truncated education, limited job opportunities, and reduced earning power for the mother and her male companion (cf.: Baldwin & Cain, 1980). Teenage parents have fewer material and nonmaterial resources to invest in their child's development. However, within the age range studied and given equal SES levels, the more optimal status of the young mother's reproductive system (at least for first births), her enthusiasm, and her energy tend to provide more optimal conditions for child development than that provided by older mothers. It is also important to consider the relative size of the contribution of age and SES to the various aspects of childbearing and Childrearing (Figure 12). In every case, the influence of SES is several times the size of the influ- ence of age. This finding suggests that the positive aspects of early motherhood are probably more than offset by the negative impact of the low SES that is associated with early motherhood. 174 The results discussed above appear to conflict with those of Oppel and Royston (1971), who found developmental differences by maternal age after matching their adolescent and older mothers on SES, birth weight of the child, maternal parity, and maternal race. The current study sug- gests that SES is the major factor for explaining the devel- opmental differences between the children of teenagers and older mothers. There are a number of possible explanations for this conflict: 1. Matching is not a satisfactory way of controlling for initial differences (Campbell & Stanley, 1963); Depending upon the rules used for matching, Oppel and Royston may have selected very special sub- groups of both age groups. If matching is done on broad categories its effectiveness is lost. If it is done on a very restrictive basis for all four variables, the samples selected may be unique; By matching on both SES and birth weight, Oppel and Royston were selecting a subgroup of adolescents who did not experience the advantage reported above (i.e., given equal SES, teenage mothers had larger children) whether or not this experience was a typical one; Oppel and Royston may have used a completely dif- ferent definition of SES than that used in the present study, a common situation in social science research; 175 5. The Oppel and Royston study had a larger, and therefore, potentially more representative, sample than the present study; 6. The Oppel and Royston study was longitudinal and its dependent variables were measures of develop- mental status while the present study was short term and its "dependent" variables were predictors of developmental status. In the present study, one can only hypothesize about the relative develop- mental status of the two groups of children at a later date. For at least these reasons, the results of the two studies may not be comparable. Several of the relationships between the endogenous variables in the model were quite unusual and appeared to challenge both theory and logic. For instance, poor pre- natal care (n1) had a positive impact on the neonatal status variables (n; and n3). This appears to directly conflict with the findings of Dwyer (1974) and others who have reported the onset of prenatal care to be predictive of improved neonatal status. The relationship between poor prenatal care and more positive maternal attitudes also appears to conflict with Leifer's contention that prepara- tion for the birth improves maternal attitudes (1977). The difference between the findings of Dwyer and Leifer and those of the present study probably stem from the types of analyses used. Both authors discussed their 176 findings in a bivariate sense. The results above are much more complex and, therefore, more difficult to interpret. For instance, consider only the relationship between maternal age (51), SES (£2), prenatal care (53), and Brazel- ton neonatal status (01)- As reported above, given equal SES, younger mothers had more positive prenatal care and their infants had better neonatal status than older mothers and their children. Also, given equal maternal ages, higher SES mothers had more positive prenatal care and their infants had better neonatal status than lower SES mothers and their children. However, the influence of maternal age and SES on neonatal status should not be considered without adding the impact of prenatal care; i.e., the estimates were generated by considering the influence of all three variables simultaneously (except for the purposes of this illustration we would also have to add the impact of physical status to the equation). Thus, if all the influences upon Brazelton neonatal status are considered, the distinct contribution of prenatal care is negative. The positive impact of prenatal care upon neonatal status reported by others may then be explained by the direct effects of maternal age and SES upon neonatal status. That is, those mothers who have high quality pre- natal care may have healthier babies (as measured by the Brazelton scales) but the enhanced neonatal status is due to their SES and age, not to the quality of their prenatal care. However, there is no apparent theoretical or logical 177 reason for prenatal care to reduce the level of neonatal status as reported here. Given this relationship, the rela- tive size of the effect of prenatal care upon neonatal status and maternal attitudes is also puzzling (Figure 12), especially when there was no apparent relationship between prenatal care and the delivery experiences of the mothers. Other apparent anomalies in the results above have similarly complex explanations that appear to contradict current theory. For example, the quality of the home envi- ronment as measured in this study was expected to have a positive impact upon mother-infant interaction. However, when home environment is considered in conjunction with maternal attitudes, age, SES, and infant temperament, home environment had a negative impact upon interaction. This result may indicate that the positive aspects of home envi- ronment reported by others are explained by the direct impact of SES (the only variable in the model shown as a cause of home environment). With the impact of SES on mother-infant interaction accounted for by the direct link- age of those two variables, the remaining impact of home environment (space, number of toys) is negative. Again the exact reason for the direction of the remaining impact of home environment is not readily apparent. It is also pos- sible that home environment has a direct impact upon child development and little or no indirect impact upon child development via mother-infant interaction as specified by the model. In this event, the model is misspecified. 178 Simply stated, the apparent contradictions between theory and the results of the LISREL analysis of the model reported above may be explained by the differences in the complexity of previous analyses and the current one. Bivariate analyses do not fully represent causal complex- ities. Those bivariate analyses that attempt to control for the effects of one or more variables are still dis- torting the richness of the transactions that are often the causes of behavior. As an example, consider the various impacts of SES in the current research model. In every case in which SES is a causal variable it has a larger impact than any other variable in the equation (Figure 12). By controlling for the impact of SES, by whatever means, researchers are strongly distorting the causal realities of the phenomenon under consideration. When one considers the simultaneous impacts of several related variables upon another, the dis- tortion implied by bivariate research becomes even more apparent. The transactions among the variables may produce results that are grossly different from those that occur when each variable is treated singly. Simply stated, when analyzing a system, the results-very much depend upon whether one examines the whole system simultaneously or the pieces separately. To date, most research in the social sciences has paid lip service to examining systems while continuing to analyze the pieces of the system separately. 179 Summary Numerous differences were found between the child- bearing and Childrearing experiences of teenagers and older mothers. Older mothers generally provided a more Optimal milieu for child development than did the teenage mothers. However, in many ways the two groups of mothers and infants were quite similar and the teenagers were found to have easier deliveries than the older mothers. The teenage mothers in this study did not exhibit the negative or ambivalent attitudes toward motherhood that had been reported in the literature to be characteristic of them. However, the results of this study do support the claim that the children of teen- age mothers are more likely than the children of older mothers to have deve10pmental problems. In the analysis of the system of childbearing and Childrearing, it was found that SES was the most pervasive and strongest influence upon the predictors of infant devel- opment. When SES was controlled for, young mothers or their infants were rated higher on the predictors of child devel- opment than older mothers or their infants. It appears, then, that maternal age, pg; 53, is not the critical vari- able in the finding that teenage childbearing has a negative impact upon infant deve10pment. Rather, the correlates of teenage childbearing that result in reduced SES may be the critical factors that explain adverse development among the infants of young mothers. 180 The results of the analysis of the system of child- bearing and childrearing must be interpreted cautiously. There were a number of indications that the model may not be a correct or complete representation of the phenomena being studied, The use of larger samples or the testing of alter- native models is necessary to determine the value of the model used. Several recommendations for respecifying the model were made. CHAPTER V SUMMARY AND CONCLUSIONS Summary Over one-half million teenagers give birth each year despite efforts by educators, family planning agencies, and parents to reduce these numbers. Researchers have consis- tently reported that teenage childbearing is strongly associ- ated with truncated education, low SES levels, high proba- bility of needing financial support from public assistance programs, high completed fertility and extremely high rates of marital instability. Furthermore, longitudinal studies have reported that the children of teenage mothers are characterized by a number of social-emotional, behavioral, and intellectual deficits by the time they are of school age. The present study followed 62 primiparous mothers, aged 15-32, from the last trimester of their pregnancies through the third month postpartum. These mothers were interviewed in their homes, the medical records of their deliveries were reviewed, their infants were tested shortly after birth, the mothers and their infants were observed monthly for three months, the home environments these mothers provided for their children were evaluated, and the 181 182 mothers completed scales assessing their infants' tempera- ments and their own attitudes toward their infants. A comparison of those mothers below 20 years of age (n = 14) with those over 20 (n = 48) revealed a number of differences. The teenage mothers were living in relatively low SES conditions and almost half of them were on public assistance (cf.: David, 1972). Only half of the teenagers maintained a relationship with the infant's father and half of the teenage mothers continued to live with their parents or another relative. Though the teenage mothers began pre- natal training later and were less likely to participate in birth preparation classes than were the older mothers (cf.: Dwyer, 1974), the younger mothers had easier deliveries (cf.: Mednick _£._l-r 1979). As has been reported in numerous other studies, the infants of teenage mothers in this study tended to have slightly shorter gestations and to be slightly smaller at birth than the infants of older mothers (see Table 2, p. 21). There were no differences between the Apgar scores of the two groups of infants (cf.: Sandler, 1979) and only one of seven Brazelton Scale summary scores (motor perform- ance) differentiated between the two groups, with the infants of older mothers doing better (cf.: Lester, 1978; Sandler, 1979). In contrast to the reports of Crumidy (1966) and Mercer (1980), no differences were found between the maternal attitudes of the two groups of mothers. No differences were 183 found between the temperament ratings of the infants either. The homes of teenage mothers were more crowded than those of older mothers and were less likely to offer a quiet place where the infant could be put during naps to escape from stimulation. Teenagers' homes also provided fewer nonhuman sources of audio-visual stimulation than those of older mothers. The differences in the home environments were strongly related to the SES differences between the mothers (see Wachs 35 al., 1971; Tulkin & Kagan, 1972). There were numerous similarities in the mother- infant interaction patterns of the two groups. However, as reported by Sandler (1979), the older mothers spent more time talking to their infants and talking to their infants during mutual gaze than did the younger mothers. The older mothers were also more responsive to their infants' dis- tress signals than the younger mothers, supporting the con- tention of de Lissovoy (1973, 1975), Williams (1974), and Mercer (1980) that teenage mothers are less sensitive than older mothers to an infant's signals. There were no sig- nificant differences between the behaviors of the two groups of infants. As suggested by the research questions posed in Chapter I and by the literature review, the background characteristics (primarily SES), the home environments, and the caretaking patterns of teenage mothers differed signifi- cantly from those of older mothers. These differences 184 potentially are linked to the developmental deficits for the children of teenage mothers reported by Oppel and Royston (1971) and Hardy gt gt. (1978). However, none of the sus- pected differences in maternal attitudes, neonatal status, or infant temperament were found, presumably removing these variables from consideration as causes of the cited deve10p- mental differences. Furthermore, since the births of teen- agers were easier than those of older mothers, there is no special reason to suspect the birth experience of teenagers to negatively influence maternal attitudes or to be related, directly or indirectly, to the cited deve10pmental differ- ences. On the contrary, the easier births may have played a role in enhancing the young mothers' attitudes toward their infants and parenting in general. It needs to be emphasized that the childbearing and childrearing experiences of the teenagers in this study were much more positive than one would expect after reviewing the literature on teenage parenting. There were no statistically significant differences between the two groups of mothers on almost 75% of the bivariate tests that were performed. Furthermore, the teenagers in this study had easier deliveries than the older mothers. It is far too easy to concentrate on the negative aspects of teenage parenting while overlooking the positive aspects. Though the caretaking situations of teenage mothers may indeed be less optimal than those pro- vided by older mothers, it is apparent that there are numerous important similarities between these two groups of mothers. 185 The data from the subjects in this study also were used to evaluate a multivariate model relating several theoretically crucial components of child development to maternal age, SES, and each other. Using LISREL, estimates were made for the causal linkages between the variables in the models and for evaluating the fit of the estimated model to the original data. However, there were numerous weak- nesses in the model and it was determined that the estimated model probably does not adequately represent the data. The model could be improved by adding theoretically important variables, choosing more homogenous variables as indicators of the theoretical variables, and by developing better measurement instruments and instruments with more compatible scales. Therefore the results of this analysis must be used judiciously. On the other hand, the estimated model indicated that maternal age and SES each make independent impacts upon variables that are important to child development. In general, SES is positively related to the intervening variables in child deve10pment used in this study while maternal age is negatively associated with these same vari- ables. That is, given equal SES, younger mothers do better than older mothers. However, SES was shown to have a much stronger impact than maternal age, probably enough to overcome the benefits associated with young motherhood. Furthermore, since teenage motherhood is strongly related 186 to low SES, there is little reason to emphasize the positive contributions of early motherhood. Conclusions Limitations Before discussing the conclusions of the study, it is necessary to outline its limitations, its strengths, and its generalizability. Most of the limitations of the study are provided by the sample studied. The subjects of the study were volunteers, not a random sample, and the number of teenagers in the study is quite small. The data provide no apparent indications that either group of mothers was unique or nonrepresentative, but one cannot overlook the special nature of a volunteer sample. There are two other weaknesses to the study. First, the study was short term and therefore one could only specu- late about the long term effects of the differences found. Secondly, because the study was short term, no logical con- cluding variable, such as IQ or social-emotional status, was measured. It must be left to future research to repli- cate the methodology applied here in a longitudinal study which should produce more conclusive results. A longitu- dinal study was needed also to determine whether the findings of Oppel and Royston (1971) and Hardy gt_gt. (1978) apply given today's medical and social programs. 187 Strengths There are several strong points which, despite short- comings, make the findings of the present study valuable. Research in adolescent parenting has been characterized by an almost complete lack of comparison groups. When com- parison groups were used, they were often extremely unusual groups. In the present study, teenage mothers and their children were compared to a group of women from the norma- tive age range for reproduction. This comparison is much more meaningful than simply describing the behavior of teenage mothers or comparing such behavior to that of other special samples. Furthermore, the sample studied, although mostly a white sample, was widely diversified in all other demographic aspects. A handicap of most research in adoles- cent parenting is the use of all black, urban, and poor samples, or samples that are unique in some aspect. Thus the sample studied is probably more representative of the U.S. population than that used in similar studies. Another strength of the present study is its inter- disciplinary, multivariate and systemic approach. Socio- logical, psychological, ecological, and medical data were gathered and analyzed. These data were organized into a systemic causal model based upon current theories of child development and analyzed as a whole. The method of analysis used was designed for such holistic and systemic analysis. Such an approach provides a more realistic View of the 188 phenomena under consideration and avoids the bias of "multiple- bivariate" research which is common in the field. Implications Both the major strengths and weaknesses of the present study are related to the samples used. For this reason the data gathered are suspected of being representa- tive but no faith can be put into any generalization that goes beyond the present sample. However, the findings of the present study should provide a firm foundation for further, hopefully longitudinal, research, which seeks answers to similar problems. For the groups studied, several critical differ- ences were found between the childbearing and childrearing experiences of teenagers and older mothers, differences that were theoretically related to developmental differ- ences for the children involved. Thus, the results of this study provide support for earlier researchers who reported developmental deficits for the children of teenage mothers (Oppel & Royston, 1971; Hardy gt gt., 1978). However, in the present study specific causal mechanisms were suggested for the developmental differences that were reported. SES was implicated as the major source of differences between teenagers and older mothers with its influence distributed via neonatal status, maternal attitudes, home environments, and mother-infant interaction. If the findings of this study are supported in future studies, the identification 189 of the specific causal mechanisms will be an important link in developing intervention programs to work with teenage mothers and their infants or to provide direction to existing programs. Furthermore, if the independent contributions of SES and maternal age to infant development are as reported above, there are direct implications for programs concerned with adolescent parenting. Programs designed to prevent adolescent pregnancy would need to publicize the impact of early parenthood on SES and the impact of SES on child development. Secondly, programs working with young mothers should be encouraged to use the strengths of young mothers (apparently energy and enthusiasm) to overcome some of the weaknesses discussed above. Adolescent parenting is an important social phenom- enon. If the children of teenage parents, approximately one-fifth of all children in the U.S., will be handicapped in life by deve10pmental problems, more resources need to be focused upon this phenomenon. However, before more funds are poured into ever more complex intervention pro- grams, more research such as the present study, using larger samples and longitudinal designs, are needed to determine the appropriate content and procedures for existing pro- grams. The need for confirmatory research before accept- ing or applying the findings of the present study cannot be overemphasized. 190 Finally, one of the more intriguing implications of the present study is the dire need for more multivariate, systemic analysis in child development. A number of rela- tionships reported for the present study were difficult to explain and contrary to previous findings from studies that were primarily bivariate. Were these unusual findings simply artifacts of the model and sample used or do they represent our ignorance about the transactions that occur among vari- ables in multivariate causal models? Only by applying both bivariate and systemic analyses to our research data can we begin to answer this question. When we begin to understand the complexities of the transactions among the variables used in child development research, our ability to predict developmental outcome will increase tremendously. REFERENCES REFERENCES Ainsworth, M. D. S. Object relations, dependence, and attachment: A theoretical review of the infant- mother relationship. Child Development, 1969, 59, 969-1025. Ainsworth, Mary D. S. The development of infant-mother attachment. In Bettye M. Caldwell and Henry N. Ricciuti (Eds.), Review of child development research, Vol. III. Chicago: University of Chicago Press, 1973, 1-94. Ainsworth, M. D. S. & Bell, S. M. Mother-infant inter- action and the development of competence. In K. S. Connolly & J. S. Bruner (Eds.), The growth of competence. New York: Academic Press, 1973, 97-119. Ainsworth, M. D. 8.; Bell, S. M.; & Stayton, D. J. Infant mother interaction and social development: Social- ization as a product of reciprocal responsiveness to signals. In M. Richards (Ed.), The integration of the child into the social world. Cambridge: Cambridge University, 1974, 99-135. Alan Guttmacher Institute. 11 million teenagers: What can be done about the epidemic of adolescent pregnancies in the United States. New York: Planned Parenthood Federation of America, 1976. Als, H.; Tronick, E.; Adamson, L.; & Brazelton, T. B. The behavior of the full-term but underweight newborn infant. Developmental Medicine and Child Neurology, 1976, it, 590-602. Als, H.; Tronick, E.; Lester, B. M.; & Brazelton, T. B. Specific neonatal measures: The Brazelton Neo- natal Behavioral Assessment Scale. In J. D. Osofsky (Ed.), The handbook of infant development. New York: Wiley, 1979, 185-215. 191 192 A13, H.; Tronick, E.; Lester, B. M.; & Brazelton, T. B. The Brazelton Neonatal Behavioral Assessment Scale. Journal of Abnormal Child Psychology, 1977, §(3), 215-231. Amiel-Tison, C. Neurological evaluation of the maturity of newborn infants. Archives of Disease in Child- hood, 1968, 3;, 89-93. Apgar, V. A. A proposal for a new method of evaluation of the newborn infant. Current Research in Anesthesi- ology and Analgesia, 1953, g3, 260-267. Bacon, L. Early motherhood, accelerated role transition and social pathologies. Social Forces, 1974, gt, 333-341. Bakeman, R. & Brown, J. V. Behavioral dialogues: An approach to the assessment of mother-infant inter- action. Child Development, 1977, gt, 195-203. Baldwin, W. H. Adolescent pregnancy and childbearing-- growing concerns for Americans. Population Bulletin, 1976 (revised 1980), §t(2). Baldwin, W. & Cain, V. S. The children of teenage parents. Family Planning Pergpectives, 1980, 13(1), 34-43. Ballard, W. M. & Gold, E. M. Medical aspects of repro- duction in the adolescent. Clinical Obstetrics and Gynecology, 1971, 13, 338-366. Bavolek, S. J.; Kline, D. F.; McLaughlin, J. A.; & Publicover, P. R. Primary prevention of child abuse: Identification of high risk adolescents. Eau Claire, Wisconsin: Department of Special Education, University of Wisconsin-Eau Claire, 1978. Bayley, N. Comparisons of mental and motor test scores for ages 1-15 months by sex, birth order, race, geographical location, and education of parents. Child Development, 1965, gg, 379-411. Bayley, N. Manual for the Baylgy scales of infant devel- opment. New York: The Psychological Corporation, 1969. Beckwith, L. Personal Communication, 1979. 193 Beckwith, L. Prediction of emotional and social behavior. In J. D. Osofsky (Ed.), The handbook of infant development. New York: Wiley, 1979, 671-706. Beckwith, L. Relationships between attributes of mothers and their infants' I.Q. scores. Child Development, 1971, gt, 1083-1097. Beckwith, L.; Cohen, S. E.; Kopp, C. B.; Parmelee, A. H.; & Marcy, T. G. Care giver interaction and early cognitive development in preterm infants. Child Development, 1976, 21, 579-587. Bell, R. Q. Contributions of human infants to caregiving and social interaction. In M. Lewis & L. A. Rosenblum (Eds.), The effect of the infant on the caretaker. New York: Wiley, 1974, 1-19. Bemis, J.; Diers, E.; & Sharpe, R. The teenage single mother. Child Welfare, 1976, 3(5), 309-318. Benedek, T. The psychosomatic implications of the primary unit: Mother-child. American Journal of Ortho- psychiatry, 1949, it, 642-654. Bennett, S. L. Infant-caretaker interactions. In E. N. Rexford, L. W. Sander & T. Shapiro (Eds.), Infant psychiatry: A new synthesis. New Haven: Yale University Press, 1976, 79-90. Berg, M.; Taylor, B.; Edwards, L. E.; & Hakanson, E. Y. Prenatal care for pregnant adolescents in a public high school. The Journal of School Health, 1979, 49, 32-35. Bibring, G. S. A study of the psychological processes in pregnancy and of the earliest mother-child rela- tionship. Psychoanalytic Study of the Child, 1961, _l__6_' 9-72. Blank, M. The mother's role in infant development: A review. In E. N. Rexford, L. W. Sander & T. Shapiro (Eds.), Infant psychiatry: A new synthesis. New Haven: Yale University Press, 1976, 91-103. Blehar, M. C.; Lieberman, A. F.; & Ainsworth, M. D. 5. Early face-to-face interaction and its relation to later infant-mother attachment. Child Development, 1977, gt, 182-194. Bolton, F. G. Jr. Theppregnant adolescent: Problems of prematuregparenthood. Beverly Hills, California: Sage, 1980. 194 Bonem, H. Measurement of infant temperament. Unpublished Masters Thesis, Michigan State University, East Lansing, Michigan, 1978. Bonem, H. M. & Zucker, R. A. Unpublished Scoring Key. Michigan State University, June 1979. Bowlby, J. The nature of the child's tie to his mother. International Journal of Psycho-Analysis, 1958, t2, 350-373. Brackbill, Y. Obstetrical medication and infant behavior. In J. D. Osofsky (Ed.), The handbook of infant development. New York: Wiley, 1979, 76-125. Bradley, R. H. & Caldwell, B. M. Early home environment and changes in mental test performance in children from 6 to 36 months. Developmental Psychology, 1976, tt(2), 93-97. Brazelton, T. B. Neonatal Behavioral Assessment Scale. Clinics in Developmental Medicine No. 50, Spastics International Medical Publication, Philadelphia: Lippincott, 1973. Brazelton, T. B. The parent-infant attachment. Clinical Obstetrics and Gynecology, 1976, ;g(2), 373-389. Brazelton, T. B.; Als, H.; Tronick, E.; and Lester, B. M. Specific neonatal measures: The Brazelton Neonatal Behavior Assessment Scale. In J. D. Osofsky (Ed.), Handbook of infant development. New York: Wiley, 1979, 185-215. Brazelton, T. B.; Koslowski, B.; & Main, M. The origins of reciprocity: The early mother-infant interaction. In M. Lewis & L. A. Rosenblum (Eds.), The effect of the infant on the caretaker. New York: Academic Press, 1974, 49-76. Brazelton, T. B. & Tryphonopoulou, Y. A. A comparative study of the Greek and U.S. neonates. Unpublished manuscript. Referred to in T. B. Brazelton, Neonatal Behavioral Assessment Scale. Clinics in Developmental Medicine No. 50, Philadelphia: Lippincott, 1973. Broman, S. H.; Nichols, P. L.; & Kennedy, W. A. Preschool 19: Prenatal and early developmental correlates. New York: Wiley, 1975. 195 Broussard, E. R. & Hartner, M. S. S. Maternal perception of the neonate as related to development. Child Psychiatry & Human Development, 1970, t, 16-25. Broussard, E. R. & Hartner, M. S. S. Further considerations regarding maternal perception of the first born. In Jerome Hellmuth (Ed.), Exceptional infant, Vol. 2, Studies in abnormalities. —New York: Bruner7Mazel, Inc., 1971. Buckley, W. Sociology and modern sygtems theory. Englewood Cliffs, N.J.: Prentice Hall, 1967. Burchinal, L. G. Trends and prospects for young marriages in the U.S. Journal of Marriagg and the Family, 1965, tl(2), 243-254. Burstein, I.; Kinch, R. A.; & Stern, L. Anxiety, pregnancy, labor, and the neonate. American Journal of Obstetrics and Gynecology, I974, ll§(2), 195-199. Butman, J. W. & Kamm, J. A. The social, psychological and behavioral world of the teenage girl. Report of Research from the Center for Research on Utiliza- tion of Scientific Knowledge, Institute for Social Research, June, 1965. Cairns, R. B. Beyond social attachment: The dynamics of interactional development. In T. Alloway, P. Pliner & L. Kramer (Eds.), Advances in the study of communication and affect, Vol. 3: Attachment behavior. New York: Plenum Press, 1972, 1-24. Call, J. D. & Marshak, M. Styles and games in infancy. In E. N. Rexford, L. W. Sander & T. Shapiro (Eds.), Infant psychiatry: A new synthesis. New Haven: Yale University Press, 1976, 104-112. Campbell, D. T. & Stanley, J. C. Experimental and quasi- experimental designs for research. Chicago: Rand McNally, 1963. Cannon-Bonventre, K. & Kahn, J. Interviews with adolescent parents: Looking at their needs. Children Today, 1979, t(5), 17-19. Caputo, D. V.; Taub, H. B.; Goldstein, K. M.; Smith, N.; Dalack, J. D.; Pursner, J. P.; & Silberstein, R. M. An evaluation of various parameters of maturity at birth as predictors of development at one year of life. Perceptual and Motor Skills, 1974, gt, 631-652. 196 Card, J. J. & Wise, L. L. Teenage mothers and teenage fathers: The impact of early childbearing on the parents: Personal and professional lives. Family Planning Perspectives, 1978, tg(4), 199-205. Carew, J. V. Social class, experience and intelligence. In H. McGurk (Ed.), Ecological factors in gtman development. New York: North-Holland, 1977, 189- 204. Carey, W. B. A simplified method for measuring infant temperament. Journal of Pediatrics, 1970, 11, 188-194. Center for Disease Control. Teenage childbearing and abor- tion patterns--United States, 1977. Morbidity and Mortality Weekly Rgport, 1980, tt(14), 157-159. Chavez, A.; Martinez, C.; & Yaschine, T. Nutrition, behav- ioral development, and mother-child interaction in young rural children. Federation Proceedings, 1975, t£(7), 1574-1582. Chilman, C. S. Adolescent sexuality in changing American society: Social and psychological perspectives. Washington, D.C.: U.S. Government Printing Office, 1980. Churchill, J. A. The relationship between intelligence and birth weight in twins. Neurology, 1965, t§(4), 341-347. - Clarke-Stewart, K. A. Interactions between mothers and their young children: Characteristics and conse- quences. Monographs of the Societyyfor Research in Child Development, 1973, tt(6-7, Serial No. 153). Coates, J. B. Obstetrics in the very young adolescent. American Journal of Obstetrics and Gynecology, 1970, 108(1), 68-72. Cohen, R. L. Some maladaptive syndromes of pregnancy and the puerperium. Obstetrics and Gynecology, 1966, t1, 562-570. Condon, W. S. Multiple response to sound in dysfunctional children. Journal of Autism and Childhood Schizophrenia, 1975, §(1), 37-56. Condon, W. S. & Sander, L. W. Neonate movement is synchron- ized with adult speech: Interactional participation and language acquisition. Science, 1974a, 183, 99- 101. 197 Condon, W. S. & Sander, L. W. Synchrony demonstrated between movements of the neonate and adult speech. Child Development, 1974b, 5;, 456-462. Coombs, L. C.; Freedman, R.; Friedman, J.; & Pratt, W. F. Premarital pregnancy and status before and after marriage. American Journal of Sociology, 1970, _t, 800-820. Cornell, E. H. & Gottfried, A. W. Intervention with pre- mature human infants. Child Development, 1976, 31, 32-39. Costner, H. L. & Schoenberg, R. Diagnosing indicator ills in multiple indicator models. In A. S. Goldberger & O. D. Duncan (Eds.), Structural equation modelstg the social sciences. New York: Seminar Press, 1973, 167-199. Crumidy, P. M. & Jacobziner, H. A study of young unmarried mothers who kept their babies. American Journal of Public Health, 1966, §§(8), 1242-1251. David, H. P. Unwanted pregnancies: Costs and alternatives. In G. F. Westoff & R. Parke (Eds.), Commission on population growth and the American future, research reports, Vol. 1, Demqgraphic and social aspects of population_growth. Washington, D.C.: U.S. Govern- ment Printing Office, 1972, 441-466. Davids, A. & DeVault, S. Maternal anxiety during pregnancy and childbirth abnormalities. Psychosomatic Medi- cine, 1962, tt, 464-470. Davies, P. A. & Stewart, A. L. Low-birth-weight infants: Neurological sequelae. British Medical Bulletin, 1975, tt(l), 85-91. deLissovoy, V. Child care by adolescent parents. Children Today, 1973, July-August, 22-25. deLissovoy, V. Concerns of rural adolescent parents. Child Welfare, 1975, §g(3), 167-174. Delsordo, J. D. Protective casework for abused children. In Jerome E. Leavitt (Ed.), The battered child. Morristown, N.J.: General Learning Press, 1974, 46-51. Dott, A. B. & Fort, A. T. Medical and social factors affecting early teenage childbearing. American Journal of Obstetrics and gynecology, June 15, 1976, tt§(4), 532-536. 198 Drage, J. S. & Berendes, H. Apgar scores and outcome of the newborn. Pediatrics Clinics of North America, 1966, lg, 635-643. Drillien, C. M. The growth and development of the pre- maturely born infant. Baltimore: Williams & Wilkens, 1964. Dryfoos, J. G. & Belmont, L. The intellectual and behavioral status of children born to adolescent mothers. Final report to NICHD, November, 1979. Dubowitz, L. M. S.; Dubowitz, V.; & Goldberg, C. Clinical assessment of gestational age in the newborn infant. The Journal of Pediatrics, 1970, 11(1), 1-10. Duenhoelter, J. H.; Jimienez, J. M.; & Baumann, G. Preg- nancy performance of patients under fifteen years of age. Obstetrics and Gypecology, 1975, g§(1), 49-52. Duncan, 0. D. Introduction to structural equation models. New York: Academic Press, 1975. Dunn, J. F. Mother infant relations: Continuities and discontinuities over the first 14 months. Journal of Psychosomatic Research, 1976, tg, 273-277. Dwyer, J. F. Teenage pregnancy. American Journal of Obstetrics and Gynecology, 1974, 118(3), 373-376. Edwards, N. The relationship between physical condition immediately after birth and mental and motor per- formance at age four. Genetic Psychology Mono- graphs, 1968, gt, 257-289. Elardo, R.; Bradley, R.; & Caldwell, B. M. The relation of infants' home environments to mental test perform- ance from six to thirty-six months: A longitudinal analysis. Child Development, 1975, gt, 71-76. Elkind, D. Egocentricism in adolescence. Child Develop- ment, 1967, gt, 1025-1034. Elmer, E. Children in jeopardy:, A study of abused minors and their families. Pittsburgh: University of Pittsburgh Press, 1967. Epstein, A. S. New insights into the problems of adolescent parenthood. Bulletin of the High/Scope Foundation, 1980, g, 6-8. 199 Epstein, A. S. Pregnant teenagers knowledge of infant development. Paper presented at the Biennial Meeting of the Society for Research in Child Development, San Francisco, March 15-18, 1979. Epstein, A. S. & Evans, J. Parent-child interaction and children's learning. The High Scope Rgport, 1979, (4), 39-43. Erikson, E. H. Identity and the life cycle. Psychological Issues, 1959, g(1), Monograph l. Ferreira, A. J. Emotional factors in prenatal environments: A review. Journal of Nervous and Mental Disease, 1965, 141, 108-118. Ferris, T. F. Toxemia of pregnancy: A model of human hypertension. Cardiovascular Medicine, September 1977, 877-897. Field, T. M.; Demosey, J. R.; Hallock, N. H.; & Shuman, H. H. The mother's assessment of the behavior of her infant. Infant Behavior and Development, 1978, t, 156-167. Fielding, J. E. Adolescent pregnancy revisited. The New England Journal of Medicine, 1978, 299(16), 893-896. Fink, E. L. Unobserved variables within structural equa- tion models. In P. R. Monge & J. Cappella (Eds.), Multivariate Techniques in Human Communication Research. New York: Academic Press, 1980, 111-141. Fink, E. L. & Mabee, T. I. Linear equations and nonlinear estimation: A lesson from a nonrecursive example. Sociological Methods & Research, 1978, 1(1), 107-120. Fontana, V. J. Which parents abuse children. In Jerome E. Leavitt (Ed.), The battered child. Morristown, N.J.: General Learning Press, 1974, 195-199. Francis-Williams, J. & Davies, P. A. Very low birthweight and later intelligence. Developmental Medicine and Child Neorology, 1974, gt, 709-728. Freedman, D. G. Human infancy: An evolutionary perspec- tive. New York: Wiley, 1974. Freedman, D. S. & Thornton, A. The long-term impact of pregnancy at marriage on the family's economic circumstances. Familerlanning Perspectives, 1979, gg(l), 6-21. 200 Freedman, R. & Coombs, L. Childspacing and family economic position. American Sociological Review, 1966, gt, 631-648. Fries, M. E. Psychosomatic relationships between mother and infant. ngchosomatic Medicine, 1944, g, 159- 162. Furstenberg, F. F. Jr. Burdens & benefits: The impact of early childbearing on the family. Paper presented at the Family Impact Seminar: Teenage Pregnancy and Family Impact: New Perspectives on Policy. Washington, D.C., 1979. Furstenberg, F. F. Jr. Unplanned parenthood: The social consequences of teenage childbearing. New York: The Free Press, 1976. Furstenberg, F. F. Jr. & Crawford, A. G. Family support: Helping teenage mothers to c0pe. Family Planning Perspectives, 1978, gg(6), 322-333. Gewirtz, J. L. & Boyd, E. F. Experiments on mother-infant interaction underlying mutual acquisition: The infant conditions the mother. In T. Alloway, P. Pliner, & L. Kramer (Eds.), Advances in the study of communication and affect, Vol. 3: Attachment be- havior. New York: Plenum Press, 1977, 109-143. Glass, G. V. & Stanley, J. C. Statistical methods in edu- cation and psychology. Englewood Cliffs, New Jersey: Prentice-Hall, 1970. Goldberg, S. Premature birth: Consequences for the parent- infant relationship. American Scientist, 1979, March-April, 214-220. Golden, M. & Birns, B. Social class and infant intelli- gence. In M. Lewis (Ed.), Origins of intelligence: Infancy and early childhood. New York: Plenum Press, 1976, 299-351. Goldstein, K. M.; Caputo, D. V.; & Taub, H. B. The effects of prenatal and perinatal complications on develop- ment at one year of age. Child Development, 1976, 51, 613-621. Gottschalk, L. A.; Titchener, J. L.; Piker, H. N.; & Stewart, S. S. Psychosocial factors associated with pregnancy in adolescent girls: A preliminary report. Journal of Nervous and Mental Diseases, 1964, tgg, 524-534. 201 Hack, M.; Fanaroff, A. A.; & Merkatz, I. R. The low-birth- weight infant--evolution of a changing outlook. The New England Journal of Medicine, 1979, ggt(21), 1162-1165. Hamblin, R. L. Social attitudes: Magnitude measurement and theory. In H. M. Blalock, Jr. (Ed.), Measure- ment in the social sciences. Chicago: Aldine, 1964, 61-120. Hanushek, E. A. & Jackson, J. E. Statistical methods for social scientists. New York: Academic Press, 1977. Hardy, J. B. The Johns Hopkins Collaborative Perinatal Project: Descriptive background. The Johns Hopkins Medical Journal, 1971, 128, 238-243. Hardy, J. B.; Welcher, D. W.; Stanley, J.; & Dallas, J. R. Long-range outcome of adolescent pregnancy. Clinical Obstetrics and Gynecology, 1978, tt(4), 1215-1232. Harper, P. A. & Weiner, G. Sequelae of low birth weight. Annual Review of Medicine, 1965, gt, 405-420. Hollingsworth, D. R.; Moser, R. J.; Carlson, J. W.; & Thompson, K. T. Abnormal adolescent primiparous pregnancy: Association of race, human chorionic somatomammotropin production, and smoking. American Journal of Obstetrics and Gynecology, 1976, gtg(2), 230-237. Honzik, M. P. Value and limitations of infant tests: An overview. In M. Lewis (Ed.), Origins of intelli- gence: Infancy & early childhood. New York: Plenum Press, 1976, 59-75. Hook, N. C. & Paolucci, B. The family as an ecosystem. Journal of Home Economics, 1970, gt(5), 315-318. Hulka, J. F. & Schaaf, J. T. Obstetrics in adolescents: A controlled study of deliveries by mothers 15 years of age and under. Obstetrics & Gynecology, 1964, tg(5), 678-685. Hunt, J. McV. Environmental programming to foster com- petence & prevent mental retardation in infancy. In R. M. Walsh & W. T. Greenough (Eds.), Environ- ments as thergpy for brain dysfunction. New York: Plenum Press, 1975, 201-256. 202 Hunt, J. V. & Rhodes, L. Mental development of preterm infants during the first year. Child Development, 1977, gt, 204-210. Illsley, R. Family growth and its effect on the relation- ship between obstetric factors and child function- ing. In R. Platt and A. S. Parkes (Eds.), Social and genetic influences on life and death. London: Oliver and Boyd, 1967, 29-42 (a). Illsley, R. The sociological study of reproduction and its outcome. In S. A. Richardson & A. F. Guttmacher (Eds.), Childbearing--its social and psychological aspects. Cambridge: Williams & Wilkins, 1967, 75-141 (b). Johnson, B. & Morse, H. A. Injured children and their parents. In Jerome E. Leavitt (Ed.), The battered child. Morristown, N.J.: General Learning Press, 1974, 18-23. Jones, A. E. & Placek, P. Teenage women in the U.S.A.: Sex, contraception, pregnancy, fertility, and maternal and infant health. A paper presented at the Family Impact Seminar for its study, Teenage Pregnancy and Family Impact: New Perspectives on Policy, Washington, D.C., 1979. JoreskOg, K. G. A general approach to confirmatory factor analysis. Psychometrika, 1969, gg, 183-202. Joreskog, K. G. A general method for estimating a linear structural equation system. In A. S. Goldberger and O. D. Duncan (Eds.), Structural equation models in the social sciences. New York: Seminar Press, 1973, 85-112. Jareskog, K. G. Analyzing psychological data by structural analysis of covariance matrices. In D. H. Krantz, R. C. Atkinson, R. D. Luce, & P. Suppes (Eds.), Contemporary developments in mathematical pgychology, Vol. 2. San Francisco: Freeman, 1974, 1-56. Joreskog, K. G. & Sorbom, D. LISREL structural relationships by the method of maximum likelihood. Chicago: National Educational Resources, Inc., 1978. Kagan, J. Overview: Perspectives on human infancy. In J. D. Osofsky (Ed.), The handbook of infant devel- opment. New York: Wiley, 1979, 1-25. 203 Kantor, D. & Lehr, W. Inside the family. Washington, D.C.: Jossey-Bass, 1976. Kennel, J. H.; Voos, D. K.; & Klaus, M. H. Parent-infant bonding. In J. D. Osofsky (Ed.), The handbook of infant development. New York: Wiley, 1979, 786- 798. Klaus, M. H. & Kennell, J. H. Maternal-infant bonding. St. Louis: C. V. Mosby, 1976. Klein, L. Antecedents of adolescent pregnancy. Clinical Obstetrics and gynecology, 1978, tt(4), 1151-1159. Kopp. C. B. & Parmelee, A. H. Prenatal and perinatal influ- ences on infant behavior. In J. D. Osofsky (Ed.), The handbook of infant development. New York: Wiley, 1979, 29-75. Korner, A. F. Conceptual issues in infancy research. In J. D. Osofsky (Ed.), The handbook of infant develop- ment. New York: Wiley, 1979, 769-785. Korner, A. F. The effect of the infant's state, level of arousal, sex, and ontogenetic stage on the care- giver. In M. Lewis and L. A. Rosenblum (Eds.), The effect of the infant on the carggiver. New York: Academic Press, 1974, 105-121. Korner, A. F. & Grobstein, R. Individual differences at birth: Implications for mother-infant relationship and later development. In E. N. Rexford, L. W. Sander & T. Shapiro (Eds.), Infant psychiatry: A new synthesis. New Haven: Yale University Press, 1976, 68-78. Lakin, M. Personality factors in mothers of excessively crying (colicky) infants. Monographs of the Society for Research in Child Development, Inc., 1957, 22 (Serial No. 64, No. 1). Land, K. C. & Felson, M. Sensitivity analysis of arbitrarily identified simultaneous equation models. Sociologi- cal Methods and Research, 1978, t, 283-307. Leibowitz, A. Education and home production. American Economic Review, 1974a, tg, 243-250. Leibowitz, A. Home investments in children. In T. W. Schultz (Ed.), Economics of the family. Chicago: University of Chicago Press, 1974b. Leifer, M. Psychological changes accompanying pregnancy and motherhood. Genetic Psychology Monographs, 1977, gt, 55-96. Lester, Lester, Lester, Lewis, Lewis, Lewis, Lewis, Lewis, 204 B. M. The continuity of change in development. Paper presented in a symposium on "Child Develop- ment and National Development: Interdisciplinary Perspectives," at the XVII Interamerican Congress of Psychology. Lima, Peru, July 1-6, 1979. B. M. Relations between teenage pregnancy and neo- natal behavior. Progress report to NICHD, March, 1978. Referred to in W. Baldwin & V. S. Cain, The children of teenage parents, Family Planning Pergpec- tives, 1980, gt(l), 34-43. B. M.; Als, H.; & Brazelton, T. B. Scoring criteria for seven clusters of the Brazelton scale. Unpub- lished manuscript, Child Development Unit, Children's Hospital Medical Center, Boston, Massachusetts, 1978. M. What do we mean when we say "Infant Intelligence Scores"? A sociopolitical question. In M. Lewis (Ed.), Origins of intelligence--infancy and early childhood. New York: Plenum Press, 1976, 1-17. M. & Freedle, R. The mother and infant communication system: The effects of poverty. In H. McGurk (Ed.), Ecologycal factors in human development. New York: North-Holland, 1977, 205-215. M. & Goldberg, S. Perceptual-cognitive development in infancy: A generalized expectancy model as a function of the mother-infant relationship. Merrill-Palmertguarterly, 1969, gt, 81-100. M. & Lee-Painter, S. An interactional approach to the mother-infant dyad. In M. Lewis & L. A. Rosen- blum (Eds.), The effect of the infant on the care- takers. New York: Academic Press, 1974, 21-48. M. & Starr, M. D. Developmental continuity. In J. D. Osofsky (Ed.), The handbook of infant devel- opment. New York: Wiley, 653-670. Lobl, M.; Welcher, D. W.; & Mellits, E. D. Maternal age Lozoff, and intellectual functioning of offspring. The Johns Hopkins Medical Journal, 1971, 128, 347-361. B.; Brittenham, G. M.; Trause, M. A.; Kennell, J. H.; & Klaus, M. H. The mother-newborn relationship. Limits of adaptability. The Journal of Pediatrics, 1977, 21(4), 1-12. 205 McAnarney, E. R. Adolescent pregnancy--a pediatric concern? Clinical Pediatrics, 1975, gg(l), 19-22. McCall, R. B. The development of intellectual functioning in infancy and the prediction of later I.Q. In J. D. Osofsky (Ed.), The handbook of infant devel- opment. New York: Wiley, 1979, 707-741. McCall, R. B.; Appelbaum, M. I.; & Hogarty, P. S. Develop- mental changes in mental performances. Monographs of the Society for Research in Child Development, 1973, gt(3, Serial No. 150). McCarthy, B. J.; Abbott, B.; & Terry, J. S. Teenage preg- nancy in Georgia: Defining the problem. Journal of the Medical Association of Georgia, 1979, gt, 373-378. McDonald, A. D. Intelligence in children of very low birth weight. British Journal of Preventive Social Medicine, 1964, gt, 59-74. McDonald, R. L. The role of emotional factors in obstetric complications: A review. Psychosomatic Medicine, McDonald, R. L. & Parham, K. J. Relation of emotional changes during pregnancy to obstetric complications in unmarried primigravidae. American Journal of Obstetrics and Gynecology, 1964, gt, 195-201. McFarlane, A. If a smile is so important. New Scientist, 1974, 25 April, 164-166. McGurk, H. Ecological factors in human development. New York: North Holland, 1977. McKenry, P. C.; Walters, L. H.; & Johnson, C. Adolescent pregnancy: A review of the literature. The Family Coordinator, 1979, tt(l), 17-28. Martinez, G. A. & Nalezienski, J. P. The recent trend in breast-feeding. Pediatrics, 1979, gg(5), 686-692. Mednick, B. R.; Baker, R. L.; & Sutton-Smith, B. Teenage pregnancy and perinatal mortality. Journal of Youth and Adolescence, 1979, t(3), 343-357. Mellor, S. & Wright, J. D. Adolescent pregnancy. The Practitioner, 1975, 215, 77-82. 206 Menken, J. The health and social consequences of teenage childbearing. Family Planning Perspectives, 1972, Mercer, R. T. Teenage motherhood: The first year. Part I: The teenage mother's views and responses. Part II: How the infants fared. Journal of Obstetric, gynecologic & Neonatal Nursing, 1980, t(l), 16-27. Messer, S. B. & Lewis, M. Social class and sex differences in the attachment and play behavior of the year-old infant. Merrill-Palmer Quarterly, 1972, gt, 295- 306. Meyerowitz, J. H. & Malev, J. S. Pubescent attitudinal correlates antecedent to adolescent illegitimate pregnancy. Journal of Youth and Adolescence, 1973, t, 251-258. Michigan Department of Public Health, Office of Vital Statistics, Lansing, Michigan, 1979. Moore, K. A. The social and economic consequences of teenage childbearing for women, families, and government welfare expenditures. Testimony to the Human Resources Committee of the United States Senate, June 14, 1978. Moore, K. A. & Caldwell, S. B. The effect of government policies on out-of-wedlock sex and pregnancy. Family Planning Perspectives, 1977, 2(4), 164-169. Moore, K. A. & Hofferth, S. L. Factors affecting early family formation: A path model. Washington, D.C.: The Urban Institute, 1978. Moore, K. A.; Hofferth, S. L.; Caldwell, S. B.; & Waite, L. J. Teenage motherhood: Social and economic consequences. Washington, D.C.: The Urban Insti- tute, 1979. Moore, K. A.; Hofferth, S. L.; & Wertheimer, R. II. Teen- age motherhood: Its social and economic costs. Children Today, 1979, t(S), 12-16. Moore, S. Mother-child interactions and competence in infants and toddlers. Young Children, 1977, March, 64-69. Moore, T. Language and intelligence: A longitudinal study of the first eight years. Part II. Environmental coorelates of mental growth. Human Development, 1968, gt, 1-24. 207 Muller, P. F.; Campbell, H. E.; Graham, W. E.; Britain, H.; Fitzgerald, J. A.; Hogan, N. A.; Muller, V. H.; & Ritterhouse, A. H. Perinatal factors and their relationship to mental retardation and other parameters of development. American Journal of Obstetrics and Gynecology, 1971, gtt, 1205-1210. National Center for Health Statistics. Monthly Vital Statistics Report: Natality Statistics. Washington, D.C.: Department of Health, Education & Welfare, tt(S), Supplement September 8, 1977. Nelson, S. A. School age parents. Children Today, 1973, March-April, 31-33, 40. Nie, N. H.; Hull, C. H.; Jenkins, J. G.; Steinbrenner, K.; & Bent, D. M. Statistical package for the social ggggnces. Second Edition. New York: McGraw-Hill, 1975. Nortman, D. Parental age as a factor in pregnancy outcome and child development. Reports on Population/ Family Plannipg, 1974, (16). Nye, I. F. School-age parenthood: Consequences for babies, mothers, fathers, grandparents and others. Extension Bulletin #667, Cooperative Extension Service, Washington State University, Pullman, 1976. Oppel, W. C. & Royston, A. B. Teen-age births: Some social, psychological, and physical sequelae. American Journal of Public Health, 1971, tt(4), 751-756. Osofsky, H. J. The pregnant teen-ager, a medical, educa- tional & social analysis. Springfield, Illinois: Thomas Publishing, 1968. Osofsky, J. D. & Connors, K. Mother-infant interaction: An integrative view of a complex system. In J. D. Osofsky (Ed.), The handbook of infant devel- opment. New York: Wiley, 1979, 519-548. Osofsky, H. J. & Osofsky, J. D. Adolescents as mothers: Results of a program for low-income pregnant teen- agers with some emphasis upon infants' development. American Journal of Orthopsychiatry, 1970, gt(S), 825-834. Overton, W. F. & Reese, H. W. General models for man- environment relations. In H. McGurk (Ed.), tgg- logical factors in human development. New York: North-Holland, 1977, 11-20. 208 Paolucci, B.; Hall, 0. A.; and Axinn, N. W. Family decision making: An ecosystem approach. New York: Wiley, 1977. Parker, R. Perspectives on father infant interaction. In J. D. Osofsky (Ed.), The handbook of infant develop- ment. New York: Wiley, 1979, 549-590. Parmelee, A. H. & Beckwith, L. The Child Home Observation Scale. Unpublished manuscript. University of California--Los Angeles, Los Angeles, California, 1972. Pasamanick, B. & Knoblock, H. Retrospective studies on the epidemiology of reproductive casualty: Old and new. Merrill-Palmer Quarterly, 1966, gt, 7-26. Persson-Blennow, I. & McNeil, T. F. A questionnaire for measurement of temperament in six-month-old infants: Development and standardization. Journal of Child Psychology and Psychiatry, 1979, 20, 1-13. Pollitt, E. Failure to thrive: Socioeconomic, dietary intake and mother-child interaction data. Federation Proceedings, 1975, gg(7), 1593-1597. Powell, L. F. The effect of extra stimulation and maternal involvement on the development of low-birth-weight infants and on maternal behavior. Child Develop- ment, 1974, gt, 106-113. Prechtl, H. The neurological examination of the full-term newborn infant, 2nd editiop, Clinics in Develop- mental Medicine No. 63. Philadelphia: Lippincott, 1977. Prechtl, H. & Beintema, D. The neurological Examination of the full term newborn infant. Clinics in Develop- mental Medicine, No. 12. London: Spastics Society with Heinemiann Medical, 1964. Presser, H. B. The social and demographic consequences of teenage childbearing for urban women. Final report to NICHD, National Technical Information Service, Washington, D.C., 1980. Ramey, C. T.; Stedman, D. J.; Borders-Patterson, A.; & Mengel, W. Predicting school failure from infor- mation available at birth. American Journal of Mental Deficiency, 1978, tt(6), 525-534. 209 Record, R. G.; McKeown, T.; & Edwards, J. H. The relation of measured intelligence to birth order and maternal age. Annal of Human Genetics, 1969, gg, 61-69. Reese, H. W. & Overton, W. F. Models of deve10pment and theories of development. In L. R. Goulet & P. B. Baltes (Eds.), Life-span developmental_psychology: Research and theory. New York: Academic Press, 1970, 115-145. Reiss, A. J.; Duncan, O. D.; Hatt, P. K.; & North, C. C. Occupations and social status. New York: The Free Press of Glencoe, 1961. Ricciuti, H. Adverse social and biological influences on early development. In H. McGurk (Ed.), Ecological factors in human development. New York: North- Holland, 1977, 157-172. Robson, K. S. The role of eye-to-eye contact in maternal- infant attachment. Journal of Child Psychology & Psychiatry, 1967, t, 13-25. Rolfe, D. J. & Roosa, M. W. Court-ordered evaluation of young teenage applicants for marriage licenses: Follow-up of a pilot study. Conciliation Courts Review, 1979, g1(2), 25-29. Roosa, M. W. The differences between dissolved and intact pregnancy-provoked marriages; a follow-up study. Unpublished master's thesis. Michigan State Univer- sity, East Lansing, Michigan, 1977. Russ-Eft, D.; Sprenger, M.; & Beever, A. Antecedents of adolescent parenthood and consequences at age 30. The Family Coordinator, 1979, tt(2), 173-178. Ryan, G. M. Jr. & Schneider, J. M. Teenage obstetric com- plications. Clinical Obstetrics and Gynecology, 1978, tg(4), 1191-1197. Sameroff, A. J. Early influences on development: Fact or fancy? Merrill-Palmertguarterly, 1975, tg(4), 267- 294. Sameroff, A. J. & Chandler, M. J. Reproductive risk and the continuum of caretaking casualty. In Frances Dugan Horowitz (Ed.), Review of child development research, Vol. 4. Chicago: University of Chicago Press, 1975, 187-244. 210 Sander, L. W. Infant and caretaking environment: Investi- gation and conceptualization of adaptive behavior in a system of increasing complexity. Developmental Research. In E. J. Anthony (Ed.), The child psychia- trist as investigator. New York: Plenum, 1977, 129-166. Sander, L. W. Issues in early mother-child interaction. In E. N. Rexford, L. W. Sander & T. Shapiro (Eds.), Infantypsychiatry: A new synthesis. New Haven: Yale University Press, 1976, 127-147. Sander, L. W. Regulation and organization in the early infant-caretaker system. In R. Robinson (Ed.), The Brain and early behavior. London: Academic, 1970, 313-331. Sander, L. W.; Julia, H. L.; Stechler, G.; & Burns, P. Continuous 24-hour interactional monitoring in infants reared in two caretaking environments. Psychosomatic Medicine, 1972, gg(3), 270-283. Sandler, H. M. Effects of adolescent pregnancy on mother- infant relationships: A transactional model. Progress reports to NICHD, June 1977, January 1978, May 1978, and May 1979. Referred to in W. Baldwin & V. S. Cain, The children of teenage parents, Family Planning Perspectives, 1980, gt(l), 34-43. Schafer, H. R. The growth of sociability. Hammondsworth, England: Penguin Books, Ltd., 1971. Schaffer, R. Behavioral synchrony in infancy. New Scientist, 1974, 4 April, 16-18. Scrimshaw, S. C. M. Infant mortality and behavior in the regulation of family size. Population and Devel- opment Review, 1978, g(3), 383-403. Sepkoski, C.; Coll, C.; & Lester, B. The effects of high risk factors on neonatal behavior as measured by the Brazelton scale. Paper presented at the Symposium on Research with Latin American Infants, XVI Inter- American Congress of Psychology. Miami, Florida, December, 1976. Shaw, C. A comparison of the patterns of mother-baby interaction for a group of crying, irritable babies and a group of more amenable babies. Child: Care, Health and Development, 1977, g, 1-12. 211 Shinn, A. M. Relations between scales. In H. M. Blalock, Jr. (Ed.), Measurement in the social sciences. Chicago: Aldine, 1974, 121-155. Sigman, M. Early development of preterm and full term infants: Exploratory behavior in eight-month-olds. Child Development, 1976, 31, 606-612. Smith, P. B.; Mumford, D. M.; Goldfarb, J. L.; & Kaufman, R. H. Selected aspects of adolescent postpartum behavior. The Journal of Reproductive Medicine, 1975, gg, 159-165. Smith, S. M.; Hanson, R.; & Noble, S. Parents of battered children: A controlled study. In A. W. Franklin (Ed.), Concerning child abuse. London: Churchill Livingston, 1975, 41-48. Social Research Group. Toward interagency coordination: FY '76 federal research and development on early childhood, sixth annual report. Washington, D.C.: The George Washington University, 1977. Solkoff, N. & Matuszak, D. Tactile stimulation and behav- ioral development among low-birthweight infants. Child Psychiatry and Human Development, 1975, t(l), 33-37. Sostek, A. M. & Anders, T. F. Relationship among the Brazelton Neonatal Scale, Bayley Infant Scales, and early temperament. Child Development, 1977, gt, 320-323. Sroufe, L. A. Socioemotional development. In J. D. Osofsky (Ed.), The handbook of infant development. New York: Wiley, 1979, 462-516. Stern, G. G.; Caldwell, B. M.; Hersher, L.; Lipton, E. L.; & Richmond, J. B. A factor analytic study of the mother-infant dyad. Child Development, 1969, gt, 163-181. Stern, D. N. Mother and infant at play: The dyadic inter- action involving facial, vocal, and gaze behaviors. In M. Lewis and L. A. Rosenblum (Eds.), The Effect of the child on the caregiver. New York: Academic Press, 1974, 187-213. Stickle, G. & Ma, P. Pregnancy in adolescents: Scope of the problem. Contemporary Obstetrics and Gynecology, 1975, §, 85-100. 212 Stott, D. H. Follow-up study from birth of the effects of prenatal stresses. Developmental Medicine and Child Neurology, 1973, gt, 770-787. Stott, D. H. Physical and mental handicaps following a disturbed pregnancy. The Lancet, May 18, 1957, 1006-1012. Sugarman, M. Paranatal influences on maternal-infant attachment. American Journal of Orthopsychiatpy, 1977, 51(3), 407-418. Sutton-Smith, B. Teenage pregnancy and birth problems found entirely social. Behavior Todsy, 1979, May 7, 1-2. Tanner, J. M. Variability of growth and maturity in newborn infants. In M. Lewis and L. A. Rosenblum (Eds.), The effect of the infant on the caregiver. New York: Academic Press, 1974, 77-103. Taylor, J. The special needs of school-age parents and their infants. SharingyAmong Those Concerned with School Age Parents, Winter 1975, 6-8. Thoman, B. Development of synchrony in mother-infant interaction in feeding and other situations. Federa- tion Proceedings} 1975, gg(7), 1587-1592. Thoman, E. B. Some consequences of early infant-mother- infant interaction. Early Child Development and Care, 1974, g, 249-261. Thoman, E. B. The role of the infant in the early transfer of information. Biological Psychiatry, 1975, gt(2), 161-169. Thomas, A.; Chess, 8.; Birch, H. G.; Hertzig, M. E.; & Korn, S. Behavioral individuality in early childhood. New York: New York University, 1963. Thompson, A. M. Pregnancy in adolescence. In J. I. McKigney & H. N. Munro (Eds.), Nutrient require- ments in adolescence. Cambridge, Massachusetts: MIT Press, 1976, 245-256. Tronick, E. & Brazelton, T. B. Clinical uses of the Brazel- ton Neonatal Behavior Assessment. In B. Z. Fried- 1ander and L. Rosenblum (Eds.), Exceptional Infant, Vol. III. New York: Brunner/Mazel, 1975, 137-152. Trussell, J. & Menken, J. Early childbearing and subse- quent fertility. Family Planning Perspectives, 1978, gt(4), 209-218. 213 _ Trussell, T. J. Economic consequences of teenage child- bearing. Family Planning Perspectives, 1976, t(4), 184-190. Tulkin, S. R. & Kagan, J. Mother-child interaction in the first year of life. Child Development, 1972, gg, 31-41. Uzgiris, I. C. Organization of sensorimotor intelligence. In M. Lewis (Ed.), Origins of intelligence--infancy and early childhood. New York: Plenum Press, 1976, 123-163. Vandenberg, B. 1976 reported in Chilman 1980. Vincent, C. E.; Haney, C. A.; & Cochrane, C. M. Familial and generational patterns of illegitimacy. Journal of Marriage and the Family, 1969, gg, 659-667. Wachs, T. D.; Francis, J.; & McQuiston, S. Psychological dimensions of the infants physical environment. Paper presented at the Midwest Psychological Associ- ation, Chicago, May, 1978. Wachs, T. D.; Francis, J.; & McQuiston, S. Psychological dimensions of the infants physical environment. Infant Behavior and Development, 1979, t(2), 155- 161. Wachs, T. D.; Uzgiris, I. C.; & Hunt, J. McV. Cognitive development in infants of different age levels and from different environmental backgrounds: An explanatory investigation. Merrill-Palmer anrterly, 1971, gt, 283-317. Waite, L. J. & Moore, K. A. The impact of an early first birth on young women's educational attainment. Social Forces, 1978, gt(3), 845-865. Walters, J. Birth defects and adolescent pregnancies. Journal of Home Economics, 1975, 23-29. Walters, J. & Stinnett, N. Parent-child relationships: A decade review of research. Journal of Marriage and the Family, 1971, 70-111. Weiner, G.; Rider, R. V.; Oppel, W. C.; & Harper, P. A. Correlates of low birth weight. Psychological status at eight to ten years of age. Pediatric Research, 1968, t, 110-118. 214 Werner, E. E.; Honzik, M. P.; & Smith, R. S. Prediction of intelligence and achievement at ten years from twenty months pediatric and psychologic examina- tions. Child Development, 1968, gt, 1063-1075. Wheaton, B.; Muthen, B.; Alwin, D. F.; & Summers, G. F. Assessing reliability and stability in panel models. In D. R. Heise (Ed.), Sociological methodology 1977. Washington, D.C.: Jossey-Bass, 1977, 84-136. White, J. L. & Labarba, R. C. The effects of tactile and kinesthetic stimulation on neonatal development in the premature infant. DevelOpmental Psychobiology, 1976, 2(6), 569-577. Willems, E. P. Relation of models to methods in behavioral ecology. In H. McGurk (Ed.), Ecological factors in human development. New York: North-Holland, 1977, 21-35. Willerman, L. & Churchill, J. A. Intelligence and birth weight in identical twins. Child Development, 1967, gt(3), 623-629. Williams, T. M. Childrearing practices of young mothers: What we know, how it matters, why it's so little. American Journal of Orthopsychiatty, 1974, gg(l), 70-75. Wohlwill, J. F. & Heft, H. Environments fit for the devel- oping child. In H. McGurk (Ed.), Ecological factors in human development. New York: North Holland, 1977, 125-132. Wolff, P. H. Mother-infant interactions in the first year. The New England Journal of Medicine, 1976, 295(18), 999-1001. Yang, R. K.; Zweig, A. R.; Douthitt, T. C.; & Federman, E. J. Successive relationships between maternal attitudes during pregnancy, analgesic medication during labor and delivery, and newborn behavior. Developmental Psychology, 1976, gt(l), 6-14. Yarrow, L. J.; Rubenstein, J. L.; Pederson, F. A.; & Jankowski, J. J. Dimensions of early stimulation and their differential effects on infant develop- ment. Merrill-Palmer Quarterly, 1972, gt, 205-218. Zachau-Christiansen, B. & Ross, E. M. Babies: Human devel- opment during the first year. New York: John Wiley & Sons, 1975. APPENDIX A SUBJECT RECRUITMENT LETTERS APPENDIX A SUBJECT RECRUITMENT LETTERS Letter to Mothers--Tri-County Area MICHIGAN STATE UNIVERSITY Mother-Infant Project Institute for Family and Child Study East Lansing, Michigan 48824 To the Expectant Mother: If you are between 15 and 30, and will be having your first baby in the next few months, I would like to include you in my study of mothering behavior. There are a large number of ways to care for children in our culture with vast differences between these ways. I am studying the many ways of caring for children and recording the dif- ferences. In order for all types of mothering behavior to be studied, it is important for mothers of all ages and from all walks of life to be included in the study. The study includes: 1. Qgg brief interview in your home before you give birth. 2. gtg brief examination of your child a few days after birth in the hospital. 3. Three visits by a trained observer who will come to your home once a month to watch you and your infant, to ask a few questions, and to have you fill out some questionnaires. ALL DATA THAT IS COLLECTED IS STRICTLY CONFIDENTIAL If you participate you will receive a one week supply of diapers (60 diapers) after each of the three home visits. You will get an additional week's supply of diapers for completing all three visits. 215 216 If you are interested in being part of this study or if you want more information, please call me at 353-6617 or fill out and send in the attached postcard. The study ends in June 1980. Thank you, Mark W. Roosa Project Coordinator Principal Investigators Dr. H. Fitzgerald, Psychology Dr. N. Carlson, Family and Child Sciences 217 Letter to Mothers--Alma Area MICHIGAN STATE UNIVERSITY Mother-Infant Project Institute for Family and Child Study East Lansing, Michigan 48824 To the Expectant Mother: If you are between 15 and 30, and will be having your first baby in the next few months, I would like to include you in my study of mothering behavior. There are a large number of ways to care for children in our culture with vast differences between these ways. I am studying the many ways of caring for children and recording the differences. In order for all types of mothering behavior to be studied, it is important for mothers of all ages and from all walks of life to be included in the study. The study includes: 1. Qgg brief interview in your home before you give birth. 2. th brief examination of your child a few days after birth in the hospital. 3. Three visits by a trained observer who will come to your home once a month to watch you and your infant, to ask a few questions, and to have you fill out some questionnaires. ALL DATA THAT IS COLLECTED IS STRICTLY CONFIDENTIAL WITH THE EXCEPTION THAT THE DATA WILL ALSO BECOME PART OF YOUR MEDICAL RECORDS. If you participate you will receive a one week supply of diapers (60 diapers) after each of the three home visits. You will get an additional week's supply of diapers for completing all three visits. If you are interested in being part of this study or if you want more information, please call me at 353-6617 or fill out and send in the attached postcard. The study ends in June 1980. Thank you, Mark W. Roosa Project Coordinator Principal Investigators Dr. H. Fitzgerald, Psychology Dr. N. Carlson, Family and Child Sciences APPENDIX B INFORMED CONSENT PROCEDURE MATERIALS Houmcwcuoou pommoum mmoom .3 gum: vmmmv ammH£Oflz .mcwmcmq ummm >uwmum>wcs mumum savanna: 33% 320 can. Deana 6m 3:335 Homnoum usmmcHlumnuoz pawno mo ummu 3mw>uwucH HMfluwcH .m Hmucmam0a6>ma .v newsman mEuom meMCGOAHmmSO .m muwmccoflummsa .m cowmmwauom new ucmmcoo poauom:H .m 3mw>uwucH .N 3wfi>uoucH .m 3mfl>umucH .m spawn mo mcuoowu Hmuflmmon 3mw>mm .m mcofiummsv mac: CH 050: cw mac: ca Hmzmcm new mCOHum>meno .H m:0flum>uwmno .H mc0flum>uomno .H meo unamaH .H hpsum :memxm .H mnucoz m mango: m :ucoz H Assam nuufim whommm mmwua>fluod mo masomnum uomnoum ucmmcHuuwnuoz mfldeMBflz mmDDmvomm BZMmZOO GuzmQhZH m XHDZNmmd 218 219 Informed Consent--Eaton, Clinton and Ingham Counties Mother-Infant Project Michigan State University Institute for Family and Child Study 1. I have freely consented to take part in a scientific study being conducted by Mark W. Roosa under the supervision of Dr. H. Fitz- gerald, Professor of Psychology and Dr. N. C. Carlson, Assistant Professor of Family and Child Sciences. The study has been explained to me and I understand the explanation that has been given and what my participation will involve. I understand that I am free to discontinue my participation in the study at any time without penalty. I understand that the results of the study will be treated in strict confidence and that I will remain anonymous. Within these restric- tions, the general results of the study will be made available to me at my request. I understand that my participation in the study does not guarantee any beneficial results to me or to my infant. I understand that, at my request, I can receive additional explana- tion of the study after my participation is completed. I understand that I will receive a four week supply of diapers (240 diapers total) if I participate in the complete study. Signed Date Age (Husband--if applicable) Date (Parent/Probate Judge--if applicable) Date Physician's Name Hespital 220 Informed Consent--Gratiot County Mother-Infant Project Michigan State University Institute for Family and Child Study 1. Name of Physician I have freely consented to take part in a scientific study being conducted by Mark W. Roosa under the supervision of Dr. H. Fitz- gerald, Professor of Psychology and Dr. N. Carlson, Assistant Professor of Family and Child Sciences. The study has been explained to me and I understand the explanation that has been given and what my participation will involve. I understand that I am free to discontinue my participation in the study at any time without penalty. I understand that the results of the study will be treated in strict confidence and that I will remain anonymous. I understand that my physician is cooperating with this project and he has requested that the information obtained from me will become part of my medical records. Within these restrictions, the general results of the study will be made available to me at my request. I understand that my participation in the study does not guarantee any beneficial results to me or to my infant. I understand that, at my request, I can receive additional explana- tion of the study after my participation is completed. I understand that I will receive a four week supply of diapers (240 diapers total) if I participate in the complete study. Signed Date Age (Husband--if applicable) Date (Parent/Probate Judge--if applicable) Date 221.‘ Consent Form for Hospital Records and Brazelton Examination MICHIGAN STATE UNIVERSITY Mother-Infant Project Institute for Family and Child Study East Lansing, Michigan 48824 I am participating in a study of mothers and infants being con- ducted by the Institute for Family and Child Study. It is generally believed that experiences during pregnancy and birth may have an effect upon both the mother and her child. Therefore the researchers are including obstetric infermation in their study. The obstetric informa- tion that they are collecting (length of gestation, size at birth, medications given during labor and birth, etc.) is that which the physician normally places in medical records at the hospital at the time of the birth. It is also important to understand the baby's capabilities at birth. The Brazelton Neonatal Behavioral Assessment is designed to measure these capabilities within a few days after the birth. The Brazelton Assessment tests such things as the baby's response to a bright light or a ringing bell, how the baby's eyes follow an object or a face, and some of the baby's reflexes. I have discussed the request of the researchers and I fully understand that request. I freely consent to allow the researchers from the Institute for Family and Child Study to obtain information from the hospital records of my delivery and to perform the Brazelton Assessment. Study Participant H. Fitzgerald N. Carlson Principal Investigators M. Roosa Project Coordinator APPENDIX C INSTRUMENTS Accum ammhuav mumsoeaem Aocnm «mwhuav Hoonom ca maucmmmum A.ouw .pmum mmmHHooumH .omum .m.mume lumuwea uaoo Hmw> unmav sewumosom Auw\>mo\oz ..m.mv mumofluuwm ANNV ANNV AHNV AHNV flame Acme Aomlmav Aomlmav Amalmflv Amalmav Amfilmav AhHINHV mm vm mo .NHV Away 2 0 mm NS Hm HE Z O m m m AHHV AHHV AHHV ucmnmm mam: Hmnuoz m N H Aoanmv mama «quay DH sma>umucH HmauacH ”somfloum ucmmcHuumauoz mBZMZDMBmZH U xHszmmfl unmpcommwm 222 223 Ammo COOm :men on mco ou pmuufisfioo .m 32 o: .N mm» .H 2595 amnmmc momuommxo hams we conz Acme Hmsuo .5 ucoESOHmess .o huwusomm HMflUOm .m mummamz .v o>fiumaos “mayo .m mama .m mucoumm .H Avmnomv Avmnomv Ammnmmv Ammummv Ammlmmv Ammnwmv Avmlmmv Avmnmmv Ammnmmv vacuum mam: Macao: msmumoum mewcflmuu Hmumcmum m ca pmummwofiuumm so» m>mm .m Au>\oEV mucmsmoum mEooon 90% new sump wags Saoumafixoummm :0 .h .m0 m0 monsom .no wEOUCH Hmnuo .mo Amt Somme mEoocH Amazed .0 ACOAumwuommc mumamfioov cowummsooo .m 224 Amvv umsuo .o pmzopfl3 .m coouo>fip .v .mve cmumuMQmm .m some lame Ammsau lie cmauume .N pmwuuma um>wc .H Aves umnuo .m mm>flumamu umnuo can mama sue: .m mw>wumawu Monuo 59H: .b mucwumm man can mama nuwz .m mucmumm no: new mama saws .m mpcmumm man spas .v mucwumm no: nuw3 .m mama saws .N wcoam .H Amvv umnuo .m oasflao .N umcoflufiuomum wum>flum .H Aevuovc ucmumm mam: umeuoz msumum Hmufiumz ucmawmcmuum mafi>fiq «semen mumo Hmumc Iona twp mocmcmwum mo races noes; mcausa .am wumo Hmumcwum mo wmhe .mm Amuflww> cmmsuon mxmm3 we mmuwmfl> Hmumcmum How uouoop on» 0» oo so» 00 smumo so: .HH .OH 225 "mucoEEOU w v m N H own hum> can mmmuo>m p000 poem mum> Ammo amneom mocmcmoum Hso> ma 30m Ammv umsuo .m Monumm H50» .m Hm£u05 usoa .v negro mo “wrung uo:\wcmaumson .m omsomm u0:\paw£o mo umzumm .N mmsomm .H won: .mo» MH .mMH isms asum>aewo unsung so» saws on mcomEOm Hafiz Ammo «same sacs comm ummmun on swam 50% on Amvumvv onnmvc Ammummc Aumm>\osc ucm>m mo pump .m .v .m* NH .oHH Assummc lmvnmmc lemumvc Aumms\osc «magnuma umma mo dump .~* MH .QHH «menace m wave on» cw muums o» swam 50> cm .H# MH .mHH ucwumm can: uwsuoz .VH .MH .NH 226 ”mucwfiaou m e m m H Hemp ummum an boom mHuuHH mHuuHH mum> one: 83 5: H3330 fit, on; sos o>mn mocoHummxo none 30: ”mucosaoo m v m m H Home ummum an poem mHuuHH mHuuHH huw> mac: AHov whocmcmoum one woch commune uoHo Mao» mm: none 30: "mucwEEou m e m m H Hump umoum an coom oHuuHH mHuuHH >um> wcoc 33 2.3 88 «we: 86> 8820 aocmcmmum on» mm: nose 30m assume mHmz Honuoz .5H .0H .mH 227 Amev AHmV "mzmH>HoDCH so mucmEEOU Home Hmmv vav Homv Anvv Amvv ucoumm mHmz Hague HmucoHuo CMHpeH OHCMQmHm MOMHm wanz vav 3ocx H.206 mHmE meEmw Amov umpumE u.:mo00 mHmE mHmEmm umnuoz .00 HNMVLDQ "szmH>umucH msoum UHCQHm no menu .mH $22 23 sos xfifi 90% on heme xwm umsz .mmH mucm3 90> on hang xmm umsx .mH l. 2. 10. 11. 228 Mother-Infant Project Obstetric Information Form Date of Birth of Infant ID (1—4) Date (5-10) (ll-l6) Hospital of Birth 1. Sparrow 2. St. Lawrence 3. Lansing General 4. Gratiot Community 5. Eaton Rapids Community 6. Hayes-Green-Beach (Charlotte) 7. Ingham Medical 8. Mason General 9. Provincial ‘ 0. Other (17) Gestational age of the infant weeks (18-19) at birth or Date of onset of the mother's (20-25) last menstrual cycle Dubowitz score (26-27) Weight of the Infant at Birth gms (28-31) Height of the Infant at Birth cm (32-33) Head Circumference cm (34-35) Length of Labor (e.g., 01.5) hrs (36-38) First Apgar Score (39-40) Second Apgar Score (41-42) Mother's Weight at the Beginning of Pregnancy lbs (43-45) Mother's Weight at Delivery lbs (46-48) 229 12. Was an obstetric medication used? a. If yes, what type? (Please check the appropriate column in the following list.) Obstetrical Medications L D Labor (49-50) Delivery (51-52) Premedication Agents 22 Uterine Muscle Stimulants oxytoxics ergot alkaloids - ergonovine (Ergotrate) _ _ 01 methylergonovine (Methergine) __ _ 02 oxytocin (Pitocin, Syntocinon) _ _ 03 prostaglandin E2 _ _ 04 Narcotic Analgesics alphaprodine (Nisentil) _ _ 05 meperidine (Demerol) _ _ 06 morphone _ _ 07 Narcotic Antagonists levallorphan (Lorfan) _ _ 08 nalorphine (Nalline) _ _ 09 naloxone (Narcan) _ _ 10 Sedative-Hypnotics barbiturate pentobarbital (Nembutol) _ _vll secobarbital (Seconag)a _ _ 12 amobarbital (Amytal) _ _ 13 nonbarbiturate diazepam (Valium) _ _ l4 scopolamine (hyoscine) _ _ 15 Tranquilizers chlorpromazine (Thorazine) _ _ l6 hydroxyzine (Atarax, Vistaril) _ _ l7 promazine (Sparine) _ _ 18 promethazine (Phenergan) _ _ 19 propriomazine (Largon) _ _ 20 triflupromazine (Vesprin) 21 aSeconal and Amytal are often combined as Tuinal. 230 PB General Anesthetic and Related Agents Inhalant Anesthetics cyclopropane _ _ 22 enflurane (Ethrane) _ _ 23 halothane (Fluothane) _ _ 24 methoxyflurane (Penthrane) _ _ 25 nitrous Oxide _ _ 26 trichloroethylene (Trilene) _ _ 27 Intravenous Anesthetics barbiturate ' thiopental (Penthothal) _ _ 28 methohexital (Brevital) _ _ 29 nonbarbiturate ketamine (Ketalar, Ketaject) _ _ 30 Neuromuscular Blocking Agents d-tubocurarine (Curare) _ _ 31 gallamine (Flaxedil) _ _ 32 succinylcholine (Anectin, Quelicin, Sucostrin) _ _ 33 Antimuscarinic Agents atropine _ _ 34 Scopolamine (hyoscine) _ _ 35 Local Anesthetic and Related Agents Procaine Analogs and Related Compounds bupivacaine (Marcaine) _ _ 36 chloroprocaine (Nesacaine) _ _ 37 etidocaine (Duranest) _ _ 38 lidocaine (Xylocaine) _ _ 39 mepivicaine (Carbocaine) _ _ 40 prilocaine (Citanest) _ _ 41 tetracaine (Pontocaine) _ _ 42 Vasopressors/Vasoconstrictors (Antihypotensives) ephedrine _ _ 43 epinephrine (Adrenalin) _ ‘_ 44 mephentermine (Wyamine) _ _ 45 metaraminol (Aramine) _ _ 46 methoxamine (Vasoxyl) _ _'47 norepinephrine, levarterenol (Levophed) _ _ 48 phenylephrine (Neo-Synephrine) _ _ 49 13. Type of delivery 1. normal vaginal (53) 2. forceps 3. breech 4. Caesarean section 5. other 14. Number of complications during pregnancy (54-55) 231 15. Number of complications during delivery 16. Comments on Labor and delivery, the severity of any complications, etc.: (56) l7. Medications given during pregnancy: RECORDER (57) 232 Mother-Infant Project: Brazelton Behavioral and Neurological Scale ID DATE RATER (1-4) (5-10) (11) Sex 1 F Age days (13-14) 2 M (12) Apparent Race 1 W (15) 2 B 3 H 4 I 5 O 6 Other Time of Exam (16-19) Time of Last Feed (20-23) Type of Feeding 1 Br (24) 2 Bo Initial State: Observe 2 minutes 1 2 3 4 5 6 deep light drowsy alert active crying (25) Predominant states (mark two) 1 2 3 4 5 6 (26) 233 Elicited Responses 1 2 3 4 0* L M H A+ (27) Plantar grasp l 2 3 Hand grasp 1 2 3 Ankle clonus 1 2 3 Babinski 1 2 3 Standing 1 2 3 (32) Automatic walking 1 2 3 Placing l 2 3 Incurvation l 2 3 Crawling 1 2 3 Glabella 1 2 3 (37) Tonic deviation of head and eyes 1 2 3 Nystagmus 1 2 3 Tonic Neck reflex l 2 3 Moro l 2 3 Rooting (intensity) 1 2 3 (42) Sucking (intensity) 1 2 3 Passive movement Arms R l 2 3 L l 2 3 Legs R 1 2 3 (46) L l 2 3 0* = response not elicited (omitted) A+ = asymmetry Descriptive Paragraph (optional) Attractive 0 1 2 3 (47) Interfering variables 0 l 2 3 Need for stimulation 0 1 2 3 What activity does he use to quiet self? (50) hand to mouth sucking with nothing in mouth locking onto visual or auritory stimuli postural changes state change for no observable reason U‘lthJNH COMMENTS: 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 234 Response decrement to light (2, 3) Response decrement to rattle (2, 3) Response decrement to bell (2, 3) Response decrement to pinprick (1. 2. 3) Orientation inanimate visual (4 only) Orientation inanimate auditory (4, 5) Orientation animate visual (4 only) Orientation animate auditory (4, 5) Orientation animate visual & auditory (4 only) Alertness (4 only) General tonus (4, 5) Motor Maturity (4, 5) Pull-to-sit (3, 5) Cuddliness (4, 5) Defensive movements (4) Consolability (6 to 5, 4, 3, 2) Peak of excitement (6) Rapidity of buildup (from 1, 2 to 6) Irritability (3, 4, 5) Activity (alert states) Tremulousness (all states) Startle (3, 4, 5, 6) Lability of skin color (from 1 to 6) Lability of states (all states) Self-quieting activity (6, 5 to 4, 3, 2, 1) Hand-mouth facility (all states) Smiles (all states) (51) (52) (53) (54) (55) (56) (57) (58) (59) (60) (61) (62) (63) (64) (65) (66) (67) (68) (69) (70) (71) (72) (73) (74) (75) (76) (77) 235 Mother-Infant Project Home Observation Form (1 of 20 pages, each with 24 observation periods) (Adapted from Parmelee/Beckwith, 1971) BABY STATE ACTIVITY” FEEDING CONTACT ID (1-4) DATE____(5-10) 013$ #___(11) v. #____(12) P. # OBS PER stress musc./adjust position affectionate touch/pat/rock active awake feeding bottle feeding breast feeding solids bottle propped held/pick up interfering touch eyes closed without movement cry eyes closed with movement drowsy quiet awake burping diapering bathing 0 g... 0 N O to O A O U1 C 0‘ 5.: H H N [_I u N H N N N w M b w H DJ M w w w b 001 002 003 004 005 006 007 008 009 010 011 012 013 014 015 016 017 018 019 020 021 022 023 024 236 m B. B. OBJECTS co OOSMHHQQM csoH so oconm uo OHUMH co >u mmHHEm mxooH mxo=m Ho:\mumm no: >MHQ uownno >MHQIMHmm space GOHumNHHmoo> mmmuumelcoc Amuov mmflm hmHm HMHUOm pownno nqu mmmmu .nno\3 m£u00m ..EHum .mmsoum mumsuo 0» meu wNHHmoo>\meu 0>Hummmc mcsmaaoo m>HuHmom mUCMEEoo 51 52 5 61 62 71 72 73 7 81 82 91 92 93 41 42 43 237 Know Your Baby Scale ID Although this is your first baby, you probably have some ideas of what most little babies are like. I am going to ask you several questions about your baby and I would like you to compare your baby to the average baby. For each of the questions I ask, let's say the average baby scores 20 units. I want you to give me a number that shows how your baby reacts compared to the average baby. For instance, if your baby does something twice as often as the average baby, you would say 40. If your baby does something three times as much as the average baby, you would say 60. If half as much as the average baby, you would say 10; if your baby does not do something at all, you would say zero. You may use any number you wish that shows how your baby com- pares to the average baby. Any questions? If the average baby is 20? a. How often is your baby hungry? (Probe meaning of response.) b. How much illness has your baby experienced? (Probe meaning of response.) c. How much does your baby sleep? d. How much milk does your baby take during a feeding? e. How much does your baby respond to strangers? 10. 11. 12. 13. 14. 238 ID DATE VISIT RATER If the average baby is 20? How much crying has your baby done? How often does your baby smile? How much trouble has your baby had feeding? How often does your baby coo, babble, make pleasant sounds? How much spitting up or vomiting has your baby done? How well does your baby play with you? How much difficulty has your baby had in sleeping? How much does your baby enjoy his/ her feedings? How much difficulty has your baby had How much does your baby enjoy it when you play with him/her? with bowel movements? How much trouble has your baby had in settling down to a pre- dictable pattern of eating and sleeping? How often does your baby sleep through the night? How much does your baby "cuddle?" Respondent: 1. mother 2. spouse 3. father of child/not spouse 4. boyfriend/not father 5. her mother 6. her father 7. his mother 8. his father 9. other (1-4) (5-10) (11) (12) (13-15) (16-18) (19-21) (22-24) (25-27) (28-30) (31-33) (34-36) (37-39) (40-42) (43-45) (46-48) (49-51) (52) Date 239 Mother-Infant Project: Home Environment Form ID (5-10) Hour (ll-12) Rater Age of Infant mo. People at home with child: I. Interview items: 1. The number of times the child was taken out of the neighborhood during the past week (shopping, etc.). The number of times that mother and child visited the neighbors during the past week. The number of adults living at home. The number of cats and/or dogs gt the home. The number of adults who actively take care of the child. Person A hours per day (22-23) (average) mother Person B hours per day (24-25) (average) (26) Person C hours per day (27-28) (average) (29) Person D hours per day (30-31) (average) (32) With how many adults does child interact daily? (1-4) (13) (14) (15-16) (17-18) (19) (20) (21) (33-34) II. 10. 11. 240 Does the child have a regular naptime (1), does he/she sleep whenever tired (2), or is this variable or in transition (3)? Does the child eat at regular times (1), does he/she eat on a demand schedule (2), or is this variable or in transition (3)? How comfortable is home for mother and child? (20 = average; 0 = totally unacceptable) How satisfied is mother that child is girl (boy)? (20 = average; 0 = totally dissatis- fied) Did some non-medical person attend the delivery? (1 = Yes; 2 = No) ll(a). If Yes, who? Spouse Father of child/not spouse Boyfriend/not father of child Her mother Her father His mother His father Other QOUI-bWNH Static Observation Items (where applicable 20 = average; 0 = bottom): 12. 13. 14. 15. Rate maternal speech for: a. Intelligibility b. Speed Number of rooms in house (excluding bathrooms & closets) Number of people in house Rate sound level inside home during observation Rate activity level in home during observation (35) (36) (37-39) (40-42) (43) (45-47) (48-50) (51-53) (54-56) (57-59) III. 16. 17. 241 Total number of people in house during observation How many toys does child have that make a definite audio/ visual response when activated? Descriptive Items Only (1 = yes; 2: 18. 19. 20. 21. 22. 23. No; 9 = DK) Is there a place where the child is put where he/she is out of earshot of noises of home and away from other people? Does the home have several small manipulatable items other than commercial toys that could be used as toys (i.e., ashtrays, bricabrac)? a. Is there a mobile over or near the crib? b. Does it move when crib is gently rocked? Is the child's room or the place where he/she sleeps decorated with pictures that stand out from the background of the wall? Are the child's toys kept in one place (1) or scattered all over the home (2)? 3 = None; 9 = DK Are the shades or curtains drawn all the time? (3 - half & half) (60-61) (62-63) (64) (65) (66) (67) (68) (69) (70) 242 M.I.T.S. (Bonem, 1978) Today's Date ID This questionnaire is designed to find out how your infant behaves in a variety of everyday situations. Read each item carefully and decide whether it is true or mostly true, or false or mostly false for your child. Mark your answer by putting a T (true) or F (false) on the line beside the item. Example: 1A. Infant often plays with his/her food. If your infant often plays with his or her food, then you would mark your answer like this: T 1A. Infant often plays with his/her food. Should your child not play with food while eating, then you would put an F beside the question like this: F 1A. Infant often plays with his/her food. Please read and try to answer all items. If a question is completely inappropriate, then you may omit it (be sure to skip the space on the questionnaire). If your child has outgrown an activity or behavior mentioned in an item, answer the item according to how he/she used to act. That is, if your child drinks only from a cup, answer items regarding breast and bottle-feeding from your memory of his/her feeding habits. We have worked very hard to make all items equally appli— cable to infants of both sexes and to those who have been breast or bottle-fed. Please do the best you can in answer- ing as accurately as possible. All information on this questionnaire is confidential, and will only be handled by the research staff, with no names attached. Please be sure to put today's date at the top of this page. If you have any questions, please contact one of the examiners. REMEMBER, ANSWER ALL QUESTIONS DATE 10. ll. 12. 13. 14. 15. 243 ID (1-4) RATER (5) VISIT CARD 1 (6) (7) Baby is irritable or cranky after sleep. (8) There is a great deal of fussing and crying with any illness. When waiting to be fed, baby is generally still. Responses to diapering and dressing are usually intense with much laughing or crying. When lights are turned on in the room, infant is usually not awakened. Child does lots of squirming or kicking while being diapered or dressed. Stops eating if hears noise such as bell, radio, etc. During play the infant is usually very active and vocal. When playing with one toy, the infant is easily distracted by another. Child kicks, splashes or wiggles throughout bath. Child usually fusses during diapering and dressing. (18) There is no clearly evident pattern in the time for child's bowel movement, it varies from day to day. Child's initial reaction to most foods, solids, liquids or vitamins, is to accept them without much fussing. Child is usually willing to be held and cuddled by strangers. Infant generally appears happy upon waking up. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 244 If playing with one toy, the infant does not usually become distracted by others. Infant's times for liquid feeding are unpredictable--vary more than 1 hour. Child's reaction to bath, whether she likes or dislikes it, is mild and not very excited. Infant's time of waking is not consistent from day to day (times vary more than half an hour). When child is with one person, she/he will easily go to another person. During diapering and dressing, child's expressions are mild--little smiling or fussing. (28) The infant shows discomfort with changes of place and situation even after con- tinued exposure. Child is generally happy when left alone in a room; will occupy herself. Child is a heavy sleeper; it takes a loud noise to wake him/her. Baby reacts to an undesired food in a mild way. Child notices and reacts to small amounts of urine in diaper. When crying from getting a shot, infant can easily be calmed by milk, pacifier, etc. If bath is given in new place, infant readily accepts change. Infant is active or playful on a fairly regular schedule. Child protested when put into bath for the first time. During feeding, the child will continue to suck even if there is much activity around him/her. (38) 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 245 If left on the floor, infant will usually move to another area. Before going to sleep, child is often fussy. Infant's general reaction to familiar people is intense-—crying or laughing. Child can be left on couch or chair for period of time without moving very much. Baby usually does not accept company (visitors). Child liked his/her first tub bath. During diapering and dressing, child is generally pleasant and smiling. During milk feedings, child is not easily distracted and continues to suck undis- turbed. Child does not seem to mind changes in amounts, kinds, or tastes of solid foods. Baby often consumes close to the same amount of food during a feeding. (48) Infant's general reaction to familiar people is mild—-frown or smile. Child generally indicates that he/she has soiled. Infant's play involves much movement and exploration. Infant is easily distracted during breast or bottle-feeding. Infant does not like to be bathed by different people. Infant does not adjust easily to efforts at changing feeding schedule. While playing with one toy, child can easily be distracted by another. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 246 Child will rarely allow strangers to hold or cuddle him. Child protested considerably to first bath. When going to sleep, infant is usually happy. (58) While playing, the infant is easily dis- tracted by everyday occurrences like the ringing phone or doorbell. Infant reacts to slight temperature changes (in room or outside). When given a food which he does not want, he reacts in a strong manner—- response is intense or powerful. Infant exhibits regular, easily identi- fiable actions around meal time. Infant falls asleep at about the same time most nights-—within half an hour. When lying in crib, infant moves around a great deal. Does not readily accept changes in types or characteristics of foods. (65) Child continues to object to grooming procedures (combing, washing, nail cutting, etc.) even after experiencing them several times. Sudden appearances of strangers will cause crying and/or a turning away. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 247 CARD 2 (7) When in carriage or stroller, baby is usually quiet and still. (8) Infant's general reaction to familiar people is friendly with laughing and smiling. Changes in lighting will not stop the baby's crying. Whether he likes or dislikes bathing, the infant's reaction is usually intense or energetic. When being washed or dressed, baby is generally pleasant, smiling, etc. Infant initially accepts new foods. When infant is full, he/she simply turns head away and lets food drool out of mouth. If left on the floor, child rarely moves from spot. Diapers must be heavily soiled before infant reacts. Baby often wakes during naps. When playing, baby will respond to hearing his/her name called. (18) Child is a light sleeper. It takes only the slightest noise to wake him/her. Infant shows a mild reaction to light or sound with little or no crying. Child has a loud response to a wet or soiled diaper. Child initially does not accept most new procedures; usually cries, fusses or does not cooperate. Infant takes nap at approximately the same time each day--within a half hour. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 248 Infant is generally fussy during play. After receiving a shot, it is difficult to stop baby's crying. In playpen or on floor, infant is active; gets into things, pulls at objects, or puts nearby objects in mouth. When the lights are turned on in his/her room, child is easily awakened. When infant cries because of hunger, she will usually stop for at least a minute if she is picked up, given a pacifier, etc. (28) Does not follow a regular nightly sleep- ing pattern. If there is any activity around him, child stops sucking during feeding. Infant can be fed at same time each day. When she is hungry, almost nothing can make infant stop crying. Infant notices and reacts to slightly soiled diapers. When there are interruptions in solid or milk feedings, the child generally remains happy- Diapers are usually very wet before baby shows any reaction. Even after several trials, infant con- tinues to reject most new foods. The baby will eat his meals at varying times during the week. Baby drinks a predictable amount of milk (If bottle-fed, varies less than 2 ounces; if breast fed, time sucking does not greatly vary). Infant usually lies still while being diapered or dressed. (39) 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 249 Child cheerfully tries new foods whether he/she ultimately likes them. Initial reaction to strangers is rela- tively mild such as a frown or smile. Child generally takes milk around the same time of day; does not vary more than 1 hour. Child still exhibits strong reactions even after repeated contacts with bright light or loud sound. Infant does not become easily accustomed to changes in caretakers--babysitters, grandparents, etc. Whether liking or disliking a food, baby's response to it is dramatic. Infant initially accepts any new procedure. Infant generally cries when solid or milk feedings are interrupted. In her stroller, infant is usually quite active or noisy. (48) Child makes himself at home most any— where; appears comfortable in new situ- ations. It is difficult to predict infant's activity or play time. Child resists going to different persons. When brought to the doctor for a well baby check-up, child is usually fussy. Baby adjusts easily to different care- takers. When left lying in crib, infant usually lies quite still. Infant's reaction to animals is intense with much laughing or crying. While playing child does not notice or react to his/her name being called. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 250 Baby readily accepts bathing by a new or different person on the first or second time. When diaper is wet or soiled, child makes no fuss or whimpers slightly. Infant cannot be left for very long on couch or bed because he might wiggle off. (59) Child wakes up from napping at approximately the same time every day (within half an hour). After 1-2 tries baby adjusts easily to changes in feeding schedule. The infant initially tolerates or enjoys new places and situations. When left alone for more than 5 minutes, child generally fusses or cries. Baby shows little reaction to bright lights or loud noises. Child usually indicates that diaper is wet. Child initially reacts to strangers with much laughing or crying. Infant will readily accept bathing by a different person or in a different place. (67) 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 251 CARD 3 (7) Child seldom seems to notice or react to differences in the taste, consistency, or temperature of foods. (8) Changes in sound (voices, TV, radio) will not stop baby's crying. Child usually rejects new foods. Infant takes 3 or more days to adjust to changes in daily schedule. At the doctor's for a well baby check-up infant is usually friendly and smiling. Baby smiles, gurgles, or plays with new people. Child seemed to dislike his/her first car ride. Infant does not readily tolerate or enjoy new places and situations. Child has low tolerance for pain. Baby lies fairly still while he/she sleeps. Child seldom or never indicates that diaper is wet. (18) Infant has no regular time pattern for napping each day (varies more than l/2 hour). While playing, infant generally does more quiet observing than active exploring. ' When engaged in play, baby is usually actively moving and making sounds Child seemed to enjoy his first car ride. Child seldom or never indicates that he/ she has soiled (b.m.). 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 252 Child will usually lie and watch a hanging mobile for just a short period of time (30 seconds or less). If an object is out of reach, infant con- tinues to reach for it for several minutes. Baby does not persist in attempts at turn- ing over, crawling, or walking. Gives up easily. When given a toy, the infant plays with it for many minutes. Baby usually fusses during bath. (28) Baby persists in attempts to crawl or walk until a few steps are taken. When seeing a new animal, child's initial reaction is one of interest and attrac- tion. When being washed or dressed, infant gene- rally cries or fusses. If child wants a toy that is on the other side of the room, he will crawl until she reaches it; or, if unable to crawl, child will continue to show an interest in it for quite a while. Infant cannot occupy himself in crib or playpen for more than a few minutes. The infant often continues playing no matter what goes on around him. Infant does not notice or react to changes in voice quality or level. When napping, baby almost always sleeps through without waking. Infant usually stops trying for a toy out of reach in less than 1 minute. When trying to turn over or crawl, child tries for a minute or two, then gives up. (38) Child is not very active during play. He displays few movements and makes few sounds. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 253 When given a food, he/she does not like, infant protests briefly but soon takes it anyway. Infant plays with one toy for only about 1-2 minutes. Diapering is often a battle. Even after first 2 weeks of bath, child continued to protest. Child will not crawl across room to another toy if there are other toys nearer to him/her. Baby shows little reaction to animals (dogs, cats, etc.). Infant explores very little; needs help to find play objects. In a crib or play pen, infant can amuse him/herself for quite a while. Infant usually plays quietly and calmly. (48) For at least 3-5 minutes, child will lie and watch a hanging mobile. Infant is usually cheerful during play; laughing, smiling, etc. If fussy during grooming procedures (nail cutting, hair brushing, etc.), baby is not easily distracted. (51) APPENDIX D SCORING AND INSTRUCTION MANUALS cow: mumum mo suaaanmq suflflanmuauuH mzoflasm mo suwofimmm ucwEwuHUxm mo xmmm mumum mo wmcmm com: 3002 mum“vum\¢AMHM\uumuN\m“HuH\m mnmumumuvnwumum\h «Num\m «Hua\m >ua>fiuo< mmcmmmo ufimuouuaasm augusumz mQCOB mmmcuuoad wuouwpzm can Hmsmfl> wuouwcsm mumsflc< Hmsmfl> mumEH24 muoufiosm mquHcmcH Hmsmw> mumswcmcH mocmsuowumm Mono: cowumucmfluo cow: xowumcwm Hamm mauumm unmaq coaumsuflnmm muoom Hmumsao wooomm Amhmd .COunNmum w EmuH wamom covawumum .mad .Hmummqv umumaau mamom couamumum may mo mcwuoom umumsao cm>mw u0w mwumuwuo 254 255 .mmuoom xmammu HmEuocnm mo “mason Hmuou u whoom xmammm .Hmfiuocnm ma m can amfiuo: mum N can .H .o mum£3 mzs Ho .mssmmummc .mscoHo ummoxm mmxwammu Ham now m no .H .0 mm conflwmo ma whoom HmEuocnw ad mwxmammm .5 humumuv.mumum«vnmumuhumumuaua.m cflxm mamom unwomum so mum uuw>cfl wmfl3umnuo “mono .H ma mmauumum :oflumHsmmm cmwz .ouw “.mumv mumxAmuHV Hua "uum>cH macaque owEocound .m I: cusoEIounocmm .. mcfiumfisvumamm nu mundaanomaoo mumum mo com: :1 mmwcfiacosu :oflumHsmmm .m muoom mooomm EmuH wamom couamnmnm umumsao HmDmSHU 256 The Instructional Manual For Assessing Infant- Caretaker Interaction in the Home was developed by Parmelee (1972) and Beckwith (1973). It is used in this study with changes that were suggested to the author (Beckwith, 1979, personal communication). All changes are enclosed in brackets ([ ]). Instructional Manual for Assessing Infant- Caretaker Interaction in the Home This manual defines procedures for assessing infant— caretaker interaction in the home. The observations are based on continuous 15 second time sampling. The behavior in the following category is mutually exclusive: baby's state. Once a behavior in a mutually exclusive category is chosen, that category is not marked again until a change occurs. The more transitory behaviors are not mutually exclusive and are marked within each time period that they occur. The coding assumes that the inter- actions are between the baby and [its mother]. A set of modifiers, to be added to the coded behavior, is used when the baby interacts with someone other than the [mother]. In order to record dyadic interplay a modifier is used to indicate the contingency of the response for either member of the dyad. A group of modifiers is available for use when an individual's behavior is in response to the baby's signals. For example, modifiers indicating con- tingency are used when the primary caretaker imitates the 257 baby's cooing or when the primary caretaker picks up the baby in response to the baby's fussing. In similar manner, a set of modifiers is available to indicate when the baby's behavior is in direct response to another person's behavior. For example, if the primary caretaker says "say baba" and the baby babbles, or if the primary caretaker interferes with a baby's playing and the baby starts to fuss, the contingency of the baby's response may be indicated in such a way that they are both mutually involved in the same act, then a modifier indicating mutual- ity is used. For example, when mother and baby are engaged in mutual visual regard or when primary caretaker and baby are cooing to each other, the code for mutuality is used. Definitions of the modifiers are given after the defini- tions of the behavioral predicates. The occurrence of a behavior without a modifier is indicated by checking in the appropriate box on the obser- vation form. To indicate a modifier rather than check, note the number of that modifier in the appropriate box. Only one modifier may be used per box. If [mother] and some other person interact with the baby in the same way in the same time period, score only [mother]. State A. These categories are mutually exclusive. Score once and do not score again until state changes. 258 01. Eyes closed without movement. Score when eyes are 3/4 or more closed and when there is no movement of head, trunk or extremities. There may be movements of face, e.g., grimacing, sucking, smiling. 02. Eyes closed with movement. Score when eyes are 3/4 or more closed and there is any movement of head, trunk, or extremities. There may or may not be movements of the face. 03. Drowsy. Score when eyes are neither 3/4 open nor 3/4 closed or are opening and closing with or without body movement. 04. Quiet awake. Score when eyes are 3/4 or more open and there is no movement, at all, of head, trunk or extremities. There may or may not be facial movements such as sucking or smiling. 05. Active awake. Score when eyes are 3/4 or more open and there is any movement of head, trunk or extremities. There may or may not be facial movements such as suck- ing or smiling. 06. Cry. Score when baby emits sustained distress vocali- zations associated with vigorous, diffuse motor activity. Fuss (#61) and cry are mutually exclusive. They may not be scored in the same time period. Activity [A. These categories are mutually exclusive. Score a l to begin and score again, with a 2, only when an activity 11. 259 stops. Includes only direct interaction with the infant; does not include preparation or caretaking of the environment. Burping. Primary caretaker pats infant vigorously on back while infant's torso is in an upright position. 12. Diapering. Primary caregiver is involved in all those 13. activities (removing diaper, wiping/washing infant, powdering, putting on diaper, etc.) associated with changing diapers. Also includes dressing. Bathing. Scored from the time that infant comes into 21. [22. contact with bath water. Also scored for sponge bath or other extensive washing of the infant. Ends after drying is completed.] Feeding These are marked each time they occur. Feeding bottle. Primary caretaker must be holding bottle but need not be holding baby. Feeding breast. Score when caregiver offers a breast 23. to the infant.) Feeding solids. Score when caretaker offers solids. 24. Primary caretaker can either feed with spoon or offer finger foods. Bottle propped. Primary caretaker holds neither baby nor bottle and baby does not hold bottle himself. 31. 32. 260 Contact These categories are not mutually exclusive. Score any category which occurs in the time period. Score each time period the behavior occurs, even if the behavior continues from the previous period. Hold/Pick up. [categories combined] Hold: Caretaker holds infant in any position so that the primary support for the infant is provided by the caretaker's body or extremities. If the infant is held in a sitting or standing position and partially supports himself, also code stress musculature (32). If mother uses hold to interfere with baby's activity also score interfering touch./ Pick Up: Caretaker lifts baby more than momentarily. If caretaker holds or places infant in sitting or standing position so that infant is required to at least partially support his own weight, also score stress musculature (32). Stress musculature/Adjust position. [categories com- bined] Stress Musculature: Mother holds or places infant in sitting or standing position so that infant is required to at least partially support his own weight./ Adjust Position: Turn baby from side to side, adjust in infant seat, move from caretaker's hip to shoulder, or move quickly from one place to another in room. Include adjusting position that is part of social play (e.g., throwing in air). 33. 261 Affectionate touch/Pats/Rocks. [categories combined] 34. B. 41. Affectionate Touch: Kiss, pat, hug, nuzzle, etc. If affectionate touch occurs within social play (53) score both./ Pats: Caretaker gently pats baby with repeti- tive motion./ Rocks: Score if caretaker provides any rocking or jiggling motion by using her own body or holding the baby in a rocking chair, or moving a carri- age back and forth, etc. Interfering touch. Caretaker touches baby to distract or inhibit ongoing activity. This may include removing object from hand, hitting, or pulling back, etc. Verbal These categories are not mutually exclusive. Score all which occur in any time segment. Score each time period the behavior occurs. Command positive. A clear request to initiate action. A categorical imperative offering essentially no option for responding in any other way than that called for by the request. Indicates an act which the child must do but is not now doing, i.e., "Sit still"; "Come here"; Put that down"; ”Make nice to the doggie“; "Wave bye- bye." Suggestions to initiate action which are stated as questions or offers are scored as [talk/vocalize] #43--i.e., "Have a cookie, honey"; or "Are you ready to put your nightgown on now?" or "It is time for lunch." 42. 262 Command negative. A clear request to terminate action; 43. for example, "Don't touch"; "No! No!"; "Don't cry"; "Sh!" may be coded as a negative command or as a noise (43) depending on tone of voice. The difficult dis- tinction between a positive and negative command is based on the presence of an inhibitory word in the negative command. For example, "Be quiet" would be scored as a positive command while "Don't cry" would be a negative command. [Talk/verbalize] [categories combined] This category includes all verbalization that are not commands. It may include any of the following: Criticism and threats. Score when verbalizations are critical, hostile, derogatory, accusatory, belittling, or taunting. For example, "Your father thinks you are a moron," "You spit up on my dress," "The lady is going to take you back to the hOSpital if you are not good," "Can't you do anything right?" The decision to score a statement as criticism is made according to content of the statement, not the tone of voice in which it is made. Praise or approval. This category includes state- ments which enhance self-esteem. Score statement of praise, approval, and encouragement. For example, "You are getting so big," "You smell so sweet," "I love you," "You can do it." 44. 51. 263 Comments. This category includes all verbaliza- tions which are not commands, criticism, praise, or instructional comments. Instructs. This category includes statements which label. Also includes reading or showing pictures with discussion. For example, "That's a doggie," "This is your nose." Noise. This category includes sounds made to the baby which do not contain words. For example, humming, clicking, imitation of baby noises. Score also for singing. Laughs. Score only when caretaker laughs during face to face interaction with the baby. Talks to others. Score only when caretaker talks to others (including observer or animals) when holding baby or immediately adjacent to baby. Objects These categories are not mutually exclusive. Score all which occur in any time period. Score each time period the behavior occurs even if it is a continuation of a transaction from a previous time period. Arouse/stimulate/soothe with objects. Caretaker pre- sents object to infant in a manner which encourages 52. 53. 264 visual and/or tactile exploration. Also score if care- taker presents an object which is not within sight or reach but provides auditory stimulation. [Includes offering pacifier.] [Also includes sustained, shared play, or interaction between baby and other person involving an object. This category may include demon- stration of the uses of an object (e.g., building a tower with blocks or putting a spoon in a cup).] Tease with object. Playful or hostile interference. Score when baby reaches for object and caretaker moves it just out of reach. When baby is laughing and seems to be enjoying the exchange, add a modifier (4) to indicate the teasing is mutual. Social play. Caretaker arouses or stimulates infant without using an object. Does so by engaging in ani- mated play with a fun-like quality where the apparent attempt is to amuse the baby or to play a game with recognizable rules (e.g., pat-a-cake). Do not score if interaction is purely verbal. In addition to scoring social play, score all component behaviors. If baby responds, use a modifier (4) to indicate that the play is mutual. Baby Vocalization These categories are not mutually exclusive. Score each category which occurs in a given time period. Score each time period the behavior occurs. 61. 62. 265 Fuss. Score when baby emits distress vocalizations of lower amplitude than a cry, not characterized by vigorous diffuse motor activity. Do not score in same time period as cry #06. Non-distress vocalization. Score when baby emits 71. vocalizations such as cooing, grunts or babbling but which excludes crying, fussing, or vegetative sounds such as burps, coughs, hiccoughs, etc. Baby Activity These categories are not mutually exclusive. For any time period score each category which occurs. Mouths. This category is scored when baby introduces 72. part of self or an object into his mouth. Self plgy. Self play is attention directed toward parts 73. of own body. This is focused exploratory behavior directed at oneself and is defined as looking at or exploration of one's body. Object play. Score when an intentional interaction 74. with an object occurs. Must include visual as well as haptic exploration, i.e., looking at a rattle and shaking it as opposed to stepping on it or shaking it without visual regard. Not eats--not sucks. Score when baby is offered food or pacifier and baby refuses by spitting out, closing mouth, or not sucking. B. 81. 266 Nonverbal Communication These categories are not mutually exclusive. Score each category which occurs in a given time period. Score each time period the behavior occurs. 82. 91. 92. 93. Looks. Score only when baby is looking at a person. Use modifier to indicate person at whom baby is look- ing. Mutual visual regard with caretaker is scored with modifier 4. Smiles. This behavior can be scored when baby is awake or asleep. Use modifier to indicate recipient of smile. Smiling subsumes looking. [Environmental Stimulation These categories are not mutually exclusive. Score a l to begin and score again, with a 2, only when an activity stops. Television on. Scored when television is turned on, with or without sound. Radio or Phono on. Score when radio, stereo, tape deck, etc. are emitting sound. Score when television sound is on without a picture. Other noisy appliances on. Score when mixer, blender, vacuum cleaner, or other appliance which creates an intensive noise is on.] 267 Modifiers Behaviors directed toward the baby The codes for the modifiers to be used with behavior directed toward the baby (categories 21-24, 31-34, 41-44, 51-53) are as follows: 1. [Mother] contingent to distress. Score when behavior occurs in response to baby's fussing (61) or crying (06). May also be scored if behavior occurs in response to other instances of distress such as choking, fall- ing, etc. 2. [Mother] contingent to nondistress. Score when behavior occurs in response to baby's smiling (82) [or other non- distress behavior.] 3. [Mother] contingent to vocalization. Score when behav- ior occurs in response to baby's nondistress vocaliza- tion (62). The response may or may not be directly imitative. 4. [Mother] mutual. Score when baby and primary caretaker are both engaged in [any mutual activity] regardless of who initiates it. 5. Other [person.] Score when [person] other than [mother] is interacting with baby. 6. Other [person] contingent. Score when [person] other than [mother] responds to any behavior of baby such as fussing (61), crying (06), smiling (82), gesturing (86), or nondistress vocalizing (62). 268 Emitted by baby The modifiers to be used with behaviors emitted by the baby (behaviors 61, 62, 71, 81, 82) are as follows: 1. Baby's behavior directed to [mother]. Score when baby looks at (81) or smiles at (82). Use modifier 4 when behavior is mutual (such as visual regard or smiling) or when baby acts in compliance with or response to primary caretaker's behavior. [2. Baby's behavior directed to other person. Score when baby's behavior is directed at person other than [mother] or observer.] 3. Baby's behavior directed to observer. Score when baby initiates behavior to observer, such as: looks at (81), smiles at (82). 4. Baby's action is contingent to [mother's] behavior or mutual with it. Score 4 for mutual visual regard or mutual smiling. Also score 4 when baby complies with a command, or responds to primary caretaker's behavior. For example, use 4 when baby offers object in compli- ance with a positive command, or when baby vocalizes in response to primary caretaker's talking to him. 5. Self. Score when baby mouths (71) part of self, such as thumb sucking. 269 Definition of duration of observation at l! [2], and 3 months Observation should start when baby is waking up and go [to the beginning] of the next sleep period. The obser- vation must include a milk feeding and awake time before or after feeding. 270 Michigan Infant Temperament Scale Scoring Key (Bonem & Zucker, 1979) Below is the scoring key for the Michigan Infant Tempera- ment Scale (Bonem, 1978). This scale consists of 164 true— false items and yields scores on 10 temperamental attributes. These attributes are based on those proposed by the New York Longitudinal Study (Thomas 33 31., 1963, 1969, 1977). All items on the scale are to be scored with a "l" assigned to items marked True and a "0" given to items marked False by the informant. Items that have been skipped should not be scored. Items listed below that are followed by an * should be scored in the reverse manner: False = 1 and True = 0. Items for each scale should be summed and divided by the number of items in the scale; divide only by the number of items endorsed by the informant. Thus if the scale contains 17 items and 2 of them were skipped, the summed total should be divided by 15. Please note that this scale has been designed to be used with infants age three to twelve months. The M.I.T.S. has been found to be a statistically reliable and temporally stable instrument with subjects of this age. Use of sub- jects older than 12 months is, at this time, not recom- mended. Below are the M.I.T.S. items organized according to the scale on which they are scored. Note that a * means that reverse scoring is to be used on that item. Activity 3, 6, 10, 35*, 44, 57, 61*, 68*, 79, 92*, 101, 107*, 112, 130*, 133*, 134 N = 16 Mood 1*, 2*, 11*, 15, 23, 33*, 38, 51, 62, 65, 77*, 87, 93, 100*, 105*, 116), 125 N = 17 Intensity 4, 8, 18*, 21*, 25*, 34, 42*, 54, 64, 67*, 73*, 74, 94*, 98, 108, 111*, 119 N = 17 Threshold 5, 24, 26*, 43*, 53*, 69, 72*, 80*, 86*, 88, 102, 117, 118*, 121, 129, 131, 136 N = 17 Distractibility 7, 9, 16*, 27, 31*, 39*, 45, 48, 52, 63*, 71, 78*, 81, 83, 85*, 109*, 122* N = 17 271 Rhythmicity 12, 17*, 19*, 29, 41, 55, 56, 70*, 76, 82*, 84, 90*, 91, 95, 103*, 113, 132 N = 17 Approach 13, 14, 30*, 36*, 37, 49*, 50*, 60*, 66, 75*, 99, 115, 123*, 126, 127*, 128*, 135 N = 17 Adaptibility 22, 22*, 28, 40, 46*, 47*, 58*, 59*, 89*, 96*, 97*, 104*, 106, 110, 114, 120, 124* N = 17 g (an experimental scale attempting to measure a form of "persistence") 138, 138*, 140, 141*, 142, 143, 145, 148*, 150*, 151*, 152*, 153*, 155, 156*, 157*, 158*, 159*, 163 N = 18 2 (same as P , an experimental scale looking at an aspect of "persistenée") 137*, 144*, 146*, 147, 149, 154*, 160, 161, 162, 164* N = 10 lSee unpublished masters thesis (Bonem, 1978) for complete discussion of "persistence." APPENDIX E CORRELATION TABLES FOR THE RESEARCH MODEL AND GLOSSARY OF TERMS 51 52 APPENDIX E CORRELATION TABLES FOR THE RESEARCH MODEL AND GLOSSARY OF TERMS Glossary of Terms Used in Computer Analysis maternal age x1 = mother's age at the time of delivery (AGEMO) socioeconomic status x2 = the highest occupational prestige rating among the household members (FAMOCCUP) x3 = maternal education level (EDMO) x“ = family income, total income from all sources of all family members in the household (divided by 1,000) (FAMINC) prenatal preparation y = did the mother participate in a birth preparation class (1 = yes, 2 = no) (PRENATMO) yz = month of pregnancy in which prenatal care began (PRENATF) neurological and behavioral status at birth ya = Brazelton Neonatal Behavioral Assessment Scale (BNBAS) orientation summary score (ORIENT) y“ = BNBAS motor performance summary score (MOTORPER) ys = BNBAS range of state summary score (RANGSTAT) ya = BNBAS regulation of state summary score (REGSTAT) y7 = BNBAS autonomic regulation summary score (AUTOREG) 272 03 05 273 ya = BNBAS reflexes summary score (REFLEXES) Note: The BNBAS habituation summary score was removed from the model due to an excessive amount of missing data physical indicators of neonatal status Y9 Ylo Y11 Y12 Yla Y1» Yls Yls Y17 birth weight of infant, in grams (divided by 1,000) (WTINF) gestational age at birth, in weeks (GESTAGE) 5-minute Apgar score (APGARZ) - maternal attitudes toward child positive attitude index from month (POSATTl) negative attitude index from month (NEGATTl) positive attitude index from month (POSATTZ) negative attitude index from month (NEGATTZ) positive attitude index from month (POSATTB) negative attitude index from month (NEGATT3) infant temperament Yle Y19 Yzo Y21 Y22 Y23 qu st Michigan Infant Temperament Scale (MITS) activity score (ACTIVITY) MITS mood score (MOOD) MITS intensity score (INTENSE) MITS threshold score (THRESH) MITS distractibility score (DISTRACT) MITS rhythmicity score (RHYTHM) MITS approach score (APPROCH) MITS adaptability score (ADAPT) 274 n6 = home environment stimulation st = the number of rooms per person measured at 3 months (RATIO) Y27 = number of caregivers reported at 3 months (ADULTS) ng = is there a quiet place where the child can be put when sleeping? 1 = yes 2 = no (QUIETPL) ng = number of audio/visual response toys at 3 months (TOYSl) n7 = quality of the mother-infant interactions (all 3 months) Yao = percentage of time mother held child while he/ she was awake (HELDMRZ) Y31 = percentage of time mother talked to child (TALK3MR2) Y32 = percentage of time child vocalized (NDVOCR) y33 = percentage of distress periods responded to by mother in 45 seconds or less (FUSSRAT) yau = percentage of time spent in mutual gaze (LOOK4R2) Y35 = percentage of time spent in feeding and care- taking activities, i.e., feeding, burping, bathing and diapering (FEDCRRZ) A's (lambda) = regression coefficients relating unobserved or theoretical variables to observed vari- ables ¢ (phi) = covariance of the exogenous theoretical vari- ables and y's (gamma) = regression coefficients relating exogenous, theoretical variables ( 's) to the endogenous variables ( '3) 8'5 (Beta) = regression coefficients interrelating the endogenous theoretical variables 6's and e's (delta and epsilon) = measurement errors in observed variables, x and y, respectively c's (zeta) = errors of prediction 2'755 Ono. Ono. o.~.. ..... -unauua. .m.. ~o~. .c.. “Nu. nn.. 0... .ou. n... a... .s.. .N.. no.. a... cw... .0... no... .0..- ..N. no..- 3:... kn... n... no... 00.. Oh... .NN. u...- ..n. s... ..... cuaouau. .ou. .m..- o... m...- ..... .s.. o... .s.. .NN. 50.. .N.. N... .... n... usu. 5... cs..- nn.. .... a... o... nn.. 0... l...- N...- n... .... .u... .... a...- 0... ON... .n.. J... a... ..... 0.0.3... N... ..n. on~.. onn. .No. ..... .u.uuuu. .5..- ..... n...- kn..- an.. .1... ..N. n.... m...- .0... o... .0... c... 0.... on... n... .9... n...- N.n.. .... n...- no.. O...- O...- n... .n... .a... Q~.. n.c.. .N.. u...- ..... «unmana- .s.. a... .sm. no... man.- .n... ..... -nuucuu. a... awn. s~.. JNN. .N... ~0.. s...- .n.. 00.. n...- .... N...- n... -.. s...- O... O... a... to... N...- s... a... 0.... on... no... 5n.. .no. .N.. n... a... N...- ..... -uunaun. .-.. no... u...- m... on... .0... .nn.o ..... un...anu omm. n... naw. a... .n.. -.. .~..o a... .s.. as... m...- ..~.. s...- ..~. .m.. .n~. a... a... no.. now.- u... no... nn.. u...- .n.. n...- ..... as... .sw. o... mu... nu... no... ..... c...m.«. 40.. 5... .~.. on... onn.- N... mac. ..n.. ..... -Nuudaua sn.. 5... n... on.. u... no..- a... h...- a... nu..- 5N.. .0..- on... so... an..- ~N.. .n... n... .0..- so... .n... nnu. N...- n... ....- a... NO..- on.. c...- n.~. a... O...- n... a... ..... udunaaom 0‘ 0. n8.u ..~.. ..~. 9.... ....- ..~.. m... .~..u 9.... uuuucnau m... g... nn.. an». .n..- sun. can. a... mop. ....- a... snn.- n... nun. 5n... .... an... sn.. 4... .n.. o...- a... a... n.~. a... .0.. an..- nan. a.~. a... a... .n.. m...- a... nu.. ..... momma... a... ..~. :~.. 9.... .5... .~.. .cm. nnn.- um». ~.n.. ..... -dH..uu. on... .0.. o...- .... a... so... .0.. nu... n...- N...- .0... .0... a... a... ..N.. s...- s... n...- .N.. NN..- no... .n~.. ....- o... o...- .o.. s.n.- o... s...- a... -.. a... a... s... .0... ..n. ..... oo....~a .na.. 0...- an... .0.. sow.- .J... ONQ. ohm. ..... mom. .m... ..... -«uucmau Noe. 21h)!“ .~.. hunnbwno swo. Swans. d... umlwpln «on. 9098 .~.. >r>~>u¢ N... MPhC¢U8 ..n. nbhnm96 o...- whhccul .~.. NrbnnOt .~.. .rhncwl O... .hhcmOC ..... Ntuat. IINuduu. Omupoctq a. or andbct .sw.- .5N.. .swo. :.n.. u...- on... u...- n... on... on..- 0...- nnn. s...- o.~.- .on.- on~.- CON.- o.n.- 0-.. o... .n... no... any. non. a~.. mJJ. m...- 0... sn.. 0.... an..- n... nan.- nu... on... a... «n..- ..... 0......» s»..- 98.1.. smo.a oxau nan.- saoocxcs .~..o cumu- o... usage»; 5.... wcaxoo. on... pecans. .o~.o coo>az .n~.. «urns... 0... ~23! n...- .m>op mom. Jcpunao a... npaao. Cu: or! o... pg... .0... zoos... s.~.o rxppxu oo~.u rum.o o...- smusxp .o.. untapxn a... 999: so... ppauru. on... nrh.uuz s... nrpqmos o~n.o ~ppqcuz sou. whpcmot on~.c .rwgcaz a... .p».mot .-.a naguga on... accumuo we... can»: ma... nuxwauw. on... nutcrac os~.a pcuwcuc s... pcrn08¢c s-.o .utcoro: u... rzuu.u .n.. urns... .9... otpgxuan 9......» .000: nouuouom as» new 00.nda co.u¢.vuuoo hxut.aam>mn admxo :0 ma. Juzuwptt no bonus. .2» no auubcaat>w 8C 27765 .0..- nn... a... ..... o... .n.. a... ..... n.N. .9... ounce... .o... .a..- .o... nu..- .~... .n.. ....- .~.. .~.. n.~. ~... ~... u... .e... .m... s»..- ..... -.u....u N... naN. n... 0nN. 0...- .... asN. 0.N. 0... .0N.- ..... ---.w... s... so..- 0N... n00. n...- .oN. .0.. .n.. .n.. .o.. .o... NN..- n... 00.. .Nn. n$~. n00. ..... o.--ana. 00N.- .N0.- 0.N.. .N..- .n.. 0.N.. 9.n.. n...- a.n.- 0n... 0.N.. ..... nauuuus. 0... n... a... n... .s.. 0n.. .0.. Nn..- on.. 0...- w.N. .55. $0.. 0.9. 05.. ~0N. NNu. a... ..... Nana...» --w.aaa. ..... N... a... .0.. .0... o... .0... .N..- 0.N.. Noe. Hoe. .m... No... Nn... Os... N... s... nn.. .0.. ..... .uoonum. ~00. No0. o.n. man. a... .N... a... .... .0.. 0.... n... .N0.. .N... ..... -o-..... nnN.. n...- .0... N...- .NN.. ..... .mN. m... .nN. no... N... n... s... .nNo. m.... .d«.. a... n...- .n... NN.. ..... -Nmuuamu a... n... .0N. n...- n...- 0nN. .s.. N.N.. .... .5... .N.. n...- n... 0... ..... -a..u.n. .... 0N.. a... n... Nmuo. n... .N..- .N.. 0.... us... nn.. so..- 0.... a... s... .N.. ~0N. 90.. .N.. a... n... .9... -N....aa n... nN.. .m.. 0... on.. nan. a... 0...- s... .N.. .0... .0..- .a.. .m.. 00.. .9... .n~.- ~m..- a.... an... m.~.- so..- a... a... ~od. .~.. ~...- s... _...- 3.... ..~.. m-.. .on. 9.0. 0.N. osn. ....- .o.. 9.... .u..nwam s... ..N. ..N. N... .o... ~.N. 0N.. n...- .N.. 0... s... N...- s...- ..N. .n0. nNN. ..... .-.au.a. ....o a... o... ~m... .n... a... ....- nu..- ..~.. me... ~... as..- m.... ....- 0.... .w... a... r... so... o... .a..- s... n... 9.... --.oo.aa ou~>..z. o...- 00.. 00.. 0... .so. NN... 00... N... O... h...- N... .N... s...- 0... .o.. a... 0n.. 9.... no.3Hmua aun»..2. No... a... owe. N... 0.... N... .... s...- on~. s... «we. ..~.. 0.... mm... om. .- on..- o... sa~. a... oo~. a... a... a... emu. .o... NupnuJou 92.3.. 01.. .38.... came. wrcuowu «assoc. h..ww:u 3069.8 ~¢tnx..p a. or u....: s... 0.N. .0.. 00.. a... .N.. N... .N.. ....- o... N... 00... s...- nnN. NoN. .N.. .nn. .... .o... -oawuua. u:.t.u atom aauoox88 po¢¢haua :nu.x. w. a» .uubct m... 00.. o... n... .0... .nu. am... an.. o... a... .N... .N... ::~.. .N... o... a... 92.x.u oxau uaouotcu at... .muucwb ugcxuo. p.4wmau .oo>ux ~wtnx... ~1¢oawz .mtoh asbu.:o my 4.3. 6.».- be... 136333. APPENDIX F LISREL ANALYSIS OF RESEARCH MODEL APPENDIX P LISREL ANALYSIS OF RESEARCH MODEL ANALYSIS OF LINEAR STRUCTURAL RELATIONSHIPS BY THE METHOD OF HAXIHUH LIKELIHOOD DISTRIBUTED BY- INTERNATIONAL EDUCATIONAL SERVICES 1525 EAST SSRD STREET CHICAGO- ILLINOIS 6015 (312) 684-4920 THIS COPY AUTHORIZED FOR USE ONLY BY SITE 1161 UNTIL HONTHIDAYIYEAR = 4/ 1/81 PROGRAM COPYRIGHT HELD BY NATIONAL EDUCATIONAL RESOUR CES- INC.- 1972 DISTRIBUTED OR USE UNAUTHORIZED BY NATIONAL EDUCATIONAL RESOURCES. INC. IS PROHIBITED. 277 278 L I S R E L I V AN EVALUATION OF THE IMPACT OF MATERNAL AGE ON CHILD DEVELOPMENT NUMBER OF INPUT VARIABLES 39 NUMBER OF Y - VARIABLES 35 NUMBER OF X - VARIABLES 4 NUMBER OF ETA - FACTORS 7 NUMBER OF KSI - FACTORS 2 NUMBER OF OBSERVATIONS 62 MODEL SPECIFICATION LAMBDA Y FULL - FIXED LAMBDA X FULL - FIXED BETA FULL - FIXED GAMMA FULL - FIXED PHI SYMM.- FIXED PSI DIAG.- FREE THETA EPS DIA6.- FREE THETA DELTA DIA6.- FIXED OUTPUT REQUESTED MATRIX TO BE ANALYZED YES TECHNICAL OUTPUT NO STANDARD ERRORS YES MATRICES OF T - VALUES NO CORRELATIONS OF ESTIMATES NO SIGMA. RESIDUALS- ETC. YES FACTOR SCORES RESRESSIONS NO FIRST ORDER DERIVATIVES NO STANDARDIZED SOLUTION YES 279 a... h... 000.- 0.h.N 1.00.... 000.0 .00.. 000.0 .0... .00. .N.. .00. .N.. ..h. 0.N. .00.. NN.. hnN. 0.N.- 000.- ..No- 000.- 00.. 5N0.- ..0.- N...- .N.. .0..- 000.. 000.- 0.N.N N00.- 00..N .0.. h00oN luudnwu. ..h. ..N.- .00. .0..- 0.0.N III-a... 000.N. 00N.» 00h.bN N0..0 0.0.. 0... ..0.0 00N.. .0..N 00... .N..N .N..- .p.. .N... N0... 00... 0.0.. 0.9.N n...- 000.- 00.. 00s. 00..N- 00..N .0..N- 000.0 ..0.- s0... 00.. 00..0 #0.... Ilium... 0.N. F0... .0..- .0.. 00.. .00.N .50.- an..- 000.0- .0..- 00.. .0..- .00. .0..- ..0.- 0.N.- 0N0. 0...- 00.. 0.0.- .0N.- 0... .00.- .0N.- 0.0.- .N.. 0.N.- N... .0..- .0..- 0... 0...- N»..- 0... 0.N.- .0N. ..0.- .0... manna... 000. N00. .N0.. 0N..- 0N0..- 000.- 00... 0.... N0... .00.. 0.... 0.N.- 00.. 00..- .0.. 0N.. 0...- .00. .0..- .0.. 05.. 00..- .0.. 0.0. 0.0. .N..- N0..- 00.. h... 0.0. 0.0.- .0..- 0.N..- 0... N0.. b... 0... 000. h...- ..0.. uuuuaaa. 00..N- .hn.N- unnoN- 0.0. 000.- 00..- 0....- .0..Nn -HHH<.qu ..0.0 0.». 0s... .0.. 00.. .N.. ..0.- 0... .0.. N.0.- N.0.- 0...- N0..- 0.... 000. 000. 00.. .... 000. 0.0.- N0.. 000.- .... 00..- .0.. .N..- 0.N.- .00.- N.N. 0... 0...- 0NN.- 0n..- N0... I...uuu. 0.... N0... .0.. .0..- 000..- .0.. Nnh.0 000.0- 000.0 N00. 00.. .00. 0.0. 0.N. .0..- 0... 00..- 0.0. 00..- 00.. 0...- 00..- 0NN.- 00..- N... .0..- h.N. 0.N.- hNN.- 00..- .00. .hN.- 000. b0N.- 000. n...- 0N.. 00..- 0.0. .00. 0...- h... 00.. .0.. ..H..... 0.0.N- N.0.N- 000.N- 0.0.. .0..- 000.- 000.0- 000.0N N0..0- .9..N0 -NH..aqu 00... .00. 000.N 0.... n0..- 0... 000. 00.. .0.. 0...- 0.0. .0..- N... 0... 00..- n... .0..- ..N. 00.. .0.. 0...- 0... 0.N. n... 0N0. 0N0. hp..- .0.. 0... 00.. .00. NN.. ....- .0.. 0... .00. 00...... .00. 0.N.. 00.. 05..- 0...- 0.0. 000.0 0.0.0- h00.0 .0..0- .NN.. -.HH..ua ..0.- 000. 00..0- o... .0.. .0..- .00. 00..- 0N..- 0...- 0.0.- 0...- N.N. ..N.. 000.- 0N0.- h... 0.0.- .0.. .0..- .0..- 000.- N...- 0.N.. .00.- N... 0.N..- 0.0. .0..- 0N0. ..0. N... 000. 00.. N0..- 00N. .0... -H.u... 22p>z¢ .00....0 :wusz mutant. coo: >h>~h0¢ nhhcouz nhh000a th00uz thcmON .hhdouz .hbdmo. N00... 08.... a... 5:00.... .20.. Nccuouu N¢0xoaJ b.1003. ..0.0: N¢t0140h Nuxaauz .0>.b ashu.:. 0.43.. 0.... ha... 8...... £2.08- 00...... 20.82. unauhz. a... >h>~huc 0hp0008 0...... Nh»¢.ua Nhhdmon .hhcout .0..... N80... 0...... 58... .uxuguuc 0.00.20 #0.»... h0h008.¢ cusxohOt haw... uh.zu¢. .lhcluct hoNoc 00.. 0.0..- N... .Nno- 0a.. .00. 0... .N0..- N.N. 0.0.- 0... 00N. N... 000.N. 00.. .N..- 0...- 0.0.0. 0H0. N.N..l .0.. 000.0. .hN. ~00. -.H.408< u. or x.¢b.t nn..0- 00..»- 0»..- 0...- 0hh..- .00..- 0.0..- ..0..- .N0.- .0.. N0..- 0...- 00..- .0N.- .0N. 0.N.- ..0.- .N..- .N..- .0N. 0N0..- ..0.- 0... .0.. h...- 0... .00..- 000.- 000.- 00.. N00.- 00..- ..0.- 00N.- .0..- .N0.- 000.. .N..- ..N. 0... 00N.- .N.. 0.N.I n0..- 0... 00N.- .00.N 00k. 0... 0...- N.0.N .n... 0»... ..0.- 005. .00. .N.. h...- ....- 0.N.- 0nN.- 000.- N... 00..- 0.0.- 0...- 00..- .NN.- N0..- h... .0.. 0...- .0..- 0... 0.0.. 0... ..0. -......u a...au.u ou~>4¢x< u. .h x.xh.t hauta04u0uo 04—10 80 um. 402¢uhut a. h90a8~ at. no 2...<:.<.u z. 280 0.N..! ....! N.N. ....! 000.. 0... 0.0. 000.! 0N.. ..... Imuuuaun .0..N! ..0.! 0.N.0! 0N...! N0..! 0... 0...! 00.. .0.. .00. .N.. 0... .... .....! 0.0.! ..N.. 00..0.. Ila....u 0.0... 0...0 00.... .N... 0....! ..0. .00.. ..0.. 000.. .0..N! 0..... .0... ....! .0...! N... ....! ..N. .0.. .0.. .00. .N.. ..0.! ....! .0.. .0... 000. 0... 000.0. .00.. Iliad... .00.! 0...! 0N0.N! .0..! .N.. .0..! ....! 0...! .0..! .N..! 000.! 0... aauuudau 00... 00.. 0.... .0.. ..0. 0... 00.. ..N.! .00. 0...! .0..N .... 0.N. .... .00. .00. ..N..NN .0..00 0.N.... manna... N0..! ....! .0...! .0..! .0..0 00.. .0.. .0.. ....! N.0.! 0.0.. 00.. .... tlanaaad ..0.! N... .0..N. ..0. 00N.! 00.. ..N.! 00..! 0NN..! N00. 000.. .N..! ..N.! .00.! ..0.! 0... 00..0N ..0.0 00..0. 00...N ullqnuufl 0....N ..0.0 .0..00 ..0.. 00... 0.N.! 0.0.. 0.0. .00.. 0N0.! 0... 0.0.! .00.! 0...0N I:|..H.. 00..0.! .0..! 0.0..N! ....! ..0..! .0..! ..0.N N... ....N 0...! ..0. 0... 0... .00.0! .0N..! ....! .0.. 0NN.! 000..! .0.. N...N ..0.0 .... N0.... 0N..! 0...! .... 0.N. ..0.! 00.. 0.N.! 0... ....! .0N. 0N0. 0.... II!..... N...0 ..... 0...0. ....N 0.N.! 000. ....! 0... .00.! 0....! ..0.N 00..! 0...! .00. 00.. .... ....N .0.. 0.0.. ..0. 0... .00. N.... ..0. ..0.0 .00. .... .... 0NN. ....! 00.. .0.. N...! 00..! N... ..... N.... 0.0.. Inuauuu. 0...0.! 0N..! 000...! 0.0.N! ..0..! .00.! N...N .00. .0..N 0.0.. ..0..! .0.. 0.0.! ..0.0! .0..N! ..0..! 00.... .0N.N .0..N. 0.0.. 0.0.! 0.N. .0..0 IH....au N0... 0.... .0..0 .00. 0.0.! ..N. ..0. .0..! 000. 0... 0... N0..! 0...! 0.... .00.. ..0. N.0.N II....:. ..0..! 0... 000.. 0.0.! .0..! ..N. N0..! .00.! ..0..! 00..! ..... N0..! ....! ..0.N! .N0.! .0..! ~nn.~ ....n~ ecu. .9... ..~.- «no... a... pa... on..- a... a... n... a... a»... cup.“ a... -~.o unuuquaa «unna4.. au~»4..:. u. up x~gp¢¢ o.~.~u can.“ con." nun. op... ow... -o. a... one... n.... c... nun. oe~.~ u...- «.~.- a... a...“ on". «an. ..N. n-.~ «cu... on..- ma... 9.... on..- n~..~- co..- ..u.«. p... nu... «mg. 02.... ..0. .auu.t.. at... Nuguou. N808... ..c..:. 89...: N...u4.. 08.... ..0. .5006... .8... N8..... N808... ..1003. 89...: N8t.x4.. thoau: .0... ...u.:. 0.43.. o.... ...o. 8...... t:..:. ...8.... 2.0.8. 02.... .10. .30..... .1... Nexuau. N808... ..c..:. ..0..! N.:.xd.. Nutaau: .0... 4...... 0.43.. ..... ..... 1...... AN EVALUATION OF THE INPACT OF HATERNAL AGE ON CHILD DEVELOPMENT PAIAIETEI SPECIFICATIONS LAIOOA V KIA. PREMATNO PREMATF ORIENT HOTORPER RANGSTAT REGSTAT AUTOIEG IEFLEXES ITINF DESTAGE APOAI2 POSATTI NEGATTI POSATT2 NEOATTZ POSATTS NEGATTS ACTIVTV HOOD INTENSE THRESH DISTRACT RHYTHM APPROCH ADAPT RATIO ADULTS OUTETPL TOYS! HELDHI2 TALKSIR2 NDVOCI FUSSRAT LOOKle FEDCRIZ OOGOOOOOQOOOOOQOOOOOOOOOOOOOOOOOOO‘OL LANODA I ...E AGEIO FANOCCUP EDHO FAHIIC BETA KIA-J ‘0. D ‘9. £00 £0. E0. E0. 3. DGOUMO‘ E l I» 00069960909609BOOGOOGOOGCOQOUOUNOQO 851.: 29 30 {IA—2 35 39 281 E 000090099°°°°OO°°°°OQOOOOflOOOOCOODO .............L ......L E u OQQCO. NM” E OOOCOOUNHOOOOOOOOOOOBOOOCOQOOOBOOOQ u E 282 no“ can at . gigging: or no 00 no on gnu 34 I33 undue; «mad-.3 no No no .0 on O. h. 0. an g Ida tag '34:: ling III—Maud Quad 312.3 was and ««««««« no me no on on on up eh a» tug lg :3 gm Indiana lug sumac“ nag. sang «a: n» uh ah an to .0 no 00 no Jaggggaagagagu no no no on on a la «'13 «Ila «Ila all-in N ... o 3 «luau 14am 283 AN EVALUATION OF THE INPACT 0F NATERNAL ADE DN CHILD DEVELOPMENT STAITTNO VALUES LAHBDA T UL-1 ‘n--2 LIA“: $15-4 m-.§ EIA-J HA.-1 PREN‘T'D 10DDD DODDD D0000 0.000 D0000 D0000 DoDDD PRENATF AoDTQ DoDDD D0DDD D0DDD DoDDD DoDOO D0DDD DITENT D0DDD 10DDD DooDD DODDD D0000 D0000 D0DDD HDTOIPER DoDDD 90862 D0000 DODDD D0DDD D0000 0.000 KANDSTAT DODDD '1‘0399 DODDD D0... DODDD DODDD DODDD REDSTAT DODDD D30D94 D0DDD DODDD DODDD DODDD D0000 AUTDRED DODDD 28.008 DoDDD DODDD D0DDD D0DDD D0DDD IEFLEXES D0DDD '320D91 D0DDD DODDD DODDD DODDD D0DDD UTTNF D0... DODDD 10DDD DODDD D0DDD D0DDD DODDD DCSTADE D0DDD DODDD 10‘96 DODDD DODDD DODDD DODDD APDAIZ DODDD DODDD ‘1023D DODDD DODDD DODDD D0DDD PDSATTI D0DDD DoDDD D0DDD 10DDD D0DDD DoDDD D0DDD NED‘TTA D0DDD D0DDO D0DDD ‘0... DODDD DODDD DODDD POSATY2 DODDD DODDD DODDD 20DD1 DODDD DODDD DODDD ~EDATT2 DODDD DODDD DODDD '0‘5. DODDD DODDD DODDD 'OSATTD DODDD DODDD D0DDD 202A: DODDD DODDD DODDD N‘DATTD DODDD DODDD DODDD °0DD2 DODDD DODDD DODDD ‘CTTVT' D0DDD D0DDD DODDD DODDD 10DDD D0DDD D0DDD .000 D0... DODDD DODDD .0... 20.6. DODDD DODDD ANTCND‘ DODDD D0DDD DODDD DODDD .IOD3D DODDD D0DDD THICSN DoDDD D0DDD DoDDD DODDD 30D55 DODDD DODDD D‘STIICT DODDD DODDD DODDD DODDD :0DD‘ D0DDD DODDD .HTTHI D0DDD DODDD D0DDD DODDD 3099. DODDD D0DDD APPIDCN DoDDD DoDOD DODDD DODDD 30355 DODDD DODDD ODO'T DODDD DODDD DODDD DODDD D01D3 DODDD DODDD NAT‘D DODDD DODDD DODDD DODDD DODDD AODDD DODDD IDULTD DODDD DODDD DODDD DODDD DODDD '0D3D DODDD DUTET'L DODDD DODDD DODDD DODDD DODDD -0D7O DODDD VOYSI DODDD DODDD DODDD DODDD DODDD .595 DODDD "CLONIz DODDD DODDD DODDD DODDD DODDD D0DDD 10DDD TALKDIRZ D0DDD DODDD D0DDD DODDD D0DDD D0DDD AD02,3 NDVDCR DODDD D0DDD DODDD D0DDD DODDD DODDD 1025‘ FUSSI‘T DODDD DODDD DODDD D0DDD DODDD DODDD 1102.1 LDDKQIz DODDD DODDD DODDD DODDD D0... DODDD I073‘ "DCIIZ DODDD DODDD DODDD DODDD DODDD DODDD 'A0172 LOHDD‘ N 53L-1 DEA—3 ADIHD 10DDD DODDD "HDCCUP DODDD 10DDD ‘DHD D0DDD 01D, FIDIUC DODDD 0.5N DETA .1 LIA—.3 ‘15—} m-.1 m...) nu Ell—.1 £00 A 10DDD DoDDD D0DDD D0DDD D0DDD DODDD DODDD [.0 2 -1D0252 10DDD Q0967 DoDDD DoDDD D0DDD D0DDD [D0 3 '50750 DODDD 10DDD DODDD DODDD DODDD DODDD ED. Q -777052D -350887 2390933 TODDD 7038D DODDD ’270625 E00 5 D0DDD '60‘19 0286 DoDDD 10°00 DODDD DODDD [00 D D0DDD D0DDD DoDDD DoDDD DODDD 10DDD DODDD 284 canon ~no.~ nuuauauu ono.~ uuu¢nuuu oo9.nn ulnauuqn .——.~ .ulldaau ....«n unauauu: N00. «snow -~.~ nunuauun ”up. poo-0 knead launuHad suaon N.J.: coo.— nan... anuqnu4u on... unnuauqu ”pw.‘ un¢~auuu «no.~ acnquu coco! $0835 uaun to Zuhhuca «(a 29.50 .43.; to 298:3 u4~u to 8983; want 80 sabhuca ”(a 203:: new: to 8983.. want to Ion-N Gono— con-Q IMHHduuu .00. Hquuuaad 909. «.4: owcooc on~.co usuann0u no.- «so. uuuaauu4 ono.o nuuudaau .00. uuudqnafl ..o. «onqqu anacnl anhh~¢a m4: Zahhuxa «(a “0".“ lurk—ca ac: u turbaca «£3 a npc.~ nuuuauau no..« uuqunnau pco.~ ununuuua pee.~ naun4uua non.“ unuauuaq ..... nauqau aaoxa no ma maoxn no :0 $3010 to do s:o¢¢ no no 530.0 to :4 caoco no >4 can.n~c unuuqn U 8 hhbmbdux awhhhbhhh mzctu mt <000 nwoo no~0 oon0ul uO~0 «un0nn 00N0 0000! n—e0u Oueo «~90 «nu0l nau0 nauou IHQkuum ooa0l 0000 0090 non0n 060050 nau0ou IMHHduua «n00. Q~N0I Nu—0NI -n0o n.000 0900 one0 ”:90 hNo0 no.0 nn~0l Ouo0 .000 nh~0o 00000 0N00I una0o 0na0l 0N00I 0990 «N900 09.00 00.00 5000 90900 «000 00900 ~0u0 0990 00900 0.900 0090 onool swoon 0900 HQHnand O¢n0 nanoo ocwol «coon- oohonn oneoht .ho0nc IMHHdnuu sn00 can0 #000” nuuo no.0 noe0| «N000 «.90: 00000 «0000 no~0 nu00| ~900- 0000 unoo 00.0 «~90 .090 eu°0 0990! nao0 neeo $0.0 o~n0I #090 Non00 no.0 unnoo noo0I 0990 no.0 onooo 0Ne0 «090 00°00 ou00 «unwamuu 00000 0000 0000 n~m0u shuobo unno— hooool «.000 Ianuddua 0000 0u00 ”~00 nueo 90.0 99°00 «.90- .9900 «900: 9.000 0090 «9900 «a..- ”~00 00.0 N900 ”900 «.90 no.0 «9900 no.0 «090 no.0 on¢0o 0090 on00c 09.0 Oncol 0990. 9990 no.0 0090! nae0 0990 «9900 «990 cho0u IIHauHuq 000. 000.- 000002. 000.- 0.0. 0000 000.- 000. 000.000 000.0- 00.. 000000: 000.00 000.. 0000000 000.0- 000. «000000 000.00 000.- 0000000 000.0- 00.. 0000000 000.00 000.- 0000000 mac. 000000 naun0uuu «ndaud 00000 000.0- 000.». 00.000 0.0... 000.. 0000 000.00- 000.0. 00000000 000.0- 000.0- 00000 000. 000. «000000 000.. 000.- 0000000 000.. 000.- 0000000 000.. 000.- 000000 000.. 000.- 00000000 000.. 000.- 0000000 000.. 000.. 00000 000. 000. 0000000 000. 000. «00000 000.0. 000.- 00000 «00.. 000.- 00000 000.. 000.- 0000000 000.. 000.- 000000 000.- «00.. 00000000 000.- 000.- 000000 000. 000. 0000000 000.. 000.- 0000 000.. 000.- 0000000 «00.- 000.- 0000000 000.0 000.0 0000000 000.- 000.- 0000000 000.0 000.. 0000000 000.. 000.- 0000000 .00. «00. 0000000 000.. 000.- 000000 000. 000. 0000000 000. 000. 00000 000. 000. 00000000 000.- 000.. 0000000 000.. 000.. 0000000 000. 000. 00000000 000.. 000.- 00000000 000.- 000.- 000.00 000.0 000. 0000000 000. 00000000 IUHdauuu mundauau 00000 ouhu 0m4<:a~mu¢ 0¢to~m 0Zutaoau>uu oa~zu zo uu< 402x00¢t 00 buu 24 292 quQkuu 0««0« 00«0 0000« 0«00 00000 0000 0000 0000 0000 «0000 00000 0««00 00«0«0 00«00 0000 ««000 00000 00000 «0000 00000 00000 ««00 0000 00«00 0«000 00«00 0««00 07000 00«00 unluaHHa «000n0« Iluauudu 00«0«« 0000« 00000« 00000 00«00 0000 h000« 00«0 0000« «««0 0««0.» 0000« 0000 ««000 00«0 ««000 0000 0««0 0«00 00«0 0«00 «0«0 00000 0«000 0000 0000 0000 «cm. nhmo 0000 nuNU4uuu «0000« ««000 00000000 0000«0 00000 ««0000 00000 0000 00000 0«000 00000 «0«00 0«000 «0000 0000 0«00 «««0 00«0« 0000 00000 ««00 0000 0000 0000 000. 0000 0«000 00000 «0«0 0«00 00«0 00«0 000. 00«0 uuumuHaq 000000« 000000 0000000 unuuauau 00000 00«00 0000«0 «0000 0«00 0«000 00«00 0«000 00«00 0«000 00«00 ««00 0000 IIwH4aaq 00000 «000« «««00« 0000 0«00 0«000 00«00 ««000 «0000 00000 00«0 0«000 0«000 0«00 0000 «««0 00«0 «N00 00«0 tanwuuu 0000«n 00000 000000 00000« allqnuufl 0000«« «0«00 000000 00000 00000 0«00 00000 0«00 «0«00 0000 00000 ««000 ««000 00000« 0000NHqu 0h«0««0 0«0000 «0«0000 0000«0 00000 0000 0500 0000 0000 0000 0««0«0 00«0 0000 0«00«0 0«00«0 0000«0 00«0«0 «0000 00000 anwuaqd 0hn0«0 0«000 0««000 0«000 0«00« nuuduauu 0««00 0000 00000 0000 00000 0000 0000 «000 0«00 0000 0«00 00000 00000 ««00 00000 InIHuqad «0000 0000« 00000« 0000 0«00 0«000 00000 ««000 00000 00000 00«0 0«000 0«000 0«00 «000 0««0 «0«0 «~«0 «0«0 uuuuuHan 0000« 0000 0«000 0000 «0000 0000 nuunnaqa ««00« 0000 50000 ««00 0«000 0000 «0«0 0«00 0««0 ««00 «0«0 00000 0«000 0000 0000« 0000« uauumuuq 000.0«0 0n0000 «00.000 000.00 00000 000. 000. one. 000. 000. «00.«0 0n0. 0000 n00.«0 000.00 000.0 000.00 000.0 «00.0 uquuuaa «0«0n« 0««00 00000« 00000 00000 «000 00000 unduaaau «000« «000 «0000 0000 «0000 0000 00«0 0000 00«0 0«00 0000 «0000 ««000 0000 00«0« 0000 0000« InuuHHnn 0«000 0«00« 0«000 0000 0«00 0«000 00000 ««000 00000 00000 0««0 0«000 0«000 0«00 «000 0««0 00«0 0««0 00«0 000.- nunqauuu 00000 - 0 0000« 0000 00000 «000 «0000 0000 0000 0000« 00000000 «000 «««0 00«0« 00«0 0«000 ««00 00«0 0«00 0««0 ««00 00«0 0«000 0«000 00«0 «000 0«00 «000 00000 Hudduwna 000000 0000«0 «~000«0 00000 0«000 0000 00«0 ««00 00«0 0«00 00000 0000 0n00 00000 «0000 00000 0«000 0n~00 00000 00000 uh0zucm 0«00 Oth0h «0000 4&bu«:0 0«000 004000 «000 0«000 0000 00000 «000 2900000 0000 82h>30 «0«0 000000«0 «0000 Zauczh InnuuunH 0:0«0 0000 uz«t0— 00000 4mhu«:0 00000 004300 0««0 0«000 ««00 00004 ««00 2000000 0«00 trb>xx 0000 butcbm—o ««00 xmucxh 293 000. 009. 0«0. «00. 990.00 990.0 000.0 0«0. 00«. 0«0.0 0«0. 10000000 000.0 000.0 0«0.0 000.0 900. 000. 090.0 000. «00. 00«. 000.0 «00. «0«. 000.0 9«0.0 000.0 900.0 909. 000.0 000.0 000.0 0«0. 009.0 00«o0 000.0 000.0 000.0 000.0 00«. «00.0 000. «00.0 000. «00. 009. 000. 000. 0«0. 090.0 0«0. 000. 0«0. nulldqan 000.«0 000.0 «00.0« «00.0 00«.0 0«0. 000.0 00«.0 000.« 990.0 «00.« 000.0 000. 00«.0 «00.0 00«.0 000.0 009.« 009.0 000.0 090. 009. 0«0.00 0«0.« 000.«0 000.0 «90.0 000.« 00«. 000.0 000. 0«0.0 000.« 000. 00«. 0«0.0 0««. 900. 000. «00. 000.0 0«0. 000.0 0«0. 900.0 900. 000. «00. 000. «00.0 000. 000.0 «90.0 00«.0 000. 000.0 «00. «00.0 000.0 090. 09«.0 0««.0 000.0 «00. 00«.0 000. 000. 000.00 000. 900.00 000. 000.0 00«.0 «0«. 000.0 000.0 000. 00«.0 000.0 00«.0 000.0 000.0 00«.0 090. 0«0. 900.0 000. 000.00 0«0.0 000.0 009.0 0«9. 090.« 009. 00«.0 000. 000. 900. «00. 000.0 000. 000. 900. 000.0 00«.0 090. «00. 900. 900.0 900.0 «00. «00. ««0. 00«. 9«0.0 000.0 000.0 900. 0«0. 090. 000. 000. 090.0 000.0 000.0 9«0. «00. 990.00 900. 9««. 000.0 009.00 000.00 «00. 000.0 000.0 000.0 000.«00 000.0 000. «00.« 000.0 900. 900. 00«.0 000. 09«. 000.0 090.00 «00.0 000.0 000. 00«. 09«. 090. 000. 00«. 000.0 0«0. 990.0 000.0 000.0 000.0 000. 090.0 0««. 00«. 900.0 0«0.0 900.0 00«.0 900. 090.00 000.« 9«0.0 0«0.0 0«0. 900.0 «00.0 099. 090.0 000. 000.0 000.00 009. 000.0 0«0.« 000.0 090.0 000. 000.« 000. 00«. 000.0 090. 090.0 99«. 000.0 000. «00.0 000.0 «00. 000.0 000. 000.0 0««. 00«.0 00«.0 090.0 000. «90.0 «00.0 000.0 900. 000.0 900.0 000.0 090. «00. 000.0 0««. 000. 000.0 000.0 990.«0 000. 000.0 099.00 000.0 000.0 000.00 000.00 000.«0 000.0 «90. 000.0 0««.« 0«0.90 00«.0 «00.00 10000000 909. 000. 000.0 090. 000.0 ««0. 000. 000. 000. «00.0 0««. 000.0 000. 900. «00.0 000. «00.0 000. 900. 000. 000.0 000. «0«. 000.0 900. 000. «00.0 0««. «00. 000. 000. 000. 9«0.0 900. 0«0. 000.0 000.0 000.0 000.0 0«0.0 9«0.00 000.0 000.0 000. 000.0 000.0 900.0 «90.0 000. 000.0 «90.0 «00.0 000.0 000.0 00000000 900.0 000. 00«.00 000. 000. 000.0 000. 000.0 0«0.0 000.0 900.0 000.0 00«. 000.0 000.0 9«0.0 000. 000.0 000. 000.0 000.0 000.0 009.0 00«.0 009.0 000. 00«.00 000. 000.0 0«0. 090. 000. «00. 000. 000.0 00«. 000.0 000. 000.« 000.0 090.0 009.00 000.0 00«.0 000. 000.00 090. 00«. 000.0 009.0 «00.0 «09.00 000.« 000.0 «0«.0 909.0 00004000 090. 00«. «00.0 00«. 000.0 000. 009. 000. 000. 900.0 000.. 000. 900.0 09«.0 000.0 0««.0 000.0 900.0 900.0 «00. 900.0 00«.0 090. 000. 000. 0«0.0 «00. 000. 900. «00.0 09«.0 000.. 900.. 000. 000.. 000. 000.0 0««. 0«0.00 000. 000.0 «00.0 0«0. 0«0. 000. 000. 0«0. 900. «0«. 000. «00. 000. 000.0 090. 090.0 000. 000.0 nunnduuc n 3333.. 009.0 090.0 000. 000.0 000.0 000.0 000. 090.0 090. 90«. 000.0 «00.0 0«0.0 000.0 00«. 0«0. 000.0 090.0 000. 000. 900. 090.0 000.0 000.0 090.0 000. 09«.0 000. 000.0 00«.0 000.0 000.0 000.0 000.0 000. 000.0 000. «00040: 00909 4090000 094000 00900 90000 3000000 829910 90009000 300029 0080910 0000 9900900 0990008 0990000 «990008 «990000 0990008 0990000 «00000 080000 0000 03000040 00000 «000000 «000004 9000000 000008 «0000409 «08040: 00909 4090000 094000 00900 90000 8000000 089920 90009000 800029 0080980 000! 9900900 099000: 0990000 «990008 «990000 0990002 0990000 «00000 0009000 02093 00004000 0000900 9090000 90900200 0000090: 920000 294 non.NI nho.l ”ON.N0 «00.0 ann.u 0N“. nno. Nan.l n50. so..- luundaua o¢n.n| Nan.l nhc.no 0NN.—I 959.0 one. uu~.| nae. on". haN. nun.u Ono.u hoc.unl «00.x duo.l «on. moo. can." noo.a hen.n| 0.0.0 can.N .sn.l oNc.cI has. "no.0 noN. ono.l Ono. cON. ooo.«u one. hon. ono.l van. on..0 cue.0 no... and. neo. man. 099. 09«.0 .99.. God.» nON. man.h ban. «Ne. hNo. «co. OpN.o “kn. nno. NNn.~n oom.o ono.nl nauuua¢u not. nuu.u onN.ul NuN. ouN.l can. .0N. how. can. 000.0 can." 0N.. noo.l IluHaaqo 0Nu.nu aen.n hh0.N uNu. non.0 con. ch~.0 coe.l enu.~o pan. ”NO.Q 00..». can.n h—n.a .9... Innuendo Dn¢.o nun. tho.nl hph.| fine.“ noo.nl nNo.no cuu.l n°~.NI nN¢.I 000.00 ”N... an.l anN.a Illuanu hhn.c boo.o onn.cN hon.u 000.“: oNu.l noo.N OhN. haN.N wee. Nan. Nae. aaidHu u00.« 00°. 00¢.“ hue.“ Nn°.I Imuduauu N00. 9.5.- uNu.0 090.0 hoNol NON. Non.l 000.! Nhn.l ON».0 09«.0 cos. nON. a... ecu. IIIHMQQd New.“ cNo. onc.c nuo.u OON.I «we. ono.n oNu. onc.l hue. enu.l onn.l Noo.o EON. N0N.I coo.nl o—n.o sec. «00.0 lunduuaa «on.0 .N.. o¢p.u .th. hue. NON. NNe.I ane.| coo.l one. one.0 .N..I ON“. 006. nNn. Dee. Iauqduud «an.» Nno.n Ono.c« onu.ul ohh.nl o~n.u coo.u n00. uno.u 03¢.I ouu.l noa.l «NN. .N.. n||ou cahctuhmu no..— sON.I non.nu an... h0N.I Nu..- Nnn.u oNo.o unuuauuu ...n. u a ruNouI nno. «00.0 kna.l DON. ene.l 005.0 can. on..- ”DNol 900.— No... No..- onN.o coo. nauon ran. hNo. Hudunaqq .so.» c a non.« ono.n no¢.NN has." nae.nl can. too." toN.I nap.u ...... ...... ....._- «up... n... on..- —.~.~. .... ~—~.~. unnua4¢u u aa<30nuuc OwN. nmu.u «an. .hN.u 00.. N... ..N. «no. N.N.! 0NN. can. no... th.o an.N O-.I OoN.I u¢h. OGN. ONA. Inauu-nH u nadaanauc Ont. and. nNo.¢ nun. one.l 0’0. epn.l sue. one. ut~2¢u Otou tauu°xoz NCInxJaz Nutnxadb (TA ETA ETA ETA ETA ETA ETA “IFIIJDUIDH PIEIATHO PIEIATF ORIENT IOTOIPEI IA~$STAT REGSTAT AUTOIEE REFLEXES UTINF BESTASC APSAI? POSATTI IEBATTI POSATT2 NEGATT2 POSATTS NEGATTS ACTlVTT HOOD INTEISE THIESH OISTIACT RHYTHM APPIOCN ADAPT RATIO ADULTS QUIETPL TOVSI MELDllz TALKSIIZ NOVOCR FUSSRAT LOOKQI2 FEDCRIZ PIENATIO PREMATF ORIENT HOTOIPEI IANGSTAT RCGSTAT AUTOIEG REFLEXES UTINF 0£STAOE 551..1 “1.196 .025 .170 “.700 .096 9.750 .515 V “ [TA £11..1 .132 .202 “.003 “.025 .037 “.110 “.071 .002 .025 .033 “.027 .052 “.223 1.037 “.229 1.090 “.250 “.023 “.065 .020 “.070 “.002 “.091 “.076 “.139 “.910 .053 .062 “.507 “.007 “.501 “.061 “.509 “.005 .057 T - K81 851.-1 “1.196 “1.020 .025 .215 “.311 .931 .606 “.693 .170 .220 051..2 “6.906 .173 “.007 “19.071 1.160 09.956 2.556 “.003 “.005 .000 .001 “.002 .005 .003 “.000 .000 .000 “.000 “.016 .000 “.036 .000 “.030 .000 .001 .002 “.001 .002 .001 .003 .002 .000 .023 “.001 “.002 .013 “.000 “.002 “.000 “.002 “.000 .000 051..Z “6.906 “10.677 .173 1.007 “2.112 6.520 0.115 “0.700 “.007 “1.130 295 u .025 .030 .000 .003 “.000 .012 .000 “.009 .020 .027 “.022 .221 “.109 .506 “.112 .536 “.112 “.003 “.009 .003 “.010 “.006 “.015 “.011 “.020 “.110 .000 .000 “.061 “.000 “.000 “.010 “.092 “.010 .010 011-.1 .052 .691 “.016 “.132 .192 “.576 “.375 .029 .221 .296 “.235 6.062 “2.990 15.902 “5.060 10.731 “3.000 “.130 “.572 .135 “.390 “.239 “.519 “.035 “.792 “.050 .030 .057 “.063 .017 .170 .022 .195 .030 “.020 £11..§ “.023 “.035 .001 .005 “.000 .023 .015 “.017 “.003 “.000 .000 “.150 .060 “.299 .066 “.517 .066 .052 .109 “.050 .159 .096 .207 .170 .317 .152 “.005 “.010 .005 .006 .069 .000 .075 .012 “.000 £11..6 “.910 “1.390 .023 .100 “.275 .023 .536 “.613 “.110 “.100 .110 “.030 .010 “1.903 .020 “2.016 .025 .152 .035 “.157 .062 .279 .603 .506 .921 0.923 “.322 “.612 0.975 .397 0.251 .510 0.650 .717 “.003 £11..1 “.007 “.072 “.000 “.002 .003 “.000 “.005 .006 “.000 “.011 .000 .017 “.000 .030 “.000 .000 “.000 .006 .010 “.007 .020 .012 .026 .021 .039 .597 “.010 “.027 .221 .033 .350 .005 .303 .059 “.000 APGARZ PDSATTI NEGATT1 POSATT2 NEOATT2 POSATTS NEOATTS ACTIVTY IOOD INTEISE THRESN DISTRACT RHYTHM APPROCH ADAPT RATIO ADULTS OUIETPL TOYSI HELDHRZ TALKBRR2 NDVOCR FUSSRAT LOOKARZ FEDCRRZ AGEHO FAIOCCUP EDIO FAIINC AOEIO FAIOCCUP EDIO FAHIIC ESTIMATED REGRESSIORS ETA ETA ETA ETA ETA ETA ETA QIAUIOCAADH ETA ETA ”H “.102 “.700 .505 “1.700 .595 “1.095 .597 .096 .275 “.100 .290 .177 .505 .522 .505 9.750 “.552 “.660 5.056 .515 5.500 .671 6.056 .950 “.626 I O ETA KIA-.1 “1.196 “6.906 “.726 “5.070 Bil-.1 20.500 70.075 7.700 52.906 ETA OR RSI (3 551..1 .029 “.001 .059 .501 “.025 0.000 .005 .900 “19.871 9.016 “05.572 10.057 “00.291 10.122 1.160 5.522 “1.205 5.551 2.159 0.652 5.007 7.070 09.956 “1.005 “5.022 27.000 2.556 27.509 5.550 29.960 0.615 “5.110 KIA..Z .025 .175 .010 .076 651..Z 70.075 505.067 59.060 160.950 031..3 “.020 .001 “.010 “.150 .000 .150 .006 296 .170 “.007 “.000 “.575 RESRESSIOA RESIDUAL EOVARIANCES £IA..1 -.ooo .ooo ‘1‘..2 .000 ..5 “.700 “19.071 “2.066 “0.750 ‘11..) ElA..3 ..2 .096 1.160 .121 .515 E1A..5 9.750 09.956 5.191 22.000 EIA_.£ .515 2.556 .266 1.126 297 999. 999.0 N99. 990.~ Jmoau N99. 999.0 «99. «99.0 wonoun J3: «moo- ~09. nno. 999.0 nn9.l 9nd." 9N9. 909.0 ~n9.| 999. 0N9. 9N9.I Ono.— gag: 099. «99. 999.0 999.. 999. 9.9. «99.0 9n9. 999.9 39999 909. 9N9.N 999.I 999. 999. 999.. 999.I .99. N99.- 099.0 «0.. gang: «99.0 999.0 099. 099. 999.- 999.0 ”—9. 999. 999.0 999. 999. 999.0 «09. 9999939 999. 909.9 999. nmn.n «99. «99.- «99.I N99. 999. «99.0 «99. N99. «99.0 999. 999.9 Janna n99.l 999. 999. n99.l «99. «99. 999999: «99.0 «99.I 9999999 999. 999. ~99999 «99.0 999.! 9099990 999.0 999.0 98998 099.0 999.- «unadmuu 999. 099. 9990939 «N9. 9—9. 999muu9 999.0 699.0 99990299 699. ”99. 9Ut¢099t «99. 999. 919999 999.9 999.- 9998u99 awn. 98998u99 7:990 99.999999 9&919999999 49=o~mu¢ lo—uau99uc 999.0 ”99.0 ~¢9uoun «99. 999. «992994 «99. «99. 99999:; 999. 099. 900998 909. two. ~98n8499 099. n99. «Ctaau: 999. 999.9 99999 099.0 999.9 4&9u~:o 999.. 999.. 994399 9nd. 999.9 69999 999. unu.l 99999 0N9. 999.0 8099999 999. 999.0 829919 999. No9.I 99999999 ”N9. 999.0 tnu¢29 999.0 999. unZU9au «N9. 009.0 969: 999. ”No.0 999-909 999. 09«.I n9999ut non.o Can.“ 9999999 999. 99N.I ~999ou8 «in.c N.N.u «999999 999. 999.0 n999ouz 999.0 «on. «999999 999. no..- «99099 999.: 999. H.99muo 999.0 099. 58993 9N9.o 999. «unaduuc 999. nn9.o 9990939 9&9. nn9.l 999wouz 099.. 999. 999M9299 099. «~9.I 9u9999oz «99. 999.! 989999 9n9ol 9.9. b992u99 o~9.o ow9. 0:998u99 M «an «II—d9 um: 89 9 «99.0 «99.0 9 99m 999.: 999.9 9 999 999.0 N99. n 999 «99.. «99.0 c 999 ~99. 999.! n 999 ~99.I 298 umnnqAMn no... No.- Nu.- a... «no. a... 0... ...-o ....I .... ..Nol ban-I ..no. no... ten-o ...-u nuuauuau n... 0.9.0 on..- 0.... mean! 0.9.0 ....o .... .... ....I ha.- «no. 9... hue. ow.- ouool .N.. a... n... ego-I a... nan-I «quH n... no..- nnoou ...-I once! nee-I ...-I .... .... 0...! ..ho one. Guns ..N. hon. onnol Gwnou .Iluaaun .... 0...! fine-0 0...! Oneal no.0! ...ol .... .... ...ol «0.. mac. .N.- nae. «N.- ha... .N.. h... n... Nu... no.- «noon IJuHuna. a... N...0 Nae-l u..ol «no-I «.900 ...-o .0.. .... .9900 hp". «an. «can .... Ona- he..- and. cuoofl IHHu—Huc ...-I a... .N.. .0.. con. .... .... ....I ....I .... «ac. h... .... o... a... N...0 o... N... «wool unu- «ma-I non. Ileaaad h... .N—ou ...-I at..- carol up..- 00..! a... N... ...-I ...-I «...- N...0 no..- no.0! .... «..0. ....I .0509 IuHHquua a... 0...! chm-I No.00 «one- «n.00 ...-I .... .... ....I ...-O 0...- 9...! no..- 00.00 N... no..- N...0 .N.. «nu-o .N.. nwuoo .0..0— IIIq~Hzc bu.¢bm~a :mUCZh «ccuouu «.0309; b<¢mm=u 80...: chnxJCh «xxoau: ~m>oh Jahunzo ”hang. Gupta ht... tuccma. tzh>3¢ pub>~bu< nhhcuuz nhhcmon thdouz Nhhdwoc ubhx¢ hu<¢bmue annex» Hutu»!— one: >h>~huc nphdouz nbp9u 49:99muc zoummuauua 999. 999. .99. 999.- 999.- 999.- 999.- 9~9.- 999.- 999.- ~99.- .99.- ~¢9.- 999.- ~99.- 999.- ~99.- 999.- 999.- 999.. 999.- 999. 999. 999. 999. 999. 999. Nxxuowu Nxcxccq h<¢mmzu ¢uo>oz Nztnxqoz NztnxaOh Alba—Dc mFJDO< 300 Ad EVALUAYION OF THE IMPACT OF NATERN‘L AGE Oh CHILD CCVELOPHENT SYANDARDIZED SOLUYION LIHBDI Y m”; sn--z £15-; an": an“: £15-4- :1...1A PRENAYHO 036‘ 00000 00000 00000 00000 00000 00000 PRCNAYF 0556 00000 00000 00000 00000 00000 00000 ORIENT 00000 0012 00000 00000 00000 00000 00000 NOTORPER 00000 0096 00000 00000 00000 00000 00000 IANGSYAT 00000 .0191 00000 00000 00000 00000 00000 “£037.? 00000 0021 00000 00000 00000 00000 00000 ‘UYOIEG 00000 0270 00000 00000 00000 00000 00000 REFLEXES 00000 ‘0319 00000 00000 00000 00000 00000 “TIN? 00000 00000 0102 00000 00000 00000 00000 0:37.00 00000 00000 0190 00000 00000 00000 00000 IP6002 00000 00000 -0152 00000 00000 00000 00000 P050771 00000 00000 00000 20062 00000 00000 00000 “£01771 00000 00000 00000 -10216 00000 00000 00000 P050772 00000 00000 00000 00606 00000 00000 00000 “£00772 00000 00000 00000 .00206 00000 00000 00000 P030773 00000 00000 00000 50903 00000 00000 00000 ”£00773 00000 00000 00000 '10200 00000 00000 00000 ICT!VTY 00000 00000 00000 00000 0220 00000 00000 HOOD 00000 00000 00000 00000 0652 00000 00000 INVE~SC 00000 00000 00000 00000 -0236 00000 00000 VHRESH 00000 00000 00000 00000 0607 00000 00000 DISTIAC' 00000 00000 00000 00000 0020 00000 00000 IHY'HH 00000 00000 00000 00000 0900 00000 00000 APPROCH 00000 00000 00000 00000 0762 00000 00000 ‘0‘?! 00000 00000 00000 00000 00307 00000 00000 I‘TIO 00000 00000 00000 00000 00000 20907 00000 IOUL'S 00000 00000 00000 00000 00000 .0000 00000 0UXET'L 00000 00000 00000 00000 00000 “0205 00000 TOYS! 00000 00000 00000 00000 00000 10665 00000 H£L0002 00000 00000 00000 00000 00000 00000 0101 YILK3II2 00000 00000 00000 00000 00000 00000 10930 IOVOCR 00000 00000 00000 00000 00000 00000 0236 FUSSI‘T 00000 00000 00000 00000 00000 00000 20119 LO0KQI2 00000 00000 00000 00000 00000 00000 0326 FEDCII2 00000 00000 00000 00000 00000 00000 .0220 L‘IUDI X “L4 Bil—2 00:00 00510 00000 FIHOCCUP 00000 190502 [000 00000 20036 'IHIUC 00000 00627 00" “-1 0n--2 ‘10—} {15-4 011...: KIA-.§ HLJ £00 I 30000 00000 00000 00000 00000 00000 00000 [00 2 -6270005 10000 620773 00000 00000 00000 00000 [00 3 '130591 00000 10000 00000 00000 00000 00000 [.00 0 -1600079 0123 300220 10000 0271 00000 .2000. C00 5 00000 '0277 0150 00000 {0000 00000 00000 [00 6 00000 00000 00000 00000 00000 10000 00000 [0. E00 E00 ‘00 £00 £00 E00 E00 K51 K81 ETA ETA ETA ETA ETA ETA ETA ETA ETA ETA ETA ETA ETA ETA ~lOclb(dfilw N00 ~00¢la>unnu filOCIJHHIDfl 00000 SAHHA £51.-1 .365 ’1120057 '30093 ‘230525 00000 00000 .70000 PHI 531.-1 10000 0040 PS1 £00.-1 '0000 00000 851..2 -10290 6920673 150655 1730004 00000 0050 160361 551..2 1.000 ‘01..2 '120175 C (STANDARDIZED) III--1 10000 .0717 0006 0505 .0275 .0030 -0713 E11..3 10000 0205 .0555 0260 0655 .0101 D C8TANDAIOXIED) 001-.1 0365 -0560 10062 0992 -0052 00000 0069 THE PIOBLEI COULD HAVE BEEN SOLVEO 01TH CI: 21:39:42 07/09/00 [31..2 -1.290 10209 ‘10003 '10253 0603 0056 0660 T090131 a..§ 0009 EIA..3 10000 0631 -0101 .0260 -0307 3(3]. -8.001 ‘03-.3 20320 10000 .0232 ‘0113 0037 1135 Ltucs PRINT. 00.2 Ens-.§ .916 KIA..1 10000 0222 0156 079000 27 PAGES PR1NT0 6.612 £11--§ 0271 £11..§ 10000 0736 COST AT 061 18 8 10000 s..1 100900 EIA..1 10000 0069