'I ||fllill|1111121|fllflmlMIMI!“Mlufllfllflll fl; University This is to certify that the thesis entitled FATHER ABSENCE AND THE SOCIAL BEHAVIOR OF PRESCHOOL CHILDREN presented by Tito Fidel Reyes has been accepted towards fulfillment of the requirements for . Ph.D. degree in Family and Child Sciences Major professor (5/ Date November 18, 1977 0-7639 FATHER ABSENCE AND THE SOCIAL BEHAVIOR OF PRESCHOOL CHILDREN By Tito Fidel Reyes A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Family and Child Sciences [i 4‘ ‘l . I l 4 1 ' .‘ d lllll‘ii 'lh'll. I'l..il1‘l“lllu1§"|l‘liu ABSTRACT FATHER ABSENCE AND THE SOCIAL BEHAVIOR OF PRESCHOOL CHILDREN BY Tito Fidel Reyes The primary purpose of this study was to investigate group differences in the affective and social involvement be- haviors of father-absent and father-present preschool children. The secondary purpose was to gain further descriptive informa- tion concerning the father-absent family. Length of absence, defined as short-term (1-2 years) and long-term (3-5 years) was examined as was the involvement of father surrogates. Utilizing a video assisted observational methodology employing small group experimentally contrived play sessions, a broad range of social behaviors were observed including aggression, cooperation, and activity level. Expressed self concept was also measured. A Parent Information sheet was used in surveying the families to secure descriptive demographic data pertaining to the child and his family. These data included the child's age, ethnicity, ordinal position, playmates, number of years of father absence, number and relationship of father surrogates, type of dwelling, number of females in the home, and day care experience. The demographic data also included the mothers Tito Fidel Reyes and/or fathers age, ethnicity, education, occupation, number of working hours, family income level, type of transportation, and type of dwelling. The sample consisted of 127 children ages 3% to 5 years and their families. They were randomly selected from 8 day care centers in southern lower Michigan. Three instru- ments were employed to collect data on the social and affective behaviors of the children - The Observation of Social Behavior, The Classroom Socio-observation, and The Brown Self-Concept IDS Referent Test. Two-way multivariate analysis of covariance was employed to test the hypotheses of interest. The study focused on three research questions concerning father-absent and father-present children. What is the effect of father absence on the social behaviors of preschool children? Does the length of father absence affect the development of social behaviors? Do older males or siblings in the family affect the development of social behaviors in children from father-absent families? The results of the study indicated no significant differences between father-absent and father-present children with regard to the variables of interest. Also, no significant differences were found with relation to the variables of long- term and short-term father-absent children or between those father absent children with a father surrogate and those without a father surrogate. Results of the demographic survey, however, indicated Tito Fidel Reyes that father absent mothers were not that different from father-present mothers when considering the hours they worked, and attained educational level. The majority of father-absent children were from families on public assistance (ADC) with a weekly income level of $125 or less, having a mother between the ages of 20 and 29 with a semi-skilled job requiring little responsibility. Slightly fewer father-absent mothers had not completed high school, attended college or held a degree, than father-present mothers. Almost all father-absent mothers received day care assistance in addition to working nearly the same number of hours (30-40) a week as father-present mothers. ACKNOWLEDGEMENTS Many people have been of great support in my graduate work during the past and many times long years. Some have contributed knowingly while others will not know their con- tributions. These acknowledgements are a small attempt to thank some of those persons. I would like to thank Dr. Paolucci for being one of the first persons at Michigan State to believe that a young man in Utah could come to MSU and earn a doctorate. Also as a faculty and committee person her advice and guidance will be remembered. To Eileen Earhart with whom I have worked over the past years, thanks for being a model of persistence when I might have not been otherwise. Her help during the final stages of this dissertation.will always be appreciated. Thanks to Don Melcer for helping me face some difficult decisions both in my graduate career and my personal life. Also thanks to Ellen Strommen as a committee member and faculty person who provided stimulating seminar environments in which to explore areas of learning. Bob Boger has not only been my major professor but also another person who was willing to trust my judgment and creativity in many research endeavors. Knowing that others believe in one's work has been a most gratifying experience. ii To the secretarial staff at the Institute for Family and Child Study who over the years encouraged me continually by asking, "Are you still here?" Nancie Wood, Coleen, Each, and ‘Mary Voth provided assistance in the typing of preliminary drafts and deserve a special thanks. I have truly appreciated the technical assistance in the data analyses and suggestion given throughout the course of this dissertation by Mary Andrews. As a fellow graduate student and friend she was always willing to lend her assistance. As was mentioned previously many people unknowingly contributed to my graduate work. To Mr. Bell at Bell's Pizza House, the folks at Fireball Pinball Palace, and countless others who provided jobs when there were none; Thanks. My mother and father are perhaps the happiest people to see this work completed. Now they can look to their new life of retirement without the fear of receiving any more of my grade reports! They were also models of persistence and were the first to recognize that I could undertake graduate work and be successful. Thanks for believing in me. Andrea, Matt, Eric, and Bobby also "helped" in this endeavor by providing different forms of input during many stages of this dissertation. Thanks for the humor and dis- tractions that kept my graduate work in proper perspective. To the very best friend I have ever had over these long years; thanks, Nancy for being understanding. The many times you listened to my outrageous ideas will always be iii remembered. I‘m sure my thanks to you is in the form of our finally reaching the end of a long road and being able to plan other adventures. iv TABLE OF CONTENTS ACKNOWLEDGEMENTS . LIST OF TABLES . CHAPTER I. II. III. INTRODUCTION . . Statement of the Prbblem Conceptual Framework Objectives . Assumptions . Definitions . REVIEW OF LITERATURE Scope of the Review . . . . The Impact of the Father on Personality Development . Father Absence During the Preschool Years . . . Length of Father Absence Surrogate Father Effects The Interface of Father Absence and Social Behaviors Summary of the Literature . METHODOLOGY . . . Research Design . . . Operationalization of Variables . Hypotheses . . . . . . . Sampling Procedures . Sample Selection . . . Description of the Sample . Instrumentation . . The Observation of Social Behavior The Classroom.Socio- Observation . . The Brown IDS Self-Concept Referent Test. Parent Information Sheet . . . . Data-Gathering Procedures . . Data Reduction and Analyses . Measures of Association . vii \lNO‘nPLOH . 10 . 10 . ll . 14 . 17 . l9 . 20 . 23 . 26 . 26 . 28 . 34 . 38 . 38 . 4O . 43 . 44 . 48 . 50 . 51 . 52 . 53 . 53 TABLE OF CONTENTS (CONT'D. . . ) CHAPTER IV. VI. RESULTS OF DATA ANALYSES . Section 1 . . . Results Section 2 . . . Demographic Data . . Summary of the Family Demographic Data . DISCUSSION . Section 1 Section 2 SUMMARY, CONCLUSIONS, AND IMPLICATIONS . Summary Conclusions . Implications . APPENDIX A Instruments . APPENDIX B Levenes' Test for Equality of Dispersions . BIBLIOGRAPHY . vi ' 100 100 107 133 141 TABLE 10. ll. 12. LIST OF TABLES Sample Distribution of Father Absent and Father Present Children by Sex . Sample Distribution of Father Absent Children by Length of Absence by Sex . Sample Distribution of Father Absent Children by Father Surrogate by Sex Sample Description by Father Status by Age by Sex . Length of Father Absence by Sex Length of Father Absence by Age Older Unlike Sex Siblings by Father Status by Sex . . . Summary of Levenes' test for Equality of Dispersion Across levels of the Independent Variable, Length of Father Absence on all dependent variables . . . . . . . . Summary of Levenes' test for Equality of Dispersion Across Levels of the Independent Variable, Presence of a Surrogate, on all dependent variables . . . . . . . Summary of Levenes' test for Equality of Dispersion Across Levels of the Independent Variable, Sex of Father absent children, on all dependent variables Summary of Levenes' test for Equality of Dispersion Across levels of the Independent Variable, Family Status, on all dependent variables . . . . . . . . . . . . Summary of Levenes' test for Equality of Dispersion Across levels of the Independent Variable, Sex of Father Absent and Father present children, on all dependent variables. vii 27 41 41 42 43 . 134 . 135 . 136 . 137 138 LIST OF TABLES (Cont'd. . .) TABLE 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Results of TwoAWay MANCOVA on affective social behaviors of father absent and father present children . Results of Two-Way MANCOVA on social involvement behaviors on father absent and father present children . Variables contributing to main effects of sex on father absent and father present children . . . . . . . . . . Adjusted cell means of variables contri- buting to father absent and father present main effects of sex . . . . . . . Results of Two-Way MANCOVA on social behaviors of long and short-term father absent children . variables contributing to main effects of sex on long and short-term.father absent children . Adjusted cell means of variables contri- buting to sex effects on long and short- term father absent children . . . . Results of Two-Way MANCOVA on Social Involvement behaviors of surrogate and non-surrogate father absent children . Variables contributing to main effects of sex on surrogate and non-surrogate father absent children . . Adjusted cell means of variables contri- buting to surrogate and non-surrogate father absent main effects of sex Results of TwoAWay MANCOVA on affeCtive social behaviors of surrogate and non- surrogate father absent children . Father Status by Ethnicity . Center by Father Status . viii . 62 . 63 64 66 67 67 69 70 70 72 74 76 LIST OF TABLES (Cont'd. . .)‘ TABLE 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Ordinal Position by Father Status . Age of Playmates at Home by Father Status . Relationship of Playmates at Home by Father Status . . . . . Age of Fathers and Mothers by Father Status . Education of Parents by Father Status . Parent's Working Hours by Father Status . Occupation of Parents by Father Status Income Distribution by Father Status-- Center Report . . . . . . . . . . Weekly Income Distribution by Father Status-- Parent Report . . . . . . . . . . . . . Family Dwellings by Father Status . Transportation to Center by Father Status . Transportation Time to Center by Father Status ix . 78 . 19 . 80 . 81 . 82 . 83 . 84 . 84 . 85 . 86 CHAPTER I INTRODUCTION Approximately 17 percent (10 million) of the children in the United States under the age of eighteen were being raised by a female head of the household, and 25 percent of these children were of preschool age (Bureau of Census, 1975). The proportion of divorces and homes broken by separation involving children has been steadily increasing since 1958. The numbers involved "serves warning that we would do well to consider care- fully what we do and do not know about the effects on children growing up in a fatherless home" (Herzog & Sudia, 1973, p. 141). In this study the differences between the social behaviors of father absent and father present children during the early period of sex-role adoption between the ages of 3 and 5 years are examined. Several theoretical approaches have been conceptualized which focus on the father as he influences the development of personality. The psychoanalytic theory of Freud (1924), the role theory of Talcott Parsons (1955), and identification theory of Sears (1965), Bandura (1963), and Lynn (1969) have cast the father as a primary determinant of positive sex-typed behaviors for girls and positive sex-typed behaviors for boys. 2 Lynn (1969) conceptualizes the father's influence dur- ing the preschool years (3-5 years) as a period of sex-role adoption, when the child begins imitating behaviors and differ- entiating masculine and feminine behaviors. Developmentally the preschooler is learning how to interact with peers and other adults (Havighurst, 1952), and is developing social behaviors ranging from solitary to cooperative play (Parten, 1932). During this period the father is "instrumental" in facilitating inter- actions between family members and society (Parsons & Bales, 1955). Parsons and Bales posed the question, "What would be the effect on personality development for boys and girls without a father?" The question has taken on added import in the 1970's with increasing numbers of preschool children growing up without a father consistently in the home. Research on the consequences of father absence in later years has focused on adolescent behaviors, delinquency, and academic performance. Carlsmith (1964) in a study of academic performance in college found reversals in usual Mathematics- Verbal patterning on the College Entrance Examination Boards. Long-term, early separated, father absent boys showed relatively higher verbal scores than mathematics scores. Hetherington (1972) found that girls showed an inability to relate appropriat- ely to men and male peers. Despite the relative importance of peer relationships through childhood and adolescence, observational data of father absent preschool peer interactions is almost non-existent. Also lacking is family demographic data of father absent children 3 which could contribute to a further understanding of the impact of variations in the father absent environment. Family characteristics such as length of father absence and availability of a father surrogate have been suggested as avenues for further research (Biller, 1967). Statement of the Problem The primary purpose of this study is to investigate group differences in the affective and social involvement be- haviors of father absent and father present preschool children. The secondary purpose is to gain further descriptive information concerning the father absent family. In this study length of absence, defined as short-term (6 months-2 years) and long-term (3-5 years), is examined as is the involvement of father surrogates. Utilizing video assisted observational methodology using small group experimentally con- trived play sessions, a broad range of social behaviors are observed including initiation, aggression, cooperation, activity level and expressed self-concept. Demographic variables described and discussed include: (a) age, ordinal position and playmates of the children, (b) education, age, occupation and number of working hours of the mother and/or father, (c) length of father absence in years, number and relationship of father surrogates, and (d) income level, ethnicity, type of dwelling, and day care experience. 4 Conceptual Framework In this study the framework for investigating father absence is based in part on Lynn's (1959) conceptualization of sex role development and a definition of father absence. To- gether these provide support for the measurement of behavioral differences at the appropriate age within a defined context of father presence and father absence. The development of sex role from birth to approximately 10 years is characterized by three stages: (1) sex-role orientation; (2) sex-role preference; and (3) sex-role adoption. Sex-role orientation relates to how the individual views himself as masculine or feminine, while sex-role preference is the individual's desire to adhere to culturally defined sex roles. Sex-role adoption or the display of sex appropriate be- haviors is the particular concept relevant to this study for several reasons. First, Biller (1969) has further defined this conceptual stage as developing between the third and fourth year when a sex role model is critical to the imitation of sex appropriate behaviors. Bandura and Walters (1963) provide evi- dence for imitative behaviors deve10ping at these ages. The prevalence of imitative behaviors makes the pre- school age group a most suitable population for investigating father absence and its affects on a wide range of social be- haviors. Biller (1967) emphasizes that the father is crucial as a model for behaviors during this period of sex-role adoption. If this is conceptually correct, the absence of a father during 5 this period should have an effect. A wide range of behaviors would then reflect differences as a result of father absence or father presence. The definition of father absence is an important part of the conceptual framework in this study. Herzog and Sudia (1973) generalize father absence as "continuing" or "temporary", given the wide range of reasons for father absence such as employment, death, separation, divorce and military. An extensive survey by Coleman (Coleman, 1966) in which children were asked, "Who is now acting as your father?" pro- vided a conceptual basis for the categories of father status. Three categories were generalized in the Coleman survey: (1) Real father or stepfather living at home, (2) foster father, grandfather, or other relative, and (3) no one acting as father. In this study similar categories were formed: (1) Real father living at home, (2) father surrogate or any male older than the child living at home, and (3) no one acting as father or any older male. A distinction lacking in most previous research has been length of absence. The categories of short-term father absence (0-2 years) and long term father absence (3-5 years) are con- ceptualized for this study. This is based upon the rationale that the first two years of father absence are disruptive for the family and cause major adjustments. The third, fourth, and fifth years are years of increasing stability as the length of separation increases and living patterns are adapted. The in- tegration of father presence, father absence, father surrogate, 6 length of absence, and sex role adoption then provides the frame- work for the investigation of father absence and its effects on preschool social behavior. Objectives The primary objectives of this study are: 1. To determine whether sex of the child, presence or absence of a father, presence or absence of a father surrogate, or length of absence, differentiates the expression of the following social behaviors: Affective Social Behavior Self Concept Initiation Autonomy Response Adult Dependency Aggression Social Behavior Involvement Activity Level Heterogeneity of Initiation Tolerance 2. To describe the following characteristics of father present, father absent, short-term, long-term, father surrogate, and non-father surrogate father absent homes: Ethnicity Education of Father Child's Day Care and Mother Experience Weekly Work Hours of Center Size Mbthers and Fathers Child's Ordinal Occupation of Fathers Position and Mothers Females in the Home Income - Center Report Playmates at Home Income - Parent Report Age of Father and Family Dwelling Mother Transportation to Center Assumptions 1. It is assumed that other pertinent factors such as hereditary influences are randomly distributed across the sample of interest and that differences in children's verbal and non— verbal behaviors can be considered random except as related to the dimensions of interest. 2. Family dynamics including father absence and father presence impact upon the observed social behaviors of preschool children. Definitions Father absent family: A family in which the father was absent, as reported by the mother. Short-term father absence: A family in which the father was absent from 3 months to 2 years. Long-term father absence: A family in which the father was absent from.2 years to five years. Father surrogate: A family in which an older male rela- tive or older non-family male was living in the home of the child. Father present family: A family in which the father of the child had been consistently present as reported by the mother or father. Family: A corporate unit of interacting and interdepend- ent personalities who have a common theme and goals, have a com- mitment over time, and share resources and living space (Hook & Paolucci, 1970). 8 Self-concept: An individual's impression formed ". . . of himself, of his characteristics, and of his capabil- ities from.information which he receives from.referents (mother, teacher, peers, and self) about the ways in which they see him." (Brown, 1952, p. 36). Autonomy: A behavior expressed in degrees of explora- tion, choice and interest (Banta, 1969), as well as independence, persistence and tolerance. Adult dependency: The frequency of self initiated interactions with an adult in a small group play situation. Initiation: The introduction of self or change in an activity prompted by the self. Response: "Acceptance" was the covert or overt awareness and acceptance of another's initiation and "rejection" was the covert or overt awareness and rejection of another's initiations. Aggression: "Verbal" was the negative affect conveyed by voice tone. "Physical" was the behavior which was not socially accepted or was negative in connotation. An example ‘was pushing or hitting. Heterogeneity of Initiation: The proportion of initi- ations directed to the unlike-sex peers versus like—sex peers. Activity level: A range of passive to active initiations or responses. Social Involvement: Unoccupied, onlooker, solitary, parallels, associative, or cooperative social behavior (Parten and Newell, 1943) were the behaviors included. 9 Associative play: The child who plays with other children borrowing or trading materials with no common goal expressed or connotated between the children (Parten & Newell, 1933). Cooperative play: The child playing with other child- ren using similar materials with an expressed or implied goal (Parten & Newell, 1943). CHAPTER II REVIEW OF LITERATURE Scope of the Review A comprehensive discussion of all father absent literature from.birth through adolescence is not applicable to this review due to the limited sc0pe of this study. A review which focuses on the preschool literature is therefore presented. A particular problem.in establishing an integrated direction from the literature relates to the number of independ- ent variables involved in most studies and the methodological approaches utilized. In this review of literature the author will attempt to set the parameters of the review by first presenting a theoretical base for understanding the effects of the father on the personality development of the preschooler. Secondly, the literature on father absent preschool children is introduced, followed by the literature on length of absence and availability of father surrogates. A point is made to review those studies which have gained prominence although methodological procedures used raise serious questions regarding the reliability of the instruments and the validity of the findings. These studies are consequently examined in detail to alert the reader to the current quagmire 10 ll of the father absent literature. The Impact of the Father on Personality Development In theory, the father as a significant influence on the development of the child, is especially present in the psycho- analytic writings of Freud (1924). Social learning theory departs from the issue of the resolution of the Oedipal Complex by explaining identification in terms of the withholding of love and dependency (Sears, Rau and Alpert, 1965), fear of punishment and substitute behavior (Mower, 1959), or through modeling and imitation (Bandura & Walters, 1963; Lynn, 1959). One of the initial theoretical constructs regarding the effects of father absence lies in the work of David Lynn (1969) on sex-role development. Sex-role development seems to be an interaction of cultural reinforcements for traditional masculine and feminine role expectations, biological determinants which predispose males and females towards certain behaviors, and parent-child relationships which predispose each sex toward certain roles (Lynn, 1974). Three categories of sex—role are conceptualized: (l) sex-role orientation, (2) sex-role prefer- ence, and (3) sex-role adoption. Sex-role orientation is the manner in which the indivi- dual begins to view himself. The development of sex-role preference (the desire to adhere to culturally defined masculine roles and social symbols) is probably influenced by discriminat- ing tasks involved in sex-role orientation. Sex-role adoption is the display of masculine and/or feminine behavior (Lynn, 1969). 12 Using Lynn's framework, Biller (1967) has suggested a ‘multi-aspect conception of masculine development in boys, focusing on two key variables: (1) the degree to which the father is available, masculine, and the setter of limits, and (2) the degree to which the mother encourages masculine behavior. Biller also postulates the beginning of sex-role orientation is the second year with the most critical period of awareness between the first and third years. The time between the third and fifth years is when imitation of a masculine model is the major process in sex-role adoption. The availability of the father, or another significant older male, is crucial. A nurturant father, one who sets limits, makes decisions and generally is a model of masculine behaviors provides interaction for the young child to imitate (Biller, 1967). Studies of intact families have shown that boys reared by masculine, dominant, and warm fathers were themselves more masculine in their stated preferences for various sex- typed activities and in their overt behavior (Biller & Borstelmann, 1967; Hetherington, 1967). Research prior to the 1970's has shown that both child- ren (Hartley, 1964) and adults (Jenkins & Vroegh, 1969) maintain traditional sex typed conceptions of femininity and masculinity. Congruent with this is Parsons and Bales (1955) classification of the male role as basically instrumental and the female role as basically expressive. In our society females are viewed as submissive, nurturant, and sensitive in social situations. In 13 contrast, males are expected to be independent, dominant, assertive and competent in dealing with problems of the environ- ment. Instrumental behaviors, such as working with different segments of society and interacting with many other individuals and groups, place the father in a unique role. He not only brings society to the family by his interactions, he also brings the family into society. He is expected to supply neutral, objective and sound judgment as well as authority and discipline, within the family (Lynn, 1974). Lynn further suggests that the father's instrumental role and his concern for his family's entrance into society en- courages his willingness to participate in his child's sex-role development. Interaction with society would require the development of appropriate sex-role behaviors and the father would thus be the means by which this would be accomplished. The concern for children developing sex appropriate be- haviors would seem to be an equally shared parental responsibil- ity, but there is some reason to believe otherwise. If one considers parental interaction based on Parson's expressive- instrumental model, the father would be the one most concerned with the child's behaviors outside the home, while the mother's concern would be towards emotional support and maintaining family group member functioning. Differences between mothers and fathers in their rein- forcement of sex appropriate behaviors are evidenced in several 14 studies. Sears, Maccoby, & Levin (1957) and Goodenough (1957) used interviews in.which mothers emerged as generally in- different to distinct sex roles in early childhood for boys and girls while fathers were more concerned. Fathers some- times treated their sons and daughters differently. Fathers interviewed in the Goodenough study of two to four year olds were opposed to opposite-sex behavior in little boys but were not Opposed to masculine behavior in their girls. Further evidence suggests that the sex of the child effects parental attitudes and behaviors. In a study by Lansky (1967) parents of preschool and kindergarten children were asked to tell what their reaction would be if a child expressed a preference for a certain masculine or feminine activity, such as doll play for boys and rough play for girls. Parents of boys expressed more negative reactions when boys made feminine choices. In later childhood Bronfenbrenner (1961) found that childrearing practices of the father reflected differential treatment by sex. Affection, attention, and praise was received more by girls than boys, while boys were subjected to greater discipline and pressure. Father Absence During the Preschool Years Father absent literature concerning the preschool child during the last two decades has been almost non-existent. Dur- ing the 1940's the father absent child became important as a result of investigators developing and refining projective doll play techniques. 15 George Bach (1946) was one of the first to publish findings on father absent school age children (6-10 years). At the time, he was using the projective doll play technique of Sears, Piltner, & Sears (1946), who also were beginning to publish their own findings on father absent children. The Sears' study is of importance as it has since been cited and used extensively by other investigators. In the Sears study 126 preschool children from day care centers who had fathers in the military (length of absence not reported), played in two twenty-minute doll play sessions. Based on these observations, boys from.father absent homes por- trayed much less fantasy doll play aggression than boys from father present homes. Sears (1951) in a more detailed analysis of the same data later reported little difference in the frequency of fan- tasy aggression between three year old father absent boys and girls. Interestingly, young,father absent,boys (three years old) emphasized the maleness of the father and boy dolls less than father present boys. Two other studies using doll play as a dependent vari- able have influenced the father absent literature. Tiller (1958) and Lynn & Sawrey (1959) studied Norwegian school age children eight to nine years old whose fathers were sailors and away from home for long periods of time. Socio-economic status ranged from officers to businessmen and white collar workers. These father absent children showed more compensatory masculinity 16 (at times behaving in an exaggerated masculine manner, at other times behaving in a highly feminine manner) based on projective tests and maternal interviews. The maternal interview consisted of 60 questions such as, "Does he (child) often get into fights? Has he talked to you about wishing to be like his father?" The doll play test was a modification of a projective test developed by Lynn (1959). For example, in the Crib or Bed Choice Situation, the child is presented with a doll-house crib, bed, and child doll. The child then chooses a bed for the doll to sleep in. If the child chooses the crib he is characterized as immature. Lynn and Sawrey then conclude that after a "significant" number of children chose the crib that "the following hypothesis was made and generally supported by the findings: “More father absent boys would show immaturity" (Lynn & Sawrey, 1959, p. 260). They also concluded that father absent boys demonstrated poorer peer adjustment than father present boys, and than father absent girls.“ Father absent boys were also found to react to their in- secure masculine identification with compensatory masculinity. Regarding the maternal interview methodology, Eron Banta, Walden and Laulicht (1961) raise the question of reliability of observational data collected from only one parent. They found that there was only occasional agreement between mother and father on their ratings of their child's behavior or interac- tions with their children. In fact the father was found to be more accurate in describing the behaviors! 17 In one of the few studies of Black children, Santrock (1970) used a methodological approach similar to previous investigations. A maternal interview schedule originally con- structed by Sears, Mbecoby and Levin (1957) was revised and administered. Santrock's methodology also included a modified doll play interview for children which was a combination of previous methods used by Emmerich (1959). In a description of the potpourri of instruments Santrock writes that he "assumed that a combination of three tested methods might provide a more comprehensive and accurate assessment of sex typing" (p. 265). Although no behaviors were directly observed, he con- cluded that there were no significant differences between Black father absent and father present girls on dependency, masculinity- femininity, nor aggression. Black boys who were father absent were more feminine, dependent and less aggressive than father present boys. If a father substitute was present boys were less dependent and if they had older male siblings, but no other father substitute, they were more masculine. _L§pgth of Father Absence Some studies strongly suggest that the effects of father absence are more pronounced in the younger child, and the younger the child at father separation, (Hetherington & Deur, 1971; Carlsmith, 1964). The consequences may be more pronounced in males than females (Nash, 1965; Seplin, 1952). 18 Early separation in preschool boys before the age of five has resulted in less aggressive behaviors, less masculine self concepts, and moredependent and less masculine game pre- ferences. In doll play these father absent boys exhibited more verbal aggression, with play patterns more characteristic of preschool girls. They also exhibited less physical aggression than females from intact families (Bach, 1946; Sears, 1951; Sears, Piltner, and Sears, 1946). Stolz (1954) provides some interesting and somewhat con- tradictory evidence regarding effects of length of father absence.*’Four to eight year old boys, who for approximately the first two years of their lives had been separated from their fathers, were less assertively aggressive and independent in their peer relations than boys who had not been separated from.their fathers. In doll play they were more aggressive than non-separated boys. They were however more often overly sub- missive and/or reacted with immature hostility." ”In studying school-aged boys Hetherington (1966) reported father absent boys scored less masculine on a projective test of sex role preferences, were more dependent on peers, less assertive and engaged in fewer physical contact activities than father present boys.“ This finding was present only if separation occurred before the age of five. In family situations where father separations occurred at the age of six or after no effects were evidenced when the boys were compared to other boys in father present families. 19 Surrogate Father Effects The development of sex-typed behavior in children from intact homes has been shown to be closely related to the presence of the father and of older males (Steimal, 1960; Patterson, Littman & Bricker, 1967). Other studies of intact families have shown that there is an association between the presence of male siblings and masculinity (Brim, 1958; Sutton- Smith, Roberts & Rosenberg, 1964). The presence of sibling males or adult males adds another dimension to the complexity of the father absent environment. Fauls & Smith (1956) in studying sex role preferences, found that four and five year old boys who had older brothers had more appropriate sex role preferences than those boys who were only children. Brown (1956) reported that five and six year old boys with older brothers made more masculine sex role preferences on projective measures. Based on teachers' ratings on a nine point masculinity- femininity measure, Brim (1958) in a later analysis of Koch's (1956) study on siblings and masculinity found that boys with same sex siblings, as compared to boys with opposite sex sib- lings, were more sex appropriate in their behaviors. Based on projective measures these early studies report boys with male siblings preferred and exhibited more sex appropriate behaviors than boys without male siblings. Recent studies (Santrock, 1970; WOhlford, Santrock, Berger, & Liberman, 1970) found that father absent four and five year old black boys with older brothers were more masculine 20 than those with older sisters, even though these boys were also more dependent on adults, less aggressive and less masculine than father present boys. Father absent girls showed more aggressive behavior and were less dependent if they had only male sibling. Biller (1968) suggests that although the inter- action with male siblings may reduce the effects of the father absent environment, the presence of the father may be a more salient factor in the development of masculinity. In summary, it is suggested by some authors that at least for those children having older brothers the effects of father absence on boys especially, is less, as the compensatory interaction of the older siblings partially fills the void created by the absent father. The Interface of Father Absence and Social Behaviors The absence of the father during the child's preschool years may be viewed from a perspective which focuses on the socialization process within the family. Relying heavily on the work of Bales, Talcott Parsons suggests that socialization is one of the requisites for a perservering social system, and con- sists of permissiveness, establishment of love of the socializa— tion agent, introduction and presentation of adult norms, and rewards and punishments which bring the child into conformity with adult norms (Parsons & Bales, 1955). In the broadest sense, socialization is "adopting society's rules of behavior" first through social responses, and later through social controls (McCandless, 1967) W.I. Thomas has 21 called "the desire for intimate response" one of the basic things a family gives its members (Thomas, 1956). The interaction analysis of Bales (1957) provides a framework for viewing a broad range of social interactions in small groups. Four types of social responses and initiations may be present during a group interaction: (1) Positive reactions-acceptance (2) Problem solving attempts-suggestions and opinion (3) Questions-ask for opinion (4) Negative reactions-disagreement, tension increase According to Bales and Parsons (1955) these categories of behavior are predictable results of a small group's attempt to problem solve in an activity, and thus the individual's be- havior will vary. In a classical study of social behavior Parten (1933) delineated the individual and group social behaviors of pre- schoolers into six categories: (1) unoccupied-child is not playing but watches anything of momentary interest, (2) solitary - child plays alone with toys different from.those around him with no effort to speak, (3) onlooker - watches others play and may offer suggestions but does not enter the play, (4) parallel - independent play beside others and with similar interests and materials, (5) associatiVe - child interacts with others borrow- ing and lending play materials in often identical activity, (6) cooperative - child plays with an organized group striving to attain a common goal. 22 Louis Murphy (1937) in an extensive study on coopera- tion and sympathy, found an increase in all forms of social interactions of the preschool child. Interaction, initiation, and social participation was greater in four year olds than three and two year olds. These findings have led Murphy and several other investigators to suggest a continuum of personal- ity from.passivity to activity (Murphy, 1937; Andrus & Horowiz, 1938). An increase in activity level may also result from cooperative play as peer interactions and experiences are in- creased. Through these peer experiences the child begins to interpret himself in light of his status within the peer group (Ausubel, Schiff & Glasser, 1952) and begins to be more aware of himself and others. Brown (1956) recognized that a child's awareness of himself or self concept is multidimensional, taking into account many perceptions including perceptions of the self, actual perceptions others have of him, and perceptions of the ways in which he is seen by others. Referents are significant others from whom the child continually draws information about himself. Significant others such as mother, teacher, peers, and self thus shapes the child's self concept. As a measure of self concept Brown developed an instrument which elicits the expressed self concept of a child using mother, teacher, self and peer as referent categories. A series of 14 questions are structured for each referent and 23 asked of the child while he is looking at his own photograph. For example, while the child is looking at his picture the tester asks, "Is Eric good or bad?" Later in another referent category, "Does Eric's teacher think he is good or bad?" The complete set of questions are presented in Appendix A. Through the procedure the child expresses his self concept in terms of himself as well as others (Brown, 1956). Conspicuously lacking in the instrument, however, is the father as a signifi- cant other, or referent. Although the Brown IDS Self-Concept Referent Test has been widely accepted (Coller, 1971), the main criticism.has been its reliance on verbal and conceptual skills that may be re- flected in addition to or instead of feelings about self, rais- ing an additional concern regarding the young child's ability to verbalize. The concern is applicable to this study as only children 3% to 5 years of age were included in the sample. Summary of the Literature If the period between the ages of three years and five years is a period of adopting behaviors of the father (Biller, 1969), then there should be differences in behaviors of father absent and father present preschoolers of these ages. After summarizing the father absent literature however, one must con- clude that few consistent behavioral differences have been found. Also, of the studies that do show differences, few have drawn conclusions from.observed social interactions of children. 24 The inconsistent findings and lack of observational data has led this investigator to ask the following research question: ARE THERE DIFFERENCES IN THE SOCIAL PLAY BEHAVIORS OF PRESCHOOL FATHER ABSENT AND FATHER PRESENT CHILDREN? The available literature suggests that the younger the child, the more pronounced are the effects of father absence. For children below six years, however, does the length of father absence really make a difference in observed social be- haviors? Since little evidence is available, the following research question is posed: DOES THE LENGTH OF FATHER.ABSENCE INFLUENCE THE SOCIAL BEHAVIORS OF FATHER ABSENT PRESCHOOL CHILDREN? The effect of a father surrogate has been found to be a compensatory force in the personality development of the pre- schooler. General masculine and feminine traits have been studied through projective measures while observations of more general social behaviors have not been made. If a father surro- gate is a compensatory force in the development of personality, the following question should be answered in the affirmative: DOES THE PRESENCE OF A FATHER SURROGATE INFLUENCE THE DEVELOPMENT OF SOCIAL BEHAVIORS IN FATHER ABSENT PRESCHOOL CHILDREN? From.the review of the literature one can conclude that little has been written describing the home environment of the father absent family. Family demographic information such as income, occupation, ages of family members, type of dwelling has not been forthcoming from.previous research. More information 25 is needed to assess the total impact of the father's absence. A survey of father absent families to collect demographic data would be most useful. CHAPTER III METHODOLOGY Research Design This is a comparative and descriptive study. The pri- mary sample consisted of 127 children whose father had been absent from.the home for at least 3 months but no more than 5 years, or whose father had not been absent from the home. Two sample groups were formed post hoc which consisted of 80 father absent and 47 father present boys and girls as presented in Table 1. Table 1 Sample Distribution of Father Absent and Father Present Children by Sex Father Present Father Absent N Boys 25 37 62 Girls 22 43 65 Total N 47 80 127 Within the father absent sample, two sub-sample groups ‘were formed based on: (1) length of father absence, and (2) availability of a father surrogate. The first sub-sample (N-80) 26 27 as presented in Table 2 is of children whose father had been absent from 3 months to 2 years (short-term.father absent) and children whose father had been absent from.3 years to 5 years (long-term.father absent). Table 2 Sample Distribution of Father Absent Children by Length of Absence by Sex Short-Term Long-Term N Boys 21 16 37 Girls 24 19 43 Total N 45 35 80 The second sub-sample (N-BO) consisting of father absent children was divided post hoc into two groups consisting of children who had an older male living with them or who did not have an older male living with them, Surrogate availability is presented in Table 3. Table 3 Sample Distribution of Father Absent Children by Father Surrogate by Sex Father Surrogate Non-Father Surrogate N Boys 21 16 37 Girls 25 18 43 Total N 46 34 80 28 A multivariate analyses of covariance (MANCOVA) model in.which age was the covariate, was used for the primary analyses of the sample groups. Operationalization of Variables The independent variables include father absence, length of absence, presence of a father surrogate, sex, and age. The dependent variables include three affective and eight social involvement variables of three and four-year-old child behavior. The affective variables are self-concept, autonomy and adult dependency. The social involvement variables include initiation, response, aggression, social behavior, involvement, activity level, heterogeneity of initiation, and tolerance. The relative degree of social behavior may be inferred from.a combination of specific observed behaviors. While each behavior is observed at one point in time, multiple observations taken across time in small and large play group situation, provides the basis for drawing generalizations. Operationalization of Independent Variables The primary independent variables as described below were gathered from.information collected on the parent informa- tion sheets. S35 - Male and female children were included in the sample. Agg - The children's ages were determined as of January 1, 1974, and are reported in months. 29 Father absence - The mother's report of the father's absence in years was used as the basis for determining father absence. Short-term father absence - Short term father absence was determined on the basis of father absence of 2 years or less as reported by the mother. Longrterm father absence - Father absence of 3 years to 5 years as reported by the mother was categorized as long-term. Father surrogate - A male (relative or non-relative) who was older than the child was considered a father surrogate. Included in this group were older male siblings, other relatives, and male friends of the mother. Ethnicity - Most of the children in the sample were Caucasian or Black. Too few were American Indian or Mexican to comprise a group. The term "Black" was ascribed when either parent or one parent was Negro. Operationalization of Social Affective Behaviors Self-concept - A self-concept score represents the ex- pressed self-concept of the child through the mother and self referent category as the sum of positive responses divided by the total number of scorable responses. Autonomy - A dimension of socio-emotionality is autonomy. The behavioral cues range on a five point scale from self directedness to dependent. Adult dependency - The average number of intervals of child-initiated interactions with adults was the basis for adult 30 dependency. Interactions with adults were recorded as were interactions with peers. Operationalization of Social Involvement Variables InitiatiOn - The frequency with which the child initi- ates interactions regardless of whether or not that behavior prompts a response from others is the basis for initiation scores. The logit1 of the proportion of intervals with initi- ations versus the proportion of intervals without initiations is the conceptual formula. Response - Similar to initiation scores, response is the frequency with which the child responds to others' interactions. The logit of the proportion of intervals with responses versus the proportion of intervals without responses is the conceptual formula. Aggression - Behaviors which are generally unacceptable to society (e.g., hitting, pushing) are coded as aggressive. verbalizations rated as aggressive were negative voice tones. Verbal and physical negative categories were combined to form the variable for aggression and the subsequent score. The con- ceptual formula is the logit of the proportion of intervals with negative scores versus the proportion of intervals without 1To stabilize proportional data for use in parametric analytic models, the natural log of the proportion is formed and this new figure is used in the analyses. 31 negative scores. Social behavior - The average rating of the child's social behavior over all intervals reflects the child's characteristic behavior. Parten's (1933) six point scale of social behavior was the basis for rating behaviors. Behaviors range from unoccupied to cooperative play behavior. Involvement - This variable is identical to the above social behavior variable but was rated on the classroom socio- observations. Activity level - The relative amount of activity in- volved in initiations and responses is determined by the mean level of activity occurring with all initiations and responses. The low score reflects a high mean level of activity while a high score reflects passive activity. Heterogeneipy of initiation - An indication of the child's willingness to initiate to the opposite sex is reflected in scores of heterogeneity of initiation. The score is the logit of the proportion of initiations with unlike sex peers over initiations with like sex peers. Tolerance - Similar to heterogeneity of initiation scores, tolerance scores indicate the child's willingness to respond to the initiations of unlike sex peers. A high score indicates the child's willingness to respond to the initiations of unlike sex peers while a low score indicates few responses to the initiations of unlike sex peers. The score is the logit of the proportion of responses to unlike sex peers over in- itiations with like sex peers. 32 Operationalization of Family Demographic Characteristics In addition to the independent and dependent variables mentioned above, the following descriptive dimensions were operationalized: (1) (2) (3) (4) (5) (6) (7) ethnicity — A child was considered Black if either or both parents were Negro. He was considered Anglo if both natural parents were Caucasian. day care experience - A half day equivalent was considered 4 hours of day care. Ten to twelve half day equivalents were considered a week. center size - Franchised centers were privately owned and administered. Non-franchised centers were non-profit public and privately administered day care centers. Size is reported in numbers of children enrolled. ordinal position - Ordinal positions were first through ninth child. females - Adult and sibling females were females in the home. playmates at home - Playmates in the family environ- *ment included older, younger, ageemates, and combi- nations of these children. A second category in- cluded siblings, relatives, friends, and neighbors, siblings and relatives, and relatives and friends. age of mothers and fathers - When available, ages of absent fathers were reported. 33 (8) education of mothers and fathers - Categories in- cluded: less than 12 years; high school with no occupational training; less than 12 years plus training; high school and occupational training; some college; college degree of a BA; advanced college degree of a MA or Ph.D. (9) hours working - Categories included: no work; less than 20 hours per week; 20-29 hours per week; 30-39 hours per week; and 40 plus hours per week. (10) occupation - Parents were categorized into six occupational types: (1) semi-skilled, (2) clerks, service, (3) sales work, foreman, (4) managers, (5) nurses, teachers, and (6) professionals, lawyers, doctors. (11) income by center report - Fees for day care services ‘were paid in several ways. Aid to Dependent Children or ADC involved the total payment of day care ser- vices by the State of Michigan Social Services. A second method included partial payments from.ADC, the family, and scholarships in covering the total cost. The third alternative was for the family to pay the full amount. (12) income by parent report - weekly income categories included: ADC, $50-$75, $76-$100, $101-$125, $126- $150, $151-$175, $176-$200, and $200 and over. 34 (13) dwelling - Categories included single dwellings which were single houses and trailers, and multiple dwellings which were apartments and duplexes. (l4) transportation to center - Categories included: walking, family car, public transportation, day care bus, and friends as transportation of child to center. (15) transportation time to center - The length of time spent transporting the child to the center was categorized as either 5 to 15 minutes, 20 to 30 minutes or 50 to 60 minutes. Hypotheses The research hypotheses involve three areas of inquiry: (1) Are there differences in the social behaviors of father absent and father present preschool children? (2) Does length of father absence influence social behaviors and (3) Do father surrogates in the family of father absent children influence their social behaviors? The following specific research ques- tions (A through C) and hypothesis statements were based upon these general research questions. The null hypotheses (HO1 through H018) are identified following each directional hypothesis (H1 through H18)' A. WHAT IS THE EFFECT OF FATHER ABSENCE ON THE AFFECTIVE BEHAVIORS OF PRESCHOOL CHILDREN? H1: Father absent children will have lower self-concept scores than father present children. B. INVOLVEMENT H4: 35 H01: There are no differences in the expressed self-concept scores of father absent and father present children. : Father absent children will have higher adult dependency scores than father present children. H02: There are no differences in the dependency scores of father absent and father present children. : Father absent children will have lower autonomy scores than father present children. H03: There are no differences in the autonomy scores of father absent and father present children. WHAT IS THE EFFECT OF FATHER ABSENCE ON THE SOCIAL BEHAVIORS OF PRESCHOOL CHILDREN? The activity level of classroom social involvement will be lower for father absent children than for father present children. H04: There are no differences in the classroom social involvement scores of father absent and father present children. : The activity level scores of father absent child- ren will be lower than father present children. H05: There are no differences in the activity level scores of father absent and father present children. : Father absent children will have higher initiation scores than father present children. H06: There are no differences in the initiation scores of father absent and father present children. : Father absent children will have higher response scores than father present children. H07: There are no differences in the response scores of father absent and father present children. H10: C. 36 : Father absent children will exhibit more verbal and physical aggression than father present children. H08: There are no differences in the verbal and physical aggression scores of father absent and father present children. : Father absent females will initiate to the opposite sex less than father present females. H09: There are no differences between father absent girls and father present girls in their initiations scores to boys. Father absent males will initiate to male children more than father present boys. H010: There are no differences between father absent boys and father present boys in their initiations to girls. DOES THE LENGTH OF FATHER ABSENCE INFLUENCE CERTAIN SOCIAL BEHAVIORS OF FATHER ABSENT CHILDREN? H11: Long-term (3-5 years) father absent children will have higher autonomy scores than short-term (2 years and less) father absent children. H011: There are no differences in autonomy scores of short-term (2 years and less) and long-term (3-5 years) father absent children. : Short-term (2 years and less) father absent child- ren will have higher adult dependency scores than long-term (3-5 years) father absent children. H012: There are no differences in adult dependency scores of short-term (2 years or less) and long-term.(3-5 years) father absent children. : Long-term (3-5 years) father absent children will have lower self-concept scores than short-term 12 years and less) father absent children. H013: There are no differences in self-conce t scores of short-term (2 years and less) and long-term.(3-5 years) father absent children. : Long-term (3-5 years) father absent children will have higher ag ression scores than short-term (2 years and less father absent children. D. 37 H014: There are no differences in aggression scores of short-term (2 ears and less) and long-term (3-5 years father absent children. DO FATHER SURROGATES AFFECT THE DEVELOPMENT OF SOCIAL INVOLVEMENT BEHAVIORS? H H16: E. 15‘ Father absent children with a father surrogate will have higher peer associative and coopera- tive behavior scores than father absent child- ren without a father surrogate. H015: There are no differences between child- ren with a father surrogate and children without a father surrogate on peer associative and cooperative behavior. Father absent children with a father surrogate will have higher initiation and response scores than father absent children without a father surrogate. H016: There are no differences between children with a father surrogate and children with- out a father surrogate on initiation and response scores. DOES A FATHER SURROGATE AFFECT THE DEVELOPMENT OF AFFECTIVE SOCIAL BEHAVIORS? H H18: 17‘ Father absent children with a father surrogate will have lower adult dependency scores than father absent children without older males in the home. H017: There are no differences between children with a father surrogate and children without a father surrogate on adult dependency scores. Father absent children without a father surrogate will have higher self-concept scores than father absent children without a father surrogate. H018: There are no differences between children with a father surrogate and children without a father surrogate on self-concept scores. 38 Sampling'Procedures The present investigation is part of a larger study sponsored by the Office of Child Development and implemented through the Institute for Family and Child Study, Michigan State University. The study, entitled "Early Social Development: ' was an evaluation of the relative Parent and Child Programs,‘ effectiveness of short-term parent and classroom programs on day care children's self concept and social involvement with peers (Roger & Andrews, 1975). Using a quasi-experimental design, eight day care centers were sampled and randomly assigned to one of four treatment conditions. Data were collected before and after the program implementation period during the fall of 1973 and spring of 1974. A part of the initial data collected for the larger study was used in this research. Sample Selection Initial information concerning potential cooperating centers was secured through the State Department of Social Services Day Care Licensing Division and area Community Coordi- nated Child area (4-C) coordinators. The present investigator was involved in the screening of potential centers against basic criteria and assisted in contacting and visiting potential centers. The original criteria for center eligibility included the following: 1. Distance from.Michigan State University--maximum of 70 miles. 39 2. Listing with the licensing division of the State Department of Social Services. 3. Offering a full-day program 4. Comparable philosophy, program, and staff quali- fications 5. No simultaneous participation with other research or program obligations 6. Heterogeneous enrollment of children to meet the following needs: a. age range -- 3% to 5 years b. enrolled for four half-days per week c. minimum of eight children in each of four groups (excluding kindergartners): low and mid SES males and females d. racial balance across SES groups or all one race. To secure an adequate number of children within the age and enrollment range, medium-to large-sized centers were approached. All centers considered met the first five criteria. The distribution of children across sex, SES, and ethnic groups was the most difficult sampling criterion to satisfy. The centers selected offered the best balance to enrollment of those centers available and willing to participate. The eight centers selected were located in four large cities in lower Michigan. Their enrollments ranged from.76 to 166 children per center. The characteristics of the children enrolled in these eight centers 40 varied somewhat. All children within the age of 3% to 5 years (as of January 1, 1974) who were enrolled for at least four days per week at each center were considered eligible to be included in the sample. The parents of the children were informed by the center director or head teachers as to the nature and content of the research. A parent permission and information sheet (Appendix A) was sent home with each child or filled out at the center by the parent. Follow-up telephone calls from the center and per- sonal contact were made to ensure return of the information sheets. Of those parents who responded (200) only those child- ren (127) who had the most complete demographic information in this first stage of data collection, were included in the sample. Descrtption of the Sample The distribution of the children by father status, age, and sex is described in Table 4. The sample was rather evenly divided by sex. More four- year olds were included than three and five-year-olds. In Table 5 the frequency distribution of father absent children and number of years of father absence is presented. Children who had been father absent for up to two years were categorized as short-term.father absent while those who had been father absent for more than two years were defined as long-term father absent. About half of the children were father absent from three to five years. 41 Table 4 Sample8 Description by Father Status, Age and Sex Characteristic N Z Father Status Father absent 8O 63 Father present 47 37 Short-term father absence 45 56 Long-term father absence 35 44 No father surrogate 34 43 Father surrogate 46 57 Age Three year olds 28 22 Four year olds 94 74 Five year olds 5 4 Sex Boys 62 49 Girls 65 51 “iv-127 Table 5 Length of Father Absence by Sex Length of Absence Years of Absence l 2 3 4 5 Males ll 10 8 6 2 Females 12 12 4 12 3 n 23 22 12 18 5 TOTAL N - 80 42 Half of the four-year-olds were short-term father absent and half were long-term father absent. Twenty-three percent were father absent since birth and a little more than half had a father present for one year. A description is provided in Table 6. Table 6 Length of Father Absence by Age Length of Absence Short-term Long-term Years of Absence 1 2 3 4 5 3 year olds 5 6 l 1 0 4 year olds l6 l6 8 16 5 5 year olds l 0 1 1 0 n 22 22 10 18 5 TOTAL N=772 Another independent variable of interest was the pre- sence of a father surrogate. The presence of older males, brothers, uncles, grandfathers, or friends of the mother living in the home were considered father surrogates. Few (8%) of the father absent children had male adults as father surrogates, although one father absent boy and two 2Data'were not reported by three families. 43 father present girls had two male adults living in their homes. Half of the father absent boys and half of the father absent girls had older brothers. Twenty percent of the father absent boys and 14% of the father present girls had older brothers, as reported in Table 7. Table 7 Older Unlike Sex Siblings by Father Status by Sex Number of Unlike Siblings 0 l 2 3 4 5 Father Absent Males 25 8 3 0 1 0 Females 28 ll 2 2 0 0 Father Present Males 20 4 0 0 0 1 Females 19 2 0 0 1 0 N = 127 Instrumentation Three instruments were employed to collect data on the social and affective behaviors of preschool father absent and father present children: (1) The Observation of Social Behavior (OSB) (Boger, Cunningham, Andrews, 1974), (2) Classroom Socio- Observation (Cunningham, Reyes & Andrews, 1973), and (3) The Brown Self-Concept IDS Referent Test (Brown, 1966). A parent information sheet was used to survey the families participating in the study. 44 Instrument Description: The Observation of Social Behavior To study small groups of children in a naturalistic setting the small group social-interaction, a revised version of the methodology (The Observation of Social Behavior) deve10ped by Boger and Cunningham (1969) was used. Observa- tional data were gathered during a ten minute play session in which four children (one male and one female from low and middle socioeconomic groups) were randomly selected and asked to play in a mobile playroom. The playroom.was in a specially designed mobile trailer equipped with audio microphones and a video tape recording camera. The dimensions of the carpeted playroom were approx- imately 11 feet by 8 feet. The space was divided by a low partition behind which the examiner could sit and observe. An additional space, partitioned by a full wall with a one-way observation window was provided for an audio-video technician. During the play session the only materials provided were large cardboard boxes of different sizes. These boxes are similar to those found in a supermarket. The children were not directed in any way in their play. "There is no overt indica- tion of behavioral expectations, and there is no attempt co guide, limit, or structure behavior." (Boger and Cunningham, 1971). The children were brought to the trailer and familiar- ized with the room, (The mobile unit was parked in the parking lot of their day care center so they were somewhat familiar with 45 it.) A statement was read about the length of time they would be there and that they could play any way as long as they did not hurt anyone. The examiner then told the children she had some papers to read and would be sitting in a chair behind the low partition. The ten minute play session was videotaped and later rated and coded using the revised version of the Observation of Social Behavior Instrument (Boger, Cunningham, & Andrews, 1974). "The revised OSB instrument is an observation technique using a combination time and event-sampling procedure. At ZO-second intervals a mechanical beep is superimposed on the audio portion of the tape. Raters record the first play behavior at each 20- second mark, thus securing a time sampling of behaviors across the ten minute play session for each child." (p. 9) The instrument is designed to measure 14 behavioral dimensions. For the purposes of this study however, behavioral dimensions were chosen. They include interaction, object of in- teraction, level of involvement, voice tone, physical behavior, physical tone, social behavior and autonomy. The description of these eight behavioral dimensions included in this study were drawn from the original instrument description of Boger and Andrews (1975). InteraCtion and Involvement--Each interaction is rated as initiation or response. Initiation is defined as an intro- duction of self or change in activity. A response included acceptance of another's interactions (A), rejection of another's initiation (R), no acknowledgement nor awareness of another's 46 initiation (N), ongoing interaction (0), and behavioral transition or eminent initiation (X). Following a response, an initiation may or may not occur. Three levels of involve- ment ranging from.intense to passive are rated for each of the two major categories. The object of the involvement group (undifferentiated, adult, individual, or pairs of individuals, materials, or environment) are also recorded. VerbalizatiOn-- The measure of quantity of verbaliza- tion and the behavioral context, through time sampling, permits analysis of the relationship between verbalizations and other behaviors. The coding of verbalizations is based on the Bales (1951) Interaction Process Analysis. Twelve mutually exclusive and exhaustive categories plus mumbling (unintelligible) are possible with a complete verbal interaction being considered in the unit. The affective delivery in voice tone is rated on a three-point scale of posi- tive, negative or neutral. Physical Behavior-- The non-verbal behavior of the child may be rated in respect to the object of physical contact. The human interaction is considered first when the objects of interaction are materials and people simultaneously. Direct or indirect physical touching or touching through the medium.of play materials (boxes) is considered. As with voice tone, an aspect of physical behavior is its positive or negative quality in reference to the affective nature and social acceptability of the behavior. Positive 47 qualities are tapping, patting, and caressing. Negative quali- ties are hitting, pushing, and kicking. Behaviors such as running or building are considered neutral non-affective activ- ities. Social Behavior-- Categories adopted to measure the child's social behavior are based on the ordinal scale developed by Parten (1932). These categories include unoccupied play, solitary play, onlooker, parallel play, associative play and cooperative play. Included as criteria for the categories is spatial proximity to other children, similarity of materials, and nature of the interaction, and goal-directedness of the play. Socio-emotional Dimensiopg-- Specific behavioral cues help define the general tone of the child's social and emotional behavior in four areas: autonomy, social leadership, social competency and emotionality. Each general behavioral tone is rated on a five point scale. The extreme ratings of positive (5) and negative (1) represent the overt behaviors in the seg- ment of interaction. The neutral position (3) represents non- observable behavior. The two intermediate positions (4) and (2) represent covert behavioral cues or mild overt behavioral indications of the dimension. For the purpose of this study the primary variables were formed based on frequencies, means, and proportions of time spent in various behavioral categories. Secondary variables concerned with (l) contingent frequencies of one behavior occurring simultaneous with another (e.g. verbal command with physical contact of a negative nature). 48 Reliability-- To maintain a high degree of agreement between raters, behavioral units must be objectively encoded and recognizable to reflect the validity of the categories of behavior. Inter-rater reliability is a form of establishing and maintaining high inter-rater agreement. The minimal level of 85% on the total recordable posi- tions was adopted for this study with an actual percentage of agreement ranging from 86% to 98% agreement between any two raters over a ten minute sequence of play activity. To control for "instrument decay" or the gradual departure from.ooncensus, regular group discussions and inter-observer checks were con- ducted. The internal consistency of behavioral units coded may be assessed by an analysis of variance method which indicate consistency over sampled items at the same point in time. In this study only those variables requiring a code during each interval were analyzed. Instrument Description: The Classroom Socio-Observation To observe children and their peer interactions in a classroom environment, the sociometric measure developed by Cunningham, Reyes, and Andrews (1973) was employed. Twelve children (3 low SES girls, 3 mid SES girls, 3 low SES boys and 3 mid SES boys) were randomly chosen to play together in their classroom or a portion of a classroom. Since classroom parti- cipating in this study did not have equal representation from all demographic groups, this procedure was implemented to allow 49 each child the same probability of associating with a like vs. an unlike peer in regards to SES and sex. A teacher was present only to organize the environment but not to supervise the play with manipulative toys, dramatic play materials or art activities. The materials and arrange- ‘ment of the room.were not changed as the classroom of the child- ren was used for the observation. The variables derived from.the c1assroom.socio- observations are: 1. level of social involvement - mean of social behavior ratings over all intervals. 2. peer proximity and associations - average number of children in proximity or in inter- action with S over all intervals. 3. heterogeneity of peer associations - number of intervals S'is in interaction with a peer of a different sex or SES. 4. consistency of play behavior - the duration of play with each peer in relation to level of social involvement over three consecutive intervals. A series of three consecutive observations was taken at the beginning and toward the end of 30 minute play periods. Each child in the sample was observed on two and sometimes three separate days. Behaviors are rated using Parten's (1932) scale of social development relating to six levels of activity. The 50 levels are: unoccupied, solitary play, onlooker behavior, parallel play, associative play and cooperative play. A more detailed description is presented in Appendix A. Reliability - To establish reliability, observers were trained and practiced at the Laboratory Preschool on Michigan State University's campus. A 90% inter-observer agreement was required prior to data collection. The actual inter-observer agreement attained was 99%. On the variable, level of social behavior, an internal consistency coefficient of .81 was obtained over three consecu- tive observations. Instrument Description: 'The Brown IDS Self—Concept Referent Test The Brown test was developed to assess the self-concept of children by using their photographs. The photograph is used to induce the child to take the role of another towards himself. The feelings of the child towards himself (self-as-subject), and his perception of his mother, teacher and peers (self- as-object) are measured. In this study only the mother and self referent were administered. A separate room.in the day care center was used to administer the test which lasted approximately 5 to 10 minutes. After a self developing picture was taken of the child and the tester was certain the child recognized the picture as that of himself, the child was asked to respond to 14 bipolar items (e.g., Does (child's name) think he is good or bad?) All 14 self referent items are presented followed by the same items 51 in the mother referent context (e.g. Does (child's name) mother think (child's name) is good or bad?) Scoring consists of positive or negative on each item. The sum of the positive responses divided by the total number of scorable responses represents the self and mother referent scores. The analysis was performed on the mother score, total number of omits, and discrepancy score (sum of items with differences between responses for the self and mother referents.) Reliability - The reliability of the Brown has been evaluated in previous research. Coefficients of .76 and .81 (Boger, Kupiers, Cunningham, & Andrews, 1974) were reported for the mother and self respectively based on a sample of 3% to 5 year olds in day care settings. The 1971 National Follow- Through Evaluation reported an internal consistency coefficient of .82 but test-retest reliability for 632 S's after a 2-3 week interval was only .55 (Shipmen, 1972). Instrument Description: Parent Information Sheet To obtain demographic data a General Information Sheet developed by Andrews and Reyes as presented in Appendix A was sent home or given to each parent by the day care director. Follow-up phone calls by the director were made as well as per- sonal contact at the day care center to secure the questionnaires. Thirty-four general questions were structured in a "non-threatening" manner regarding ages, sex, and relationships of persons living in the home. Additional information pertained to income level, type of employment, number of hours worked and educational levels of the parents. 52 Information sheets were sent out to 200 families and 180 were returned. Of these 127 had complete data relating to the presence or absence of the father. A detailed analysis is presented in Chapter V. The protocol is in Appendix A. Reliability - No attempt was made to determine the reliability of the data reported by the parents except in one instance. Day care centers receiving Aid to Dependent Children (ADC) from the State of Michigan, reported those figures as additional data to the parent information sheet. The discrep- ancies in information reported by the parent and reported by the day care center are discussed in Chapter V. Data Gathering Procedures Data collection was the responsibility of the trained research staff of the Institute for Family & Child Study, who were graduate students from the Department of Family & Child Sciences. Their activities were coordinated by a senior research assistant. The present investigator participated in the selec- tion of day care centers, sample selection and subsequent pre- testing treatment input reported in this study. The video-tape play sessions, classroom socio-observations, self-concept test, and collection of information sheets were the responsibility of the research staff. Undergraduate students 'with previous experience working with young children and who were pursuing degrees in the social sciences were trained in rating the OSB. Training sessions, discussions, and simultaneous ratings by more than one rater provided experience in rating the tapes. 53 An inter-rater agreement of 85% total recordable positions was required. Data ReduCtiOn and Analyses Data ReduCtion'Procedures The computer programs used in data analysis were those of the 6.0 version of the Statistical Package for Social Sciences (Wie, Bent & Hall, 1970) and the adapted version of Finn's Multivariance program (schmidt & Scheifley, 1972). The analyses were completed on the Control Data Corporation 6500 computer at the Michigan State University Computer Laboratory. A chance probability of .05 was used as the decision rule for hypothesis testing. Correlation, multiple step-wise regression, and two-way multivariate analysis of covariance were required to test the hypotheses of interest. The three research questions concerning father absent and father present children were: (1) What is the effect of father absence on the social behaviors of pre- school children? (2) Does the length of father absence affect the development of social behaviors? (3) Do older males or sib- lings in the family affect the development of social behaviors in children from.father absent families. Frequency distributions and descriptive statistics are presented for a number of demo- graphic variables. Measures'of‘Association CorrelatiOn: The degree to which two variables vary together may be 51 in the mother referent context (e.g. Does (child's name) ‘mother think (child's name) is good or bad?) Scoring consists of positive or negative on each item. The sum of the positive responses divided by the total number of scorable responses represents the self and mother referent scores. The analysis was performed on the mother score, total number of omits, and discrepancy score (sum of items with dif- ferences between responses for the self and mother referents.) Reliability - The reliability of the Brown has been evaluated in previous research. Coefficients of .76 and .81 (Boger, Kupiers, Cunningham, & Andrews, 1974) were reported for the mother and self respectively based on a sample of 3% to 5 year olds in day care settings. The 1971 National Follow-Through Evaluation reported an internal consistency coefficient of .82 but test-retest reliability for 632 S's after a 2-3 week interval was only .55 (Shipman, 1972). Instrument Description: Parent Information Sheet To obtain demographic data a General Information Sheet develOped by Andrews and Reyes as presented in Appendix A was sent home with each parent by the day care director. Follow- up phone calls by the director were made as well as personal contact to secure the questionnaires. Thirty-four general questions were structured in a "non-threatening" manner re- garding ages, sex, and relationships of persons living in the home. Additional information pertained to income level, type of employment, number of hours worked and educational levels of the parents. 52 Information sheets were sent out to 200 families and 180 were returned. Of these 127 had complete data relating to the presence or absence of the father. A detailed analy- sis is presented in Chapter V. The protocol is in Appendix A, Reliability - No attempt was made to determine the reliability of the data reported by the parents except in one instance. Day care centers receiving Aid to Dependent Children (ADC) from the State Of Michigan, reported those figures as additional data to the parent information sheet. The discrepancies in information reported by the parent and reported by the day care center are discussed in Chapter V. Data Gathering Procedures Data collection was the responsibility of the trained graduate research staff of the Institute for Family & Child Study. Their activities were coordinated by a senior research assistant. The present investigator participated in the selection of day care centers, sample selection and sub- sequent pretesting treatment input reported in this study. The video-tape play sessions, classroom socio- observations, self-concept test, and collection of information sheets were the responsibility of the research staff. Under- graduate students with previous experience working with young children and who were pursuing degrees in the social sciences were trained in rating the OSB. Training sessions, discussions, and simultaneous ratings by more than one rater provided experience in rating the tapes. 54 expressed by computing a correlation coefficient. The Pearson product-moment correlational coefficient may be computed when both variables are measured on a linear interval scale. The formula for this correlation coefficient is: When the covariance of x and y are divided by the square root of each individual variance, the quotient of correlation coefficient is obtained (Glass & Stanley, 1970). The strength of the relationship between the two variables is indicated by the value from a perfect relationship (1), to no relationship (0). The direction of the relationship is indicated by the sign, either minus (-) or inverse, or positive (+) and direct. The two basic assumptions of the statistics are: l. A linear prediction of the relationship. 2. Equal appearing intervals of the measurement scale. Based on the nature of the hypothesis the first assumption is satisfied. All of the dependent variables in this study could be considered continuous and of an interval scale. The social interaction variables were either logs of proportions or means, both based on intervals derived by a divisor. Many of the in- dependent variables such as length of father absence, age, education, etc. were also continuous and interval. The metric varied from.one independent variable to another such as months, years, and levels of school. These were equal appearing in all cases . 55 Regression: Hypothesis stated in the predictive mode require an extension of the correlation coefficient in regression analy- sis. A regression line may be predicted based upon the relationship between one variable and another. Regression analysis produces a regression line that minimizes the devi- ations between the observed and predicted values of the depend- ent variable for each value of the independent variable. Least squares is the method implemented with the following equation for the regression line: Ye = a + bX The dependent variable value (Y) is the sum.of the Y intercept plus b times X (the beta weight or constant representing the slope of the line times the observed value of the independent variable.) Multiple independent or dependent variables may be used also. Multiple Regression: To produce a linear combination of independent variables that will correlate as highly as possible with the dependent variable, multiple regression analysis is implemented. The pre- diction equation of multiple regression is: Ye = blxl + bZX2 + b3X3 ... + a + R The beta coefficients are chosen to make the residual or error (R) as small as possible. 56 If one wishes to choose independent variables to give the best prediction, then stepwise rggression is the statistical tool. A limitation of this procedure is that the results of the regression analysis are dependent on the ordering of the independ- ent variables. A strength is that specific orderings may be chosen to test specific hypothesis. The coefficient which indicates how closely the two or more variables are associated is R2. The formula is: R2 = SS linear regression SSitotal The proportion of the total variation in Y associated with X is R2. The assumptions of multiple regression and stepwise regression are: l. Normality 2. Homoscedasticity -- the variation in Y is constant for all changes in the value of X. 3. Linearity -- the relationship between the independent and dependent variables is linear. There is no reason to believe that these assumptions have been violated. Multiple regression was used to determine which independent variables were the strongest predictors. ‘Analysis of Variance: The statistical tool best suited to studying group differ- ences is the analysis of variance model (ANOVA). Interactions 57 and main effects may be tested as well as extending the basic model to include multiple dependent variables and covariates. The assumptions of the basic model are: 1. Measures are continuous with equal-appearing intervals. 2. Normality -- the samples have been chosen from populations that are normally distributed. 3. Homogeneity of variance -- the variance within groups are statistically alike. 4. Independence -- observations are independent of one another both within and between groups. The first two assumptions have previously been discussed and satisfied. To check the assumption of equality of variance across levels of the independent factors, the Levenes' Test for Equality of Dispersion (Glas & Stanley, 1970) was used. With this test, the absolute amount of variance in each observation relative to the group mean is entered as the score value into the analyses of variance computations. The null hypothesis of no differences across levels is accepted when equality of vari- ance exists on that dependent variable. The results of these analyses and a discussion of the consequences of violation of equal variance is presented in Appendix B. Children wererandomly selected to play in small group situations. The natural large play situation also provided for 58 randomly formed play groups and thus the fourth assumption of independence of observation is satisfied. Multivariate Analysis of Covariance: When there is reason to believe that an independent variable may be the primary source of variation between groups, analysis of covariance is used to adjust for those initial differences. To explore the question of whether father absence significantly influences social behaviors, age was used as a covariate in the multivariate analysis of covariance. The assumptions for analysis of covariance are the same as those for analysis of variance and these have previously been discussed and shown to be satisfied. CHAPTER IV RESULTS The results from two sets of data are presented. First, in Section 1 the results for each hypothesis are presented and grouped by general research questions. In each case the general research question and hypothesis are again presented to aid the reader. The results of the descriptive analysis of the family demographic data are presented in Section 2. Section 1 A. WHAT IS THE EFFECT OF FATHER ABSENCE ON THE AFFECTIVE BEHAVIORS OF PRESCHOOL CHILDREN? H1: Father absent children will have lower self-concept scores than father present children. H01: There are no differences in the expressed self concept scores of father absent and father present children. H2: Father absent children will have higher adult dependency scores than father present children. H02: There are no differences in the dependency scores of father absent and father present children. H3: Father absent children will have lower autonomy scores than children from.father present homes. H03: There are no differences in the autonomy scores of father absent and father present children. Initial regression analysis indicated that age was associated with the affective social variables at the .02 level 59 60 of chance probability. Based on these results, age was employed as the covariate in the multivariate analysis of co- variance CMANCOVA) testing for group differences between" ~ father absent and father present children. The results are presented in Table 13. Table 13. RESULTS OF TWO-WAY MANCOVA ON AFFECTIVE SOCIAL BEHAVIORS OF FATHER ABSENT AND FATHER PRESENT CHILDREN Factors 1 4. F Ratio ...... Degrees of.Freedom .. Probability Main Effects Father Status 1.3862 3 & 120 .2504 Sex 1.6857 3 & 120 .1738 Interaction .5018 3 & 120 .6818 There were no interaction effects between sex and family status nor were there main effects of sex and family status. Therefore the null hypothesis of no differences in the affective social behaviors of father absent and father present children, H01, H02, H03, and H04 is not rejected. B. WHAT IS THE EFFECT OF FATHER.ABSENCE ON THE SOCIAL INVOLVEMENT BEHAVIORS 0F PRESCHOOL CHILDREN? H4: The level of classroom social involvement will be lower for father absent children than for father present children. H5: H6: H7: H8: H9: H10: 61 H04: There are no differences in the classroom social involvement scores of father absent and father present children. The activity level scores of father absent child- ren will be lower than father present children. HO : There are no differences in the activity 5 level scores of father absent and father present children. Father absent children will have higher initiation scores than father present children. H06: There are no differences in the initiation scores of father absent and father present children. Father absent children will have higher response scores than father present children. HO : There are no differences in the response scores of father absent and father present children. Father absent children will exhibit more verbal and physical aggression than father present child- ren. H08: There are no differences in the verbal and physical aggression scores of father absent and father present children. Father absent girls will initiate to the opposite sex less than ather present girls. H09: There are no differences between father absent girls in their initiation scores to boys. Father absent boys will initiate to boys more than father present boys will initiate to boys. H010: There are no differences between father absent boys in their initiations to boys. To detemmine the strength of association between age and the social involvement variables, multiple regression analysis was performed. The results indicate that age was associated with social involvement variables at the .02 level of chance 62 probability. Based on these results age was the covariate in the multivariate analysis of covariance testing for group differences on social involvement variables between father absent and father present children. The results of these analyses are presented in Table 14. Table 14 RESULTS OF TWO WAY MANCOVA ON SOCIAL INVOLVEMENT BEHAVIORS ON FATHER.ABSENT AND FATHER PRESENT CHILDREN Factors ‘ F Ratio Degrees of Freedom Probability Main Effects Sex 3.6075 8 & 115 .0009* Father Status .8154 8 & 115 .5905 Interaction .7218 8 & 115 .6720 *P5.05 There were no interaction effects and no effects of father status. There were sex effects for the involvement vari- ables as noted in Table 14. The social involvement variables of social behavior, activity level, heterogeneity of initiation, and tolerance accounted for the significant differences between boys and girls, as presented in Table 15. 63 Table 15 VARIABLES CONTRIBUTING TO MAIN EFFECTS OF SEX ON FATHER.ABSENT & FATHER PRESENT CHILDREN Variable Univariate F Ratio Degrees of Freedom. Probability Involvement 2.1797 8 & 115 .1425 Social Behavior 4.2952 8 & 115 .0404* Activity Level 6.0624 8 & 115 .0153* Initiation 1.6470 8 & 115 .2019 Heterogeneity of Initiation 14.7053 8 & 115 .003* Response 1.0697 8 & 115 .3031 Tolerance 4.3560 8 & 115 .0390* Aggression 1.6819 8 & 115 .1972 *P 5.05 Adjusted cell mean differences between girls and boys (Table 16) on social behavior indicate that girls interacted less at associative and cooperative play than boys and were more passive (l-active, 3=passive) than boys. In their initiations with others as measured by heterogeneity of initiation, girls initiated more with boys than boys initiated with girls. Girls were also more tolerant of initiations with the opposite sex than were boys. 64 Table 16 ADJUSTED CELL MEANS OF VARIABLES CONTRIBUTING TO FATHER.ABSENT & FATHER PRESENT MAIN EFFECTS OF SEX Variable Boys Girls Total Social Behavior 4.447 4.2253 4.3336 Activity Level 2.122 ~ 2.2270 2.1759 Heterogeneity of Initiation .7139 1.4780 1.1050 Tolerance .0611 .4682 .2695 Since the results of main effects for father absence were not significant, the null hypothesis 4 through 10 of no differences in social involvement, activity level, initiation, response, physical and verbal aggression, and heterogeneity of initiation cannot be rejected. C. DOES THE LENGTH OF FATHER ABSENCE INFLUENCE CERTAIN SOCIAL BEHAVIORS OF FATHER ABSENT CHILDREN? H11: Children of father absence of more than 2 years ‘will have lower autonomy scores than children of father absence of less than 2 years. H011: There are no differences in autonomy scores of short-term (2 ears & less) and long-term.(3-5 years father absence. H12: Children of father absence of more than 2 years will have higher adult dependency scores than children of father absence of less than 2 years. 65 H012: There are no differences in adult dependency scores of short-term (2 years & less) and long-term (3-5 years) father absent children. H13: Children of father absence of more than 2 years will have lower self-concept scores than children of father absence of less than 2 years. H013: There are no differences in self-concept scores of short-term (2 years & less) and long-term (3-5 years) father absence. H14: Children of father absence of more than 2 years will have lower scores of aggression than children of father absence of less than 2 years. H014: There are no differences in aggression scores of short-term (2 years & less) and long-term (3-5 years) father absent children. Group differences between short-term (2 years and less) and long-term (3-5 years) father absent children were explored through hypotheses 11, 12, 13, and 14. To determine the strength of association between age and the behavioral variables, multiple regression analysis was performed. The results indicate that age was associated with the social behavioral variables at the .0435 level of chance probability. Based on these results, age was the covariate in the multivariate analysis of covariance testing for group differences between short and long-term father absence. The results of these analyses are presented in Table 17. 66 Table 17 RESULTS OF TWO-WAY MANCOVA ON SOCIAL BEHAVIORS OF LONG AND SHORT-TERM FATHER ABSENT CHILDREN Factors F Ratio Degrees of Freedom. Probability Main Effects Length of Absence 1.5811 11 & 65 .1256 Sex 3.3602 11 & 65 .0005* Interaction .2109 11 & 65 .9963 *pf.05 There was no interaction effects of sex and length of absence. Main effects of sex were significant at the .0005 level of chance probability and there were no effects of length of absence. Self-concept, activity level, initiation and heterogeneity of initiation accounted for significant sex differences as reported in Table 18. The adjusted cell means for those variables contri- buting to significant sex differences are presented in Table 19. These results indicate that girls were more passive in their activity levels, than boys, and made fewer initiations than boys. In their expressed self concept, girls were less positive than boys and initiated more to the opposite sex. 67 Table 18 VARIABLES CONTRIBUTING TO MAIN EFFECTS OF SEX ON LONG AND SHORT-TERM FATHER.ABSENT CHILDREN Variable Univariate Degrees of Freedom Probability F Ratio Self concept 4.5750 11 & 65 .0357* Adult dependency .8449 ll & 65 .3610 Autonomy 3.7552 11 & 65 .0565 Social Behavior 3.7004 11 & 65 .0582 Involvement 1.6012 11 & 65 .2097 Activity Level 5.6312 11 & 65 .0203* Initiation 4.4732 11 & 65 .0378* Heterogeneity of Initiation 8.1280 11 & 65 .0057* Response .1616 ll & 65 .6888 Tolerance .7767 ll & 65 .3810 Aggression 2.7947 11 & 65 .0988 *p§-.05 Table 19 ADJUSTED CELL MEANS OF VARIABLES CONTRIBUTING TO SEX EFFECTS ON LONG AND SHORT-TERMIFATHER.ABSENT CHILDREN Variable Boys Girls Total Self-concept 8.3947 7.6379 7.9879 Activity Level 2.1124 2.2295 2.1753 Initiation -.7400 -l.2991 -1.0405 Heterogeneity of Initiation .7233 1.4526 1.1153 68 Based on these findings the null hypothesis of 11, 12, 13, and 14 of no differences in the social behaviors of long and short-term father abent children cannot be rejected. D. DO FATHER SURROGATES AFFECT THE DEVELOPMENT OF SOCIAL INVOLVEMENT BEHAVIORS? H15: Father absent children with a father surrogate will have higher peer associative and coopera- tive behavior scores than father absent children with a father surrogate. H15: There are no differences between children with a father surrogate and children with- out a father surrogate on peer associative and cooperative behaviors. H16: Father absent children with a father surrogate will have lower initiation and response scores than father absent children with a father surrogate. H016: There are no differences between children with a father surrogate and children with- out a father surrogate on initiation and response scores. Hypotheses 15 & 16 investigated the effect of older males as father surrogates in the father absent family. To determine the association of age with the dependent variables, multiple regression analysis was performed. Age was associated with the dependent variables at the .0051 level of chance probability. Proceeding with multivariate analysis of co- variance, age was the covariate. Group differences were tested between father absent children with a father surrogate and father absent children without a father surrogate in the home. 69 There were no interaction effects of sex and surrogates, and no main effects for surrogates. There were effects of sex at the .004 level of chance probability as reported in Table 20. Table 20 RESULTS OF TWO-WAY MANCOVA ON SOCIAL INVOLVEMENT BEHAVIORS OF SURROGATE AND NON-SURROGATE FATHER ABSENT CHILDREN Factors F Ratio Degrees of Freedom Probability Main Effects Sex 4.3033 8 & 68 .0004* Surrogate .9479 8 & 68 .4838 Interaction 1.3860 8 & 68 .2185 *p5.05 The social behaviors of activity level, initiation, and heterogeneity of initiation accounted for significant sex differences among father absent children as reported in Table 21. Based on the adjusted cell means presented in Table 22, girls' activity levels were passive as compared to boys, and fewer initiations were made by girls than boys. In hetero- geneity of initiation, girls, initiated more to the opposite sex than boys initiated to the opposite sex. 70 Table 21 VARIABLES CONTRIBUTING TO MAIN EFFECTS OF SEX 0N SURROGATE AND NON-SURROGATE FATHER.ABSENT CHILDREN Variable Univariate F Degrees of Freedom. Probability Ratio Involvement 1.6345 8 & 68 .2065 Social Behavior 3.4585 8 & 68 .0669 Activity Level 5.5144 8 & 68 .0215* Initiation 4.5792 8 & 68 .0357* Heterogeneity of Initiation 8.1727 8 6 68 .0056* Response .1770 8 & 68 .6752 Tolerance .7649 8 & 68 .3840 Aggression 2.8080 8 & 68 .0980 * p5.05 Table 22 ADJUSTED CELL MEANS OF VARIABLES CONTRIBUTING TO SURROGATE AND NON-SURROGATE FATHER.ABSENT MAIN EFFECTS OF SEX + Variable Boys Girls Total Activity Level 2.1128 2.2312 2.1764 Initiation - .7410 -1.3063 -l.0448 Heterogeneity of Initiation .7180 1.4459 1,1093 71 The null hypothesis 15 and 16 of no differences between father absent children with a surrogate and father absent children without a father surrogate on involvement behaviors cannot be rejected. E. DOES A FATHER SURROGATE AFFECT THE DEVELOPMENT OF AFFECTIVE SOCIAL BEHAVIORS? H17: Father absent children with a father surrogate will have lower adult dependency scores than father absent children without a father surrogate. H017: There are no differences between children with a father surrogate and children without a father surrogate on adult dependency scores. H18: Father absent children with a father surrogate will have higher self concept scores than father absent children without a father surrogate. H018: There are no differences between children with a father surrogate and children without a father surrogate on self-concept scores. Hypotheses l7 and 18 focused on the effects of older males as father surrogates in the father absent home and the development of affective social behaviors. To determine the association of age with the dependent variables of interest, multiple regression analysis was performed. Age was associated with the dependent variables at the .0284 level of chance prob- ability. Proceeding with multivariate analysis of covariance ‘with age as the covariate, there were no interaction effects, and no surrogate effects, although sex effects approached significance as reported in Table 23. 72 Table 23 RESULTS OF TWO-WAY MANCOVA ON AFFECTIVE SOCIAL BEHAVIORS OF SURROGATE AND NON-SURROGATE FATHER ABSENT CHILDREN Factors F Ratio Degrees of Freedom. Probability Main Effects Sex 2.5961 3 & 73 .0589 Surrogate .4068 3 & 73 .7486 Interaction .3962 3 6 73 .7562 Based on these results the null hypotheses of 17 & 18 of no differences between father absent children with a father surrogate and father absent children without a father surrogate and their affective social behaviors, the null hypothesis cannot be rejected. 73 RESULTS Section 2 In this section the results of the descriptive analysis of the demographic data are presented. The following dimensions were investigated: (1) socioeconomic status, (2) ethnicity, (3) child's day care experience, (4) center size, (5) child's ordinal position, (6) female adults in the home, (7) playmates at home, (8) age of mothers and fathers, (9) education of mothers and fathers, (10) hours working, (11) occupation, (12) income by center report, (13) income by parent report, (14) dwelling, (15) type of transportation to center, and (16) time in transport- ation to center. It should be noted that some families did not respond to all parts of the Parent Information Sheet. Therefore table totals are not consistent in all cases and are noted. Socioeconomic Status (SES) Occupation, education, and source of income of the primary wage earner which are included in the short form of the McGuire & White (1955) instrument were selected as demographic variables. It was deemed appropriate to examine the individual components of SES rather than an SES score, since SES was not an independent variable in this study. Also a more comprehensive examination ‘may be made of the factors contributing to family life of the father absent child. 74 Ethnic Background Although ethnicity was not a design variable, it is an important element in describing the sample. For the purposes of this study, a child was considered Black if either or both natural parents were Negro. He was considered Anglo if both natural parents were Caucasian. As illustrated in Table 24, 72% of the sample were Anglo, and 26% Black and less than 2% Mexican or Indian. There were 63% Anglo father absent children and 63% Black father absent children. The distribution of centers across ethnic group represents the best balance in enrollment of those centers available and willing to participate. One center had a 90% enrollment of Black children, three centers were 90% or more Anglo and the other four< centers enrolled mixed populations of between 60-80% Anglo, and 20-40% Black. Table 24 Family Status by Ethnicity Ethnicity Black Caucasian Other Father Absent 21 57 2 Father Present 12 35 0 TOTAL 33 92 2 75 Day Care Experience During the school year (1972-1973) prior to the data collection (1973-1974) 83% of the father absent and 73% of the father present children had been in day care 10-12 months. At the time of data collection 90% of the father absent and 80% of the father present children were in day care situation 10-12 half day equivalents a week (1 half day equiva- lent - 4 hrs.). Therefore, during the year prior to and during the year of data collection a higher percentage of father absent children were in day care. Center Size The licensed daily capacity of the centers ranged from 47 to 120 with an average of 87. The actual enrollments ranged from 70 to 166 with an average of 118. There was a relatively even distribution of the sample between franchised (51%) and non-franchised (49%) centers. Non- franchised centers were non-profit public and privately administered day care centers. Franchised centers were privately owned and administered facilities. All centers were located through the central part of lower Michigan. Two centers were in Flint, three centers in Grand Rapids, two centers in Battle Creek, and one center in Lansing, Michigan. Father present and father absent sample totals by center are presented in Table 25. 76 Table 25 Center by Father Status Center Type Franchised .Non-Franchised Center 1 2 3 4 5 6 7 8 Father Absent 9 10 l 18 5 12 9 16 Father Present 4 6 9 8 4 6 2 8 TOTAL 13 16 10 26 9 18 11 24 N - 127 Ordinal Position The frequency of ordinal positions for father absent and father present children is presented in Table 26. Father absent children more often had siblings while father present children were more often only children. Half (48%) of the father absent and three-fourths (78%) of the father present children were only children. Table 26 Ordinal Position by Father Status Ordinal Position 1 2 3 4 5 6 7 8 9 Father Absent 38 26 7 4 3 0 0 2 0 O O O O O H Father Present 37 6 3 TOTAL 75 32 10 4 3 0 0 2 l N - 127 77 Adult Females in Home In 92% of the father absent and 95% of the father present families no other female adult was living in the home with the mother. A third (32%) of the father absent boys and one fourth (20%) of the father present boys had older sisters. About one fifth (23%) of the father absent girls and father present girls (18%) had older sisters. Playmates at Home A majority of the children (76%) played with older play- mates or agemates at home as presented in Tables 27 and 28. This was true of a majority (72%) of the father absent as well as father present children, (76%). Table 27 Age of Playmates at Home by Father Status Age Older Younger Same age Mixed ages Father Absent 29 3 25 14 Father Present 23 3 l4 5 N - 1163- 3 Data were not reported by eleven families. 78 Table 28 Relationship of Playmates at Home by Father Status Father Status Father Absent Father Present Siblings l4 3 Relatives 8 3 Friends & Neighbors 29 21 Siblings & Neighbors 3 1 Siblings & Friends 13 9 Relatives & Friends 7 5 N - 1164 Ago of Fathers and Mothers Almost all (95%) of the present fathers were younger than 40 years, with an even number in both the 20-29 year old range and 30-39 year old range. As presented in Table 29, two-thirds (66%) of the father absent mothers were between 20 and 29 years as were over two- thirds (69%) of the father present mothers. 4Data‘were not reported by eleven families. 79 Table 29 Age of Fathers and Mothers by Father Status Age 19 or 20-29 30-39 40-49 50 & over younger Father Absent Fathers 0 5 2 0 2 Mothers 2 53 21 4 . 0 Father Present Fathers 0 22 21 l l Mothers 1 32 12 1 0 N = 1255 Education of Parents Education as reported in the Parent Information Sheets is shown in Table 30. The educational attainment of those absent fathers reported is evenly distributed between less than high school and college, while three-fourths of those fathers present had been educated in college and/or held a degree. Very few father absent mothers (10%) had not completed high school and over half (56%) had gone to college and/or held a degree. Similarly, very few father present mothers (4%) had not completed high school but over two-thirds (70%) had gone to 5Data‘were not reported by one family. 80 college or completed a degree. Table 30 Education of Parents by Father Status Father Status . Father Absent Father Present Mother Father Mother Father 12 years 5 2 2 1 12 & training 3 0 0 1 High school 16 2 5 3 High school & occupational training 11 0 7 6 Less than 4 years of college 32 2 ll 11 4 years of college 12 0 16 13 Advanced degree 1 l 5 10 N - 1266 Parents Working Hours A high percentage (86%) of the father absent mothers and father present mothers (91%) worked 30 to over 40 hours a week outside the home, as presented in Table 31. The majority of father absent and father present mothers worked over 40 hours a week. Almost all (93%) of the present fathers worked 40 or more hours as did six absent fathers. 6Datawere not reported by one family. 81 Table 31 Parents Werking Hours by Father Status Hours of work 0 1-20 21-29 30-39 40 & over Father Absent Mothers 6 2 3 14 52 Fathers 0 O 0 l 5 Father Present Mbthers 2 O 2 10 31 Fathers 1 0 l 1 41 N - 1227 Occupation of Parents As presented in Table 32, a majority of present fathers (62%) held managerial positions that required supervision or added responsibility, and of these few were professionals (13%). Over three-fourths (78%) of the father absent mothers and less than half (46%) of the father present mothers held jobs requiring semi-skilled, clerical, or sales activities. Over half (54%) of the father present mothers and few (22%) of the father absent mothers held jobs requiring added responsibilities of management or professional services. 7Data‘were not reported by 5 families. 82 Table 32 Occupation of Parents by Father Status Father Status Father Absent Father Present Semi-skilled Clerical & Service Sales & Foreman Managers Nurses & Teachers Mother Father 13 38 10 4 12 Professionals - Doctors, 1 Lawyers, etc. N - 1248 2 4 l 0 1 0 Mother Father 2 4 l4 7 5 6 5 6 18 16 2 6 Income Distribution--Center Report Families receiving public financial assistance for day care, Aid to Dependent Children (ADC), as reported by the center director are illustrated in Table 33. Famdlies receiving other aid in the form of reduced fees is reflected in the partial fees (PF) figures. Those who paid their own fees are listed as Private. 8 Data were not reported by three families. 83 Table 33 Income Distribution by Father Status--Center Report Source of Income ADC. PF Private Father Absent 45 20 15 Father Present 2 13 32 N - 1269 Income Distribution--Parent Report Income distribution as reported by the parent is pre- sented in Table 34. Inconsistency between the center's report and the Parent Information Sheet regarding ADC should be noted. Centers reported 45 father absent and 2 father present families receiving ADC while only father absent mother reported.ADC assistance. Father present families reported a much higher percentage (71%) of weekly income above $200 than father absent families (7%) and father absent families reported a higher percentage (62%) of weekly incomes below $125 than father present families (2%). 9Data‘were not reported by one family. 84 Table 34 Income Distribution by Father Status--Parent Report Weekly Income ADC $50- $76- $101- $126- $151- $176- $201+ . 75 100 . 125 . .150 ..175 200 Father Absent 1 12 20 14 8 7 9 5 Father Present 0 0 0 1 3 3 6 33 N = 12210 Dwellings Half (52%) of the father absent children lived in a single family dwelling while the other half (47%) lived in multiple family dwellings as presented in Table 35. Almost all (87%) of the father present children lived in single family dwellings while few (13%) lived in multiple family dwellings. Table 35 Family Dwellings by Father Status Dwelling Single Apartment Duplex Trailer Father Absent 39 28 8 1 Father Present 39 2 4 l N - 12211 10,11 Data were not reported by five families. 85 Transportation to Center All of the father present and almost all of the father absent families used their own car for transportation to the center as presented in Table 36. A small percentage (14%) of the father absent families used transportation other than a private car. None of the families were within walking distance of the centers. Table 36 Transportation to Center by Father Status Transportation walk Car Public Center Bus Friends Father Absent 0 69 2 8 1 Father Present 0 47 0 O 0 N - 127 Time in Transportation to Center There were small differences in the numbers of father absent children and father present children requiring more than one-half hour travel time to the center, as presented in Table 37. Most children arrived at their centers within half an hour from home. 86 Table 37 Time in Transportation to Center by Father Status ...... Minutes V , 5-15. p.20-30,, 35-45...50'60. Total N Father Absent S9 17 0 2 78 Father Present 38 7 l 0 46 N = 12412 Summary of the Family Demographic Data 1. ETHNIC BACKGROUND: The proportions of father absent and father present children among Caucasian families were very similar to the proportions of father absent and father present children among Black families. 2. CHILD'S DAY CARE EXPERIENCE: Father absent children had more day care experience than father present children during the year of data collection and the year prior. 3. CENTER SIZE: The average center size was 87 with all centers ranging from 47 to 120 children. There was an even distribution of father absent children between franchised and non-franchised centers. 4. CHILD'S ORDINAL POSITION: Most father absent children had siblings, while most father present children were only children. 12Data‘were not reported by three families. 10. 11. 87 ADULT FEMALES IN THE HOME: A third of the father absent boys and one-fifth of the father absent girls had an older sister. About one-fifth of the father present children had an older sister. Most father absent boys had an older sister. PLAYMATES AT HOME: Playmates at home varied greatly. MOst children played with friends and neighbors rather than sib- lings. AGE OF PLAYMATES AT HOME: The vast majority of father absent and father present children played with older or age- mate children at home. AGE OF FATHER.AND MOTHER: Two thirds of both father absent and father present mothers were between the ages of 20 and 29 years. Almost all of the present fathers were between the ages of 20 and 40 years. EDUCATION OF PARENTS: Very few father absent mothers had not completed high school and over half had gone to college or held a degree. Most fathers in father present families had attended college or held a degree. Very few fathers present mothers had not completed high school and over two-thirds had gone to college or held a degree. PARENTS WORKING HOURS: Almost all of the mothers regardless of family status, worked 30-40 hours a week outside the home and a majority of the mothers worked more than 40 hours a week. OCCUPATION OF PARENTS: Most of the father absent mothers held jobs requiring semi-skilled activities, while over half of the father present mothers held jobs requiring added 12. 13. 14. 15. 88 responsibility. The majority of present fathers held jobs requiring supervision skills or added responsibility. Few fathers were professionals. INCOME DISTRIBUTION: Over half of the father absent families received Aid to Dependent Children (ADC) in the form of day care payment assistance. Only two father present families reported incomes over $200 per week while two-thirds of the father absent families reported weekly incomes of $125 and below. DWELLINGS: Half of the father absent children lived in multiple family dwellings while most all of the father pre- sent children lived in single family dwellings. TRANSPORTATION TO CENTER: All of the father present and al- most all of the father absent families used their own car for transportation to the center. None of the families were 'within walking distance to their centers. TIME IN TRANSPORTATION TO CENTER: The center was within 30 minutes of most children's homes. CHAPTER V DISCUSSION The discussion will be presented in two sections. Section 1 will involve the discussion of the main independent variables, including father absence, father presence, length of father absence, presence of a father surrogate, sex, and age. Section 2 of the discussion will focus on the family demographic dimensions. These dimensions are ethnicity, child's day care experience, center size, child's ordinal position, females in the home, playmates at home, age of mothers and fathers, education of mothers and fathers, hours working, occupa- tion, income by center report, income by parent report, dwelling, transportation, and time in transportation to center. 'Section 1 The results of this study indicate that no significant differences existed between father absent and father present children with regard to the variables of interest. Also, no -significant differences were found with relation to the variables of long-term and short-term, father absent children, or between those father absent children with a father surrogate and those without a father surrogate. 89 90 Previous father absence research (Robinson, 1946; Sears, Piltner & Sears, 1946; Sears, 1951; Tiller, 1958; Lynn & Sawrey, 1959; Santrock, 1970) has focused on aggression dependency, and self concept. Definitions of these behaviors are in contrast to those made in this study. For instance, self concept in previous studies has been defined as masculine and feminine while in this study self concept was defined as either positive or negative. Aggression and dependency were previously operationalized through projective doll play and maternal interviews. Aggression (Santrock, 1951), as an example, was operationalized in doll play as an act by the sub- ject using the same sex doll to hit the opposite sex doll. In the present study, aggression was operationalized and carefully quantified as verbal (tone of voice) or specific physical be- havior such as hitting or pushing as observed in a play setting with other children. The present investigator strongly suggests that peer interaction observed in play groups is more indicative of the child's true behavior than the projected behavior in a doll play episode. With the exception of one (Robinson, 1946) all of the studies mentioned above used maternal interviews in finding father absent effects. Eron (1961) concluded that in an inter- view situation the father is more accurate than the mother in describing the actual behavior of their child. With this methodological consideration in mind those studies utilizing maternal interviews as sources of finding behavioral differences 91 in father absent children, should be questioned. In contrast to previous research, the basic quantity of a wide range of preschool behaviors have been observed and carefully quantified in this study. This observational methodology is an important departure from.most previous research in this area, and the lack of father absent effects have particular import in light of this. The developmental age of the children in this study cannot be overlooked as a factor in the findings of no differences. At the preschool age it is possible that differences are not observable due to the significant amount of growth and develop- ment of social (Parten, 1932) as well as physical (Gesell, et a1., 1940) skills. The impact of father absence may be secondary to the impact of other factors affecting the development and sub- sequent behaviors of the preschooler. Biller (1967) has suggested that during this period of sex role adoption, ages three to five, father absent differences could be measured through observer ratings of the child's behavior. In the practical sense this is possible at the pre- school age since sex role adoption relates to the individual's ability to imitate sex appropriate behaviors. It is also generally accepted that many observable social skills are developing during this age period (Parten, 1933). It would seem.that the observational methodology used in this study for 'measuring differences in social behaviors as a function of father absence is not only practical but also conceptually sound. 92 It should be emphasized that the absence of the father may be an insignificant determinant of the preschoolers social behavior. This would account for no observable differences in father absent and father present social behaviors in this study. However, the absence of observed variations in behavior does not necessarily rule out the possibility that differences developmentally may be occurring at the preschool years which could affect behavior at a later developmental period. This con- clusion is supported elsewhere, (Biller, 1970). Studies of the father absent adolescent have noted the inability of father absent girls to interact appropriately with males in social situations (Hetherington, 1972) and that father absent boys have difficulty in establishing and sustaining peer relationships (Miller, 1961; Mitchell & Wilson, 1967). Con- sidering this evidence and the widespread assumption of many scientists, caseworkers, psychologists, and the general public that juvenile delinquency and extramarital pregnancy are characteristics associated with father absent children (Herzog & Sudia, 1973) it is suggested here that father absence does have an impact upon the child but that it may not be an observable effect during the preschool years. All long-term father absent children in this study did not have a father for at least the first three years of their life. The fact that no differences were found between long and short-term father absent children has other interesting implica- tions in addition to the notion that the father's influence does not translate into observable social behavior is measured in 93 this study. One implication is that the mother may be the most important influence on the child's development of social be- haviors during the first three years of life. In support of this, Goldfarb's work (1943) on the develop- ment of attachment strongly suggests that if an infant is kept in a depriving institution up to the age of three and one-half years, a transfer to a foster home causes extreme difficulty in the infant's ability to form an attachment to his foster mother. Provence and Lipton (1962) observed infants eighteen to twenty-four months old experiencing difficulty forming attachment behaviors. If the early years are uniquely important for maternal attachment purposes, does the same hold true for the father and are there paternal attachment behaviors? If there are attachment behaviors related to the father's interaction during the first three years of life and if differ- ences in behavior cannot be determined during the preschool years, one might then expect later childhood to be the stage in which these differences in behavior become most apparent and observ- able. The previously discussed literature would support this. The attachment process is a most complex one. It is surprising that in a recent review of literature on attachment behavior by Ainsworth (Herzog & Sudia, 1973) few studies investi- gated the father's role and implications drawn did not include the father as a possible intervening factor in the development of attachment behaviors. 94 The results of this study indicate that there were no differences between long-term.and short-term father absent preschoolers. The report by the parent(s) on length of absence was used in determining long-term and short-term father absence. Understandably the parent may not have been totally candid, for personal reasons, in giving the actual length of absence. Fear of possibly losing ADC support by reporting the presence of the father or not wishing to reveal family circumstances in a written form may have been a confounding circumstance in the determination of the father's short or long-term absence. Methodological differences aside, one would expect differences in behaviors of long and short-term father absent preschool children based upon a review of previous research (Biller & Bahm, 1971: Stolz et al., 1954; Leichty, 1960). Methodological shortcomings and inadequate controls, however, (Biller, 1970) also raise questions concerning the validity of those findings. The independent variable of father surrogate was not found to affect the social involvement or affective dimensions of preschoolers behavior. This is not a surprising finding considering the previous discussed findings of no father present effect. The fact that there were no father surrogate effects, however, further supports the conclusion that the presence of absence of the father may not appreciably impact on the develop- ing social behaviors of the preschooler. The mother's presence as a compensatory agent in encour- aging masculine behaviors especially in boys has been suggested 95 as a source of change for the father absent environment (Biller, 1967). For girls the mother might also encourage interaction with male adults such as boy friends, relatives, teachers, or other men they both come in contact with during errands and other activities. Encouragement of interaction with males would also be accompanied by a positive attitude towards males as expressed by the mothers' comments to other females, to her own children, and to herself as overheard by the child. Television or movies viewed with the mother or by the child alone might also be a source of positive feelings about male adults. Therefore, a combination of, providing opportunities for the child to inter- act with adult males, the child watching positive models of male-female interaction, the child imitating male-female interactions, and the mother reinforcing her child's adult male interactions, could be the basis for compensating for the loss of the father. Section 2 Family Demographic Characteristics Family demographic data in twenty categories were collected through a Parent Information Sheet prior to the obser- vations of social behavior. Center directors contacted parents who had not returned the Information Sheets and collected a total of 180 of the 210 administered. Of the 180 families responding, 127 participated in the research. The 53 families returning information but not included in the sample were 96 eliminated due to the child's sporadic attendance at school which did not permit the collection of observational data. The following generalization could be made based on family demographic data including source of income, income level, age, and type of occupation. The majority of father absent children come from.families on.ADC, with a weekly income of $125 or less; having a mother between the agos of 20 and 29, holding a semi-skilled job requiripg little responsibility. This generalization could envoke the image of an un- employed mother sitting around the house all day while she collects welfare. However, if one considers educational level, hours working per week and type of welfare assistance, the following generalizations could also be made. Slightly fewer father absent mothers have not completed high school, attended college or hold a degree than father present mothers. Also, almost all father absent mothers received day care assistance while working nearly the same number of hours (30-40) a week as father present mothers! Father absent mothers were not that different from father present mothers when considering the hours they work, attained educational level, and assistance available. Father absent mothers, with their larger families (they often had two or more children.while father present mothers had one child) and depressed income level, would be expected to take advantage of ADC while working 30 to 40 and over hours a week. The present investigator suggests that the reason father absent mothers do not have a job of responsibility is not 97 because of their incompetence but their inability to accept added responsibility in view of their family's functioning. Since educational levels of father absent and father present mothers were similar, skills in responsibility would not seam to be the problem, Rather, overtime working hours and the added physical and psychological energy required to cOpe with job related responsibilities would seem.to be most pertinent factors. Jobs of responsibility do involve risk in task comp pletion as compared to semi-skilled positions. The chance of losing a job of responsibility may be greater than the chance of losing a semi-skilled job by a mother who is over qualified for the position. The risks involved in a high responsibility job may not be as readily accepted by the father absent mother, for if she fails, all income may be lost. But if the father present mother fails, all family income is not lost. The author poses the following question relevant to the father absent mother's family position. If you were a person between the ages of 20 and 29, responsible for two or more child- ren, with no mate to share your problems and family responsibil- ities, making $125 or less a week and working 30 to 40 plus hours weekly, would you want additional responsibility? The question illustrates the dilemma of the father absent mothers in this study. In many respects the father absent families in this study were not as different from the father present families as one might expect. In addition to no differences in the social 98 behaviors of their children there is additional information to suggest few differences between these two groups. There was an even distribution of father absent and father present families among racial groups (Black and Caucasian). There- fore, ethnicity would not seem.to be a contributing factor to father absence. Almost as many father absent families as father present families had their own transportation to the center. All families lived within a 30 minute ride of the center. Father absent children (although they often had sib- lings) played with older or agemate friends and neighbors as did the father present children. Some differences between father absent families and father present families were found. Father absent children often had siblings while father present children were frequently only children. Half of the father absent children lived in multiple family dwellings while almost all of the father pre- sent children lived in single family dwellings. CHAPTER VI SUMMARY, CONCLUSIONS, AND IMPLICATIONS Summary The primary purpose of this study was to investigate group differences in the affective and social involvement be- haviors of father absent and father present preschool children. The secondary purpose was to gain further descriptive informa- tion concerning the father absent family. One-hundred and twenty-seven children were randomly selected from.sight day care centers in four large cities in lower Michigan. Their ages ranged from.3% to 5 years and were enrolled for at least four half-days per week. There was an even distribution of boys and girls. Family demographic data ‘were gathered through a Parent Information Sheet sent home and returned to the center. Descriptive information concerning the family included: (a) age, ordinal position, and playmates of the child, (b) edu- cation, age, occupation, and number of working hours of the mother and/or father, (c) length of father presence or absence, number and relationship of father surrogates, and (d) income level, ethnicity, type of dwelling, and day care experience of the child. Father absence or presence, length of father absence, and the availability of a father surrogate were the dimensions 99 100 of interest in the observation of social behaviors. Utilizing video assisted observational methodology and small group con- trived play sessions, social behaviors including initiation, aggression, cooperation, activity level, and expressed self concept were observed. The results of this study indicate that father absence does not make a significant impact on social behaviors in pre- school boys and girls. Differences in social behaviors were not found between children with a father and children without a father. Length of father absence regardless of whether it was short (six months to two years) or long (three to five years) was of little consequence in affecting observed social behaviors. The presence of a father surrogate was not an intervening factor in the expression of social behaviors by father absent children and is consistent with the findings of no differences between father absent and father present children. Conclusions and Implications This study has led to the following limited conclusions and implications: (1) The father has limited impact on the preschool child's expression of a wide range of social behaviors. That no differences were found between father absent and father present children at an age which many theorists regard as criti- cal in terms of imitating father behaviors, places these results in a conflicting position with established theoretical models. 101 Either the father is not a critical model for the child's social behaviors or his influence is not apparent until later childhood. The former would certainly be in contrast to most previous research (Lynn, 1974; Biller, 1971) and theo- retical positions (Parsons & Bales, 1955; Lynn, 1959; Bandura & Walters, 1963) while the latter would be generally supported (Hetherington, 1972; Biller 1971; Lynn 1974). To suggest that at the preschool age the father is not crucial in the develop- ment of the child's ability to exhibit certain social behaviors would be a more accurate statement, however, one cannot reject completely the notion that the father is not as influential at this age as has been previously suggested (Biller, 1971; Lynn, 1974; Hetherington & Deur, 1971). The findings in this study could be in support of the life style and family makeup of the father absent family. Although a father absent family is limited in many ways the positive aspects especially the apparent compensatory behaviors of the mother, should also be recognized by educators who seek to provide information to parents living in a father absent family environment. (2) The presence or absence of a father surrogate also does not seem to impact upon the observable social behaviors of the preschool child. No differences were found between children with a father surrogate and children without a father surrogate. As a model, the father surrogate would also be expected to influence the behaviors of the child. This, however, 102 was not the case and is added support for the conclusion that fathers or other males do not influence overt child behavior at this age. This conclusion is also in a difficult position con- sidering the literature pertaining to the influence of the surrogate father on the development of masculine behaviors (Santrock, 1970; Wohlford, Santrock, Berber & Liberman, 1970) which would support the father surrogate as an important model, especially for boys. The surrogate father may play a more important suppor- tive role to the mother than to the young child as a model of masculine behaviors. Although no evidence was gathered in this study to confirm or deny this suggestion, one cannot disregard the psychological, physical and economic support of a second person. Even with the contradictory evidence concerning the in- fluence of the surrogate father, parent education programs should contain a guarded amount of information concerning the role of the surrogate father and the contributions he may make. (3) Length of father absence does not seem to be a differentiating factor in the expressed social behaviors of preschool children. While other investigations (Hetherington, & Deur, 1971; Carlsmith, 1964) have concluded the longer the separation the more pronounced the effects, especially in boys (Nash, 1965; Seplin, 1952) the results of this study do not support those conclusions. 103 The contradictory nature of the previous findings presents a problem in drawing implications although it would seem that any length of father absence is too long. The implication should not be that length of absence does not make a difference but more accurately a wide range of intervening factors, not measured in this study, may have accounted for no differences. (4) A unique aspect in this study of young father absent children is the methodological approach in which large and small group preschool peer interactions were carefully observed and quantified. This is a significant departure from previous preschool father absent research which has centered on projective doll play methodology, historical accounts and maternal interviews concerning the child's behavior. In this study peer social behaviors were observed in a naturalistic as well as contrived play setting and as such provides a wider continuum.upon which father absence effects may be studied. That no differences were found between father absent and father present children in view of this recent methodology raise questions as to whether or not projective, historical, and parental interview instruments are conceptually suited for ‘measuring differences in behaviors of father absent children. It is suggested that because observational methodology captures a wider range of peer interaction behavior, the ability of the investigator to describe and predict behaviors is greatly in- creased over other measurement models. 104 Limitations and Spggostions for Future Research 1. A lack of detailed information concerning the mother's behaviors and attitude towards father absence was a limitation in the interpretation of the father absent mothers' compen- satory behaviors or lack of compensatory behavior. Attitudinal data from.an interview could be verified through observation of behavior in a mother-child interaction play session or through direct observations in the home. 2. A similar limitation involved the lack of specific information about the compensatory behaviors and attitudes of significant others such as day care teachers and staff. Direct observations of teacher-child interactions, activities in the day care program, number of male staff, and father status of the group of children with which the father absent child plays, would have been useful in the interpretation of the results. 3. The reasons for father absence were not determined. This presented another limitation in the interpretation of the results. General categories that could explain the father's absence such as death, separation, employment, and military could be included in an expansion of the Parent Information Sheet. 4. A longitudinal study would be of particular strength in investigating the question of whether developmental differences during father absence affects father absent child behavior. If the sample in this study were followed into grade 105 school, achievement and success in school could be investigated as affected by father absence. Also, social behaviors of adolescent boys and girls could be compared with their social behaviors during preschool. The long-term effect on the developing social behavior could then be assessed. 5. The child's expressed self concept in reference to his father was not a part of the data on self concept. A methodological revision of the Brown Self Referent Self-Concept instrument would be most useful in obtaining data concerning the impact of the father's absence or presence. Presently the Brown utilizes mother, teacher, peers, and self as referents. If the father's effect is to be studied, his role as a referent should be included in the instrument. 6. Perhaps the most important contribution to further study this research makes is that of providing supportive evi- dence that father absent mothers are capable of being a positive influence during their child's preschool years. The results of this research support the contention that undue concern for the well-being of the young child in families where the mother is the sole adult member, may be inappropriate. 7. A.limitation inherent in the measurement of social behaviors was that one ten minute observation was made for each child. Observations over a period of weeks or months in varied settings would be preferred. The cost in time and money of gaining and rating these observations, however, necessitated a more lhmited approach. 106 8. Finally, the fact that more questions have been generated than answered in this study points to the need for increased research on the father absent family. New methodological approaches are needed that are tied to direct behaviors and which have longitudinal potential to determine the existence or absence of delayed effects. A careful methodological approach is necessary to include relevant factors such as reason for father absence, length of father absence, role of the mother as a compensatory factor and the role of significant others as compensatory agents. APPENDIX A 107 108 Classroom Socio-Observation The classroom.socio-observation technique was developed to assess the social involvement and play activity of children in the c1assroom.setting. It was developed by Jo Lynn Cunningham and Tito Reyes, Family and Child Study Center, Michigan State University.1’2 The present procedures are an adaptation of the original instrument.3 General Procedures The children will be grouped at the time of the obser- vation in order to establish balanced groups of 12 children that include: 3 Low SES Boys, 3 Mid SES Boys, 3 Low SES girls, and 3 Mid SES Girls. Additional groups of 12 children will be formed until all the children in the sample are observed. Children may be included in more than one group in order to establish balanced groupings. Three (3) consecutive observations (one set) are made near the beginning of the free play period and another set of three (3) observations are made toward the end of the period. Approximately 10 minutes should lapse between sets of observa- tions. 1Cunningham, J.L., and Reyes, T.F. The sociometry of preschool children. Unpublished paper, Michigan State Univer- sity, 1969. 2Special thanks are given to Kristin Anderson for her help with the preliminary testing of this technique. 3The present adapted version was developed by Mary Andrews, Institute for Family and Child Study, Michigan State University, 1973. , 109 The setting for the observations will be a classroom that includes a variety of activities for free play (i.e. blocks, house corner, manipulative toys, etc.). This setting should be familiar to all of the children. One (1) teacher will be present to supervise the children during the observation. Her interaction with the children should be minimal. Name tags or a number or letter code should be placed on each child (taped or pinned) prior to the observations. Such tags will aid the examiner in identifying the children. Form The form.used for recording observations is a drawing of the floor plan of the preschool classroom with major play areas indicated. It is suggested that a list of all children in the class with their identifying code letters be attached. Recording,Observations For each observation, a systematic recording is made of the play location and involvement of each child. Start at one end of the room.and record each individual as quickly as possible. Each child must be recorded once and only once. There- fore, if a child moves to another group after an observation is recorded of his activity, he is not recorded again, even though the other children in the new group are recorded if they have not been previously observed. As soon as the entire class has been recorded and checked, proceed with the second and then third in the set of 110 three consecutive observations. Codes The recording of each item is as follows: % Major activity areas are indicated on the observation form. INDIVIDUAL A...N - Subjects (unique identifying letters are assigned to each child) X = Teacher Y = Other adult PLAY INVOLVEMENT 1 - Unoccupied behavior: The child apparently is not playin at all, at least not in the usual sense, but occupies himseIf ‘with watching anything which happens to be of momentary interest. When there is nothing exciting taking place, he plays with his own body, gets on and off chairs, just stands around, follows the teacher, or sits in one spot glancing around the room. 2 - Solitary Play: The child plays alone and independently with toys that are different from those used by the children with- in speaking distance and makes no effort to get close to or speak to the other children. His interest is centered upon his own activity, and he pursues it without reference to what others are doing. 3 = Onlooker Behavior: The child spends most of his time watch- ing the others pIay. He often talks to the playing child- ren, asks questions, or gives suggestions, but does not enter into the play himself. He stands or sits within speaking distance of the group so he can see and hear all that is taking place. Thus, he differs from.the unoccupied child, who notices anything that happens to be exciting and is not especially interested in groups of children. 4.x II Parallel Play: The child plays independently, but the activity he chooses natura 1y brings him.among other child- ren. He plays with toys which are glike those which the children around him.are using, but he plays with toys as he sees fit, without trying to influence the activity of the 111 children near him, Thus, he plays beside, rather than with, other children. This activity is characterized by physical proximity god similarity of activity with refer- ence to other children. Associative Play: The child plays with other children. They may be BOrrowing and lending play material or follow- ing one another with trains and wagons. There are mild attempts to control which children may or may not play in the group. All en age in similar, if not identical, activity. Each ch 1d acts as he wishes and does not sub- ordinate his interest to the group. There is interaction between children, but no common goal. Cooperative Play: The child plays within a group that is organized for the purpose of making some material project, of striving to attain some competitive goal, or dramatiz- ing situations of adult or group life, of playing formal games. There is a marked sense of belonging or not be- longing to the group. The control of the group situation is in the hands of one or two members who direct the activity of others. The goal and the method of attaining it necessitate a division of labor, the taking of differ- ent roles by various group members, and the organization of activity so that the efforts of one child are supple- mented by those of another. The critical distinction is the goal-directedness of the group. 112 CLASSROOM SOCIO-OBSERVATION Date Class; Time Observer P Piano Blocks Art Puzzles mush: Hose: hoqmz: uumZi wooed: .U.Mn mason: .93 ..H oaouomn owuooouu xomumnm oasm: coouonun mow:u:o oaonom: souoa