'6th ‘I’y—y— ,_ '.r .._..v-...- - 1-7-.“ .v..--n .:_.‘J~ .._»,. 4 1""..2. :34“. . ‘ : 1.1.1,. .f. _. u- - z - 'w—m . ‘ ,. If; T.;. ‘ - -w— -._. . l- , «.5 ., ~1st 'f— .~ . i..}= - [-1. o' .174 0:3';": 'A,‘l ' “2 his}! ‘ 5“?“ A “A“. 3 I ." l ' . f A .. b fizzzA-im’fz ,‘ “Hz. . 9‘. AAA ,c .:A At 1-7 e31. 1‘ s, may A: a .5: {k‘AEAi-‘hw 5; “ ; AA‘JEA 1'" "l L. AfillEA' AAAAIv'LA‘A‘AAAE‘: ' . AAAXA- ‘ £33! A. .11 A A]? “Alfr- A I ‘ '3‘- .‘. ,.-'v . . .,? ,- .., A 4:? ‘ ‘A‘v AAAAAAHAt-A Ali} “AA AA A! AAA-“512% é, )lgufhl sat“ v." a“: x‘ i AAA; ‘ m r .—-o- . '. _. mi}: ;1' A1,. M "1,’ .~ - A». .4 .‘r ._’V ‘:."._...\; .3... - ‘ ’33]. . ,— .- 3"" “A HA A LIAA"A{‘§L. -L 3". 0' AAv'A‘; “1‘! «Hz! O .-| £1 a. . ‘A’A’AAA: y u' u A. . . ' ' . . ~ A V I :36 1"; v v I. . i 1 t, ' 1‘ 3,. A. :A"(' 6'3 .v' AAA' A h H ‘ - l GLA‘V"AAA§.' . 'A; L!“ 31"},{1' mi". ' “5154‘“: ‘JA ’2“! ' ‘ I f “I, I Thu?!» 1:1'?‘ l . ~‘.- THESIS WIHIIIIIIIHI 23 01691 8843 This is to certify that the dissertation entitled THE INFLUENCE OF SOCIAL-ECOLOGICAL FACTORS ON THE GENDER ROLE ATTITUDES 0F FIFTH GRADE AFRICAN AMERICAN FEMALES presented by Margaret Goodiy has been accepted towards fulfillment of the requirements for Ph.D. degree in Famiy and ChiIdEcoIogy 9%me Major professor Iii/51M?" MSUL: an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE MTE DUE DATE DUE (”“5 n In» ""-' 041‘“? .. . ~< . )"5‘ Jfigj-ad 3’3 1/93 chlHCIDuthpGS—p.“ THE INFLUENCE OF SOCIAL-ECOLOGICAL FACTORS ON THE GENDER ROLE ATTITUDES OF FIFTH GRADE AFRICAN AMERICAN FEMALES By Margaret Goodly 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 Ecology 1997 ABSTRACT THE INFLUENCE OF SOCIAL-ECOLOGICAL FACTORS ON THE GENDER ROLE ATTITUDES OF FIFTH GRADE AFRICAN AMERICAN FEMALES By Margaret Goodly The purpose of this study was to determine the relationship between selected social-ecological factors and the total, occupational, and activity gender role attitudes of fifth grade African American females. Urie Bronfenbrenner’s Ecological Systems Theory and Albert Bandura’s Social Cognitive Theory influenced the conceptual model used in this study. The social-ecological factors investigated for this study were parents’: education, gender attitudes, income, occupation, occupational prestige, familial status, employment status, family structure, religiosity, spirituality, and church attendance. Additional social-ecological factors investigated were the child’s television viewing and observation of activities in their social environment. Forty-three fifth grade African American females and their mothers were included in the sample. The sample was drawn from the Lansing School District in Lansing, the capital city of the state of Michigan. Data analyses included descriptive statistics, the Pearson Product Moment Coefficients, multiple regression, and path analyses. Multiple regression analyses were used to determine the best model to predict total, occupational, and activity gender role attitudes of the sample. Path analyses were used to determine possible indirect effects of predictor variables. The results of the study showed that income of parents and the child’s observation of activities in their social environment were most influential in predicting the gender role attitudes of the fifth grade subjects. The income of parents and child’s observation of activities influenced total, occupational, and activity gender role attitudes of fifth grade African American females. Capyright by MARGARET GOODLY 1997 I hereby dedicate this work to parents, Norris and Melba Goodly, my husband and partner, Jose Antonio Garcia, my son Diego, and to my God. iv ACKNOWLEDGEMENTS .Much appreciation is extended to the members of my dissertation committee, Dr. Harriette McAdoo, Dr. Tom Luster, Dr. Robert Griffore, and Dr. Muria Suvedi. Their support and guidance throughout this research is sincerely appreciated. I will be forever grateful for their valuable time, scholarly advice, and encouragement. Special thanks are due to Dr. Harriette McAdoo for her mentorship, advice, and consistent support throughout my doctoral program. Despite an extremely busy schedule the continuous support by Dr. Harriette McAdoo have been unwavering. Special thanks are also due to Dr. Tom Luster for his consistent academic advice and mentorship throughout my doctoral program. I am thankful to all of the many professors in the Department of Family and Child Ecology who mentored me during my program. Particular thanks are extended to Norma Bobbitt, whose classes and mentorship will be forever remembered. I am also thankful for the support of the College of Human Ecology and the . Department of Family and Child Ecology. I am especially appreciative of Ruth Sedelmaier and Mary Faloon for their friendship and detailed knowledge of the essential paperwork necessary for a doctoral program. Finally, I started this program as a single woman and end it as a married woman with an infant son. My deepest love and appreciation is extended to my husband for his support and my son who have been the spark for the completion of this dissertation. TABLE OF CONTENTS LIST OF TABLES ............................ LIST OF FIGURES ........................... LIST OF APPENDICES ........................ CHAPTER 1 - INTRODUCTION ................... Purpose of the study ......................... Significance of the study ....................... Conceptual Model . . . . ...................... Significant Theories of the Study ................... Ecological Systems Theory .................. Social Cognitive Theory ................... CHAPTER H - REVIEW OF LITERATURE ............. Social Cognitive Theory, Ecological Systems Theory, and Gender Role Development ..................... Knowledge of Gender Role Sterotypes . .............. African Americans and Gender Role Socialization ........ Children’s Occupational and Activity Stereotypes .......... Factors Influencing Gender Role Stereotyping ........... Ethnicity ........................... Family Structure ....................... Socioeconomic Status ..................... wok-Hum») oo Parents’ Educational Level ............ Parents’ Occupation ................ Parental Employment ............... Parental Status ................... Parents’ Gender Role Attitudes ........... Church Attendance .................. Religiosity . . .................... Spirituality ...................... Television ...................... Summary . .............................. Research Definitions ......................... Operational Definitions ....................... Research Hypotheses . . . . . ....................... Research Hypotheses Tested ..................... Research Hypotheses Not. Tested .................. CHAPTER III - RESEARCH METHODOLOGY .......... Sample Selection .......................... Data Collection and Procedures ................... Sample .............................. . Research Instruments ........................ Data Analyses ............................ Limitations of the Study ....................... CHAPTER IV - RESULTS ....................... Research Instrument Findings ................... Missing Data ............................ Relations among the Independent Variables ............. Relations among the Dependent Variables .............. Relations between the Independent and the Dependent Variables . . Multiple Regression Analyses .................... Exploratory Analyses: Child Observation of Activities ....... Zero-order Pearson Correlation Analyses ......... Multiple Regression Analyses ............... Path Analyses ............................ Summary of Results ......................... CHAPTER V - DISCUSSION Summary of the Study . ............ . .......... Contributions of the Study ...................... Implications of the Study ..... . . . . ............. Suggestions for Future Research .................. Recommendations of the Study ................. APPENDICES REFERENCES ................... . ......... TABLES l . 2 3 4. 5 .°‘ 10. 11. 12. 13. 14. 15. LIST OF TABLES Schools of Fifth Grade Subjects . . . . . . . . . ...... Demographic Characteristics of Parents ........... Demographic Characteristics of Families ........... Research Instruments ............. . ...... Item modifications of the Shepard Hess Stereotypes of Occupations and Activities .................. ' Shepard Hess Stereotypes of Occupations and Activities . . . Sex-typed items of Child’s Shepard Hess Stereotypes of Occupations and Activities . . . . . ..... . ..... Items Appropriate for both Men and Women . . . ..... Activity items of Exploratory Variable: Child’s Observation of Activities ........................... Zero-order Correlations: Relations among Independent Variables ........................... Zero—order Correlations: Relations among Dependent Variables and Subscales of Dependent Variables ............ Zero-order Correlations: Relations between Independent Variables and Dependent Variables . . . . . . . ....... Multiple Regression Analysis: Dependent Variables ..... Zero-order Correlations: Relations between Exploratory Variable and Dependent Variables .................... Multiple Regression Analysis: Exploratory Variable and Dependent Variables .................. PAGE 29 3O 3 1 32 36 42 43 46 48 49 49 51 53 54 FIGURES l. 2 3 4. 5 LIST OF FIGURES Conceptual Model: Social-Ecological Factors . . ...... Bronfenbrenner’s Ecological System Theory Model Steps of Path Analyses Model ............... Statistical Results of Path Analyses . . . ....... . . . Social-Ecological Model of the Study ........... APPENDICES A. Lansing School District Consent .............. B. Letter of Consent ...................... C. Letter of Introduction ......... . .......... D. Research Directions . . . . . . . . . . . . . . . . ..... E. Shepard Hess Stereotypes of Occupations and Activities F. Child Observation of Activities . . . . ........... G. General Parental Family Questionnaire ........... H. Lansing School District Ethnic Report ........... H LIST OF APPENDICES Crosstab Table: Family Structure by Mother’s Employment . . CHAPTER I INTRODUCTION The focus of this research study is to explore the impact of gender role stereotyping on African American children. This research study specifically addresses two problems. First, research has not focused solely on an African American population to determine how liberal their gender role attitudes are. Second, research has not indicated where children's gender role stereotypes are most pronounced. This research explores whether gender role stereotyping is most pronounced in occupational roles, activity roles, or both. A gender role stereotype, usually considered cognitive, is a set of beliefs; it deals with what men and women are like, and it is shared by the members of particular groups (Williams & Best, 1982). In other words, gender role stereotyping is simply an overgeneralization by society concerning expectations about sex appropriate activities, attributes, preferences, and the desire to adhere to these concepts. Exploration of the relationship between selected predictors and gender role attitudes and stereotypes is conducted with a sample of fifth grade African American children females. A fifth grade population was chosen because past research found younger children to be more stereotypical (Schau & Busch's, 1982 and Billings, 1992). This study also explores whether Urie Bronfenbrenner’s Ecological Systems Theory and Albert Bandura’s Social Cognitive Theory are useful for understanding the influence of those selected predictors to gender role stereotyping. The ecological systems theory serves as the overarching theory, explaining the interrelationships and interactions of 2 gender role attitudes among systems. The primary theory of the study is social cognitive theory explaining gender role attitude development via modeling. Factors found to influence gender role stereotyping in previous research are explored in this study with African American subjects . They are family structure, parents’ income, education, occupation, employment, gender attitudes, church attendance, religiosity, spirituality, and child’s television viewing (Francis, 1996, Willetts—Bloom, 1994, Tuck, 1994, Billings, 1992, Signorielli & Lears, 1992, Pellett, 1993, Jensen, 1993, Fagot, 1992, Barak, 1991, Lytton & Romney, 1991, Nelson, 1990, Eccles, Jacobs, & Harold, 1990, Logan, 1988, Morgan, 1987, Bardwell, Cochran, & Walker, 1986, Hageman & Gladding, 1983, Zuckerman & Sayre, 1982, Perloff, 1982, MacKinnon, Stoneman, & Brody, 1981, and Hensley & Borges, 1981). The study also included predictors that have not previously been studied. These predictors are parental status and child’s observation of activities. Gender role stereotyping is not a new phenomenon in our society and that it can touch our daily lives through various processes. It is important to explore whether influential predictors of gender role stereotyping found in earlier studies are significant for an African American population. PURPOSE OF THE STUDY The overarching purposes of this study are to: 1) Expand the knowledge of gender role stereotyping by focusing on an African American female population; 2) To investigate how social cognitive and social ecological systems factors influence gender role attitudes among African American females; and 3) To modify and assess an instrument designed to test occupational and activity stereotypes. SIGNIFICANCE OF THE STUDY The results of this study are significant because it has educational and policy implications within the United States for curriculum development, career education, and gender issues within our families, school systems, and society at large. In addition, research on an African American population will contribute to the study of gender role stereotyping in our multiracial society. CONCEPTUAL MODEL The conceptual model for this study, as shown in Figure 1, is based upon both Urie Bronfenbrenner’s Ecological Systems Theory and Albert Bandura’s Social Cognitive Theory. Ecological systems theory is seen as being on the outside of the model, explaining the many systems with which the child comes into contact on a daily basis potentially influencing the child. Social cognitive theory, which is seen as directly explaining the influence of the social-ecological factors to the child is on the inside of the model. The independent social-ecological variables are divided into four main categories and are viewed as influencing the child’s gender attitudes (liberality), which is the dependent variable of the study. These categories not only influence each other, they are part of the child’s macrosystem, exosystem, mesosystem and microsystem. The categories of independent variables are personal, work, family, and religion. These categories contain the independent variables identified as parental in nature Two independent variables, considered to be mediating variables, are identified as child factors and placed within the child circle. They are television viewing and the exploratory variable, Child Observation Of Activities. Ecological Systems Theory income TV viewing family structure prestige observation employment .» A religiosity s intualit tiitendancy Figure 1 - Conceptual Model: Social-Ecological Factors SIGNIFICANT THEORIES OF THE STUDY WW. Urie Bronfenbrenner's concept of the natural setting for behavior includes not only the environmental influences the developing individual comes in contact with on a regular basis, but several more distant layers of society and culture also. Social forces that surround the developing child are the focus of the theory (Steuer, 1994). Social forces are not limited to the family. Development in this theory is defined as the individual's growing ability to understand and influence his or her environment. The environment is characterized as a set of nested structures, each inside the next. The theory sees infants and children developing within a multilayered social environment. The ecology of human development is defined by Bronfenbrenner as, “the 5 mutual accommodation between an active, growing human being and the changing properties of the immediate settings in which the developing person lives, as this process is affected by relations between these settings, and by the larger contexts in which the settings are embedded” (Bronfenbrenner, 1979). Key concepts of this multilayered social environment are the microsystems, mesosystems, exosystems and the macrosystems (see Figure 2). The microsystems form the layer closest to the developing child and represent the settings and social relationships he/she experiences on a regular basis. The mesosystem is the next layer, which is the relationships between and among microsystems. The mesosystem is viewed basically as a linkage between two or more microsystems. They serve as important determinants of the individual's development. The exosystems refers to the layer further removed from the child. The child does not actively interact with exosystems but may be influenced by them indirectly. The macrosystems refer to the outer layer of the ecological environment consisting of the beliefs, ideologies, and accepted behavioral practices of the subcultures and general culture to which the child belongs. The macrosystem's influence becomes obvious when one moves from culture to culture and sees how relationships and institutions are restructured from one culture to the next (Steuer,l994). MESOSYSIEM MICROSYSTEM Figure 2 - Bronfenbrenner's Ecological, Systems Model 529W. Albert Bandura and his colleagues have been the primary architects of social learning theory since the 1960s. Bandura demonstrated that a child does not have to make a response or to experience rewarding consequences in order to learn a behavior. He believed that simply observing another person engaging in the behavior is enough. Bandura expanded the theory to cover nonobservable mental, or cognitive, features as well as observable behaviors (Steuer, 1994). Social learning theory is now referred to social cognitive theory, focusing on the child's tendency to learn by acquiring knowledge rather than simply by acquiring behavior through societal influences. Bandura's notion of abstract modeling captures this idea. When a child learns through 7 abstract modeling, she learns a general rule for behavior, rather than a precise set of actions. She can then transfer the rule to a variety of situations and act as the model would have acted under similar circumstances. Modeling, according to Bandura refers to imitative or observational learning (Steuer, 1994). Bandura uses the word modeling, observational learning, and vicarious learning to mean that the child adds to his repertoire of actions by seeing or hearing someone else perform the behavior rather than by overtly carrying out the behavior himself (Thomas, 1992). Societal influences involve differential gender labeling and the structuring of activities in ways that teach sex roles traditionally favored by the culture. Modeling serves as a major conveyor of sex role information. Children are continuously exposed to models of sex-typed behavior in the home, school, and in the media representations of society. On the basis of these sources of sex role information young children learn the behaviors appropriate for their own sex. Social sanctions and observed consequences also make outcomes partly dependent on the sex-appropriateness of actions. Through these direct and vicarious experiences children learn to use sex-typing information as a guide for action (Bussey & Bandura,l984). Social Cognitive theory posits that parents’ modeling behavior is the primary source of sex role development (Juni, Rahamin, and Brannon, 1983). Social Cognitive theory also posits that sex role development is promoted through a vast system of social influences with modeling serving as a major conveyor or sex role information (Vasta, 1992; Bussey & Bandura, 1984). In Social Cognitive theory, gender-role development is largely a psychosocial phenomenon (Steuer, 1994; Bandura, 1986). Using abstract modeling as a basis, one such rule that children often seem to follow is to attend more carefully to the behaviors of their own sex than to that of the other sex. As a result the learning of sex-role behaviors is amplified (Steuer, 1994). CHAPTER I I REVIEW OF LITERATURE ' Current literature on sex-role development and gender role stereotyping will be summarized and studies related to this topic will be reviewed in order to better understand the influence of these factors on children's understanding of gender roles. 0x: cm, ‘ IO'OI . o 9'1. 'H Ivor 2!. for {o ‘ DI‘V‘ Him Sex role development occurs early in the lives of children. In a 1988 study at the University of Rhode Island, an age/sex interaction effect indicated that the tendency to demonstrate a higher degree of same sex stereotyping was most pronounced among four year olds (Albert, 1988). Bussey and Bandura (1984) reported the results of studies which extended previous research in demonstrating the viability of same-sex modeling as a mechanism of sex role development. In a 1978 study at Harvard University, it was indicated that children as young as two-years-old possess substantial knowledge of sex-role stereotypes prevailing in the adult culture (Kuhn, 1978). In a study by Nadelman (1974), it was discovered that by the age of five, children in our culture hold sex role stereotypes similar to adults (Nadelman, 1974). Nelson's (1990) study, contrasting female and male early adolescent sex-role attitudes and behavior development in an ecological context, found that the level of traditionalism in females’ sex-role attitudes was significantly influenced by maternal employment, father's sex-role traditionalisrn, and adolescent's age. In a study on the 9 influence of parents and family context on children's involvement in household tasks, Antill, Goodnow, Russell, and Cotton (1996) found that gender of the child was the strongest contributing variable. Girls were doing more feminine tasks than boys and boys doing more masculine tasks than girls. MW Reis and Wright's (1982) study produced evidence of sex-stereotyping appearing at 3.5 years and increasing both across different age groups and within the same children as they aged. The children's knowledge encompassed both feminine and masculine traits, and was related to general intellectual development. Historical circumstances surrounding enslavement and post-slavery days have led to the stereotype of a matriarchal African American family. The Black woman is stereotyped as being domineering and castrating, while the man is seen as irresponsible and unavailable (Davenport and Yurich, 1991). Unfortunately, most studies of the African American families have focused on low-income and working-class communities (Barnes, 1983). Today, researchers have discovered that there is not one family configuration or set of gender roles consistent across socioeconomic class, geographical region, and educational level among African American families (Cazenave, 1983; Zeff, 1982). The stereotypical matriarchal family, according to Barnes (1983), is most likely to exist among low-income and working-class communities, while patriarchal structures are most likely to exist when the husband has a high occupational status and equalitarian structures when both partners work out of the house. The equalitarian family structure is also believed to be increasing among middle-class families (McAdoo, 1988). Research on the parenting practices among Black families show that more 10 importance is placed on getting a job done than on the sex of the child for a task. Task assignments are based on the child‘s competence and age than the gender of the child (Peters, 1997). The female role for Black women has also been defined as one of strength, independence, and resourcefulness, in contrast to the Anglo, Hispanic, or Asian traditional female role (Reid, 1985). Proof of that resourcefulness could be recent statistics showing that the percentage of Black women who have at least a high school education increased substantially between 1970 and 1993. These same statistics show more Black women than Black men with four or more years of college (Costello and Krimgold, 1996). Q'H'Q 'IIE"S O'Keefe and Hyde's (1983) study investigated occupational sex-role stereotypes of nursery school, kindergarten, third-grade, and sixth-grade children and the effects of acquiring gender constancy. They discovered that children chose stereotyped occupations for themselves even before they had a concept of gender stability. In addition, boys' personal aspiration responses were more stereotyped than girls. The results of Gettys and Cann's (1981) study of female and male children ranging in age from 2.6 to 8 years indicated that the children at each age level made a significant distinction between occupational groupings by gender, with the extent of the distinction increasing with age level. The results were interpreted as indicating the leaming of adult stereotypes concerning the sex appropriateness of occupations occurred among children as young as 2.6 years old. Tremaine, Schau, and Busch's (1982) study found that age accounted for more variance in the occupational sex-typing measures than occupational classification. Younger children were also found to possess more occupational sex stereotyping than older children. A study by Billings (1992) on occupational sex-role stereotyping in elementary 11 students found that second graders had significantly higher sex-stereotyping scores than sixth graders. Pellett (1994) found that boys had more stereotypical perceptions of physical activities than girls. In addition, children in grades two and eight were more stereotypical than those is grades four, six, ten, and twelve. E I U . E 1 E I S . Ethnicim, Bardwell, Cochran, and Walker's (1986) study examined the role of race on the sex-role occupational stereotyping of five-year old kindergartners. The results showed that White children gave more stereotyped responses than did Black children. The study also noted a significant interaction between educational level of parents and race. White children tended to give more stereotyped responses as the educational level of their parents increased. Researchers believed that this occurred because white families become less egalitarian as educational level increases. No other research was found in this area on ethnicity and sex-role occupational stereotyping. W. Research has not been consisted relative to sex role of family structure. In a study by Logan (1988), it was discovered that children from two-parent families were more aware than those from one-parent families of what is traditionally considered appropriate for one's own sex. This differs from MacKinnon, Stoneman, and Brody's (1981) previous research which found that children of single parent families were more liberal in gender role orientation. Billings’ (1992) study however found that family structure was not significantly related to occupational stereotyping. W. A study by Billings (1992) on occupational sex-role stereotyping in elementary students found that socioeconomic status (SES) has a significant 12 relationship with sex-role occupational stereotyping. Girls with low SES parents had higher sex-stereotyping scores than girls with high SES parents, but boys with low SES parents had lower sex-stereotyping scores than boys with high SES parents. Children from lower socioeconomic backgrounds were found to be more conservative and preferred males in the occupations of doctor, engineer, chef, carpenter, school principal, scientist, astronaut, and elementary teacher (Hageman and Gladding, 1983). Another study, involving middle class children, found that the children demonstrated nonstereotypical attitudes toward occupations presented by the researchers ( Zuckerman & Sayre, 1982). MW. Bardwell, Cochran, and Walker's (1986) study examined the impact parents' educational level has on the sex-role occupational stereotyping of kindergartners. Children whose parents were in the middle and high educational levels gave more sex-role stereotyped responses than did children whose parents were in the low educational level. The Billings (1992) study, however, showed that parent educational level was not significantly related to occupational sex stereotyping. W. Tuck's (1994) study on New Zealand adolescents found significant associations between sex role attitudes, gender of the student, and maternal occupation. Females and students with mothers working in careers, were found to have more egalitarian attitudes towards sex roles, than those whose mothers were primarily caregivers. Barak (1991) found that the traditionality of mother's not father's occupation was found to be significantly correlated with the traditionality of the vocational interests of both boys and girls. Wm. Research in this area has yielded conflicting results. Hensley and Borges (1981) found that children of employed mothers were more gender l3 stereotyped. Nelson (1990) found female sex role development positively influenced by maternal employment, and Barak (1991) did not find maternal employment to be significantly related to the traditionality of children vocational interests. W. No research studies were found regarding the maternal or paternal parental status to a child on sex-role occupational and activity stereotyping. This variable of interest is viewed as exploratory, due to the various parental familial status within intact, divorced, single, and blended family structures (Darden, 1992; Jarrnulowski, 1985). W Pellett (1993) found significant correlations between parents' and their children's stereotyping of physical activities. In Fagot's (1992) study of 24, 30, and 36-month-old children, mothers whose children had mastered labels for boys and girls endorsed more traditional attitudes toward women and toward sex roles within the family. These mothers also initiated and reinforced more sex-typed toy play with their children. Willetts-Bloom (1994) found that only young adult females held more egalitarian gender role attitudes than yound adult males. In a meta-analysis of 172 studies, Lytton and Romney (1991) found that in North American families, the only socialization area of nineteen to display a significant effect for both parents is the encouragement of sex-typed activities for boys and girls. Eccles, Jacobs, and Harold (1990) summarized the results of a series of role studies, which showed that parents distort their perceptions of their own children in gender role stereotypic activities, that the child's gender affects parents' causal attributions for their children's performance in gender role stereotypic activities, and that these perceptual biases influence the child's own self-perceptions and activity choices. 14 W. One study by Willetts-Bloom (1994) found that attitudes towards gender roles were linked to religious attendance. In this study, people who reported frequent attendance at religious services were more likely to endorse traditional gender roles. Religigsjty. Three studies found that attitudes towards gender roles were linked to religiosity. Religious devoutness was found to consistently predict five dimensions of gender role attitudes. Those dimensions were familial roles, extrafamilial roles, male/female stereotypes, social change, and gender-role preference (Morgan, 1987). Francis (1996) found that higher levels of religiosity among men and women were a function of gender orientation. Jensen (1993) discovered that highly religious people from Protestant, Catholic, and Mormon denominations supported a more traditional female role. W. No research studies were found regarding the influence of spirituality of a person on gender role stereotyping. This variable of interest is viewed as exploratory. Spirituality is viewed in this study as a possible determinant of religion. [claim Past research have shown that television viewing can reinforce, strengthen, or cause the development of sex-typed perceptions and behaviors. Past research have also suggested that repeated exposure to sex-typed television may reinforce stereotyped achievement aspirations and self image among children (Perloff, 1982). Significant relationships were found by Signorielli and Lears’ (1992) research examining the relationship of television viewing to fourth and fifth graders’ attitudes towards sex-stereotyped chores. In their research children who watched more television were more likely to say that girls and boys should do those chores traditionally associated with 15 women and men. In a previous study, Signorielli (1990) found that the images of both men and women on television tend to be stereotypical. Stroman’s (1991) review of gender related literature indicated that television has contributed to the socialization of African American children. According to Stroman (1991), this socialization may be positive, but it may also be providing examples and role models that negatively affect Black children's attitudes and behaviors. Summary The review of the literature shows how the results of gender role stereotyping studies are similar to each other. The literature supports the idea that children acquire stereotyped views of gender roles at an early age and that views of gender roles become less stereotyped as children become older. Social cognitive theory and ecological theory explain to some degree how these gender roles are learned. The review also includes studies that demonstrate the effects that ethnicity, parental status, income, education, occupation, employment, church attendance, gender roles, and family structure may have on gender role stereotyping. Despite, an attempt at a careful research review, more gender role research is needed on people of color. Most of the research reviewed, consisting of solely white or predominantly white samples, may have little relevancy for people of color, except in a general sense. How influential these factors are on the gender role stereotyping of an African American population, cannot be determined from previous research on solely white or predominantly white samples. A study of gender roles focusing on an African American population is needed in child development research. This is increasingly apparent as we move into the future, where a majority of the population will be people of color and fifty percent of the workforce will be women. 16 RESEARCH DEFINITIONS The following key temrs are used in this study. 1. W Theory developed by Albert Bandura, that views human functioning in a model of triadic reciprocality in which behavior, cognitive and other personal factors, and environmental events all operate as interacting determinants of each other (Bandura, 1986). 2. WW Theory developed by Urie Bronfenbrenner, which is viewed as the mutual accommodation between an active, growing human being and the changing properties of the immediate settings in which the developing person lives, as this process is affected by relations between those settings, and by the larger contexts in which the settings are embedded (Thomas, 1992). 3. Wing An overgeneralization by society concerning expectations about sex appropriate activities, attributes, preferences, and the desire to adhere to these concepts. 4. “finality: The number of "either" responses given in the Shepard Hess Stereotypes of Occupations and Activities, which represents a willingness to deviate from a traditional gender role vieWpoint of occupations and activities. The either response is “men and women.” .. 5. W Maternal and paternal parental status to the fifth grade female 6. W Household and martial arrangement of parental figure(s) 7. WW Parent attendance at church 8. W Home and job related activities in the social environment 17 OPERATIONAL DEFINITIONS The following independent variables are used in this study. 1. Family Structure Theoretical: any adults and child(ren) living together for at least a year Operational: household or marital arrangement of parental figure(s) WW 1. two parent home, married 2. living with partner, not married 3. one parent home, female headed 4 . one parent home, male headed 2. Parental Status Theoretical: maternal and paternal role of female/male adult in home to child Operational: status of mother/father figure in the home towards fifth grader Iijg levels; 1. Birth/biological mother/father 2. not birth/biological mother/father 3. Parents' Education Theoretical: amount of education experience Operational: highest grade completed by mother/father 4. Parents' Occupation Theoretical: current job of mother and father Operational: Occupational Prestige Scale 18 5. Parents' Employment Theoretical: time devoted to work outside home Operational: mother/father commitment to work outside home Wed: 1. employed 2. unemployed 6. Income Theoretical: amount of family income Operational: total amount of reported family income per year 7. Mother's Gender Role Beliefs and Attitudes Theoretical: overgeneralization by society concerning expectations about sex appropriate activities, attributes, preferences, and the desire to adhere to these concepts Operational: mother's occupation, activity, and total liberality scores from the Shepard Hess Stereotypes of Occupations & Activities 8. Parent Church Attendance Theoretical: attendance at religious service Operational: how often parent attends church in a given month 19 9. Religion a. Religiosity Theoretical: religiousness of a person Operational: how religious parent is EIIEI'°"III' 1. Very religious 2. Somewhat religious 3. A little religious 4. not religious b. Spirituality Theoretical: spiritualness of a person Operational: how spiritual parent is E IVIEH I' .1”. 1. Very spiritual 2. Somewhat spiritual 3. A little spiritual 4. not spiritual 10. Television viewing Theoretical: amount of television watched by child Operational: duration of television viewing by fifth grader in a given week 20 *Exploratory Variables 11. Child’s Observation of Activities (COA) Theoretical: home and job related activities Operational: activities observed by fifth grader in their social environment by men and women 12. Father's Gender Role Beliefs and Attitudes Theoretical: overgeneralization by society concerning expectations about sex appropriate activities, attributes, preferences, and the desire to adhere to these concepts Operational: father's occupation, activity, and total liberality scores from the Shepard Hess Stereotypes of Occupations & Activities The following Dependent Variables are used in this study. Children's Gender Role Beliefs and Attitudes Theoretical: overgeneralization by society concerning expectations about sex appropriate activities, attributes, preferences, and the desire to adhere to these concepts Operational: children's occupation, activity, and total liberality scores' from the Shepard Hess Stereotypes of Occupations & Activities 21 RESEARCH HYPOTHESES In order to accomplish the research purposes, several specific research hypotheses, based on the review of literature will be addressed. Research hypotheses are divided by hypotheses tested and hypotheses not tested. Research hypotheses tested include the predictors found to be significantly viable for statistical testing. Research hypotheses not tested include the predictors eliminated due to missing information, lack of variability, and a small sample size. These assumptions are: Wm H 1. Fifth grade females from a two parent family structure will have higher total, occupation, and activity liberality scores than fifth grade females from a one parent family structure. HA1 . Fifth grade females from a two parent family structure will have lower total, occupation, and activity liberality scores than fifth grade females from a one parent family structure. H2. As total income goes up so does the total, occupation, and activity liberality scores of fifth grade females. HA2. As total income goes up the total, occupation, and activity liberality scores of fifth grade females goes down. H3. Mothers with a high level of education will have daughters with higher total, occupation, and activity liberality scores than mothers with a low level of education. HA3. Mothers with a high level of education will have daughters with lower total, occupation, and activity liberality scores than mothers with a low level of education. ”M 22 H4. Fathers with a high level of education will have daughters with higher total, occupation, and activity liberality scores than fathers with a low level of education. HA4. Fathers with a high level of education will have daughters with lower total, occupation, and activity liberality scores than fathers with a low level of education. H5. Fifth grade females with an employed mother will have higher total, occupation, and activity liberality scores than fifth grade females with an unemployed mother. HA5. Fifth grade females with an employed mother will have lower total, occupation, and activity liberality scores than fifth grade females with an unemployed mother. H6. Fifth grade females with a biological father in the home will have higher total, occupation, and activity liberality scores. HA6. Fifth grade females with a biological father in the home will have lower total, occupation, and activity liberality scores. *Exploratory Variable H7. Fifth grade females who view both men and women performing the same activities in their social environments will have higher occupation, activity, and total liberality scores. HA7. Fifth grade females who view both men and women performing the same activities in their social environments will have lower occupation, activity, and total liberality scores. 23 Wed (HNT) HNTl. As the occupational prestige of mothers goes up so does the occupation, activity, and total liberality scores of fifth grade females. HNT2. As the occupational prestige of fathers goes up so does the occupation, activity, and total liberality scores of fifth grade females. HNT3. Fifth grade females with an employed father will have higher occupation, activity, and total liberality scores than fifth grade females with an unemployed father. HNT4. Fifth grade females whose mothers have a high occupation, activity, and total liberality score will have higher occupation, activity, and total liberality scores. HNTS. Fifth grade females with low parent church attendance will have higher occupation, activity, and total liberality scores than fifth grade females with high parent church attendance. HNT6. Fifth grade females with a biological mother in the home will have higher occupation, activity, and total liberality scores. HNT7. Fifth grade females with a very religious parent will have lower occupation, activity, and total liberality scores. HNT8. Fifth grade females with a very spiritual parent will have lower occupation, activity, and total liberality scores. 24 HNT9. Fifth grade females who view less television per week will have higher occupation, activity, and total liberality scores. CHAPTER III RESEARCH METHODOLOGY This research design chapter is divided into the following parts: 1) Sample selection; 2) Data collection and Procedures; 3) Sample; 4) Research instruments; 5) Data analyses; and 6) Limitations of the study. SamnILSelectian The sample population for this study was African American families in a Mid-Michigan school district. The sampling frame and the unit of analysis consisted of the total population of fifth grade African American elementary females and their mother or mother figure from the Lansing school district. This correlational research design specifically included the sampling of the total population of fifth grade African American elementary females. As stated by the Lansing School District's Research and Evaluation Services department, the population of fifth grade African American females at the time of the study was 211. The Lansing school district was contacted regarding permission and information pertinent to the study. Upon receiving research approval by the Lansing School District, the parent(s) of all fifth grade African American females were contacted regarding permission and participation in the study. Once permission was granted by the parent(s), researchers sent the questionnaires to the participating parents and administered the necessary questionnaire to the girls at their particular schools. The girls were given the opportunity to assent to the participation of the study at the beginning of the questionnaire 25 26 administration. MW Data collection took place during the Spring of the 1997 school year. Data collection was also conducted in a three stage process. The first stage involved parental contact in writing regarding their interest in the study, via the Lansing school district. Parental contact included a letter of introduction, a promise of ten dollars for participation, and a returned stamped envelope. The second stage involved sending research questionnaires to parents who indicated interest in the study in the that stage. The parental research package included another letter of introduction, consent form, research directions, all pertinent parental questionnaires, and a stamped envelope for returning the questionnaire. Parental questionnaires included the general parental family questionnaire and a Shepard Hess Stereotypes of Occupations and Activities questionnaire for the mother. In addition, a Shepard Hess Stereotypes of Occupations and Activities questionnaire for the father was supplied for all fathers who chose to participate. The third and final stage involved administering the child questionnaire individually or in a group setting to the students whose parents’ granted permission at their particular school. As stated above, the students were given the opportunity to assent to participation in the study at the beginning of the interview. Finally, as stated in the consent form, parents were paid a fee of ten dollars upon the completion and return of all pertinent questionnaires (see Appendices) to the researcher. This included the General Parental Family questionnaire, the Shepard Hess Stereotypes of Occupations and Activities for mother and child, and the Child Observation of Activities. 27 Sunnis The initial response rate of the parents contact from the Lansing School District was 30% of the total fifth grade African American female population. However, the final sample consisted of forty-three African American mother—child pairs. This change was due to the lack of returned mother questionnaires (see Table 4) necessary for data entry. The fifth grade subjects represented 20% of the fifth grade African American females in the Lansing School District. The fifth grade subjects also represented twenty-four of the thirty- four elementary schools in the Lansing School District (see Table 1). A Lansing School District, October, 1996 Ethnic Report, found that there were 1,212 African American females attending these schools (see Appendix H). The report also noted the total percentage of African Americans in these schools (see Table 1). Characteristics of the total sample are presented under three categories: child characteristics; parental characteristics, and family characteristics. Wes. Demographic information obtained from mothers found that the age of the girls ranged from ten years to thirteen years with a mean age of eleven (SD .627). The girls reported viewing five hours of television per day on average (SD 2.97). W. Parental characteristics, including ages, education, employment status, religiosity, spirituality, and church attendance are presented in Table 2. The mean age of the parents was 37 years (SD 6.30) for mothers and 39 years (SD 6.56) for fathers. Both mothers and fathers had a mean education level of 13 years. Thirty (70%) of the mothers’ reported that they were working. The mothers also reported that twenty-one (49%) of the fathers were working and three were not. Nineteen (49%) of the mothers reported that they did not know if the father figure living outside the home was 28 working. Thirty-two (74%) of the mothers reported that one or both parents attended church with their child and thirty-four (80%) reported that they were very or somewhat religious or spiritual. W Familial characteristics, including number of children in home, income, family structure, marital status of mother, father in home, and mother and father figure, are presented in Table 3. The sample consisted of twenty-seven (63%) two-parent families and sixteen one-parent (37%) families. Two-parent families included six mothers currently living with male partners. Thirty-nine (91%) of the fifth grade girls were the biological child of the participating mother figure, while only seventeen (40%) had a biological father figure in the home. Twenty-three (54%) of the mothers reported having a father figure in the home. There was an average of three (512 1.33) children in the home. The mean income of the families was $31,362 (812 $24,606). 29 Table l - Wm African American School School 11 = 43 Females % _n_ African American Allen 3 67 38 Attwood l 32 24 Averill 2 59 57 Bingham 2 35 32 Cavanaugh 2 51 29 Chrmberland 4 5 1 24 Fairview 1 24 18 Grand River 1 37 30 Gunnisonville 1 25 20 Harley Franks 2 47 46 Kendon 2 41 29 Lewton l 55 37 Maple Grove 2 34 24 Maplewood 2 29 21 Moores Park 1 37 37 Mount Hope 2 39 19 North 2 94 37 Northwestern 1 25 35 Pleasant View 1 68 46 Post Oak 1 43 20 Riddle l 54 58 Verlinden 2 67 48 Wainwright 1 74 53 Woodcreek 5 126 63 Total 1,214 30 Table 2 - WWW Characteristics % M 5]; Age of Mother 37.30 years Age of Father 39.42 years 6.30 Mother's level of education 12.81 years 6.56 Father's level of education 12.61 years ' 1.71 Working 70 Unemployed 30 2.28 Working 49 Unemployed 7 missing data 44 very or somewhat 80 little or none 20 ChmhAttendance one/two parent 74 no attendance/missing 16 31 Table 3 - WWW Characteristics % M 512 Number of children 3.12 1.33 Income $31,362 $24,606 two-parent 63 one-parent 37 Never married 30 Married 37 Divorced 16 Other 17 Biological 91 Other 9 Biological 40 Other 14 Missing 46 Father in home 54 No Father in home 46 32 W This study involved the use of three research instruments (see Appendicies e, f, and g). The instruments were the Shepard Hess Stereotypes of Occupations and Activities, the Child Observation of Activities (COA), and the General Parental Family Questionnaire (see Table 4). The intention of the study was to gather research data from the fifth grade subjects, the mother figures of the fifth grader, and the father figure of the fifth grader. The Shepard Hess Stereotypes of Occupations and Activities was given to the fifth grade subject, the mother figure, and the father figure as shown in Table 4 . Unfortunately, there was an insufficient number of father figures completing the research instrument for inclusion in the study. The Child Observation of Activities was given only to the fifth grade subjects and the General Parental Family Questionnaire was given only to the participating parent. Table 4 - Researchinsmrments Instrument Child Mother Father Shepard Hess Stereotypes of Occupations and Activities x x x Child Observation of Activities x General Parental Family x Questionnaire We: The Shepard Hess Stereotypes of Occupations and Activities was developed in 1975 by Winifred O. Shepard and David T. Hess to measure liberality in terms of stereotypes of occupations and 33 activities (Beere, 1990). The instrument was chosen because of its inclusion of both occupational and activity items and the ease of administration to both children and adults. Permission was given to use and modify the Shepard Hess of Occupations and Activities by David Hess. David Hess was contacted at S.U.N.Y. College at Fredonia, in the summer of 1996. An unsuccessful attempt was also made to contact Winifred O. Shepard for permission. Development of the measurement tool involved a set of 44 occupations and activities given to 50 college men and 50 college women. College students were asked to indicate for each item whether the traditional American viewpoint in general would recommend it for a man, woman, or both. Sixty percent of both the male and female judges agreed to the categorization of the items. Forty-three of the items were retained on the scale. The measurement tool consists of .a list of forty-three occupations and work-related activities. The scale can be administered individually or with written instructions to indicate who "should" perform each occupation or activity. Each occupation is accompanied by 9’ ‘6 three response options. The options are: "only men, only women, ” or “men and women" (Beere, 1990). The instrument has two scoring methods. The first scoring method, used by Shepard and Hess (1975), involves finding a liberality score by counting the number of times the respondent selects an answer that differs from the traditional or stereotyped response. The second scoring option, used by Archer (1984), involves finding a liberality score by counting the number of "either" responses (men and women). Either responses are rated as " 1" and other responses are rated as "0". Both scoring methods yield scores that ranged from 0 (very traditional) to 43 (very liberal). Validity of the measurement was found for eight graders, college students, and adults. In all three cases, females were more liberal. Eight graders were found to be more liberal than kindergartners, and college student the most liberal (Shepard & Hess, 1975). 34 Similar validity was found by Archer (1984) at the kindergarten level. Archer (1984) also found that liberal scores were associated with increasing age among fifth graders and eleventh graders. Researchers employed the use of nonparametric statistics with the original data due to what they felt was the ordinality of the response measures (Shepard & Hess, 1975). However, it appears as if Shepard and Hess (1975) utilized both ordinal and interval methods. For example, the liberalin score was defined in terms of the number of items on which each participating subject deviated from traditionality. These scores ranged from 0 to 43 (Shepard & Hess, 1975). Modifioafions. The measurement tool was modified to obtain three scores, as opposed to one total score, utilized by Shepard and Hess (1975) and Archer (1984). The modified Shepard Hess Stereotypes of Occupations and Activities instrument used in this study included a total, occupational, and activity score. The activity and occupational items are treated as subscales of the instrument and their score viewed as subscale scores. The activity subscale consists of seventeen items and the occupational subscale thirty-four items. The total, occupational, and activity scores are treated as individual variables. The scores of the research instrument were also used as both dependent variables and independent variables. The scores from the Shepard Hess Stereotypes of Occupations and Activities given to the child were treated as dependent variables and scores from the Shepard Hess Stereotypes of Occupations and Activities given to the mother were treated as independent variables. The scoring method used was the method employed by Archer (1984) which involved finding a liberality score by counting the number of "either" responses (men and women). Scores were also treated as interval data, as opposed to ordinal data, as employed by both Shepard and Hess (1975) and Archer (1984). Support of this 35 modification was that the data for the study showed that the dependent variables (child’s Shepard Hess Stereotypes of Occupations and Activities) had a normal distribution unlike the findings of Shepard & Hess (1975). Furthermore, interval data allows for more in- depth analysis of the factors influencing liberality. Addition modifications made to the Shepard Hess Stereotypes of Occupations and Activities research instrument were the deletions of items, a change of the language, and the addition of new items. As shown in Table 5, four activities and five occupations were delated, the language of four previously delated items was changed, and eight activities and three occupations were added to the instrument. The final modified instrument had a total of fifty-one items. Thirty-four were occupations and seventeen were activities (see Appendix E). TableS- "111.10..- Cu 0 .t'i‘ - 1! .. Item E I . . E 'v' . knit a sweater paint a picture repair shoes fix a car President of US. banker High School principal kindergarten teacher repair a car beauty parlor/hair salon clean the bathroom plant a flower garden mow the lawn drive a tractor take out garbage shop for groceries do house repairs go hunting airplane attendant basketball coach repair computers '.‘0M,IC 101 O . 0Q -- , 37 W This exploratory instrument based on the Shepard Hess Stereotypes of Occupations and Activities was used to measure who the fifth grade subject observed in the last month performing certain activities in her social environment. The measurement tool consisted of a list of fourteen work-related activities (see Appendix F). The scale could be administered individually or with written instructions to indicate who "the child saw in the last month performing each activity.” Each activity was ’9 ‘6 accompanied by three response options. The options are: "only men, only women,” or “men and women”. The scoring method utilized was similar to the Shepard Hess Stereotypes of Occupations and Activities. However, this instrument involved finding an observation score, rather than a liberalin score, by counting the number of "either" responses (men and women). Either responses were rated as "1" and other responses were rated as "0." Camoralfiarmtalliamilxflnstioman The third instrument used in this study was a demographic parent questionnaire developed by the researcher to gain important descriptive information. This included questions relating to education, parental status, relation of mother figure, marital status, family structure, income, occupation, employment, and parent church attendance (see Appendix G). W. The Occupational Prestige tool was used to rate the occupation of the parents in this study. It was developed at the National Opinion Research Center in 1963-1965 in a project on occupational prestige directed by Robert W. Hodge, Paul S. Siegel, and Peter H. Rossi. The concept of prestige was defined by the researchers as the respondents' estimation of the social standing of occupations (Hodge, Siegel, and Rossi, 1965). 38 llataAnalxsis Measures of central tendencies were used to describe the sample. Pearson correlations were used to characterize the relationship among and between the dependent and independent variables. Reliability analyses was performed on the revised Shepard Hess of Occupations and Activities measurement tool and the exploratory variable, Child’s Observation of Activities. Reliability analysis was used to detemrine the internal consistency of the dependent variables and the exploratory variable. Hypotheses were tested at the p < .05 chance probability level. Three multiple regression analyses were used to determine a multiple correlation between the dependent variables and the social-ecological variables. Variables found to be significantly correlated with the dependent variables by way of the Pearson correlations analyses were entered into the multiple regression analysis. Occupation, activity, and total gender role scores of the Shepard Hess were the criterion or dependent variables. Multiple regression were also used to determine a multiple correlation between the dependent variables, correlated predictor variables, and the exploratory variable, Child’s Observation of Activities (COA). The goal of the regression analysis was to find the best set of predictors which may influence a child's gender role attitudes. As a result of the aforementioned statistical analyses, path analyses was employed. Path analyses was used to determine if an indirect relationships exist between the predictor variables and the dependent variables. Analysis of Variance (ANOVA) and crosstabs were also utilized in the research study to describe associations of the sample. 39 I . °| Ii I ll SI 1 There appear to be six major limitations affecting the results of this study. Those limitations are sample size, generalizability, missing information, respondents, measurement tools, elirrrination of variables, and the inability to assess the influence of the father figure on child’s liberality. The first and most serious limitation of this study is a sample size of forty-three mother-child pairs. This sample size from the Lansing, Michigan geographical area may be too small to allow the findings to be generalize to any other geographical area outside Lansing, Michigan or the midwest. In addition, a random sample could not be used due to constraints by the Lansing School District. A substantial amount of missing information altered the use and viability of some variables. In this study, two parent families were most likely to participate than one parent families. These two parent families were most likely to have more education, income, and a biological father figure in the home, which evidently affects the outcome. Therefore, respondents self-selection could be perceived as a limitation. The General Parental Family Questionnaire and the Child Observation of Activities measurement tools used in the study are also a limitation. These measurement tools were not piloted with another population for comparison. Although, the Shepard Hess Stereotypes of Occupations and Activities was modified for this study it is not considered a limitation due to similar findings with the original tool (Shepard & Hess, 1975). However, the restricted range of scores on the mothers Shepard Hess is a limitation. Additionally, the completion of the Shepard Hess and the COA by the child could be viewed as a limitation. Shared method variance may account for the relationship between these two instruments. As stated previously, nine variables were eliminated from the analyses due to missing information, lack of variability, and small sample size. The elimination hindered the testing of nine hypotheses to the research study. 40 The last limitation, the inability to attain the influence of the father figure to child’s liberality is considered serious also. As we move towards the millennium, research should also move from the mother only focus to a more egalitarian focus on mothers and fathers. Unfortunately, the attempt to gather the gender attitudes of the father figures in this study fell dismally short of expectations. A low number of completed questionaire from the father figures could not be used. CHAPTER IV RESULTS The presentation of results is this chapter is divided into the following parts: 1) Research instrument findings; 2) Missing data; 3) Relations among the independent variables; 4) Relations among the dependent variables; 5) Relations between the independent variables and the dependent variables; 6) Multiple regression analyses; 7) Exploratory analyses; 8) Path analyses; and 9) Summary of results. The purpose of these analyses was to determine if the predictor variables are related to the gender role liberality of the fifth grade females. ResearoLInstmmenLEindings WWW. Total liberality scores for the fifth grade subjects ranged from eleven to fifty-one with an average liberality score of 30 (SD 11.21). Occupation liberality scores ranged from eight to thirty-four with an average liberality score of 20 (SD 7.24). Activity liberality scores ranged from two to seventeen with an average score of 10 (SD 4.29) (see Table 6).. Due to a small sample size, scores were counted manually for seven subjects with missing data in order to maintain sample size. Reliability coefficients for the child’s Shepard Hess are also shown in Table 6. Cronbach’s alpha was .94 for the total, .91 for the occupation, and .85 for the activity subscale. Cronbach’s alpha for mothers Shepard Hess was .96 for the total, .94 for the 41 42 occupation, and .89 for the activity subscale. Table 6 - WW .M SD Reliability V . Child total score 30.26 11.21 .94 (n = 39) Child occupation score 20.26 7.24 .91 (n = 37) Child activity score 10.00 4.29 .85 (n = 41) Mothers' total score 45.47 8.84 .96 (n = 40) Mothers' occupation score 30.26 5.98 .94 (n = 41) Mothers' activity score 15.21 3.01 .89 (n = 42) A careful review of the participating children’s Shepard Hess items found thirteen occupational items and five activity items to be sex-typed by a majority of the children. Items were identified as sex-typed if the respondants chose the “only men” or “only women” choices. The activity of fishing was equally sex-typed by the fifth grade subjects as “only men” or “only women.” As shown in Table 7, the occupations of perfume salesperson, secretary, cheerleader, nurse, beauty parlor/hair salon worker, decorator, daycare worker, ballet dancer, broadcaster, truck driver, car mechanic, plumber, and solider were sex-typed by a majority of the participating children. The activities of sewing, hunting, driving a tractor, and planting a flower garden were also sex-typed by a majority of the participating children. Table7- ‘.-tu'!;u 0 i!" i":.'!' "one; 0 9 A...” l i 'v' . Item only women only men 11 11 sell perfume 30 secretary 30 cheerleader 29 nurse 28 work in a beauty parlor/hair salon 25 decorate a room 25 daycare worker 23 ballet dancer 22 drive a truck 21 broadcast a football game 32 repair a car 29 plumber 27 solider 23 E . . . sew . 28 go hunting 28 drive a tractor 25 plant a flower garden 22 go fishing 21 21 Twenty-one occupation and twelve activity items were identified as being appropriate for both men and women by the fifth grade subjects in their Shepard Hess instrument. These items, as shown in Table 8, are varied and reflect who the fifth grade subjects believed should perform these occupations and activities. 44 Table 8 -.Itcm_amrootiatefoLboth_Men_andR/omon Item men and women n deliver mail 41 write a book . 40 own a supermarket 39 lead a band/orchestra 38 doctor 38 high school principal 37 lawyer 36 fly a plane 33 own a factory 32 ride a horse 29 basketball coach 29 kindergarten teacher 28 direct traffic 27 repair computers 27 librarian 26 put out fires 25 airplane attendant 25 sell refrigerators 24 install a telephone 24 telephone operator 22 engineer 22 E . . . taking care of young children 37 wash dishes 36 clean the bathroom 35 vacuum 35 cook dinner 33 iron clothes 31 check out groceries 30 take out garbage 29 shop for groceries 28 laundry 27 house repair 26 mow the lawn 24 ' ' . Liberality scores for participating mothers showed less variability than the participating children. Total liberality for the mothers ranged from twelve to fifty-one with an average liberality score of 46 ($118.84). Occupation liberality scores ranged from ten to thirty-four with an average score of thirty (SD 5.98). Activity liberality scores ranged from two to seventeen with an average score of 15 ($113.01) (See Table 6). Results from a review of items from the Shepard Hess of the participating mothers found none to be sex-typed by this group. Items are identified as sex-typed if a majority of the respondents chose the “only men” or “only women” choices. Participating mothers overwhelmingly chose the “men and women” response choice, therefore producing no variability in the variable. Furthermore, with limited variability, there was no relation between child’s liberality scores and the scores of the participating mothers. In addition, due to mothers’ highscores, the distribution was not normal. WWWL Total Observation scores for the fifth grade subjects ranged from zero to fourteen with an average observation score of 6 (SD 4.08). A reliability coefficient of .86 (n =36) was found. Scores were added manually for six subjects with missing information to maintain the sample size. Four of the subjects missed two questions and one subject missed two. One of the participating subjects did not answer the question regarding the use of the vacuum in the home. This was because no one in her environment had carpet to vacuum. Similar to the Shepard Hess, items were identified as sex-typed if a majority of the respondents chose the “only men” or “only women” choices. A review of the items found six of the activities performed by women, three by men, three by both men and women, and two seen equally performed by men and women and only men in the social environment of the fifth grade subjects. As shown in Table 9, the activities of sewing, 46 dusting, cooking dinner, shopping for groceries, sweeping, and ironing were performed by women. The activities of house repairs, taking out garbage, and moving of furniture were performed by men. The activities of vacuuming, washing the dishes, and cleaning the bathroom were performed by men and women. Taking care of young children and doing laundry were the only activities the subjects saw in their social environment, performed by both men and women and only men. Item only women only men men and women [I D n sew 32 household dusting 24 cook dinner 23 shopping for groceries 23 sweeping 20 iron 20 house repairs 26 taking out garbage 21 move furniture 19 vacuum 25 wash dishes 24 clean bathroom 23 care of young children 21 21 laundry 20 20 W W. Missing information, a lack of variability, and a small sample size hindered the use of some independent variables. Those variables were mothers’ gender attitudes (total, occupation, and activity), father employment, 47 biological mother in home, parent church attendance, religiosity, and spirituality. A review of the central tendencies showed a large amount of missing information for the father’s employment variable and only one large group for biological mother, and church attendance. In addition, the average income were given to seven subjects with missing information in order to maintain sample size. W. A small sample size as well as missing and insufficient information regarding the specific occupation of both the mothers and fathers hindered the use of the Occupational Prestige Tool as an independent variable for this study. The instrument lacked specific categories or did not fit into a category for some of the information provided by the participating mothers regarding occupation. Zero-order correlational analyses of the independent variables indicate that two-parent families have higher incomes, mothers and fathers with higher education have higher incomes, mothers with higher incomes and education are more religious and spiritual, and employed mothers are more spiritual. The spirituality and religiosity variables were found to be so similar in nature (see Table 10) that it was not appropriate for use in a multivariate analysis. No significant associations were found with additional Analysis of Variance (AN OVA) analyses conducted. Significant results of the crosstab analyses found that twenty-one (49%) of the mothers in the two-parent family structures were employed and 6 ( 14%) of the mothers were not (see Appendix I). Table 10 - _ .- Farnily Income Mother's Father’s Fathers Mother’s Structure Education Education in home Employment Family - .30“ .83 ’ " " Structure Income - .5 6 * Mother’ 5 Education .56“ - .41 * Fathers Education .45* .41 * - Father in home 83*" 42" - Mother’ 5 Employment - Note: *p<.05 I""'p< .01 "*p<.000 WWW Results of the zero-order correlations verify that the subscales of the modified Child’s Shepard Hess Stereotypes of Occupations and Activities are highly correlated with total score. Table 11 shows that the occupation subscale is correlated at .98 and the activity subscale at .95 with the total score. Table 11 also shows that the occupation subscale is correlated at .88 with the activity subscale. Table 11 - . Child Child Shepard Hess Shepard Hess Occupation Activity Walrus Child Shepard Hess Total .98* .95* W Child Shepard Hess Occupation .88* Note: I"p < .000 Vt” ,.‘..1 .. ”hwy; a”. H, ..,.,.H.! 2.2-:- Zero-order correlations between the independent variables and the dependent variables indicate that mothers with higher education have more liberal sex-role attitudes (see Table 12). Mothers with higher incomes had higher liberality scores on the Shepard Hess total as well as the occupation and activity subscales. The correlations suggest the as mothers’ income goes up so does the liberality among the children. None of the other correlations were large enough to be statistically significant. Table 12- .10‘0'7 Irr' «.Jr ‘ i' 2.0!, 01‘ “t inure/r Vat-.07 1.1. W Income Mother’s Education Child Shepard Hess Total .47* Child Shepard Hess Occupation .48* * * .3 1* Child Shepard Hess Activity .42* * Note: *p < .005 ** p < .01 "*p < .001 50 Multinfloareasionmlxses Findings of the zero—order correlations yielded only two significant predictor variables to be used in the regression analyses. These predictor variables were income and mothers’ education. The mothers’ education predictor variable was converted into a dummy variable for the multiple regression analyses. A series of three stepwise regression analyses was then used to identify how income and mothers education would uniquely influence the total, occupational , and activity liberality among the fifth grade African American females. Due to a small sample size and a low number of significant predictors found during the zero-order correlational analyses, stepwise regression was utilized for the regression. Utilization of the stepwise regression method incorporated a larger sample size for analyses. The first regression analysis included the two significant predictor variables with the total liberality score as the dependent variable. As shown in Table 13, the predictors accounted for 22% of the variance in total liberality (F (1,41) = 11.54 = o = .01) for the fifth grade females. The income variable was significant at .001. The second regression analysis, utilized the occupational liberality score as the dependent variable with the two predictor variables. This regression analysis accounted for 23% of the variance in occupational liberality (E (1,41) = 11.97 = o = .01) for the fifth grade females. The income variable was significant at .001 (see Table 13). In the third and final regression analysis, the activity liberality score was the dependent variable. Income accounted for 18% of the variance in activity liberality (E (1,41) = 8.88 =_o = .01) for the fifth grade females. The income variable was significant at .005 for a third time (see Table 13). 51 Table 13 - WWW Standardized [3 R2 = .22 Income .47“ Qcounafionflienariflemflfl Income .23 .48* Income .18 .42* Note: Mother’s education was eligible for entry in each of the analyses but was not a significan predictor of any of the three outcomes. * p < .05 In summary the results indicate that income was the only significant predictor of liberality in all three regression analyses. Eighteen to twenty three percent of the variance was explained by income. The occupational liberality subscale had the most variance explained by income, followed by total liberality, and the activity subscale. In other words, the results imply that fifth grade females who live in homes with higher incomes tend to have higher total, occupational, and activity liberality scores. Although the correlation analyses showed that mothers education was significant, the relation between education and liberality scores was reduced to nonsignificance when income was controlled. 52 E I I ll 'Clillfll l' I! |'°|' The exploratory predictor variable was an observational score of gender activities in the girl’s social environment. In other words, fifth grade females were asked who did they see in the last month performing certain activities in their environment. WW2: Zero-order correlations among the dependent variables and between the independent variables was conducted with the exploratory variable. Results of the correlations among the independent variables found the exploratory variable significantly correlated with only the income variable at .38. This finding shows that fifth grade females who live in a family with a high income are more likely to see men and women doing similar activities in their social environment. Results of the correlations between the dependent variables and the exploratory variable, showed that the observation measure was significantly correlated with all dependent variables. As shown in Table 14 the exploratory variable was significantly correlated with the Shepard Hess total at .48, the occupational subscale at .45, and the activity subscale at .48. This finding implies that fifth grade females with high total, occupational, and activity liberality scores are more likely to see men and women doing similar activities in their social environment. The significance of the correlational findings also allows the exploratory variable to be included in a series of regression analyses. 53 Table 14 - Child Observation of Activities Child’s Shepard Hess Total .48" Child’s Shepard Hess .45"I Occupation Child’s Shepard Hess .48" Activity Note: *p < .005 " p < .001 NW. A series of stepwise multiple regression analyses similar to the previous analyses was utilized with the exploratory variables. In the first regression analysis, the Shepard Hess total was the dependent variables. Three predictor variables, however, were utilized for these analyses. They were income, mothers education, and the exploratory predictor variable, Child Observation of Activities (COA). As shown in Table 15, the predictor variables accounted for 32% of the variance in total liberality (E (2,39) = 9.29 = o = .001) for fifth grade females. Income and the COA were significant predictors. The predictor variable accounted for 31% of the variance in occupational liberality (E (3,39) = 8.83 = p = .001) for fifth grade females. Income and the COA were significant predictors again. In the third and final regression analysis, the predictor variables explained 23% of the variance in activity liberality (E (1,40) = 12.01 =_o = .001) for the fifth grade females. The only predictor variable found to be significant was the COA. 54 Income .34“ Child Observation of Activities .35* Mom education WW -31 .35"l Income .32“ Child Observation of Activities Mom education 'v' = .23 Child Observation of Activities .48"‘ Income Mom education Note: I"p < .05 The results of the regression analyses with the Child Observation of Activities, suggest that it is just as important as the income variable in predicting gender attitudes. In other words, the income of the parent and what the fifth grade girls see in their social environment are good predictors of their total and occupational liberality. Furthermore, fifth grade girls who live in homes with higher incomes tend to see more men and women performing the similar activities in their social environment. This could also be due to the possibility that these are the homes with men living there. As shown in Table 15, the COA and the income predictor variables together accounted for more variance in total and 55 occupational liberality than income alone. Thirty-two percent of the variance was explained for total liberality and 31% for occupational liberality. In the first regression analysis, income explained 22% of the variance for total liberality and 23% of the variance for occupational liberality. Similarly, the COA predictor variable alone accounted for more of the variance in activity liberality than the income variable alone for the fifth grade subjects. The COA predictor variable explained twenty-three percent of the variance while the income predictor variable explained eighteen percent. EatLAnalx2: Path analyses were conducted to determine possible indirect effects of the mothers’ education and family structure on the gender attitudes of the fifth grade subjects. A series of regression analyses were conducted to deterrninethe direct and indirect effects of the predictor variables. As shown in Figure 3, there were three stages in this path analyses model. In step one, income was used as the dependent variable with mothers education and family structure as independent variables in a regression model. In step two, COA was used as the dependent variable with mothers education and family structure as independent variables in a regression model. In the third and final step, mothers education, family structure, income, and COA were independent variables. The enter method was utilized in the regression analysas. In step one, only the mother education predictor accounted for 36% of the variance in income (E (2,40) = 11.42 = o = .000). Findings suggest that there is a significant relationship to income and some indirect effect of mothers education to gender attitudes. In step two, no significant relationship was also with either variable. The predictor variables accounted for only .05 of the variance of variance in COA (E (2,39) = 1.01 = o = .374). In the third and final step of the analyses, the predictors accounted for 36% of the variance in total liberality (E (4,37) = 5.30 = o =.002) for fifth grade subjects. Income and 56 the COA were significant predictors in the regression. The predictors accounted for 36% of the variance in occupational liberality (E (4,37) = 5.11 = o = .002). Income and COA were significant predictors once again. In the last regression analysis, the income and COA predictor variables were significant once again. The predictors explained 35% of the variance in activity liberality (E (4,37) = 4.87 = o = .003). Figure 4 shows a model of the final statistical results of the path analyses. In this model, the predictor mothers education points to a direct relationship to income and the indirect effect to the total, occupation, and activity gender attitudes of the fifth grade subjects. 57 Gender Attitudes Figure 3 - Steps of Path Analyses Model 58 Mothers Income Education Family Structure Mothers Education Gender Attitudes @/ Activity Note: Standardized Betas presented on figure Figure 4 - Statistical Results of Path Analyses SummarLoLBoflrlt: In this section, a summary of the results of the study is presented in terms of the research hypotheses. H1 . F ifth grade females from a two parent family structure will have higher total, occupation, and activity liberality scores than fifth grade females fi'om a one parent family structure. HA 1. Fifth grade females from a two parent family structure will have lower total, occupation, and activity liberality scores than fifth grade females from a one parent family structure. 59 Zero-order correlations between the family structure variables and the dependent variables were not significant. Based on these findings, the family structure predictor variable was included in the multiple regression analyses. Therefore, the hypothesis was not supported in this study. The findings of the data analyses are sirniliar with Billings’ (1992) study, which found family structure not significantly related to occupational stereotyping. Results of path analyses also found no significant indirect effect of the family structure variable. H 2 . As total income goes up so does the total, occupation, and activity liberality scores of fifth grade females. HA2. As total income goes up the total, occupation, and activity liberality scores of fifth grade females goes down. Results of the zero-order correlations between the income variable and the dependent variables indicated that mothers with higher incomes had higher total, occupation, and activity liberality. The income variable was placed into the multiple regression analyses, based on the findings of these correlations. Results of the multiple regression analyses supported the hypothesis that fifth grade females living in homes with higher incomes tended to have higher total, occupation and activity liberality scores. Results are consisted with Hageman and Gladding’s (1983) study in which children from lower socioeconomic backgrounds were more conservative in their gender beliefs. The children were also found to have preferred males in the occupations of doctor, engineer, and principal. The results are also similar to Zuckerman and Sayre’s (1982) study. In this study nonstereotypical attitudes toward occupations were found among 60 middle class children. H3. Mothers with a high level of education will have daughters with higher total, occupation, and activity liberality scores than mothers with a low level of education. HA3 . Mothers with a high level of education will have daughters with lower total, occupation, and activity liberality scores than mothers with a low level of education. Mother’s education had a positive relationship with the occupational liberality dependent variable. This findings indicated that mothers with more education are more liberal in occupational liberality than mothers with low education. However, results of the multiple regression analyses did not support this hypothesis. The maternal education variable was not found to be significantly related to the total, occupation, or activity dependent variable when income was controlled. Results of the multiple regression are consisted with Billing’s study (1992). However, an indirect effect was found between mothers education and gender attitudes when path analyses were conducted. Mothers’ education was significantly related to income. Income was significantly related to the gender attitudes of fifth grade subjects. H4. Fathers with a high level of education will have daughters with higher total, occupation, and activity liberality scores than fathers with a low level of education. HA 4. Fathers with a high level of education will have daughters with lower total, occupation, and activity liberality scores than fathers with a low level of education. 61 No relationship was found between father’s education and the dependent variables. As a result, this variable was not included in multiple regression analyses and the hypothesis not supported. Findings are consisted with Billings’ (1992) study, that found parent educational level not significantly related to occupational sex stereotyping. H5. F tfih grade females with an employed mother will have higher total, occupation, and activity liberality scores than fifth grade females with an unemployed mother. HA5. Fifth grade females with an employed mother will have lower total, occupation, and activity liberality scores than fifth grade females with an unemployed mother. No relationship was found between this variable and the dependent variables. As a result, this variable was also eliminated from multiple regression analyses and the hypothesis not supported. H6. Fifth grade females with a father in the home will have higher occupation, activity, and total liberality scores. 0 HA 6. F ifth grade females with a biological father in the home will have lower total, occupation, and activity liberality scores. 62 Zero-order correlations between the father in the home variable and the dependent variables were not significant. Based on these findings, the father in the home variable was not included in the multiple regression analyses. Therefore, the hypothesis was not supported in this study. H 7. Fifth grade females who view both men and women performing the some activities in their social environments will have higher occupation, activity, and total liberality scores. HA 7. Fifth grade females who view both men and women performing the some activities in their social environments will have lower occupation, activity, and total liberality scores. Data analyses showed that this variable was positively related to the total, occupation and activity dependent variables. Results suggests that fifth grade females who view men and women doing similar activities tended to have higher total, occupation, and liberality scores. Multiple regression analyses found the variable to be significantly related to total, occupation, and activity liberality when income was controlled. Thus the hypothesis that fifth grade females who view both men and women performing the same activities in their social environment will have higher total, occupation, and activity liberality scores is supported in this study. Conclusion, Only three predictor variables (mothers education, income, and the Child Observation of Activities (COA)) out of seventeen were used in the final regression model of this study. Only two of these predictor variables, income and Child Observation of Activities (COA), were found to be significant predictors in the regression analyses. 63 Therefore, two of the hypotheses from the study were not rejected. Those hypotheses are: H 2. As total income goes up so does the total, occupation, and activity liberality scores of fifth grade females. H7. Fifth grade females who view both men and women performing the some activities in their social environment will have higher total, occupation, and activity liberality scores . CHAPTER V DISCUSSION ‘In this final chapter, the purposes of the study are summarized, the contributions of the study are stated, implications of the study are presented, suggestions for future research in the area of gender roles on African Americans are discussed, and recommendation of the study are given . SummarLQLLhLSmdL Stereotyping is not a new occurrence in society. A gender stereotype, usually considered cognitive, is a set of beliefs; it deals with what men and women are like, and 'it is shared by the members of particular groups (Williams & Best, 1982). The impact of gender role stereotyping reaches into the very core of our daily lives through various processes. Although society is slowly expanding its roles for men and women, same sex stereotyping and knowledge of sex-role stereotypes prevail among very young children (Kuhn, 1978 & Albert, 1988) and older children (Archer, 1984). Young children in different culture settings also seem to hold sex role stereotypes similar to those of adults (Bardwell, Cochran, and Walker, 1986). The gender role attitudes of forty-three mother-daughter pairs in Lansing, Michigan were studied in the Spring of 1997. The social-ecological factors investigated initially as possible predictors of gender role attitudes were; parents’ education, gender attitudes, income, occupation, occupational prestige, parental status, employment status, family structure, religiosity, spirituality, and church attendance. Additional social-ecological 64 65 factors were; the child’s television viewing and observation of activities. The three research instruments used to collect the data were: The Shepard Hess Stereotypes of Occupations and Activities, Child Observation of Activities, and the General Parental Family Questionnaire. The three major purposes of this study were to: 1. Expand the knowledge of gender role stereotyping by focusing on an African American female population; 2. Investigate how perceived social cognitive and ecological factors influence gender role attitudes among African American females; and 3. Modify and assess an instrument designed to test occupational and activity stereotypes. 29329325. Elm The first purpose of the study was to expand the knowledge of gender role stereotyping by focusing on an African American female population. Although a major task, the results of the study show that this purpose was met. Data results with this African American population confirmed past research that children from higher income families were more liberal in their gender role beliefs (Zuckerman & Sayre, 1982 & Hageman & Gladding, 1983). Results also showed that children’s observation of activities was equally influential in gender role attitudes. In other words, among this population, the activities children saw adults performing was just as influential to gender role beliefs as the income of the family. This finding allows us to expand gender role stereotyping research by studying what is actually going on in the social environments of our subjects. W- The second purpose was to investigate how perceived social cognitive and ecological factors influence gender role attitudes among African American females. This study initially studied seventeen social-ecological factors aspossible influential predictors of the gender role attitudes of fifth grade African American females. However, 66 final results of the data analyses found only two social-ecological factors to be influential. The social - ecological factors found to significantly influence the gender role attitudes among the fifth grade African American females were the parents’ income and the child’s observation of activities. Hume}, The third purpose to modify and assess an instrument designed to test occupational and activity stereotypes was met. As stated in Chapter 3, several modifications were made to the original Shepard Hess Stereotypes of Occupations and Activities for statistical purposes. The instrument was also modify to update the terminology of the occupations and activities (see Table 5). Modifications included the obtaining of three scores, as opposed to one total score. Unlike the original instrument, the modified Shepard Hess Stereotypes of Occupations and Activities instrument used in this study included a total, occupational, and activity score (Shepard and Hess, "1975 & Archer, 1984). The activity and occupational items were treated as subscales scores. Another modification included the treatment of scores as interval data, as opposed to ordinal data, as employed by both Shepard and Hess (1975) and Archer (1984). In addition, the deletions of items, a change of the language and the addition of new items were made to the instrument (see Table 5). The final modified instrument had a total of fifty-one items as oppose to forty-four in the original (Shepard and Hess, 1975 & Archer, 1984). Data assessment of the modified instrument found it to have substantial reliability. Reliability coefficients of .94 for the total, .91 for occupation, and .85 for activity were found with this population. A final model of the study shown in Figure 5 attempts to incorporate the importance of both hypotheses to the gender roles of fifth grade females. The income and the COA predictor variables share in explaining the gender roles of the fifth grade subjects. The 67 variables also represent the two major theories of the study. The income predictor variable represents Bronfenbrenner’s Ecological Systems theory and the COA predictor variable represents Bandura’s Social Cognitive theory. In Bronfenbrenner’s Ecological Systems theory the environment is characterized as a set of nested structures, each inside the next. Behavior, therefore, includes not only the environmental influences the developing child comes in contact with on a regular basis, but several more distant layers of society and culture (Steuer, 1994) . In this study, the income variable is viewed as being part of the exosystem, a layer further removed from the child. The income of the parent does not actively interact with the child’s liberality, but the child’s liberality is influenced by it indirectly. The income of the parent can influence the experiences of the child in the environment. Parents .with higher income are able to afford extracurricular school activities, trips, or social events that may allow the child to see men and women in various occupations and activities. The major thrust of Social Cognitive theory by Bandura is that children add to their repertoire of actions by seeing or hearing someone else perform the behavior rather than by overtly carrying out the behavior themselves. The modes of leanring these actions are via modeling, observational, or vicarious learning. The COA variable, which is essentially an observation of activities in the child’s social environment therefore, influences the child’s liberality this way. This theory shows that the child’s observation of activities is a direct and influential aspect of the fifth grade subjects attitudes in the study. 68 Child's Gender Attitudes Child Observation of Activities Social Cognitive Variable Figure 5 - Social-Ecological Model of the Study 69 W This study made two significant contributions. First, the results showed that family income and the observation of activities are related to gender roles. Fifth grade subjects who lived in families with higher incomes were more likely to have liberal gender roles than fifth grade females living in families with lower incomes. Similarly, fifth grade females who observed men and women performing the same tasks were also more likely to have liberal gender roles than females who did not. The income of families also were related to the task distributions in families. Fifth grade females who lived in higher income families were more likely to see men and women performing the same tasks in their social environment than fifth grade females living in families with lower incomes. These fifth grade females were also more likely to have a father figure in the home. The second contribution of the study centers around the measurement tools. They are the modifications of the Shepard Hess Stereotypes of Occupations and Activities measurement tool and the development of the Child Observation of Activities (COA) measurement tool. The modifications made to the Shepard Hess Stereotypes of Occupations and Activities measurement tool had similar findings as the original, and attempted to add to the knowledge of gender-role stereotyping, by identifying where possible stereotypes exists. Because it is a new instrument, more research needs to be conducted with the Child Observation of Activities (COA) measurement tool. The Child Observation of Activities (COA) attempted to add to the knowledge of gender-role stereotyping by focusing on what activities are being performed by men and women in the social environment of children. Results of the COA suggests that females, at least in this study are not seeing enough men and women performing the same tasks in their social environment. A majority of the fifth grade females in this study came from a two-parent home; therefore family structure does not explain these results. The average mean score of the COA was only 6 out of a total 70 score of fourteen. The findings of the COA also show that a majority of the activities are being sex-typed in their social environment (see Table 9). Only three of the activities were seen by a majority of the subjects performed by men and women. Only two-of the activities were seen as being equally performed by men and women or only men. ImuliszatlnnuthLStndx A major implication of this study can be summarized by the adage, “don’t talk the talk, walk the talk.” In other words, although the participating mothers overwhelmingly believed that men and women should perform all the occupations and activities in the Shepard Hess Stereotypes of Occupations and Activities, it is not being displayed in their home. Additionally, they do not appear to expose their daughters to that liberal view in their social environment. If the mothers truly believe in their liberality, then they should either model their belief in the home or provide opportunities for their daughters to see men and women performing the same task. In other words, they should "walk the talk.” Possible reasons for the overwhelming liberal outcome of the participating mothers’ Shepard Hess Stereotypes of Occupations and Activities scores are: mothers may not be revealing their true feelings; mothers may believe in their liberality about occupations and activities, but are not able to act out their belief in their home; and lastly, the fifth grade subjects may not be revealing what in actually going on in their social environment. Another implication is that the income of parents remain an influential entity to family dynamics. Families with higher incomes in this study were most likely to have a male figure in the home; therefore, they are able to expose their children to more opportunities and situations. The families with higher education had higher incomes which influenced their gender attitudes. Lastly, the study implies, that if parents are not performing non-sexist activities nor providing their children the opportunity to view men and women performing non-sexist 71 activities, then the schools should fill that gap W This study involved a sample of only forty-three mother-daughter pairs in the city of Lansing, Michigan during the Spring of 1997. Follow —up studies are needed on a larger sample to see if results would be similar. Follow-up studies are also needed with different age groups and populations of color. In addition, future research could include participating subjects’ teachers’ gender attitudes or the addition of another female or male population to serve as a comparison group. Furthermore, consideration made for the inclusion of fathers or gender-related research which focuses solely on fathers, would extend research from the a predominately maternal research focus. The study should be replicated using samples of children from other geographical regions and schools, such as private, charter, and home schools. BecammsndafinnuLthLSmdrt Based on the research findings of this study, recommendations of the study focuses on three groups most influenced by the results. These groups are the child, the parent, and the school. Wong. Recommendations for the child are: the encouragement of multiple experiences so they can decide what career path the want to take; the participation in various activities in their social environment; and the encouragement of responsible television viewing, which show men and women participating in various activities and occupations. 72 W5. Recommendations for the parent are: provide opportunities for children to see men and women participating in various activities and occupations; monitor the television programs children are viewing in the home; assign similar activities to male and female children in the home; and model various gender-role activities in the home for children to imitate and learn. W. Recommendations for the school are: provide opportunities for gender-role education for parents to incorporate in their home; model various gender-role behaviors and activities in the school for children to imitate and learn; incorporate career awareness programs which show men and women in different occupations; incorporate programs where men and women in various occupations visit the schools; and incorporate mentoring programs for children with adults in various occupations. One such program being used is the “Girls Re-Cast TV” gender equity based media literacy program sponsored by Girls Inc in cooperation with the Center for Media Literacy (“Girls Recast TV,” 1995). Girls Incorporated is an organization which supports girls in thirty-four states. The thrust of this four-part program is for girls ages eleven to fourteen to become more aware of how they are portrayed on television and how such portrayal can affect their self concept. The issues analyzed in this program are gender role stereotyping, sex and power, body image, and gender violence. School systems can also utilize a technique tested by Tozzo (1990) for modifying children’s occupational stereotyping in schools. Tozzo (1990) developed a program in which men and women from a variety of nontraditional occupations visited the classroom, described their work, and taught an activity associated with their career to the children. Results of this program showed that children’s perception of occupations can be modified. In addition, to supporting programs, counselors within school systems can be 73 trained to confront the gender stereotypes that pervade school systems and increase student awareness of societal attitudes regarding gender stereotypes (Bartholomew, 1994). APPENDICES APPENDIX A Lansing School District Consent LANSING SCHOOL DISTRICT Committed to Quality January 29, 1997 Margaret Goodly 107 Human Ecology Michigan State University East Lansing, MI 48824 Dear Ms. Goodly: In regard to the proposed study, "Social-Ecological Parental Factors: Influence on Gender Roles of Fifth Grade African American Females”, the request to conduct the study in the Lansing School District has been approved. The following comments apply to the study: Participation in the study is strictly voluntary. Parent consent forms must be on file at the school prior to any student contact. Please contact me regarding identifying voluntary schools and changes to the parent consent letter. I can be reached at 325 -64 60. Thank you. W/flxuéofl "7 ’ \ Marian Phillips Supervisor MP/mlc cc: Research Review Committee Members 74 APPENDD( B Letter of Consent Dear Participant: I would appreciate greatly your agreement to participate in this project. Please indicate your willingness to participate by signing and dating the lines below. I agree to participate in the research project, " Social-Ecological Parental Factors: Influence on Gender Roles of Fifth Grade African American Females" being conducted by Margaret Goodly, graduate student in the Department of Family and Child Ecology. The purpose of this project is to gain a better understanding of parental factors on the occupational and activity choices of African American females. I agree and understand the following conditions:s (a) Part I of this study will involve completing questionnaires that apply to my family situation, regarding family situations and personal attitudes. (b) Pan II of this study will involve my fifth grade child or guardian completing a questionnaire with the researcher, regarding occupational and activity choices. I understand that the researcher will make clear to my child that he or she may or may not choose to participate by completing the questionnaire; his or her choice to do so is voluntary. If my child should volunteer to participate, I give my permission for him or her to do so. (Please provide name and school of child below) (c) The participation of myself and my child is voluntary, we may refuse to answer any question and we can terminate our participation at any time during the course of the project. (d) All information provided will be confidential, my child and I will remain anonymous in any report of the research findings, and any report of the research findings will be available to me if I would like to review them. (e) Payment of $10 will be given to my family only upon completion of the general parental questionnaire, at least one parent (Mother or Father) questionnaire, and the child questionnaire. Parent Signature Address Date Name of child School of child 75 APPENDIX C LETTER OF INTRODUCTION Dear Parent(s): My name is Margaret Goodly and I am a graduate student at Michigan State University in the Family and Child Development Department. I would like to take this opportunity to invite you to participate in the research study, "Social-Ecological Parental Factors: Influence on Gender Roles of Fifth Grade African American Females ." The purpose of this study is to gain a better understanding of parental factors on the occupational and activity choices of fifth grade African American females. Dr. Harriette McAdoo is my supervising professor for this study. Participation in this study involves two parts. For part I, you will receive questionnaires that asks personal and detailed questions regarding your family situation and personal attitudes. You are asked to please fill out all questionnaires that apply to your specific family situation. The estimate time of these questionnaires will be about thirty nrinutes. Part II, will involve your fifth grade child/guardian completing a questionnaire at her school. Your child will be given the option of completing the questionnaire alone or with the researcher. Your child will also be given the choice of participating. The estimate time of this questionnaire is about ten to fifteen minutes. All the information provide from these questionnaires will be WW. Neither you or your child will be referred to by name in any report. Your involvement is strictly on a volunteer basis and you may decide at any point in time to no longer be involved with this project. This information and the information obtained from others who participate will be used to write a report. This report will be available to you if you would like to review them. From your participation in this study your family may gain a greater understanding of how the family influences a child's occupational and activity choices. W W. If you would like to participate in this study, please complete the consent form that rs enclosed with this letter and return them to me in the envelope provided. Questionaires will be mailed to you upon delivery of the consent form. I understand that your time is valuable, sol would like to thank you in advance for your participation, as well as, encouraging your promptrress with this study. If you have any questions or concerns, you may call me at 432-3328. Sincerely, Margaret Goodly Harriette Pipes McAdoo Michigan State University Michigan State University Department of Family and Child Ecology Department of Family and Child Ecology 107 Human Ecology East Lansing, MI 48823 76 APPENDIX D RESEARCH DIRECTIONS Dear Dr. Harriette McAdoo and I would like to take this opportunity to thank you for your interest in the research study, "Social-Ecological Parental Factors: Influence on Gender Roles of Fifth Grade African American Females ." The purpose of this study is to gain a better understanding of parental factors on the occupational and activity choices of fifth grade African American females. As stated in the introduction letter, participation in this study involves two parts. For part I, you will receive questionnaires that asks personal and detailed questions regarding your family situation and personal attitudes. You are asked to please fill out all questionnaires that apply to your specific family situation. The estimate time of these questionnaires will be about thirty minutes. Part II, will involve your fifth grade child/dependent completing a questionnaire at her school. Your child will be given the option of completing the questionnaire alone or with the researcher. Your child will also be given the choice of participating. The estimate time of this questionnaire is about ten to fifteen minutes. Included in this packet are a general parental questionnaire and a mother and father questionnaire. Questionnaires that need to be completed are the general parental questionnaire and either the mother or father questionnaire. It is not necessary that the family answer both the mother and father questionnaire. However, you are encourage to fill out both the mother and father questionnaire if your personal family situation allows participation by both parents or guardian. Information from the mother and father adds to the richness of the study. Please complete the questionnaires and return them to me in the envelope provided. Your payment of $10 will be sent to your family upon completion of the general questionnaire, at least one parent (Mother or Father) questionnaire, and the child questionnaire. Thank you very much for your interest and participation in this research study. Sincerely, Margaret Goodly Harriette Pipes McAdoo, Ph.D. Michigan State University Michigan State University Depart. of Family and Child Ecology Depart. of Family and Child Ecology 107 Human Ecology East Lansing, MI 48823/ Phone: 432-3328 77 APPENDD( E Shepard Hess Stereotypes of Occupations and Activities Direction: Please indicate for each item who should perform the occupation or activity (circle only one choice) 1. Who should W5? only women men and women only men 2. Who should muslin? only women men and women only men 3. Who should Mane? only women men and women only men 4. Who should clamthahathrmm? only women men and women only men 5. Who should lead a Whom? only women men and women only men , 6. Who should W? only women men and women only men 7. Who should am? only women men and women only men 8. Who should Mm? only women men and women only men 9. Who should Wanamaker? only women men and women only men 10. Who should W1? only women men and women only men 11. Who should sellmrfmne? only women men and women only men 12. Who should W? only women men and women only men 13. Who should m? only women men and women only men 14. Who should gofishing? only women men and women only men 78 15. 16. 17. 18. . Who should W? Who should Wm? only women Who should WWW? only women Who should MW? only women Who should W? only women . Who should Wm? only women . Who should Wm? only women only women . Who should WW? only women . Who should WM? only women . Who should W? only women . Who should W? only women 79 men men men men men men men men and women and women and women and women and women and women and women and women and women and women and women only men only men only men only men only men only men only men only men only men only men only men . Who should gamma-an? only women . Who should Wager? only women . Who should Mm? only women only women . Who should W? only women men men . Who should WW? and women and women and women and women and women only men only men only men only men only men . Who should Wis? only women and women only men 80 32. Who should W? only women men and women only men 33. Who should W? only women men and women only men 34. Who should Mamba? only women men and women only men 35. Who should beam? only women men and women only men 36. Who should WW? only women men and women only men 37. Who should W? only women men and women only men 38. Who should MM? only women men and women only men 39. Who should delimmafl? only women men and women only men 40. Who should MW? only women men and women only men 41. Who should Wk? only women men and women only men 42. Who should W? only women men and women only men 43. Who should WW? only women men and women only men 44. Who should W? only women men and women only men 45. Who should W? only women men and women only men 46. Who should W? only women men and women only men 47. Who should W? only women men and women only men 48. Who should 1211mm? only women men and women only men 49. Who should 311W? only women 50. Who should W? only women 51. Who should W? only women 81 men and women men and women men and women only men only men only men APPENDIX F Child Observation of Activities Part 2: For each item circle WW do the following: 1. iron clothes only women men and women only men 2. sew only women men and women only men 3. clean the bathroom only women men and women only men 4. vacuum the house only women men and women only men 5. laundry only women men and women only men 6. move furniture only women men and women only men 7. cook dinner only women men and women only men 8. house repairs only women men and women only men 9. taking care of young children only women men and women only men 10. wash dishes only women men and women only men 11. shopping for groceries only women men and women only men 12. taking out the garbage only women men and women only men 13. household dusting only women men and women only men 14. sweep the floor only women men and women only men 82 A. List four of your favorite television programs? B. Who are your favorite actors/singers on television? C. Who are your favorite actresses/singers on television? D. How much television do you watch in a day (hours or minutes)? APPENDIX G GENERAL PARENTAL FAMILY QUESTIONNAIRE ID Date Name of Child Child's Age School Directions: Please circle and write your response choice. A. Mother's age B. Father's age C. How many children are in your household? D. What is your relationship to the child? 1. Who is the mother figure in the household? Birth/biological mother..l Adoptive mother ..... 2 Stepmother ......... 3 Foster mother ........ 4 No mother figure in the household. . . . 5 Other (specify) 6 2. What is your current marital situation? Please circle all that apply. Never married ...... l Divorced ....... 5 Married .......... 2 Widow/er ....... 6 Separated ......... 3 Remarried ....... 4 2a. What is your current family structure? Two parent, married ...... 1 Living with partner, not married ...... 2 One parent home, female headed ....... 3 One parent home, male headed ........ 4 84 Old 85 3. Who is the father figure in the household? Birth/biological father.. 1 Adoptive father ...... 2 Stepfather ......... 3 Foster father ........ 4 No father figure in the household. . . 5 Other (specify) 6 4. How do you describe your race/ethnicity? Black/African American ........ 1 Other ..................... 2 (specify other ) 5. What is the mother/mother figure main job outside the home? ____No mother/mother figure _No main job outside the home 6. What is the mother/mother figure current work situation? 1. work full-time 2. work part-time 3. unemployed 4. homemaker 7. What is the father/father figure main job outside the home? _No father/father figure _No main job outside the home. 10. What is the father/father figure current work situation? 1. work full-time 2. work part-time 3. unemployed 4. homemaker 11. What is the highest grade (or number of years) of education completed of the following persons... A. Mother/Mother figure B. Father/Father figure 86 12. List any university degree(s) of the following persons... A. Mother/Mother figure B. Father/Father figure 14. What is your current total family income including all sources? Please round to the nearest thousand 15. What is your family's religious affiliation? l._no religious affiliation 16. In a given month how many times does your 5th grade daughter/dependent attend religious services with a parent? (Which parent most often? ) 17. How religious are you? (circle one) 18. How spiritual are you? (circle E one) 1. Very religious 1. Very spiritual 2. Somewhat religious 2. Somewhat spiritual 3. A little religious 3. A little spiritual 4. Not religious 4. Not spiritual APPENDIX H LANSING SCHOOL DISTRICT Ethnic Report, October 7, 1996 Unofficial headcound report generated from Pupil Accounting Worksheets SCHOOL STUDENTS BY GENDER BY ETHNICITY Total ROUNDED African American PERCENT m / f 87 APPENDIX I Crosstab Table: Family Structure by Mother’s Employment NEWFAM new family structure by MEMPLOY mother's employment MEMPLOY Page 1 of 1 Count Row Pct unemploy employed Col Pct ed Row Tot Pct 0 1 Total NEWFAM O 7 9 16 single parent fa 43.8 56.3 37.2 53.8 30.0 16.3 20.9 1 6 21 27 two parent famil 22.2 77.8 62.8 46.2 70.0 14.0 48.8 Column 13 30 43 Total 30.2 69.8 100.0 Number of Missing Observations: 0 88 REFERENCES REFERENCES Albert, A. A. (1988). Children's gender-role stereotypes. A sociological investigation of psychological models. W, 1, 184-210. Archer, C. J. (1984). Children's attitudes toward sex-role division in adult occupational roles. Molesjfl, 1-10. Bandura, A. (1986). ' ' Englewood Cliffs, NJ: Prentice- Hall, Inc. Barak, A., Feldman, S, & Noy, A. (1991). Traditionality of children's interests as related to their parents' gender stereotypes and traditionality of occupations. Moles, 24, 511-524. Bardwell, J. R., Cochran, S. W., and Walker, S. (1986). Relationship of parental education, race, and gender to sex role stereotyping in five-year-old kindergartners. W. 15. 275-81. Barnes, A.S. 91983). 1 c s wives: as e ' s ' ' In C.E. Obudho (Ed.), Black marriage and family therapy (pp. 55-73). Westport, CT: Greenwood Press. Bartholomew, C. G. (1994). Gender equity: Suggestions for broadening career options of female students. WM, 245-255. Beere. C. A. (1990). WES (1313- 227-228). New York: Greenwood Press. Billings, S. K. (1992). A, ' - (Report No.85). Fort Hays State University. (ERIC Document Reproduction Service No. ED352119). 89 9O Bronfenbrenner, U. (1979). v W- Massachusetts: Harvard University Press. Bussey, K. and Bandura A. (1984). Influence of gender constancy and social power on sex-linked modeling WW& 1292- 1302. Cazenave, N. A. (1983). Black male-black female relationships: The perceptions of 155 middle-class black men. MW 341- 350. Costello, C. and Krimgold, B. K. (Ed). (1996). W m6. W. W. Norton: New York Darden, E. C. (1992). Blended families: A decade review, 1979 to 1990. Family Therapy. 12. 25-31. Davenport, D. S. and Yurich, J. M. (1991). Multicultural gender issues. Journals); Cormalinaét W 64-71- Eccles, J. 5., Jacobs, J. E., and Harold, R. D. (1990). Gender role stereotypes, expectancy effects, and parents' socialization of gender differences. 166mm 5mm] [3,5665 46, 183- 202. Francis, L. J. (1996). Religion and gender orientation. Egrsonality a, Individual W 119- 121 “Girls re- -cast tv” builds on girls knowledge of tv and media (media-literacy program). (1995 Spring) W6 Gettys, L. D. and Cann A. (1981). Children's perceptions of occupational sex stereotypes. SEW. 301-308. Hageman, M. B. and Gladding, S. T. (1983). The art of career exploration: Occupational sex- -role stereotyping among elementary school children. W661 We 17 280-87 Hensley, K. K. & Borges, M. A. (1981). Sex role stereotyping and sex role norms: A comparison of elementary and college age students. Esygbglggy 6f Women Quanstlxi. (4). 543-554. 91 Hodge, R. W. Siegel, P. M., and Rossi, P. H. (1964). Occupational prestige in the united states, 1925- 63. MW, 70, 286- 302. Jarrnulowski, V. (1985, February). The blended family: Who are they? We, 33-34. Jensen, L. C. (1993). Family values, religiosity, and gender. W3, 429—430. a. Juni, S. Rahamin, E. L., and Brannon, R. (1985). Sex role development as a function of parent models and oedipal fixation. WM, 146, 89- 99. Kuhn, D. (1978). Sex-role concepts of two and three year olds. Wm, 49, 445-451. ."- Logan, L. (1988). Gender, family composition and sex-role stereotyping by young children. (Report No. 56 ). ERIC Document Reproduction Service No. ED290564). Lytton, H. and Romney, D. M. (1991). Parents' differential socialization of boys and girls: A meta analysis. W, 267- 297. MacKinnon, C. E, Stoneman, Z., & Brody, G. H. (1984). The impact of maternal employment and family form on children's sex—role stereotypes and mothers' traditional attitudes. W8, (1), 51-60. McAdoo, H.P.(Ed.). (1991). Wigs (3rd ed.). Newbury Park: Sage Morgan, M. Y. (1987). The impact of religion on gender-role attitudes. mm ngen Quamdxull. 301-309. Nadelman, L. (1974). Sex identity in american children: Memory, knowledge, and preference tests. Wm, 10, 413-17. Nelson, C. and Keith, J. (1990). Comparisons of female and male early adolescent sex role attitude and behavior development. Adglesceng, 24, 183-204. O'Keefe, E.S.C. and Hyde, J. S. (1983). The development of occupational sex-role stereotypes: The effects of gender stability and age. Sex R6165, 4, 481-492. 92 Pellett, T. ”(1993). Relationship between children's and parents stereotyping of physical activities WW 1283- 1289 Perloff, R. M. (1982). Mass media and sex- typing: Research perspectives and policy implications WW5. 265 273 Peters M F (1997) Hareminaefleungebildieninfilaekfamilles InH P McAdoo (Ed.,) Black families (pp. 167- 182). Newbury Park: Sage :- Reid, P. (1985). Sex-role socialization of Black children: A review of theory, family, and media influences MW 211 -212 Reis, H. T. and Wright S. (1982). Knowledge of sex-role stereotypes in children aged 3 to 5. W, 8, 1049-1056. '.‘ 1!. . \' Signorielli, N. (1990). Children, television, and gender roles: Messages and impact. loumfloLAdolessenLHealthfiareJl. 50-58. Signorielli, N. & Lears, M. (1992). Children, television, and conceptions about chores: Attitudes and behaviors. W1, 157-170. Shepard, W. O. & Hess, D. T. (1975). Attitudes in four age groups toward sex role division in adult occupations and activities. Wm, 6(1), 27- 39. Steuer, F. B. (1994).Ih§_E§)LQhQIQ£1£aLdfl£lQnm§m_Qf£hfldmn- CalifomiazBrooks/Cole. Thomas, R. M. (1992). mmmmmmmnmmm California: Wadsworth. Tozzo, S. G. ( 1990) Playing nurse and playing cop: Do they change children’s perceptions of sex-role stereotypes Wm 4, 123- 129. 9 Tremaine, L. S., Schau, C. G., and Busch J. W. (1982). Children's occupational sex-typing. W. 8, 691-709. 93 Tuck, B. (1994). Adolescents' attitude toward gender roles within work and its relationship to gender, personality type, and parental occupation. M, 547-559. Vasta,R.(l992). '. - ' ° - isms. London: Jessica Kingsley Willetts-Bloom, M. C. and Nock, S. L. (1994). The influence of maternal employment on gender role attitudes of ment and women. W 371-389. Williams, E. J. and Best, D. L. (1982). W. California: Sage Publications, Inc. Zeff, S. (1982). A cross-cultural study of mexican american, black american, and white american women at a large urban university. WW W2.) 245- 261 Zuckerman, D. M. and Sayre DH. (1982). Cultural sex role expectations and children's sex-role concepts. W, 8, 853-62. lrfl'I-‘. .5.-"'. -.