LIABR AIRES \\\\\\\\\\\2\\\\\l a, , Millillllllllll This is to certify that the thesis entitled Socioeconomic Status and Self-Esteem Among Black Adults presented by Shu-Yao Hsu has been accepted towards fulfillment of the requirements for Master's degree in Sociology Major professor Date JM /5L/ 7y¢ 0.7639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Mlchigan State University PLACE N RETURN BOX to remove this checkout from you: record. TO AVOID FINES return on or before date duo. DATE DUE DATE DUE DATE DUE Wain-9.1 i -_.__._ i___—___ SOCIOECONOMIC STATUS AND SELF-ESTEEM AMONG BLACK ADULTS By Shu-Yao Hsu A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Sociology and Urban Studies 1 994 ABSTRACT SOCIOECONOMIC STATUS AND SELF —ESTEEM AMONG BLACK ADULTS By Shu-Yao Hsu The self-perception theory is used to examine the roles of personal and parental socioeconomic status on personal self-esteem among black adults. This study differentiates the effects of personal and parental socioeconomic status on. personal self—esteem. The pivotal distinction concerns whether various levels of personal self-esteem result fiom personal more than parental socioeconomic status, or result from parental more than personal socioeconomic status. /Multiple regression: andipath analysis; are used to explain the relationship between socioeconomic status and self-esteem, to test whether self-perception theory is applicable to black adults, and to understand patterns of self-esteem. This study concludes that personal socioeconomic status is more strongly associated with personal self-esteem than is parental socioeconomic status. The hypothesis of self—perception theory is supported. ACKNOWLEDGMENTS A number of people have helped me in my thesis and I cannot thank them all here, but I would like to express my most grateful thanks to my adviser, Dr. Clifford L. Broman, for his encouragement, detailed instructions, and patience. I would like to thank Dr. Janet Bokemeier, Dr. Marvel Lang, and Dr. Stan Kaplowitz for serving on my thesis committee. Finally, I would like to express my appreciation and gratefulness to my family, especially my husband, Chun-Hao Li, for their enthusiastic support and care. iii TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES Introduction . Self-Esteem. . . . Defining Self-Esteem . Studies m Self-Esteem among Blacks . Self-Perception Theory . Research Hypotheses . Data and Method . Data . . Measures Analyses. Results Conclusion . List of References . iv Page 12 12 12 14 16 21 27 LIST OF TABLES Table 1. Descriptive Characteristics of the Sample. 2. Regression of Self-Esteem on Socioeconomic Variables Page 23 24 LIST OF FIGURES Figure 1. Path Model of SES and Self-Esteem . . . . 2. Alternative Path Model of SES and Self-Esteem . Page 25 26 CHAPTER I INTRODUCTION Socioeconomic status (SES) has been argued as a crucial factor in predicting self-esteem (Demo and Savin-Williams, 1983; Gordon, 1969; see Porter and Washington, 1979; Poter, 1971; Rosenberg and Pearlin, 197 8; Rosenberg and Simmons, 1971; Samuels, 1973; Wiltfang and Scarbecz, 1990; Yancey, Rigsby and McCarthy 1973). The effects of SES on self- esteem among children, adolescents, or adults have been examined in a number of studies (Demo. and Savin-Williams, 1983; Porter, 1971; Rosenberg and Pearlin, 1978; Rosenberg and Simmons, 1971). Rosenberg and Pearlin (1978) found that the relationship between SES and self-esteem was strongest in the adult group than was that for children or that for adolescents. This is because SES is one particular kind of social experience for adults, but this kind of social experience is not endowed with the same meaning for children or for adolescents. SES is particularly important and salient for adults' self-esteem. SES can be observed as the outcome of one's behavior to represent one's social position and relative prestige, and serve as the standards of evaluating how worthy one is as a person (Rosenberg, 197 9; Rosenberg and Pearlin, 197 8; Wiltfang and Scarbecz, 1990; Yancey, Rigsby and McCarthy 1973). The status one achieves is regarded as the outcome of one's effort which is a reflection of self-esteem. Very few studies focus on SES and self-esteem among black adults. I 2 Research literature on black's self-esteem is very controversial (Broman, Jackson, and Neighbors, 1989; Cross, 1985; Jackson, McCullough, and Gurin, 1981; Simmons, 1978), more understanding is necessary. When dealing with the causal relation between SES and self-esteem, it should be distinguished, whose SES affects one's self-esteem? Is it parents' SES, one's own SES, or both? The purposes of this study are to examine the relationships among parental SES, personal SES, and self-esteem, to test the applicability of self-perception theory to the black adult group, and to provide an interpretation for patterns of black adult self-esteem. It also needs to be understood how strong parental SES impacts on personal SES and self- esteem among black adults, and whether SES can be handed down from parents to children. We begin by defining the concept of self-esteem; then discussing the findings of previous research on blacks' self-esteem in various time periods and the reasons why a further study on SES and self-esteem for blacks is necessary. Self-perception theory is adopted to explain the causal relation between SES and self-esteem among black adults. An empirical examination is performed to test whether self-perception theory is applicable to the black adult group. CHAPTER H SELF -E STEEM Defining Self-Esteem Self-esteem is defined as self-worth; that is, how individuals feel about themselves as a person of worth (Rosenberg, 197 9; Rosenberg and Simmons, 1971). Self-worth does not mean feelings of superiority or overweening pride, but it means feelings of adequacy or importance as a person. Individuals who have high self-esteem may not feel superior than others, but they consider themselves worthy and accept themselves as persons regardless of whether they are good or bad. Self-esteem is the most important aspect of self-concept (Rosenberg, 197 9). Individuals with high self-esteem have more stable self-concept than those with lower self-esteem (Baumgardner, 1990; Campbell, 1990). Self- esteem also serves a buffering function to relieve one's anxiety (Greenberg, Solomon, Pyszczynski, Rosenblatt, Burling, Lyon, Simon, and Pinel, 1992). Since self-esteem plays such a crucial role for a person, it is important to explore the critical factors which cause various levels of self-esteem among blacks. In next section, we will briefly review the research literature in black‘s self-esteem and discuss the limitations of the past research. Studies in Self-Esteem among Blacks Few studies in self-esteem among blacks were conducted from 1939 to 1960 (Cross, 1985). Most studies in this period focused on black group 3 4 identification, and the research findings reported a negative black identity trend (Cross, 1985; Jackson, McCullough, and Gurin, 1981). The negative black identity trend was considered as a result of the rejection to the oppressed social status of black group, and group rejection was automatically linked to low self-esteem. The empirical studies in the period were mostly around choices and reactions to black and white dolls among young children (see Jackson, et al., 1981; Clark and Clark, 1947; Radke and Trager, 1950; Goodman, 1952; Landreth and Johnson, 1953; Stevenson and Stewart, 195 8). The behavior of choosing a white rather than a black doll by a majority of black children was viewed as a reflection of group rejection as well as self-rejection. The limitation of doll studies is that the observation of choices to black and white dolls only represented racial group identification or personal preference to dolls, but not self-esteem. The argument that blacks suffered from low self-esteem is advanced in the interpretation of these findings. Self-esteem was gradually given more attention after the Civil Rights Movement because it was considered to reflect the change of blacks' perception from self—rejection to self-worth, which was thought to be affected by the movement. It has been criticized that racial group identification and self-esteem were not separately measured in most of the earlier studies (Cross, 1985; Jackson et a1. 1981; Porter and Washington, 1979). However, this sparked a great controversy in the research literature concerning self-esteem among blacks (Broman, Jackson, and Neighbors, 1989; Cross, 1985; Jackson et a1. 1981; Simmons, 1978). The issue that attracted most attention and. was most frequently investigated is the comparison of black-white differences in self-esteem (Bachman and O'Malley, 1984; Heiss and Owens, 1972; Porter, 1971; Rosenberg and 5 Simmons, 1971; Samuels, 1973; Yancey et a1. 1973). Blacks are much more economically disadvantaged as compared to whites due to structural inequality (Hacker, 1992). The economic disadvantage has led to the unequal bases between blacks and whites in terms of SES. According to the inequality between black and white SES, two issues should be raised: first, whether the comparisons of black-white differences contribute to theories of self—esteem for blacks; second, whether it is appropriate to assume whites as the comparative group for blacks (Cross, 1985; Porter and Washington, 197 9). This is because the values which provide blacks with criteria to evaluate objects can be different from those for whites. One's comparative group can be defined by race, sex, and class; the group can change over time. The key point is that the comparative group should be meaningful to individuals. Hence a comparison to other blacks may be more meaningful than that to whites for blacks. Self is not important if individuals do not live in human society. However, once individuals are born, they are parts of the society and are expected to follow the social rules. In order to enhance self-esteem, individuals may behave the way defined by the society. It has been argued that social comparison is an important way individuals use to learn about themselves by comparing themselves to others based on similar socioeconomic background (Pettigrew, 1967; Rosenberg and Pearlin, 197 8; Yancey et al. 197 3). The key point is with whom a person actually compares his SES. A number of studies in personal self-esteem among blacks have stressed that the black group itself is utilized as a reference group for personal comparison rather than the white (see Porter and Washington, 1979, for review). The frame of reference provided by the black group may offer an alternative and more achievable criteria of success for blacks; therefore, 6 the assumption that blacks utilize whites as a reference group in social comparisons is a problem. The Rosenberg and Simmons study (1971) indicated that black children demonstrated equal to or higher levels of self- esteem than did whites even though the black children showed an explicit preference for light skin. Their conclusion was that blacks were very likely to use other blacks as the comparative others rather than to use whites. Therefore, the present study focuses on a group of black adults, and the major concern is what the effects of SES are on self-esteem in this group. CHAPTER III SELF-PERCEPTION THEORY Self-perception theory has been used to explain why SES is an important predictor of one's self-esteem because SES is regarded as the outcome of one's own behavior which serves as standards of self-evaluation (Rosenberg and Pearlin, 1978; Demo and Savin-Williams, 1983; Wiltfang and Scarbecz, 1990). Wade and colleagues (1989) compared black and white adolescent self-esteem. They found that the variable "seeing things happening as one's own doing" was a significant predictor of self-esteem for black adolescents, but not for white adolescents. Therefore, personal accomplishment can be a critical determinant to self-esteem, especially for blacks. Self-perception theory was one of the four important theories-the others are social comparison, reflected appraisal, and psychological centrality-used by Rosenberg and Pearlin (l 97 8) to emphasize the importance of personal SES to self-esteem, and to explain why the association between SES and self-esteem was weak among children, but modest among adolescents; and moderate among adults. They argued that all of the four theories could be equally applied to children and adults, but the role of SES was particularly important to predict self-esteem for adults. Their study was applauded as a valuable model for investigation on the linkage between social structure and personality (House, 1981; also see Wiltfang and Scarbecz, 1990). SES was most relevant to adults' self-esteem 7 8 because it could lead to different social experiences for adults and make them much more "class-conscious" than children and adolescents. Therefore, SES is very crucial for adults to evaluate how worthy they are as a person. Rosenberg and Pearlin applied Bern's self-perception theory (1965, 1967) to explain the psychological impacts of social structural facts to personal feelings of self worth. Bem proposed this theory to contest dissonance theory's fundamental motivational assumptions. He (1965) argued that individuals must rely on their own behavior as a guide to understand and interpret their own inner states such as attitude and intentions. When persons estimated their own behavior, they were fimctionally at least in the same position as an outside observer. Hence, individuals observed their own behavior just like what others did. In other words, individuals got to know about how to evaluate themselves by means of observing and learning from how outside observers treated them. According to self-perception theory, we come to know ourselves by observing our own behavior. We learn about ourselves as others learn about us. What we have done is much more important than what our parents have done. Therefore, the feelings of self worth primarily stem from our own accomplishment, neither from our fathers' nor our mothers‘. As Rosenberg and Pearlin suggested: "Our self-regard depends primarily on what we have done, secondarily on what our ego—extensions have done ...... Children’s self- esteem ..... is probably just as dependent as adult self-esteem on achievement; but this achievement is their own, not their parents'." (1978, p. 66; also quoted by Wiltfang and Scarbecz, 1990, p.176). For adults, SES is regarded as the outcome of our efforts and SES is achieved and earned. What our parents have achieved is not as important as what we have achieved in terms of the effects on our feelings of self worth. Although parents‘ status may 9 assist one to get ahead, it is not equal to one's own success, and parents' status may not directly affect one’s self-esteem. The reason why SES affects adult self-esteem is that the status individuals achieve is regarded as the outcome of their efforts. Following this reason, two hypotheses are drawn. The first one is that the higher one's education, the higher one’s self-esteem. The second one is that the higher one's family income, the higher one‘s self- esteem. Respondents' education in the first hypothesis and family income in the second hypothesis are used to reflect personal SES. More details concerning measures will be provided in the later section. ‘ Rosenberg and Pearlin concluded that their findings supported self- perception theory. Parental SES did not function as a strong factor in the levels of personal self-esteem because SES was achieved rather than ascribed for adults. Wiltfang and Scarbecz (1990) extended the Rosenberg and Pearlin study and focused on the examination of self-perception theory for explaining the relationships among parental SES, adolescents' own achievements, and adolescents' self-esteem. They argued that Rosenberg and Peariin's conclusion of self-perception theory was premature because Rosenberg and Pearlin failed to compare the impacts of parental SES and subjects‘ achievements on self-esteem empirically. However, self-perception was supported in Wiltfang and Scarbecz's research. They concluded that the effects of adolescents' achievements on self-esteem were stronger than those of parental SES. Although findings in the Wiltfang and Scarbecz study supported the self-perception hypothesis, a critical problem noticed in the two previous studies is that their data were frequently from samples of children and adolescents. These samples were not representative of either adults and/or blacks; therefore, the generalizations derived from samples of children and 10 adolescents to the adult black population deserve further testing. In this present study, the sample selected from the population of black adults is different from those in the two former studies. An examination which compares the effects of personal SES on self-esteem to that of parental SES on self-esteem among black adults is performed. Another two hypotheses are drawn: (1) the relationship between fathers’ education and self-esteem is weaker than that between respondent's education and self-esteem; (2) the relationship between mothers' education and self-esteem is weaker than that between respondent‘s education and self—esteem. Whether self-perception theory is useful for the study of black self-esteem among adults is studied here. We seek to discover whether parental SES and/or personal SES have an effect on black adult's self-esteem. Some research emphasizes that age is a'critical intervening variable in the relationship between SES and self-esteem (Demo and Savin-Williarns, 1983; Rosenberg and Pearlin, 1978; Wiltfang and Scarbecz, 1990). Adults are aware of class more than children and adolescents due to different social experiences. The research in this area pays much attention to the comparison among the three age groups, and whether variations exist in each age group. Therefore, we examine the effect of age on self—esteem in the black adult group. In addition, although we expect that the levels of self-esteem are directly affected by the personal SES variables, direct effects from parental SES variables to personal SES variables and indirect effects from parental variables to self-esteem should not be neglected. Therefore, the examination of the pattern of relationships among parental SES variables, personal SES variables, and self-esteem is performed. ’11 Research Hypotheses Four hypotheses linking parental and personal SES to self-esteem among black adults are generated. Specifically, these hypotheses are: 1. The relationship between fathers' education and self-esteem is weaker than that between respondent's education and self-esteem. 2. The relationship between mothers' education and self-esteem is weaker than that between respondent's education and self-esteem. 3. The higher one's education, the higher one's self-esteem. , 4. The higher one's family income, the higher one's self-esteem. CHAPTER IV DATA AND METHOD Data The data used in this study were collected by the National Survey of Black Americans (NSBA) in 1979-1980. The NSBA is a cross-sectional study. The sample including 2107 black Americans at ages of 18 years and older was selected by a multistage sarnplin g procedure by the Survey Research Center at the University of Michigan. Face to face household interviews were conducted with respondents in their homes by all black professional interviewers. The response rate is approximately 69%. Taylor (1986) performed a comparison of the NSBA sample with United States Census Bureau data. The educational level of the NSBA samme is similar to Census Bureau data, but family income of the NSBA is , slightly higher that of Census Bureau data. The NSBA respondents' demographic profile is available in Broman, Neighbors, and Jackson (1988). Further information may also be found in Jackson, Tucker, and Gurin (1987). Measures . SES consists of two dimensionsfémental SES variabl‘éflandyersonal SES variables: Parental SES is operationalized as fathers' years of education and mothers'\“years of education. The traditional measures of parental SES usually include father’s education, occupation, and income. We use both father's and mother's years of education instead of the traditional SES 12 13 measures because of the following reasons. First, the NSBA does provide the data, father‘s current occupation and income, but the information such as how long respondent's father has worked on the current job and what kind of job he did before the current one is deficient. The stability of father’s occupation is relevant to the levels of father‘s income as well as its effects on child‘s self-esteem. If one‘s father has worked on the current position for only one year, the effect of father‘s occupation on personal self-esteem may be very small. Because of the lack of father's occupation history, the use of traditional SES measures is not proper here. Second, a mother plays a very important role in a family, and the effect of mother‘s education on personal self-esteem may not be less than that of father‘s among blacks. Third, the use of modern racism as social and political mechanism is a critical barrier to black‘s economic status (McConahay, 1986). A black with high level of education is very likely to get lower wages or lower skilled jobs compared to a white with the same educational level. Hence the measures which can really reflect black parental SES are father’s and mother‘s education, but not their income and occupation. We use father‘s education and mother‘s education instead of traditional measures of SES. The question asked about fathers' education was: "How many years of school did your father complete?“ The question asked about mothers‘ education was the same as that asked about fathers‘ education. Both fathers‘ and mothers‘ levels of educational attainment were recorded in 18 categories from 0 through 17 or more. Respondents' years of education and family income are included as indicators of personal SES. Respondents‘ education was coded as for the fathers‘ and mothers‘. We use education and family income here because education is viewed as a prestige indicator (Glenn, 1963) and family income 14 is more meaningful to individuals among black Americans. The measures of self-esteem are constructed by a six-item scale. Two items are from Rosenberg‘s (1979) self-esteem scale: “I feel that I‘m a person of worth," and “I feel I do not have much to be proud of. “ Two were from Bachman and Johnston (197 8): “I feel that I can‘t do anything right,“ and “I feel that my life is not very useful." The additional two questions asked were: “I am a useful person to have around,“ and “As a person I do a good job these days.“ There were three items stated negatively; the negative items are receded before computing the average scores of the scale so that all items were in the same and positive direction. These items were coded (4) almost always true, (3) often true, (2) not often true, or (1) never true. The range of average scores on all items for each respondent is from 1.67 to 4.0 with a mean of 3.53. The reliability coefficient Alpha for this scale is .66. The higher the value of the reliability coefficient, the more reliable is the scale. For a six-item, the reliability of this scale can be sufficient. One way to improve reliability is to increase the items of the scale (Bohmstedt, 1983). A reliability coefficient of .95 is considered as the desirable standard (Nunnally, 197 8), but increasing reliability as high requires other long number of items in the scale and needs to cost a great deal of time and fimds. For the purpose of saving time and frmds, the modest reliability of .66 for our self-esteem scale which consists of six items can be considered to suffice. Analyses LMultiple regression analyses are used to analyze the relationships between the predictor variables-age, sex, father‘s education, mother‘s ’fiv-_...~_~_n--—-_»..,__> .- ~~..—----- education, respondent‘s education, family income—and self-esteem (see Table — 2). A series of regression equations were estimated. All equations included 15 demographic variables—age and sex. Although significant relationships between measures of SES and self- esteem have been well documented, it is important to understand more about the causal dynamics underlying these relationships and to find the plausrbility of specific causal patterns between these measures. A causal model, using path analyses, is estimated to ascertain the direct and indirect effects of parental SES on self—esteem and on personal SES, as well as the effects of personal SES on self-esteem (see Figure 1 and Figure 2). The analyses can help us to interpret the patterns of causation between the measures of SES and self-esteem and to examine the plausibility of causal patterns. By using path analyses, all variables are treated in standard score form, 2 score. Because path analysis involves in interpretation of multiple independent variables which are measured on different units, standardized coefficients allow us to compare the relative effect of each independent variable on the dependent variable (Pedhazur, 1982). CHAPTER V RESULTS . The data analysis was conducted as an attempt to test the self- perception theory. Table 1 presents descriptive characteristics of the sample used in this study. The bivariate analysis demonstrates that personal SES variables, education and family income, are significant predictors of self- esteem among blacks, while neither fathers‘ education nor mothers‘ education significantly predict self-esteem. In general, the higher educated people have higher levels of self-esteem. The levels of self-esteem tend to be higher while family income increases. The results of the bivariate analyses are merely suggestive. Controlling for the demographic variables, age and sex, is needed to examine the relationship between SES and self—esteem. Table 2 presents the results of two regression equations of self-esteem: the first equation consists of the measures of fathers‘ education, mothers‘ education, respondents‘ education, and family income using age and sex as control variables (see the left three columns of Table 2); the second includes those in the first equation except father’s and mother’s education (see the right three columns of Table 2). In Table 2, the significant predictors are respondents‘ education, family income, and age for both equations. The coeflicient of determination only slightly decreases after father‘s and mother‘s education are deleted. The deletion of father’s and mother’s education from the first regression equation does not significantly affect the results. Hypothesis 3 is supported; that is, the higher educated people have higher 16 17 levels of self-esteem. This finding is consistent with the Yancey, Rigsby, and McCarthy study (1 97 3) that respondent‘s education has strong effects on self-esteem. The hypothesis that the higher one‘s family income, the higher his self-esteem is also supported by these data. The relationship between age and self-esteem is significant and positive. The finding confirms Rosenberg and Peariin‘s finding and indicates the importance of age to self-esteem among the black adult group. Age is crucial to self-esteem not only among different age group but also within groups. Based on the results of two tables, fathers‘ and mothers‘ education cannot significantly predict respondents‘ self-esteem. In order to understand the patterns of relationships among parental SES, personal SES, and self-esteem, a path model is examined. The analyses in the path model concerning fathers‘ education, mothers‘ education, respondent‘s education, family income, and self-esteem address the issue of whether parental SES variables directly affect self—esteem or their effects are direct to personal SES variables but indirect to self-esteem. This model allows for examining the effects of parental SES on personal SES and self-esteem, as well as the effects of personal SES on self-esteem. Figure 1 demonstrates the path model for SES and self-esteem. The measures of fathers' education, mothers‘ education, respondent‘s education, family income, and self-esteem are represented by rectangles. Causal paths are represented by straight arrows, while relationships in which causal direction is unclear are represented by double-headed curved arrows. The results of these analyses show that respondent‘s education is a significant predictor of self-esteem and family income. Family income also significantly predicts self-esteem. Neither the path coefficient between father‘s education and self-esteem nor that between mother‘s education and 18 self-esteem attain the significant level. In short, none of the relationships between parental SES variables and self-esteem, and those between parental and personal SES variables is significant. Because of past racial discrimination, it is not surprising that parental education is not linked to respondent's education. Black parents of these respondents were denied opportunity that the respondents themselves may have had. Path analyses enable us to decompose the relations between parental SES and self-esteem, and between personal SES and self-esteem into direct, indirect, spurious, and unanalyzed effects (Asher, 1983). In this section, we focus on the direct and indirect effects of parental and personal SES on self- esteem. The sum of the direct and indirect effects of an independent variable on a dependent variable is equal to the total effect of the independent variable on the dependent variable (Asher, 1983). The total effect of father‘s education on self-esteem is -.046 which is equal to the sum of direct and indirect effects of father‘s education on self-esteem. The direct effect of father‘s education on self-esteem is -.038, which is 83% of the total effect for self-esteem. Most effect of father‘s education on self-esteem is direct, but the correlation between father‘s education and self-esteem is not high at all. The total effect of mother‘s education on self-esteem is .012. The direct effect of mother‘s education on self-esteem is equal to the total effect, .012, but the correlation between mother‘s education and self-esteem is even weaker than that between father‘s education and self-esteem. In general, the relationship between parental SES and self-esteem is not strong based on the analyses of causal effects. The total effect of respondent‘s education on self-esteem is .165. The direct effect of respondent‘s education on self—esteem is .12, which is 73% of the total effect for self-esteem. The correlation between respondent‘s 19 education and self-esteem is stronger than both that between father‘s education and self-esteem and that between mother‘s education and self- esteem. The total effect as indicated by the correlation of family income on self—esteem is .101. The direct effect of family income on self-esteem is .101 which is equal the total effect of family income on self-esteem. Again, the correlation between family income and self-esteem is stronger than both that between father‘s education and self-esteem and that between mother‘s education and self-esteem. These results indicate that respondent‘s education and family income are important factors to self-esteem. For a simpler model and testing self-perception theory, the incremental test to explained variance is performed using the same number of cases as the former model. Five paths are deleted based on the consideration of using the data to form our hypotheses (McPherson, 197 6; also see Pedhazur, 1982). Figure 2 demonstrates the new path model for SES and self—esteem. The measures of fathers‘ education, mothers‘ education, respondent‘s education, family income, and self-esteem are represented by rectangles. Causal paths are represented by straight arrows, while relationships in which causal direction is unclear are represented by double-headed curved arrows. The results of incremental test to explained variance show that the probability Of getting those differences of fit between the original and the new models by sampling error alone is greater than .05. The new model is acceptable. Based on self-perception theory, SES among black adults is achieved and is the product of personal accomplishment. Parental SES variables do not significantly affect either self—esteem or personal SES directly. On the contrary, personal SES variables had more significant effects on respondents‘ self—esteem than parental SES variables. Hence, the results support the hypothesis that the relationship between parental education and self-esteem is 20 weaker than that between respondent‘s education and self—esteem, and the data support self-perception theory. These findings are consistent with those of Rosenberg and Pearlin‘s study (1978), Demo and Savin-Williams (1983), and Wiltfang and Scarbecz‘s (1990). Self-perception theory is applicable to explain the relationships between SES and self-esteem for black adults. CHAPTER VI CONCLUSION Our findings indicate that personal SES variables, respondent‘s education and family income, are important factors to affect one‘s self-esteem among black adults. This supports 1 the hypothesis of self-perception theory that the feelings of self-esteem primarily stem from one‘s own accomplishment. Personal SES is more strongly correlated to self-esteem than is parental SES. The implication is that socioeconomic status is meaningful to a person because it is regarded as personal accomplishment and is earned by one‘s effort. In order to enhance one‘s status, personal accomplishment is particularly critical for blacks. Blacks have been ranked lower than whites in terms of socioeconomic status. The mobility of blacks based on parental SES is not as effective and efiicient as that of whites. Personal accomplishment is important to upwardly mobile blacks because this is the only possibility, even though the structure offers no opportunity. Self-perception theory is applicable for blacks and is important to explain why the effects of black‘s personal SES on self-esteem is greater than those of parental SES on self-esteem. Some limitations of this research are the lack of advanced data of father‘s and mother‘s occupation history and much missing data on father's and mother‘s education. The weakness of these data may aflea the validity of measurement on parental SES and make the test of intergenerational job mobility unable to be performed. 21 22 The causal patterns of effects of SES for self-esteem found in the present research hold for a black adult sample need for further study. However, even if the patterns can reflect the black adult population, their implications are highly significant. These findings suggest that the frame of reference provided by black group may offer important criteria of success for black adults. Self-perception theory is an important theory, more understanding about its power to explain the various levels of self-esteem between men and women and among different age groups of blacks is an issue for future research. 23 Table l. DESCRIPTIVE CHARACTERISTICS OF THE SAMPLE Self-Esteem Variables It Mean Education Grammar school (08 years) 456 3.49 9-11 years 459 3.48 High school graduate 650 3.53 Some college 334 3.58 College graduate 183 3.67 Total 2,082 3.53“ Family income $0-2,999 239 3.39 $3,000-4,999 222 3.50 $5,000-6,999 213 3.48 $7,000-9,999 255 3.60 $10,000-l4,999 272 3.54 $15,000 or more 632 3.58 Total 1,833 3.53" Father‘s education Grammar school (0-8 years) 634 3.54 9—11 years 151 3.54 High school graduate 209 3.50 Some college 51 3.58 College graduate 177 3.54 Total 1,222 3.53 Mother‘s education Grammar school (0-8 years) 596 3.54 9-11 years 308 3.53 High school graduate 369 3.53 Some college 66 3 .42 College graduate 111 3.56 Total 1,450 3.53 *p < .05. **p < .01. 24 Table 2. REGRESSION OF SELF-ESTEEM ON SOCIOECONOMIC VARIABLES Self-Esteem Equation I Equation [1 Variables b B r b B r Constant 3.054 -- -- 3.083 -- -- Age .004“ .155 .146 .004M .163 .141 Sex (1=rnale) .032 .037 .036 .015 .017 .016 Education .017" .119 .101 .014" .112 .088 Family income .010" .101 .087 .011" .116 .099 Father‘s Education -.001 -.O39 -.O36 -- -- -- Mother‘s Education .000 .012 .01 1 -- - .- R2 .049 .038 n 1,000 1,822 Fratio 8.591 17.988 89' (6, 993) * (4, 1817) Note: b = unstandardized regression coeflicient, B = standardized regression coefficient, and r = partial r. When b is significant, B and r are significant at the same level. Degrees of freedom (regression, error) for F ratio in parentheses. *p < .05. "p < .01. 25 Figure 1. Path Model of SES and Self-Esteem \ Respondent Education Father / Education .377 Mother Education Self-Esteem Family Income Note: All coeficients net of sex and age. **p<.01. 26 Figure 2. Alternative Path Model of SES and Self-Esteem Respondent Education -.05 Father .122“ Education 37" Self-Esteem Mother Education «“5“ ~10 “ Family Income Note: All coefficients net of sex and age. ** p < .01. LIST OF REFERENCES Asher, Herbert B. 1983. Causal Modeling. Beverly Hills, CA: Sage. Bachman, Jerald G. and Patrick M. O‘Malley. 1984. “Black-White Differences in Self-Esteem: Are They Affected by Response Styles?" American Journal of Sociology, 901624-39. Bachman, Jerald G. and L. D. Johnston. 197 8. The Monitoring the Future Project: Design and Procedures. Ann Arbor: University of Michigan, Institute for Social Research. Baumgardner, A. H. 1990. “To Know Oneself Is to Like Oneself: Self- Certainty and Self-Affect.“ Journal of Personality and Social Psychology, 58:1062-1072. Bem, Daryl J. 1965. “An Experimental Analysis of Self-Persuasion.“ Journal of Experimental Social Psychology, 1:199-218. 1967. “Self-Perception: An Alternative Interpretation of Cognitive Dissonance.“ Psychological Review, 74(3):]83-200. Bohmstedt, G. W. 1983. “Measurement.“ Pp. 69-121 in Handbook of Survey Research, edited by P. H. Rossi, J. D. Wright, and A. B. Anderson. Academic Press. Broman, Clifford L., Harold W. Neighbors, and James S. Jackson. 1988. "Racial Group Identification Among Black Adults." Social Forces, 67:146-158. '27 28 Broman, Clifford L., James S. Jackson, and Harold W. Neighbors. 1989. “ Sociocultural Context and Racial Group Identification among Black g- Adults.“ Revue Internationale de Psychologie Sociale, T.2, N.3:367- 78. Carnpbell, J. 1990. “Self-Esteem and Clarity of the Self-Concept." Journal of Personality and Social Psychology, 59:538-549. Clark, K. B. and M. P. Clark. 1947. “Racial Identification and Preference in Negro Children.“ Pp. 169-17 8 in Readings in Social Psychology, edited by T. M. Newcomb and E. L. Hartley. New York: Holt, Rinehart & Winston. Cross, William D., Jr. 1985. “Black Identity: Rediscovering the Distinction between Personal Identity and Reference Group Orientation.“ Pp. 155- 71 in Beginnings: The Social and Aflective Development of Black Children, edited by Margaret B. Spencer, Geraldine K. Brooklins, and Walter R. Allen. Erlbaum. Demo, David H. and Ritch C. Savin-Williams. 1983. "Early Adolescent Self-Esteem as a Function of Social Class: Rosenberg and Pearlin Revisit .“ American Journal of Sociology, 88(4):763-74. Glenn, Norval D. 1963. “Negro Prestige Criteria: A Case Study in the Bases of Prestige.“ American Journal of Sociology, 682645-57. Goodman, M. E. 1952. Race Awareness in Young Children. Reading, MA: Addison-Wesley. Gordon, C. 1969. Looking Ahead. Washington, DC: American Sociological Association. Greenberg, J ., S. Solomon, T. Pyszczynski, A. Rosenblatt, J. Burling, D. Lyon, L. Simon, E. Pinel. 1992. “Why Do People Need Self-Esteem? Converging Evidence That Self- Esteem Serves an Anxiety-Buffering Function.“ Journal of Personality and Social Psychology, 63:913-922. Hacker, Andrew. 1992. Two Nations. New York: Charies Scribner’s Sons. 29 Heiss, J erold and Susan Owens. 1972. “Self-Evaluation of Blacks and Whites.“ American Journal of Sociology, 782360-70. House, James S. 1981. “Social Structure and Personality.“ Pp. 525—61 in Social Psychology: Sociological Perspectives, edited by Morris Rosenberg and Ralph J. Turner. New York: Basic Books. Jackson, James S., Wayne R. McCullough, and Gerald Gurin. 1981. “Group Identity Development within Black F arnilies.“ Pp. 252-63 in Black Families, edited by Harriet P. McAdoo. Sage. Jackson, J. S., M. B. Tucker, and G. Gurin. 1987. National Survey of Black Americans 1 979-80 (MRDF). Distributed by Inter-University Consortium for Political and Social Research, Institute for Social Research, Ann Arbor, MI. Landreth, C. andB. C. Johnson. 1953. “Young Children‘s Responses to a Picture and Inset Test Design to Reveal Reactions to Persons of different skin color.“ Child Development, 24:63-79. McConahay, J. B. 1986. “Modern racism, ambivalence, and the modern racism scale.“ In J. F. Dovidio & S. L., Gaertner (Eds), Prejudice, descrimination and racism, pp. 91-125. New York: Academic Press. McPherson, J. M. 1976. “Theory Trimming.“ Social Science Research, 5:95-105. Nunnally, Jum C. 197 8. Psychometric Theory. McGraw-Hill. Pedhazur, Elazar. 1982. “Path Analysis.“ Pp. 577-635 in his Multiple Regression in Behavioral Research. The Dryden Press. Pettigrew, Thomas F. 1967. “ Social Evaluation Theory: Convergences and Applications.“ Nebraska Symposium on Motivation, 15:241-311. Porter, Judith R. 1971. Black Child, White Child: The Development of Racial Attitudes. HarVard University Press. Porter, Judith R. and Robert E. Washington. 1979. “Black Identity and Self- 30 Esteem: A Review of Studies of Black Self-Concept, 1968-1978.“ Pp. 53-74 in Annual Review of Sociology. Vol. 5, edited by Alex Inkeles, James Coleman, and Ralph H. Turner. Annual Reviews. Radke, M. J. and H. G. Trager. 1950. “Children‘s Perceptions iof the Social Roles of Negroes and Whites.“ Journal of Psychology, 29:3-33. Rosenberg, Morris. 1979. Conceiving the Self Basic Books. Rosenberg, Morris and Leonard I. Pearlin. 1978. “Social Class and Self- Esteem among Children and Adults.“ American Journal of Sociology, 84(1):53-7 7 . Rosenberg, Monis and Robert G. Simmons. 1971. Black and White Self- Esteem: The Urban School Child. American Sociological Association Rose Monograph Series. Samuels, S. 1973. “An Investigation into the Self-Concepts of Lower and Middle-Class Black and White Kindergarten Children.“ Journal of Negro Education, 42:467-472. Simmons, Robert G. 197 8. “Blacks and High Self-Esteem: A Puzzle.“ Social Psychology Quarterly, 41:54-57. Stevenson, H. W. and E. C. Stewart. 1958. “A Developmental Study of Race Awareness in Young Children.“ Children Development, 29:399- 409. Taylor, Robert J. 1986. “Receipt of Support from Family among Black Americans: Demographic and Familial Differences.“ Journal of Marriage and the Family, 48:67-77. Wade, T. J ., V. Thompson, A. Tashakkori, and E. Valente. 1989. “A Longitudinal Analysis of Sex by Race Differences in Predictors of , Adolescent Self-Esteem.“ Personality and Individual Diflerences, 10:7 1 7 -7 29. ' Wiltfang, Gregory L. and Mark Scarbecz. 1990. “Social Class and Adolescents‘ Self-Esteem: Another Look.“ Social Psychology 31 Quarterly, 53(2): 174-183. Yancey, William L., Leo Rigsby, and John D. McCarthy. 1973. “Social Position and Self-Evaluation: The Relative Importance of Race.“ American Journal of Sociology, 78: 338-359. MICHIGAN STATE UNIV. LIBRARIES illll“illNlllllilllilllllWillilllllllllllmlWHIIHI 31293010289332