OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. 1,! .. r © 1979 ARTURO TOMAS R I 0 ALLRIG'ITS RESERVED DEFENSIVENESS, SELF CRITICISM AND SELF CONCEPT IN A SAMPLE OF BLACK, MEXICAN, AND WHITE AMERICAN ADOLESCENTS BY Arturo Tomas Rio A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services, and Educational Psychology 1979 ABSTRACT DEFENSIVENESS, SELF CRITICISM AND SELF CONCEPT IN A SAMPLE OF BLACK, MEXICAN AND WHITE AMERICAN ADOLESCENTS By Arturo Tomas Rio The purpose of this study was to investigate differ- ences in total self concept, self criticism and defensive- ness among Black, Mexican and White American adolescents. The relationship between ethnic group membership and other personality variables was also examined. The research was conducted in a medium-sized, Mid- western, heavily industrialized and labor intensive urban school setting. A sample was drawn from the population of tenth, eleventh and twelfth grade students in the two secondary schools of the district. One school was pre- dominantly White, the other was predominantly Black. The Mexican American population was less than 10 percent in each of the two schools. Three major null hypotheses were formulated: Hypothesis 1. Thre will be no difference in mean Self Concept scores of Black, Mexican and White American secondary school students. Hypothesis 2. There will be no difference in mean Self Criticism scores of Black, Mexican and White Arturo Tomas Rio American secondary school students. Hypothesis 3. There will be no difference in mean Defensiveness Positive scores of Black, Mexican and White American Secondary school students. Subhypotheses: There will be no interaction between race and all other independent variables. The Tennessee Self Concept Scale (TSCS) and an Infor- mation Questionnaire developed by the author were adminis- tered to a sample of 301 students. There was a total of five independent variables (race, sex, grade, achievement, and socioeconomic status) and twenty—nine dependent variables (TSCS scores). School was not treated as an independent variable. An SPSS file was established and the following statis- tical procedures were programmed: 1) one-way-ANOVA's between all dependent and independent variables, and 2) two- way ANOVA's of all hypothesized relationships between dependent and independent variables. The alpha level for statistical significance was set at f .05. Hypotheses one and three were not rejected. Hypothe- sis two was rejected at the .02 level. Ethnic group member- ship was found to be a variable significantly associated with the following TSCS scores: 1) Psychosis, 2) Personality Disorder, 3) Personality Integration, 4) Number of Deviant Signs, 5) General Maladjustment, 6) Social Self, 7) Moral- Ethical Self, 8) Physical Self, 9) Variation (External), and 10) the five scores derived from response characteris- tics of subjects. Arturo Tomas Rio Findings indicated that White adolescents scored more favorably than Blacks and Mexican Americans in all but four of the fourteen scales. Black and Mexican American students obtained more desirable scores than White students in the following scales: 1) Physical Self, 2) Personality Dis- order, 3) General Maladjustment, and 4) Psychosis. ACKNOWLEDGEMENTS The writer wishes to express his deep appreciation to Dr. Louise M. Sause, dissertation committee chairperson, for the immeasurable assistance and guidance extended throughout this project. The writer is also fortunate to have had Professors John P. McKinney, William A. Mehrens, and John H. Schweitzer as members of the committee. The valuable contributions extended by these individuals made the completion of this study possible. Thanks are also due to Drs. T. Clinton Cobb and Alexander Kloster for their guidance and concern during the first year of the doctoral program. The suggestions and direction of David Solomon on questions pertaining to statistics and research design are gratefully acknowledged. The contributions of Barry E. Quimper in the areas of sampling and instrument administration were of crucial importance. Without his c00peration and diligent parti- cipation throughout the data gathering procedure, this study could not have materialized. I am fortunate to have had unconditional support from the Department of Racial and Ethnic Studies on this project. Special thanks are due to Dr. Jack M. Bain and the departmental staff for their consideration and encour- agement while I was a member of the faculty. ii Gratitude is extended to my fellow graduate student, Richard Gonzalez, for his assistance in the handling of the data; and to Ms. Patricia Beach and Kathie B. Alcoze for their patience and cooperation in the editing and typing of the manuscript. I feel particularly grateful to my family for the continued support extended during graduate school. I would like to express special thanks to my brother, Dr. Raul A. Rio, for advice and direction on personal as well as pro- fessional matters during this period. Also, to Elaine and Michael for their tolerance in receiving an answer to the question: "When will you be finished with school?" To my parents, Dr. and Mrs. Raul Rio-Leon, I wish to extend my deepest love and appreciation for all they have unselfishly contributed toward my development. Without the assistance, understanding and encourage- ment of those individuals mentioned here this major goal in my life would never have been realized. iii TABLE OF CONTENTS ACKNOWLEDGEMENTS . LIST OF TABLES . Chapter I. II. III. INTRODUCTION The Problem . . Need for Study . Purpose of Study REVIEW OF THE LITERATURE Introduction Studies on Ethnic Identity and Preference : Group Membership . Self Concept and Defensiveness in Self Report . . Self Concept and Related Variables Grade Level Sex Socioeconomic Status Achievement METHODOLOGY . Hypotheses Self Concept . Self Criticism . Defensiveness Description of Population . Sampling Procedure Instrumentation . Tennessee Self Concept Scale . Information Questionnaire Variables in the Study Statistical Procedure . Page ii iv Chapter Page IV. RESULTS . . . . . . . . . . . . . . . . . . . 48 Test of Research Hypotheses . . . . . . . . . 48 Relationships Between Other Variables in the Design . . . . . . . . . . . 58 Self Concept From an Internal Frame of Reference . . . . . . . . 58 Self Concept from an External Frame of Reference . . . . . . . . . . . 60 Empirical Scales and Race . . . . . 62 Response Characteristics of Subjects . . . 65 V. DISCUSSION, IMPLICATIONS AND CONCLUSIONS . . 68 Discussion . . . . . . . . . . . . . . . . . 68 Research Hypotheses . . . . . . . . . 68 Relationships Between Other Variables in the Design . . . . . . . . . 72 Implications . . . . . . . . . . . . . . . . 76 Conclusions . . . . . . . . . . . . . . . . . 77 APPENDICES A. Nature and Meaning of Tennessee Self Concept Scale Scores . . . . . . . . . . . . 80 B. Hollingshead Two Factor Index of Social Position . . . . . . . . . . . . . . . 89 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . 99 Table 10. 11. 12. 13. 14. 15. Anova of LIST OF TABLES Total Positive Self Concept Scores and Race Anova of Total Self Concept Scores by Sex and Race Anova of Race and Anova of Race and Anova of Race and Anova of Anova of and Sex . Anova of and Grade Anova of Total Self Concept Scores by Achievement . . Total Self Concept Scores by Grade . . Total Self Concept Scores by Socioeconomic Status Self Criticism Scores by Race Self Criticism Socres by Race Self Criticism Scores by Race Self Criticism Scores by Race and Achievement Anova of Self Criticism Scores by Race and Socioeconomic Status Anova of Anova of Race and Anova of Race and Anova of Race and Anova of Race and Defensive Positive Scores by Race Defensive Positive Scores by Sex Defensive Positive Scores by Grade Defensive Positive Scores by Achievement Defensive Positive Scores by Socioeconomic Status iv Page 49 49 SO 50 51 52 53 53 S4 54 55 56 S6 57 S7 Table Page 16. Anova of Self Concept Scores from an Internal Frame of Reference by Independent Variables . . . . . . . . . . . . . 59 17. Anova of Self Concept Scores from an External Frame of Reference by Independent Variables . . . . . . . . . . . . . . . . . . . 61 18. Anova of Empirical Scales of the TSCS by Independent Variables . . . . . . . . . . . 63 19. Anova of Subjects' ReSponse Characteristics on the TSCS by Independent Variables . . . . . 66 20. Response Distribution of Subjects by Ethnic Group Membership . . . . . . . . . 67 CHAPTER I INTRODUCTION Educational researchers have been concerned with the relationship between the self concept and other factors such as learning, academic achievement, socioeconomic sta- tus, sex, racial/ethnic group membership, school climate, home environment, social competency, teaching methods, and curricular design. It has been extensively argued that the self concept of a student is an important variable in the educational process. The influence of self concept is a complex topic of potential importance to educational theory and practice. An individual's self concept is indicative of what type of a person that individual believes that he/she is. The self concept is typically acquired from what "signifi- cant-other" people think of that individual. A person adopts, and eventually incorporates into the self-structure the beliefs that he/she feels others have of him/her. A person acquires a notion of a personal self from what other people think of or act toward that person. A voluminous amount of research has indicated that children reveal varying degrees of self-acceptance, self- esteem, and self-worth, suggesting self-conceptualizations which.may be viewed as positive (high) or negative (low); and which are not only reflective of social-emotional ad- justment but also interact with other aspects of psycho- logical functioning, including academic achievement and other school-related outcomes. The Problem A review of the literature indicates that a low (nega- tive)self concept has been observed among children belong- ing to racial and ethnic minority groups in the United States. Mexican American (Chicano) and Puerto Rican Ameri- can students have been found to have significantly lower self concepts than their Anglo American (and Black American) counterparts (e.g. Coleman, 1966). In recent years the re- search outcomes have been contradicting. A number of stud- ies over the past decade have reported that racial minority group students scored higher on self concept measures than white students. Further research on this problem is needed in order to determine the relationship between racial group membership and self concept. The process of identification is one which begins at a very early age and is considered crucial in the development of the self-image or self concept. One of the earliest forms of identity evident in American children is ethnic and/or racial identity (Clark and Clark, 1939; Goodman, 1952; Morland, 1958; Porter, 1971). The establishment of racial and ethnic identity is influenced by factors such as skin color (i.e., racial characteristics) and ethnic group mem- bership. These have been considered two of the most impor- tant factors in self-identification among minority group children. It is possible that American society reflects a racial and ethnocentric ideology which young children per- ceive at a very early age. The preschool child eventually learns that racial/ethnic group membership is an important variable in identification and self-evaluation. Ehrlich (1973) outlined a comprehensive review of the literature on the social psychology of prejudice in support of his theoretical (propositional) perspective. In this book he states that "ethnicity is a major characteristic by which people code themselves and other persons in society" (p. 128). In an examination of the effects of prejudice and discrimination on self-attitudes, Simpson and Yinger (1972) emphasized the role of social learning in the forma- tion and development of the self concept: Our analysis will be built largely around the exper- ience of learning that one is of an inferior color, but this should be seen as illustrative of the whole experience of learning that one belongs to a minor- ity group. It is in the context of slights, rebuffs, forbidden opportunities, restraints, and often vio- lence that the minority-group member shapes that fundamental aspect of personality--a sense of one- self and one's place in the total scheme of things. (p. 192) American education has been primarily a process of socialization aimed at assimilation into a dominant White society. The child who enters school with a different lin- guistic, cultural and socioeconomic background than that fostered by the total school environment.may be expected to confront very real and serious problems in attempting to adapt to a foreign and narrow monolingual-monocultural sys- tem. This may be the case regardless of the variable of race. Becker (1972) conducted a study on teacher attitudes toward their inner-city, lower socioeconomic class (and pri- marily minority) pupils and found their perceptions to be very negative. This sample of teachers viewed their stu- dents as dirty, promiscuous, violent, unambitious, diffi- cult to teach, and virtually uncontrollable. There are other indications that Anglo teachers do not adequately know or understand Mexican American children (Thurston, 1957; Ulibarri, 1960). Mexican American children are con- sidered less important and less favorable than Anglo chil- dren. Parsons (1966) studied the opinions held by Anglo teachers toward Mexican Americans in the state of California. He reports: "In general, the Anglo informants characterized the Mexicans as immoral, violent, and given to fighting, dirty, unintelligent, improvident, irresponsible, and lazy." A related finding by Jacobson (1966) is that Anglo teachers of Mexican American students of high IQ tend to perceive them as being more American (Anglo or White) than Mexican. Need for Study A comprehensive review of the literature on self con- cept and Mexican Americans has yielded conflicting evidence. Comparisons of Mexican American students' reported self con- cept to self concept measures obtained from samples of non- Mexican American students have revealed a number of differ- ent relationships. Most studies report a lower self concept for Mexican Americans than for Anglos. Other studies have failed to reveal significant differences in self concept measures comparing groups of Mexican American students to their Anglo and/or Black counterparts. A smaller number of studies have yielded higher self concept scores for Mexican Americans than for Anglo Americans. A review of the re- search literature on self concept and racial/ethnic group membership is presented in Chapter II of this study. It has been suggested that the contradictory research findings are indicative of changing social, economic, poli- tical, and educational conditions over the last decade or two. Zirkel (1973) has suggested that the factor of time is of potential importance in the context of changing so- cial and political conditions of American minorities. An examination of reported findings and corresponding publication dates suggests that there has been an increase (improvement) in the self concept scores of Mexican Ameri- cans over the last decade. It is possible that a change in self concept among American minority students is at least in part related to the Civil Rights Movement of the 1960's and subsequent changes in educational practice which have ultimately led to improved conditions and opportunities for minority group students. Fantini and Weinstein (1968) have attributed the re- ported rise of self concept scores of Black students to the Civil Rights Movement and the subsequent emphasis placed on Black pride by the Black community and school systems. How- ever, this interpretation has been ruled out by Cicirelli (1977) on the basis of research findings that are inconsis- tent and often contradictory. There exists the possibility that the reported posi- tive change in the self concept of American minority stu- dents may be an artifact. One possible factor which may account for the recent and divergent research findings on self concept and ethnic group membership (i.e., a higher self concept for minority children or a lack of differences in self concept scores obtained from comparable samples of minority and majority children and youth) has been termed "defensiveness" on the part of minority subjects. Researchers have commented on this notion of defensive- ness. DeBlassie and Healey (1970) pointed out that Mexican American students are defensive in self-report measures and attempt to convey an overly positive picture of themselves. In a subsequent publication Healey and DeBlassie (1974) found significantly higher Defensive-Positive and lower Self-Criticism scores for both Mexican American and Black American samples than for the Anglo American sample. The data were based on the TENNESSEE SELF CONCEPT SCALE. The authors hypothesized that the "artificially elevated" scores obtained from both minority samples were due to defensivee ness in responding. Long (1969) contended that defensive responding on the part of minority students on self concept instruments oc- curs in order to convey an overly positive image of the self. Greenberg (1970) considered the possibility that de- fensiveness may account for high self ratings of low achievers. Samuels (1977) has called attention to the notion of defensiveness as a potential measurement and re- search problem. The tendency of subjects to reveal an overly favorable and thus inaccurate picture of the self in personality in- ventories has been a long-recognized problem in the measure- ment of personality traits and research based on such data. Defensiveness and lack of response reliability were given by Williams and Byars (1968) as factors to be taken into account in the interpretation of self concept scores obtained from subjects of minority group status. Soares (1969) found that low socioeconomic background students re- ported a higher percentage of favorable personality charac- teristics than middle class students. Findings were inter- preted in terms of aspiration level, social pressure, and meeting expectations of others and self. Eisenberg (1967) believes that lower class children develop defense mechan- isms against repeated failure and do not care if they suc- ceed in the future. Such defensive dynamics may create new referents for self-evaluation. The evidence suggests that the notion of defensiveness is one which is worthy of systematic investigation since its operation may affect true self concept scores and thus threaten the validity of comparative self-concept research findings and instruments. Purpose of Study The primary purposes of this study are: l) to deter- mine if there are differences in self concept among Black, Mexican and White American secondary school students; and 2) to determine if there are differences in defensiveness among Black, Mexican and White American secondary school students. This study will also assess the relationship of sex, grade level, academic achievement, and socioeconomic status to the obtained total self concept scores and two different measures of personality defensiveness. The statistical analyses will test for the effect of interaction between scores on a standardized self concept measure and the inde- pendent variables identified above. Specific hypotheses to be tested are listed in Chapter III. CHAPTER II REVIEW OF THE LITERATURE Introduction Sullivan (1949), Rogers (1951), Jersild (1952), Combs and Snygg (1959), Hamachek (1971) and other phenomenologi- cal personality theorists have emphasized the role of the individual's perception of situational conditions as well as the importance of the judgments of "significant-other" people in the formation, development, and modificability of the self concept. The role of the family in early development and found- ation of the self concept has been long remembered (Wylie, 1961). Christian (1976) believes that the family is the primary source of socialization for the child and by the time the secondary socialization process begins, "in which the school and the written word play a major role, the self concept is considered to have been formed in a manner and to a degree that it will not be changed fundamentally there- after." He stresses the fact that during the primary socialization period "the child internalizes the world of family members and friends," and this has "transcendent significance in the creation of the child's identity during this period" (pp. 18-19). Coopersmith (1967) has also 10 emphasized the importance of the family, particularly the relationship between the child and parents in terms of emo- tional closeness and discipline or control over behavior. As the child enters school, teachers, peers and school materials will have a significant impact on the child's de- veloping self concept (see, for example, Combs, 1962). Kinch (1963) postulated that the self concept develops as a result of social interaction and is an important variable in guiding future behavior. The child attempts to judge and value the self in relation to peers. As the child develops socially, cognitively and emotionally, he/she gains a better conception of self as reflected from significant others; a better understanding of position in society; and the signi- ficance of racial/ethnic group membership. Studies on Ethnic Identity and Preference The early studies of Clark and Clark (1939) on racial identification and preference among Black children have gen- erated much interest and subsequent research in the last four decades. Simpson and Yinger's (1972) review of such research led them to conclude that there is substantial sup- port for the following generalizations on race awareness and preference. They summarize as follows: Race awareness begins at an early age, particularly in societies where the race line is important; in the United States, a large majority of both black and white children express preferences of white dolls, puppets, or play-group members; boys may be more likely to favor white than girls, by age seven or eight the expressed preference of black children for white begins to decline; these patterns have remained quite constant through the last several decades. (p. 193) 11 Goodman (1952) utilized the doll technique with a sample of nursery school children. She found that Black children, on the average, perceived racial differences at an earlier age than their White counterparts; and that both Black and White children manifested more hostility toward Black dolls than they did toward White dolls. Werner and Evans (1968) found that four- and five-year- old Mexican American children were able to discriminate be- tween "good" and "bad" dolls on the basis of skin color, classifying good dolls as white and bad as dark. Moreover, boys tended to perceive the white male adult doll as larger than the dark adult doll of exactly the same size. Results showed that feelings of prejudice and ethnic identification were evident in preschool years. A more recent study of racial awareness and identifi- cation was conducted by Durrett and Davy (1970). They used a sample of low socioeconomic background kindergarten chil- dren from San Jose, California. Subjects were Black, Mexi- can American and White children in an integrated setting. Using the doll technique (Black and White only), they found that: l) Anglos expressed the highest own-race preference in both identification and choice of playmate; Z) Blacks showed the least own-race preference (as above); 3) Mexican Americans valued the Anglo group over the Black; 4) Blacks showed less own-race rejection than that suggested by com- parable research reported eleven years earlier; and 5) there was less evidence of self-derogation and hostility on the 12 part of Black preschoolers than had been previously reported in the literature approximately one decade ago. Hraba and Grant (1970) examined racial preferences of Black children aged four to eight and found preference for the doll of their own race, suggesting that black children in inter-racial settings are not necessarily white-oriented. The race of the interviewer did not affect choice of doll, and Black doll preference was found to be positively asso- ciated with age. Harris (1962) reported that Black males were more likely to identify in terms of race than Black females; whereas White female subjects more often identi- fied themselves in terms of race than White males. Rice, Ruiz and Padilla (1974) presented a set of three (Anglo, Black and Mexican American) photographs of adult males to groups of preschool and third grade students from the three different groups. Results indicated "a well de- veloped knowledge of racial differences," although Mexican American preschoolers had difficulty in differentiating be- tween Anglo and Mexican American photographs. Anglo pre- schoolers displayed a significant preference for the photo- graph of their own ethnicity and the other two groups did not. At the third grade, however, only Mexican Americans showed a clear preference for their own ethnic group. A significant number of Black preschoolers chose the Anglo photograph as the one they would prefer to grow up to be like. 13 A review of studies of racial awareness and identifica- tion in young children has been published by Stevenson (1967). He concluded: The results from this sample of studies on racial awareness with preschool children are sobering. It is clear that children as young as three can dis- criminate racial differences and that negative at- titudes towards members of another race may be evi- denced by children as young as five. The problems for healthy personality development arising from membership in a minority group are also revealed. (p. 213) The available research studies clearly indicate that ethnic identity and preference of North American children is established at a very early age. Hawk (1967) further added that among the socially disadvantaged, "the self con- cept is formed through assimilation of external labels that are applied to the person" (p. 199). Self Concept and Ethnic Groungembership A review of the literature on self concept and ethnic group membership and mixture has been published by Zirkel (1973). He noted that although there have been numerous studies on the subject (most of which appeared in the 1960- 1970 decade), "Wilie's (1961) landmark review of the self concept research did not mention any empirical studies re- lating self concept to ethnic group membership" (p. 211). Zirkel states that "the effects of ethnic group mixture in the school on the self concepts of students is a subject which has evoked perhaps much heat but little light" (p. 214). He also notes that the reported findings for minority 14 groups other than Blacks are "less numerous but somewhat similar" (p. 213). Chang (1975) notes that while most of the research on the relationship between the self concept and ethnicity have centered around Blacks and Mexican Americans, "the results of these studies have been conflicting and unclear." The majority of the published research studies on the relation- ship between self concept and ethnic group membership in the United States has utilized samples of Black students in com- parison to White counterparts. A brief review of findings emanating from such studies follows. There are a number of studies which show no significant differences in self concept for Blacks and Whites (Coleman, 1966; Clark et al., 1967; Gibby and Gabler, 1967; Powers, 1971; Zirkel and Moses, 1971). Several investigators have reported that Blacks have higher self conceptualizations than Whites (Soares and Soares, 1969, 1970; Trowbridge, 1970,1972; Cicirelli, 1977). However, the majority of studies in the past have revealed significant differences in favor of White students. Among the studies reporting that Black students exhibited lower self concepts than White students are those of Deutsch (1960), Ausubel and Ausubel (1963), Keller (1963), Radke-Yarrow et al., (1965), Brown (1968), Long and Henderson (1968), and Williams (1968). Studies comparing self concept measurements of Mexican American and White students reveal findings similar to Black- White comparisons. Among the studies showing no significant 15 difference in self concept between Mexican American and White subjects are those of Carter (1968), DeBlassie and Heahnr(1970), Klein (1975), and Little and Ramirez (1976). Studies indicating significant differences between the two groups are more numerous. Coleman (1966) found that Mexican American students had significantly lower self concepts than did Anglo and Black students. McDaniel (1967) reports the same phenomenon for Mexican Americans with re- ference to Anglos, but notwhen compared to Blacks. Palo- mares and Cummings (1968a; 1968b), Evans (1969), and Hishiki (1969) also report lower self concepts for Mexican American students than for White American counterparts. A study involving Mexican and Anglo American students from kindergarten through the fourth grade was undertaken by Muller and Leonetti (1974) in order to determine if there were differences in self concept, as measured by the PRIMARY SELF CONCEPT SCALE (Leonetti, 1973). The 24 pictorial-item instrument was administered to ten Anglo and ten Mexican American students at each of the five levels. Subjects were from low socioeconomic backgrounds ranging in annual family incomes from $2,500 to $4,000. The results indicate that at kindergarten level, Anglo students had significantly higher self concepts than Mexican Americans but differences were not significant at the other four grade levels. There was an upward trend in self concept scores for the Mexican American sample across grades but this was not true for the Anglo American sample. The lack of an overall difference 16 between the two groups "suggests that (they) do not differ to the extent previously suspected," and that "the school experience seems to have a positive effect on the self con- cept of the Spanish—surnamed child" (pp. 58e59). The lack of differences may be due to: 1. increased teacher sensitivity toward Mexican Americans in the school district; 2. control of socioeconomic status of subjects; and/or 3. properties of the instrument. Maldonado (1972) administered the TENNESSEE SELF CON- CEPT SCALE to a group of male and female Mexican American tenth, eleventh and twelfth graders from low socioeconomic backgrounds. The scores did not reveal a sense of infer- iority and worthlessness typically attributed to members of this particular group. The overall scores on the TSCS were not significantly different for males and females. Little and Ramirez (1976) administered a semantic dif- ferential scale (MYSELF AS I REALLY AM) to a group of Mexi- can American and Anglo fifth through eighth grade students. No significant ethnic differences in self concept were found, but the ethnicity of the testers significantly in— fluenced self esteem scores of seventh and eighth graders, and not the elementary school students. The Anglo tester was able to elicit more positive self-esteem scores than the Mexican American tester. Although the results could be attributed to different personality characteristics 17 between the two scale administrators, the authors suggested that the older students may have more accurately perceived the Anglo as a typical evaluator (i.e., an individual with power and authority) and thus described themselves more positively. This interpretation is subject to an empirical test. However, the finding suggests that the ethnicity of the tester should be taken into account in the design of a self concept study utilizing a sample of middle school age students. Zirkel and Moses (1971) found that self concept was af- fected by ethnic group membership but was not influenced by the ethnic group composition or mixture in a particular school. A study was conducted by Petersen and Ramirez (1971) in order to determine if Mexican and Black American children experience a greater discrepancy between their real and ideal selves than Anglo American children. Subjects for the study were low socioeconomic class nine- to fourteen- year-olds. The instrument was a modified version of Block's (1964) inventory consisting of two rating scales. One rat- ing scale was used for assessing the real self and the other rating scale for assessment of the ideal self. The results indicated a significantly greater disparity between real- ideal self scores for both Mexican American and Black Amer- ican than for Anglo American subjects. Sex differences within groups were not found to be significant. Observed feelings of self-rejection and anger were interpreted to be 18 the result of frustration experienced by minority group members due to social discrimination. In addition, it was noted that the Mexican American may have identity problems due to the "internalization of conflicting sets of values" (p. 26). The finding that Mexican American children had higher discrepancy scores than both Blacks and Anglos on self-rejecting measures (e.g. "dull" and "lazy" items) sug- gests that they may have internalized common social stereo- types promulgated by the dominant Anglo culture toward Mexicans. Swartz (1971) found a less optimistic orientation to- ward the future among Mexican American students. Anthro- pological reports suggest that this is associated to a fatalistic element inherent in the Mexican American culture; but other reports have suggested that the effects of poverty and failure are the major factors in the interpretation of this finding. A fatalistic orientation, reduced motivation, lowered ambition and expectations, decreased confidence and aspirations as well as feelings of apathy have been asso~ ciated with low self-conceptualizations among individuals. Indicative of a low self concept is the high frequency of self-derogation which has been observed among Mexican American students, particularly adolescents. For example, Mexican American high school seniors were found by Coleman et a1. (1966) to be more self-depreciating than their Anglo counterparts. 19 Related findings by Gibby and Gabler (1967) were that Black sixth graders' self-ratings on intelligence were less accurate when compared with intelligence test scores than the intelligence self-ratings obtained from White peers. The Black students' tendency to overrate personal intellec- tual abilities has also been interpreted by Samuels (1977) to be supportive of the defensiveness point of view. Healey and DeBlassie (1974) conducted a comparative self concept study in order to determine if differences existed among Black, Anglo and Mexican American adolescents attending the ninth grade in a Southwestern middle school system. Findings based on the TENNESSEE SELF CONCEPT SCALE (TSCS) revealed that Mexican Americans scored higher on the Self-Satisfaction and Moral-Ethical Self scores than both Black and Anglo American subjects. Although no significant differences were found between socioeconomic class and the Total Positive scores, the Social Self scores tended to in- crease as socioeconomic position increased. In addition, there were significant differences attributed to sex, with male subjects scoring higher; but no sex-ethnicity interac- tion was found. Self Concept and Defensiveness in Self Report The Healey and DeBlassie (1974) study noted above revealed that the Total Positive scores did not differ sig- nificantly among the three groups. Anglo Americans had the lowest mean score (317.75) and Mexican Americans the high- est (328.70). Blacks had a mean score of 321.69. The 20 Self-Criticism subscores for Blacks (34.40) and Mexican Americans (34.16) in the above study were found to be sig- nificantly lower than those obtained by Anglos (37.01) on this scale. The Defensive-Positive subscores for Blacks (57.45) and Mexican Americans (56.21) were also signifi- cantly higher than the mean for Anglos (49.40). This was indicative of defensiveness and the authors hypothesized that this factor was responsible for the "artificially ele- vated" Total Positive scores obtained for Black and Mexican American subjects in the study. Long's (1969) contention of defensive responding on the part of minority students on self concept instruments so that they may present a more favorable personality pro- file prompted Samuels (1977) to consider it a potential measurement and research problem. The practice of present- ing an overrated self-description in order to hide personal feelings of inadequacy is one of the problems of self-report measures of personality. 1 Wenland (1968) compared TSCS scores obtained from a sample of almost seven hundred Black and White eighth graders and reported significantly more positive self con- cepts for the Black group in the sample. Although there was no significant difference in Self-Criticism scores for the two groups, the Black subsample was found to have sig- nificantly higher Defensive-Positive, Conflict, and Varia- bility scores. Thompson (1972) reviewed Wenland's study and others making similar comparisons. He concluded that 21 Blacks "appear to be more defensive in describing themselves than their white peers, and that they present self concept profiles characterized by high Conflict and Variability Scores" (p. 26). Thompson added that it is difficult to evaluate the meaning of a Total Positive score due to the presence of defensiveness; and that in cases where there is no significant difference between Black and White samples on Self-Criticism or Defensive-Positive scores, Blacks have been shown to have lower Total Positive scores than Whites. With particular reference to Mexican Americans and de- fensiveness, Thompson (1972) cites only one study-~Healey's dissertation at New Mexico State University. Healey (1974) found that both Mexican and Black Americans were signifi- cantly more defensive than Whites. The two minority groups obtained significantly lower Self-Criticism and higher Defensive-Positive scores on the TSCS. The TSCS was also utilized by Williams and Byars (1968) in comparing the self concepts of Black and White high school seniors. They reported that Blacks scores signifi- cantly higher on the Defensive-Positive Scale and signifi- cantly lower on the Self-Criticism Scale, indicating more defensiveness in self report. Self Concept and Related Variables Grade Level The research literature on the relationship between age and self concept suggests that there is a decline in 22 self concept scores associated with chronological age, as students progress through school. Hamachek (1971) contended that self-perceptions of stu- dents tend to become progressively more negative with age. Cicirelli's (1977) study reported a decline in self concept scores from grades one to three on the PURDUE SELF CONCEPT SCALE. This particular study controlled the variable of socioeconomic status. Chang (1975) reported that scores on the PIER-HARRIS CHILDREN'S SELF CONCEPT SCALE (PHCSCS) de- clined with age. The findings were based on samples of Black and Korean American third through sixth grade students. A number of other studies have reported a negative cor- relation between self concept and age. Among them are those of Piers and Piers (1964), Long and Henderson (1968), Trow- bridge (1972), and Klein (1975). Morse (1963) reported that confidence about school work is negatively associated with chronological age. On the other hand, Carpenter and Busse (1969) reported that obtained self concept scores for two samples of elementary school students were significantly lower for first graders than fifth graders. Similar find- ings were obtained by Rogers (1977). Thompson's (1972) review on the relationship between the self concept (as measured by the TSCS) and age led him to conclude that "for researchers, these findings mean that age is a variable which must be controlled or accounted for in some fashion” (p. 20). 23 Sex Studies on the relationship between general self concept and sex suggest that females as a group have higher self con- cepts than males. Hamachek and Conley (1968) found that fe- male adolescents' self-perceptions became more positive from grades six to twelve, while for males the progression from the sixth through the twelfth grade was correlated with de- clining self concept scores. Campbell's (1966) findings re- vealed more positive self concepts among females than males; and Perkins (1958) reported that fourth and sixth grade fe- males had greater self-ideal congruence than males. With reference to achievement and self concept of abi- lity, the effect of sex is not clear. Fink (1962) found a stronger correlation between achievement and self concept for males than for females. A similar finding was reported by Bledsoe (1967) in a sample of fourth and sixth grade stu- dents. Using a different measuring instrument, Brookover, Patterson, and Shailer (1962) found that seventh grade fe- male students had significantly higher self concept scores than male counterparts on a self concept of ability measure. Black males have been reported to have higher self- esteem than Black females (Long and Henderson, 1968). Car- penter and Busse (1969) found that first grade Black females had significantly lower self concept scores than White fe- males in that same grade, and that Black females showed more negative perceptions than White females, indicating an ethnic rather than a sex difference. 24 The research findings indicate that the variable of sex is one to be taken into account in the design of self concept studies since it can potentially contaminate findings and conclusions. Socioeconomic Status The relationship between self concept and socioeconomic status has not been clearly established. Wylie's (1961) re- view of the research failed to reach any conclusion regard- ing the two variables. The author contended that the avail- able studies provided no conclusive evidence of relationship. However, two years later Wylie (1963) reported that self concept differences between Blacks and Whites are no longer present when the factor of socioeconomic status is controlled. This would lead the researcher to think that there is indeed a relationship, and that the socioeconomic variable may even be more important than that of racial or ethnic group mem- bership. In a chapter entitled "Race, Social Class and Child Self Concept," Samuels (1977) reviewed studies which had at least some degree of control for race and social class. This author concluded that social class appears to be a more significant factor in determining self concept than race. Kinch (1963) and Gale (1969) have emphasized the rela- tionship between self concept and social, cultural, and eco- nomic variables. Leonetti and Muller (1976) stated that 25 factors such as skin color, language, socioeconomic variables, and cultural characteristics seem to be responsible, at least partially, for distorted self- perceptions in this culturally different group (Mex- ican Americans." (p. 250) Cicirelli (1977) administered the PURDUE SELF CONCEPT SCALE in an attempt to determine the relationship between ethnicity, socioeconomic status and self concept. A sample of 345 lower elementary school pupils of low socioeconomic status (about half of whom were Black Americans) was used. The results showed that: 1) Blacks had higher self concept scores than Whites; 2) self concept scores declined with ascending grade levels; 3) children (primarily Black) from families receiving welfare assistance scored significantly higher than non-welfare children; and 4) the self concept scores of non-welfare Blacks were almost precisely the same as for non-welfare Whites. Cicirelli eliminates the possibility of the often pre- sumed generalized effect of the Black Power Movement on Black pride due to the finding that only welfare status Blacks showed higher self concept scores than Whites. How- ever, it was not possible to determine the effects of wel- fare status on self concept since there was not a sufficient number of welfare status Whites in the sample. The author also rules out segregation by socioeconomic status as a variable because students in the sample shared residential neighborhoods. Lower parental and teacher expectations was suggested in the above study as a possible interpretation, but 26 Cicirelli (1977) added that this must be empirically tested. Results were interpreted in terms of "defensiveness in the testing situation." External locus of control scores gen- erally observed with low socioeconomic status samples was also a factor considered in the interpretation (Ducette and Wolk, 1972). The author further notes that by holding external forces responsible for their situation, these children are able to defend the self against the negative evaluation which would result if they held themselves responsible for their poor situation and achievement. A study comparing Black and White elementary school children of welfare mothers on measures of self concept was undertaken by Carpenter and Busse (1969). Subjects were forty first grade children and forty fifth grade children. One-half of each grade sample was White, the other half was Black. Subgroups were equally divided for sex. All sub- jects were from father-absent, welfare recipient families and lived with their natural mothers (only one child per family was eligible for random selection). The instrument used was the WHERE ARE YOU GAME which is based on a five- point scale and is composed of seven bi-polar dimensions (Engle and Raine, 1963). Carpenter and Busse (1969) reported that: 1) race dif- ferences on overall self concept scores were not significant; 2) in the first grade, Black females had significantly lower self concept scores than White females; 3) Black females had more negative self-perceptions than Black males; 4) self concept scores were significantly lower for students in 27 grade one than for those in grade five. Results showed that when socioeconomic status is equated, race differences on self concept measures tend to disappear. Chang (1975) conducted a self concept study using Korean American (U.S. born) children whose parents were first gener- ation immigrants and Black American children of both sexes who were in grades three, four, five and six in a large metropolitan area school system. The sample was drawn from a middle class section of the city and included 151 Korean American and 144 Black American children, with almost equal proportions of males and females. The PHCSCS was the instru- ment administered. The results of the above study revealed that Korean children had higher average scores on three of the six mea- sures (Behavior, Intellectual and School Status, and Happi- ness and Satisfaction). Black children scores higher on two measures (Popularity, and Physical Appearance and Attri- butes); on the sixth measure (Anxiety), there was no signi- ficant difference between the two ethnic groups. The mean scores of both minority groups in Chang's study exceeded the mean of the norm group (51.84). Korean Americans had a mean score of 62.14 while Black Americans showed an average score of 57.20. The authors interpreted these findings in terms of more favorable attitudes of Korean parents toward education and child rearing practices. They concluded that "Korean American and Black American children do not necessarily suffer from lower self-concept 28 and a lower sense of personal worth than children in the norm groups" (p. 57). A study by Davidson and Greenberg (1967) reported the finding that high achieving low socioeconomic status stu- dents viewed themselves more positively in terms of per- sonal, social, and academic competence than their low achieving peers. Wylie (1963) and Long and Henderson (1968) both re- ported a positive relationship between the socioeconomic status of the family and the self concept of children. They found that children from economically disadvantaged home backgrounds tend to have significantly lower self concept scores than children from more affluent home environments. Contrary to the above findings, more recent investiga- tions have reported that children of low socioeconomic sta- tus were found to have higher self concepts than middle class children. Among them are studies by Soares and Soares (1968, 1970), Green and Rohwer (1971), Zirkel and Moses (1971), and Trowbridge (1972). The presently available research does not consistently indicate a directional relationship between socioeconomic status and self concept. Nevertheless it is evident that this variable must be taken into account in the design of every study of self concept since there are indications of its potential contribution to results. On the other hand, it is possible that the causes for negative self concept development are, as Purkey (1970) has noted, "more psycho- logical than economic" (p. 19). 29 Achievement The relationship between self concept and academic achievement has been one of considerable interest to edu- cational psychologists during the last two decades. In ad- dressing this particular relationship, Gill (1969) concluded that "the importance of the self concept in the educational process seems to need more emphasis than is presently given to it" (p. 6). Purkey's (1970) review of the literature pointed out that a positive relationship exists between self concept and academic achievement. Unsuccessful students are more likely to hold negative self concepts than those who are successful in school. Hamachek (1971) asserted that "there is a mounting body of evidence to suggest that a student's performance in an academic setting is influenced in both subtle and obvious ways by his self-concept" (p. 184). Among the first to relate self concept development to educational achievement were Prescott (1938) and Lecky (1945). Prescott believed that the best guarantee that we have that a person will be able to deal with the future effectively is that he has been essentially successful in the past. People learn that they are able, not from failure, but from success. (p. 53) Lecky emphasized students' perceptions of self-confidence and their relationship to achievement in areas such as spel- ling. Lecky (1945) indicated that a student's self concept determines his level of achievement in school, and that 30 early school experiences are important in the development of the self concept and subsequent academic performance. His interpretation suggests that there is a causal rela- tionship, although in actuality the underlying causes are not clear. It is possible that the self concept is a re- flection of past achievement (success or failure) or that achievement is at least partially determined or affected by self concept. Most research suggests, however, that the self concept does have at least some effect on future aca- demic performance (Purkey, 1970; Hamachek, 1971). Purkey (1970) points out that although the data do not provide clear-cut evidence about which comes first--a positive self-concept or success, a negative self-concept or scholastic fail- ure--it does stress a strong reciprocal relationship and gives us reason to assume that enhancing the self- concept is a vital influence in improving academic performance. (p. 27) A positive significant correlation between self concept and grade point average (GPA) among third and sixth graders has been reported by Bruck (1958). Using the DRAW-A-PERSON- TEST as the self concept instrument, Bodwin (1957) found a negative and significant relationship between self concept and reading disability in a sample of elementary school sub- jects. Underachievers have been found to have lower self con- cepts than achievers (Shaw, 1961). Shaw, Edison and Bell (1961) reported that male underachievers were found to have lower self concepts than male achievers, although the same phenomenon was not evidenced among females. This finding 31 was obtained from a sample of high school students and in- telligence was a controlled variable. Fink's (1962) study suggested that among adolescents, achievers have higher self concepts than underachievers, as determined by a number of different self concept measures. A study by Farquhar (1968) reported that in a high school sample of students with junior standing, those students con- sidered high in academic productivity exhibited higher self concepts than those measuring low on this criterion. Irwin (1967) reported a positive relationship between self concept and achievement in a sample of first-year college students. Research by Campbell (1966) in an elementary school setting also revealed a positive relationship between self- concept and academic achievement; and Dyson (1967) reported that young adolescents who were considered high achievers had higher self concepts than their low achieving counter- parts. This same finding, based on a similar sample, was reported by Farls (1967). The research of Brookover and his associates at Michigan State University has had a significant impact on the study of the relationship between self concept and achievement. Rather than referring to the self concept in general or global terms, they have made a significant advance in de- fining the self concept as a construct and in clarifying its relationship to academic achievement. Brookover et al. (1965) defined the self concept of ability as "those definitions a student holds of his ability 32 to achieve in academic tasks as compared to others" (p. 13) and postulated that the "self-concept-as-a-learner" is a better predictor of achievement than intelligence. More importantly, modification of this construct through positive feedback from others leads to improved achievement in aca- demic work. Brookover, Erickson and Joiner (1964) found that there is a positive and significant relationship between self con- cept and academic success. An earlier study (Brookover et al., 1962) reported that the average self concept of ability for high achieving groups was significantly higher than for low achieving groups. In both studies intelligence was a controlled variable. Brookover has asserted that by enhancing the self con- cept, achievement can be raised to higher levels for most students, and rejects the notion of fixed limits in student potential or capacity (Brookover et al., 1966). Basing his reasoning on the theoretical orientation of Mead (1934), Sullivan (1947), Combs and Snygg (1959), Kinch (1963), and other theorists, Brookover emphasized the development of a "self-concept of ability acquired in interaction with sig- nificant others" (parents, teachers, friends) and "the gen- eral hypothesis is that the functional limits of one's abi- lity are in part set by one's self-conception of ability to achieve in academic tasks relative to others" (1966, p. 2). The research of Brookover and his associates indicates that a student's negative self concept of ability to achieve 33 academically can detrimentally affect his/her intellectual performance and that directly working with academic signi- ficant others (e.g. parents) can lead to self concept en- hancement which, in turn, positively affects academic achievement. Brookover and Gottlieb (1964) have contended that If the child perceives that he is unable to learn mathematics or some other area of behavior, this self-concept of his ability becomes the function- ally limiting factor of his school achievement. (p. 469) The voluminous research of Brookover and his associates suggests that a high self concept of academic ability is a necessary but not sufficient condition for superior academic performance. A large number of those high on self concept of ability do not perform well on achievement measures, and students with low self concepts of ability seldom perform high on achievement measures. Thus, a high or positive self concept of ability will not, in itself, guarantee achievement although it can facilitate it. A significant number of other studies have consistently indicated a positive relationship between self concept and achievement. Among them are those of Coopersmith (1959), Brookover and Thomas (1963), Coleman (1966), Davidson and Greenberg (1967), Williams and Cole (1968), and Chaplin (1968). The evidence is not totally conclusive, however, Holland (1959) reported that underachievers were found to have more positive self concepts than achievers. 34 It appears that the self concept of ability is related to factors other than academic achievement. For example, Harding (1966) found that it was an important variable in- fluencing high school drop-out rates for White males. There have been a small number of published studies which have addressed the relationship between self concept and academic achievement using samples of Mexican American children. A study conducted by Pruneda (1974) compared the upper and lower twenty-seven percent of scores obtained by Mexican American students (in grades six, eight and ten) on the HOW I SEE MYSELF SCALE and found that students with high self concept scores had significantly higher scores on stan- dardized achievement tests than those with low self concept scores. This was true in the sixth and eighth grades but not in the tenth. This author also reports that there was no significant difference between achievement scores and degree of Anglo acculturation, as measured by the ACCULTUR- ATION SCALE. Zimmerman and Allebrand (1965) reported that low socio- economic status Mexican American fourth and sixth graders who evidenced low reading achievement had unfavorable self concepts, as determined from profiles of the CALIFORNIA TEST OF PERSONALITY. A study conducted by Anderson and Johnson (1971) re- vealed the complex relationship between self concept, achievement outcomes, motivation, and parental practices for Mexican American students. Findings of this study 35 suggest that the student's self concept of ability, based on perceptions of evaluation from peers (and not parents or teachers) is the most predictive variable in mathematics and English achievement. The finding reported in the above study is not consis- tent with those of Brookover et a1. (1965) and Thomas (1964). These two studies reported that parental evaluations of aca- demic ability of their children were found to be more impor- tant than those of an expert or counselor in affecting self concept of ability and GPA of students. They observed posi- tive outcomes after parental perceptions were positively modified. However, the positive effect no longer operated after the treatment period elapsed. Kleinfield (1972) studied the relationship between aca- demic self concept and students' perceived evaluation by parents and teachers using a sample of Black and White high school students. She found that for Black students (espe- cially females), the relationship between teachers' per- ceived evaluation and academic self concept is higher than for Whites. On the other hand, White subjects' perceived evaluation by parents showed a higher correlation with the academic self concept measure. Thus, White students' aca- demic self concept appears to be more related to parental evaluation; while for Blacks, perceived teacher evaluation seems to have a greater effect on the academic self concept. White students may place greater importance on the evalua- tions of their parents, who, on the whole, have attained 36 higher levels of education than parents of Black pupils. It was suggested that through positive reinforcement and non-threatening atmosphere, teachers, rather than parents, may be more powerful agents in improving the academic self concept of Black students. One further study by Coleman et a1. (1966) is worthy of note. An important finding was that although the self concept was a good predictor of achievement for White stu- dents, this was not the case for Black students in the sample. This indicates that ethnic group membership may influence the relationship between self concept and achieve- ment. The available research on the relationship between self concept, achievement, and racial/ethnic group member- ship is inconclusive. There exists the need to examine the relationship between these variables and to determine if there are significant interactions. CHAPTER III METHODOLOGY This chapter defines: l) the hypotheses to be tested; 2) description of the population; 3) sampling procedure; 4) instrumentation; 5) independent variables; and 6) sta- tistical procedure. Hypotheses The following major hypotheses are tested in this study: Self Concept (Total Positive) 1. There will be no difference in mean Self Concept' scores of Black, Mexican and White American sec- ondary school students (P:.05). 1a. There will be no significant race by sex interaction (p:.05). . 1b. There will be no significant race by grade interaction (p:.05). lc. There will be no significant race by achievement interaction (p1.05). ld. There will be no significant race by socio- economic status interaction (p:.05). Self Criticism 2. There will be no difference in mean Self Criti- cism scores of Black, Mexican and White American 37 38 secondary school students (pi.05). 2a. There will be no significant race by sex interaction (p:.05). 2b. There will be no significant race by grade interaction (p:.05). 2c. There will be no significant race by achievement interaction (pi.05). 2d. There will be no significant race by socio- economic status interaction (P:.05). Defensiveness 3. There will be no difference in mean Defensive Positive scores of Black, Mexican and White American secondary school students (p1.05). 3a. 3b. 3c. 3d. There will be no significant race by sex interaction (p:.05). There will be no significant race by grade interaction (pi.05). There will be no significant race by achievement interaction (p:.05). There will be no significant race by socio- economic status interaction (Pi-05)- Description of Population The study was conducted in a medium-sized Midwestern, labor-intensive and heavily industrialized metropolitan area. Samples were drawn from the population of Black, Mexican and White American high school students enrolled in that public school system. 39 The total student population of the above school dis- trict is slightly over 18,000 students. The high school population (grades 10 through 12) for the district is 3,741 students. The percentages by grade are: tenth--43 percent, eleventh--30 percent, and twelfth--27 percent. The ethnic composition of the high school population in the district is as follows: White-~48 percent, Black-~ 37.4 percent, Latino (Hispanic) American--8 percent, and other-~6.6 percent. It should be noted that the minority enrollment for this district is significantly higher than the total minority proportion for the state. SamplingAProcedure The sample was selected from the population of students in the tenth, eleventh, and twelfth grades during the mid- part of the 1978-1979 academic year. The sampling proce- dure included both high schools within the school district described above. The basis for selection of classrooms was aimed at achieving a balanced sample with reference to sex, ethnic group and curriculum. Class lists for each high school classroom in the district were examined and two classrooms from each of the two schools at each of the three grade levels were selected from the population. Particular at- tention was given to insure adequate representation of Spanish-surnamed students in the sample since in actuality they constituted less than eight percent of the total high school population of the district. 40 The following classes were selected from the general (mainstream) education curriculum for the district: health (2); science (2); government (2); history (2); economics (2); vocational education (2). Each selected classroom had at least four Latino students formally enrolled. Instrumentation Shavelson, Hubner and Stanton (1976) have addressed the major problems inherent with self-report techniques. Among them were response sets and the unwillingness to reveal pri- vate information. Zirkel and Gable (1977) added that there was a lack of reliability and validity data on self concept measures, particularly when used with students who are from different ethnic groups. The authors warned that consider- able attention must be given to the selection of self con- cept evaluation instruments by researchers. Zirkel and Gable also cautioned researchers about cultural considera- tions. There are a number of self concept instruments which have been utilized in its assessment. Some are commercially available measures on which psychometric data has been ob- tained and have to some degree been standardized. Among these are the TENNESSEE SELF CONCEPT SCALE, the CALIFORNIA PSYCHOLOGICAL INVENTORY, the SELF ESTEEM INVENTORY, the THOMAS SELF CONCEPT VALUES TEST, the SELF APPRAISAL SCALE, the HOW I SEE MYSELF SCALE, the WAY I FEEL ABOUT MYSELF SCALE, the PIERS-HARRIS SELF CONCEPT SCALE, and the MICHIGAN STATE SELF CONCEPT OF ABILITY SCALE. 41 There are also a number of instruments which have been developed "locally" for use in certain research applications or for immediate administration in educational programs. A serious problem with these measures is that.they do not pro- vide the user with sufficient pertinent theory or data. In addition to the above there are other methods such as the Q-sort and Semantic Differential Techniques which may be applied for a particular purpose or situation. A common problem confronting the researcher then becomes one of deciding which of the available instruments will be utilized in a particular study. TENNESSEE SELF CONCEPT SCALE The TENNESSEE SELF CONCEPT SCALE (Fitts, 1965) is a self-administered instrument consisting of 100 items on which the individual is requested to respond to self- descriptive statements by choosing one of five categories ranging from ”completely true" to "completely false." There are two forms of scoring and obtaining scores or profiles: the Counseling Form (C), and the Clinical and Research Form (GER). The CGR form provides a larger number of scores than does the C form. This inventory requires approximately ten to twenty minutes of the subject's time to complete (mean time is thirteen minutes). It is intended to be used with indivi- duals twelve years or older, since it requires approximately a sixth-grade reading level. Computer scoring services are available through the publisher. 42 Peter M. Bentler and Richard M. Suinn have reviewed the TSCS in Buros' The Seventh Mental Measurements Yearbook (1972). Bentler stated that "the various content areas are well-conceived, and the scale yields a vast amount of in- formation" (p. 366). Persons varying in age, sex, race, and socioeconomic status were used in the norming sample, although those aged 12 to 30 years were overrepresented. Test reliability is in the high .80's and Bentler reported that this is "sufficiently large to warrant confidence in individual difference measurement" (p. 366). Suinn argued that "although the scores seem to have certain content validity, there has been little work dir- ected toward empirical validation of individual scores" (p. 368) and thus, behavioral correlates have not been clearly established. He concluded that the "TSCS ranks among the better measures combining group discrimination with self concept information" (p. 369). The TENNESSEE SELF CONCEPT SCALE is a particularly useful instrument for the measurement of defensiveness. A measure of overt defensiveness is obtained through a Self Criticism Score. This measure is derived from ten items which are part of the MINNESOTA MULTIPHASIC PERSONALITY INVENTORY's L-Scale. In addition, the TSCS provides a second scale which is purportedly a more subtle measure of defensiveness than the Self Criticism Score (Thompson, 1972). This second scale is referred to as the Defensive Positive Scale (DP), and is based on 29 items. Low Self Criticism 43 scores and High Defensive Positive scores are both indica- tive of defensiveness in self-report. There are other scales (e.g. Variability, T/F and Conflict Scores) which are helpful in the interpretation of data. _The reader should refer to Appendix A for a listing of scores and their interpretation. Information Questionnaire Each student in every selected classroom was adminis- tered an Information Questionnaire which requested the fol- lowing information: 1) I am: Female Male (circle one) 2) My age is: years months 3) I am in the: 10th grade 11th grade 12th grade (circle one) 4) My ethnic group membership is: Black, Afro or Negro American Spanish, Hispanic, Mexican or Latin American White, or Anglo American Please specify (if you cannot classify your- self in above groups) 5) I live with my: mother only father only mother and father guardian-~please specify relationship to you: 6) 7) 8) 9) 44 What is the occupation of your: (be as specific as you can) Father Mother Guardian The highest school year completed by my father was: (circle one) Elementary School Middle School High School 4 5 6 7 8 9 10 ll 12 College 1 2 3 4 5 6 7 The highest school year completed by my mother was: (circle one) Elementary School Middle School High School 4 5 6 7 8 9 10 ll 12 College 1 2 3 4 5 6 7 The highest school year completed by my guardian: (circle one) Elementary School Middle School High School 4 5 6 7 8 9 10 ll 12 College 1 2 3 4 S 6 7 I plan to complete the following grade: (circle one) 8 9 10 11 12 13 Vocational-Technical Degree 14 Junior College Degree 16 Bachelor's Degree 18 Master's Degree 20 Ph.D., L1.D., or M.D. Grade Point Average (GPA): 4S Variables in the Study There was a total of five independent variables and 29 dependent variables. The dependent variables were the scores on the TSCS. The three hypothesized dependent var- iables were: 1. Total Positive Score on the TSCS 2. Self Criticism Scores on the TSCS 3. Defensive Positive Score on the TSCS The remaining 26 dependent variables were the other scores on the Clinical and Research (CGR) version of the TSCS. The independent variables were: 1. Racial/Ethnic Group Membership (three classifica- tions) 2. Sex (two classifications) 3. Grade (three classifications) 4. Achievement (eight classifications) 5. Socioeconomic Status (nine classifications) Students indicated their racial/ethnic group member- ship, sex, and grade level on the Information Questionnaire. The achievement measure was obtained from computerized records which indicated career grade point average (GPA) for each student. The district in which this study was con- ducted has compiled the GPA measure of achievement since the beginning of the seventh grade for each student, or since the student was enrolled in the district. Students in the sample were provided with this information upon hav- ing completed the TSCS and the Information Questionnaire 46 and this measure was entered at a designated location on each Information Questionnaire. Socioeconomic status was determined through the appli- cation of the HOLLINGSHEAD TWO FACTOR INDEX.OF SOCIAL POSI- TION (1956). This measure is also referred to as the INDEX OF SOCIAL POSITION (ISP). The ISP uses the occupational role and formal educational level of the head of the house- hold in determining the social position of an individual or member of a family unit. In cases where both parents were employed outside the home, the head of the household was defined as the parent with the higher score. The range of obtainable scores on this scale is from 11 to 77. Scores are negatively correlated with socioeco- nomic status. The score distribution was truncated in in- tervals of seven score points, yielding a total of nine functional categories. The resulting distribution was negatively skewed due to the narrow range of occupational and educational status of the population under investiga- tion. Statistical Procedure Data for each subject on all dependent and independent variables were punched on computer cards. There were two cards per subject. A Statistical Package for the Social Sciences (SPSS) file was established for convenient access to data for statistical treatment. The analysis was con- ducted at the computer laboratory facilities of Michigan 47 State University which uses a Control Data Corporation sys- tem (Models 6400 and 6500). The following statistical procedures were programmed: l. One-way analysis of variance (ANOVA) between all independent and dependent variables 2. Two-way ANOVA's of all hypothesized relationships between variables. The results from these analyses are presented in the following chapter. CHAPTER IV RESULTSl This chapter reports the statistical findings in rela- tion to: l) the three research hypotheses, and Z) other variables included in the study. Test of Research Hypotheses Hypothesis 1: There will be no difference in mean Self Concept scores of Black, Mexican and White American secondary school students (p<.05). Table 1 reports the result of a one-way ANOVA procedure between the Total Positive Scores on the TSCS and ethnic group membership. The null hypothesis was thus not rejected on the basis of this statistical analysis. The analysis does not indicate a significant statistical difference between the means obtained by the three groups on Total Positive Scores. However, the reader may note that: l) the mean score for Mexican Americans was lower than for Blacks or Whites, and 2) the mean scores for all three groups were lower than the normative data (X=34S.57) provided by the test TSCS publisher in the Manual (see Fitts, 1965, p. 14). Tests for interaction were conducted using the two-way ANOVA model. Tables 2 through 5 report the result of the 1A complete listing of the nature and meaning of TENN- ESSEE SELF CONCEPT SCALE scores is found in Appendix A. 48 49 TABLE l.--ANOVA of Total Positive Self Concept Scores and Race. N Mean SD Black 100 332.03 27.79 Mexican 49 327.73 35.78 White 152 334.39 30.59 Total 301 332.52 30.60 F=.89 Sig.=.41 e2=.006 TABLE Z.--ANOVA of Total Self Concept Scores by Sex and Race. Source SS df MS F Sig Sex 2884.02 1 2884.02 3.12 .08 Race 1367.04 2 683.52 .74 .48 Sex X Race 3394.67 2 1697.34 1.84 .16 Within 272909.87 295 925.11 Total 280865.11 300 50 TABLE 3.--ANOVA of Total Self Concept Scores by Race and Achievement. Source SS df MS F Sig Race 397.69 2 198.84 .22 .80 Achievement 19022.36 7 2717.48 3.03 .004* Race X Ach. 11051.50 14 789.39 .88 .58 Within 247250.50 276 895.84 Total 278599.28 299 931.77 TABLE 4.--ANOVA of Total Self Concept Scores by Race and Grade. Source 88 df MS F Sig Race 1344.62 2 672.31 .75 .48 Grade 14671.17 2 7335.58 8.14 .001* Race X Grade 1238.19 4 309.55 .34 .85 Within 263279.20 292 901.64 Total 280865.11 300 936.22 51 TABLE 5.~-ANOVA of Total Self Concept Scores by Race and Socioeconomic Status. Source SS df MS F Sig Race 736.59 2 368.29 .40 .67 SES 10935.86 8 1366.98 1.50 .16 Race X SES 15755.87 13 1211.99 1.33 .20 Within 252496.83 277 911.54 Total 280865.11 300 936.22 analyses of Total Self Concept scores by independent varia- bles. The reader will note that: 1. there were no significant interactions for the four sub-hypotheses; 2. achievement and grade level accounted for a signi- ficant proportion of the Total Positive Self Con- cept score variance, while race and socioeconomic status were not found to be significant main ef- fects. Hypothesis 2: There will be no difference in mean Self Criticism scores of Black, Mexican and White American secondary school students (p<.05). The result of a one-way ANOVA procedure indicates that there is a significant statistical difference across Self Criticism and races, as reported in Table 6. The mean Self Criticism score for Mexican Americans was lower than the mean scores for Blacks and Whites; the mean scores for Black 52 and White subjects were not significantly different. TABLE 6.--ANOVA of Self Criticism Scores by Race. N Mean ' SD Black 100 36.23 5.22 Mexican 49 34.59 6.05 White 152 36.95 5.17 Total 301 36.33 5.39 F=3.63 Sig.=.021* e2=.023 Two-way ANOVA analyses (Tables 7-10) revealed that: l. the variables, race, sex and achievement had a significant effect on Self Criticism scores; 2. there was no significant interaction between race and the other four independent variables; 3. the variables of grade level and socioeconomic status were not found to be significantly asso- ciated with Self Criticism scores. Hypothesis 3: There will be no difference in mean Defensive Positive Scores of Black, Mexican and White American secondary school students (p<.05). A one-way ANOVA of Defensive Positive Scores by race resulted in non-significant F ratios (Table 11). Although Mexican Americans had the highest Defensive Positive Scores (and a higher standard deviation) and Whites the lowest, 53 the F ratio did not approach significance and the null hy- pothesis was thus not rejected. TABLE 7.--ANOVA of Self Criticism Scores by Race and Sex. Source SS df MS F Sig Race 179.47 2 89.74 4.00 .04* Sex 186.40 1 186.40 3.18 .01* Race X Sex 5.63 2 2.82 .10 .91 Within 8313.09 295 28.18 Total 8712.09 300 TABLE 8.--ANOVA of Self Criticism Scores by Race and Grade. Source 55 df MS F Sig Race 190.70 2 95.35 3.33 .04* Grade 76.33 2 38.17 1.33 .27 Race x Grade 71.80 4 17.95 .63 .64 Within 8356.99 292 28.62 Total 8712.09 300 54 TABLE 9.--ANOVA of Self Criticism Scores by Race and Achivement. Source SS df MS F Sig Race 78.03 2 39.01 1.40 .25 Achievement 402.05 7 57.44 2.06 .05* Race & Ach. 411.42 14 29.39 1.06 .40 Within 7691.29 276 27.87 Total 8706.67 299 29.12 TABLE 10.--ANOVA of Self Criticism Scores by Race and Socio- economic Status. Source SS df MS F Sig Race 106.78 2 53.39 1.90 .15 SES 360.74 8 45.09 1.61 .12 Race X SES 378.97 13 29.15 1.04 .41 Within 7765.42 277 28.03 Total 8712.09 300 29.04 55 TABLE ll.--ANOVA of Defensive Positive Scores by Race. N Mean SD Black 100 56.81 ' 9.62 Mexican 49 58.24 13.52 White 152 ' 54.91 9.99 Total 301 56.08 10.57 F=2.22 Sig.=.11 e2=.014 The two-way ANOVA procedure (Tables 12-15) revealed the following statistical findings: 1. race was not found to be a significant variable affecting Defensive Positive scores; there was no significant interaction between race and the four other independent variables on the dependent variable measure; none of the four independent variables was found to be significantly related to Defensive Positive scores. With reference to the relationship between the four re- maining independent variables and the three dependent varia- bles above, the findings were: 1. Total Positive scores were found to be significantly related to achievement (Table 3) and grade level (Table 4). In both cases the relationship was positive; S6 2. Self Criticism scores were significantly related to sex (Table 7) and achievement (Table 9). Scores for males were found to be significantly lower than for females. Self Criticism scores were positively related to grade point average. 3. Defensive Positive scores were not significantly related to any of the four independent variables. TABLE 12.—-ANOVA of Defensive Positive Scores by Race and Sex. Source SS df MS F Sig Race 434.51 2 217.25 1.97 .14 Sex 263.78 1 263.78 2.40 .12 Race X Sex 280.08 2 140.04 1.27 .28 Within 32465.31 295 110.05 Total 33500.92 300 111.67 TABLE l3.--ANOVA of Defensive Positive Scores by Race and Grade. Source SS df MS F Sig Race 471.67 2 235.83 2.15 .12 Grade 491.11 2 245.56 2.24 .11 Race X Grade 527.44 4 131.86 1.20 .31 Within 31990.61 292 109.56 Total 33500.92 300 111.67 57 TABLE l4.--ANOVA of Defensive Positive Scores by Race and Achievement. Source SS df MS F Sig Race 613.09 2 306.55 2.82 .06 Achievement 640.73 7 91.53 .82 .55 Race X Ach. 2208.73 14 157.77 1.45 .13 Within 29980.62 276 108.63 Total 33377.68 299 111.63 TABLE 15.--ANOVA of Defensive Positive Scores by Race and Socioeconomic Status. Source SS df MS F Sig Race 479.36 2 239.68 2.15 .12 SES 580.04 8 72.50 0.65 .74 Race X SES 1558.50 13 119.88 1.08 .38 Within 30870.63 277 111.45 Total 33500.92 300 111.67 58 Relationships Between Other Variables in the Desigg As noted in Chapter III, the design involved a total of five independent and 29 dependent variables. The independent-dependent variable matrix yielded a total of 145 cells. Of this total, 113 units are reported in this section and 58 were found to be statistically significant (the overall alpha level was not controlled and thus ap- proximately seven or eight of these could be expected to be significant). Since these relationships were not hypothe- sized, they are not discussed individually in this chapter. Tables 16 through 18 report the F ratios and significance levels of results obtained through one-way ANOVA procedures for those cells where p:.05. For purposes of logical organization the findings have been grouped as follows: 1. Self Concept From an Internal Frame of Reference (Table 16); 2. Self Concept From an External Frame of Reference (Table 17); 3. Empirical Scales and Race (Table 18); and 4. Response Characteristics (Tables 19 and 20). Self Concept From an Internal Frame of Reference The four scales within this category are: Basic Iden- tity, Self Satisfaction, Behavior Perception, and Variation across the three scales. Table 16 reports that race was a significant variable only with regard to Variation (Black students' scores were significantly more variable). 59 .Ho>o~ oomum no“: ommonoew mohoom .mHHoo acmowmficmwm Ham sou N Benefimaemem eozH ooo. uNo oooo. um om.o~um em.ofinz mo.o~n2 .mnz .m.z .m.z .m.z Ho.n~um cofiumflsm> ocouu flow -mewemoe 0: who. "N0 BMC. "N0 moo. um moo. n newuaooaoo .m.z o.mum Nw.mnm .m.z .m.z pom>mnom oM:o< Novena xom comm -oocH An oocohomom mo osmhm Hangouco cm .moaomwhm> uaoocoo scum mouoom Homoeoo wHom mo <>oz<--.oH momo~ ocean gnu: omeohoew mouoom .m~_oo useufiuueuum -a you ~ ueauuuueuwm noz— o~o. IND mnoo. um oo.o nu _o.nnum .m.z .m.z .m.z oo.o~nz .m.z cofiuauue> mum gun: owgu< wooeuo new one: .mouaamhe> “zoo -eoooo:_ so ooeououox me cause _aetouxm so so»; museum Homoeoo ouom we <>oz<--.s_ m4oo«:u< compo xom eon: .moaeeaee> seaweeaeee_ A; moms use me museum senataasm we <>oz<--.a~ mam nno. n~o eve. .. ee.. .a see. .No en.e~_.z Aaom ee.e~_.a Race. .8 ea.en_.: emeeanezo .m.z .m.z .m.z ~m.en~.z an.a .a se.ae~.n ease: a\e mmm acolo>omso< compo xom cum: .munnumhw> UEOVSOQDQGH >5 mUWF OSH CO mUMUmMHOUuahdgu OmGOQmflm m.uUflH£=m HO <>Cz