Capyright by DAVID ALLEN PAYNE 1961 A DIMENSION ANALYSIS OF THE ACADEMIC SELF-CONCEPTS OF ELEVENTH GRADE UNDER- AND OVERACHIEVING STUDENTS BY DAVID ALLEN PAYNE A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1961 Chapter II. III. TABLE OF CONTENTS TABLE OF CONTENTS . LIST OF TABLES LIST OF FIGURES . . ACKNOWLEDGEMENTS. ABSTRACT FORMULATION AND DEFINITION OF THE PROBLEM . Purpose of the Study . Need for the Study . . Statement of the Problem . Academic Self- -Concept Theory . The Hypotheses . . Statement of the Hypotheses . Organization of the Study. REVIEW OF RELATED RESEARCH LITERATURE . Theoretical Considerations of Self-Concept . Definition of Self-Concept Measurement of Self-Concept Dimensions of Self-Concept . . Self— —Concept Related to Academic Achievement . Sex Differences in Self- -Concept . Self- -Concept, Motivation, and Level of Aspiration Behavior . Conclusions and Interpretation . Summary DESIGN AND METHODOLOGY Instrumentation Sample Selection . . Item.Ana1ysis Procedures Selection of Appropriate Multivariate Model. Multiple Scalogram Analysis Procedures Advantages of MSA . . . Criteria for Homogenity of Sub-Matrices . Assumptions of MBA Model Reproducibility (Reliability) of Sub-Matrices . Procedural Steps in MBA . Summary . . . . a . . . . ii Page ii iv vi vii . viii OOQJ-‘WNN 10 10 12 13 14 16 23 25 28 29 3O 30 31 34 35 36 37 38 39 41 41 43 Chapter IV VI. ANALYSIS OF THE DATA Item Analysis Results Multiple Scalogram Analyses . Results of Multiple Scalogram.Ana1ysis - Males Results of Multiple Scalogram Analysis - Females. Orthoganality of the Dimensions Summary . . . . . . . INTERPRETATION AND DISCUSSION OF THE DIMENSIONS . Discussion of Male Dimensions Discussion of Female Dimensions Discussion of Sex Differences Interpretation of Dimensions In Relation to Theory . Summary SUMMARY, CONCLUSIONS AND IMPLICATIONS . BIBLIOGRAPHY APPENDIX A. Tables of Summary Characteristics of Students Representing Academic Extremes . B. The Word Rating List . iii Page 44 44 - 46 47 52 57 58 6O 60 63 65 67 7O 72 79 87 119 Table No. 2. 1 .10 .11 .12 .13 LIST OF TABLES Dimensions of Self-Concept Isolated in Previous Investigations . . . . . . . . . . Summary of Mean "Self Concept of Ability" Scores of Different Achievement Groups (Brookover e.a1.) Summary of Correlations Among "Self Concept of Ability"; Grade Point Average; and "Importance of Grades" for Males (N-513) and Females (N-537) (Brookover et.a1.) . . . . . . . . . . Summary of Sample Sizes Used in Validation and Cross-Validation of the Word Rating List . Summary of Item.Ana1ysis of the Word Rating List Summary of Hoyt's Reliability Estimates on the Word Rating List for Different Samples by Sex . Item Content of Dimension I (Males). Item Content of Dimension II (Males) Item Concent of Dimension III (Males) Item Content of Dimension IV (Males) . Item Content of Dimension I (Females) Item Content of Dimension II (Females) . Item Content of Dimension III (Females) Item Concent of Dimension IV (Females) . Item Concent of Dimension V (Females) Intercorrelations Among Four Male Dimensions Obtained by Multiple Scalogram Analysis(N-100) . Intercorrelations Among Five Female Dimensions, Obtained by Multiple Scalogram.Ana1ysis (N-100). Summary Dimensions in the Academic Self-Concepts of Male Under- and Overachievers (N=298) Summary Dimensions in the Academic Self-Concepts of Female Under- and Overachievers (N=312) . iv Page No. 15 18 19 44 46 49 50 51 52 53 54 55 56 56 57 58 75 76 Table No. A.1 A.2 A.3 Summary Student Characteristics of Academic Extremes . . . . . . . . . . . . . . . . Summary Personality Characteristics Associated with Academic Extremes . Summary Characteristics of Parent Child Relationships and General Home Environment of Over (High) and Under (Low) Achievers Page No. 88 100 113 LIST OF FIGURES Figure No. Page No. 3.1 Methodological Selection of Individuals with Stable Measured Aptitude . . . . . . . . . . . . 32 3.2 . Method of Selecting Under- and Overachievers . . 33 5.1 Schematic Representation of Academic Self- Concept Types in Relation to Hypothetical Generic Variables . . . . . . . . . . . . . . . 69 vi ACKNOWLEDGMENTS Patience is a virtue manifest at the time of thesis writing. This was particularly true in the completion of the present study. Acknowledgement is greatly expressed to the following "virtuous" individuals, without whose constant encouragement and guidance could this study have been completed: To Dr. William W. Farquhar, the writer's major advisor,"thera- pist":and chairman of his guidance committee. To the other members of the guidance committee, Dr. Willard G. Harrington, Dr. Gregory A. Miller, and Dr. Bill L. Kell. To Dr. Charles F. Wrigley, Donald M. Wilkins, Marion D. Thorpe, Wayne H. Chubb, John J. Paterson, and Ronald G. Taylor for technical and statistical assistance. And to his wife, Mary Ann, whose invaluable assistance took the form of "aid above and beyond the call of duty". A DIMENSION ANALYSIS OF THE ACADEMIC SELF-CONCEPTS OF ELEVENTH GRADE UNDER- AND OVERACHIEVING STUDENTS BY DAVID ALLEN PAYNE AN ABSTRACT OF A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1961 Approved 161w Lfl MflU/[I 4/ ABSTRACT A DIMENSION ANALYSIS OF THE ACADEMIC SELF-CONCEPTS OF ELEVENTH GRADE UNDER- AND OVERACHIEVING.STUDENTS by David Allen Payne The study was concerned with l) the development of an objective ”Imeasure of academic self-concept, and 2) a dimension analysis of the discriminating items. Item discrimination was determined on samples of statistically defined under- and overachieving eleventh grade students of each sex. The following assumptions were made, 1) self-concept is a functionally limiting factor in school achievement, 2) the student learns what he perceives he is able to learn, 3) significant others, particularly the teacher, have important influences on the development of a student's self-concept, in the form of expectancies, which in turn affects his ability to perform in the academic setting, and 4) under- and overachieving students have significantly different academic self-con- cepts. A one-hundred and nineteen item rating scale was developed which purported to measure academic self-concept. Using a four point.8¢81€: the student was asked to rate one, two, or three word concepts and phrases as he thought his teacher would in describing him as a student. It was found that 48 items significantly discriminated between under- and overachievers for each sex after cross-validation. Analysis of variance reliability estimates ranged from .90 to .93 for males and from .88 to .93 for females in various samples. A multiple scalogram dimension analysis of the discriminating viii David A. Payne items for each sex was performed. This technique, a generalization of Guttman's scalogram procedures, yielded four interpretable dimensions for males and five interpretable female dimensions which accounted for 96% and 98% of the cross-validated items for males and females respectively. The reproducible males dimensions were labeled as follows: D1 Achievement via Traditional Academic Role Taking D2 Achievement via Academic Conformity D3 Achievement via Intrinsic Motivation D4 Achievement via Unique Accomplishment The reproducible female dimensions were labeled as follows: D1 Achievement via Traditional Academic Role Taking D2 Achievement via Peer Normative Competition D3 Achievement via Academic Independence D 4 DS Achievement via Intellectualizing Achievement via Meeting Teacher Expectations The research was supported by funds granted by the U.S. Office of Education, as part of a study under the direction of William W. Farquhar, entitled, A Cgmprehensive Study pf the Motivational Factors Underlying Achievement g; Eleventh Grade High School Students, Research Project No. 846 (8458). ix CHAPTER I FORMULATION AND DEFINITION OF THE PROBLEM The obvious fact that a command of basic academic skills is necessary for successful scholastic performance has been repeatedly demonstrated. However, estimates of academic success based on general ability, academic skill or aptitude measures only define part of the predictive variance. In an attempt to isolate factors which might account for more variance, researchers have increased their investiga- tions of non-intellectual or personality variables related to academic achievement. Educational psychologists and sociologists have attempted to delimit the study of academic dynamics by investigating the dimension of personality labeled self-concept.1’2 The academic self-concept research trend is demonstrated by a cursory review of articles in recent psychological and educational journals. Also indicative of the growing research interest in self-concept is the increase in research grants being made available by public and private agencies, e g. the Cooperative Research Branch of the U.S. Office of Education is encouraging research relating self-concept to academic achievement by 1Merville C. Shaw, K. Edson & H. M. Bell, "The Self-Concept of Bright Underachieving High School Students as Revealed py an Adjective Check List", Personnel and Guidance Journal, 1960, Vol. ‘9, pp. 193-196 2R. M. Roth, "Role of Self-Concept in Achievement", Journal of Experimental Education, 1959, Vol 27, pp. 265-25. 2 2 3 financial support of investigations by Combs,1 Bledsoe, and Brookover. Purpose of the Study Despite the fact that researchers have attempted to relate gross measures of self-concept to academic performance, few definitive investigations have been made. Too few researchers of academic self-. concept have made the content of their self concept referents plausibly relevant to the criterion variables under investigation. The purpose of this study was to explore the academic self-concepts of eleventh grade high school students with an attempt to overcome the above limitations. Need for the Study The construdtion and analysis of such a measure of academic self-concept can be justified solely because it permits better educa- tional selection and placement. A dimensional analysis, by revealing the underlying structure which accounts for the effective functioning of the proposed instrument, may provide the foundations upon which an even T v v ~vvv ‘— 1Arthur W: Combs, The Relation§h12.g§ Child Perceptions 59 W Begvior 1.3 the garly Sghool Years, Research Project No. 814, Supported by the U. 8. Office of Education in cooperation with the University of Florida,,1959 2Joseph C. Bledsoe, 325 Self-Concepts 2; Elementary School Children ‘3 W 33 Their Academicrghievement, Intelligence, Interests. and Manifest Anxiety, Research Project No. 1008, Supported ' by the U.S. Office of Education in cooperation with the University of Georgia, 1960 3Wilbur‘W. Breakover. Relhtionship'gg Self-Images £2 Achieve- ment in Junior High 1 fighjegte. Research Project No. 845, Supported by the U. 8. 0 ice of Education in cooperation with Michigan State University, 1959. 3 more successful measure can be built. Moreover, the subsidiary value of strengthening the curriculum, teaching methods, and counseling pro- cedures by better understanding the nature of what dimensions are inherent in the academic self-concepts of students, further emphasizes the need for this study. The value of such an instrument for further research, particularly experimental research, is obvious. Statement of the Problem It was the problem of this investigation to construct, validate and cross-validate an instrument which purports to measure the academic self-concepts of male and female eleventh grade high school students. The problem was also concerned with the determination, by multiple scalogram analysis, of the psychological dimensions of such an instrument. The investigation developed as part of a research project sponsored by the United States Office of Education, under the direction of William W. Farquhar.1 This'project, hereafter referred to as the Farquhar Motivational Research Project, was concerned with the development of an objective battery of tests which measured various aspects of motivation for academic achievement. In the final analysis, toe problem.ef'the pre- sent investigation was to determine the number, nature, magnitude and structure of the dimensions of academic self-concept. The theory on which this study is based is developed as follows. v F‘— 1William W. Farquhar, .A Cmrehensive Study 9'; 5h; Motivational Factors Underlyigg_Achievement g; Eleventh Grade High School Students, Research Project No. 846 (8458); Supported by the U.S. Office of Educa- tion, in cooperation with Michigan State University, 1959 4 Academic Self-Concept Theory The four basic tenets of a theory of academic self-concept are presented in this section. Following each tenet is a discussion of its theoretical and research base. Although it is recognized that innate factors may set limits on learning ability, it is also recognized that few students achieve near the level set by inherent capacity. One factor which may func- tionally limit the learning of many students is a low or negative self- concept as a school learner. The basic theory that self-concept is a functionally limiting factor in school achievement emerges from the perceptual approach to 1 and Brookover,2 individual behavior as expressed by Combs and Snygg, and the symbolic interaction frame work of social psychology derived from George H. Mead,3 and Charles H. Cooley.4 The basic tenets of the academic self-concept are summarized by Brookover.5 I: Persons learn to behave in the ways that each perceive as appropriate to himself. 1Arthur W. Combs and D. Snygg, Individual Behavior, New York: Harper & Brothers, (Revised Edition), 1959 2Wilbur B. Brookover, "A Social Psychological Conception of Classroom Learning", Schgol and Society, 1959, Vol. 87, pp. 84-87 3George H. Mead, Mind, Self, and Society, Chicago: University of Chicago Press, 1934 4Charles H. Cooley, Human Nature and the Social Order, New York: Scribner's, 1902 5Wilbur B. Brookover, "A Social Psychological ..... ", pp. 86 5 Combs and Snygg have postulated that all behavior, "...without exception, is determined by the perceptual field at the moment of action".1 This perceptual field constitutes the universe of experience for each individual. The perceptual system gives meaning to each indi- viduaIs experiences so that people do not behave according to the facts as others see them, but according to the facts as they see them. II. Appropriateness of behavior is defined by each person through internaflization of the expecta- tion of significant others. ' Tenet Two necessitates a consideration of the development of the self-concept. The "self" arises in the process of social experience. The develOpment occurs within a matrix of social interaction. The influence of the "significant other" within this matrix is described by Head;2 It is by means of IGEexiveness--the turning back of experience of the individual upon himselfL-that the whole social process is brought into the experi- ence of the individuals involved in it; it is by such means, which enable the individual to take the attitude of the other toward himself, that the individual is able consciously to adjust himself to that process.... The mechanism by which the influence is assimilated is described by Cooley as a "looking-glass-self". The reflection of self is hypothesized as consisting of three principle parts: 1Arthur W. Combs & D. Snygg, Individual Behavior, New York: Harper & Brothers, (Revised Edition), 1959 2George H. Mead, Mind, Self, and Society, Chicago: University of Chicago Press, 1934, p. 134 6 ....the imagination of our appearance to the other person; the imagination of his judgment of that appearance, and some sort of self- feeling....1 Empirical evidence has been presented by Helper to support the hypo- thesized relationship between "significant others" and the development of an individuals self-concept.2 He found that children's self-concepts were similar to the self-concepts of their parents. Significant others are defined as those individuals from whom self-relevant data is abstracted. They may exert either a positive or a negative influence .on the individual. III: The functional limits of one‘s ability to learn are determined by his self-conception or self-image as acquired in social inter- action. 3 would From a perceptual viewpoint, if intelligence, as Combs define it, is the capacity for effective behavior, the intelligence of an individual will be dependent upon the richness and variety of per- ceptions possible to him at a given moment. However, perception is a selective process. One’s self-concept is a vital factor in determining the type of perception selected. An example of the operation of thisv' 4 tenet is presented by Combs. 1Charles H. Cooley, Human Nature and the Social Order, New York: Scribner’s, 1902, p. 152 2MalcolmM. Helper, "Learning Theory and Self-Concept", Journal _£ Abnormal and Social Psychology, 1955, Vol. 51, pp. 184-194 3Arthur W. Combs, "Intelligence From a Perceptual Point of View", Journal g£.Abnormal and Social PsychologY, 1952, Vol. 46, pp. 662-673 ifirf“ r v, Cow‘s, ”latelligcnce From ------- ” p. 168 7 ...The child who believes himself unable to read, confronted with the necessity for reading, is more likely than not to do badly. The external evaluation of his teachers, and fellow pupils, ...all provide proof to the child of how right he was in the first place. The possession of a particular self-concept tends to produce behavior that corroborates the self-concept on which the behavior originated. IV: The individual learns.what he believes significant others expect him to learn in the classroom. Theoretically the self-concept is viewed as a learned structure, growing mainly from comments made by other people and from inferences drawn by children out of their experiences in home, school and other social groups. Teachers, therefore, being significant others, can be seen as potentially having an extremely important influence on the development of a child's self-concept, and thereby his school perfor- mance. Staines, for example, demonstrated that teachers, through their roles as significant others to students, can cause actual changes in the self-concepts of their students by giving positive comments and creating an atmosphere of greater psychological security.1 Furthermore, Staines found that the interaction of teacher and student had an effect on self-concept which improved achievement. The four basic tenets of the above academic self-concept theory may be summarized as follows: 1) the child learns what he perceives he is able to learn, and 2) the teacher, as a significant other, has an 1J.‘W. Staines, "SelfePicture as a Factor in the Classroom", British.lenrnal.hf.Educatinnal.Baxcholngx, 1956, Vol. 28, pp. 97-111 8 Q important influence on the development of a child’s self-concept, which in turn affects his ability to perform in the academic setting. From the above academic self-concept theory a number of hypo- theses are generated. The Hypotheses The basic assumption was made that under- and overachieving students hold different conceptions of themselves as learners, manifested in the fact that.their academic performance is not commensurate with their measured scholastic aptitude. (See Chapter III for a statistical definition of under- and overachievement.) The further assumption was made that an instrument could be developed to measure academic self-concept, and that this instrument could be validated and cross-validated. The discriminative power of such an instrument could be determined on the basis of a dispropor- tionate selection of certain items by a given discrepant achievement group. ' It is hypothesized that psychologically meaningful dimensions exist within a measure of academic self-concept. In as much as research literature indicates sex differences in self-concept, the determination of sex differences in the academic self-concepts of males and females will form a secondary phase of this study. The over-all approach of the present investigation is not toward the classical form of hypothesis testing. Concern is generally with exploration, with a view toward redefinition and expansion of 9 _ theory. The only significance test applied will be those involved in item analyses. Statement 9: the Hypotheses Major Hypothesis I: Under- and overachieving students will differ significantly on a measure of academic self- concept. Major Hypothesis II: .The dimensions of a measure of academic self-concept can be determined.‘ Minor Hypothesis I: The items selected from the measure of academic self-concept will be different for males and females Minor Hypothesis II: The dimensions of academic self-concept will be different for males and females. Organization of the Study The over-all plan of this dissertation is as follows: In Chapter II a review of research literature related to current investi- gation is presented. A discussion of the general design of the study, together with consideration of instrumentation, sample selection, and the analytic procedure used in dimension analysis of a measure of academic self-concept is presented in Chapter III. The analysis of the data is reported in Chapter IV, while the summary, conclusions, and implications for further theory development and research appear in Chapter V. CHAPTER II REVIEW OF RELATED RESEARCH LITERATURE In recent years, the research in self-concept has proliferated. Silver reports that in the 1939 Psychological Abstracts, six references pertaining to the self, self-concept, or ego are found.1 In 1949 nine references are listed, and 1956, 54 articles are indexed. In the 1959 issues 122 references pertaining to this area of research are present. -. Theoretical Considerations of Self-Concept Most contemporary self theories derive directly or indirectly from William James discussion of what he called the "Empirical Me".2 James' "self" was considered to be the sum total of all that a man can call his: including his body, traits, and abilities; his material possessions; his family, friends and enemies; his vocation and avocation. The.term 221; as used in contemporary social science literature has two generally accepted meanings.3 In the first definition, self-as- an-object, reference is made to an individual's attitudes, feelings, per- ceptions, and evaluations of himself as he views himself. In this sense, self is what a person thinks of himself. In the second definition, 1A. W. Silver, The Self Concept: Its Relationship 22 Parental and Peer Acceptance, Doctoral Dissertation, Michigan State University, 1957 2William James, Principles 2£_Psychology, New York: Holt, 1890 Chapter 10 ’ ' 3Calvin S. Hall and G. Lindzey, Theories 2f Personality, New York: John Wiley, 1957, p. 468 10 11 self-as-a-process, self is regarded as a group of psychological pro- cesses which govern behavior and adjustment. The self is considered a functioning object which consists of an active group of processes such as thinking, remembering, and perceiving. ' Comprehensive reviews of the psychology of the self are presented 2 Symonds,3 and Serbia.4 It is apparent by Hamachek,1 Hall and Lindzey, from the literature that there is no agreement on the use of the term "self". However, a predominant trend is concerned with relating self- concept theory to perceptual theory. Reviews of this emerging area of speculation and investigation are presented by Gordon and Combs,5 and Wylie.6 The theory that the self-concept of an individual is deve10ped through interaction with significant others has been expressed previously. (See Chapter I). Two recent investigations give credence to the operationalism of such a theory. Chertok presents research findings 1Donald E. Hamachek, Relationships Between the Self-Images 2; Elementary School Children and Certain Measures 2f Growth, Doctoral Dissertation, University of Michigan, 1960 2Calvin S. Hall 6 G. Lindzey, Theories 2; Personality, New York: John Wiley, 1957, pp. 469-489 3P. M. Symonds, The Ego and The Self, New York: Appleton-Century- Crofts, 1951 4Theodore R. Sarbin, "Role Theory", in G. Lindzey (Ed.) Handbook 2; Social Psychology, Reading: Addison-Wesley, 1954, pp. 238-258 5Ira J. Gordon 8 A. W. Combs, "The Learner: Self and Perception", Review gf Educational Research, 1958, Vol. 28, pp. 433-444 6Ruth C. Wylie, The Self Concept, Lincoln: University of Nebraska Press, 1961 12 which suggest that the interactional or behavioral aspects of relation- ships are more important in the development and maintenance of self- conception than are the more purely affective factors of such relation- ships.1 In an important study by Mannheim, the following assumptions were made, 1) self- mage should be affected by the frame of reference which the individual adopts from his major identification group, 2) the frame of reference relevant to the self-image is the consensus of the reference group about the self-image, and this consensus is presumably perceived by the individual as his "looking-glass-self", and 3) it is through the “looking-glass-self" that the identification reference group is assumed to influence the self-image.2 Mannheim foundwthat the self- image tended to be similar to the "looking-glass-self" reflected from the reference group, regardless of whether the indiyidual_did or did not consider himself a member of the referent group”. Results from both of the above studies demonstrate that the symbolic interactionist theory of self-concept developed by Cooley and Mean can be empirically tested with an acceptable validity. Definition 2; Self-Concept As has been noted, self-concept may be conceptualized as referring to a global attitude toward the total self, or it may refer to certain 1Ely Chertok, The Social Process 2; Self-Conception, Doctoral Dissertation, University of Washington, 1955, Dissertation Abstracts 1955, Vol. 15, p. 2330 2Betty F. Mannheim, .A_n Investigation 3f the Interrelations -o_f_ Reference Groups, Membership Groups, and the Self Image: ‘A Test pf the CooleyeMead Theory 2; Egg Self, Doctoral Dissertation, 1957, Dissertation Abstracts, 1957, Vol. 17, pp. 1616-1617 l3 traits or areas of functioning. Some investigators assume that specific self-concept measures are interchangeable indices of global self-concept. This is nota valid assumption as the developing individual reacts to his environment not with his total personality, but as he sees the . specific situation, and relates them to similar previous experiences. The variable definitions which have been attached to self-concept, ranging from prescribed to broad meanings, has resulted in much confusion not only within research studies, but also across investigations. For the purposes of this review and the present investigation, self-concept will be defined as: The person as known to himself, particularly the stable, imporiant and typical aspects of himself as he perceives them. Academic self-concept, following logically then, will be defined as: The total aggrigate of those aspects of the way an individual views himself which might be related to academic achievement. These aspects are perceived as being reflections of expectations of his academic "signi- ficant other", particularly the teacher. Measurement and Self-Concept The problems of reliability and validity of self-concept measures, and the design difficulties of studies using self-concept measures have 2,3 been competently reviewed elsewhere. It was found that five general 1Arthur‘W.’Combs & D. Soper, "The Self, It's Derivative, Terms, and Research", Journal pf Individual ngchology, 1957, Vol. 13, pp. 134-135 2Ruth C. Wylie, The Self Concept, Lincoln: University of Nebraska Press, 1961, pp. 23-113 3Theodore R. Sarbin, "Role Theory", in G. Lindzey (Ed.) Handbook ‘gf Social Psycholggy, Reading: Addison-Wesley, 1954, pp. 244-245, 253-255 14 types of instruments have been used--l) rating scales, 2) questionnaires, 3) adjective check lists, 4) coded interviews, and 5) Q-sorts. In con- sideration of the five types of measures, wylie states: At present no general conclusion can be drawn con- cerning any one instrument or type of instrument, because the array of studies is too widely scattered across instruments. It is concluded that no investigator to date has satisfactorily conceptualized or coped with these diffi- cult measurement problems. Quite a few have indicated that they make no claims for ....... validity and are content to let the reader' beware , ....... 2 In the majority of studies no reliability estimates are given, and those that are presented are mostly of the split-half or interjudge variety, ....... 5 The present investigation is designed to overcome the pitfalls wylie highlights. Dimensions of Self-Concept Investigators of self-concept have primarily been concerned with global or gross measures. Few attempts have been made to deter- mine the basic dimensions of self-concept. Furthermore, when such dimensions have been determined, the researcher has not made further investigations of them. A summary of dimensions, as determined in previous self-concept invetigations, might provide clues of what results 1Wylie, The Self Concept, p. 37 2Wylie, The Self Concept, p. 39 3Wylie, The Self Concept, p. 39 15' might be obtained from a dimension analysis of the measure of academic self-concept developed in the present investigation. A summary of these dimensions is presented in Table 2.1. The investigations summarized are those of; Martire,1 Dawkins,2 Smith,3 and Osgood, Suci, and Tannenbaum.4 Table 2.1 Dimensions of Self-Concept Isolated in Previous Investigations Investigator ‘Martire11956) ,Qagkins(1957) Smith(l960) Osgood, et.al.(l957) Intelligence Intellectual Independence Evaluative Motivation Motivational Self-Esteem Potency Creativeness Emotional Anxiety-Tension Activity Initiative Social Estrangement ‘ Success - Physical Body Image All the dimensions, with the possible exception of those deter- mined by Osgood, et. al., are self-explanatory. The dimensions in this 1J. G. Martire, "Relationships Between the Self-Concept and Differ- ences in the Strength and Generality of Achievement Motivation", Jgpzpal .2: Personality, 1956, Vol. 24, pp. 364-375 2Peter B. H. Dawkins, The Construct Validity 22 g Self-Rating Scale, Doctoral Dissertation, University of Texas, 1957,_Dissertation Abstracts, 1957, Vol. 17, p. 2678 3Philip'A. Smith, "A Factor Analytic Study of the Self-Concept", Journal 2: Consulting Psychology, 1960, Vol. 24, p. 191 4Charles E. Osgood, C. J. Suci, & P. H. Tannenbaum, The MeaSure- ment 2; Meaning, Urbana: University of Illinois Press, 1957 l6 later study were derived by factor analysis and were viewed as descrip- tions of the meaningfulness of concepts, rather than strictly unique to‘ self-concept. It was found that these three factored dimensions accounted for 66% of the variance in a factor analysis of a bipolar adjective check-' list. An evaluation of Table 2.1 indicates that those aspects of an individuals self-concept which are prominent, are those characterized by concern with intelligence, motivation, body image, and emotions. The summarized dimensions are to be accepted with caution because each investigator used a different measure and analytic procedure. It seems obvious that if psychological interpretation is to be made of self- concept, and academic self-concept, the dimensions inherent in this structure must be determined. Self-Concept Related to Academic Achievement Evaluation of research relating self-concept to academic achieve- ment is difficult for three reasons; 1) the lack of consistency in type of self-concept measure used, 2) the lack of comparability of samples, and 3) the uniform lack of agreement in the method of determining rela- tionships. Taking these limitations into consideration an attempt will be made to integrate the most significant and relevant research to the present investigation. 7 'The three investigations most directly related to the problem 17 1 under consideration are those of Brookover, et. al., Davidson and 2 and Shaw, et. al.3 In the Brookover study evidence is presented Lang, for the functional operation of the "looking-glass-self" conceptualiza- tion of self-concept, as it relates to academic achievement. Two experi- mental rating scales were developed to meet Guttman's criterion of unidimensionality.4 These were labeled Self Concept of Ability and Importanfliof Grades. A summary of the research results currently completed on the first scale are presented in Table 2.2. The preliminary results in Table 2.2 demonstrate that high achievers had a significantly higher mean self-concept score than the underachievers, even though these two groups have comparable I.Q.'s. The same inter- pretation is true of the difference between the mean self-concept of ability scores of overachievers and low achievers, who also had compar- able I.Q.'s. Further results from the Brookover study are to be found in Table 2.3. These results are self-explanatory, with one exception. Changes in the correlations of self-concept and grade point average, with importance of grades, when the importer!2 of grades was partialed 1Wilbur B. Brookover, A. Velinsky, & S. Thomas, "Relationship of Self-Images to Achievement in Junior High School Students", Paper read at Annual Meeting of American Educational Research Association, Chicago, February, 1961 2Helen H. Davidson, & G. Lang, "Children's Perceptions of Their Teacher's Feelings Toward Them Related to Self-Perception, School Achieve- ment and Behavior", Journal 2; Experimental Education, 1960, Vol. 29, pp. 107-118. 3Merville C. Shaw, K. Edson, & H. M. Bell, "The Self-Concept of Bright Underachieving High School Student as Revealed by an Adjective Check List", Personnel and Guidance Journal, 1960, Vol. 39, pp. 193-196 , 4Louis Guttman, "The Basis for Scaleogram Analysis", In S. A. Stouffer et. al. Measurement and Prediction, Princeton: Princeton University Press, 1950, pp. 60-90 18 out, lends support to the conclusion that the self-concept of ability scale is measuring something other than concern about achievement. It Table 2.2 Summary of Mean "Self-Concept of Ability" Scores of Different Achievement Groups (Brookover et.a1.) Significagze _J[_ High Achievers N Underachievers Level Males 172 29.94 26 26.17 .001 High 1.9. Females 171 30.74 26 27.27 .001 Overachievers Low Achievers Males 38 27.67 160 24.58 .001 Low 1.9. Females 35 29.58 158 25.70 .001 High GPA Low GPA was also found that teachers were chosen as significant others by high and overachievers more often than they were chosen by low and under- achievers. A Chi square analysis indicated that this difference was significant at the .05 level. These results must be accepted with some degree of caution as no evidence for reliability or validity of the experimental scales has yet been reported. The criticism just made of the Brookover study is not applicable to the research reported by Davidson and Lang.1 The extensive pre-testing 1Helen H. Davidson & G. Lang, "Children's Perceptions of Their Teacher's Feelings Toward Them Related to Self-Perception, School Achievement and Behavior", Journal pf_§gperimental Education, 1960, Vol. 29, pp. 107-118 19 of their experimental instruments is to be commended. The results may be summarized as follows: 1) a significantly positive relationship between favorable perception of teachers’ feelings toward the student and academic achievement was found, 2) children who had a more favorable Table 2 . 3 Summary of Correlations Among "Self Concept of Ability" Scores; Grade Point Average; and "Importance of Grades" Scores for 'Males (N-513) and Females (n-537) (Brookover et. a1.) .m Males Females Self Concept and GPA .555 .548 Self Concept and Importance of Grades .456 .334 GPA and Importance of Grades .237 .006 Self Concept and GPA with Importance of Grades Partialed .517 .559 or a more adequate self-concept, i.e. those who achieved a higher self- perception score, also perceived their teachers' feelings toward them more favorably, and 3) favorable perception of teachers' feelings was associated with desirable classroom behavior. The results are inter- preted as indicating that when the teacher’s feelings of acceptance and approval are communicated to the child, these are perceived as positive appraisal by the child. The perceived positive appraisals stimulate the child to seek further teacher approval by achieving well and demon- strating acceptable classroom behavior. The interrelationships among self-perception, perception of teacher approval, and academic achieve- ment has been demonstrated. However, these findings should not be inter- 20 preted as indicating cause and effect relationships. Also, it is noted that the sample was limited to those students who were the better known readers, the poor readers being eliminated. Such a sampling procedure would result in higher homogeniety among subjects, and therefore more conservative estimates of possible differences. Although important relationships were demonstrated in this study, the lack of meaningful statements about the psychological dimensions of self-perceptions is evident. An investigation by Shaw, Edson and Bell attempted to overcome the short-coming described above.1 Using the Sarbin 200 word adjective check-list, significant differences were found in the self-concepts of achievers and underachievers, for each sex, with intelligence controlled. Thirteen adjectives were significant discriminators at the .05 level for [males, e.g. stable, reliable, intelligence, and mischievous. Seventeen adjectives were significant discriminators at the .05 level for females, e.g. ambitious, pleasure-seeking, moody, and easy-going. The authors drew the following Conclusions from their findings; 1) male underachievers seem to have more negative feelings about them- selves than do male achievers, and 2) female underachievers tend to be ambivalent in their feelings toward'themselves. No evidence has been presented to demonstrate whether differences in the self-concepts of students manifesting achievement extremes, is a cause of or a result of their scholastic experiences. 1Merville c. Shaw, K. Edson, a. n. M. Bell, "The Self-Concept of Bright Underachieving High School Students as Revealed by an Adjective Check List", Personnel and Guidance Journal, 1960, Vol. 39, pp. 193-196 21 The_three studies just reviewed gain sigdficance in light of the research findings of several other investigations. A study by Stevens determined that the saliency of self-concept is its most important 1 In studying a group of successful and unsuccessful characteristic . college students Stevens found that significant differences existed between these groups in terms of the salience, or distinctiveness, of the characteristics that they attach to their own personalities. Self- characteristics that involve energy, productivity and efficiency were more highly valued by the academically successful student than the unsuccessful student. It was also found that self—insight and self- acceptance was greater in the academically successful student. This was especially true when such students evaluated their own intellectual abilities. It was suggested that the poorer self-evaluation of the unsuccessful students, which.was reflected in a distorted self-picture of their intellectual ability, had been due to defense mechanisms used to compensate for academic failure. The influence of defensiveness is further clarified by Roth.2 In studying college students enrolled in reading improvement courses, Roth found significant differences between the self-perceptions of three groups; 1) drop-outs, 2) improvers, and 3) non-improvers. The differences took the form of general defen- 1Peter H. Stevens, An Investigation_ of the Relationship Between Certain Aspects of Self-Concept and Students Academic Achievement, Doctoral Dissertation, New York University, 1956 2R. M, Roth, "Role of Self-Concept in Achievement", Journal 2: Experimental Education, 1959, Vol. 27, pp. 265-281 22 siveness, distortion of perceived self and self as seen in relation to Iauthority, and self as a reader. Drop outs were found to be most defensive in respect to learning, the improvers least defensive. The investigators did not control for continuity of testing at the time the self-concept measures were obtained.. This fact might significantly have affected the results. An pertinent study reported by Passow and Goldberg focalizes the 1 Before considering the results significance of self-acceptance scores. of the study two serious methological shortcomings should be noted. The problems are 1) a compounded sampling procedure by selecting criterion groups from different grades within the high school with each representing a different achievement classification, and 2) a lack of control for sex. It was found that the total self-acceptance scores of underachievers were somewhat lower than those of high achievers. The most consistently differentiating characteristics were those of high achievers, of an intellectual or task-oriented nature, i.e. underachievers felt they were less capable of successfully performing in these types of activities. No significant differences were found on personalésocial traits or on special talents and aptitudes. Although the underachievers viewed themselves as less adequate in intellective and task related behavior, their expressed "wishes" in these areas were not different from the high achievers. Thus, the psychological distance which the underachiever 1A. Harry Passow &‘M. L. Goldberg, 225 Talented Youth Project: ‘5 Progress Report 1961, New York: Horace; Mann-Lincoln Institute of School Experimentation, Teachers College, Columbia University, (Mimeographed) 23 must travel to attain his desired status is far greater than the gap confronting the high achiever. The underachiever is faced with a psychological task which he may perceive as being beyond his efforts to accomplish and consequently either lowers his aspirations to the point where he is satisfied with "just getting by" or searches for some fantasy device which will enable him to achieve his goal. He may also completely divorce himself psychologically from the academic aSpect of school, and obtain his self-satisfactions in-non-academic areas. An investigation which summarized the types of relationships that have been discussed in this section, is that of Reeder.1 Using pupil self-ratings corroborated by teacher and peer ratings of middle grade children, and a comparison of two groups matched on intelligence test scores, but having extremes of self-concept, Reeder found that children with low self-concept scores had; 1) school achievement lower than expected based on a given level of "potential", and 2) frequently were classified as having classroom behavior problems. An interesting trend emerges from the preceeding review of literature relating self-concept to achievement, namely sex differences. Sex Differences in_Se1f—Concept Aside from the sex differences already noted in the studies by V 1Thelma A. Reader, A Study 3E Some Relationships Between Level of Self-Concept, Academic Achievement, and Classroom Adjustment, Doctoral [Dissertation, North Texas State College: 1955, Dissertation Abstracts, 1955, Vol. 15, p. 2472 ' 24 Brookover et.al.1 and Shaw et.al.,2 several other investigators have demonstrated that significant differences exist in analyses of the responses of males and females to measures of self-concept. Sarbin and Rosenberg,3 using a modified Gough Adjective Check List, found in studying a normal college p0pulation, that men exceeded women in checking such adjectives as resourceful, mature, logical, adventurous, realistic, deliberate and efficient. Women exceed men in checking feminine, emotional, affectionate, pleasant and temperamental. However, the groups were not specified as being equal in size and the greater response total of the men was not controlled in the statistical analysis. Perkins also found sex differences in self- ideal congurence in favor of females in a study with elementary school children.4 Davidson and Lang, again in studying elementary school children, found that sex differences existed in their subjects' "looking-glass" self-concept.5 The females perceived their teacher's v 1Wilbur B. Brookover, A. Velinsky & S. Thomas, "Relationship of Self-Images to Achievement in Junior High School", Paper read at Annual Meeting of American Educational Research Association, Chicago, February,l96l 2Merville C. Shaw, K. Edson & H. M. Bell, "The Self-Concept of Bnght Underachieving High School Students", Personnel and Guidance Journal 1960, Vol. 39, pp. 193-196 3T. R. Sarbin & B. G. Rosenberg, "Contributions to Role-Taking Theory: IV. A Method for Obtaining a Qualitative Estimate of Self“, Journal of Social Psychology, 1955, Vol. 42, pp. 71-81 4H. V. Perkins, "Factors Influencing Change in Children's Self- Concepts", Child Development, 1958, Vol. 29, pp. 221-230 5Helen H. Davidson & G. Lang, "Children's Perceptions of Their Teachers' Feelings Toward Them Related to Self-Perception School Achievement and Behavior", Journal of Experimental Education, 1960 Vol. 29, pp. 107-118 '_'—"" 25 feelings toward them as being more favorable than did the males (Meanf = 2.60, Meanm ‘-’ 2.52, 1:32.41, p(.02). In two studies by 1’2 using college subjects, sex differences were found in self Rosen, appraisal. The weight of evidence in the literature supports a hypothe- sized difference in the self-concepts of males and females. This is particularly true if the self-concept measures are further related to academic achievement. In this situation, however, the differences might have been due to initial sex differences in achievement. Self-Concept Related to Motivation and Level of ASpiration Behavior Academic performance may be viewed as a function of ability and 'motivation. Evaluations of research studies which have attempted to relate self-concept and academic performance have been presented earlier in this review. Evidence has been presented elsewhere that motivation and academic performance are related.3 It would seem logical, then, to posit a relationship between self-concept and motivation. Only in one published study was an attempted to test such a relationship. This might be explained in part by the lack of reliable and valid measures of motivation. The one study which was uncovered, however, has particu— lar relevance for the purpose of this review. Martire obtained thematic lE. Rosen, ”Self-Appraisal and Perceived Desirability of MMPI Traits", Journal of Counseling Psychology, 1956, Vol. 3, pp. 44-51 2E. Rosen, "Self-Appraisal, Personal Desirability and Perceived Social Desirability of Personality Traits", Journal of Abnormal and m“ Social Psychology, 1956, Vol. 52, pp. 151-158 3Marion D. Thorpe, The Factgred Dimensions of_ an Objectixg_ Inventory 2f Academic Motivation Based 22_Eleventh Grade Male Over- and Unaer Achievers, Doctoral Dissertation, Michigan State University, 1961 26 apperception measures of achievement motivation (n-Achievement) from 53 male college students under both "neutral" and "achievement- motivating" instructions.1 Self-Ideal and Self measures were obtained by having students rank 26 traits according to their importance in' general, and again in terms of how characteristic of them each trait was. Wishful and realistic levels of aspiration were obtained for a specific pencil and,paper task. In general it was found that individuals with high motivation scores under both instructive conditions had a significantly greater discrepancy between their self-ideal and self ratings on five achievement-rated traits. It was also found that self- concept and level of aspiration measures were related to achievement, motivation, but not to each other. A logical interpretation of this finding would lead to the conclusion that measures of these two vari- ables are each concerned with some distinct and independent aspect of an individuals "achievement syndrome". A related study is that'bf Pyne.2 In this study, conducted at the University of Minnesota, Pyne demonstrated that successful and unsuccessful competitors could be distinguished on the basis of their self-concept, level of sepiration, and motivation as measured in terms of observable behavior. Several investigators have limited their research to studying only relationships between self-concept and level of aspiration. Level V v 1John G. Martire, "Relationships Between the Self-Concept and Differences in the Strength and Generality of Achievement Motivation", Journal g£_Personality, 1956, Vol. 24, pp. 364-375 2Francis P. Pyne, The Relationship'of Measures of Self-Concept, Motivation and Ability tb'Success in Compefitive Athlefics, Doctoral Dissertation, University—of Minnesofa,’l956, Dissertation Abstracts, 1957, Vol. 17, p. 559 H 27 of aspiration was defined as‘a standard by which a person judges his own performance as a success or a failure, or as being up to what he expects of himself.1 Gilinsky has demonstrated a positive relationship between self- concept and level of aspiration (r = +.67).2 No information, however, was available to indicate what cognitive and/or emotional variables were involved in producing the relationship. A similar finding is reported by Sears.3 She found an association between size of positive discrepancy score in an experimen- tal level of aspiration task and the size of stated (self-self ideal) discrepancies on academic tasks. The "reality dimension" was held constant across subjects, so the‘discrepancy scores were a function of individuals differences in stated level of aspiration. It could not be determined from the data whether the high (self-self ideal) discrepancies were due to statements of unusually high ideals, low realistic self, or both.. In this study, level of aspiration was operationally defined as "goal setting behavior". The same definition was used in a study by Steiner in studying college students by using a measure of various self- perception ratings.4 Steiner concluded that persons with uncertain self- v ~r 1Horace B. English & A. C. English, A Comprehensiye Dictionary of Psychological and Psychoanalytical Terms, New York: Longmans, 1958 2A. S. Gilinsky, "Relative Self-Estimate and the Level of Aspira- tion", Journal of Experimental Psychology, 1949, Vol. 39, pp. 256-259 3Pauline Sears, “Level of ASpiration in Relation to Some Variables of Personality", Journal of Social Psychology, 1941, Vol. 14, pp. 311-336 4I. D. Steiner, "Self Perceptions and Goal Setting Behavior", Journal of Personality, 1957, Vol. 25, pp. 344-355 28 perceptions tend to set goals which are high relative to their past performance. They expect their performance scores to vary considerably over time, and are more likely, than a "certain" group, to over-estimate their future performance. They are less certain than others that their goals are realistic. Persons pessimistic in self-appraisal stated that they felt because they were pessimistic,this in some way affected their performance. This later group was also intrepunitive in their explanations of their performance. Steiner often used terms which held unique meaning for him. This fact in combination with the elaborate measurement proce- dure, which was susceptible to faking and malingering on the part of subjects, casts some serious doubt on the research results. DeSpite the indicated short-comings in the studies reviewed in this section, some support for posited relationships between self-concept and level of aspiration is present. Conclusions and Interpretation Although few definitive’studies have been made in the area of self-concept, there are sufficient positive findings to support further use of self-concept theory as a research base. The paucity of research is even more evident when searching for investigations which have attempted to relate aspects of self-concept to an achievement setting. Self-concept does appear to be related to academic .achievement, motiva- tion, and level of aspiration. To date, indications are that under- and overachieving students differ significantly on these last three variables. Therefore, it is logical to assume that statistically defined discrepant achievers would differ in self-concept. This would be 29 particularly true if the self-concept is academically referent. There is also research evidence to justify an analysis of academic self-concept separately for males and females. A need exists to identify the dimensions of self-concept which are related to academic achievement. Summary In the review of related research presented in this chapter, a discussion of various theoretical orientations of the psychology of the self was made. Following from this discussion, definitions of self- concept and academic self-concept were presented. The types of and pro- blems in, the measurement of self-concept were next discussed, followed by a review of studies in which the primary concern was to discover the underlying structures or dimensions of self-concept. These dimensions were summarized as follows; intelligence, emotions, body image, and motivation. Evidence was offered in support of hypothesized relationships between self-concept, and academic achievement, motivation, and level of aspiration behavior. Sex differences in self-concept were noted. For example, male underachievers tend to have a negative and pessimistic self-concept, whereas female underachievers are ambiva- lent about their feelings toward themselves. As a total group, females had more positive self-concepts than did males. Finally, a discussion of the research impalications of the total review was presented. CHAPTER III DESIGN AND METHODOLOGY Considerations of the general design and methodology of this study are made under five general headings; l) instrumentation, 2) sample sel- ection, 3) item analysis procedures, 4) selection of apprOpriate multi- variate model, and 5) multiple scalogram analysis procedures. Instrumentation An instrument was designed and constructed which required the stu- dent to respond to a series of concepts. These concepts were-assumed ta“ be related to both the student's self-concept and academic achievement. The student was to respond to these concepts as he thought his teachers would, when viewing hig as a student. 2 Items for the censtruction of an instrument to measure academic self- concept were deve10ped from information drawn from three primary sources; . 1) the academic self-concept theory developed in Chapter I, 2) the re- view of related self-concept literature found in Chapter II, and 3) sum- mary personality, motivational, intellectual, and emotional characteris- tics, which had been used by the Farquhar team for over-all item develop- ment in the Motivational Project. (See Appendix A) Items were deve10ped in the form of one, two, or three word concepts and phrases. Some of the concepts were develOped as "slang" expressions frequently used by adolescents. The resulting instrument consisted of 119 concepts. The student was asked to rate each of these concepts on a four point scale as he thought his teachers would. The rating scale used was as follows; 30 31 l=never, 2=sometimes, 3=usually, and 4=always. This format was used to bring into play the "looking-glass-self" theory discussed in Chapter I. ReSponses to the instrument, labled the Word Rating List (hereafter referred to as WRLl), were obtained on IBM answer sheets to facilitate data analysis. The WRL was administered to eleventh grade under- and overachieving students of each sex, hypothesizing that these in- dividuals would differ significantly in their academic self-concepts. The procedure for the selection of statistically defined under- and over- achieving students is outlined in the following section. Sample Selection Validation and cross-validation of the Word Rating List was accom- plished by contrasting under- and overachieving students who were iden- tified in the following manner: 1) A survey of high schools in the 100 largest pOpulated cities was made to determine the nature of their testing programs. 2) Nine high schools in eight Michigan cities having 9th grade Differential Aptitude Test scores available on their current 10th graders, were contacted and asked to co-operate in the study. 3) A second aptitude measure was obtained so that reliable esti- mates of academic aptitude could be made. California Tests of Mental Maturity were administered while the students were in the 10th grade. Administration was necessary in all but one of the schools. l A c0py of the Word Rating List is to be found in Appendix B. 4) 5) 6) 32 Grade point averages (GPA) were calculated using grades in 9th and 10th grade subjects. Only academic subjects were included, i.e. those requiring homework. Activity courses were eliminated from the calculations. The DAT-Verbal Reasoning and CTMM-Language sub-test scores were used in obtaining a stable estimate of academic aptitude after empirically examining possible DAT and CTMM sub-score combin- ations. Regression lines were calculated for each school and sex assum- ing a correlation of +1.00 between DAT-VR and CTMM-L. Separate equations were calculated because a pilot study indicated that one was not applicable across schools. Only those individuals who fell within one standard of error of estimate above and below the regression line were included in the study. Because it was important that the criterion groups be classified with little chance of making a Type II error (accepting when should have re- jected), it was decided to run the risk of a Type I error (re- ject when should have accepted)even if sample were lost in the process. (see Figure 3.1) (High) 5 ’x D / / / / K CTMM-L ’ s D (LOW) x / _ DAT-VR X=individuals selected for the study Figure 3.1Methodological Selection of Individuals with Stable Measured Aptitude 33 7) Regression equations predicting GPA from the DAT-VR scores were calculated for each sex in each of the participating schools. The DAT-VR was used as the dependent variable be- cause it was found to correlate consistently higher with GPA than the CTMM-L scores. (e.g. in one pilot study the cor- relation of VR with GPA was found to be +.65, and CTMM-L with GPA to be +.50) Underachievers were defined as those indivi- duals whose GPA fell at least one standard error of estimate 2212! the regression line prediction of achievement. Similarly overachievers were designated as falling one standard error of estimate above the regression line. (see Figure 3.2) I ’./’ X X GPA ‘ ’,.- x X ’,/ I"’ N E a (Low) (High) l-Overachievers ' flrUnderachievers Figure 3.2Method of Selecting Under- and Overachievers By using the above indicated method, under- and overachieving students were selected from the full range of academic ability. Approx- imately 12% of the sample was classified in one of the extreme groups. This procedure resulted in obtaining 167 male overachievers, 131 male underachievers, 191 female overachievers, and 170 female under- achievers, on whom.WRL results were available. Each of these four groups was divided in half, the first becoming the validation sample 34 and the second, the cross-validation sample. A summary of the re- sulting sample sizes is presented in Table 3.1. Table_3.l Summary of Sample Sizes Used in Validation and Cross Validation of the Word Rating List* Overachievers Underachivers Validation , 87 62 MALES' _Cross-Validation 80 ' 69 Total 167 131 Validation 95 84 FEMALES Cross-Validation 96 86 Total 191 170 *Drawn from anfioriginal sample of 4200 eleventh grade students Item Analysis Procedures The four point response continum of the URL was dichotomized to facilitate the tabulations and calculations necessary for item analysis. Responses of "1" and "2" became "0", and responses of "3" and "4" became "1". The "0" scoring direction indicating that the student felt his teachers would n2; use a particular concept in describing him, and the "1" response indicating that he felt that his teachers would use a particular concept in describing him. Response frequencies for every item were obtained. These frequencies were then entered into a 2X2 contingency table (under- and overachievement by "0" and "1"), and the 35 Chi square values determined.1 This analysis took the form of finding which items significantly discriminated between under- and overachievers. The level of significance was set at .10 for the validation and cross- validation of the URL. The stringent significance level was used in order to minimize rejecting the null hypothesis when it should have been accepted (Type I error). Selection of Appropriate Multivariate Model Several multivariate analytic procedures are available for the determination of underlying structure for a given set of data. A recent study by Lingoes presents empirical results to aid in the selec- tion of an appropriate procedure.2 In this investigation a hypothetical binary response matrix (dichotomous responses) of Congressional voting behavior was analyzed using 1) multiple scalogram analysis,3 2) Guttman's scalogram analysis, as discussed by Stouffer,4 3) multiple factor analysis 1This analysis was accomplished on a high speed electronic computer (MISTIC) at Michigan State University, by having the observed frequencies for Chi square analysis punched on computer tape and analyzed with the K6M program. 2James C. Lingoes, ' lti 1e Scalogram Analysis: _A Generalization 9f.gg;;m§g;§ §ca1glénalysis", Doctoral Dissertation, Michigan State University, 1960 3James C. Lingoes, "Multiple Scalogram Analysis ----- " 4S. A. Stouffer (Ed.),Measurement and Prediction,Princeton: . Princeton University Press, 1950 36 (principle axes extraction1 and quartimax rotationz), and 4) Loevinger's method of homogeneous tests. The purpose of this study was to discover the underlying dimensions of the response matrix, such that the whole matrix could be perfectly reconstructed on the basis of ranked scores. . The analysis results were validated and cross-validated for each of the analytic procedures. Multiple scalogram analysis was found to be the only technique among the four which could directly, simply, objectively, uniquely, and parsimoniously accomplish the aim as proposed. Loevinger's technique was found to be second in meeting the criteria. On the basis of this research it was decided to use multiple scalogram analysis pro- cedures to determine the underlying dimensions of the response matrix obtained from eleventh grade students on a measure of academic self-con- cept. A description of multiple scalogram analysis is presented in the following section. Multiple Scalogram.Ana1ysis Procedures A generalization of Guttman's scaling method has been presented for dichotomous data by Lingoes.3 This method, multiple scalogram analy- sis (MSA), is designed'to yield a number of unidimensional scales in a single analysis. 1J. 0. Neuhaus and C. F. Wrigley, "The Quartimax Method: An Approach to Orthogonal Simple Structure", British Journal 2; Statistical Psychology, 1954, Vol. 7, pp 81-91. 2Jane Loevinger, "The Technique of Homogeneous Tests Compared With Some Aspects of "Scale Analysis" and Factor Analysis", Psychological Bulletin, 1948, Vol. 45, pp. 507-529 3James C. Lingoes, ulti 1e Scalogram Analysis: A Generalization Qfi,Qu§tman’s Scale Analysis , Doctoral Dissertation, Michigan State University, 1960. 37 Items are said to be perfectly scaleable or to constitute a uni- dimensional scale if and only if the same response pattern is given by any to 3'3 having the same score. Stated in another way: a unidimen- sional scale exists if one is able to reproduce each S's responses pat- tern on the basis of each S's score and a knowledge of the item .ordering. In a broad sense, the reproducibilityof a scale is its reliability. Furthermore, for any S having a higher score than another S, it is true that the higher ranking S ranks as high or higher than the lower ranking S on every item in the set, i.e., passes or endorses these items. Multiple scalogram analysis is an objective and empirical technique for partitioning a dichotomous response matrix into a number of sub- matrices, such that each sub-matrix tends to be maximally homogenous. The method is founded on the premise that there must exist some formal criteria for item selection as an alternative to the subjectivism im- plicit in Guttman's concept of a "universe of content or attributes". Advantages 9i _M_§_Ag_ The chief advantages of MSA are as follows: 1) It allows for a wider latitude in complexity of the items included for analysis, e.g. personality items as broadly defined rather than some specific personality variable. In effect, the data are permitted to "speak for themselves", without imposing restrictions in advance of exploring the underlying structure of a given set of data. 2) It is a unique solution for the data. 3) 4) 5) 6) 38 It is economic in terms of time and effort. If the assumptions of.a scale model underlie the data, the results of MSA will clearly indicate the existence of a scale or scales. This is true only if such a scale exists, as opposed to factor analysis which yields factors even if only chance relationships are present. A clear picture not only of what scale or scales exist, but where each individual fits on the scale is presented. MSA is able to handle a number of items and subjects limited only by the capacity of the computer being used. One limitation of the MBA program at present is that it will handle only dichotomous data. In relation to the present data this is not considered a limitation because of its dichotomous nature. Criteria for Homogenity 2f Sub-Matrices Homogenity of sub-matrices (sub-scales or dimensions) resulting from MSA is achieved by three formal criteria: 1) 2) Each set or group of responses to each item is maximally related to items or sets which are adjacent to it by the use of a set-measure rather than conventional measures of correlation. Each item is allowed to contribute only its proportional share of error to the sub-matrix of items and people, where error is defined in terms of deviations from perfect reproduction of the sub-matrix. . All II | Ill 0 39 3) E.ch person is similarly permitted to contribute only a certain percentage of error to the sub-matrix. This is accomplished by eliminating S's from the response matrix ghen their response patterns fall below a given critical value,as determined by an iterative procedure. Criteria 2 and 3 above have implications for the determination of reliability of the resulting dimensions. This fact will be discussed in another section of this chapter. . Assggtions g; _M_S_A_'l_go_cl_e_l_ The major assumptions underlying the multiple scalogram model are as follows: 1) The items being analyzed either are or can be made to conform to the characteristics of cumulative items. 2) Item direction for a particular sub-matrix has not been determined on an 3 mpasis. 3) The items being analyzed are simple, equipotent, and additive. Furthermore they are invariant in validity and meaning for a given sample of subjects and items: 4) Prediction, understanding, and psychological meaningfulness of resulting sub-matrices will result to the extent to which constancy does 325 exist across items and/or people in terms of observed relationships. 5) The error for subjects is reliable within the same domain of subject matter or content. 6) The response matrix to be analyzed is based on a dichotomous response measurement. 40 The data of this study meet the above assumptions for the follow- ing reasons: 1) The rating scale used in the Word Rating List can be justified as meeting the requirements of an interval scale which allows for additivity and cumulation. 2) The dimensions of the response matrix have not been deter- mined on anlg priori basis. 3) The response continum has been dichotomized. 4) The assumption of equipotency and invariance of validity is met due to the fact that all items to be analyzed must have reached or exceeded the .10 level of significance for discrimination. 5) All item-word-concepts are assumed to be measures of the same broad content area of academic self-concept. 6) Individuals within a particular achievement classification (under- or overachievement) are assumed to be homogenious and stable._ Some indication of the stability of at least one of the achievement classifications over time, namely underachievement, is presented by Shaw and McCuen.1 They found that that their underachievement classifications of both males and females had stabilized by the tenth grade, after following these individuals from the first grade. 1Merville C. Shaw and J. T. McCuen, "The Onset of Academic Under- achievement in Bright Children", Journal 2f Educational Psychology, 1960, Vol. 51, pp. 103-108. 41 Reproducibility (Reliability) 2; Sub-Matrices Reproducibilities of the resulting dimensions from multiple scalogram analysis correspond to the value:l§_e l- _j£(error§) , mn where m- the number of items in the sub-matrix, n= the number of sub— jects, and errors are counted as deviations from the predicted response pattern.1 Before the data is fed into the computer, a parameter is read in which indicates a predetermined error level for allowable percentage of variation of items by subjects, i.e. deviation from perfect scalability. In the world of reality, perfect scales are almost never found, but reproducibilities around .90 are considered as being close approximations. If the 10% criterion for scalibility is used, the resulting reproducibili- ties will range from .90 to .70,if the 201 criterion is used, the range will be from .80 to .60.2 Procedural m i}; M_S_A A comprehensive discussion of the actual procedural steps in multiple scalogram analysis is presented by Lingoes.3 Several of these steps should be discussed as they are unique to the data of the present study. fiv— v—v—vr 1W. H. Goodenough, "A Technique for Scale Analysis", Educational and Psychological Mbasurement, 1944, Vol. 4, pp. 179-190. 2Donald‘M. Wilkins, Statistician in the Computer Laboratory, Michigan State University, personal communication 3James-C. Lingoes, Multiple Scalogram.Ana1ysis:‘A_Generalization ,2; Guttman's Scale Analysis, Doctoral Dissertation, Michigan State University, 1960, pp. 13-18. \ ' 42 1) Multiple scalogram analysis was carried out only on those items which had met the .10 significance level criterion set for cross-validation. 2) Following from the discussion of sex differences in self- concept as presented in Chapter II, separate analyses were completed for males and females. 3) In order to maximize the dimensions to be obtained from.MSA, under- and overachievers were combined within each sex. Due to computer limitations, 49 females had to be randomly elimi- nated,with the resulting sample sizes for MSA as follows, Males 8 298 and Females - 312. 4) For each of the cross-validated items, a dichotomous response matrix (using "0’s" and "l's"), was built for males (combined under- and overachievers) and females (under- and over-achievers). 5) These response matrices for each item by sex were punched into item cards and fed into MISTIC, the high speed electronic computer at Michigan State University, and the multiple scalogram analysis completed.1 6) Multiple scalogram analysis will be completed using the 101‘ criterion for scalibility. If all dimensions do not meet this criterion the error level will be moved to 20%. This latter step will still allow for dimension reliability of- .80. If the resulting dimensions from the second analysis do not meet the criterion, analysis will be terminated, it being assumed that reliable dimensions are not present. lursnc Library Routine for MSA is x914 43 Summary A discussion of the development of an experimental instrument to measure academic self-concept has been presented. This instrument was labeled the-Word Rating List. In order to validate and cross-validate the instrument, criterion groups of under-and overachieving eleventh grade students were selected for each sex. It was assumed that such discrepent achievers would differ significantly in their academic self- concepts. Chi square analysis procedures were presented. To determine' the psychological meaningfulness of an inventory of academic self-con- cept, it was felt that a multivariate analysis was necessary; the specific purpose being to determine the underlying dimensions of such an inventory. A review of the research literature indicated that multi- ple scalogram analysis would best meet this purpose. The nature of, assumptions necessary for, and procedural steps involved in this analytic method were discussed. CHAPTER IV ANALYSIS AND INTERPRETATION OF THE DATA In this chapter the data pertaining to each of the major and minor hypotheses described in Chapter I will be presented. Following each data presentation will be an interpretation of the analysis as it relates to the hypothesis. Item Analysis Results Major Hypothesis I concerned with the question of whether or not significant differences exist between under- and overachieving ‘ students on a measure of academic self-concept. The tests of signifi- cant differences on such a measure, were determined by Chi square analyses. A summary of the item analyses for the Word Rating List, is presented in Table 4.1. TABLE 4.1 Summary of Item Analyses of the Word Rating List* A J h Chi Square Significance Level .10 .05 .01 Total Males ll 19 37 67 Validation Females 10 27 50 87 Males 4 8 36 48 Cross-Validation Females 7 7 34 48 :Originally 119 items. Inspection of Table 4.1 indicates that 402 of the original 119 items 44 45 held up after cross-validation, for each sex. It is also noted that 751 of the male items and 711 of the female items were significant discriminators at the .01 level. It must be remembered that the items were required to meet the .10 discrimination criterion in both valida- tion and cross-validation to be acceptable for the final scale. All of the above strongly support the hypothesis of significant differences between under- and overachieving students on a measure of academic self- concept. ‘Minor Hypothesis I was concerned with sex differences in academic self-concept. It was found that of the 48 discriminating items, 35 or 73% were in common to both sexes.1 It was also found that all of the items in common to both sexes were significant discriminators in the same scoring direction. All of the findings thus far presented, give evidence for the rejection of the hypothesis of sex differences in academic self-concept. More data concerning this hypothesis will be available when the results of the dimension analyses are presented. Such results will indicate to what extent the 13 unique items influence the resulting dimensions. A summary of the reliability estimates for the total scale scores, for the cross-validated items, are presented in Table 4.2. The 2 estimates, obtained by Hoyt’s method, are for underachievers, over- achievers, a random sample of the general population, and a sample of A 1The discriminating items for males and females are indicated on the copy of the Werd Rating List found in Appendix B. 2Cyril J. Hoyt, “Test Reliability Estimated by Analysis of Variance", Psychometrika, 1941, Vol. 6, pp. 153-160. 46 combined under- and overachievers. An evaluation of Table 4.2 indicates TABLE 4.2 Summary of Boyt's Reliability Estimates on The Word Rating List for Different Samples by Sex Reliability _jL_ Estimate MALES Random Sample 66 .92 Overachievers 50 .92 Underachievers , 50 .91 Combined Under- and Overachievers 100 .93 FEMALES Random.8ample ' 66 .90 Overachievers < 50 .90 Underachievers 50 .88 Combined Under- and Overachievers 100 .93 1 that all reliability estimates are of an acceptable magnitude. The reliability estimates obtained on the samples of combined under- and overachievers are of crucial importance, as these groups constituted the samples used in the multiple scalogram analyses. The reliabilities of the total URL scores for these samples are also of an acceptable magnitude. Multiple Scalogram.Ana1yses The results of the multiple scalogram analyses will be considered in two sections, the first for males, the second for females. The primary purpose of this study was not to develop a measure of academic self-concept which was scalable, but to determine the underlying structure, or dimensions of academic self-concept. With a view toward 47 expansion and/or modification of the academic self-concept theory pre- sented in Chapter I. Primary concern, then, will not be with meeting 1 and the "scale-type“ criteria which have been suggested by Guttman, Edwards.2 To increase the interpretablity of the analyses, the following points should be considered: 1) The rep¥educibility of a dimension is a measure of the degree of accuracy with which the item responses can be duplicated from knowledge of the total scores alone. The coefficient of reproducibility is calculated by the method described by Goodenough.3 2) The minimal marginal reproducibility of a dimension (an average of the modal response preportions), when compared with the reproducibility indicates the improvement in predictions of response patterns from the knowledge of total scores. 3) Sometimes the modal response proportion for "q" is greater than "p", i.e. a greater percentage of people responded in the direc- tion opposite than that which was indicated. When this occurs, the particular item is "reflected" and is interpreted in the direction opposite from the scored direction. 4) Multiple scalogram analysis yields information not only about the grouping of items, but also about the ranking or placement of individuals. Therefore, the individuals who primarily determine the 1Louis Guttman, "The Basis for Scalogram Analysis", In S. A. Stouffer, et. a1, MEasurement and Prediction, Princeton: Princeton University Press, 1950, p. 60-90. 2Allen L. Edwards, Techniques of Attitude Scale Construction, New York: Appleton-Century-Crofts, 1957, pp. 172-200. 3W. E. Goodenough, "A Technique for Scale Analysis", Educational and Psychological Measurement, 1944, Vol. 4, pp. 179-190. 48 nature of a particular dimension may be characterized by the item con- tent of the dimension they have determined. Such a fact aids in the interpretation of dimensions because it allows for characterization of individuals and thereby their academic self-concepts. Results 9_f._ Multiple ScalogLram Analysis - 99.1.92 The first analysis, using the multiple scalogram program with an error level of .90 (allowing for reproducibilities between .90 and .70), for combined male under- and overachievers did not yield sufficient dimensions for psychological interpretation. The error level was therefore moved to .20 (allowing for reproducibilities between .80 and .70) and a second analysis completed. This analysis yielded four dimensions which accounted for 46 of the 48 cross-validated male items. Only items eight (confident) and 89 (contented) did not scale on any of the dimensions. The item content of the four dimensions is presented in Tables 4.3, 4.4, 4.5, and 4.6. Following each of these tables is a brief description of the dimension. A more comprehensive interpretation of the dimensions will be presented in Chapter V. A content evaluation of Dimension I as presented in Table 4.3 indicates that this is a general or global dimension of academic self- concept. Dimension I contains 54% of the scaled items for males. Dimension I also characterizes the individual who seeks academic achievement by using the usual academically sanctioned means, e.g. by being interested, ambitious, careful, orderly and intelligent in his classroom behavior. This is the type of individual who wishes to learn what is presented in class because he knows he must do so in order to obtain an acceptable grade. He is the type of individual who 49 Va, TABLE 4.3 Item Content of Dimension I (Males) Reproducibility - .786 ‘Minimimal Marginal Reproducibility = .715 F. r t J Scoring . Modal Response Item_§2. Direction* . Item Content** Proportion*** 4 - Inefficient (efficient) .869 19 - Uninterested (interested) .859 45 - Reckless (careful) .852 54 - Stubborn (flexible) .829 53 - Lazy (ambitious) .839 39 - Rebellious (conforming) .826 106 - Impatient (patient) .819 50 - A Person who Postpones (A Person .789 Who Does Things Immediately) 40 - Nervous (calm) .842 113 - Passive (active) .829 102 - Inconsistent (consistent) .819 103 + Teachable .732 118 + Reliable .748 48 + Dependable .752 29 + Responsible .705 17 + Orderly .641 16 + Thorough .550 83 + A Thinker .581 85 + Ambitious .621 97 + An Achiever .560 77 + Productive .554 76 + Above Average .507 12 + Smart .507 13 + Successful .587 34 + Intelligent .651 *Scoring is in overachievement direction **For items which yielded a negatively scored direction, the positive meaning of the item content is presented in parentheses to aid in comprehension. ***Either "p" or "q" which ever is larger would fit the teacher’s sterotype of the "good" student. His under- standing of and interest in the subject matter, however, may tend to be superficial. Two other trends are apparent from an evaluation of this dimension 50 ‘ First,the modal response proportions for the highest ranking eleven ’ items are for items which are scored in the negative direction. A ten- tative interpretation might be that the male, particularly the male overachiever, knows best what his self-concept is not like, rather than what it is like. Secondly, the characterization of "ambitious" appears strong on this dimension, because two items, number 53 and 85, were found to scale on this dimension. TAHLE 4.4 f Item Content of Dimension II (Males) Reproducibility - .749 Minimal' Marginal Reproducibility - .661 m c J m. Scoring Modal Response Item.No. Direction* Item.Content** 4Proportion*** 23 - Different (not different) .742 59 - Carefree (concerned) .654 117 - Easily Distracted (not easily .688 distracted) 10 + Logical .604 5 + Practical .638 v—. F'— *Scoring is in overachievement direction **For items which yielded a negatively scored direction, the positive meaning of the item.content is presented in parentheses to aid in comp prehension. ***Either “p" or "q“ which ever is larger An evaluation of Table 4.4, which contains the item content of Dimension II, indicates a description of an individual who responds in a socially acceptable manner in the academic setting. His responses are not motivated by a desire to make academic inquiries, but to impress on his peers and teachers that he is like everyone else, that he conforms to the norm.group behavior. He logically determines the practical ramifications of his behavior before he acts or reacts. 51 a f TABLE 4.5 Item Content of Dimension III (Males) Minimal Marginal Reproducibility - .584 Reproducibility - .743 fi Scoring Modal Response Item 113; Direction* Item Content _ Proportion“ 15 + Careful .708 1 + Patient .658 33 + Consistent .644 22 + Studious .550 115 + Efficient .557 41 + Systematic .517 101 + Competent .547 52 + Exacting .594 67 + Intellectual .621 2 + Talented .570 32 + Original 3523 36 + In-the-know , .527 *Scoring is in the overachievement direction . **Either "p" or "q" which ever is larger Dimension III, as presented in Table 4.5, describes an indivi- dual whose achievement is motivated by a desire to completely understand or to master the subject matter material as presented in class. Although both Dimension I and Dimension II are concerned with obtaining knowledge of subject matter, they differ in the motivations for and. the degree to which such knowledge is obtained. The fourth male dimension in Table 4.6, characterizes an individual who obtains his achievement by meeting his own academic or intellectual needs. He competes with standards which are his own, and which may or may not be those of his teachers or peers. 52 TABLE 4.6 Item Content of Dimension IV (Males) Reproducibility - .800 Minimal Marginal Reproducibility - .619 Scoring ‘ Modal Response Item.No. Directign* {Item Content Proportion** 71 + Alert .701 119 + Serious .658 98 + A Planner .523 18 + Purposeful .597 *Scoring is in overachievemant direction **Either “p" or "q" which ever is larger Results QEyMultiple Scaloggam.Analysis - Females As was the case with the males, the second scalogram.analysis yielded the most psychologically interpretable dimensions, with the error level set at .20. This analysis yielded five dimensions on which all but one of the 48 cross-validated female items scaled. Only item 85 (ambitious) did not scale for the females. Ambition was a trait which was particularly evident in the male academic self-concept. The five female dimensions present a different picture of the structure of academic self concept than did the male dimensions. 53 TABLE 4.7 Item Content of Dimension I (Females) Reproducibility - .811 Minimal Marginal Reproducibility 8 .758 a 1 - I; I Scoring Modal Response Item.No. Direction* ItemeContent*** Proportion**** 50 - A Person Who Postpones (A .907 Person Who Does Things Immediately) 20 - A Procrastinator (not a .888 procrastinator) 24 - Discontented (contented) .894 106 - Impatient (patient) .891 107 + Friendly .901 26 - Flighty (stable) .888 35 + Distractsble .846 54 - Stubborn (flexible) .878 40 - Nervous (calm) .801 104 + Reasonable .824 118 + Reliable .785 29 + Responsible .776 17 + Orderly . .795 71 + Alert .760 103 + Teachable .721 115 + Efficient .612 34 + Intelligent .516 12 + Smart .571 76 + Above Average .583 77 + Productive .612 67 + Intellectual .699 52 + Exacting .740 55 + Perfectionistic .782 2 + Talented .705 . 23 -(R)** Different .760 42 ~(R) Daring ' .734 59 ~(R) Carefree .644 117 -(R) Easily Distracted .760 5 +(R) Practical (impractical) .718 *Scoring is in overachievement direction. **Items 23, 42, 59 and 117 are negatively scored items which have been 'reflected", and are therefore interpreted in their original form. Item 5 is a positively scored item which has been "reflected" and is inter- preted negatively. ***For items which yielded a negatively scored direction, the positive meaning of the item content is presented in parentheses to aid in comprehension. ****Either "p" or "q" which ever is larger. 54 Dimension I for females in Table 4.7, appears as a general or global self-concept factor as did Dimension I for males. Dimension I contains 29 items or 621 of the 47 cross-validated items that scaled on any of the five female dimensions. This dimension seems to describe an individual who conforms to the teacher's wishes. She does what she is told, and does it immediately, and is generally responsive to the demands of her teachers. The teacher’s academic standards are the ones in which she is interested. Dimension 11, however, presents evidence for the existence of another standard. TABLE 4.8 __ Item Content of Dimension II (Females Reproducibility - .756 ‘Minimal Marginal Reproducibility 8 .592 Scoring Modal Response Item No. Direction! Item Content Proportion** 96 + Concerned .705 56 + Accepting .619 16 + Thorough .631 18 + Purposeful .574 97 + An Achiever .519 98 + A Planner .561 60 + Competitive .593 101 + Competent .564 41 + Systematic .564 *Scoring is in overachievement direction **Either "p" or "q" which ever is larger Dimension II for females might tentatively be labeled socially acceptable competition, or competition via conformity. An individual so labeled, would use the usual academic modes of behavior, but only to the degree to which she was forced to by virtue of peer competition. She is accepting of and competes within the peer value frame of refer- ence . _ 55 The last three female dimensions, although their reproducibilities are of an acceptable magnitude, must be interpreted'with caution, as only three items scaled on each dimension. This fact might seriously affect the stability of such a dimsion. was» , a w 1m Content of Dimension III (Females) Reproducibility I .750 Minimal Marginal Reproducibility - .639 W Scoring Modal Response Item No. Direction* Item Content" Profit-tion“ 81 +(R) Persuadeable (not per- .696 suadeable) . 9 +(R) Average (not average) ~ .638 89 + Contented .583 fingering isrin *overachifavemanA t direction ”Items 81 and 9 had modal response proportions which caused them to "reflect", i.e. "q" was larger, and are therefore interpreted in the negative direction. Wither "p" or "q" which ever is larger Dimension III appears to characterise an individual who con- ceives of herself as being not average and not persuadeable, and therefore not conforming. She is content to be independent and func- tions in the academic situation accordingly. Female Dimension 11‘ is somewhat similar to male Dimension IV. Table 4.10 the item content of Dimension Iv is presented. The item content of this dimension would tend to characterise an indivi- dual who is not competitive with standards of the teachers of her posts, but "plods" along‘doing what is required of her and no more. 56 TABLE 4.10, Item Content of Dimension IV (Females) Reproducibility - .823 Minimal Marginal Reproducibility = .596 1. ‘r# . EH Scoring . Modal Response Item No. Direction* Item Content Proportion** 10 + Logical .638 33 + Consistent .609 22 + Studious .542 *Scoring is in overachievement direction **Either "p" or "q" which ever is larger If the academic requirements and assignments are structured she will consistently attempt to bring them to a logical conclusion. TADLE 4.11 - Item Content of Dimension V (Females) . Reproducibility I .801 Minimal Marginal Reproducibility - .549 r:l: . 1‘ Scoring ‘Mbdal Response Item No. Direction* Item.antent Porpgrtion** 119 + Serious .628 83 + A Thinker .516 13 + Successful .503 *Scoring is in overachievement direction **Either "p" or "q" which ever is larger The last female dimension is summarized in Table 4.11. This dimension is difficult to distinguish from Dimension IV as both place an emphasis on a basic seriousness of approach to academic studies. Dimension V, however, describes an individual who is successful because she is serious and doesn't present behavior problems in the classroom, which if present, would tend to lower the teachers evaluation of her. 57 Orthogonality of the Dimensions An inspection of the item content of the male and female dimen- sions obtained from.mn1tiple scalogram analysis indicates that the inter- pretability is difficult because of the overlap of certain concept meanings. To determine the orthogonality or independence of the various dimensions, all the dimensions were intercorrelated, obtaining product- moment correlation coefficients. Intercorrelations were calculated for males and females separately using random porportional samples. The results of the intercorrelations of the male dimensions are presented in Table 4.12. One-hundred males were used in these intercorrelations (55 overachievers and 45 underachievers). TABLE 4.12 ‘ Intercorrelations Among Four Male Dimensions Obtained by Multiple Scalogram.Ana1ysis* ' N-lOO f I M Dimension 2; 22 '93 D2 .56 Dimension .23 .84 .46 D4 .75 .31 .68 *Values are positive unless otherwise indicated The range of the intercorrelations for the male dimensions was from .31 to .84. The results of intercorrelations of the five female dimensions are summarized in Table 4.13. The sample again contained 100 subjects, I 58 but with 53 overachievers and 47 underachievers. TABLE _4. 13 IntercorrelatiOns”Among Five Female Dimensions Obtained by Multiple Scalogram.Ana1ysis* N=100 Dimension D1 '22 ‘2; 24: D2 .72 D3 .13 .ll Dimension ‘24 .69 .67 .17 25 .61 .67 -.02 .41 4 *Values are positive unless otherwise indicated: Summary The item discrimination and multiple scalogram analyses results of a measure of academic self-concept were presented in this chapter. It was found that 48 items from the original 199 item word Rating List remained after cross-validation. These items, 35 of which.were in common to both sexes, were significant discriminators between statistically defined under- and overachieving eleventh grade students. Estimates of internal consistency reliability ranged from .90 to .93 for males and from .88 to .93 for females in various samples. Multiple scalogram analyses yielded four male dimensions and five female dimensions which accounted for 96% and 98% of the cross-validated items respectively. The four male dimensions had average reproducilities of .77 and average minimal marginal of .64. The five female dimensions had average 59 reproducibilities of .80, and average minimal marginal reproducibilities of .63. Descriptions of these nine dimensions were presented, together with the interdimension correlations for each sex. The interpretations, discussion, and labeling of the obtained dimensions of academic self-concept will be presented in the next chapter. CHAPTER V INTERPRETATION AND DISCUSSION OF DIMKNBIONS The dimensions of academic self-concept reported in the previous chapter, to have theoretical or practical usefulness, must be made psychologically interpretable and meaningful. The purpose of this chapter will be to interpret, label and discuss the interrelationships of the dimensions of academic self-concept for each sex. The item contents and dimension intercorrelations reported in Chapter IV will form the basis for this presentation. Discussion of Male Dimensions L Dimension I for males (see Table 4.3,‘ggggé.p. 49) gains partie cular significance in the present study, due to the fact that multiple scalogram analysis indicates the ranking of items as well as individuals on a particular dimension. This dimension is a global or general factor and includes all those aspects of an individuals self-concept which con- ceivably could be related to achievement. The significance of Dimension I lies in the comparative item rankings. The first ten items appear to be concerned with a general personality factor which is characterized by how the individual controls his behaavior. The next five items tend to characterize behavior which is "people oriented", i.e. behavior that is motivated by a desire to respond to an action or reaction of another individual.. In the academic setting the individual producing the response would be the teacher. The last ten items could be evaluated as intellectual characteristics which are associated with academic success. 61 It is interesting to note that this last set of items, the "intellectual group", are ranked at the bottom rather than at the top of the dimension as one might expect. The reverse in ranking indicates that students feel the evaluations of their school performance is not primarily a function of their aptitude. Because the concepts describe an individual who is interested in learning the class subject matter,“ because he knows this is required to obtain a grade, and because the defined behavior indicates the operation of extrinisic.mntivation, Dimension I is labeled, Achievement 1;; Traditional Academic 5213 gm. The items on male Dimension II (see Table 4.4, supra p. 50), characterize an individual who is a conformist, and is concerned with the practical ramifications of his behavior. This behavioral concern is related, not only to relationships with teachers, but is also generaliza- ble to peer contacts. He is not concerned with seeking knowledge, but achieves because of his conformity to the usual and acceptable mode of academic classroom behavior. Dimension II is labeled, Achievement via Academic Conformity. Dimension III (see Table 4.5, 32252, p. 51) characterizes the individual who is careful, intellectual, efficient, studious, and com- petent. He is interested in obtaining a complete understanding of sub- ject matter. He wants to master the basic concepts which underlie the reasons for,classroom presentations. Furthermore, he is interested in application of subject matter. The item content implies motivation to learn for learnings sake. Dimension III is labeled, Achievement via Intrinsic_ggtivation. Dimension IV for males is seen as a ranking of items which 62 describe the goal orirCnted or goal directed individual. 1 The goals may or may not be those of the teacher. A student who would choose these items is concerned with demonstrating his "brightness". Ila may undertake self intitiated projects which might be presented to the. class. He might be called an “intellectual show-off". logically following from a discussion of the items on Dimension IV, the label of Achievement 1i; m-Accmlisbment is given to this dimension (See Table 4.6, w, p. 52) Considered in totality, the four male dimensions suggest several interrelationships. Because Dimensions I and II are both concerned with the teacher ”8 evaluation of the student’s motivation a high intercorrela- tion between these two dimensions would be expected. Because of the interpretation emphasising social conformity, rather than an intellectual or academic ranking of items, it is expected that Dimension II will correlate highest with all other male dimensions. Both of the above predicted correlational trends are supported by the results reported in Table 4.12 (£2232: p. 57). The correlation of Dimension I with Dimen- ' sion II was .84. This fact might justify the visualizing of these two dimensions as combined on a "motivational cmtinum", being characterized at one extreme by extrinsic and at the other by intrinsic motivation. The intercorrelations of Dimension II with the other three dimensions justifies it's being considered as a relatively independent dimension of academic self—concept. The correlation were found to be .43 with Dimension III and .31 withrDimension IV. 63 Discussion of Female Dimensions Dimension I for females tends to characterize an individual, (see Table 4.7 22253, p. 53) who does what she is told, immediately, and in an orderly, efficient and intelligent manner. .This type of female student is responsive to thecdonforming demands of the teachers. She has experienced positive reinforcement from her teachers for this type of behavior. Ber self-concept has incorporated an expectancy of pleasing the teacher. Females characterized by Dimension I would represent the total range of academic ability. Dimension I is labeled, Achievement via .Tgaditijojtlgl Academic Role-Taking. The second female dimension describes an individual who is con- cerned with what people think of her. She desires to be accepted by her peers. She is also academically competitive. These two seemingly ' contradictory trends are resolved by being competitive but using socially acceptable, or peer acceptable behavior. In Dimension I the emphasis was on conformity to teacher's demands, in Dimension II, how- ever, the emphasis is on conformity to the expectations of the peer group. If peer conformity is characteristic of the achievement of females as a group, it would be expected that homogenity would be found in their academic performance. Research has frequently indicated that such an observation is justified. Dimension II for females will be labeled, Achievement V¥§.Z£SE Egrmative Competition (See Table 4.8, £222 9. 54). Female Dimension III (see Table 4.9, 33233, p. 55), describes a nOrv-conformist. She is independent and considers herself as not being avarage. She is content with her independence. In the academic setting, 64 she pursues academic interests which may or may not be similar to those of the teacher or her classntes. There is, however, apparently suffi— cient overlap with teacher's goals to allow for an acceptable level of achievement. Such an individual is probably above average in intelligence. Dimension III is labeled, Achievegt 1143 Academic Independence . Dimension IV for females, as presented in Table 4.10 (supra, p. 55), characterises an individual who meets the teachers' sterotyped conception of the "good" student, i.e. logical, consistent, and studious. A student so characterised would be primarily concerned with meeting the teacher's specified requirements, and in general with functioning within a structured classroom.program. It might be hypothesised that such an individual would be of average intelligence. Dimension IV is labeled, Achievement gig £29.38 Teachsr Mentions. ‘ The fifth female dimension (see Table 6.11,;259‘1’. 56), describes the student who is content to think, contemplate and investigate acade- mic problems. Although the concept ef "creative" did not scale on this dimension, it is felt that the content of item 83 (A Thinker), together with item 119 (Serious), tends to carry this conetation.‘ Creativeness in female behavior is traditionally and frequently valued more highly by teachers, than is similar behavior in males. The item content of this dimension also implies an "intellectual" type of individual. An intellectual, however, who is able to maintain academic endeavors within the teacher and peer group social context. In this respect, this dimension closely approximates the traditional conception of the "gifted" student. Following from the above discussion, Dimension V for females is labeled, Achievement via Intellectualizing. 65 With the exception of Dimension III, the female dimensions appear to have more common than unique elements. It would be expected, there- fore, that the correlations of Dimension III with the other four dimen- sions would be relatively small. This prediction finds statistical support in Table 4.13 (32255, p. 58), because the intercorrelations between Dimension III and the remaining dimension range from +.l3 to -.02, a low and restricted range. Discussion of Sex Differences Sex differences in academic selfnconcept are apparent not only in the dimensions obtained by multiple scalogram.analysis, but also from a comparison of the items in the word Rating List selected by males and females (See Appendix B). An evaluation of the discriminating items selected by each sex yields the following trends: 1) Females present strong indications that they do not see theme selves as procrastinators. Two items, number 20 (procrastinator) and number 50 (a person who postpones), describing this characteristic were significantly avoided by them (discriminated in negative direction). Males avoided only item 50. t 2) Females do consider themselves as being distractable. Items 35 (distractable) and 117 (easily distracted) were significant discrimi- hnators between female under- and overachievers. Only item 35 was a sig- nificant discriminator for males. These two above findings indicate sex differences are of degree rather than kind. The following indicate fairly marked sex differences: 5 3) Males evidence consideration of themselves as "ambitious". 66 Item.85 (ambitious) did not scale for females. 4) Males did not evaluate themselves as being "carefree", females did evaluate themselves in this way. 5) Heles, not females consider themselves as being "different". 6) Item 5, "practical", discriminated between males and females, in favor of the males. 3 In addition to the above noted item.differences, several inter- pretable differences in the dimensions obtained from scalogram analysis for males and females should be considered. Both the male and female Dimension I contained item content which implied taking the traditional academic role in order to obtain scholastic success. In the male dimension, emphasis was on extrinsic motivation, whereas, the female dimension emphasized intelligent con- formity. Dimension II for males is similar in content to the female Dimension I. Dimension III for females is similar to male Dimension IV, in that both describe individuals who are academically independent. The emphasis in the female dimension, however, is on intellectual indepen- dence, and not on unique academic independence. The significance of why a peer related dimension did not appear for males is difficult to evaluate. It may be that peer related academic behavior is not characteristic of males, or this type of behavior was not significantly discriminating between male under- and overachievers. Another explanation might be found by investigating the grouping of the items on scale, i.e. the absence of this type of dimension might be an artifact of the analysis procedure. A more plausible evaluation 67 might be that due to the nature of the male role, as it is perpetuated in our society, with emphasis on individualistic or competitive types of behavior, the frequency of male group behavior in the school setting is significantly decreased. In summarizing the discussion of this section, the following trends might be highlighted: l) Males tend to view themselves as being ambitious, concerned, conforming, and practical. 2) Females tend to see themselves as not being ambitious, but carefree, different and impractical., ‘ 3) The keynote of differences between male and female academic self-concepts is the selection of an identifying "significant other". Females tend to choose the teacher'pr peers, males tend to select teachers, or some other individual or group not identifiable in the present study. Academic self-concept is not a singularly generalizable or uni- dimensional trait . The male and female dimensions of academic selfcconcept, as interpreted and labeled in the previous three sections of this chapter, gain significance because of their similarities and differences, when related to the academic self-concept theory presented in Chapter I (supra, p. 4-7). Interpretation of Dimensions in Relation to Theory The four basic tenets of Brookover's social psychological concep- 68 1 which were used as the theoretical ception of academic self-concept, base of the present study, may be summarized as follows: 1) the student learns what he perceives he is able to learn, and 2) his perception of his ability to learn is a function of the expectations significant others have for him. The present investigation has demonstrated possible refinements of Brookover’s academic self—concept theory. Several types and sub-types of academic self-concepts were de- rived from the empirically constructed item dimensions. This finding is interpreted as demonstrating the statistical validity for the theoretical assumption that persons behave in ways that each perceive appropriate to himself. Appropriateness of behavior is determined by each person through internialization of the expectations of significant other. It had been assumed that teachers were the primary significant others in the academic setting. Interpretation of the dimension analyses data indicates, however, the presence of at least two other categories of significant others. One of these categories was identified as peer groups. The other category, which was associated with the "individualis- tic-independent" or "ego" self-concept type, was not identifiable by name. Logically following from this finding, is the hypothesis that variations in value orientations or behavioral expectancies on the part of a significant other for a given individual. 1* 1Wilbur B. Brookover, "A Social Psychological Conception of Classroom.Learning", §ghggl and Society, 1959, V01. 87, pp. 84-87. 69 Tenet III (see Chapter 1, supra, p. 6), of the basic academic self-concept theory is that; "The functional limits of one's ability to learn are determined by his self-conception or self-image as acquired in social interaction". Evidence for the empirical and theoretical validity of this tenet was found in the present data. The influence of social interaction was seen in the dimensions of both males and females. This trend was particularly evident in females, usually being indicated by conforming type behavior. A schematic diagram dipicting possible refinements of the basic academic self-concept theory of the present study is found in Figure 5.1. Self-Concept 1222 Significant Other Refined as Source 2f Values and Expectations W ,1 Actiyitz Content Traditional Academic Role-Concept (Global Dimension) /|\ J. Peers 4L Group I Scholastic Scholastic Teacher ‘ Ego 1 Teachers J, Inter- Personal 1 Individuals other teachers or peers d Individual n. Intellectual Figure 5.1 Schematic Representation of Academic Self-Concept Types in Relation to Hypothetical Generic Variables The three self-concept types are all seen as emerging from the Traditional Academic Role Concept. This global or general dimension was 70 found to be present in both males and females. The self-concept-type lables are arbitrary, but were chosen with a view toward theoretical integration of self-concept-types and hypothesized generic variables indicated in Figure 5.1. A note of caution is offered in summary. At no time in the pre- sent investigation was an attempt made to deliniate cause and effect relationships. Furthermore, relationships discussed in the immediately preceding section are of a hypothetical nature.~ Summary In chapter V the item content of the four male and five female dimensions of academic self-concept were hypothetically integrated. This integration took the form of, l) determining the psychological meaning of each dimension, and labeling each group of igems accordingly, 2) discussing the implications of sex differences in academic self- concept dimensions, and 3) evaluating the dimensions in relation to academic self-concept theory. The dimensions of academic self-concept determined in the present study were labeled as follows, for gglgg; D1 Achievement via Traditional Academic Role-Taking 02 Achievement via Academic Conformity D3 Achievement via Intrinsic Motivation DZ Achievement via Unique Accomplishment The five female dimension labels were; D1 Achievement via Traditional Academic Role-Taking D2 Achievement via Peer Normative Competition D3 Achievement via Academic Independence Dz Achievement via Mbeting Teacher Expectations D5 Achievement via Intellectualizing 71 After considering the basic academic self-concept theory in light of the empirical results of the present study, it was decided to retain the four basic tenets, but make the following refinements; 1) consider the academic self-concept as consisting of one global or general dimension and three sub-types (peer, teacher, and ego), 2) consider individuals other than teachers as being perceived as a signi- ficant others, 3) depending on which significant other, variations in value orientation, behavior mode, and expectations should be anticipated. CHAPTER VI SUMMARY, CONCLUSIONS AND RESEARCH IMPLICATIONS Summary The major problem of the present investigation was to determine the dimensions of academic self-concepts of eleventh grade male and female eleventh grade students. Theoretical assumptions were drawn from the psychological percep- tual theorists and the symbolic interaction frame work of social psychology. Self-concept was viewed as a functionally limiting factor in school achievement. Three therotical assumptions were made: 1) the student learns what he perceives he is able to learn, and 2) the teacher, as a significant other, has an important influence on the development of a child's self-concept,-which in turn affects his ability to perform in the academic setting, and 3) under- and overachievers will differ significantly on a measure of academic self-concept. A one-hundred and nineteen item rating scale was developed which purported to measure academic self-concept. The student was asked to rate word or phrase as he thought his teachers would, in describing him as a student. A four point rating scale was used. The instrument con- structed was one of several experimental instruments created for inclu- sions in a motivational test battery being used in an on-going research 73 project conducted at Michigan State University.1 The measure of academic self-concept, labeled the Word Rating List, was administered to samples of statistically defined under- and overachieving eleventh grade students of each sex. Using Chi square model to determine item discrimination, it was found that 40% or 48 of the original 119 items held up after cross-vali- dation for each sex, at the .10 level of significance. It was further found that of the 48 discriminating items, 35 or 73% were in common to both sexes. Hoyt’s analysis of variance technique was used to determine the reliability (internal consistency type) of the total scale scores.2 The resulting reliabilities ranged from .91 to .93 for males and from .88 to .93 for females in various samples. Using the cross-validated items for each sex,1multip1e scalogram dimension analysis was performed. This analytic procedure is a nonpara- metric, datareduction technique for maximizing interitem reliabilities, such that both subjects and items are uniquely ordered, i.e. form a uni- dimensional or Guttman type scale. Multiple scalogram analyses set at an acceptable error level of .80 yielded four male and five female dimensions which accounted for 96% and 98% of the male and female cross- validated items respectively. An index of reliability (stability) of the resulting dimensions was obtained. This index, reproducibility (R), “—*-‘—__— 1William W. Farquhar, A Comprehensive Study 2§.£h£ Motivational Factors Ugderlring Achievement g; Eleventh Grade flighFSchool Students, Research Project No. 846 (8458); Supported by the U.S. Office of Educa- tion, in cooperation with Michigan State University, 1959. 2Cyrl J. Hoyt, "Test Reliability Estimated by Analysis of Variance", Psychometrika, 1941, Vol. 6, pp. 153-160. 74 is a measure of the degree of accuracy with which the item response patterns can be duplicated from knowledge of the total score alone. The ' four male dimensions had average reproducibilities of .77 and the five female dimensions had average reproducibilities of .63. A summary of the four male and five female self-concept dimensions, together with their respective interpretive emphasis is found in Tables 6.1 and 6.2. Conclusions On the basis of the results of the present study, several conclusions are made. Following from the item analysis results of the Word Rating List it is concluded that: 1) A reliable measure of academic self—concept can be constructed. 2) Under- and overachieving students differ significantly in their academic self-concepts. 3) The functioning of an objective instrument which measures academic self-concept can be accounted for by theory 4) The items which significantly discriminated between male under- and overachievers are not essentially different than those which discriminated between female under- and overachievers From the results of dimension analyses of the cross-validated items for males and females, the following conclusions are warranted: 1) Academic self-concept is not a unidimensional trait. 2) A major dimension is found in the academic self-concepts of both males and females which holds common evaluative and interpretative significance. 3) Despite the fact that the interdimensional correlations within each sex are generally high, several relatively independent sex linked dimensions are present and interpretable. When the results of the dimension analyses are related to basic .muommoua moumwuficH >Ham3pw>fipaa mo muoHumuum 1:09mn .mocmpwmmpm meu mamow mama no Hmcomumm udpEfimHHQFooo< msvficm mH> quEm>mwsu< an .coHum>HuoE ma Hmaumuma mo asmommz .umuumE oomflnam mo muoHumowaamm cam .muamucoo mcfixaumpcs .uasouwxomn Roux ou whammn COwum>Huoz uwmafiuucH mw> ucme>mH£o< ma uofl>mnwn sooummmao mo mmazu manmumooow haamauom cu zuHEu0maou >2 ma ucw8o>ounom mo apex. muwauowcou ofiamumo< mH> unmEm>mwno< Na wcfluumw spasm u0m vmuwsku ma u“ no nouuwa uoofinsm wcwcpmma aw pmummuouaa Heapfi>qvc« mo :oqumaHUmma wcfixma oHom ogEovmo< Homaguaumufi «we udmfio>ma£u< HQ wfimmnmam o>auuumeWaH Hoes; scamcymmm null} 1 l Ammmuzv mum>oasumum>o can chops: «Hm: mo munooaoocmaom anovuu¢ «nu Ca mcowmcme«a xumEEam a. o 39% ‘JI . 1 .mofioamuumaxo EooummmHu mo 3oH> a“ huHHmo nwuomum mo aowumuwpwmcoo u505uH3 summaumo>aq xaamucoe kumcow :H can mumamaouaou .xafiau cu unsucou ma 0&3 Hmava>wvafi ca manquuoon wawuHHMSuomHHouaH mw> unwEw>mH£o< ma mucoEoufisku nmwmwommm mumnomou waaumma co mfimonmfim .Hmsuo ucmuwwacmwm mm umnommu aufis cowumUAMfiuaovH mcoHumuumdxm nonumoH wcwuwmz mH> ucme>mwsu< an mummhouCH owfiwcwom :30 mmamusm .mwwum>m no: pom ucovamawmcfi .umfiauowaouuaoz ooamwcomoch owampmom mH> uao6m>mfi£o< ma mcoHuMuommxo was oocohmmmu mo madam odaw> mo oUHDOm ma maouw poem .mmoawpmsm Hmauom cowuaummEou o>fiumeuoz ummm mw> uaoso>ofino< Na .ucwwaaamuaa mam unwauawmo .mauovuo mH .ucmEcwwmmw some hamumfipmeau xwmu use mmuuumo wawxmw oaom oaaopmo< HchguavmuH aw> unmem>ow£o< Ha mummnmem wwwuoumuouaH Ho>om.COamaoE«n {Mull} I Amanizv mum>oanuauo>o was tacos: damask mo muaoocootmaom oHEmpmot one CH maoamaeawn kumsasm. N.o m4mmm .< .n mumnavumm .3 .3 ha meflmEoo wanna mafia a on. ucmnnom an. uamnnom mm. vcmHHom mwfln< .cBoum mofiuwamno owonummsm wmcouca wood .0 sofiuonumo o>HmH=mEH .m nowumnumo Hoozom mnam> non moon .q :owumnuow AmumcHumMMUOpm Ou mocmccwu .cowuom Ou mmmcw>flm -Hume mo xumnv amnmv auh>nuu< .m cunnmnumu mucmEmuflsvmu qumwmom Ou xH03 Hoosom £ua3 unsoum .o mommmHo vmxHA .m mumnommu vmqu .o HHMM mhwflomwu UHUM .m Eooummmao was mamme mo~n< .mBoum Euomcoo Ou mmmcwcHHHHBGD .N sowumnumw .mmofim CH oEHu once ucoam .N mucoscwfimmm Hmcoauno weavmou NmSuH< Annamaomhsv sand: .H «nowuonuou was mmw>oe um «Ewu mama ucoam .H mmHmEmm van mofimz mmHmEmm com moan: . mwumfimsm mmmnwow. mmumflmsm wwmiwom MOHzH Hzmzm>MHmu< 304 nz< Mmoza mOHzH Hzmzu>mHzo¢ mem nz< mu>o m0 mUHHmHMMHU¢mo= .5H \m. Haouuom uoaamcoo mmuH ooaoauuaxm .NH magnum hvaum uduauammo AHon uoa on monu Hook .oH mm. Hamuuom uoHHmcoo mama oocoauwaxu .oH AHo o>Hu .oa mmocm>wmcommon Hozmsmow3 mach mfinmumvmoa wwo>< .MH nm. Hamuuom HmcofiuoEo vaHouucoo mood .mH i mmocm>wmcoamwu csoum o 3m£m Hmofiuwuuummkm .NH hm. Hamuuom Hmcofiuoso mo xoma wmxumz .NH mucmumm uwwcu Eoum vmuom3 zonu mafia a“ mwownu Hmflumume mcowuuoaoum Hmoaonosumm :3oum a Swnm man :muuow uo: on: moan uHoh .HH mm. Hamuuom wowsommu mmoco>wmum>ouucH .HH mmmfio cw cowusouum anchm mo sofiumnuou wcfimmm CH xufiaowmmflv uwumwuu .oH @309 mowamSUmE mo Ammuw umummuo .oH monuficnomu hvsum Lowumnumo wsHm: CH muasowmmwo Hoummuu .m wSon cowmconmumEoo wawvmou Hoswwm .m muwnmn was unscaanowu mm. vcmaao: umoum xaamHoOm mu¢ .w Hmmz .vzon zwoum umuumn “whoa wwwwaum .w an. ocmHHom mumooaoo waom w>wufimom .n awkwEowo3 “momma ucmsoouh .m MOHzH Hzmzm>mamu¢ 304 nz< mmazs mOHzH Hzmzw>MHmo< mem nz< mm>o ho mUHHmHmMHU4uow mambo was wouaoHHO >44m400m OOH .ou compose UHumHHmmm .0N nowuoonuo mcoso GOHmmmumuv cu zoaovaoa .mm on. Shanda: 4mEuom mmo4 o>m£ mucouwm .mm mc04umnaooo 44mumv umoc4osn cmmuoz. aw mcomuom mo mumououc4 o>om .qm unannmo mocwum4maoo now noon Honwam .qm om. humSOUZ uuonn3m umuoB moduHOm .m~ uumsnow “mono How noon Hocwfim .mN unannmw owcwno you vow: noanS .NN unannoo :04umooouuc4 you too: Hongm .NN unannmo GOHuM444wm< new wood uo£w4m .4N uumnnoo. uaoam>u450m you now: nonw4m .4N mm 4m: 04oz «Magda—w: nod-0.444. I awom 39W .MwMHM I 49490 muozumw mm. 44muuom ufioau ou HOHummc4 whoa 4ooh .0N mm. 44muuom AHmnmn m.uo:uoa usonm mm. 44muuom mo mwn44oow m>4mcouxo who: .m4 mm» 44muuom u044mcoo mmm4 cosmHuomxu .m4 Aum Homv AHm4com 4mc0mumm usonm mm. 44wuuo: a04unooumn Show Houoom .04 mm. 44muuom uo44maoo mmm4 mocw4umaxm .w4 m08<0HHmm>zH Hzmzm>u4m0< 304 02¢ anza m08<0HHmm>zH Hzmzw>u4mo< muHm 02¢ Mm>0 mo mUHHmHMMHU044mz. um4fiuomcou .om 44m£oufiz mmma4£uuo3ca mo mwcw4ooh .mm >044mz >4ucmsuouw whoa 00:04040m40 .nm 44onuu4: wcauuom 4oow o>4um>uomdoo .um meow , >044mz mcwwcm44m£o noo4om uoc moo0 .om 44unuu42 «>0u4um0800 uoz .om ou9w4m 4nmmmoooam xNOB uaaovouo annousu >044mz . mmm4 so vo>4oouo0 Hannah .nm 440£Uu42 :04uummm4umu owu mcwwuno .mm moHMEum mo4maoh mmww “Maw. www.mowu mWoMnmsw. Mwm4wuow: :mwuoz . o>4mm=muom .qm compo: uamcHEov whoa mu< .mm mEon mumm>40 Eoum >m3m coon coumo mmm4 0mm .Nm cmwuoz zu4u=ume umwuouCH £w4m .Nm monomunm u4onu pow %HMm mo4whom 4M400m mo mumm>40ummomn mmo4 momusoo u4mcu u4mm .4m cmmqu. 4mo40>u mumoumucw o>mm .4m 44m3 mmuoa HHoau mumm>40 ou4amwno uoc 04:00 kmnu u4wm .om cmwuoz o400mcommmm .om coxmu wuoB zwcu mow mama mcu mouoc mumo>40 uwmzu vo3m4>mu :mumo mmm4 .mm ammuoz m4nmwcm0w0 .mN xuoz um43o4uuno mumm>40 muuxo :4 muzos whoa unmam .wN cmwuoz muHmeuCH cwwm .mwN mOHZH Hzm2m>mHzU< 304 02¢ 0002: MOHZH HszMDMHmo< EUHm 02¢ Mm>o mo wUHHmHmMHU¢m4naomo0 .mq 440£ou42 mGHuuum 4mow 04uu44mou=0 .o¢ 2044mz wcmum4muom .mq . muons UHBovmoo moouuso o4Emvooo now vuwkuu 44mnou4: icon :4 atoms omo 4400490 coo .no 2044M: 4o4uonm o>4w uoc moon 2448mm .nq muo>oasoo mustM Haummoounm 2044a: numoca nonuo £u43 mouoHUOmm< .qo 2044a: mm oo>4uouo0 hunuwh .oq oucmauomuom uoH >044mz mvumaou 4m4000m mom: 2448mm .mo 2044a: umHEpowaoouso: mH .nq CH woumwuouc4 2044mz muomHASm co 24:0 mxuo3 .Nq 2044a: umoucoucM4wm 0000 .oq >044m2 unmonoo M4om uoom .4e 2044mz 4oo£um nox44 .mq 2044mz mwwmum oa4m> uo: mmo0 .0¢ 2044m2 coflumu40mm mo 4o>o4 stm .mq nonumm mo Emwo4u0uo 2044mz Eouw Moonwamo mv4o4nm nonuoz .om >044mz 24ucm=v0um mmo4 va44040m40 .4o >044mz mmm4 oum0 .mm >044mz mocmvcmmovCH mumum00 2446mm .oq mum>o4nomuum>o >044mz monsooomc4 >44m400m 4oom .mm >044m2 umnuo nu43 mqu4uomm< .om 209<049mm>24 Hzmzw>0420< 304 0z< 22020 209404Hmm>20 Hzmzm>0420< onm 02¢ mu>0 mo mUHHmHmMHU0u4c90 whoa mucoumm .nm w:4uocw4 mm. cm300 unaumEEH .om m3ou0 0cm wGHumaHEov muscugz .om vmumosvm mm. assoc monoumm cmfiuwuHuonuam o>mm .nm mouofim nouuwn mum? 0:3 mucoumm no: .mm efinno anmo no mo4oaovcuu 04uouu5oa Ho U4uonomm0 anon umHHM on was mm04wamm um. assoc 2w40mfiv >4uamaooum mug: .qm ooum4m 44mam Scum 0800 cu vcmH .qm owumanuwua_vao wa4vmou occauohumuu Hoseauooauo mm. :w300 :4 muamquEou mo xom4 .mn mouo4m nonw4£ 040: 0:3 muonuoa 0mm .mn mm. assoc hocoa was mEHu we on: uoom .Nm mouo40 unawHM ucouwm4nw4m mos Hoauwm .Nm mm. cmaow nawuvnu43 204m400m .4m ame4omm=z monoum o4aocooou0400m umame .4m 30. :m300 m4ouuooo own xmo3 .om mo44wamm “044mam .om moUHOno o4Eovmoo . wcfi404m um. cmzow was 4o:04uomaooo oamo> .mq cmEHmmmsz ummwcaox Ho umov4o 244mnm0 .m¢ OHQEmm van o4oz_ m4mEmh can o4qz mWoMnmuzm. www.mflom Mmmownw 4m.om.u.m. LN4M uWoMHMHQ 4N0M0m SMHM 209<0Hamm>2H 8222M>MH20< 304 024 «M020 209<0HHmm>2H azumWVMHmod mem 02d mubb mo moHamHmuaoauamo ho mUHHMHMNHo40m480=m mumnuo CH um~h0ufi4m40 mummumucfi mSOHHom oz «woufinam HmomUw augm. .40 .00 .mn oHMEmm 0cm m4mz Anv mm. emcee Anv mm. awaou Anv mm. nmsoo m30h0 m3ou0 moumwm mouo4m mm. smsoo 4B0 mm. ewsou ADV mm. nwsou mmoco>4uoomwo o4Ew0mom mo mason mom muwnmn zvnum 000w mom mCHEHOmcoo vocowumwsvcn mvuovcoum u490¢ GO4uMU404uam00 noon umuumm m444xm awamuwvmm4 so nonwwm muonuo mo wcwummuu¢ cofiuofium 4mc0muoaumua4 mo oucmm0¢ 2u4mos4cm 0cm 44433444 mo 4M0000 .50 .00 .00 .00 .m0 .N0 .40 .00 ..mm m4mawm 0cm o4oz 39s human moms... as. 94 208¢0H8mm>2H Hzmzm>mHmo¢ 3C4 02¢ mmnzb. mo mUHHmHmmHU¢M¢mU MOH¢0HHmm>zH HZNZM>MHEU¢ mon 02¢ 2M>o mo moHHmHmmHo¢2¢mo 000 mm. amaoo unmaunmma annmcomuom .aa 200 mm. nmsoo Au44uu4> nacho .mn 2u04o:0mum0 000 mm. awaou saw wane mpfiumumuuaH .55 2u44m004umu Any mm. awaoo 0am 2u084amaow mmuammngam .05 sauna AHHouuo .nh mawu0 «44000 .05 A00 m0. nwaou unmanm44maooom mo madam mom .mm A00 mm. nwnou «nonunm mo nuoaano4uom u4uo0 .Nn as. amaoo ououm =mnamnuoeoamm=Hmmuwmm .4“ Any mm. assoc smocflsunaHE menu .0“ H02: :0 swoon dowuo04400m uaow4440 wow 00. nwaoo :oum4>o0 040uomonohmm: nocme .05 A00 mm. nwaou 0ma4oum=m How 2u40o0m0 .05 40:: no . BooummmHU 00 mo. smaoo . mnoum :uanmz: uo£w42 .00 m30h0 avamsmc hu44mou ou um5h0¢ .m0 wuuu40 udmfiumznvawa who: .00 mouo40 mmvmuw swan o94m> .00 o4maom 0cm o4m: . oHQEmh 0am m4mz 4.92.0.0“. mWoMfimn—w. aflomow. :w4w. N .Mcmouu mWoMnmaum. 4M0mow swim moa¢0HHmm>2H Hzsz>MH00¢ 304 02¢ 00020 009¢UHHw0>2H Hzmzm>mH00¢ MUHm 02¢ mm>0 ho mUHHmH2MHU¢M¢mU . ho mUHHmHMwHU¢M¢Jw 95 mmmm4wmm o>4unmmd nnonu ou acnmmonnxo m>4uuommw mm. nnmnann unsung m>nm on mnna mama .mn mnonumw um. 44w0542 nu43 040m004uo4on o>4umwoz .Nn ”gnu: mwumnmam H Show awnm ounwnm mononm ounmnm Nunsu .AAV mm. fiwfiou ounonm mononm n¢a4uuoaz Any mm. amnoo Anv mm. nmnou ABC mm. nwsoo ucoEo>onnom now mumma nonwnm .mw 04umnuoao0 once and 003 mnwnuox m>om .00 mung: awomfimqnm nmomum awn: muons @504w44on :4 u>4uuo «no: .50 Amn>m gov nwnwnn nunnauma unmnouan 4wa04uwnauoo .00 m4uom hnouonoa nonwnm .00 4u04004 Ionuhnn vouuafivo nouuom .00 96 omuonvaon now manuooaomaoo .nm uaovnma00304oa .onuo484uao .Nw 4onunoocm4om .40 hndfibwmnm>40 0am m904m>4nm can no a04uuofim2 .00 o4maom 0cm w4mz N. Hawav- mWoMHmn—W 4M0Muw. swim. 20H¢GHHmM>ZH HZMZH>MH20¢ 304 02¢ 2002: no mUHHmHMMHU¢M¢mU o MOH¢UHHmm>zH HZMZMDMHmU¢ mon 02¢ mu>0 mo moHHmHmuHo¢m¢mU mao4non0 nunnx xmm o>mn 24uamnvonm one: .00 wananmo4 xoon swan ovum: nunsu nuna 4n03 cu mn4mm0 «no: .05 _ . mcw40 4m004uwuo> 00w nunsx 4mqonumuavo onwm> 0am 00u4544 .05 mm4mawm 0am mo4wx mWom. “Maw 4N0mow. nflfim. mwmomnw. mamagw4wl nunnx 040m 00 cananmo 0000 .qm nonuuouuu .ovnna .uuonuuan nun92 bozo 0:3 mucmnmm m>m0 .mm Bonn nommmmn 0am unnDM muconmn ommo40 cu unm3 .Na ’ mo4maom 0am mo4oz deMHmww 4Mdmww.nmwm.mmeM£m.mMWMdmame. acnuncwouon munmnm naavn>nuan non can: amnm .nm nonnmunnonuam once . mononm . and? 0:3 onwnuoa no: .55 aqe4omnsz 544oao4uo20 vouanuum nwuumm .om “w «04oaum mo4qaom mMuMBun—w. 4Nomum awnw. mwuomfimam 4N0muw. nflnm oonu4m o>4maonwwm one: .05 spans wan saga nunnnuuos no novaunuu< .nn mwan4ouw o>4mmonwwu um. nnmnsnx usonu nuunxaa was unnpo .en mm4mz awomfimsm Hmomww nmnm 208¢04Hmu>2w, Hznzu>0400¢ 3mm.02¢ 22020 MOH¢0HHmm>2H 92020>0400¢ 0040 02¢ mu>0 mo mOHHwHMMBU¢M¢=o ho moHHmHMHHU¢M¢wU 04304 0400000000 0004 .00 04304 .004 04304 400404n0 0004 .00 04304 0:0n0a00 .004 04304 0>40004um0>04 0004 .00 04304 04008 :4 00um0n0004 .004 40040004400 .004000nv 04304 0:0440num40m 0004 .50 04304 0040000 403000440004 .404 4.000n0u4800 ..mwonmv 04304 08000n=000>00 0noz .00 04304 .0080: n04n0000 aonw 0800 .004 04304 2000404000 40nzu490 0040 00000 .00 04304 00444800 n044080 .00 0040800 000 00402. 0040800 000 00402 0M0“ Raw 2M0 W0M8M4M mMuMnmn—w mam 0M8M4M 0n00808 244800 00080 Nunsx 0400000000 0n02 .00 nunnx 0008000400 0000 000800 0000n00 .00 nunsx mmmnumon unoz .mm. 00000 804044008 annmnonmun unannxm .50 nunnx 5000003 0n02 .N0 nunnx 400404n0 .00 . 0040n004 3000 00nd: 000004000000400 00 3004 unoz .40 nunnx Eonm 0040000Jm000 0>4n00 .mm 04000n000 2440040200 \ 98' 0040800 000 0040: 0040800 000 0040: mWoMHMQW. 400500 5W4m mwaonnu AME—0080 4W aflowmaw 4M0m.0.m 440.44% £MsMnm4J mwwcmaw4m. nzmzu>nnmo< 0040 824 muse no monnmnmmn0zn mo mUHHmHmmHo¢M¢mU MOHzH 99 00080304000 00403 00040404 00 00nm 400m .504 004000nnx0 400040080 00403 00 20000000 000 00 80000nm .004 0000n00 00 0040040n 00403 04 0000004000400 00 0044000 .004 0040: 0MUMHM1W.2MOMOQSNJM. 009¢04900>24 H2020>0400¢r0040 02¢ mm>0 Mnh<04Hmm>24 Hzmzu>0400¢ 004m 02¢ mw>0 mo moHBmHMMHU¢M¢mU A no mUHHmHMMHU¢d¢mo .00n000 00n0000n 00 00400040 00040800 now 2000n0044040 00 n0000N 004000 .0 .0 00 004400004 2000400000 804 20000 on 004u0n40004 0000 000 004n00n00 20000 0000 0>0nv 0n0>04000 0300 0840 00 0040004n0040 004300 .0 0300 1000 0000 0n08 20000 .0 44000042 n04>0000 0>4000000 .0 44000042 00000000800un0>0 .0 5m. 440nn00 2004000 44000040 0040 000 00400040 00n0 .0 4400004: 0040000 4000 0>4u0>n00000 .0 an. 0004400 0000 44000040 00080>04000u000 04000.40 .0 440000442 400.3300 00 00044000 0>00 .m M 1 00000>4u4000800 no on. 0004400 . 00000 000nu0 0 20 0004n0000 .44000042 00400>4008 00004 .N 4400004: tn000 000 800020 4000400>4uoz .N 00n0.04800000 0000 0n03 04800000 00:0nnu N44000042 n0000 04 00000 000 4440400 .4 N44000040. 004000004000 000 0040000 .4 0040800 000 0040: 0040800 000 0m40z 00000000 0004400 00000000 0004400 009¢04H00>24 9200000400¢ 304 02¢ 00020 009004900>24 92000>0400¢ 0040 02¢ 0000 00 0040040000¢0¢00 00 004H040090¢0<00 4002009x0 UHZMQ¢U¢ 0HH3 00H¢40000¢ 00480400H0¢0¢00 2HH4¢200000 20¢2200 N.¢ 040¢H 200040004000 0.000000 20 0000004404 2443000 0008 00 44000040. 04003 00404000 0400v 000000 0400000 000 0000 000 0044000 0008 0 04000000 0040000000 00004400n0040000 00 0000000 04000 00 004000 .04 0 000040040 20 000034008 0008030400< .04 0040004000 0 00 00008 4000000 000 0 00 00 00000000 40000000 04000000 .0000000000 404000 .00004402n0040000 00 0000000 0000n000 00000 .04 0 000040042 20 000034008 00080304000 .04 0200 0 0000000 4400 200002 :000 000 .000 0000 0040 0 4440002 10440030040000 10444440 000000 000 0000000 04000006 0000000000 .004000000 404000 0040004400 00 0800 .04 0 000040040“ 20 000034008 00080304000 .04 .00000 04000040000 0040 000 00 0000 04 0000 0003 2040 I003400 000 400000 0040 00030 :00 2040044440 04 0000000440 04000000 0008000000 004000000 40 000000 0000000 0 004000000 .44 0 000040042 20 000034008 0008030400< .44 0000400000 20000 04000006 00300 004 03400 004000000 00400 04 2040044440 0000000 .04 0 000040040. 20 000034008 0008030400< .04 00040 04 004000000 0400000 000 00004000 00\000 00000004 004000000 004200 04 2040040440 0000000 .0 .00004402n0040000 40 0000000 0000q000 000000 .0 00040 04 :4000 000000: 0 000 000000004 00000008 2000 4004 .00000000004 000 0300 0043 000004000004 04000 .0 004000000 0000040 04000 40 0008 0040 .0 800 0000000000 .0300 000 0000000 0044040 00 0008 .2 00400000o 24008000 04 00008004000 000 .2 009004H00>24 82000304000 300 024 00020 008004H00>24 Hzmzm>040o< 0040 024 0030 00 0048040080<0<00 mo 004904009040400 101 102 nsoum hufiuonusm Amano mmamav van anm cu huwawumo: mo mwcfiammm .m~ cmwuoz mmonusa mo mmemsoauwm .mN akoum Ahaao mmauav vaw 3mnm mumauo mo Hmowufiuu hauw>o .NN amwuoz anHmaommmu cam wanmvammmn .NN Azaao mwamsv mcowummsuuo mmawm mmwcamsn . sumac: a“ mnemuwn mo mummuwucH .HN amwuoz Azaao mmamsv w>Hmm=mumm .HN Akaco mwamev maoaummaooo Hauuov mmmaaman :mwuoz aw mnemumm mo mummumuaH .om amwuo: Ahaao moamav uwumfiaaumo .om mflumsuau tuawumouum can uhom a nouvaUHz. waquwm “mach vumzou vamua .mH d unannww Ahaao mmamev ucwumfimcoo .mH waux mauonuau awnaaouuuosm anuoa .uoaoan .uhom Amado uquav umwa Guam a acuafivvuz_ mo Huuaov vuwaou vamuH .wH a ununnom vam onwanwuo ou o>aun .wH hnmunz maumnuau a Haauuuz w coumavwfiz acqmum>ouuxo vumBou vcwuw .NH awake: unmafison .NH oaHaSuao wag Ahaao moauav w nouuavvfiz axmmm whammmam vuwsou vcaHH .cH ammuox Roam“: huaHSuma ummuouau .oH muonuom mumewa van commafioz mucovsum umnuo amnu wowonu Ho mofl>umm Hmaoom ca aaomuon Imafixmom coauwmsuuo mo uawvwmaoo mama .mH amwuoz mo Hmofimzu ummumuaH .nH MOHzH Hzmzm>uHmo< 304 92¢ mmnz: \MOHzH Hzmzm>mHmo¢ mem nz< Mm>o mo mUHHmHMMBUfiamauawv cam m>qu mm. Hamuuom xuwx uuunuum nasmaoo Anus awuouaononumum .om an. unmade: oHnquOmcs uu< .om moumH can mucmamufinvuu uHamvmua mauoua mmamn< .naoum ou auowcoo cu mawaafi3ca .mN an. vcmafiom mew xufiummmo «0 30¢; .mu wahx use: 834 w unmanmo moumH vcm 8:4 ucu8w>wfinom new wwwc mmawn< .naoum mannaummuuoua cu vamH .mm ammuoz uo ~>mfinom ou :ofium>auoz .mN umom Azaco woomH new a unmanmo mmawev acmumm uwnuo you moawn¢ .nsoum uom ou mmmcm>wmwumv mxumq .NN ammuoz sumacOU van «0 mmmcmum3< .NN A>Hao mmamav mmvnufiuum uoawmum hucmwowmwamuwamm cw“: .om camper ofiumaamou can Hamunwamcu .om aaq umcwmum muwamcoHuOSm swam .nu :mwuo: ucmvfiwcouuwaom .mu mm. Hamuuo: umcwMum .m vmumSvaHme tam manmumc: .qu cmwuo: Azaao mmamev uaumwumcm .¢~ mOHzH quzm>mHmu< 304 oz< mmazs MOHzH .wxm>mH:o< :UH: az< xm>o ho mUHHmHMMHU«uqaca «bin .o¢ mm. Hauuuom cannon a an «unsuonu amok ”cc anmafluxn no. mo «mum on. 282.5 K£33. Ba 3:8 «.2 .3 S. Sate: 2: 3 «238.38 :6» .8 donuuuonxo unquaou auuoaaw aw uunvnou haunu mm. aaauuom you use mnunuuuu away-Ian .wn fin. Huuuucm anon. uanauuouafio doom .om maul hm. afiuuuom nuanun uaoau :uouaoo uouuouu .nn an. dungeon ounuuouaa onanqaou m>qm .un ouonuo no awaaaoou cu o>uu um. ”awaken handcuUGau no nunqauou u>um .on on. vauaaom unonuou and ad vcunuuouam .on uni—Eva“ ununaunovaa mauvanabv uncuuuauwu nu umuvu on you nu uan noduuouuv noun: R. 3933 .8. ~26 3:38 .3 on. 3:28 2:35—33 :2, 38 .nm nouoouun an. afinuuom uoauaou uo>o uuuaunoo .«n an. vanuaoz law canuuo>uu u ouuouo .cm mm. Hauuuom and“ nunnuum huaaunOuuoa Huuauouuhm .mm on. vauaaom voaaouuaounaaoz .nm moa¢uuamm>zH quzflbummod 304 92¢ mmaz: moazH quzu>uHmo< muHm 02¢ mu>o ho moHBmHmmHo<¢uuo guinea .5; 5n..HHuuuo= nu-ououuu uaaaauaqa . an. unduaom can o>aumoouuuxo one: .o¢ um. Hfiauuom maul dawn-oumaw hanks: an. unufiaou can-uopnu ...fi . .ouaouu .nq a Agauuuz hamuax an. unmade; «>«afiaaaq .H .eq a gdduuqx 3.3.336 banks: on. vanuaoa anathema. caucus“ coda o>¢m .nc a uauuuqx moa¢unamu>zu azuzm>uamu< sea nz< xuaz . uaafimmamm>zH » ‘ «vacuumcoulmaaa uuuuqauumo .Hn .Huaom man can: :3 Pam ”8:3 a. flu uuommu ouauqa one: .on unsung: waqhauovua noon «and .mc sand-houuon doonoo. asb,u50p¢ uuafiuaoo sung .m< uuuauoanvu Haauuo>o coco .n¢ ouoac>«- Iannaow.uuao«uoao no xo-a . voxuuauooono>qunomuou . educauoao vanaouuaoo coca .o¢ mnuuavao one: .nc omauno anon: vouuouaou coca .¢¢ ouu050auununxo «.04 .n¢ MD,moHHmHmMHon~mu<,=uam az< nape no moaamuanaufiuum “masofiuuau naqum cw haaaqudeAmou mo waouumau< mcowuwmom cu ammonu uoa mum: .om mm. nwaoo Amaze mmawav mo>Hmmamsu cmnu waouum8u< uwvao muoHcngou vmuummwum .mm mm. nwsoo usuamvam can muHHHnmvaommc .cofiumummoou wcouum8u< so mwaaumu HmBOH vucfimuno .«m mm. nwsou auscuuum umvsx so :uoaamcou waqvfio>m pow moawummmum: mo mmuw «nu a“ wcouumau< mwuoum 30H whoa vmcfimuno .mm mm. amsou anneaumuo> young mo mamum Hm:0flumu=n taco «nu ma uuuouu and mo wcouumau< umaasa kumuuw voaauuno .Nm mm. nmsou .uo> uwvnx kn vmuzmwua mu mumuumuaa unacfiEov wmomusa mo mmwamaowuwm owmmm .wn mmoaH=mcmHm .nm mmwcaaumvuo .om Mu¢fi0Hmuu>Hv can anoH n£>fium msu mo coauwofiox .mm coauao>cou mo moamuamood .«m acquaoaammq adamaaav and vocamumam you huwommmo .mm Hausa saga wwuwm uoa vav gowns mm. amSou wcouumau< mamom HugoHuwuo> manusm vmm .Hm mm. nwaoc 3 Houucounmamm .Nn MOHzH Hzmxm>uHmo< zen nz< mmnz: A, MOHzH Hzmzm>MH=o< muHm 92¢ mm>o mo moHHmHmmaumwaa Ho ucuam>mwnum camcmam a Nunsx 30H saga mucouum omoono .¢o moumam mapaxmam muoz .mo Ahaco mwamav awaafimmm mHaumo: you umauao uomuuv _ mHuuHH .voonvafinu maumm muawvcoa mm. HHmnaHx aw uum>o whoa .aoammmnww4 .mo muumam nuts“ mH> uaoam>uwsu¢ .qo Amaao meuav mum>mfinum Hashes cmzu nowumo 1x. .awmaucmvw oaaaaamm vumBou luamauouaou mm. Haunaax hoauvaou uuuuoum abnm .No uuumum ma> wouafimuo unwaw>mwsud .mo ,mwao.mwwwev. Hmumaow ca Gowuwuzvo huwmoaanw mm. Hamnafix vuwaou avauauum o>aumwuz .Ho mm. amsou van Hafiaufiaa mo amazon .mo Ahaao mwamav muwuo ufiquSa :30 Hanna ca muaamn moumam u nuuuuaua ou macaxam van mm. amsou mm. Hfimaaqx mofiufiaflnu uaonu o>quaumwn .oe mm. nmsoo muosuo mo muaaumooo¢ .Ho Amado «mamav Ammaafiomm mufluofluomaa nuwav mm. Hamnaax .muaawum ouuuonwamv umoaH< .mn mm. awsou muaamua> .oo Ahaao umHmav mvwon %uaovcmm sud uaucuaoum .hu«>aaudm .hua uuuuwm hoaoaoawmu mm. Haunawu aheaummca mo owcwaouw waouum .wn mm. Awnoo amsuooaaounfl vaw auscuuum .mm moazH Hzmzm>mHmo< 304 nz< munz: «OH¢GHHmm>zH Hzmzm>MHmu¢_mUHm nz< Mu>o ho moHHmH&MHUuw>aummuu no. swoon nau>qson Hm«00n< .¢n w Nunsx vcm muaamcfiwauo .nm Axaco moaqev fimsuuwdfioucfi cu vouoaao mm newsman augmuuvnua save a cmeuuy mqmwumucw muq>auum umwcouum .MN a nuuax mo mmauaausv awnanxm .qm Axacc unauav mowumwpwuouumnu cowaosm mm>~uufimnu no man“: a xxxm vwocaoco»na m“ auqaqumom .mn w uuusx coficuno swan s~u>wuwaom .mm mega aw“; Oman ma uGQEo>oanua :Omcmzm Hchauwuo> can HuCOwuwu: accumSm Hoozun mo mvuuucmuu 6 «guns Iva mamm> can cmuasfia u>mz .an d uuusx anon: ovaoauw omoono .Nn wcficummfilxoon acmcwsm amnu mucus uwmnu and: somcoam u>auuonaan can ¢ uuuax wcHLuQEOm ow umnumu case: .0“ w uuuax Hamaucman maowumfimu uuom .Hu acmcmSm ucmEumsnvm Hmaxom w nunsx Ioumumn mo mEmHnoua w>wz .oc mm. smsoo ucosnmfiagEooom mo omcwm .05 mm. cmaou acmcmSm mm>amm a uuuax namzu CH wucmwfimcou xqu .wo mm. swsou mqumn hcaum @009 .m0 camcmsm mmmam>au ¢ Nuusx mmamnca .mo mm. cmaoo uuomwm anmvmum mo umcmm .mo cowamzm zufiawaoaumu vcm ¢ uuHDM manmmwcmno .cc mm. cwaoo muwsaamaam co mammsmEm .no quzm>uH=o< 304 az< mmoz: mOHzH Hzmzm>mHmo< :uHm Qz< mm>o MOHZH ho mUHHmHmuHo«uuuom nonwonu . Hucoauunnooo\oaauvuuu mm. ausoo ausuuunuqa oquunuuad .Nm mm. autos onwawmov van uuoao .Ahado neg-av uoouou. . . tau Hauoom cad ausuuoaauu .nn. ausoo muunuo a“ uuououaaown .am avvo a agency law uomnouuu unnabu tuna untao>uqnuu cue novaoa you Havana o>auuo ou Bonn condo has occuunmuuu . . . .uuom van mfifiaum mo muons mm. dunno vanguaomoouua .ounuuaaH .ow nuuabnanom AAHQM aw nodunmnou vac Hanan uaflu> nan macaw nonayam and can» «yes how aouu mm. nuabo ognauoonaa no Huow 02 .mm a sauna , ,Iuoavo.vu¢mou cu vaoa uuwuuuvnou condos“ madam aunuaw mm. ansoo caucusoa van uwuoauhnm was a swung ou xuoa Hoonuo.ouufiox - hunuauumunn nouaozm. canon Hideauuuo> van R. 5.25 :38 v5 Huang—Bu; .: a 3.5m 1833:? :3: 26: noncoam wauauuoanxoon aouu mm. Guano uaouuaou emu xumn .om . a nunsu acauuauuauuo o>wuun ouwono noduuauau Huaoauunsuuo vau uwaov noucyam. aoouaouau,aw humus 5n. nunbc anon uuuuuuovca van uuuaoaa .mn a uuusu .vfiuwfioaa hmaauuauvnu< moath[r. Hzmzu>uumo< BOA 924 MMQZD mOHzH Hzme>NHmu< mWHm n24 mu>o ho moHFMH¢MHoHuuu :mHm3 Ahaco mmamav mmanHmn Hmmm .qw swam: mumwuqaa ou mmum .om Ahano muamav Ahaao mmamav swam: aw vaEmn .vmuowuuawu Hmmh .mm £ma¢3 mmOfiono oxma cu «mum .mm mam: «Hm: muoafimam Numuamamau uauau>manum . 0>Humauu ou vmumawu m~m>au Huanoo lemon huauauua Huanuoau .wm mmamamm van mmauz muuufinsm Hoonum swam nwnounu mumunuauHm omwa auumam mo 3wa> vmucmaHOthuH>quuw mm. am3ou .vmuHHmHUOm .uwummwmanuau .mm mm. awaou mmmao>amusmumm .mucwcwaon .ow sowumoo mm. cmsoo sauna“ .mmmaw>umm«5umm .mw MDH<0HHmm>zH HZMZM>MHmo< 30A nz< mmnz: moazH Hzmzm>mHmo< mon 92¢ mm>o mo mUHHmHmMHU«uuwua can doaaaonuu v:«9n no .unouuu no a09u¢>o .ouauqfinaoo announu Ahado ouauav nowuu>uuuu Haw scou:0uuu can .u>«aonu:9 nm9u3 Masada .h90>auaumov and .59 Juana o>auuauuaaou a“ madman .nm ahaco Ahaao moaqav uoaqav uuuawaawcodon. snags vmuuHqu Ho wouoofiuu 9009 .09 and»: we wnaaoom uaoauum .Nm. mOH¢UHHmM>zH quzu>MHmo< 309 924 mmqr‘ moezH . Hzmzm>mHmo< muHm 92¢ mubo ho moHBmHmuHuuw> m9 Hague: .m acuuonuoun93 ov9u9 Donn cu new mucus nonuoz .9 90090» n9 mouw9m n09umo9aaaaou mv9o>u uwAqu .n aouuonuouaau 9903 0v 0» new «aunt umnuoz .m «mocmn9ma9 wv9auao u9uowuoao van u>wuuu moum9m mws9uxw cu avaou umnuo: .o aouuonuoun93 ma ou no. magma umnuoz .e Xwa :35 .99 no mumauouc9 woum99 mo x9mu monouunmam nonuo: .n aouuonuounfil mph; ou can Quads nonuoz .m ua08909m>uv n609u900v ooum99 uumuo9ouum ou mumamuum Hugues .q Ibuuonuoua93 n30 uxma cu now manna umAuoz .q acuuwuumaoo n9 muum99 hunuoa mo ouamv9o>< .m lbuuonumun93 9903 ow ou v9990 mafia: uuAuoz .m Hugues am9umu9uonu=ucaoa cums uo9ua mm. 99wna9x ou unwanouuuw «@090 .N ouum99 m:9>u«noutnwun mo oumnuqz .N «muum9m w>9mauua9 who: .9 quuowm uHuwuuoauv «was «Hanan: .9 umnuoz nonuoz m9qz m912 moazH mmu>n99990< 95.99 999929 9999.40993sz mMMBWWSV $595 MES 90 UHHmHMMHUu§o< 96.: «~52: 52 935 mo HZNZZOMH>ZM MSUE 9O mo mUHHmHMMHU900909 90:00: .09 ch90009800 .00990099990090 009900 .0000090v 909009 00090a<.9 00009800 mm. 990989M 9093 999000900909 0>900w0z .m9 0 00009 009 wc9090000 0:0 8903 909002 .m9 9990 00090:<.9 wa9a9090 000E0>09900 00 009059 mm. 990089M :00900909 000:905009 9009 .09 a 00009 n990aou 00009w9aw9m 0005 909002 .09 909009 909008 99 009000990 009099 909>9000 mo m0>0999< .M9 800000909093 90 0090090000800 90090999 .m9 #900 90090 050 9000 090900 9093 009w99 :30 m9: 909 800000900093 :00 009 000 0000 909002 .N9 800000900093 9: 00000 00 000 00003 90900: .N9 593 930 :93 . 0002.3 30 3:3 33000:. 800009900093 09990 093 000990009 90900: .99 800009900093 990 00 :00 00003 90n0oz .99 900009 009099 9000000900 0900009 909002 .09 800009900G93 0 09 00 com 00:03 909002 .09 9993 009950 0000999 009099 90099 00 00980000 90:00: .9 800000900093 030 09: 0308 00 :00 00:03 .a moe<099mm>zH mmm>9990<.NEOAM.99929 90929 mmm>9990< AmUHIV 99>o 90 UHHmHMMHU00 0800 0099009900004 .09 009099 0093 009900 0908 090000: .99 0000 0909000989 0099009900000 9903 0090809 99. 00300 90 09000 00 000 000909 .nm 009099 009309000u0090 90 0900002 .09 900009 000 9009mm 90000: 09000900 0800 8099 090809 0 300009 000000 00090 0908 900009 .09 0090099 0 9909900 900009 00 90990909 9009 .mm 00009 .0089900 98000000 000900009 .99 009990990 0099999 0 9909900 009000 000 0000909 90000: .99 00009 .0089900 0>909000 0000909 .09 0000909 00 00900909 9909909 900008 90 90090300 09009 .99 00903. 09 0000009000900 90 0099009 .09 mu 1 0000909 00 0090009900 900090080 9909909 900008 0093 00999000 09 .09 00903 90 90000000 000 8000099 090: .99 90000: - 900009 000 90000: 09080M 000 090: 090: 900 00039008 00 000900<.9 909902 0090309 9090090 0000 999809 .09 0 00009 98000000 0000900000 900009 .99 0009m0< 9 . 900009 000 900002 0 0099 00009800 0009 900009 09 0000999 90 0000909 0000 0908 000900<.9 0090009 9099099 900 00099990090 999809 09009 .09 0 0000.9 000 03909000 0009 900009 .09 mo9<09900>29 000>0900< 90000 0002: 009<09900>29 000>0900<1900900 00>o 9D UHHmHmMBo0000 mm. 9909900 0900008 00000 00999000 0009 .mm 00909900 90 000800090 09 00800 908900 000009 03900990009 000 0099009900000 8099 099909 90 00080>09000 0 03099 0908 0903 003 900008 0>00 .09 00080z 00 90990909 00800 000090 8099 000 090900 099909 90 00080>09000 900000 .99 0009000 00803 0900900800 090 0900002 .09 0000909 900002 00900 00900099 00000 0900 000 09 .90 .OHflawh VG“ OH”: 8000 09 . 00900 00999 090099 3000 0000909 .90 909902 000000090009 0900009 999809 .90 . . . 0000000 09900 000 000909 0003000 909902. 09800000 909 090309 9090090 oz .9m 0090M 009000990 90 00000000 090099 .09 0000999 0000 . 0000909 99000 909902 9900000099 0009 00099990099 .00 080 3:0 0.9200 0000 09 . 03090 000900 000 000000 000 0300 900009 000 900002 0 300m 9000 0000000 0990000 9009 .99 909902 900009 90 0000090000 09009 .99 009000000 090300 000009000900 0009909 . 90 9090000 0908 000909 .09 900009 0990910000909 90 ., 99. 00300 0900900000 .00009800 0000909 .99 009099 . 009900990000 0902 .09 0800090050 0003—0000. 0.009. 00000! M $900.990ga0 _. .. w. , , 000000000. 000.000 008 90 09000009000400 . . . 90 0090009000400 116 amnmoa a msmwa Nupax swung Nuusx uuwsoaanmm N. m. . Qw300 mm. Hamuuom mw>m4 demand a acumwum .Hamvamuo acmnmm w \ coummum .Hamvcmuo flauummuuvaao ou uoonmmu suas u>wuwcsm whoa mum>mw£um swan mo muamumm .oq muawumm uuwmmwm .nq uawuum ouuoaa cu uaws mucousum .qq unauSum ou coauowwmm van ovaua .ummumuca 305m mucmumm .mq udwfiw>mfinum afiamvmom mom uwvum: m>auum cu Hwav Ia>avca «mama mucmummuom mafiamm mo uoum ca aoamsmaoo vac apnea .Nq hua>aummuo van coaummo lupus“ .mmucw>«mmwauwm kn uwuw nuuuomuwsu unweaoua>cw maom .Hq 117 uwnumm vam Hague: Honuam nuas maowumHmu m>fiumwwz .o¢ uwnumm uwnuos o>fiuuououauum>o nuaz vmuw«00mmm unmao>manuw swam .mm uHOQQSm chowuoam How uamvcmamv mama .mm ucuEm>ofinum you vumsmu Hmaumuma uumuaa .mm mOHZH mmm>mHmu< Azoqu munz: moezH no oHemHmmeomHmo< AmuHmv am>o mo oHamHmmau no codenamau uuuuouw donoaaux_.uuuxnom hauquowmunwau «>6: uuuouam .nm vuuwwnm unuau>uanuu you wawusauuum .tn vuuwwum u>auuouounuuo>o an aunouum 3mm .mn nuauga annouuoao vuuwwum aw muaxuua on uuauuum 3am .Nm nunuuun ou vuuwmum manuaowuaaou o>uuuwua Hauuaou .an o>auwabm haaauwnom can vuuwwum wawuuufiun on aunouun voauq> .on uofiau Haunu cu maauwnvu vuuwwum hp uuaouun mo «unannouuu awuu .mq uduvnuu auupauoa ou cougauuuu mm. adsou van uuuuuuau xocu mucouum .m¢ Acquuno ou manwuuom van Human mum down: no doauauonaa onuv mauvauaov as mm. nuabu vHHAu do acqunou usn caucuum .n¢ moazH mmm>uHmu< macaw amaz: .moezH mmuwuHmu¢.AmuHmV mubo mu nHamHmwao893 >gaa £8323 £8323 Teachers feel that I am: Teachers feel that I am: *1. patient 1 2 3 4' 21. unreliable l 2 3 4 *,#2. talented I 2 3 4 *,f22. studious 1 2 3 4 3. ‘dull l 2 3 4 *,#23. different *4. inefficient l 2 3 4 #24. discontented 1 2 3 4 *,#5. practical 1 2 3 4 25. energetic 1 2 3 4 Teachers feel that I am: Teachers feel that I am: 6. clever 1 2 3 4 #26. flighty 1 2 3 4 7. curious I 2 3 4 27. very active 1 2 3 4 *8. confident 1 2 3 4 28. pessimistic 1 2 3' 4 #9. average 1 2 3 4 *,#29. responsible 1 2 3 4 *,#10. logical 1 2 3 4 30. creative 1 2 3 4 Teachers feel that I am: Teachers feel that I am: I ll. unsuccessful 1 2 3 4 31. a follower l 2 3 4 *,#12. smart I 2 3 4 *32. original 1 2 3 4 *,#l3. successful 1 2 3 4 *,#33. consistent I 2 3 4 l4. "blah" 1 2 3 4 *,#34. intelligent 1 2 3 4 *15. careful I 2 3 4 #35. distractable 1 2 3 4 Teachers feel that I am: Teachers feel that I am: *,Il6. thorough 1 2 3 4 *36. in—the-know 1 2 3 4 *,017. orderly 1 2 3 4 37. childish l 2 3 4 *,#18. purposeful l 2 3 4 38. decisive 1 2 3 4 *19. uninterested 1 2 3 4 *39. rebellious l 2 3 4 #20. a procrastie 1 2 3 4 *,#40. nervous I 2 3 4 nator 121 U) 3». Hr-lm “as a: 33:13 > clams-a 0 zzcn :3 .¢ 2: Teachers feel that I am: Teachers feel that I am: *,#41. systematic l 2 3 4 61. impulsive 1 #42. daring 1 2 3 4 62. unreasonable l 43. cold 1 2 3 4 63. dependent 1 44, below average 1 2 3 4 64. a "wheel" 1 *45. reckless l 2 3 4 65. a "grind" 1 Teachers feel that I am: Teachers feel that I am: 46. energetic 1 2 3 4 66. fool-hearty l 47. "sharp" 1 2 3 4 *,#67. intellectual l *48. dependable 1 2 3 4 68. sociable l 49. shrewd 1 2 3 4 69. retiring 1 *,#50. a person who 1 2 3 4 70. driven 1 postpones Teachers feel that I am: Teachers feel that I am: 51. a goof off 1 2 3 4 *,#71.. alert 1 *,#52. exacting 1 2 3 4 72. critical 1 '*53. lazy l 2 3 4 73. brilliant 1 *,#54. stubborn 1 2 3 4 74. casual 1 #55. perfectionistic 1 2 3 4 75. adventurous 1 Teachers feel that I am: Teachers feel that I am: #56. accepting 1 2 3 4 *,#76. above averagw l 57. persistent 1 2 3 4 *,#77. productive 1 58. submissive l 2 3 4 78. relaxed 1 *,#59. carefree 1 2 3 4 79. a "brain" 1 #60. competitive l 2 3 4 80. optimistic 1 122, Sometimes Usually Always b Teachers feel that I am: #81. persuadeable 82. motivated *,#83. a thinker 84. conforming *,#85. ambitious Teachers feel that I am: 86. unusual 87. independent 88. determined *,#89. contented 90. outsider Teachers feel that I am: 91. aggressive 92. a person who delays 93. indecisive 94. irresponsible 95. non-critical Teachers feel that I am: #96. concerned *,#97. an achiever *,#98. a planner 99. a leader 100. indifferent Never .. E: .._. as as 2 3 2 3 2 3 23 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2f. 2 3 2 3 2 3 2 3 2 3 2 3 2 3 2 3 Always Teachers *,#101. *102. *,#103. #104. 105. Teachers *,#106. #107. 108. 109. 110. Teachers 111. 112. *113. 114. feel that I am: competent inconsitent teachable reasonable inquisitive feel that I am: impatient friendly fault-finding reserved dominant feel that I am: inaccurate touchy passive pushed *,#115. efficient Teachers 116. *,#117. *,#118. *,#119. 123 feel that I am: tense easily dis- tracted reliable serious Never H H Sometimes Usually Always b) ROOM USE WU mmwg a ‘1. if 'f ’ ‘JP (.0 m‘ t). I; {I 9' ,i figs} (I? .434 I" f: it“ 5‘]. A 4 .1: ‘3.- "lllllllllmfllllMET