University of Southern California Dissertations and Theses

A Measurement Of Attitudinal Change Following A Ten-Week Guidance Class: A Semantic Differential Study

(USC Thesis Other)

A Measurement Of Attitudinal Change Following A Ten-Week Guidance Class: A Semantic Differential Study

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has been microfilmed exactly as received 69-644 SEIM, Robert Martin, 1920- A MEASUREMENT OF ATTITUDINAL CHANGE FOLLOWING A TEN-WEEK GUIDANCE CLASS: A SEMANTIC DIFFERENTIAL STUDY. University of Southern California, Ed.D., 1968 Education, guidance and counseling University Microfilms, Inc., Ann Arbor, Michigan A MEASUREMENT OF ATTITUDINAL CHANGE FOLLOWING A TEN-WEEK GUIDANCE CLASS: A SEMANTIC DIFFERENTIAL STUDY A Dissertation Presented to the Faculty of the School of Education University of Southern California In Partial Fulfillment of the Requirements for the Degree Doctor of Education by Robert Martin Seim June 1968 This dissertation, written under the direction of the Chairman of the candidate’s Guidance Committee and approved by all members of the Committee, has been presented to and accepted by the Faculty of the School of Education in partial fulfillment of the requirements for the derrree of Doctor of Education. n Z T f a * /f£/ .......... j Dean u idpn 'et Committef / / ) ’ Chairman IT v&ZLfiyi c\ ... ACKNOWLEDGMENTS My interest in this subject dates back to a workshop in 1960 led by Mary Perry. The guidance possibilities in a tenth grade course which would combine information, group procedures, and a teacher's concern for an individual's existential dilemma seemed unlimited. Since then, encouragement for an evaluation of such a course has come from Jean Henderson, Clarke Cosgrove and, most of all, from Dr. Earl P. Carnes— friend, honest critic, and Chairman of my doctoral committee. My wife, Dorothy, has been a constant source of encouragement through out this enterprise. The test instrument described herein was adapted from the Webb-Harris Word Meaning Test (copy righted) by permission of Dr. Allen Webb and Dr. John Harris. The computer program was written by Mr. Robert Uzgalis for the 360-75 computer at Western Data Processing Center. Computer time was made available without charge through the courtesy of International Business Machines. TABLE OF CONTENTS ACKNOWLEDGMENT S LIST OF TABLES Chapter I. THE PURPOSE AND SCOPE OF THE STUDY . . . . Introduction The Purpose of the Study The Scope of the Study Importance of the Study The Statistical Hypotheses Definitions of Terms Limitations Basic Assumptions Summary Organization of the Remaining Chapters II. REVIEW OF THE LITERATURE.................. Studies of Group Guidance Techniques Summary of the Relevant Literature on Group Guidance Techniques Studies of the Semantic Differential Technique Summary of the Relevant Literature on the Semantic Differential Technique III. SOURCE OF DATA AND METHODOLOGY ......... The Sample Population The Treatment Variable— The Guidance Class The Method of Measurement— The Semantic Differential Technique Chapter Page The Test Instrument The Collection of the Data The Problem of Criterion Analysis of the Data The Chapter Summary IV. RESULTS OF THE INVESTIGATION The Findings Testing the Hypotheses Other Findings The Chapter Summary V. SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summary Conclusions Recommendations SELECTED BIBLIOGRAPHY APPENDIXES ......... APPENDIX A. APPENDIX B. APPENDIX C. APPENDIX D. APPENDIX E. APPENDIX F. APPENDIX G. Instructions for Administration of the Test Instrument Directions to Students The Test Instrument: Name Card The Test Instrument: Numerical Listing of Concepts The Test Instrument: Sample Tabulating Card Demographic Data Collection Sheet Ethnic Survey of Schools Included in Investigation APPENDIX H. Description of Samples Lost iv 62 95 107 113 LIST OF TABLES Table Page 1. A Tabular Representation of the Statistical Treatment Employed in Recent Semantic Differential Studies ........................ 34 2. Median Family Income for Area Surrounding Selected High Schools.........................38 3. Description of S a m p l e........................... 40 4. Stanine Score System ........................... 41 5. Factor Analysis by Item Summary ....... 63 6. Means and Standard Deviations of Pre-Test Post-Test Change Scores for Stepwise Discriminant Analysis of Concepts by Schools............... 66 7. Discriminant Analysis of School Groups— Step 1 .......................................... 67 8. Discriminant Analysis of School Groups— Step 2 .......................................... 68 9. Discriminant Analysis of Test/Control Groups— Step 1 ..................................70 10. Discriminant Analysis of Test/Control Groups— Step 2 ..................................71 11. Experimental Group Pre-Test Post-Test Change Scores...................... 74 12. Pattern of Change Responses Among Schools for Evaluative Dimension on School-Related Concepts..................................... 76 v Table Page 13. Change Score Response Pattern for All School-Related Concepts as a Function of Guidance Class Instructor ............. 82 14. Post-Test Response Pattern of Boys/Girls for School-Related Concepts........... . 84 15. Post-Test Response Pattern for School- Related Concepts of Students in Guidance Classes by I.Q. G r o u p s.................... 87 16,. Rank Correlation Between Vocational Interest Concepts on Kuder Preference Record-Vocational and SAVII ............... 90 17. Direction of Post-Test Responses Made to School-Related Concepts by Students Other Than B l O ' s .......................... 91 18. Summary of Analysis of Variance.............. 92 Figure 1. Alternate Forms of the Semantic Differential ............................... 48 vi CHAPTER I THE PURPOSE AND SCOPE OF THE STUDY Introduction One of the pressing problems in education today is that of maintaining optimum concern for the individual de spite ever increasing student populations. Concern for the individual is, without doubt, the sine qua non of all coun seling efforts. To insure the individual the opportunity of seeking the counsel of a qualified counselor, a realis tic counselor-pupil ratio must be provided. Whenever the American School Counselor Association's recommendation of one credentialed, full-time counselor for each two hundred students is not possible because of the shortage of avail able workers or when budgetary limitations prevent the as signment of counseling personnel as suggested, the school staff should remedy this disparity by incorporating group methods of counseling as soon as possible. Most of the previous efforts in group guidance have been limited to career planning. These group methods of vocational guid ance have been employed since the 1920's, although they 2 have not always been successful or popular (44:161-181, 46) . The widespread use of group guidance programs in secondary schools today would suggest that the newer methods are successful (1, 5, 8, 9, 11, 48) . As popular ity alone does not provide proof of the value of group guidance, a study which at least indirectly evaluates the effectiveness of one of these current programs seems war ranted. Specifically, the group guidance program with which this study concerns itself is a ten-week guidance class instituted in September, 1962, throughout a large metro politan school system as a required subject for tenth grade students. Although it would be difficult to assess the ultimate worth of the class in the lives of the students who have taken the course, there should be some value in studying the course as it is generally presented and, in turn, in evaluating its effect on students. An evaluation of this class will be made by student responses on an in strument designed to measure the change in meaning assigned to certain selected concepts. It is assumed that change in meaning is related to change in attitude and that change in attitude precedes actual change in behavior. The Purpose of the Study There is no objective evidence reporting the success or failure of this particular curricular experiment, known as tenth grade guidance, which is conducted in each high school throughout this large school system. The central purpose of this study is to discover the change in meaning of certain school-related and work-related concepts as held by a sample drawn from the vast population of tenth grade students after they have been exposed to the re quired guidance class in the school system mentioned above. This study can, at best, provide only a partial clue as to the effectiveness of such a class in meeting its stated goals. Instead of using ultimate criteria or the opinions of students and teachers, an indirect method of assessment, known as the semantic differential technique, will provide the data from which generalizations will be made. In the investigation an effort will be made to dis cover whether this test instrument can detect a difference between student responses made in a counselor-taught guid ance class and those made in such a class taught by a full time teacher. Two of the twelve classes considered in the study were conducted by a counselor who would be responsi ble for the students in his guidance class as counselees throughout their three years of high school. The other classes were taught by teachers assigned the guidance class as a part of a regular teaching assignment. Hope fully, the students in the classes taught by the counselor will assign a more positive meaning to school after the completion of the guidance class than their responses indi cated at the beginning of the course. Of less importance, but of great interest to this investigator, is the anticipated capability of this par ticular test instrument to duplicate the rank ordering of vocational interests previously done by the Kuder Prefer- erence Record-Vocational. A rank order correlation will be made. Taking the Kuder inventory is an optional class room activity, and not all students in this study will take it in their guidance classes. The primary purpose, then, is to see if the ten- week guidance class is able to so affect students that a change in meaning (attitude) will be indicated on the post-test scores of a semantic differential instrument. The known variables in this study which may affect the test scores are the sex and intelligence of each stu dent, the influence of the teacher in each class, and the unique environment of each school. Although the individ ual pupil's socio-economic status is not described, a school-by-school comparison of median family income in the area served by each school broadly hints at this vari able . The Scope of the Study Twelve classes of approximately thirty-three stu dents each were selected for this study. Nine guidance classes and three driver education classes were tested in September, 1966, and again in November, 1966, at the con clusion of a ten-week cycle. Of the 382 students who were tested in September, only 297 were tested in November. The loss in student population was primarily due to pu pils' transferring to another school or class, dropping out, or being absent on the day of the testing. Selection of the three schools used in this study was made by the Committee on Research Studies in the school system previously mentioned. Randomization was attempted within the school district, but because of com peting federal studies, was not ideal. High, average, and low ability students were found in varying numbers in each of the schools. The twelve teachers who adminis tered the testing were given uniform directions. The tests were returned to this investigator through the cooperation of the principal and head counselor in the three three-year high schools from which the samples were drawn. The students were asked to complete the test as a regular part of their class work. They were not volunteer subjects. Importance of the Study The need for economy in education reinforces the demands for group procedures in guidance. If a course in group-guidance proves to be effective in helping tenth grade students adopt positive attitudes towards school, work, and themselves, then its use in a large metropolitan school district would appear to be justified. A study of this kind will provide a better descrip tion of the tenth grade student's perception of school than was possible before the semantic differential technique was introduced. The results of this study may provide information of value to other investigators of group guidance procedures in the classroom. The use of the semantic differential technique in the measurement of vocational interests provides another test of the versatility of this technique. The Statistical Hypotheses Except for the testing of the one research hypothe sis, the null hypothesis will be tested as follows: 1. It is hypothesized that there will be no signif icant difference between the experimental and control groups with regard to test-retest scores for all concepts on all factors or dimensions. It is hypothesized that there will be a positive significant change for the experimental group for those concepts which relate to general school orientation ("school,1 1 "teachers," "boys at this school, 1 1 "girls at this school") when the evaluative dimension alone is measured. It is hypothesized that there will be no dif ference between schools on those concepts which relate to general school orientation ("school," "teachers," "boys at this school," "girls at this school") . It is hypothesized that there will be no dif ference between classes taught by a regular teacher and classes taught by a counselor or grade advisor on the concepts which relate to general school orientation. It is hypothesized that there will be no dif ference between boys and girls on the concepts which relate to general school orientation. It is hypothesized that there will be no dif ference between high, low, and average ability students on the concepts which relate to gen eral school orientation. It is hypothesized that this instrument (the Seim Attitude and Vocational Interest Inventory) 8 will be unable to rank the vocational interest terras so as to correlate significantly with a pupil's scores on the Kuder Preference Record- Vocational. Definitions of Terms Guidance.— Guidance is a program designed to assist pupils in personal, social, educational, and vocational ad justment. Counselors are guidance workers. Group guidance.— Group guidance extends the guid ance functions of the school into small and large groups which may or may not be homogeneous. Tenth grade guidance class.— The tenth grade guid ance class is a required course for tenth grade students. It deals with orientation to high school and with educa tional and vocational planning. Semester.— A semester constitutes one-half of a school year. Approximately twenty weeks are in a semester. The guidance class is in session for one-half semester or approximately ten weeks. Tenth grade.— The tenth grade is the first or - ‘ freshman year for students in a three-year high school. Driver education.— Driver education is a one-half semester classroom course required of all high school stu dents in California. It constitutes the other half of the semester course for students in tenth grade guidance. The driver education classes in this study act as control groups in the experimental design. Counselor.— The counselor is the pupil personnel worker who teaches the tenth grade guidance class and is also the student's regular advisor. (The "teacher" is the instructor of the tenth grade guidance class and, of course, is not the pupil's advisor throughout the three years of high school.) Concepts.— Concepts are the stimulus words which ap' pear at the top of each of the mark-sense cards which com prise the test instrument. Seventeen concepts are used in this study. School-related concepts.— In this research instru ment the school-related concepts are: "school," "teach ers," "boys at this school," and "girls at this school." Dimensions (factors).— There are three factors or dimensions in the twelve scales. The evaluative factor consists of the adjective pairs kind-cruel, true-false, wise-foolish, and ugly-beautiful. The potency factor is composed of the following adjective pairs: hard-soft, strong-weak, light-heavy, and small-large. The activity factor is made up of the four adjective pairs passive- active, calm-excitable, moving-still, and fast-slow. No other factors are considered in this instrument. Semantic differential.— The semantic differential is a combination of association and scaling procedures de signed to give an objective measure of the connotative meaning of concepts (31). The task of the subject is to rate a number of concepts such as "me," "my ideal self," "work," and so forth along a scale which has seven posi tions between two bi-polar word opposites such as: kind : : : : : : cruel. The subject marks 1 2 3 4 5 6 7 his response in one of the extreme positions or in an in termediate position. If he responds in space number one, he indicates that he sees the concept "me," for example, as kind. Space number four is a neutral position and space number seven, in this case, is the extreme opposi-^ tion position from kind. The position of the response represents direction and intensity of one's judgment. Measurement of meaning.— The ways in which the cen tral semantic processes vary from concept to concept, from person to person in the communicative product (words) are measured at an ideational level. Words are emitted rather 11 than elicited behavior. Meaning will be measured here in an indirect manner by a meaning polarity test. Semantic space.— Semantic space is multi-dimen sional. Both direction and intensity are measured. Principle of congruity.— Osgood presented a theo retical model to predict attitude change or resistance to change. This involves reference of the original attitude toward the source of the message, the original attitude toward the object evaluated by the message, and the nature of the evaluative assertion. Degree and direction of at titude change are assumed to be functions of the congruity of these items, the degree of polarization of the atti tudes, and the degree of incredulity regarding incongruous messages (34). Attitude.--An attitude is a psychological con struct, or latent variable, inferred from observable responses to stimuli, which is assumed to mediate consistency and covariation among these responses. Attitudes may also be inferred from expressive or symbolic behavior in which overt choice is implied or indirectly expressed, as on questionnaires, in interviews, and in responses to projective techniques or by observation of overt behavior related to but not identical to the choices in question. (16:212) Attitudinal scales.— "Thurstone was the first to adapt the methods of psychological scaling originated by Fechner to judgments of favorableness and unfavorableness 12 towards various objects." Examples of attitudinal scales are scales by Thurstone, Likert, Green, Edwards, Remitters, and Cattell (16:213). Test.— In this paper, "test" is used to mean a relatively fixed set of items scored by a relatively fixed scoring system. Evaluation.— Evaluation is one of the three gen eral factors isolated by factor analysis. This accounts for the largest portion of the variance and is usually on the good-bad continuum. Potency.— Potency is one of the three general fac tors isolated by factor analysis. This is the "football player" factor. Large-small, strong-weak, and heavy-light are examples (34). Activity.— Activity is one of the three general factors isolated by factor analysis. Some relation to physical sharpness or abruptness is suggested. Examples are fast-slow, active-passive, and calm-excitable. Technique.— Technique is used to mean an approach to the measurement of some attributes— an approach that has to be modified to fit the particular subject matter under investigation. The semantic differential is a technique rather than an instrument. 13 Limitations One important limitation of this study is the lack of complete randomization which is inevitable if classroom groups are selected by subjective means. Neither students nor classes were chosen by consulting a table of random numbers so as to select subjects or groups from the entire parameter of tenth grade students in a large school sys tem. However, the selection was made by knowledgeable school administrators and reflects their desire to meet the requirements for randomization. The composition of class groups was the result of three methods of program ming students--computer selection, self-selection or "running for classes," and counselor selection by "pre programming." One concern of this experimenter was to disrupt normal school routines as little as possible. Be cause of this, students were tested in class by the regular class teacher, and administration was standard only if each teacher followed the instructions exactly. That subjects were lost is due to the same concern for the school; the giving of make-up tests would require many classroom inter ruptions. The other limitation of this study is the lack of ultimate criteria. Whether Krumboltz (23) would accept an intermediate criterion instead of observable behavior is 14 doubtful. Later researchers, especially those trained in his methodologies, may be able to evaluate the success of a course in guidance more objectively. The variables associated with "teacher" are not controlled except that teacher's orientation is dichoto mized into full or part-time counselor versus "not a counselor." Therefore, the personality of the teacher, his age, and his length of experience are uncontrolled variables. However, it is unlikely that a beginning teacher would be assigned a guidance class. It is not possible to control all the variables associated with teaching methods (lectures, demonstrations, field trips, audio-visual techniques, and group participa tion) and teaching materials. It is assumed that the counselor-taught-guidance classes would follow group coun seling techniques more often than the classes taught by regular subject matter teachers. The limitations of the instrument are its validity and reliability. The validity of the concepts selected for the study are subject to close scrutiny. Those not appropriate will probably attract only minimal responses from students. The reliability of any test rests upon the assumption that all the responses are honest and carefully made. This is especially true when the assessment instru ment is an indirect measure such as this semantic differ- 15 ential test. Basic Assumptions The following assumptions were made for this study: 1. The method of sampling minimizes such vari ables as teacher, students' previous marks, time of day. 2. The teachers are familiar with the course of study as outlined. 3. The students' responses are honest. 4. The instruments are marked properly. 5. The concepts selected by this version of the Webb-Harris test are understood by the stu dents. 6. The students in the school district tested are representative of the larger parameter of tenth grade students. a. The three selected schools reflect the various socio-economic levels. b. Girls and boys are in each class. c. Intelligence of students in the sample is representative of the larger parameter. 7. The attitudinal changes reported are primarily a result of the learning associated with the guidance class. 8. The assumptions for the tests of significance are met. 9. Equal intervals between scales are assumed. 10. Zero point falls at the center of each scale. 11. The concepts are appropriate. 12. The six-point scale was adequate. 13. Attitudinal change precedes behavioral change. 14. Test, directions were adequate. 15. Each student was able to read the test direc tions and test items. Summary This study is an attempt to evaluate the effect of a ten-week required guidance class upon the participating students by means of an analysis of responses made on a seventeen-concept semantic differential test instrument. A significant change of response toward the positive ad jectives is to be regarded as evidence that the individual has changed his attitudes toward school in a positive di rection. An analysis of group responses will be made by looking at the variables of teacher orientation/ school, sex, and intelligence. Organization of the Remaining Chapters In Chapter II of the study a review of the 17 literature on group guidance instruction will be presented in some detail, but only a brief review of the literature on the semantic differential technique will follow, as the rationale for its use in counseling has been explained quite thoroughly by Benoit (3) and others. The instrument designed for this investigation will be discussed in Chapter III along with the details of procedure employed in the collection and treatment of the data. Chapter IV presents an analysis of the important findings concerning change of meaning or attitude as well as an estimate of the ability of the Seim Attitude and Vocational Interest Inventory to rank order one's voca tional interests in the same way as is done by the well- known Kuder Preference Record-Vocational. A summary of the study, conclusions, and recom mendations, including recommendations for further study, will constitute the substance of Chapter V. I 1 i CHAPTER II ! i REVIEW OF THE LITERATURE j ] Studies of Group Guidance Techniques Until recently much of the literature regarding group guidance in occupations, freshman orientation, jun ior problems, or vocational civics had little to do with experimental studies. For example, three of the four studies most related to the group guidance class in a cer tain large metropolitan school district were concerned primarily with course content, and the conclusions drawn seemed to be based upon feelings and observation rather than objective data (15, 25, 37). A study by Coogan (9) did strive for objectivity through the use of a question naire. The move toward objectivity in research on group guidance was noted by Sinick (42). In the eighth article of a series on research on the teaching of occupations, he noted that 42 per cent of the articles surveyed were based upon experimental studies (42). Less than one-half of these were done at the high school level (6). Hoppock (17) somewhat cynically asserted that most of what had 18 19 been done in group guidance had been done only because it seemed like a good idea. Hoyt’s evaluation of group and individual programs in vocational guidance led him to believe that counseling in groups was one satisfactory solution to the counselor's concern with ever increasing student populations (18:26). He maintained that group counseling led to more individual counseling and to just as realistic vocational choices by students as did individual counseling. High school boys who had a rewarding group experience grew in self-awareness, increased their use of counseling services, recognized their need for help in career planning, made greater use of occupational literature, and augmented their understand ings of the occupation they had tentatively chosen. In one experimental group involving thirty boys, the realism of their vocational choices increased from a mean of 34.6 per cent to 57.7 per cent (18). Individual counseling pro duced the same percentage, but no significant improvement was recorded by the control group which had been exposed to neither group nor individual counseling. An earlier study by Billings (4) based upon a test, a questionnaire, and an index of congruence revealed that the resourcefulness and versatility of the teacher were more important than any one method of teaching occupations in a group setting. She suggested that the success of 20 group experience depended upon the creation of an informal democratic atmosphere, a problem-solving approach, and a variety of classroom experiences involving problems and projects (4). Billings (also known as Mildred Lincoln) worked with Dr. Brewer of Harvard on the test she used and also on a manual called How To Teach Occupations. Her well- known study revealed that when the pupil's counselor is also the group leader, the interview with the counselee can start with a background of mutual understandings and build upon these. She contended it was wasteful to cover the general information with every student counselee. The answer to Stone's query, "Are vocational courses worth their salt?" was that students who received instruction in vocational orientation in addition to indi vidual counseling showed marked gains in the proportion of optimal vocational choices made by the end of the year (44). Stone’s oft-quoted study was designed to discover if courses in occupations served as a substitute for in dividual counseling and if the desired ends of the stu dent personnel program could be achieved by referring to the counselor only those pupils considered to be problem cases. In this way the majority of students would be helped only through the medium of group guidance classes. The test of the group program was the judged realism of 21 vocational choices made by students at the end of the course.- Stone discovered that the experimental groups lowered their aspirations so that they were closer to their fathers' jobs and that the control groups held even more unrealistic goals at the end of the year. He sup ported the idea that group counseling motivates students to seek individual counseling interviews. Super (46:496) contended that the timing of infor mation giving is very crucial. He found that junior high school boys were generally unready for and uninterested in a course on occupations. Instead of presenting factual information, much of the instruction in educational and vocational opportunities should be provided in connection with other aspects of the school such as extra-curricular activities. If the timing is bad, the facts will not be acceptable. So the proper starting place, he suggested, would be in the area of attitudinal orientation. Super saw the value of group guidance in the helping of indi viduals with social values, but he doubted that personal ones could be changed. Group guidance offers a non threatening opportunity for counseling to the person who would never talk to the counselor personally. Non directive group therapy, with its permissive atmosphere, Super saw as appropriate for high school. He suggested: 22 A good program of group guidance services should forestall the development of some problems, should assist in the growth of better integrated personalities, improve social and occupational orientation, and finally should render people bet ter able to make effective use of personal and vo cational counseling when and as they are needed. (46:510) This philosophical talk is permitted by someone with Super’s respect for research. Rosengarten, in an article, "Another Job Course Pays Off," an obvious reference to a previous citation, gave rather flamboyant information regarding an experiment conducted with high school seniors. Rosengarten said: High school graduates who participated in an experimental program of vocational guidance activ ities during their senior year surpassed a com parable group of their former classmates in aver age earnings. During the same period, the experi mental groups were employed for greater lengths of time, were more satisfied with their jobs, and achieved higher merit ratings from their employers. (39:531) The results were from a follow-up questionnaire which had more than a 97 per cent return. Rosengarten asked: Were these results worth the cost in time, effort, and money? The fifty students in the experimental groups earned a total of $274,029 during the four years of the experiment compared to the fifty-two students in the control groups who earned a total of $250,037.20 during the same time. . .The ex perimental groups earned $23,991.80 more than the control groups. The program cost $6,273.75. The difference in these two figures suggests the pro gram was well worth the cost. (39:531) A successful college orientation program was re ported by Richardson and Borow (38) from studies done at 23 the University of Minnesota, where it was shown that stu dents who received group orientation to educational-voca tional counseling adopted a more effective and realistic role in counseling than did students without such prepara tion. They reminded us that success in counseling, how ever, is contingent not alone upon the competence and methodology of the counselor, but upon the client*s set toward and understanding of the counseling process as well. Test interpretation was the focus for a comparison of group and individual counseling effectiveness by Wright (52). Even if this group method proved successful, Ohlsen (28) questioned the break in confidence which is inevi table when students are coerced into revealing their test scores. Wright emphatically declared that group processes are not intended to supplant but, rather, are meant to supplement individual counseling. The effect of the group leader's reinforcement of desirable actions is one of the many topics currently be ing researched and reported by Krumboltz (23) and associ ates. Kagan (21), in his review, cited a need for more research on the counselor's role in client change. The group process has been shown to be a superior counseling method when information retained was the cri terion. Kagan's recent review of group procedures 24 mentioned that the common grade point criterion may be useful only after a three-semester waiting period follow ing experimental treatment. Kagan questioned whether the typical criterion of group effectiveness is not a bit un realistic. Possibly, the creation of insight may be a suf ficient reward, but this is hardly quantifiable. The ef fectiveness of group procedures may depend upon what Gribbons (14) called readiness for vocational planning. His study, which was built upon Super's longitudinal studies of boys, revealed that it may be only the college preparatory group which is ready for intensive vocational planning and that this readiness is discernible by the eighth grade. How to plan for the average non-committed or general student then becomes a problem of great moment. Willey (51) noted that the emphasis in general courses has shifted from detailed vocational information to wider areas dealing with life adjustment problems— to activities in which the members of the group are led to a self-inventory, a self-analysis, and to self-planning for the future. He saw that the curriculum pattern of group guidance generally considers the teacher to be the pupil's counselor as well. The modern guidance classes may stress decision-making skills rather than the making of a single vocational decision. Willey's experimental group, which was given significant local data, was superior to the control group in knowledge about the decision-making proc ess, in awareness of high school and college alternatives, and in knowledge of the probabilities involved in these alternatives. He suggested that course content must be planned with the interests and abilities of specific stu dents in mind. Baer and Roeber (2) reminded us that this general approach begins in the elementary school and em phasizes healthy attitudes toward understanding one's self and one's occupational and educational roles. This general approach is the one which is stressed in the guidance class. Decisions expected of students may be no more in volved than the rejection or continuation of one's high school major field. Summary of the Relevant Literature on Group Guidance Techniques No single experimental investigation has been re ported which is directly related to this study of tenth grade guidance in a large metropolitan school district. Four masters' projects on the course content for a tenth grade guidance class emphasized the need for such a class but failed to show the value of previous experimental classes conducted in separate schools in this school system. The results of the major studies on group instruc tional practices done by Billings (4), Stone (44), Hoppock 26 (17), and Gribbons (14) were generally favorable to a group approach. These researchers reported that such an approach created a readiness and a proper set for indi vidual counseling, provided an efficient way to deal with test and occupational information, and resulted in a high level of retention on information tests. Surveys of the practices of teaching courses in occupations were made by Sinick (42), Hoyt (18), and Kagan (21) . They indicated approval of the higher level of experimental studies now conducted in the field. Ex perimental studies now exceed position papers and philo sophical statements regarding group experiments in edu cation (21) . Not all of the groups in the reported ex periments are class size, however. Some deal with groups of only five members. If the amount of genuine student involvement is relative to group size, the class size group may be expected to be only partially effective in changing the attitudes of members of large groups. Most authorities assert that only a counselor skilled in group procedures would seem to be able to overcome normal stu dent resistance to a class which lacked the prestige of college-required courses but which might require much assigned work. The demise of many courses on occupations suggests that the effective course must meet the needs of a very heterogeneous student population. The hope for the 27 classroom approach, educational and vocational, seems to be directly related to availability of individual coun seling (18) . The unique nature of a class in educational and vocational guidance keeps it from being compared too readily with the small group experiments which study as few as five students (but which, nevertheless, abound in the current literature), or with traditional courses in occupations, or even with courses in human relations which may be oriented toward mental health. The tenth grade guidance class evaluated in this study is designed to be more closely related to the regular instructional program in other academic classes than to therapy, and, for this reason, may fail to yield the significant changes re ported elsewhere. Studies of the Semantic Differential Technique The so-called semantic differential test is in reality a misnomer. Osgood (29:197) declared that the semantic differential is a technique employed by psycho linguists and other researchers to get at the meaning of concepts and other stimulus words or ideas held by indi viduals and that it is not necessarily a test. His Uni versity of Illinois research staff has produced one author itative book. The Measurement of Meaning (34), and dozens 28 of journal articles and unpublished manuscripts relating to the semantic differential technique (29, 30, 31, 32, 33, 34, 35). A description of the semantic differential in strument used in this work is to be found in Chapter III. Whether the semantic differential technique is ap propriate to measure attitude or meaning change after a group counseling experience of ten weeks is debatable. Webb (49:260) reported a study of NDEA counselors using the Webb-Harris Word Meaning Test, a test using the same bi-polar adjectives as the instrument designed for this study. Adult counselors made significant attitudinal changes on certain concepts following a six-week NDEA sum mer institute. Webb and Harris created their test to assess changes in point of view as a result of the insti tute. They used Osgood's D (a generalized distance for mula) to measure changes in meaning. The distance for each concept was calculated from the basal concept "my actual self." Webb discovered, as have others, that the evalua tive dimension may allow more significant findings to be reported than would follow if raw score data were averaged over all three dimensions/ The NDEA group was not tested again; therefore the stability of the change cannot be known. Fewer than thirty subjects comprised his sample. In another study on the effect of class size upon students, Stempel (43), using the semantic differential technique with a sample of 225 college seniors, attempted to discover their attitudes toward the lower cost subjects. No support for higher cost subjects (small classes) was given. The semantic differential was chosen because it gave subjects a wider range of expression than usually was posed by attitude questionnaires. One might question that the concept "instruction in English at Central” (or phys ical education or political science or speech) might be a stimulus that related to intrinsic elements of the four subjects mentioned and so have little bearing on the class size issue. In the relatively new Journal of Verbal Learning and Verbal Behavior, Ross (41), in an effort to discredit Osgood’s hypothesis that a concept provokes a response which is independent of context, designed a semantic dif ferential with supposed context contamination. The con tamination was supposed to result from a cursory examina tion of all concepts before responding to the test. In Ross' study, sixty students on one side of a room were given one concept set to rate; sixty students on the op posite side of the room were given different concepts to rate. When the results were analyzed, his hypothesis re garding context contamination was not supported, and we are led to accept Osgood's idea that context has a very limited effect on one's responses to various concepts. 30 The need for rigid instructions for the administra tion of a semantic differential was investigated by Miron (26). Miron's findings lead us to believe that the influ ence of instruction modification upon the test-retest re liabilities of the semantic differential is negligible. Four groups, designated as fast-memory group, slow-memory group, fast-no-memory group and slow-no-memory group, took a twenty concept by twenty scale semantic differential twice in a single sitting within a two-hour period. One hundred and twelve college students participated in this reliability study which confirmed the wisdom of Osgood's instructions to students to work rapidly. The fast groups had scores with fewer mean deviations between administra tions and displayed better recall of previous answers. The effect of greater time lapses between administrations was not tested. The concept "school teachers," which is also used in this specific study, was the focus for an experiment by Husek and Wittrock (20:209). They were interested in the dimensionality of the attitudes about teachers held by 259 education students. One hundred and seventeen scales were used and subjected to factor analysis. In line with Osgood's earlier work, the most important factor was the evaluative one which accounted for 24 per cent of the total variance and 39 per cent of the common variance. These 31 education students reported a very favorable mean score of 2.19 (for the concept "school teachers") on a seven-point scale where "1" is the most positive position possible. The ability to generalize from this study using only one concept is limited, but we do have a reference point as to what is a very favorable attitude toward teachers. Again, the suitability of the semantic differential technique in education seems to be confirmed. That the semantic differential technique is of limited use in schools because of social desirability vari ables was questioned by Ford and Meisels (13). Much re search has been done on these variables and with the eval uative factor of the semantic differential described by Osgood (34) . The correspondence between the two was in vestigated by defining the social desirability value of a given bi-polar scale as the discrepancy between the mean social desirability ratings of its separate adjectives. The indices of evaluativeness were the evaluative factor loadings and dimension coordinates of the scales. The fifty bi-polar scales of Osgood (34) were chosen. Two sets of judges produced their social desirability values. Results indicated that the evaluative factor loadings and dimension coordinates are predictable from the social de sirability scale values. The same representational medi- ational process which is hypothesized to underlie semantic differential judgments then is basic to the responses a i j subject makes to questionnaires and other personality assessment devices so common in education. Pioneering work on attitudes was done by Thurstone (47) and Likert (24). Osgood's (34) semantic differential | t technique simply gives us another way to measure connota- | tive meaning. In 1952 he noted that no standardized methodj existed for the measurement of meaning (29:197-237). The theoretical basis explained in that early article rested upon the belief that words (often stimulus words chosen as concepts) are emitted rather than elicited behavior. i i Osgood remains the most important researcher in the use of the semantic differential technique. His asso ciates, Luria (32), Suci (30, 33), Tannenbaum (34), and others have continued to report findings using this tech nique . Summary of the Relevant Literature on the Semantic Differential Technique Charles Osgood has dominated the theoretical and experimental work done with the semantic differential technique in quantifying the meaning of words. The cau tious words of Kerlinger (22) and others reminded us that this is only another way to discover what meaning people assign to various concepts. The field of education, al ways the object of internal and external criticism, has been able to use this relatively new technique in learning more about the students, the teachers, and the curricula of the schools. Many of the studies reported in the liter ature use the familiar captive audience found in courses in psychology and educational psychology. A few studies reveal that elementary and secondary students can make appropriate responses to the stimulus words presented. Well-designed studies continue to be reported. Knowledge of the ease of scoring may result in even wider appli cation of the semantic differential technique by research ers such as school counselors, school psychologists, and directors of guidance. A sample of the statistical treatment employed in recent semantic differential studies is outlined in Table 1 to demonstrate the versatility of the semantic differ ential technique as well as indicate the level of measure ment used to test the hypotheses generated by research. 34 TABLE 1 A TABULAR REPRESENTATION OF THE STATISTICAL TREATMENT EMPLOYED IN RECENT SEMANTIC DIFFERENTIAL STUDIES Author Field of Investigation Statistical Treatment Benoit (3) Ford and Moiseis (13) Hunt (19) Huaek and Wittrock (20) Miron (26) Nelson (27) Osgood (29) Osgood and Suci (30) NDEA Institute— change in attitude of counselor- trainees "Social Desirability and the Semantic Differen tial" Measurement of self-con cept in relation to vo cational choice using Kuder (Oil) Attitudes towards teachers Effect of instruction modification on test- reteat reliabilities of the semantic differential Manual for administration of the semantic differ ential Estimate of factors using fifty descriptive scales "A Measure of Relation Determined by Both Mean Difference and Profile Information" t test between various factor score means (pre post) by groups, within groups Product moment correlations 0 score modified to C score ^Qij-D), °D Cij . ( i M (-2)) + 10 Principal components factor analysis of an 80 x 80 matrix of intercorrelations of means and standard devi ations Pearson product-moment correlation coefficients; analysis of variance re lated to speed, recall, and concepts Factor analysis Factor analysis; product- moment correlations D statistic 0 (X^ and F test suggested or D method of factoring) 35 TABLE 1— Continued Author Field of Investigation Statistical Treatment Perez (36) Rosenthal (40) Ross (41) Stempel (43) A measure of change in attitude using the Webb- Harris test "A Semantic Differential Investigation of Critical Factors Related to Achieve ment and Underachievement of High School Students*1 "Change in the Use of the Semantic Differential with a Change in Context" "The Relationship of Cost of Instruction and Atti tudes Toward Instruction" Signed-Ranks test Analysis of variance (F test) using averaged factor scores; 0 statistic Factor analysis Analysis of variance Strem (45) An analysis of drop-outs and stay-ins Analysis of covariance; F test Webb and "A Semantic Differential Harris (49) Study of Counselors in an NDEA Institute" D statistic; Wilcoxson's Paired Replicates Test CHAPTER III SOURCE OF DATA AND METHODOLOGY In this chapter the student sample, the treatment variable, and the techniques for collecting and analyzing the data are described. This information has been organ ized into the following categories: (1) The Sample Popu lation, (2) The Treatment Variable— A Guidance Class, (3) The Method of Measurement— The Semantic Differential Tech nique, (4) The Test Instrument, (5) The Collection of the Data, (6) The Problem of Criterion, (7) Analysis of the Data, and (8) The Chapter Summary. The Sample Population To discover the meanings assigned to various school-related, work-related, and self-related concepts, a semantic differential instrument was administered to twelve classes of tenth grade guidance in three public high schools. The control groups were three driver edu cation classes which switched with three guidance classes at the end of their ten-week cycle. One control group 36 37 was chosen for each school. Selection of schools to re flect differing socio-economic levels and ability levels was made by the Committee on Research Studies in the par ticipating school district. Randomization of schools was assumed in its selection. Pupils were tested in intact classes. No attempt was made to test absent pupils, and there was, therefore, an inevitable loss of subjects due to a student's absence on either of the days of testing or a pupil's transfer to another class or school. The subjects in this study, tenth grade students in guidance classes in three high schools, represented the extremes of socio-economic background as well as the aver age. The median family income for the area served by each high school is given in Table 2. The median family income for the city of Los Angeles is presented for the purpose of comparison. In an attempt to control for the time-of-day vari able, classes were tested which met during each class period with the exception of period two. The students were given the test as part of the regular class work. They did not volunteer, nor were they selected for a particular reason. Some students were assigned to their classes by grade advisors or counselors. Classes may have been grouped as far as college-bound and terminal students were concerned, but this is doubtful. 38 TABLE 2 MEDIAN FAMILY INCOME FOR AREA SURROUNDING SELECTED HIGH SCHOOLS (1960 CENSUS DATA) Income School I = $6,301 School II = $8,277 School III = $6,786 Mean = $7,121 S.D. = $ 841 (Median Family Income for City of Los Angeles = $6,896) The students in this study follow a major sequence plan which was developed in their eighth and ninth grade English classes and outlined in a workbook which became an educational guidance record. The junior high school coun selor, the parents, and the senior high school counselor were also involved in the establishment of the senior high school major sequence adopted by each tenth grade student before he entered high school. The tenth grade student is generally about fifteen years old, but he may be as young as fourteen or as old as seventeen. Most tenth graders celebrate their sixteenth birthday in the tenth grade. Because school is not 39 required after the pupil is sixteen years of age, the placement of the guidance class is most strategic. The experimental group was made up of 115 girls and 108 boys, and the control group consisted of 34 girls i j and 40 boys. There were 83 students participating from j School I, 120 from School II, and 94 from School III. The 256 BIO students make up the majority of the 297 pupils included in the investigation. The sample is described in Table 3. This table gives a description of the total sample, class by class. The mental ability of each student in each class is reported in a stanine score. Table 4 provides a brief explanation of the stanine system. The grade level and sex of each pupil are indicated. Guid ance and driver education classes are labeled as experi mental or control groups. The role of the teacher is separated into three categories— teacher of driver educa tion, teacher of guidance class only, or teacher/counselor. The Treatment Variable— The Guidance Class The guidance class meets for ten weeks and alter nates with another ten-week class in driver education— a state-mandated class in automobile safety. Students were removed from one of the original guidance classes for ad ministrative reasons. Loss of samples was also the result of absence and transfer from the school. A description of □1 02 03 □4 05 06 07 08 09 10 11 12 TABLE 3 DESCRIPTION OF SAMPLE School Ability Level (Stanines) Period of Day Grade Level Experimental Control Sex Instructor x> M u e o O —1 M c >> < 0 m e h m > TJ c c • r t -H O 3 M 3 O a cd ( - i 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 □ m a H < H H CD r ~ l H < CM H CO IM H <C a >i o m n iH U • r l 13 H e +> £ I 0 0 1 6 10 1 5 1 0 X 18 6 X 14 10 24 X I 0 2 3 6 3 4 1 0 0 X 12 4 l 0 1 1 X 10 9 19 X I 0 1 1 6 9 3 1 4 0 X 19 4 0 l 1 0 X 12 13 25 X I 0 1 6 3 5 0 0 0 0 X 14 1 X 6 9 15 X Sub-Total 0 4 11 21 27 8 7 5 0 63 15 i l 2 1 42 41 83 II 0 0 4 1 4 7 6 4 3 X 27 0 l 0 1 0 X 13 16 29 X II 0 0 4 3 4 3 8 4 3 X 28 i X 14 15 29 X II 0 0 1 0 10 9 9 1 1 X 29 2 X 15 16 31 X II 0 0 1 4 7 6 9 2 2 X 31 X 13 18 31 X Sub-Total 0 0 10 8 25 25 32 11 9 115 2 2 0 1 ~0 55 65 120 III 1 4 0 7 5 4 0 0 0 X 19 2 X 13 8 21 X III 1 2 3 6 6 1 2 0 0 X 19 1 1 X 12 9 21 X III 2 3 1 5 7 4 2 1 0 X 21 4 X 12 13 25 X III 1 3 1 9 4 4 3 2 0 X 19 7 1 X 14 13 27 X Sub-Total ~s 12 ~ 27 22 13 "7 ~3 "o 78 14 2 0 0 ~0 51 43 94 o TOTAL 5 16 26 56 74 46 46 19 9 3 0 4 1 1 3 256 31 5 1 3 1 9 3 148 149 297 3 7 2 Guidance Only TABLE 4 STANINE SCORE SYSTEM Stanine Scale C.T.M.M. I.Q. Range Per Cent in Each Interval (Rounded) Cumulative Percentages Number Included in This Sample 9 129+ 4 100 9 8 121-128 7 96 19 7 113-120 12 89 46 6 105-112 17 77 46 5 97-104 20 60 71 4 89-96 17 40 56 3 81-88 12 23 26 2 73-80 7 11 16 1 to 72 4 4 5 N - 297 ---------- v f c . f —1 42 the samples lost is included in the appendix. Students were assigned to these classes in a random fashion. Most students in the three high schools can expect to take driver education and guidance in their first semester of high school. The guidance class is a ten-week class organized around units dealing with the preparing for educational opportunities, the understanding of vocational opportuni ties, and the planning ahead for a changing world. In the first unit the student becomes acquainted with the educational program in his high school. This would include curricular and extra-curricular offerings. He would gradually learn to realize that decisions need to be made in the choosing of subject majors, electives, and other curricular presentations. The decision-making proc ess is usually discussed in this context. As the student learns to make judgments, he is expected to utilize infor mation gleaned from past performance in test experiences as well as in class work. Subject marks and test scores are the obvious focus of attention here. The pupil be gins to gain an awareness of the use of objective data in the prediction of his success in future educational and vocational experiences. It is at this time in the course that the student may expect to be administered such stand ardized tests as the Differential Aptitude Test and the 43 Ruder Preference Record-Vocational (7). The vast range of post-high school educational opportunities may then be explored by the pupils on the basis of an increased under standing of interests and aptitudes. The teacher's skill in presenting this information on testing and self-analysis is certainly a variable dif ficult or impossible to control. One can only assume that the enthusiastic teacher will be the most successful in changing attitudes toward school and the world of work. It may not be safe to say that the guidance teacher who has expressed his strong interest in the specialty called 1 1 counseling" by taking professional courses in pupil per sonnel work will be more successful in creating these positive attitudes than is the more subject-matter ori ented teacher. Two of the nine experimental classes were taught by a teacher who was also the grade counselor for many of the pupils in his classes. A second phase of this somewhat flexible curricu lum applies the learning experiences of the first unit toward understanding the vocational opportunities which are present and which will emerge in the future. The basic tools used in the classroom may be college catalogs, college handbooks, listings of job opportunities, and de scriptions of specific jobs. Unless the teacher provides an adequate preparation for the use of these tools, only 44 a cursory examination of them can be expected from the pupils. The third general area of concern in the guidance class is a culminating activity where tentative plans for the future are made. Readjustments of one's school pro gram are scheduled so that reasonable expectations may be realized. This means attitudinal adjustments. A workbook experience is ordinarily provided the student so as to objectify the matching of his personal qualifications with future schooling and the world of work. Classroom management by the teacher may provide for such diverse experiences as lectures by guest speak ers, panel discussions, group discussions, buzz sessions, debates, sociodramas, report giving, role playing, test taking, and test interpretation as well as the more tra ditional high school experiences of note taking, filling out application blanks and questionnaires, teacher-led question and answer periods, writing reports, reading texts, magazines, and other supplementary reading mate rials, viewing films, or listening to tape recorded inter views or discussions. The success of the class may very well depend upon the amount of pupil participation. It would be assumed that a counselor committed to a student-centered approach to counseling would follow a plan whereby the pupil is an 45 active learner interested in and responsible for his own decisions. In an oblique way, this investigation will test this assumption, although the sample of teachers rep resents but a small sampling of the parameter of teachers of guidance. The guidance class is generally taught in a room which has been equipped with vocational files, catalogs, and supplementary materials on occupations and occupa tional classification. In a short time the guidance teacher becomes somewhat expert in locating vocational information. Information concerning educational opportun ities is also to be found in this resource room. The guidance class is often a joint responsibility of the counseling staff and the social studies department in the high school. The guidance instructor is likely to be a part of the pupil personnel team or the social stud ies staff. The driver education class, which alternates with the guidance class, is taught by a specialist who has pre pared himself in the area of traffic safety. Passing the driver education class is one of the state requirements for a high school diploma. The class is designed to pre pare high school students to pass a test on the Vehicle Code of California should they desire an operator's li cense. The class provides preliminary instruction for 46 behind-the-wheel training which is taught by a special staff of teachers not assigned to any single high school. The driver education class is held in a self-contained classroom and does not involve any actual driving. Be cause driver education is a required course, there may be students in every grade level in the guidance class who are new to the school and are programmed to the two ten- week classes. This accounts for the twelfth graders in the sample. The emphasis in these two subjects is quite dif ferent in that the driver education teacher instructs a class which must be passed, while the guidance class in structor must persuade the students that his class is as valuable as one which serves as a prerequisite for a di ploma or a more advanced class. The successful guidance class teacher will probably capitalize on pupil perceptions and pupil participation rather than on his personal accumu lated knowledge. Because the stated purpose of the class is a change in attitude, the teacher must orient his think ing away from information giving. The emotional climate of the room should reflect the philosophy of a teacher who is emotionally mature enough to accept pupils as they are. This is no small order. The teacher of the guidance class, then, should produce a blend of the techniques of group dynamics and' traditional didactic classroom methods in 47 order to fulfill the objectives of the course. The Method of Measurement— The Semantic Differential Technique Semantic differential techniques have been used to measure meaning in many settings such as hospitals, clin ics, industries, and schools at all levels. Although some people in psychology often refer to the semantic differ ential as if it were a test with a definite set of items and a specific score, such is not the case (34). It is safer to call it a technique; in fact, it is a general way of getting at a certain type of information. Each time the semantic differential is used, the experimenter adapts the concepts and scales to meet his own needs. Standardi zation comes not through standard concepts and scales but through the allocation of concepts to a common semantic space defined by a common set of general factors. The concept acts as a "stimulus," and the subject's checking is a terminal "response." Generally, the con cepts are printed words but may be spoken words or even elements of projective tests. The experimenter has to sample from the universe of possible concepts which are germane to his research. The semantic differential often takes two forms as illustrated in Figure 1. 48 FORM I (concept) LADY ME (polar adjective) rough fair (scales) (polar adjective) smooth unfair (polar adjective) rough _ fair _ active FORM II (concept) LADY (scales) : X (polar adjective) smooth X unfair passive (Items in parentheses do not appear on forms given to subjects. X=subject's mark along scale.) Fig. 1.— Alternate Forms of the Semantic Differential After the forms have been marked, the experimenter possesses data which may be plotted on a rectangular solid with concepts forming the length, subjects forming the width, and scales forming the height. A response is usu ally recorded as a zero when the middle space is marked or as a plus or minus 1, 2, or 3 for the other positions on a seven-step scale. A matrix is formed by the number of scales times the number of concepts times the number of subjects. Each cell represents the judgment of a 49 particular concept against a particular scale by a partic ular subject. The significance tests usually applied to the semantic differential are product-moment correlation (r), analysis of variance (F), and factor analysis (u). Both non-parametric and parametric tests of significance have been used. The Test Instrument The instrument used in this study was the Seim Attitude and Vocational Interest Inventory, to be known as the SAVII throughout the remainder of this paper. This inventory follows the semantic differential technique de veloped by Charles Osgood and associates at the University of Illinois. The inventory consists of seventeen "con cept" cards which are tabulating cards. All seventeen cards are identical except for the concept at the top. The concepts in order are: "school," "discovering new facts and solving problems," "working with machines and tools," "boys at this school," "going to concerts, playing instruments, singing," "me," "doing creative work with your hands," "work," "working with numbers," "meeting and dealing with people to promote projects or things to sell," "helping people," "working outdoors," "girls at this school," "reading and writing," "teachers," "what I would — — — -— — ” — —------------- 1 50 | I ! i like to be," and "doing office work with precision and j | accuracy." All cards have twelve bi-polar adjectives with six marking spaces or scales separating them. The bi-polar adjectives are kind-cruel, hard-soft, passive-active, strong-weak, calm-excitable, true-false, light-heavy, j wise-foolish, moving-still, ugly-beautiful, small-large, | and fast-slow. The six-scaled test, unlike the seven- scaled test, does not leave a neutral or uncommitted re sponse. Each response is a decision, even though the middle positions are considered to be weak in intensity. The intensity is measured by the distance from the adjec tive which more closely indicates how the subject feels about the concept at the top of the card— the closer the mark, the more intense the feeling. The cards were marked with an electrographic pen- cil which was provided with the classroom packet of materials. This particular instrument was developed from the Webb-Harris test used in a clinical setting (50). The scales are the same, but the concepts are different. The seventeen concepts represented three clusters of related concepts. The "me" card and the "what I would like to be" card are self-concept cards. The school- related cards are "school," "boys at this school," "girls at this school, 1 1 and "teachers." The work-related concepts 51 are "discovering new facts and solving problems," "working I with machines and tools," "going to concerts, playing in struments, singing," "doing creative work with your hands," i "work," "working with numbers," "meeting and dealing with j people to promote projects or things to sell, " "helping | people," "working outdoors," "reading and writing," and "doing office work with precision and accuracy." j The SAVII was designed to be used with the Kuder Preference Record-Vocational, a test generally used in the guidance classes within the participating school district. No endorsement of this popular test is intended. The con- i cern of this investigator was simply to see if the SAVII, ! a semantic differential instrument, was able to rank one's vocational preferences so as to achieve a significant cor relation with the well-known Kuder test. The Kuder inven tory gives raw scores (convertible to percentiles) for vocational interests which are based upon the subject's responses to a triad of career-related experiences. The Kuder test, as well as the SAVII, makes the subject re spond to a forced-choice situation. The pilot study of the SAVII was done in seven junior high schools with the so-called "culturally disad vantaged" children who participated in an excursion and counseling effort designed to expand the horizons of the pupils as well as provide them with a weekly counseling 52 experience. Over 250 protocols were completed in pre-test and post-test administrations twelve weeks apart. No statistical analysis of the data is available at this time, but an inspection of the tests revealed that the directions to test administrators and pupils were adequate, although a shortage of electrographic pencils necessitated the re marking of some cards. The forced-choice nature of the test did not create a dilemma; however, some concepts may have elicited only a mild response. The bi-polar adjec tive scale light-heavy seemed to be confusing when the concept "school" was rated. Shifting from noun concepts to verbal nouns did not appear to be a source of diffi culty. The average time necessary for completing the seventeen-card test was thirty minutes. This investi gator administered the test twice to a control group of thirty-four B8 students in an English class whose members were not unlike the experimental group in intelligence, aspirations, or achievement even though they were younger. Hand scoring of some of the protocols seemed to suggest that there might be differences in the treatment groups which were related to the degree of counseling skill pos sessed by the group counselor. In short, the teacher variable seemed to be vital. The data in statistical form are unavailable, however, at this time. The pilot study, for all practical matters, 53 produced only ideas that could be incorporated in the future administrations of the test instrument. The adaptation of the Webb-Harris Word Meaning Test was made possible through the permission of Dr. Allen Webb, a co-author of the test, who is a clinical psychologist in Pasadena, California. Some use of the SAVII test instrument was made re cently by the research and development staff of the afore mentioned school district. For the district's use, the instrument was titled the Seim Semantic Differential, al though it was essentially the same as the SAVII. No re sults from these investigations have been made available. The use of the instrument was directed toward discovering a change of attitude following a treatment administered as a part of a federally-funded project. In short, over three hundred students in the age range thirteen to nineteen were administered the SAVII test before and after an experimental treatment variable which consisted of counseling experiences, planned excur sions, remediation measures, and information-gathering opportunities. The directions appeared to be adequate. The Collection of the Data The tests for each school were delivered to the head counselor prior to the first administration which was scheduled for the earliest possible date in the semester— during the first week of September, 1966. The tests were packaged in large manila envelopes. The "Instructions for Administration" sheet was inserted in each large envelope. (A sample of this instruction sheet is included in the ap pendix.) The forty business envelopes in each large en velope were marked, "Use Electrographic Pencil." Envelopes were distributed to the classroom teacher by the head coun selor in each school. The counselor in each school was asked to test three guidance classes and one driver edu cation class. At least one guidance class was to be selec ted from those which met in the early morning, one from the late morning, and one from the afternoon. The completed tests were sealed, replaced in the large envelope, re turned to the head counselor, and later were collected by this investigator. The desire to disturb the schools as little as possible prevented the testing of absentees or new enrollees. Unfortunately, it was impossible to meet with the teachers prior to testing as the teachers did not return to school until the first day of instruction of the fall semester. The second administration was completed at the end of the ten-week cycle. This testing took place in Novem ber, 1966. Each test protocol was examined for complete ness and was prepared for reading by a mark-sensing 55 machine. The names were placed on a master alphabetical list and an identification number was assigned. This identification number was placed on the reverse side of each card. A master roll of subjects listed each pupil by- identification number. The three schools were arranged in alphabetical order and were identified as Schools I, II, and III as in Table 3. The four classes in each school were numbered in order of time of day. The cumulative record of each student in the sam ple was searched for the sex of each student, his grade level, a composite IQ designated by a single stanine score, and a rank order of items on the Kuder Preference- Vocational. This information was recorded on the sheet labeled "Demographic Data", - a sample of this appears in the appendix. This additional information was added to the master roll in the order mentioned above. The header card for each subject was key-punched with the above data. Concept cards from the second administration were marked with the student's identification number followed by a hyphen and a "2" to designate the second administra tion. Some subjects were given a second number since it became apparent that the loss of subjects would create a discontinuity in identification numbers. Concept cards were pre-punched as "001" to "017" 56 before initial packaging through the courtesy of the staff at University of Southern California School of Business Ad ministration's computer facility. Data were processed at the Western Data Processing Center through the courtesy of International Business Machines. The computer program was processed by Robert Uzgalis at the Westwood facility. It was understood that all the variables concerned with test administration were not under optimum control, but that they were probably typical of most test adminis trations in intact classrooms where a skilled test super visor is not in charge. The semantic differential tech nique is not considered to be very sensitive to instruction modification (26). The Problem of Criterion Is it always necessary to wait for long-range, follow-up studies of subjects' school or work experience to evaluate the worth of a group guidance course? Should the criterion be an ultimate one such as satisfaction in one's work and life? Additional criteria might be the realism of one's educational and vocational decisions as judged by experts (14, 15), pupil appraisal of such a course at the end of high school (13, 37), teachers' eval uations of students' progress, objective measures such as improved report card marks in achievement, work habits, 57 and cooperation, increased use of individual counseling services or library facilities after exposure to the course, or school retention measures. Some less objective ways to ascertain the effect of such a course might be to use such indirect measures as change scores on questionnaires, examinations, Q sorts, self-reports, or semantic differential instruments. The criterion to be used in this study is change in meaning scores on the school-related concepts on a semantic differential instrument designed specifically for students in the tenth grade guidance class. It is assumed that those students whose change scores reflect more posi tive attitudes will be able to adapt satisfactorily to their new educational environment, to succeed in high school, and to make wise educational and vocational choices. There is little reason to expect that the develop ment of insight or the ability to communicate one's feel ings, both reasonable goals for a guidance class, would lend themselves to an objective analysis by one investi gating this area. Therefore, attitude measurement seems to be justified at this time. Analysis of the Data The major hypotheses of this investigation were concerned with the assessment of change in meaning following a ten-week class entitled "Guidance." To test the assumption that the test instrument used has the three dimensions of evaluation, potency, and activity, the data composed of twelve scale markings for seventeen concepts for 297 pupils over two administrations (12 x 17 x 297 x 2 bits of information) were subjected to a series of factor analyses. A number of techniques for rotation were em ployed. The IBM 360-75 was used to compute the results. Without factorial purity, it is presumptuous to talk about the "dimensions of meaning." The factor scores were com puted for each concept. The amount of the total variance, as well as the common variance, was determined for each factor. Only when the data support the idea of three factors, can the three dimensions of meaning be charted. Careful examination of the factor scores revealed that the expected separation of factors into the three discrete categories labeled "evaluative, potency, and activity dimensions" was only partially successful. Even though some of the bi-polar scales were matched with the appropriate dimension, the amount of the total variance explained by the first three factors was less than 50 per cent when the seventeen concept cards were analyzed for factorial purity. Only the first of the factors was simi lar to any of the three traditional factors associated with Osgood’s research. The so-called evaluative — — — ------— — — —— 1 59 dimension was represented by the bi-polar adjective pairs j kind-cruel, true-false, and wise-foolish. One method of hypothesis testing was the technique i of discriminant analysis which demands an interval scale j of measurement. This test indicated which scales or con- j cepts acted as discriminators to test whether certain | i classes or schools could be identified because of the way pupils in those classes reacted to various concepts. This test, which demands the assumptions associated with para metric tests, is a powerful test of significance. A test of considerably less power is the chi-square test for related samples. This test was used to assess the hypotheses associated with the direction of the change in meaning following the second administration of the test. Chi square can use ordinal data arranged in a con tingency table where direction, but not intensity, is indicated. Positive, negative, and no-change categories for change scores were assigned according to sex, school, treatment group, teacher orientation, and intelligence. In summary, the students' scale markings were analyzed by factor analysis, discriminant analysis, and chi square to test the instrument, the discriminators which verify discrete groups, and the significance of the difference between groups in terms of the direction of the scale markings after a second administration of the test. 60 Except for the research hypothesis of expected change in the class taught by a counselor, the null hypothesis was tested throughout this investigation. The correlation between the SAVII and the Kuder Preference Record-Vocational was tested by rho for the correlation of ranked data. The Chapter Summary The student population was composed of 297 stu dents in three high schools in a large metropolitan school district. The experimental groups were composed of tenth grade students in a ten-week guidance class which alter nated with a driver education class (control group) which met for a similar period. The test instrument was a modified version of the Webb-Harris Word Meaning Test which was based upon the semantic differential technique developed by Charles Osgood and associates at the University of Illinois. The test instrument was evaluated after a pilot testing program. The tests were administered to the students in class groups in September and in November, 1966. Distri bution and collection of tests were handled by the head counselor in each school. The statistical treatment of the data was designed 61 to assess the change in meaning for the seventeen concepts after the students had completed a course entitled "Tenth Grade Guidance." CHAPTER IV RESULTS OF THE INVESTIGATION Findings This investigation of a ten-week guidance class was conducted in September and November, 1966. The major hypotheses were concerned with a change in meaning or at titude as measured by a semantic differential instrument adapted for this specific purpose. The scale markings on this instrument provided the raw data which were first factor analyzed on an IBM 360-75 and then tested by the techniques of multiple discriminant analysis and chi square. The three usual factors of evaluation, activity, and potency were not apparent after repeated factor anal yses. Table 5 summarizes the results of the factor studies for each of the seventeen concepts. The one (kind-cruel), six (true-false), and eight (wise-foolish) pattern, which constituted factor one, appeared to have explained approx imately 40 per cent of the variance when concept sixteen was eliminated. Factor two was quite inconsistent over 62 Item 1 2 3 4 5 6 7 8~ ' 9 10 11 12 13 14 15 16 17 ALL TABLE 5 FACTOR ANALYSIS BY ITEM SUMMARY Per Cent of Maximum Var. with Per Cent of Three Factors Variance Factor I Factor II Factor III Factor IV 42 45 1,6,8 2,7 11 3,5 36 39 3,9,12 4,8 2,7 1,6,8 10 11 46 49 1,6,8 2,7 11,12 3,4,5 47 49 1,6,8 10 3,9,12 7,11 63 65 3,9,12 4,5 1,6,8 10 40 43 1,6,8 10 7,11 4,5 3,9,12 50 53 1,6,B 10 3,9,12 4,5 11 42 45 3,9,12 4,11 1,2,7 6,B 49 52 2,10 7,11 6,8 3,9 48 50 1,6,8 4,10 3,5,7 39 44 1, 6,8 9,12 5,7 46 49 6,8 4,9,12 11 3,5 45 47 1,6,8 3,5 7,11 46 48 6,8 10 7,11 3,5 9,12 46 49 1,6,8 4,9,12 11 47 49 12 7,11 47 50 1,6,8 3,5 2,7 11,12 1,6,8 7,11 3,9,12 j 2,10 c r > (The 1,6,8 pattern must explain about 40 per cent of the variance if Item 16 is eliminated.) 64 the seventeen concepts, although the seventh scale (light- heavy) and the eleventh scale (small-large) were most common over all- The third scale (passive-active), the ninth scale (moving-still), and the twelfth scale (fast- slow) were fairly common in the third factor. The fourth factor was composed of the second scale (hard-soft) and the tenth scale (ugly-beautiful). This fourth factor was a curious blend of the potency and evaluative dimensions. The first factor contained three of the four scales normally associated with the evaluative dimension. The other so-called evaluative scale (ugly-beautiful) be came a part of factor four. This evaluative factor ap peared to be consistent enough over all the concepts to warrant its use in testing some of the hypotheses concerned with the school-related concepts (school, teachers, and other pupils). In much of the analysis which follows, only the evaluative dimension will be used. In other words, the pupils' answers on only three of the twelve scales provided the data used to test the hypotheses. It is possible that concept-scale interaction accounted for the lack of anticipated factorial purity as well as the large amount of random data. Since the three usual factors of evaluation, ac tivity, and potency were not extracted by the factor anal ysis, a change in the original design appeared to be 65 warranted. Instead of considering change scores for each concept on a factor-by-factor basis, twelve dimensions of space were substituted so that each scale created a sepa rate dimension. In this way a range of change scores from zero to approximately one hundred was created for each concept. The over-all mean change score for all concepts was 30.52. Table 6 lists the means and standard devia tions for the change scores for each of the seventeen con cepts. There are only two concepts whose mean change scores would suggest that they are significantly different from the other seventeen. The technique of multiple dis criminant analysis tested these differences. The raw score data were tested by the technique of discriminant analysis so that the three schools were com pared. Table 7 reveals the fact that the best discrimi nator among schools was the concept "what I would like to be." School II was significantly different from Schools I and III when scale markings for this concept were compared. Sixty-eight and one-half per cent of the students in School II were properly classified by school. School III had as many incorrectly classified as properly placed, but School I had slightly more incorrectly placed than rightly so. The addition of the concept variable "going to con certs, playing instruments, singing," shown in Table 8 in rial inc 1 2 3 4 5 6 7 8 9 ID 11 12 13 14 15 16 TABLE 6 MEANS AND STANDARD DEVIATIONS OF PRE-TEST POST-TEST CHANGE SCORES FOR STEPWISE 0I5CRIMINANT ANALYSIS OF CONCEPTS BY SCHOOLS SCHOOL I SCHOOL II SCHOOL III Standard Standard Standard Mean Deviation Mean Deviation Mean Deviation 28.29 10.91 27.27 12.95 28.59 13.26 31.80 12.69 30.42 12.32 33.22 12.68 30.66 13.37 31.70 13.11 31.66 12.09 29.38 12.39 28.26 13.91 30.19 16.12 31.52 13.04 31.92 15.39 36.64 15.27 24.20 11.32 22.10 9.27 26.33 11.83 32.01 11.63 31.46 13.38 32.91 11.80 29.59 14.71 28.51 12.19 32.07 13.56 33.95 14.69 31.00 12.40 32.39 12.75 33.16 14.05 32.02 12.31 34.20 14.40 31.09 13.39 27.98 12.27 32.72 12.16 31.28 12.98 29.08 12.47 33.76 15.80 32.60 16.26 30.32 11.51 30.64 11.76 33.83 12.36 29.81 13.22 33.11 12.70 34.34 14.91 30.32 12.65 31.68 14.79 26.12 12.41 21.27 11.75 27.15 13.54 34.04 15.17 32.86 12.77 35.22 15.33 Mean = 30.52 sem = 3.16 TABLE 7 DISCRIMINANT ANALYSIS OF SCHOOL GROUPS STEP NUMBER 1 VARIABLE ENTERED 16 ("WHAT I WOULD LIKE TO BE") VARIABLES INCLUDED AND F TO REMOVE - DEGREES OF FREEDOM 2 294 16 6.8106 VARIABLES NOT INCLUDED AND F TO ENTER - DEGREES OF FREEDOM 2 293 1 0.0030 5 2.8544 9 0.5256 13 0.7209 2 0.3995 6 2.0554 10 0.2473 14 1.372B 3 0.7016 7 0.0741 11 2.2320 15 1.2405 4 0.0345 8 0.8439 12 1.3452 17 0.0994 U-STATISTIC 0.95572 DEGREES OF FREEDOM 1 2 ; 394 APPROXIMATE F 6.81063 DEGREES OF FREEDOM 2 294.00 F MATRIX - DEGREES OF FREEDOM 294 GROUP 2SCHL 1SCHL VARIABLE 16 CONSTANT GROUP 35CHL 2SCHL 1SCHL GROUP 35CHL 11.65269 0.29465 FUNCTION 3SCHL 0.17318 2SCHL 7.33171 25CHL D.13570 1SCHL 0.16663 -2.35089 -1.44336 -2.17640 NUMBER OF CASES CLASSIFIED INTO GROUP - 3 5CHL 2SCHL 1SCHL 47 28 42 40 83 34 7 10 6 TABLE 8 DISCRIMINANT ANALYSIS OF SCHOOL GROUPS STEP NUMBER 2 VARIABLE ENTERED 5 {"GOING TO CONCERTS, PLAYING INSTRUMENTS, SINGING") 2 293 VARIABLES INCLUDED AND F TO REMOVE - DEGREES OF FREEDOM 5 2.8544 16 6.1001 VARIABLES NOT INCLUDED AND F TO ENTER - DEGREES OF FREEDOM 1 0.1849 2 0.0798 3 1.0324 4 0.1941 6 7 8 9 U-STATISTIC APPROXIMATE F F MATRIX - 1.2933 0.1145 0.1930 1.7575 0.93746 4.B0817 10 11 12 13 0.0807 1.6230 0.7100 1.9272 DEGREES OF FREEDOM DEGREES OF FREEDOM 2 4 DEGREES OF FREEDOM GROUP 3SCHL 2SCHL 293 GROUP 25CHL 15CHL 7.59622** 2.65664 * FUNCTION 3.80372 3SCHL 2SCHL 1SCHL VARIABLE 5 0.15188 0.13405 0.12867 16 0.14954 0.11484 0.14661 CONSTANT -3.36067 -4.81239 -3.94295 NUMBER OF CASES CLASSIFIED INTO GRI 3SCHL 2SCHL 1SCHL GROUP 3SCHL 39 31 24 2SCHL 29 70 22 1SCHL 25 28 29 292 14 1.6755 15 1.8386 17 0.0003 2 294 586.00 ** p < .01 * p < .05 CO 69 step number two of the discriminant analysis, increased the number of "hits" in School I, but again, only School II proved to have more "hits" than "misses." School II was sufficiently unique to be declared significantly different at the .01 level of confidence. The other schools were unique enough to be significant at the .05 level. There were, of course, a number of students in Schools I and III who marked the concept dealing with musical interests in a manner characteristic of School II. The control and experimental groups were indistin guishable except for the concept "working with machines and tools" as indicated in Table 9. Although significance was not achieved with the single concept, the addition of the concept "school" made it possible to state that the combined score of these two concepts discriminated between the two treatment groups at the .01 level of confidence on the multiple discriminant analysis test as presented in Table 10. Adding further concepts did not increase the power to discriminate. Tables 9 and 10 show the results of step number one and step number two in the discriminant analysis of test and control groups. An F of 5,91 was necessary at the .05 level of confidence and 6.81 for the .01 level. Significance was noted in Table 10 only. Even with this statistical support of differences, it is important to TABLE 9 DISCRIMINANT ANALYSIS OF TEST/CONTROL CROUPS STEP NUMBER 1 VARIABLE ENTERED 3 ("WORKING WITH MACHINES AND TDDLS") VARIABLES INCLUDED AND F TD REMOVE - DEGREES OF FREEDOM 3 3.3798 n.s. 145 VARIABLES NDT INCLUDED AND F TD ENTER - DEGREES OF FREEDOM 1 144 1 2 4 5 6.6202** 0.2968 0.0367 1.1221 6 7 8 9 U-STATISTIC APPROXIMATE F 0.0006 1.6307 0.6807 0.9920 0.97722 3.37981 ID 11 12 13 3.7392 0.1372 0.1962 0.0174 DEGREES OF FREEDOM DEGREE5 OF FREEDQM 1 1 14 15 16 17 1 145 145.00 3.6404 3.8085 2.4093 0.4758 F MATRIX - DEGREES OF FREEDOM 145 GROUP TEST 1 VARIABLE 3 CONSTANT GROUP CNTRLl YCNTRL TEST 1 XTEST2 GROUP CNTRLl 3.37982 n.s. FUNCTION CNTRLl TEST 1 0.20022 0.17350 -3.49264 -2.62287 NUMBER OF CA5ES CLASSIFIED INTO GROUP - CNTRLl TEST 1 15 11 47 44 21 26 64 67 ** p< .01 o TABLE 10 DISCRIMINANT ANALYSIS OF TEST/CONTROL GROUPS STEP NUMBER VARIABLE ENTERED 2 1 ("SCHOOL") VARIA8LES INCLUDED AND F TO REMOVE - DEGREE5 OF FREEDOM 1 1 6.6202* 3 7.7607** VARIABLES NOT INCLUDED AND F TO ENTER - DEGREES OF FREEDOM 1 144 143 2 0.3568 4 0.7706 5 0.3021 6 0.7068 7 8 9 10 U-STATI5TIC APPROXIMATE F 0.9319 D.0138 0.0D8D 2.5964 0.93427 5.06551 11 12 13 14 0.12B8 0.0002 1.2908 1.5818 15 1.7905 16 1.0654 17 0.1722 DEGREES OF FREEDOM DEGREES OF FREEDDM 2 2 1 145 144.00 F MATRIX - DEGREES OF FREEDOM GROUP CNTRLl 144 GROUP TEST 1 VARIABLE 1 3 CON STA N T GROUP CNTRLl YCNTRL TEST 1 XTE5T2 5.06551** FUNCTION CNTRLl 0.07091 0.1688B - 3.82946 TEST 1 0.11420 0.12304 - 3.49632 NUMBER OF CASES CLASSIFIED INTO GROUP - CNTRLl TEST 1 21 14 42 47 15 23 69 64 ** P < .01 * p < .05 -o H 72 guard against overgeneralization- The ability to discrim inate does not imply that overlap is not general. There is overlap in the two groups, even though these groups appear to be quite dissimilar. In step number three, the addition of the concept "meeting and dealing with people to promote products or things to sell" did not increase the power of discrimina tion. Even though computations were made as far as step seventeen, this material was not included in the findings because it did not add anything significant. Testing the Hypotheses 1. There will be no significant difference be tween the experimental and control group with regard to test-retest scores for all concepts on all factors or dimensions. This hypothesis was not tested over all factors or dimensions because of the lack of factorial purity over both administrations. However, it was noted in the pre vious discussion that the concepts "working with machines and tools" and "school" did distinguish between the exper imental and control groups at the .01 level of confidence as evidenced in Table 10. Why this combination would pro duce a significance is beyond the scope of this investi gation. The concept "working with machines and tools" may 73 have some relation to the course content of the driver edu cation class where the automobile is the focus of atten tion. In like manner, the possibility exists that since "school" is a primary focal point in the guidance class, the concept "school" may evoke stronger feelings in this class than in the driver education class. 2. There will be a positive significant change for the experimental group for those concepts which relate to general school orientation ("school," "teachers," "boys at this school," "girls at this school") when the evaluative dimension alone is measured. This hypothesis was not confirmed. The results of an inspection of the data presented in Table 11 precluded the use of any statistical test since a one-tailed hypoth esis was stated. Instead of moving in the direction of a more positive identification with school, there was a decided trend toward assigning a less positive scale posi tion to the school-related concepts. When the response mode for the three scales of kind-cruel, true-false, and wise-foolish was examined for this population of 223 stu dents (with 2,676 possible responses), only 33 per cent of the students had responded in a more positive manner to ward these school-related concepts after the ten-week treatment period. Fifty per cent of these students marked 74 these concepts less favorably, and 17 per cent indicated no change in meaning. The concepts labeled "girls at this school" and "boys at this school" elicited fewer negative responses than those designated "school" and "teachers." Table 11 summarizes the responses to these four concepts. TABLE 11 EXPERIMENTAL GROUP PRE-TEST POST-TEST CHANGE SCORES Direction of Change Concept More Positive Less Positive No Change Total School 64 123 36 223 Teachers 60 125 38 223 Girls 82 106 35 223 Boys 87 91 45 223 Total 293 445 154 892 The scale marking patterns for the four school- related concepts provided little helpful data in support of the idea that the guidance class was responsible for a significant positive change in students' attitude toward school, teachers, and boys and girls in their schools. The concept "teachers" brought forth a more favor able response in 20 per cent of the responses from the guidance classes as opposed to 18 per cent in the driver education (control) classes. The percentage of responses which were unchanged over two administrations favored the guidance class by 9 per cent. The no-change pattern was noted in 45 per cent of the guidance class respondents, while only 36 per cent of the driver education subjects indicated like responses. The negative response pattern was more characteristic of the driver education pupils— 46 per cent to 35 per cent. The average scale point devi ation was -.47 for the students in the driver education classes in contrast to a -.37 scale point deviation for the guidance class. In other words, the concept "school" was unable to evoke a favorable response of any intensity, but rather there was a slight shift (an insignificant shift, however) in the direction of a lower evaluation of this concept when the scales kind-cruel, true-false, and wise-foolish were used. In other research studies these scales have been identified as part of the evaluative dimension or the evaluative factor. The other school-related concepts extracted simi lar response patterns. Table 12 provides a more detailed account of this, but the pattern of a slight change toward a less intense evaluation of these four concepts persists. The concept "teachers" elicited the strongest negative responses in the driver education class at School I. Students in that class made only fourteen 76 TABLE 12 PATTERN OF CHANGE RESPONSES AMONG SCHOOLS FOR EVALUATIVE DIMENSION ON SCHOOL-RELATED CONCEPTS School I School II School III Mora intense or Positive (E= 77.52) 0= 86 (E=119.56) 0=101 (E«t 95.91) 0=106 Total 293 Less intense or Negative No Change (E=120.95) 0=100 (E= 40.74) 0= 50 (E=181.59) 0=213 (E- 62.84) 0= 50 (E=145.00) 0=132 445 (E= 50.41) 0= 54 154 Total 236 364 292 892 with df = 4 X2 = 22.187*** ***p < .001 77 responses of a more positive nature in the second adminis tration to the concept "teachers," while fifty responses were more negative and thirty-six remained unchanged. The average scale point difference was -1.04, the largest change noted in any class on any of the school-related concepts. Scale point difference was the mean summed change score obtained from the scale markings for the three evalu ative items. In other words, the difference between scale markings over two administrations gave a difference score for each of the three evaluative items. These were summed and a mean was computed. This mean represented both inten sity and direction. In School I, the guidance class pattern was 41 per cent no change, 25 per cent more positive, and 34 per cent more negative. The guidance classes were very similar to the driver education classes when responses from all schools were tabulated. The difference in positive evalu ations favored the guidance class by 3 per cent— 20 per cent to 17 per cent. The less intense response accounted for 38 per cent of the guidance class responses and for 46 per cent of those made by the driver education students. It may be noted that 62 per cent of the responses made in the guidance classes were unchanged or more positive, while only 54 per cent of the responses in the driver 78 | education classes could be so categorized- The scale point deviation was -.44 for the guidance classes and -.80 for the driver education classes. j i For the concept "'girls at this school," 44 per cent] i I of the driver education pupils made less intense reactions i on the second test than on the first test. This was true for 30 per cent of the subjects in guidance classes. The percentage of more positive responses was nearly identical for both groups— 27 per cent for driver education and 26 per cent for guidance. This indicated that the no-change pattern was more common with the guidance class— 44 per cent to 29 per cent. The average scale point deviation was -.12 for the guidance classes and -.51 for the driver education classes. In six of the guidance classes a slight trend ap peared toward more positive or more intense evaluation of the concept "boys at this school." However, when all classes are considered, the usual pattern of responses persists. The positive responses favor the guidance classes with 28 per cent of the responses to 24 per cent for the driver education classes. The no-change scores again favor the guidance classes— 43 per cent to 39 per cent. The negative responses account for 37 per cent of those made in driver education, while 29 per cent made in guidance were negative. The average scale point deviation 79 per response was +.06 for the guidance classes and -.20 for the driver education classes. In class number three in School I, the average scale point deviation was +.76, the largest positive response for any class. 3. There will be no difference between schools on those concepts which relate to general school orientation. The null hypothesis was rejected at the .001 level of confidence by the test. Table 12 presents this in formation. Schools I and III were similar in the distri bution of total responses which reflected a negative change, a positive change, and no change from the initial scale markings on the school-related concepts when measured on the evaluative dimension. While School II had a larger proportionate share of negative responses than the others, it followed the trend toward having the most common change in the direction of less positive or less intense scale marking. Approximately 43 per cent of the student re sponses in Schools I and III were less intense, about 37 per cent of the responses were more intense, and approxi mately 20 per cent signified no change. In School II, the less intense responses were double the more intense ones (approximately 58 per cent to 28 per cent), while about 14 per cent indicated no change. The positive change scores for School I were 80 rather evenly divided among the four concepts. School III had 50 per cent of the positive response change scores on the combined concepts "boys at this school" and "girls at this school" and only 44 per cent on the concepts of "school" and "teachers." School Ill's positive change responses showed a decided difference which favored the age mates over the school and its staff by a margin of two to one. There was a difference of only one response be tween "school" and "teachers" in School II on both the negative and positive change scores. Even though the null hypothesis was rejected, the most common response in all schools at the time of the second administration was toward a less positive scale position for the school-related concepts when measured by the three scales usually associated with the evaluative dimension. Multiple discriminant analysis revealed those concepts which did establish significant differences be tween schools— these were "what I would like to be" coupled with "going to concerts, playing instruments, singing." 4. There will be no difference between classes taught by a regular teacher and classes taught by a counselor or grade advisor on the concepts which relate to general school orientation. 81 Two of the nine experimental classes were taught by a counselor. These classes are identified as "2306" and "2407." Both were taught by the same counselor. Sixty students comprised this portion of the total sample, while 163 remained in the experimental "teacher-taught" classes. The null hypothesis was rejected at the .01 level of confidence. Unfortunately, the discovered difference suggests that when the total responses for all the school- related concepts were summed and categorized by positive mean response, negative mean response, and no change, the students in counselor-taught classes showed significantly more negative responses than did the subjects in classes taught by the regular guidance class teacher. Even though this appears to be a statistically significant finding with a non-parametric test, it would be dangerous to assess a cause-effect relationship here. When each of the con cepts was tested by the test, there was no significant difference between groups on the concepts "boys at this i school, 1 1 "girls at this school," and even "teachers." The concept "school" indicated a difference in response pat tern which was significant at the .025 level of confidence. Table 13 illustrates this finding. The difference noted was not one of kind but of degree. The students in both groups totaled more negative response changes than 82 TABLE 13 CHANGE SCORE RESPONSE PATTERN FOR ALL SCHOOL-RELATEQ CONCEPTS AS A FUNCTION OF GUIDANCE CLASS INSTRUCTOR Guidance Class Teacher Counselor More intense or Positive Less intense or Negative No Change Total By concepts: (E = 214.1659) 0 = 230 (E . 325.2690) 0 a 303 (E » 112.5650) 0 « 119 (E « 78.8340) 0 = 63 (E = 119.7309) 0 = 142 (E « 41.4349) 0 = 35 652 with df « 2 X; 240 11.1283** "schooln df . 2 X2 = 8.98* "teachers" df - 2 x2 a 3.34 n.s. "girla at this school” df = » 2 X2 . 5.76 n.s. "boys at this school" df = 2 X2 = 1.30 n.s. Total 293 445 154 892 *p < .025 **p < .01 83 positive ones. The counselor’s students changed more to negative responses than to either a positive or no-change reaction. The difference between students was found in about 4 per cent more no-change responses and about 9 per cent more positive responses in the seven teacher-taught guidance classes than in the two classes led by a coun- selor-teacher. 5. There will be no difference between boys and girls on the concepts that relate to general school orientation. The null hypothesis was accepted at the .01 level of confidence. There was only a slight difference between girls and boys on any of the three response patterns— positive, negative, or no change— when the mean responses for the three evaluative scales were associated with the four school-related concepts. Table 14 reveals the re sults of the chi-square test of significance. Interestingly enough, the response patterns, when categorized by boys and girls for the three response methods, were exceptionally close when percentages were considered. The boys' change responses were 32.15 per cent positive compared to 33.48 per cent for the girls; the negative responses for the boys accounted for 48.57 per cent compared to 51.05 per cent for the girls; the no-change pattern characterized 19.29 per cent of the boys' 84 TABLE 14 PD5T-TE5T RESPONSE PATTERN OF BOYS/GIRLS FOR SCHOOL-RELATED CONCEPTS Boys1 Responses Girls' Responses More intense or Positive (E = 137.9496) 0 a 135 (E a 155.0404) Q = 158 Total 293 Less intense or Negative (E » 2D9.5291) 0 a 204 (E a 235.4708) 0 . 241 445 No Change {E « 72.1120) 0 = 81 (E = 81.4887) 0 = 73 154 Total 420 472 892 with df « 2 X2 * 1.0962 n.s. answers compared to 15.47 for the girls. When the change response pattern for these four school-related concepts was examined, agreement was dis covered in the direction of the most negative changes for "teachers" and "school" in that order. The girls indi cated more positive changes for the concept "hoys at this school" than for their sex-mates "girls at this school." The boys' responses changed only slightly in favor of the opposite sex. The negative response pattern appeared to be slightly different for boys and girls. The boys rejected "school" slightly more than "teachers," while the reverse was true for the girls. The boys rejected "girls at this school" slightly more than "boys at this school," but the girls rejected their sex-mates considerably more than the opposite sex. However, the rejecting or negative pattern again operated at nearly a two-to-one ratio over the posi tive pattern for the concepts "school" and "teachers." Little difference existed between girls and boys on the concept "boys at this school" as far as rejecting or ac cepting patterns, but the girls rejected "girls at this school" at a ratio of nearly three to two. The no-change response pattern was more character istic of the boys' responses to a slight degree. In summary, there was no significant difference 86 between girls and boys in the direction of their post-test responses. Following the ten-week class, both groups indi cated less intense responses for the four school-related concepts than they did in September. No cause-effect relationship is suggested, however. The post-test re sponses, although slightly less positive were, neverthe less, positive in nature. In other words, these school- related concepts were perceived by the subjects to be more "kind" than "cruel," more "true" than "false," and more "wise" than "foolish." 6. There will be no difference between high, low, and average ability students on the concepts that relate to general school orientation. The null hypothesis was rejected at the .05 level of confidence; a of 13.28 would have been sufficient to reject the null hypothesis had it been set at .01. These data are presented in Table 15. The only two dispropor tionate areas were in the positive mean responses for both the high and low stanine groups. A few more positive re sponses were made than were expected by the low ability group, and the opposite was true for the high ability group. It follows, then, that the high ability group was more inclined to assign a less positive meaning to school- related concepts during the second administration of the 87 TABLE 15 POST-TEST RESPONSE PATTERN FOR SCHOOL-RELATED CONCEPTS OF STUDENTS IN GUIDANCE CLASSES BY I.Q. GROUPS High Ability (Stanines 9,0,7) N * 55 Average Ability (Stanines 6,5,4) N = 131 Low Ability (Stanines 3,2,1) N = 37 More intense (E- 72.26) (E=172.12) (E=37.98) Total or Positive 0 a 59 0 = 179 0 = 55 293 Less intense (E=109.75) (E=260.B5) (Ea73.83) or Negative 0 = 117 0 a 260 0 a 68 445 No (E» 37.90) (E« 90.47) (Ea25.55) Change 0 s 44 0 a 85 0 a 25 154 Total 220 524 140 892 with df - 4 X2 a 12. *p < .05 06* test than was the low ability group. The three ability groups assigned evaluative mean ings to the four school-related concepts in a manner con sistent with other findings enumerated in this study. The students indicated a less positive evaluative meaning for school and teachers and, in a lesser degree, for boys and girls in their schools. Regardless of stanine category, there is a pattern which persists. About 50 per cent of the responses for all three groups were toward a less positive meaning for the school-related concepts. About 30 per cent of the responses were more positive than in the previous administration, and about 20 per cent showed no change. A student's responses did not apparently re flect the amount of scholastic ability he possessed. 7. This instrument (the Seim Attitude and Vocational Interest Inventory) will be unable to rank the vocational interest terms so as to correlate significantly with a pupil's scores on the Kuder Pref erence Record-Vocational. The null hypothesis was confirmed predicting no significant correlation between the rank order of inter ests measured by the Kuder Preference Record-Vocational and the same interests as measured by the SAVII, an interest test based upon the semantic differential 89 technique. Table 16 reveals that approximately 21 per cent of the correlations were significant, but this ratio was not sufficient to permit generalization. The SAVII concepts, paraphrased from the statements on the profile j sheet used by the students in this investigation to chart j I their Kuder scores, apparently did not elicit responses which could be rank ordered so as to correspond with those on the Kuder interest profile. Of the twenty-eight students whose ranked interest scores were significantly correlated, nine were from School I, twelve from School II, and seven from School III. Four classes accounted for twenty of these twenty-eight students? nine were girls and nineteen boys. The majority of these pupils were of average ability. Although statistical significance was not reached for a major portion of the total sample, there may have been more than a chance relationship between the interest items in these two tests. Further investigation may be justifiable. Other Findings This investigation revealed that students in grades other than BIO who were members of guidance classes assigned a less evaluative meaning to the school-related concepts after ten weeks of school. Table 17 confirms ID 1 11 14 17 24 27 40 42 51 55 59 69 73 79 81 85 87 94 96 97 103 107 109 110 111 112 113 121 123 125 120 90 TABLE 16 RANK CORRELATION BETWEEN VOCATIONAL INTEREST CONCEPTS ON KUDER PREFERENCE RECORO-VOCATIGNAL AND SAVII N * 127 r ID r ID r 10 r .49 138 .30 3 .21 134 ,91**b -.26 140 • 58*a 5 .05 136 .86**b .43 143 .23 10 .39 139 .43 -.15 145 .66*a 13 .24 142 .09 .03 146 .53 15 .06 150 -.30 .26 151 .46 16 .46 155 .85**b .45 153 .15 18 .00 157 -.45 .24 162 .35 29 .46 158 • 60*b -.12 ,, 170 .65*a 30 .36 164 .34 .00 171 -.21 37 .54 165 -.26 -.11 173 .66*a 44 .18 167 .13 • 71*a 182 -.17 47 -.56 180 .6l*b .49 187 .02 48 .33 184 .16 .53 201 .45 50 .28 213 -.36 .15 212 .67*a 56 .20 215 .77*b .51 219 .44 57 .47 220 -.14 .28 221 .6Q*a 62 -.22 226 .24 .43 225 .49 64 ,71*b 227 .17 .58*a 229 •76**a 70 .48 231 -.17 .50 234 .39 74 -.60*b 232 .37 .51 238 .64*3 80 .50 239 -.07 .66*a 241 -.03 83 .54 243 .09 .56 242 .33 64 -.74*b 250 .45 -.43 246 .55 88 -.09 254 -.48 .10 248 .53 89 .51 260 .8Q**b .25 253 .49 91 .33 264 -.25 .38 255 .52 108 .75*b 265 .25 .69*a 257 .52 118 -.24 267 .88**b .65*a 207 .07 120 -.50 269 .25 .55 292 .19 126 .51 274 .12 • 62*a 294 .71*3 127 .42 278 .09 132 .46 291 -.36 133 ,58*b * p < .05 **p < .01 a— based upon 10 concepts including "working outdoors” b— based upon 9 concepts not including "working outdoors" ID— student identification number r= rho (Spearman's renk-difference correlation) 91 this finding. The concept "school" was seen as more posi tive by 27^ per cent of the pupils not in the BIO. Fifty- one and one-half per cent of the mean responses were negative, and 15 per cent showed no change. This same pattern followed for the three other school concepts. TABLE 17 DIRECTION OF POST-TEST RESPONSES MADE TO SCHOOL-RELATED CONCEPTS BY STUDENTS OTHER THAN BIO'S CONCEPTS Direction "School" "Teachers" "Girls at "Boys at of this school” this school" Change (Per Cent) (Per Cent) (Per Cent) (Per Cent) Positive 27.5 25 27.5 25 <+) Negative 57.5 65 52.5 57.5 (-) No Change 15 10 20 17.5 Total 100 100 N = 31 100 100 Although not listed as part of the statistical hypotheses, the three schools provided mean scores over all seventeen concepts which were tested for significant dif ferences between schools. Table 18 describes the results of an analysis of variance and a resulting t test to test these differences. An F of 4.43 indicated that significant 92 differences did exist at the .05 level of confidence. The t test which followed pointed out that there was a signif icant difference at the .05 level between the mean change scores for Schools II and III but not for Schools I and II or Schools I and III. School II was the most unique of the three. School I appeared to be the least distinctive. This confirms the findings reported previously in Table 8. TABLE 18 SUMMARY OF ANALYSIS OF VARIANCE Source of Variation Sums of Squares a- • Hi • Mean Square F Among schools 65.87 2 32.94 4.43 Within schools 356.64 48 7.43 Total 422.51 50 Tests of differences between schools by use of t: School I X School II n. s. School I X School III n. s. School II X School III p < .05 Summary The research hypothesis setting forth the suppo sition that students in the nine guidance classes would assign a more positive meaning to four school-related concepts following a ten-week guidance class was not 93 accepted. The guidance classes, instead, ascribed less evaluative meanings to these concepts. The average change for all students in the guidance class who completed both administrations of the test was -.20 scale points for the four concepts which were desig nated as school-related concepts. The percentage of posi tive change responses was almost equal— only 23 per cent for the guidance group and 21 per cent for the driver edu cation class. The null hypotheses of no difference between groups held true when boys and girls were compared and when schools were compared. The null hypothesis that there would be no sig nificant difference between those classes taught by a regular teacher and those taught by a counselor or grade advisor was rejected. In fact, a barely significant find ing revealed that the students taught by a counselor as signed less evaluative meanings to school-related concepts. This was especially true for the concept "school." Significant difference did exist among students when categorized by mental ability. Above-average ability students assigned less evaluative meanings to school- related concepts than did students of below-average ability. The null hypothesis was confirmed that there would 94 not be a significant relationship between the ranked interest scores of certain vocational interest items on the SAVII and those on nine of the scores for the well- known Kuder Preference Record-Vocational. The findings of the factor analysis study failed to support the assumption that the twelve scales used to measure each concept represented the three dimensions of meaning usually labeled as evaluation, activity, and potency. Students other than BlO's also assigned a negative rather than a positive meaning to school-related concepts after the treatment of tenth grade guidance. CHAPTER V SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS Summary Two hundred and ninety-seven pupils in three high schools in a metropolitan school district were administered a seventeen-concept semantic differential test which asked for a response on a six-point scale to twelve bi-polar ad jectives such as true-false. The test was administered at the beginning and again at the end of a tenth grade ten- week guidance class. The test instrument was designed to discover the meaning assigned by students to four school-related con cepts, eleven work-related concepts, and two self-concepts. The technique of multiple discriminant analysis was used to learn which concepts were most able to dis criminate among classes and schools and between experi mental and control groups. Significant differences (p < .01) were found between test and control groups when the concept in step two, "school," was added to the first concept, "working with machines and tools." The addition 95 96 of the concept in step three did not increase the signifi cance already achieved. A second application of discriminant analysis which provided significance was the use of the concept "what I would like to be" to discriminate by schools. School II, the school with the highest percentage of above- average intelligence scores, proved to be significantly different from the other schools by this means. This test properly classified almost 70 per cent of the students in School II. School I was the least unique. Step two dis criminated even further when the scale markings for the concept "going to concerts, playing instruments, singing" were employed. With the additional information from step number two, School II was different enough to be signifi cant at the .01 level of confidence, while the other schools both reached the .05 level. A number of carefully designed factor studies failed to extract the factors usually associated with semantic differential studies. It was not possible to accept the idea that the three dimensions or factors of evaluation, potency, and activity could be assigned to the scales in such a way as to analyze the data according to factors. The pattern of scales labeled kind-cruel, true-false, and wise-foolish was consistent enough over the concepts to accept it as the evaluative dimension. 97 This dimension, which accounted for about 40 per cent of the total variance, was used in order to evaluate the change in meaning associated with the school-related con cepts . The tenth grade students considered in this study- assigned a less evaluative meaning to the school-related concepts after ten weeks of school in all three schools and in most of the classes. In only a few classes was there a slightly more favorable evaluation of the concepts of "school," "teachers," "girls at this school," and especially of "boys at this school." The null hypothesis that there would be no sig nificant difference between the experimental and control groups with regard to all concepts on all factors or di mensions was rejected. A significant difference at the .01 level was found between these groups on the concept "working with machines and tools" when combined with the concept "school" in a multiple discriminant analysis test. The research hypothesis that there would be a positive significant change for the experimental group for those concepts which relate to general school orien tation when the evaluative dimension alone was measured was also rejected. The chi-square test indicated no sig nificant positive change; in fact, the change was in a negative direction, but the hypothesis was one-tailed. 98 The null hypothesis that there would he no differ ence between schools on the concepts which relate to school orientation was accepted. However, there were significant differences between schools on the concept "what I would like to be" when it was combined with "going to concerts, playing instruments, singing." In this case a significant difference was accepted at the .01 level of confidence for School II and at the .05 level for Schools I and III when tested by the multiple discriminant analysis test. The null hypothesis was rejected that there would be no difference between classes taught by a regular teacher and those taught by a counselor or grade advisor on those concepts which relate to general school orienta tion. The chi-square test showed this to be significant at the .01 level of confidence. The students in the counselor-taught classes indicated significantly more negative responses than did subjects instructed by the regular guidance class teacher. The null hypothesis that there would be no sig nificant difference between boys and girls on the school concepts was accepted on the basis of a chi-square test. Both boys and girls tended to devalue school after the first ten weeks of the fall semester. No real differences existed in the three response patterns over all school- related concepts. While the boys rejected "school" 99 slightly more than teachers, the girls rejected "teachers" more than the concept "school." The null hypothesis that there would be no differ- i ence between high, low, and average ability students on | the school-oriented concepts was tested by chi square and rejected. At the time of the second administration, high ability students assigned significantly fewer positive meanings to the concepts than did low ability pupils. The null hypothesis was accepted that the test instrument, known as the Seim Attitude and Vocational Interest Inventory, would not be able to rank the voca tional interest items so as to correlate with pupils' scores on the Kuder Preference Record-Vocational. Conclusions The results of the testing of the hypotheses would suggest that tenth grade students assign a positive mean ing to school-related concepts when they first arrive at their new high school and that this favorable attitude (assignment of meaning) toward school-oriented concepts does not persist at this intensity for as long as ten weeks. The direction of change is negative. What happens may be an acceptance of the realities of homework, daily routines, the pressure of grades, and teacher variation in terms of classroom management, methods of presenting 100 material, pupil-teacher relationships, and systems of assigning marks. This phenomenon would seem to parallel the popular notion of the acceptance of reality as ex pressed by a newly married couple when they admit, "The honeymoon is over." If this diminution of positive atti tudes toward school is a fact of school life, then attempts to measure the improvement of a pupil1s attitude toward school should be deferred until the reality phase takes over. To determine if this change to a less intense or less positive meaning for school were peculiar to BlO's alone, a look at the change scores for the students in other grades would appear warranted. Students in grades other than BIO who were members of the guidance classes also assigned a less evaluative meaning to the school- related concepts after ten weeks of school according to the data presented in Table 17. If the majority of the stu dents who comprised the non-BlO group attended the same school the previous semester, two possible conclusions could be drawn in explanation: the decline was due to a November "letdown," or it was in some way a regression to the mean to be expected on such a testing instrument. It would be difficult to establish which alternative has the more merit. Adopting a pre-test post-test design with the guidance class for the ten-week period from February to 101 April might test these conclusions. Recommendat ions Introduction The tenth grade ten-week guidance class appears to be more than an expedient to fill the block of time which alternates with the mandatory driver education class. An evaluation of a class is much more than an examination of a course of study. It entails an analysis of the indefin- ables such as school spirit, the teacher's attitude to ward the worth of the subject, and the teacher's feelings toward each student as well as the individual student's attitude toward the subject matter, the instruction, and the instructor. The instrument employed in this investigation samples but a small portion of the complex interrelation ship of student, subject, school, and teacher. The empha sis was, however, placed on the student and his percep tions. Other techniques of measuring attitudes may be more effective than the measurement of meaning. It might be that the real value of the guidance course, or the entire high school experience, can only be measured in terms of successful adult patterns of adjustment. 102 Vocational Interests Nine of the concepts embodied in the test instru ment used in this study were intended to be suggestive of vocational interests. Undoubtedly some of the concepts were so poorly phrased as to obscure the precise intent of the test. As one searches the Sixth Mental Measurements Yearbook for evaluations of interest tests, he is bound to acquire a feeling that interest tests administered to tenth grade students will produce results which are inter esting but hardly predictive of future endeavors. Perhaps the vocational interest test score is only one of a number of bits of information which each individual must use as he attempts to assess the relative strength of a number of vocational possibilities. Because the tenth grade is not a year to make final decisions, the student may be able to accept the results of any interest test with an objectivity which sees all vocational interest indicators as merely suggestive. The School Environment Successful school experiences often follow first impressions which are positive. Those students who initially perceived school as good, active, and strong found school to be much the same after ten weeks. How ever, the evaluative scores were slightly depressed for 103 all students- Whether this slight drop is correctly assessed by this instrument is not known? possibly such factors as the reality of tests, increased homework, and a larger school environment have caused school to be slightly devalued in the eyes of the pupils tested. Students viewed the concept "school" in almost the same manner as they did "teachers," whereas they as cribed slightly higher evaluative scores to the concepts of "boys at this school" and "girls at this school." The obvious remedy for lack of school identifica tion is the development of programs within the school which increase the number of people who have successful school experiences. Removing the threat of grades might be helpful in accomplishing this end, say some critics of education. The Self-Concept Clinical studies using the semantic differential technique often cite the distance score between "me" and "what I would like to be" as a measure of self-acceptance. This was not a function of this investigation. These con cepts were included only as they might relate to vocation al concepts. The fact that the concept "what I would like to be" acted as a discriminator among schools was acci dental. This finding should, nevertheless, be tested in a further study. Recommendations for Further Study 1. An investigation of the value of a guidance class through the use of a longitudinal study appears warranted. A questionnaire should be developed for high school juniors or seniors. 2. Curricular innovations which depart from the prescribed guidance course of study should be evaluated by comparing matched treatment groups. 3. More complete randomization of sample selec tion might produce different results from those herein presented with a known selection bias. Computer programming which offers more randomization should be employed. 4. The test instrument should be adjusted to accommodate a neutral position to test the forced-choice dilemma for the test-taker. 5. The test instrument developed for this study should be administered to students who have expressed vocational preferences in test in struments other than the Kuder Preference Record-Vocational. The newer Kuder inventory might be helpful. 105 6. Future investigators should analyze the raw score data so as to achieve a better correla tion between the semantic differential tech nique and vocational interest inventories. 7. The effect of the elimination of marks for the guidance and the driver education classes should be tested; pass/fail might be substi tuted. 8. A similar study should be conducted at midyear to test whether different results (possibly less negative change) would occur. November to February testing or February to April test ing might test the assumption that the period from September to November represents the transition from enthusiastic approval of school to a more realistic attitude. 9. A study which categorized students by high school major or by career choice might suggest the nature of the student most likely to benefit from a guidance class. 10. The use of closed-circuit television to dis seminate information regarding careers might increase the value of the tenth grade guidance course. The need exists for additional re search to be done in instructional technology. 106 11. The readiness for a guidance class should be tested. Those students who feel that their future is already charted might wish to defer taking the guidance course until their senior year, if not altogether. The guidance class might then give more of its attention to the decision-making process. 12. The evaluation of a guidance class essentially demands a test which can be validated against ultimate criteria. A study of these criteria might place the class in its proper perspec tive. SELECTED BIBLIOGRAPHY SELECTED BIBLIOGRAPHY 1. 2. Baer, Max F. "Vocational Guidance in Group Activi ties," Occupations, 25:530-534, May, 1947. _ , and Roeber, E. Occupational Information. Chicago: Science Research Associates, Inc., 1964. Benoit, Robert. "An Investigation of Changes in Knowledge and Attitudes of Counselor-Trainees During the Course of an NDEA Guidance Institute and their Relation to Counseling Competence." Unpublished Doctoral dissertation. School of Education, University of Southern California, 1964. 4. Billings, Mildred Lincoln. Occupations. Scranton: Company, 1941. Group Methods of Studying International Textbook 5. Borow, Henry (ed.). Man in a World at Work. Boston: Houghton Mifflin Company, 1964. 6. , and Lindsey, Robert V. Vocational Planning for College Students. Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1959. 7. Buros, Oscar (ed.). Sixth Mental Measurements Yearbook. Highland Park, New Jersey: The Gryphon Press, 1965. 8. Catron, David W. "Educational-Vocational Group Counseling: The Effect on Perception of Self and Others," Journal of Counseling Psychology, 13:202-207, 1966. 9. Coogan, H. J. "A Course Outline for Tenth Grade Personal Guidance." Unpublished Master's project, School of Education, University of Southern California, 1960. 108 109 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. Cooley, William W., and Lohnes, P. R. Multivariate Procedures for the Behavioral Sciences. New York: John Wiley and Sons, Inc., 1962. Davis, Donald Arthur. "An Experimental Investigation of Group Guidance in a Secondary School." Unpub lished Master's project, School of Education, University of Southern California, 1958. English, H. B., and English, A. C. A Comprehensive Dictionary of Psychological and Psychoanalytical Terms. New York: Longmans, Green and Company, 1958. Ford, LeRoy H., Jr., and Meisels, Murray. "Social Desirability and the Semantic Differential," Educational and Psychological Measurement, 25; 465-475, 1965. Gribbons, Warren D. "Evaluation of an Eighth Grade Group Guidance Program," Personnel and Guidance Journal, 38:746-747, 1960. Halten, Torger. "Bridging the Gap Between Different Schools Through the Tenth Grade Junior Problems Classes in the Los Angeles City Schools." Unpublished Master's project, School of Education, University of Southern California, 1959. Harris, Chester W. (ed.). Encyclopedia of Educational Research. New York: The Macmillan Company, 1960. Hoppock, Robert. Group Guidance: Principles, Techniques and Evaluation. New York: McGraw- Hill Company, Inc., 1949. Hoyt, Donald. "An Evaluation of Group and Individual Programs in Vocational Guidance," Journal of Applied Psychology, 39:26-31, 1951. Hunt, Richard A. "Self and Other Semantic Concepts in Relation to Choice of a Vocation," Journal of Applied Psychology, 51:242-246, 1967. Husek, T. R., and Wittrock, M. C. "The Dimensions of Attitudes Toward Teachers as Measured by the Semantic Differential," Journal of Educational Psychology, 53:209-213, 1962. 110 21. 2 2 . 23. 24. 25. 26. 27. 28. 29. 30. 31. Kagan, Norman. "Group Procedures, 1 1 Review of Educational Research, 34:274-285, April, 1966. Kerlinger, Fred N. Foundations of Behavioral Research: Educational and Psychological Inquiry. New York: Holt, Rinehart, and Winston, 1965. Krumboltz, John D. "Stating the Goals of Counseling." Monograph published by the California Counseling and Guidance Association, 1966. Likert, R. A. "Technique for Measurement of Attitudes," Archives of Psychology, 22:5-55, 1932. Miracle, Betty. "A Unit in Vocational Guidance for Tenth Grade Students." Unpublished Master1s project, School of Education, University of Southern California, 1963. Miron, Murray S. "The Influence of Instruction Modification Upon Test-Retest Reliabilities of the Semantic Differential," Educational and Psychological Measurement, 21:883-893, 1961. Nelson, Ronald G. "A Manual of Instructions for Use with the Semantic Differential." Unpublished Master's project, School of Education, Univer sity of Southern California, 1964. Ohlsen, Merle. Guidance Services in the Modern Secondary School. New York: Harcourt Brace and World, Inc., 1964. Osgood, Charles E. "The Nature and Measurement of Meaning," Psychological Bulletin, 49:197-237, 1952. __________, and Suci, George J. "A Measure of Relation Determined by Both Mean Difference and Profile Information, 1 1 Psychological Bulletin, 49:251-262, 1952. __________ (ed.). "Psycholinguistics— Background for Semantic Differential— A Survey of Theory and Research Problems," Psycholinguistics, 49:171-176, 1954. Ill 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. , and Luria, Zella. "A Blind Analysis of a Case of Multiple Personality Using the Semantic Differential," Journal of Abnormal and Social Psychology, 49:579-581, 1954. , and Suci., George J. "Factor Analysis of Meaning, " Journal of Experimental Psychology, 50:325-338, 1955. , , and Tannenbaum, Percy H. The Measurement of Meaning. Urbana: University of Illinois Press, 1957. __________. "Studies on the Generality of Affective Meaning Systems," American Psychologist, 17:10-28, 1962. Perez, Patricia. "Semantic Differential Measurement of Attitudinal Changes for Students on Academic Probation." Unpublished Master1s project, School of Education, University of Southern California, 1965. Rankin, Robert D. "A Proposed Course of Study for a Tenth Grade Guidance Class." Unpublished Master’s project, School of Education, University of Southern California, 1964. Richardson, Harold, and Borow, Henry. "Evaluation of a Technique of Group Orientation for Voca tional Counseling, " Educational and Psychological Measurement, 12:587-597, 1952. Rosengarten, William, Jr. "Another Job Course Pays Off," Personnel and Guidance Journal, 41:531-534, 1963. Rosenthal, Oscar A. "A Semantic Differential Investigation of Critical Factors Related to Achievement and Under achievement of High School Students." Unpublished Doctoral dissertation, School of Education, University of Southern California, 1965. Ross, John. "Change in the Use of the Semantic Differential with a Change in Context," Journal of Verbal Learning and Verbal Behavior, 4:148-151, 1965. 112 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. Sinick, Daniel, Gorman, W. E., and Hoppock, Robert. "Research on the Teaching of Occupations 1963- 1964," Personnel and Guidance Journal, 44:591-595, 1966. Stempel, Guido H. "The Relationship of Cost of Instruction and Attitude Toward Instruction," Journal of Educational Research, 57:207-209, 1963. Stone, C. Harold. "Are Vocational Courses Worth Their Salt?" Educational and Psychological Measurement, 8:161-181, 1948. Strem, Bruce E. "An Attitude Survey of High School Dropouts by Means of the Semantic Differential Process." Unpublished Doctoral dissertation, School of Education, University of Southern California, 1966. Super, Donald E. "Group Techniques in the Guidance Program," Educational and Psychological Measure ment, 9:496-510, 1949. Thurstone, L. L., and Chave, E. J. The Measurement of Attitude. Chicago: University of Chicago Press, 1929. Warters, Jane. High School Personnel Work Today. New York: McGraw-Hill Company, Inc., 1956. Webb, Allen P., and Harris, John T. "A Semantic Differential Study of Counselors in an NDEA Institute," Personnel and Guidance Journal, 42:260-263, 1963. __________. "The Webb-Harris Test: A Semantic Differential Approach to Psychodiagnosis with Adolescents." Unpublished monograph, 1961. Willey, Roy De Verl, and Strong, W. Melvin. Group Procedures in Guidance. New York: Harper Brothers, 1957. Wright, E. Wayne. "A Comparison of Individual and Multiple Counseling for Test Interpretation Interviews," Journal of Counseling Psychology, 10:126-134, Summer, 1963. APPENDIXES APPENDIX A INSTRUCTIONS FOR ADMINISTRATION OF THE TEST INSTRUMENT 115 INSTRUCTIONS for ADMINISTRATION The Seim Attitude and Vocational Interest Inventory is one of a number of semantic differential or word meaning tests. The meaning of thB terms is to be determined by the student. No dictionary is necessary. The student reacts to the concept at the top of each card by marking how he feels about such concepts as "School," "Working with Machines and Tools," "Working with Numbers," and "Helping People." There is no neutral position, therefore the student must favor one adjective over another. The third and fourth positions on the scale are rather wBak or uncommitted positions. 5ecure the cooperation of the students by reminding them that there are no right or wrong answers but only the answers that really ex press how they FEEL about the concept at the top of the card. You might want to say something like this to your class. "The name of this class is Guidance. In this class we will discuss how we feel about many school activities and how to plan for what we would like to do after we leave high school. Before the regular work of the class begins, I would like to have you participate in an experiment. You will be given a white envelope and an electrographic pencil. Follow the simple direc tions in the envelope. There is no time limit, but most of you will be through in thirty minutes. Do not spend a lot of time on any answer, but do not hurry so much that you are careless. When you are through, please seal the envelope and return the pencil. Thank you. You may begin." (There is always the temptation to pattern answers without regard to the response needed. You might check on this. Regular test conditions should prevail.) Note: Charles E. Osgood's book, The Measurement of Meaning. (1957), gives the rationale for this technique of indirect measurement. APPENDIX B DIRECTIONS TO STUDENTS DIRECTIONS You are being asked to participate in an experiment conducted through the Univer sity of Southern California. Your answers are confidential, and no one at this school will see them. You are asked to give your name so that a follow-up study can be made. Please write your name, your school, and your room number on the blank card. There are seventeen numbered cards. Each numbered card has one or more words at the top in large letters. React to these words by making a solid black mark in one of the long circles between each pair of dBBcriptive words beneath. Indicate your feelings by mark ing near to the term which comes closest to expressing how you feel. For example, if the words were 1 1 Ice Creon," you might mark: HOT () () () {) () 0 COLO (You feel it's very cold.) SWEET () 0 () () () () BITTER (You feel it's rather sweet.) If you have great difficulty in making up your mind, mark in the space near the middle but on the side which is closer to how you feel. In 'a way the spaces say: 0 0 0 0 0 0 agree agree agree agree agree agree strongly with with with with strongly with word left-side left-side right-side right-side with word on left word word more word more word on right side fairly than right- than left fairly side strongly side word side word strongly There will be 12 marks on each card. Work rapidly. Do not spend a lot of time on any answer, but do not answer carelessly. Thank you* APPENDIX C THE TEST INSTRUMENT: NAME CARD S tu d e n t’s N am e Room N o. Period School Seim Attitude and Vocational Interest Inventory D a te 120 APPENDIX D THE TEST INSTRUMENT: NUMERICAL LISTING OF CONCEPTS 121 THE TEST INSTRUMENT: THE SEVENTEEN-CARD SAVII Concept Concept Printed at Top of Tabulating Card Number 1. School 2. Discovering New Facts and Solving Problems 3. Working with Machines and Tools 4. Boys at This School 5. Going to Concerts, Playing Instruments, Singing 6. Me 7. Doing Creative Work with Your Hands 8. Work 9. Working with Numbers 10. Meeting and Dealing with People to Promote Projects or Things to Sell 11. Helping People 12. Working Outdoors 13. Girls at This School 14. Reading and Writing 15. Teachers 16. What I Would Like To Be 17. Doing Office Work with Precision and Accuracy 122 APPENDIX E THE TEST INSTRUMENT: SAMPLE TABULATING CARD r 123 SAMPLE TABULATING CARD (The seventeen cards were identical except for the number key-punched into the card and the wording of the individual concept.) 1. School ““ 0 0 0 0 0 0 cRuEL 0 0 0 0 0 0 s o f t PASSIVE 0 0 0 0 0 0 AC™ £ STRONG 0 0 0 0 0 0 "EAK “ lu 0 0 0 0 0 0 excitable ™ e 0 0 0 0 0 0 false l i g h t 0 0 0 0 0 0 h e a w »ise 0 0 0 0 0 0 F00uSH MOVING 0 0 0 0 0 0 STILL u g l y 0 0 0 0 0 0 b e a u t i f u l 1 suall 0 0 0 0 0 0 large rv ru fast 0 0 0 0 0 0 SL0W APPENDIX F DEMOGRAPHIC DATA COLLECTION SHEET 125 DEMOGRAPHIC DATA COLLECTION SHEET (Numbers Follow Cumulative Record) 1. Student's name _____ Birthdate , — Sex Masc. Fem. File § School Cl as 8 Period 2. Marks: h i BIO English S.S. Math. For. Lang. Prac. Arts Fine Arts P. E. Scholarship society yes no Perfect attendance yes no Grad, with diploma yes no 5. Major sequence 13. Test information: Date __________ Age __________ 9. Parent information: Both parents at home yes no Parent deceased mother father Parents separated yes no Pupil lives with step-parent yes no Pupil lives with guardian (not parent) yes no Verbal N. Verbal Total Stanine R. Voc. R. Comp. Arith.R. Arith. F. L. Mech. L. Spelling _ Kuder scores: V Outdoor 1. Mechanical 2. Computational 3. Scientific 4. Persuasive 5. Artistic 6. Literary 7. Musical B. Soc. Serv. 9. Clerical 11. Family: Language spoken in home Siblings: Older (give ages) Younger (give ages) 16. Significant data: 126 APPENDIX G ETHNIC SURVEY OP SCHOOLS INCLUDED IN INVESTIGATION 127 ETHNIC SURVEY OF SCHOOLS INCLUDED IN INVESTIGATION School I School II School III White Spanish surnames 615 (2l)* 48 (2) 1076 (30) Other whites 1142 3491 953 Non-white Negro 3 (l) 1 (0) 659 (14) Oriental 42 (l) 12 (0) 58 (l) American Indian 5 0 8 Other non-whites 13 1 11 Total 1820 3553 2765 •Numbers in parentheses designate the sample of the population tested by the 5AVII. An inspection of students' cumulative folders pro vided the only opportunity to collect the above data for tabulation. The basis for the regular population figure was visual inspection by classroom teachers. The ethnic composition of pupils by grades was not available. The percentages for pupils in all junior and senior high schools in the district at the time of testing were: Spanish surnames 16.7 Other whites 60.8 Negro 18.5 Oriental 3.6 American Indian .1 Other non-whites .3 100.0 APPENDIX H DESCRIPTION OF SAMPLES LOST 129 DESCRIPTION OF SAMPLES LOST Classes 5chools Stanines Sex Classes by Grades 01-9 1-33 1 - 1 Boys 52 B10 - 68 02-6 03-5 II - 17 2 - 3 3 - 7 Girls 29 A10 - 5 04 - 13 * III - 31 4-10 Bll - 6 05-4 06-7 5-10 6 - 7 All - 0 07-3 7 - 0 B12 - 2 08-3 09-8 8 - 0 9 - 1 A12 - 0 10 - 13 11-5 No test data - 34 12-5 61 81 01 81 01 Treatment: Experimental 60 Control 21 61 "Program adjustment made by administration resulted in the lass of these subjects. Samples lost through students' moving . . 6 Samples lost— absence on second testing or program changes......................68 Samples lost— tests not usable. . . . . 7 81

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Learning Differences Between High And Low Auding Subjects

PDF

Parental Support Of Independent Free Schools. Determinant And Value Implications For Public School Counselors: An Exploratory Study

PDF

An Investigation Of Selected Correlates Of Academic Achievement Among Seventh-Grade Pupils

PDF

Measures Of Self-Actualization Changes And Their Relationship To Interaction Preferences Among Encounter Group Participants

PDF

An Investigation Of The Relationship Between Counselor Role-Function And Counselee Perception Of Help Received

PDF

An Investigation Of Leader And Member Helpfulness In Group Counseling

PDF

A Philosophical Model For Counseling Systems

PDF

The Relationship Of Self-Concept Discrepancies To Vocational Choice, Intelligence, School Achievement, And Socioeconomic Status

PDF

An Analysis Of The Role Of The School Psychologist In The State Of California

PDF

An Analysis Of The Position Of The Elementary School Guidance Counselor

PDF

An Evaluation Of Hutt Adaptation Of The Bender-Gestalt Test

PDF

A Video-Modeling Approach To Group Counseling

PDF

Relationships Between Counselors' Philosophical Consistency And Congruency With Their Effectiveness As Counselors

PDF

Experiential Group Counseling With Intern Teachers

PDF

Inservice Education Programs For High School Counselors In California With Recommendations For Improvement

PDF

A Comparative Study Of The Marathon-Group Experience In A University Counseling Center

PDF

A Methodological Investigation Of Affect Response Bias

PDF

Nonverbal Communication In Counseling: An Exploratory Study

PDF

A Comparative Study Of Divergent Thinking Ability And Academic Achievement Of Students In The Honors Program At The University Of Southern California

PDF

Effects Of Short-Term Group Counseling On Changes In Attitudes Of Flexibility, Tolerance, And Nonauthoritarianism

Asset Metadata

Creator Seim, Robert Martin (author)

Core Title A Measurement Of Attitudinal Change Following A Ten-Week Guidance Class: A Semantic Differential Study

Degree Doctor of Education

Degree Program Education

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, guidance and counseling,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Carnes, Earl F. (committee chair), Michael, William B. (committee member), Wagner, Elmer E. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-639427

Unique identifier UC11362212

Identifier 6900644.pdf (filename),usctheses-c18-639427 (legacy record id)

Legacy Identifier 6900644.pdf

Dmrecord 639427

Document Type Dissertation

Rights Seim, Robert Martin

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA