AN INVESTIGATION OF FACTORS LEADING TO THE WITHDRAWAL OF WALDORF JUNIOR COLLEGE FRESHMEN Times-Is. Ect- Hm Degree of DE. D. MICSIQAN STATE EEEVEESETY Clayton Durward Rea 1965 ABSTRACT AN INVESTIGATION OF FACTORS LEADING TO THE WITHDRAWAL OF WALDORF JUNIOR COLLEGE FRESIHvL‘EN by Clayton Durward Roa The purpose of the study was to describe and attempt identification early in the school year of waldorf Junior College freshmen who withdrew from the Class of 1963-1964 and who were judged by the faculty to have acute intrapersonal concerns. The pOpulation consisted of 252 students of which 1&4 were male and 108 female. Of these students, 82 (32%) did not return to Waldorf College for their SOphomore year. Of these 252 students, 39 (15) were judged by the faculty to have acute intrapersonal concerns which might jeOpard- ize academic success. A questionnaire devised for the purpose of this study and three standardized instruments were employed to gather data. A test of academic ability, the ACT, was provided by the college. The Minnesota yultiphasic fiersonality Inventory and the Lutheran Youth Research Inventory were the two psychometric instruments employed. A research design was develOped which consisted of two parts: 1. Description and eXploration of the subjects of interest by the use of the chi square statistic, accumulative contingency tables, and product-moment correlations. 2. Identification of the subjects of interest by the use of multiple regression equations, based on variables selected from factor analytic techniques, followed by cross—validation. The regression equations were employed after testing the hypotheses that the Clayton Durward Roa differences in the multiple R for the groups of interest were statis- tically significant. The major findings may be summarized as follows: 1. The primary period of withdrawal is after two semesters at Waldorf College. The primary reason for withdrawal is transfer to another educational institution. A secondary reason is various dissatisfactions. School mobility before college matriculation is most in evidence for female drOpouts and males with acute concerns. Students in each group accept full responsibility for the decision to leave Waldorf College. The person most influential in encouraging males to attend Waldorf College is the parish pastor. For females, parents are the primary influence. The questionnaire does not adequately discriminate items that are interpersonal and affective in nature. The ACT English test is the most useful of the five ACT tests for the purposes of this study. Five LYRI and three MMPI scales when used in the four derived multiple regression equations are the most discriminating scales from among the 36 scales selected from the factor analysis. In a cross-validated random sample the multiple regression equation derived from the first half of the dichotomized groups correctly predicts 81% of the dr0pouts and 60% of the acute 10. Clayton Durward Roa concerns. Students who elected not to release LYRI scores for research purposes are consistently higher on each of the 17 MMPI scales selected for this study. The difference in mean in each case is significant at the .01 level. AH INVESTIGATION OF FACTORS LEADING TO THE HITHDRAWAL OF WALDORF JUNIOR COLLEGE FRESHMSN BY Clayton Durward Roa A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Counseling, Personnel Services. and Educational Psychology 1965 PREFACE "Wisdom is the principal thing; therefore get wisdom: and with all thy getting get understanding." --A Proverb of Solomon ii ACKNOWLEDGMENTS The investigator expresses his appreciation to his committee chair- man and research director, Dr. John E. Jordan, and to the members of his committee. Dre. Norman Abeles. James W. Costar and Walter F. Johnson,for their skillful guidance and encouragement. Acknowledgment is also made to Dr. Carl Frost for serving on the committee before accepting a position at the University of Nigeria and to Dr. Charles A. Blackman for critical review of the thesis. For motivation to study factors leading to the withdrawal of Waldorf College freshmen the investigator is indebted to students who for six years shared their concerns with him as their chaplain. To the faculty and staff of Waldorf College the investigator is deeply indebted. For competent statistical consultation Dr. John Felty has earned the investigator's sincere gratitude. For his patient introduction to computerization.nr. Bruce Rogers deserves more than words of appreciation. Dr. Merton Strommen, developer of the Lutheren‘fouth Research Inventory, supplied encouragement and consultative competence repeatedly. A Ph.D. is known as an earned degree. The investigator's wife. Eleanor, deserves a related degree. Competent and affectionate manage- ment of a household of four children is a sufficient calling. She willingly added the tedious work of preparing this typeecript. Separa- tion from the children for five weeks became part of the sacrifice. To Darel, Paul, Janet and Mark and hospitable relatives the investigator's gratitude in ink feels highly inadequate. iii TABLE OF CONTENTS Page Pmm O O O O O O O C O O O 0 ii ACKNOWLEDGMENTS . . . . . . . . . . iii LIST OF TABLES. . . . . . . . . . . Vii LIST or ILLUSTRATIONS . . . . . ' . . . . ix LIST OF kaDICES. O O O O O O O O O X Chapter I. TEE PROPJLET‘I e e e e e e e e e 1 Introduction . . . . . . . . 1 Statement of the Problem . . . . . 3 Need for the Study. . . . . . . 4 Distinctive Role of the Junior College 4 Student Personnel Practices . . . 5 s Attrition Rates . . . . . . 9 Definition of Terms . . . . . . 10 Limitations of the Study . . . . . ll Assumptions . . . . . . . ll Generalizability . . . . . . 12 Organization of the Thesis. . . . . 13 II. REVIEW'OF RELATED LITERATURE . . . . . 14 Psychometric Theory and Practice . . . lb Historical Development . . . . 1b Measurement of Personality . . . 16 Attrition Studies . . . . . . . 17 College Freshmen . . . . . . 18 MMPI Related Studies . . . . . 20 College Screening Use of the MMPI 20 Achievement Studies with the MMPI 23 Studies on Deviant Behavior . . 28 Incorporating Conflicting Results 30 Lutheran Youth Research Studies . . . 32 Faculty Rating Studies. . . . . . 32 Summary of Related Research . . . . 34 iv Page III. DESIGN AND PROCEDURES . . . . . . 35 Introduction . . . . . . . . 35 The Sample . . . . . . . . 35 Design . . . 36 Criterion (Dependent) Variables . . 36 Dr'Opouts. . . . . . . 35 Acute Concerns . . . . . 36 Predictor (Independent) Variables 37 MMPI. . . . . . . . 37 LYRI. . . . . . . . 4O Demographic Data . . . . . . 44 Administration Procedures. . . . . 45 Scoring and Coding . . . . . . 45 Hypotheses to be Tested . . . . . 46 Analysis Procedures . . . . . . 46 MMPI and LYRI . . . . 46 Questionnaire and Demographic Data . 49 Summary of the Design and Procedures . . 49 IV. ANALYSIS OF THE DATA . . . . . . 50 Introduction . . . . . . . . 50 Subjects . . . . . . . . . 50 Questionnaire . . . . . . 5O MMPI - LYRI. . . . . 53 Analysis of the Questionnaire and SAT Tests 54 Analysis and Intensity of Feelings . 54 I’Iithdrawal Timing, Response, Current Activity. . . . . . 9 Pro-college Residence and Schools . 61 Parental Socio-economic Items . . 63 Matriculation Influences. . . . 64 Health and Study Skills . . . . 67 Class and Chapel Attendance . . . 69 Students' Perceptions of Parents. . 69 Students' Perceptions of Home Congregation 73 Campus Friendships . . . 76 Feelings in Relation to Classmates . 77 Feelings in Relation to Closest Friends 79 Estimate of Classmates' Feelings. . 81 Estimate of Closest Friends' Feelings 83 SAT Tests . . . . . . . 85 Analysis of MMPI and LYRI Results Factor Analysis . . Multiple Regression . Hypotheses Testing . Cross-Validation . . SCOPE Procedure . . LYRI Release and Nonrelease Samples Compared on the MMPI . . . Mean Differences between Criterion Groups and Normal Groups . . Summary of the Analysis of the Data . V. SUMMARY, CONCLUSIONS AND IMPLICATIONS . REFERENCES. APPENDICES. Summary of the Study . . . . Conclusions . Implications . . . Summary of the Investigation . Demographic and Attitudinal Data ACT Relationships. . . . PDTPI "' URI 0 e e e LYRI Non-release Sample . . . Differences Between Criterion Groups and Nondropouts-nonacute Concerns General Conclusions . . Problems for Further Study 101 102 102 103 103 105 105 107 107 107 108 110 112 114 122 Table 3.1 3.2 3.3 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 LIST OF TABLES Students with acute concerns identified by the faculty . . . . . . . Characteristics of selected MMPI scales . . Characteristics of selected LYRI scales . . Responses to instruments employed . . . Analysis of feelings about students, faculty, administration, regents, alumni, and the American Lutheran Church . . . . . Intensity of feelings about students, faculty, administration, regents, alumni, and the American Lutheran Church . . . . . Time of withdrawal from Waldorf College . . Locus of withdrawal decision. . . . . Analysis of current activity. . . . . Residence before matriculation at Waldorf College . . . . . . . . . Residential change of parental family during freshman year . . . . . . Different schools attended before matricu- lation at waldorf. . . . . . . Evaluation of parental social class . . . Family income, in relation to income of friends Locus of decision to matriculate at waldorf . Waldorf rated as first choice of a college . Primary2matriculation influences . . . General health during freshman year . . . Reading ability at the beginning of the freshman year. . . . . . . . vii 55 57 6O 6O 61 62 62 63 64 65 65 66 67 68 Table 4.17 4.18 4.19 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28 “.29 4.30 4.31 4.33 Deve10pmenta1 (remedial) study during freshman year . . . . . . . . Class and chapel attendance during the freshman year . . . . . . . . Parents as perceived by the student . . . Home congregation as perceived by the student Number of campus friends . . . . . . Roommate as closest campus friend. . . . Feelings in relation to classmates . . . Feelings in relation to closest campus friends Estimate of classmate's feelings about himself Estimate of closest friends' feelings about hms elf . O O O O O O O . Analysis of SAT scores for dr0pouts and acute concerns. . . . . . . . Factor analysis of MMPI-LYRI . . . . . Most useful R scales . . . . . . . Iultiple regression equations selected for cross-validation . . . . . . . Mean, standard deviation and correlation of the real I and the predicted Y for randomly dichotomized samples . . . . . . Analysis of correct and incorrect statistical identification of dropouts and acute concerns Multiple regression coefficients derived by SCOPE R computer program . . . . . viii Page 68 7o 71 74 76 76 77 79 81 LIST OF ILLUSTRATIONS Figure Page 1 Difference between mean of LYRI release and nonrelease samples . . . . . 97 2 Mean difference between nondeviate, drOpout, and acute concerns . . . . . . 99 LIST OF APPENDICES Appendix Page A DrOpout Data . . . . . . . 123 B Demographic and Socio-Economic Data. . 124 C Comparative MMPI Data . . . . . 128 D Questionnaire . . . . . . . 130 E Mean and Standard Deviation of Questionnaire Items . . . . . 137 F Code Book . . . . . . . . 142 G Comparison of MMPI Mean and Standard Deviation between LYRI Release and Nonrelease Responses . . . . 156 H Cross-validation of Variables Selected on the Basis of Repeated Occurrence in Twenty-four E eriments (A Clinical Huncig . . . . . 158 CHAPTER I THE PROBLEM INTRODUCTION On July 12, 1963, the late President John F. Kennedy allocated $250,000 from his emergency funds to help 63 selected school districts bring potential or actual drOpouts back to school. This presidential action stimulated many other communities throughout the nation to intensify their efforts to keep students in school. In March, 1965, President Lyndon B. Johnson declared in an address to delegates at a conference sponsored by the National Education Associa- tion: "You can sit in your rocking chair; but I am going to use every rostrum to tell the peOple that we can no longer afford the great waste that comes from the neglect of a single child" (Time, (arch 12, 1965, p. 22). Both national and local leadership has been aroused to concerted action in relation to the problem of student attrition. Evidence of increased interest is reflected both in papular and professional public- ations. Articles on student withdrawal reported in The Readers' Guide. to Periodical Literature indicate a striking increase in the sixties both of articulated concern and deliberated action. A comprehensive study by Iffert for the Office of Education showed that only about 40% of all college entrants graduate in four years from the institution they originally entered. This prOportion increased to 60% for those who eventually graduated from some college (McQueen, 1964). While abundant evidence exists for much contemporary concern with student attrition, earlier awareness of the problem is also of interest. 1 The President of the University of Minnesota reported to his Regents in 1900 that failure of students originated in laziness, bad habits, or lack of ability. In 1947 he reported: "In the judgment of the Adminis- trative Board, failures are the result, not of the lack of mental ability, but of emotional and personal maladjustments. We are trying to help these individual cases...(Torrance, 1963). The problem of student withdrawal in relation to the emotional needs of college students impinged acutely on the awareness of this investigator in his role as a college chaplain. It appeared necessary at Waldorf College to find economical ways to try to identify early in the school year the students with acute intrapersonal concerns so that apprOpriate prOphylactic assistance might be proferred before withdrawal became actualized. Clinical observations indicated that intrapersonal concerns tended to jeopardize academic success for some students. A positive relationship seemed to exist between such students and those who became drOpouts. Such students crowded the appointment calendars of personnel workers at waldorf College. The rapid increase in student enrollments has contributed to the problem nationally. The total national increase among college freshmen was 20% for 1964-1965. Steady increases are projected through 1970 (Parker, 1965). Students polled students at the University of Toronto reporting that 50% of the men and 64% of the women had worries for which they would like some kind of professional help (King, 1964). A middle western college indicated successful identification early in the school year of students with acute problems (Woolf and Woolf, 1953). Some studies report that psychiatric disorders among college students are increasing at an alarming rate: In those cases where drOpping out is a symptom of some kind of emotional conflict, the associated psychological conditions are varied. Apathy, unconscious rebellion, vacillation, indecision. and psychosis are often present. we still do not understand the psychodynamics involved in many cases of drOpout, even though we can classify some of the accompanying psychological states. Obviously continuing research is of prime importance (King. 1964, p. 329). anagram OF THE PRCBIEM The primary purpose of this study is the description and explora- tion of factors that relate to Waldorf (Junior) College freshmen who withdraw after the freshman year and who have acute intrapersonal con— cerns. Factors investigated are selected from standard admission data and a questionnaire deve10ped for this purpose. The second purpose of this study is the testing of two major hypotheses: 1. Selected scales from the Minnesota Multiphasic Personality Inventogy (MMPI) and the Lutheran Youth Research Inventory (LYRI) can be used to predict waldorf College freshmen who withdraw after the freshman year. Selected scales from the ninnesota Multiphasic Eersonality Inventogy (MMPI) and the Lutheran Youth Research Inventogy (LYRI) can be used to predict waldorf Junior College freshmen N 0 who have acute intrapersonal concerns, as measured by faculty judgment. NEED FOR THE STUDY The Distinctive Role of the Junior College The rapid eXpansion of junior colleges in the United States calls attention to their unique role in higher education. Students can be accepted who are less able both intellectually and financially. The urgency of meeting student needs early is reinforced by an awareness that maximum matriculation is two years, not four. Many students in junior colleges attend but one year. A possibility of accelerated maturation exists when SOphomores must assume the leadership roles re- served usually for upperclassmen in the four year colleges. These factors help to operationalize the American dream of equality of educational Opportunity. One of the distinctive roles that can be played by the relatively smaller independent colleges is the cultivation of personal rela- tionships and enrichment of interests among students and between them and their teachers. In times when numbers appear to attract public attention and the god of bigness sits high among the gods of education, it may not be entirely out of order to recall the nostalgic idea of the split log. we must beware lest the log be so long that neither end can be heard (Bogue, 1950, pp. 109-114). Tead (1954) looks to the junior college to help the student inter- relate his academic courses with his total outlook and interests, in- cluding the vocational. This enhances the character of the learning process making it more realistic and vital than is often true in liberal arts colleges that divorce themselves from vocational concerns. Less reliance on verbal glibness, a less narrowly intellectual emphasis, a deliberate effort to enrich the emotional life, a greater eagerness to extend the student's guided experience of learning beyond the c1assroom--these are among the touchstones of good junior college teaching (Ordway Tead, 1954, p. 124). It appears that the junior college needs to capitalize on its unique potential for individualizing instruction. It seems necessary to explore further the place of skilled counseling and apprOpriate diagnostic methods. Student Persgnnel Ezagtiges Medeker (1960, p. 25) observed that since the beginnings of the junior college movement it has been recognized that students need professional assistance in making sound educational and vocational choices. Before becoming an authority on junior colleges, Medsker assumed that rendering individualized personnel services would be a strong point with twoeyear schools. Subsequent study forced him to conclude that this is not the case consistently. Another leader in the junior college movement stated: If the community college recognizes the fact of individual differences, ... this will demand an organized and well-financed program of personnel services. It is difficult to believe that any professional educator can fail to make this admission of fact; it is more difficult to understand why neglect should be allowed when the fact is admitted. The task is one that is time consuming, because prOper counseling and guidance, like diagnostic practices by the physician, cannot be done en masse. It is an individualized, personalized undertaking (Bogue, 1950, p. 321). The degree of responsibility each junior college accepts in stimu- lating the total growth of each student needs clarification. An aware- ness that approximately six out of ten students fail to continue in college raises questions about present policies. McDaniel (1962) identified essential personnel practices for the Commission on Student Personnel of the American Association of Junior Colleges. He noted that the purposes of the college need to be Operationalized with C\ consistency. He concluded that systematic evaluation of all policies is the only protection against mistakes. A survey of 76 junior colleges found that administrations and boards of control had not placed student personnel services on a par with instruction and that insufficient budget appropriations had been made to ensure a full range of such services. The counseling program in many of these institutions was inadequate. An insufficient number of counselors with professional preparation was observed. Little research to obtain facts about their students was conducted and only limited effort to evaluate their programs was expended. If the junior college is to provide the first post-high ex- perience . . . to channel students into avenues consistent with their characteristics and likelihood of success, much will depend on the quality of the counseling program. The failure of the junior college in this important responsibility could well mean that the growing dependence on this institution is in error (Medsker, 1960, p. 163). A survey of guidance practices in selected junior colleges revealed the following characteristics: personal-social guidance had not been develOped extensively; lack of trained counselors is a major weakness; extensive use of untrained counselors enlisted from the faculty seems customary; little, if any, research or evaluation was evident (Starr, 1950. p. 489). An important part of the personnel program should be research to determine characteristics of entering students, the quality and duration of their college experience, and what they do after leaving the college. In spite of the value of such data, only a small amount of it is avail- able. Bogue (1950, p. 326) felt that adequate personnel services are more than a question of efficiency; it is a moral issue. Clear ~o recognition of the student as a person, the phi1030phy of individual differences, the claims of personalized education, and the reSponsibility of the college to guide students into fruitful vocations are essential. Otherwise, he college must assume the risk of society's harsh judgment of moral turpitude. He anticipated that lack of money and understanding of need would be the retort of some administrators. It is not preposed that the college assimilate the function of the therapeutic clinic or hospital. It is prOposed that colleges can meet more of the emotional needs of the students in responsible ways within an educational milieu. The counselor in an educational institution can help students become relatively free from distractions that inter- fere with learning. A broad Spectrum of possible helpfulness exists for sensitized personnel services. Referral to psychological resources outside of the college ought to be the exception rather than the rule. Academic failure does not generally seem to be associated with classical neuroticism, but rather with a failure in social . orientation (Gibbs, 1965, p. 580). Of the almost 600 cases seen by Harvard's psychiatric service in one academic year, about one-fourth were diagnosed as neuroses and one- fourth were given other psychiatric labels (mostly depression and schizoid states). The remainder of the group were given "Problem Diagnoses" after a system devised for this purpose by Konks and Heath for labeling cases in which there seemed to be no identifiable psychia- tric ailment (Elaine and McArthur, 1961, pp. 62-63). The skilled college counselor can assist in meeting the needs of this later group, the majority of students who seldom approach a psychiatric center. The consulting psychiatrist for a large liberal CO arts college in Minnesota reported that the addition of a well-trained counselor to the personnel services staff reduced his "nuisance load" considerably. This permitted psychiatric services in depth for the more acute cases.1 The Dean of the College of Education, Michigan State University, has observed that the major thrust of our educational system currently deals with so-called "rational faculties" of our pOpulation. we specialize in employing facts and logical systems of reasoning. We do a less effective job of dealing with the emotional mechanisms that really control human behavior and help students learn how to discipline these mechanisms effectively in individual and group behavior. I realize that you can't divide these matters into rational and emotional dichotomies, but, at the same time, if we could learn more about the psychiatric, psychological, sociological and cul- tural anthropological variables that create a predisposition to behavior emotionally in certain directions, I think we would be much more effective in our educational programs. Johnson (1965) has summarized a rationale for junior college per- sonnel programs. He found it difficult to defend the admission of all comers unless counseling adapted to their needs is provided. To fail in this is to develop a "revolving-door" college. By design the junior college is an open-door college. Johnson concluded that the junior college has an obligation to help students achieve a self-understanding that enhances the making of educational-vocational choices. 1 Personal conversation with the investigator. 2 Personal communication from Dr. John E. Ivey, Jr., July 7, 1964. \O Considered by some to be the "dean" of contemporary college student personnel workers, Hilliamson (1949) has prOposed that in our American democracy the student's full and balanced maturity is the major goal of education. This involves the total development of the student--socially, emotionally, intellectually, spiritually, and psychologically. Both the quality of education and the quantity of students con- tribute to the need for further research. Merlo (1964) has observed that one out of every five college students in the United States now attends a junior college. California, the leader in the field, now enrolls 8h of every 100 freshmen and SOphomores in the state in public twoayear colleges. About 25 new two-year colleges are Opening each fall, with the rate of growth expected to increase. William G. Dwyer, president of the Massachusetts Board of Regional Community Colleges, summarized: "On a university campus, a student sinks or swims but on a community college campus, we try to teach him how to swim (Time, March 5, 1965, p. 60). Ail:iii22.fiaisa. The need for study of college student emotionality has been docu- mented by the unique goals of junior college education and the inade- quate way student personnel programs have been meeting these goals. Further evidence for need will be derived from interpretation of data on withdrawals. theer (1937) studied entering freshmen at 25 universities in 1931-1932. Follow-up studies indicated that 62% left the university without a degree. or these. 17% transferred to other institutions or lO later returned to complete their degrees. The highest dr0pout rate occurred during the freshman year, exactly one—third of the total. Twenty—five years later Iffert (1958) completed a study with re- sults that are consistent with those of McNealey. Only 40% of college entrants graduated from the institution they entered. This prOportion increased to 60$ for those who eventually graduated from some college. Medsker (1964, p. 53) reported that a California survey showed a 66% loss between the first and second year in colleges in that state. Approximately one-third of entering junior college students in the nation graduate in two years. The emotional anguish eXperienced by students and parents, when added to the vast investment in "failure" by the colleges in relation to the 40% who never complete what they began, underscores the need of research on student attrition. These gross statistics cannot convey causality. Major causes of waldorf College student withdrawal are reported in Appendix A. DEFINITION OF TERMS Academic success: At waldorf College students who fail to earn a grade point average of 1.5 or above on a 4.0 scale by the end of their first semester in attendance are not permitted to continue. Excluding the first semester, any student whose grade point average falls below 1.7 is not permitted to register for the following semester. Counseling: Counseling is defined as a dynamic face-to-face relationship designed to assist persons in making significant choices. Faculty judgment: Working independently, each faculty member at Waldorf College identified students with whom they had personal ll interaction, i.e., a first-hand face-to-face encounter, not merely hearsay. The location was immaterial: classroom, office, home, or casual contact that was informal and unstructured. Intrapersonal concerns: This term was broadly defined to suggest the full range of attitudes variously labeled emotional, mental, social, psychological, or spiritual. Junior college: For the purposes of this study the following terms are used interchangeably--junior college, community college, two-year college. Standard admissions data: After considerable study the investigator concluded that much of the biographical, familial, medical, and appraisal information available in each student's admissions file was not sufficiently precise for correlational purposes. Data used is limited to American College Testing scores (hereafter identified as ACT). Withdrawal: A student who matriculated at Waldorf College as a freshman in Se tember, 196 and who did not enroll as a soohomore in o & 196A is considered a withdrawal or a dropout (interchangeable tenns). LIMITATIONS OF THE STUDY Assumptions Scientific methodology is not devoid of presuppositions. An aware- ness of the assumptions that a researcher makes tends to minimize some of the subjectivity that can distort the findings. Three assumptions are utilized in the following discussion. The use of measurement instruments in their present stage of refine- ment to assess human behavior and personality necessitates caution. It is assumed that the instruments selected for use in this study have limitations. Their use does not imply a lack of awareness that behavior is complex. Specific personality variables need to be studied in rela- tion to detailed background and behavioral data. The use of approximately thirty faculty members on a highly homo- geneous campus to make independent judgments about student emotionality may afford questionable validity. A subsequent discussion under the heading of "faculty rating" indicates attempts to be aware of this hazard. A similar question could be asked regarding variation in the understanding of items on the questionnaire. A final assumption relates to the value of a self-study by a college faculty. Learnings may be obtained by observation of students who withdraw that may be vital to the enrichment of retained students. At the conclusion of their extensive study on adolescent personality and behavior using the MMPI, Hathaway and Monachesi (1961, p. 150) observed: "Possibly the clearest data emerging from our complex of tables are evidences predicting normal adjustment. The average lives turn out to be more solidly predictable than are deviant lives." Generalizability The sample under investigation is the freshman class of one mid- western junior college. Further restriction on generalization to the total junior college pOpulation in America is the denominational nature of this independent school. It is possible that the 272 independent church-related junior colleges in America would find these results more useful than the #22 public junior colleges. If only the size of the 13 college under study is considered, additional generalization might be possible. Of the 272 independent junior colleges, Waldorf College ranks 58th in size of student body. ORGAIIZATION OF THE THESIS Chapter I has stated the problem, the purposes, and the need for the study. The need was resident in the distinctive purposes of the student personnel program of the junior college in relation to contem- porary attrition rates and reasons for withdrawal. Definition of terms and limitations of the study were followed by an outline of the thesis. Chapter II is devoted to a review of the literature related to the problem under study. Attention is given to research dealing with the historical deve10pment of personality assessment. Concentration is on studies of attrition among college freshmen. Chapter III contains the design and procedures. The instrumenta- tion used for gathering data is described and the procedures for statistical analysis are outlined. Chapter IV includes the analysis of the data and a discussion of the significant findings. Chapter V contains a summary of the study, conclusions reached, and implications for further research. CHAPTER II REVIEW OF RELATED RESEARCH PSYCHOMETRIC THEORY AND PRACTICE Historical_pevelopment As recently as one century ago psychology was still intertwined with philOSOphy. An arm-chair, non-experimental discussion of the nature of the mind did not involve measurement until the beginning of this century. Ties with philosophy yielded to new affiliations with the biological sciences. Thorndike and Hagen (1961) identify three merging contributions to measurement in psychology and education: the nineteenth century physiological and experimental psychology originating in Germany, Darwinian biology, and clinical concern for maladjusted persons. Hilhelm Wundt at Leipzig in 1879 has been credited with the first eXperimental laboratory for psychology. An outgrowth of experimental psychology has been precision of techniques, rigorous statistical designs, and an appreciation of the need to operationalize concepts. Darwin's emphasis on individual differences was applied to human behavior by Sir Francis Calton, an English biologist. Anastasi (1961, p. 7) credits Calton with being "primarily responsible for launching the testing movement on its course." Calton recognized the need to measure the characteristics of persons, related and unrelated. He pioneered in applying rating scales and questionnaires as well as statistical methods to the analysis of individual differences. One of his students, Karl Pearson, further refined these methods. 14 15 Reinforced by contact with Hundt and Calton, James Cattell, an American psychologist, merged experimental psychology and the new test- ing movement. Cattell first used the term, mental test. A student of Cattell, E. L. Thorndike, became a primary influence in the deve10pment of standardized educational tests. Clinical interest in the individual was especially strong in France. Binet experimented with direct, though crude, ways to measure complex intellectual functions. Lewis Terman in 1916 built on Binet's work to produce the influential §tanford-Binet version. Psychological measurement in America in this century appears to divide into four stages. Up to 1915 initial development of methods was explored. werld war I stimulated the need for measurement advances. Rapid classification of many recruits prompted the development of the Army Alpha and the Army Beta group tests. Standardized tests became available for many school skills and content areas. In this wave of enthusiasm for testing, uncritical use led to reactions regarding the whole philoSOphy of quantification and the use of numbers to express psychological qualities. Critical appraisal followed from about 1930 to l9h5. Measurement of a limited range of skills yielded to evaluation of the whole range of educational goals. Global projective methods of personality assess- ment came into focus. Thorndike and Hagen (1961) predict that the ;present period, since 19u5. will be characterized by test batteries and testing programs. The successful use of integrated aptitude batteries during Xforld I'Iar II and the large-scale testing programs such as the Collxege Entrance Examination Board are evidences offered. Measurement of Personality Still in its infancy is the psychological measurement of the non- intellectual varieties of human behavior such as emotional adjustment, motivation, attitudes, interests, and social relationships. Kraepelin (1895) used a free association test with abnormal patients. Another prototype was developed during Vorld War I, a self-report inventory known as the Personal Data Sheet. Hartshorne and May (1930) approached personality measurement through the application of performance or situational tests. A more recent approach has been the use among clinicians of projective techniques. Anastasi (1961, p. 18) has cautioned, "All the available types of personality test present serious difficulties, both practical and theoretical." Special difficulties in personality testing include these three: one, the possibility of faking and nalingering affects test reliability; two, specificity of responses is required yet behavior varies with time and place; three, adequate criterion data for establishing validity is quite limited. The present practice favors correlations with other tests (concurrent validity) and internal consistency (content validity). Solutions to these problems include these two possibilities: one, continued exploration in the improvement of personality inven- tories; two, recognizing for the present the crude stage of contemporary development and restrict applications accordingly. The personality theogy favored by each test constructor affects the amount of experimental verification to which the test is subjected. Psychoanalytic theory, for example, does not readily admit to testing. The principle of indeterminancy provides a serious problem for the 17 phenomenological theorist. Among the personality theories that have stimulated test development rather widely is the manifest need system of H. A. Murray. The Edwards Personal Preference Schedule is the most comprehensive inventory designed to assess the strength of such needs. This survey of the historical development of measurement and discussion of personality assessment is an encouragement to be cautious in drawing conclusions from the data gathered by instruments used in this stud . The vast variations in personality theory indicate an additional reason for accepting tentatively the conclusions drawn under any one system. It is especially important that interpretations be confined to the Operationally defined terms of the criteria against which validity was established. These cautions introduce a review of literature applicable to this study. ATTRITION STUDIES Guilford (195a, p. 3&1) stated that "no other contribution of psychology has had the social impact equal to that created by the psychological test. No other technique and no other body of theory in psychology has been so fully rationalized from the mathematical point of view." The discussion that follows partially documents Guilford's rather strong statement. A veteran dean of personnel services, E. G. Williamson (196%), has noted that Meehl established empirically that certain kinds of personality description and diagnoses are as accurately made by means of standard and generalized interpretations of personality test profiles as by interpretations of clinical psychologists. Sarbin (1960) earlier 18 reported a similar finding. Predictions of grades using a battery of tests and high school grades were compared with predictions made by psychologists with knowledge of these tests and traditional case data following personal interviews with students. The counselor's interview and knowledge of case data, including tests, added little to the statistical accuracy of prediction when the test battery only was used for prediction. College Freshmen The plethora of studies made of high school drOpouts is not duplicated when the population under study is limited to college fresh- men. This is particularly significant when compared with the finding that most college attrition takes place during and following the fresh- man year. Studies reported on junior college freshmen attrition are almost non-existent. Matson (1955) completed the most comprehensive study of twenty two-year colleges between the years l9h9-1957. In comparing a group of students who withdrew with a group who did not withdraw but had similar characteristics, Matson found no significant differences. She concluded that the student who withdraws from junior college may lack a sense of belonging or identification with the college environment. Following this hypothesis the work of Erik Erikson becomes useful in the assessment of the effects of varying college environments on personality change. The tasks of establishing a stable identity and deve10ping patterns of intimacy are of major concern to college fresh- men. Erickson (1950) understands problems to be difficulties that grow out of attempts to resolve these tasks. 19 Motivated by a high attrition rate among college freshmen, Lang gt.gl}(l962) compared the Edwards Eersonal [reference Schedule with instructors' estimates of academic achievement. For females, academic achievement correlated positively with achievement and dominance but negatively with nurturance. For males, academic achievement correlated positively with order and negatively with dominance. Lang's study was made on a limited sample of #9 females and 38 males. The academic abilities of junior college students were the interest of Seashore (1958). In a comparison of academic aptitudes as measured by the College anlification Test (CQT) of students in two and four- year colleges, Seashore found the following: one, the median score for junior college freshmen is near the 25th percentile for senior college freshmen; two, about 24 per cent of the junior college men and 20 per cent of junior college women are above the respective medians for fresh- men in fourayear colleges: three, there are many junior college students whose scores would be considered superior in senior colleges, and many low-scoring college freshmen would also rate low in junior colleges. These findings indicate the importance of research on a junior college pepulation and the need for individualized instruction through counsel- ing of junior college students. Funkenstein gt.§l,(l959) observed that extreme scores on the Brggggain fielf-Concgpt Te§t were significantly associated with visits to the psychiatric clinic during the freshmen year. King sees as one outcome of the Harvard Student Study an evaluation of tests or test items that are potent in predicting the need for help from the psychiatric service or the counseling service at Harvard: ”Already 20 our experience has indicated an association between certain extreme scores on the fixers-Briggs lype Indicator in the freshman year and later occurrence of emotional disturbance" (King, 196a, p. 330). On the subject of freshman failure the humanities have a spokesman in Ridlor (1961, p. 59). He found that the relationship of good reading to success in college is primary. A second skill is the ability to listen constructively coupled with candid, dispassionate, objective thinking. The attitudes which block successful completion of the fresh- man year "all revolve around the central pole of self-consciousness. Fear, anxiety, selfishness, aggressiveness, timidity, 1ethargy--these often spring from unwholesome conceptions of a relationship between self and society." Anxiety for the adolescent is a regular part of their traveling equipment, Ridlor concluded. alggggota Multiphagic [ergonality Inventory Related Studies Buros (1959) has documented 779 references to the use of the MMPI. An estimate of MMPI research in the past six years would increase this number. The following discussion isolates the most significant recent findings that relate to this study under these headings: College Screening Studies, Achievement Studies, Deviant Behavior Studies, and Studies Incorporating Conflicting Results. College ficzggning Use of the HEEL: Among the earliest attempts to use the MMPI as a psychometric tool for diagnosing personality disorders among college students is the work of Hampton (1947). He found the MHPI "very helpful" in determining areas of student deviation. This tended to confirm similar use of this instrument in military and industrial settings documented by Mann (l9h9).‘ 21 Roessel (1954) made an extensive study of 763 drapouts and graduates from among 2835 students in Minneapolis public schools. On nine of ten MHPI scales the drOpouts showed significantly more abnormal scores. The single exception was on the masculinity-feminity (Mf) scale. When grade level of the withdrawal was considered, pupils whose MMPI scores were higher tended to withdraw from a lower grade. Tenth grade dropouts were more abnormal on the MMPI than those who drOpped from the eleventh or twelfth grades. He also observed that, for both dropouts and graduates, with an increase in the number of children in the family the mean HMPI scores increased. Drake (195“) compared evaluations by counselors of college students with MMPI scores. Those students who were rated to be difficult (e.g., aggressive or Opinionated) had achieved elevated PsychOpathic deviation and Hypomania. A "shy“ group was characterized by high Social introver- sion. The nonresponsive group coded high. Goodstein (1954) found a striking similarity when testing regional differences in MMPI reSponses among male college students. He concluded that the deve10pment of regional norms was unnecessary. Grace (1957) found that certain MMPI scales correspond positively with attrition and test scores. Hawkes (1959) drew two eXperimental groups from among 1100 entering freshmen. MMPI scores for one group deviated two standard deviations above the mean. The control group was matched for age, sex, class rank, and other test scores within the range of the normal profile on the MMPI. The MMPI was found to be an accurate predictor to the extent that the student's curriculum was concordant with his Kuder’Preference Record interest scores. 22 The most prolific reporter of research on the identification of maladjusted college students is Kleinmuntz. In 1960 he reported an item analysis of the MMPI records of two criterion samples that yielded 43 discriminating items. The maladjusted college student appeared as an ineffectual, pessimistic, procrastinating, anxious, and worried person. He also found (1960b) that the mean.Ego Strength (Es) scores for ad- justed college students (defined as having no counseling contacts) was significantly higher than the mean for>ma1adjusted students (defined as having personal counseling). Parker (1961) studied the predictive use of the MMPI in a college counseling center. He regarded the test as a valid indicator of motiva- tion for treatment especially if it is administered at the time coun- seling is requested. He also observed that the complete form is more useful than the short form for this purpose. Kleinmuntz (1961) has developed a college maladjustment scale (Mt) designed for screening purposes. It can be used to identify persons most in need of counseling. The normative data for the Mt scale was based on 208 teachers' college candidates and 825 entering freshmen. The marked differences on Mt scales between various college groups suggested that separate norms be computed before the scale is applied to any one group. From a group of 126 freshmen at a counseling center using MHPI scores from an orientation battery two counselors judged 81 to be seeking vocational-academic counseling and 45 to be seeking “emotional” counseling. Using a cutting score of 15 on the Mt scale of #3 items, Kleinmuntz achieved 8#% accuracy. Parker (1961, p. 89) tested this Mt scale with samples of engineering students. He achieved 23 accuracy of 46% of the students who later requested "emotional coun- seling" from among students tested at matriculation. He concluded from further study that the Mt scale discriminates between adjusted and maladjusted students at the time of counseling but "it will likely not do so in a routine battery." The extensive interest of Kleinmuntz is indicated by his eXperi- ment in personality test interpretation by digital computer. A set of decision rules was devised for interpreting the profile patterns of the MMPI of "maladjusted" and "adjusted" college students. The procedure used divided students into that dichotomy based on the decisions of an expert test interpreter. His decision-making processes were tape recorded while he was thinking aloud during the sorting of the profiles of the 126 college students. The programmed decision rules yielded greater accuracy than the decisions of the original interpreter (Kleinmuntz, 1963, p. 416). Heilbrun (1963) investigated the possibility of a better system of weighting the ten clinical scales of the MMPI to enhance their usefulness as measures of adjustment level within a grossly normal college popula- tion. He used a two-group discriminant analysis, the sexes being con- sidered separately, and found that his revised system differed greatly from the standard system. Heilbrun's study is a sample of the plethora of work being done with the MMPI. Achievement Studies with the MRI. It appears that use of the MMPI for the screening of college students is both less intensive and extensive than the use of this instrument for studies of under and over-achievement. The review that follows identifies significant studies in chronological order. 24 Clark (1953) studied the grade achievement of female college students in relation to nonintellective factors. He concluded that although poor grades were not related to any syndrome measured by the MMPI in randomly selected samples of college women, a group of items can be selected which will separate achievers from nonachievers. Hackett (1955) administered the MMPI to 32 college freshmen on the assumption that nonability factors play an important role in determining college achievement. He tested the hypothesis that scores on the MMPI were related to later achievement in college. Course grades were the criteria of achievement. Hackett found that scales K, 1, h, and 9 on the MMPI were valid predictors of academic achievement.1 The number of negative correlations between grades and MMPI scores tended to sub- stantiate the assumption that good adjustment is important to college achievement. Frick (1955) tried to improve the prediction of academic achieve- ment by adding the clinical scores on the MMPI to the ACE (the American Council on Education ngchological Examination for College Freshmen) in a multiple correlation. The coefficient of determination afforded by the ACE alone was .23. Adding selected MMPI scales increased the correlation to .41. The 12% index of forecasting efficiency by the ACE alone increased to 23% by including certain MMPI scales. While this evidences a positive gain in adding the MMPI, the results are only fairly discriminating. 1Scale names and characteristics are identified in Appendix G. 25 Anderson (1956) compared an experimental group having a paranoia (Pa) T-score of #0 or more and at least one counseling contact with a randomly selected group of normal college students. The Pa group had more academic difficulty than the control group both in the non and underachiever categories. A significantly lower Pa was reported for students who had some difficulty with their parents. Anderson ventured that low Pa represents repressed or denied hostility. Bendig and Klugh (1956) reported a validation of Gough's scale in the prediction of academic achievement. Gough's scale and high school rank in class correlated .32 with self-reported quality point averages. CPA was .95. Stone and Ganung (1956) divided 317 female freshmen into two groups. Sixty-eight were placed in the normal group, T-scores being uo-éo. One hundred twenty-six were placed in the high group, defined as having a T-score on one of the clinical MMPI scales over 70. Thirty-eight per cent of the normal and twenty per cent of the high group graduated, a difference that is significant at the .05 level. The mean grade point average for the "normal" group was 1.9 and for the "high" group, 1. 5, a difference that is significant at the .01 level. Drake and Getting (1957) administered the group form of the MMPI ‘30 3%0 entering male college freshmen. The GPA for this group was compu'ted for the first semester. The MMPI was coded according to the Hathaway system and all profiles with scales 8 and 9 paired high and low on 31 were separated from the total group and divided into two--5 high and 5 not high. The GPA distributions were tabulated for the two sub- groups and the total group and tested for significance. These researchers concluded that a rather complex pattern was necessary to predict the criterion. No single scale in this pattern differentiated the subgroups from the total group. Drasgow and McKenzie (1958) in a study of college transcripts, graduation, and the MMPI found nongraduates significantly higher in the direction of maladjustment on scales 8 and 9. Zeaman (1958) compared men and women students on some of the per- sonality attributes related to achievement in college. Low achievers of both sexes were found to be significantly more hypomanic and less conventional than high achievers. Jensen (1958) administered the MMPI to z+58 students at a western denominational college. Scholastic ability was measured by the ACE and high school GPA plus college grades. The most consistent trend identi- fied was for nonachievers of low scholastic ability to obtain higher MMPI scores than the achieving students of high scholastic ability. 0n scales F, 8 and 9 the nonachieving students of low ability scored sig- nificantly higher than the achievers of low scholastic ability. The suggestion was made that scholastic ability favors adjustment and low scholastic ability obstructs it. McKenzie (1962) attempted development of MMPI scales predictive of academic over and underachievement. He found that underachieving students scored higher than normals on 1+ and 7, lower on L and K. The overachiev- 11'18 students scored higher on 2, 5, 7 and lower on 9. McKenzie concluded that. both groups were more anxious than normal students. The under- achievers, he hypothesized, internalized anxiety while the overachievers efifternalized their conflicts. 27 Drake (1962) found that scales 9 and 9 were significantly associated with low achievement eSpecially when scale 5 was not high. This pattern evidenced hostility, Drake suggested, and a consequent need for empathic counseling rather than legalistic demands. Drake favored a pattern interpretation of MMPI scores rather than individual scale interpretations. McKenzie (196h) has reported the most recent study on the dynamics of deviant achievement. Underachievers among college males were com— pared with a group defined as normal on the MMPI clinical and validity scales. N for each group was 75. The means for the underachievers were significantly higher on scales 1 and 7, lower on L and K. The higher scale 7 scores seem to indicate that underachievers are more anxious than normal achievers. The higher scale h score suggests that under- achievers tend to be antagonistic to authority, to reject socially accepted values, and to incline toward impulsive behavior. The antisocial orientation of the underachievers may be seen as a way of handling anxiety. A lower degree of defensiveness would be consonant with an attitude of antagonism toward authority figures and socially accepted values. McKenzie observed that since some need to flout these values, the underachievers are willing to admit information about themselves that may not be consistent. In such peOple underachievement could be viewed as a way of expressing rejection of parental and societal values. Under- achievers respond to a number of items in a direction that suggests harboring a good deal of aggression. Another cluster of items indicated that underachievers are dependent for direction upon peOple who may be perceived as having little reapect for their rights. This tendency to subserve their own wishes to those of other peOple may be important in generating hostility, an important dynamic with underachievers. 28 Studies on Deviant Behavior. Of the numerous studies of delinquency and abnormal emotionality using the EMPI, the following have been selected from the past decade for their value to this investigation. ’ Glasscock (1955) investigated the value of the MMPI as a prOgnostic instrument with institutionalized psychotics. Using three out of nine empirically derived identification criteria he correctly identified 77% of the subjects. Gottesman (1959) found the Ego Strength (Es) scale to broadly discriminate between psychiatric and nonpsychiatric adults and adolescents. He did not find a discrimination between different degrees of psychiatric incapacity nor between delinquent and nondelinquent behavior. Silver (1961) observed that it was characteristic of state hospital chronic patients to have extremely elevated F scales, an indicator of psychosis. Hathaway and Monachesi (1963) have published a longitudinal study of 15,300 ninth grade students. An attempt was made to isolate premoni- tory signs of delinquency for accuracy in prediction. MMPI patterns were correlated with other data like teacher evaluation, socio-economic status, school achievement, and type of residential community. "Bad conduct" was associated generally with high scores on the 4, 8, and 9 scales. It was observed that personality patterns of delinquency-prone adolescents are diverse and not monotonic. These investigators found very generally that scale 4 with its implications of rebellion and immaturity predicts delinquency most sensitively. By contrast, the neurotic and schizoid triads of scales 1, 2, 3 and 8 are more indicative of dropouts. Jurjevich (1963) used two groups of 70 and 100 delinquent girls to establish unbiased estimates of the means and standard deviations of the clinical, validational and 38 additional MMPI scales. The values obtained were compared with standard scales. Delinquent girls were found to differ from the normal pOpulation in being more anxious, less self-controlled, and less able to inhibit hostility. Their ego strength was lower. They were more excitable and more ready to admit the exis- tence of some pathological traits. They were less socially responsible and less repressive. Hathaway, one of the originators of the MMPI is quoted in a recent University of Kentucky publication, oci evianc and Adol scent Personality: "Professor Ball's findings again demonstrate that the MMPI is a useful instrument in the study of the personality patterns of nonconforming and conforming groups of adolescents” (Ball, 1962, p. VIII). Halroyd (1964) attempted to identify with a new MMPI scale (PDf), defensive paranoids. Seventy-six per cent of the females tested were properly classified; 19% misclassified. Seventy-eight per cent of the males tested were properly classified; 15% misclassified. Johnson and McNeal (1964) with data available on 316 psychiatric patients soon after admission used a multiple regression formula to predict length of stay. Demographic and MMPI predictors made possible early identification of long stay patients with greater than base rate frequency. The asymtote of prediction was reached when 6 of 24 variables had been included, suggesting that further research in the area would be redundant. Prediction with the five predictor schema was stable on a cross-validation sample of 352. 30 Incorporating Conflicting Results. Not all researchers find the MMPI so useful. Among the studies which report less positive findings are the following. Brown (1948) wanted to determine whether a University of Minnesota freshman group could be considered a sample of the original normal papulation of college students used by Hathaway and McKinley in the derivation of the MMPI scales. They found significant differences among the several college pOpulations on which the MMPI data were available. Thus a valid interpretation of an individual profile would seem to re- quire knowledge of whether an individual's group differed significantly from the original "college normal" pOpulation. Should other researchers wish to make similar comparisons, the findings by Mueller (1965) of the means and standard deviations from the administration of the MMPI to 260 males and 205 females at Michigan State University can be located in Appendix C. Sanders (1950) attempted the prediction of academic success among university freshmen in a school of education. Four hundred seventy-three students were given the ACE and the MMPI. Using honor point ratio as the criteria, it was found that neither measure alone nor a regression equa— tion was acceptable for predicting grades of individuals though they may be for group prediction. Hoyt and Norman (1954) found that grades were predicted more accurately for freshmen men with normal profiles than those with deviant profiles. No differences were found in percentages of over and under- achievers who scored high or low on any MMPI scale. 31 Patterson (1962) reported having the experience of turning up about 10 false positives for every disturbed student among those having high scores on the MMPI. It was hypothesized that there may be something peculiar about college students on clinical tests like the MMPI. Caution in interpretation except as a single part of a clinical evaluation was stated. Seegars (1962) with an N of only 25 females found indistinguishable bright and probationary students. weigand (1951) summarized studies of student success and failure: The survey of the literature in the field of prediction of academic success directs attention to four main conclusions. First, intellectual factors have been emphasized.... Second, the various measures of intellectual factors correlate .5 and _ .7 with grade point averages in college. Third, high school grades and subject matter tests are superior to intelligence tests for predicting college grades. Fourth, ... there is acceptable evidence that personal-data items, certain aspects of personality, and motivational aspects of the individual are important in identifying those students who will fail (p. 201). It is the investigation of that fourth conclusion toward which this study is directed. Quinn (1957) has observed that when intelligence was properly con- trolled the MMPI failed to define the differential personality character- istics of under and overachievers. He posited that many of the MMPI characteristics that differentiated under the overachievement in past studies could be ascribed to the influences of intelligence. The relationship of personality assessment to academic achievement surveyed here indicates inconclusive and contradictory findings. This apparent ambiguity is somewhat explained by the selectiveness and diversity in American education as well as the crudeness of contemporary measurement tools. 32 LUTHERAN YOUTH RESEARCH STUDIES The definitive study of the Lutheran Youth Research Inventory (LYRI) has been published by Strommen (1963). His scholarly report of a four- year study of 3000 Lutheran youth primarily from the midwest is the only technical literature thus far available. Technical data for that somewhat pOpularized report can be gleaned from Strommen's unpublished doctoral dissertation completed under the direction of C. Gilbert‘Wrenn (Strommen, 1960) . The LYRI was normed on a randomly selected sample of high school SOphomores, juniors, and seniors of both sexes. This investigator is the first known to use the LIRI with a college population. It has been learned that since the inception of this study nineteen Lutheran colleges have initiated the use of the LXRI with their students. Findings from these studies have yet to be published. FACULTY RATING STUDIES Studies made of faculty judgment of student emotionality are somewhat more available than LYRI studies. Billard (1962) compared teacher ratings of personality with results of an incomplete sentence blank. He tried to identify high school freshmen with acute problems. In spite of rather laborious scoring required, values were seen in the patterns that emerged. The relationship of the teacher's knowledge of the student's back- ground to the effectiveness of teaching and the relationship between teacher's knowledge of students and certain selected teacher character- istics are samples of related studies. Much of the literature records 33 pupil perception of the teacher. Apparently little study has been made of the teacher's perception of the college student. The dearth of literature is not, in itself, a hazard to the present study since the relationship of personality assessment to its instrumenta- tion is part of the purpose of this investigation. The two qualitative studies reported below tend to supply the lack of quantitative studies. Hathaway and Monachesi (1963) tested 11,319 ninth grade students in Minnesota. They designed their study to compare the MMPI with pre- dictions of behavior derived from teacher ratings. The teachers in the schools who were familiar with the students were asked to list five students they thought most likely to get into serious difficulty with the school authorities or the law within the next four*years. They were also asked to list five students most likely to show evidence of emotional maladjustment within the next four years. The names from one or more teachers of each school were then combined in final lists in the two prediction categories. No real effort was made to get exactly ten or to restrict the number named. From the vast data collected by Hathaway and Monachesi (1963), conclusions of interest to the present study follow: The variety of personality types among dropouts makes it hard to predict accurately from the MMPI categories. To generalize from scale 8, which is most generally related to drapout for boys, some of these adolescents probably isolate themselves and are not very visible, appearing drab and uninteresting in contrast to their troublemaking and much more obstreperous classmates who are potential delinquents. Teachers attempting to pick future delinquents were found to be as good, if not better, in predicting which pupils would drop out of school (p. 102). Ball (1962, p.88) anticipated that ratings by teachers would depend in learge part upon the dispositions and attitudes of the teachers. Since 34 these very dispositions are a part of the total school situation the student encounters, they are pertinent to such a study. It is not assumed that the teachers' ratings have any psychological validity other than the social situational sense described. To involve a faculty of nearly thirty teachers in making judgments about student emotionality admits of a wide variance in validity, Three factors tended to reduce this hazard. First, the school enjoyed a wide reputation for a pervasive "family spirit." Both faculty and students spoke of a "glass house" existence. Second, faculty members were dedi- cated both in principle and practice to individualized instruction. Students drOpped in without appointments at faculty offices and homes with high frequency. Third, only 23 students in this sample of 256 were commuters. The compact residential campus afforded casual and informal interaction of faculty and students. SUMMARY OF RELATED RESEARCH This chapter has surveyed the historical develOpment of psychometric theory and practice. The focus has been on personality assessment. Studies of student attrition found research insights in the studies made on secondary school samples but the major interest was on junior college freshmen students. Significant studies from the vast research on the MMPI were discussed under the subjects of screening college students, under and over-achievement, deviant behavior and conflicting results. This was followed by a discussion of the LIRI. The chapter concluded with an eXploration of literature on faculty rating studies. The study of related literature afforded stimulus and direction for the design of the study to be reported in the next chapter. CHAPTER III DESIGN AND PROCEDURES INTRODUCTION In this chapter the sample of the pOpulation studied will be described and the design of the study will be discussed. The investiga- tion will be divided into two parts. The first will consist of hypotheses testing and the second will describe and explore selected demographic data. The administration of instruments used and coding procedures will be discussed. The methodology for the design and analysis will be detailed. THE SAMPLE The 1963-1964 freshman class of waldorf Junior College was selected for this investigation. Waldorf College, located at Forest City, Iowa, is the only two year college in the U. 8. Sponsored by the American Lutheran Church. Liberal arts, pre-professional, and terminal programs in business and industrial arts are offered. Eighteen states and four foreign countries were represented in the freshman class of 1963-1964. Forty per cent of the students were residents in Iowa, Membership in the American Lutheran Church was claimed by 82%. The modal occupational background of parents was agricultural. The sample included 252 subjects of which six foreign and special students were excluded to enhance the homogeneity of the sample. Additional demographic data is presented in Appendix B. 35 36 DESIGN Criterion (Dependent) Variablgg Drogouts. Students who matriculated at Waldorf College in September, 1963, and who did not enroll as SOphomores in 1964 constitute the first criterion group. Hereafter this group will be called Dr0pouts. The percentage of drOpouts in relation to the size of the class at matricu- lation was 32%. Enrollment information was supplied by the Office of the Registrar. Faculty Judgment of the Acutely Concerned. At the conclusion of the school year, May, 1964, the waldorf College faculty was asked by the investigator to identify on a list of all freshmen those students who seemed to have acute concerns that may have tended to inhibit their academic success. An additional refinement was the limitation to students with whom the faculty member had some personal interaction. Any first- hand encounter was included. The locus of the relationship was not con- sidered: classroom, office, home, or casual campus contact. The faculty members were asked to limit their selection to 10% of the students in their classrooms or a maximum of 25 students from the whole freshman class. Three student personnel workers were asked to limit their selec- tion to 20% of the class or a maximum of 50 students. This weighted the nonclassroom contacts in response to an awareness that only about 10% of the student-week is spent in classrooms. Faculty members completed their judgments independently. All of the 29 faculty members responded. Not identified by any faculty member to have acute intrapersonal (Hancerns that may have tended to inhibit academic success were 33% of the class. Approximately 2C5 of the class were identified by four or more faculty members to have acute intrapersonal concerns. Students in the latter group constitute the second criterion group hereafter called Acute Concerns. TABLE 3.l--Students with acute concerns identified by the faculty. Number of faculty members who identified a given student 0 l 2 3 4 5 6 7 8 9 10 ll 12 l3 14 15 l6 l2 l8 9 55 2 24’ 12 9 l3 3 2 3 3 2 0 0 l 1 O O 1 Number of students identified Eredictor (Independent) Variables Minne ota Multi hasic e sonalit nvento . The MMPI was develOped in the early forties as a psychometric instrument designed to provide in a single test an objective assessment of personality characteristics that affect personal adjustment. The MMPI contains 550 items which were empirically selected to differentiate between normals and criterion groups of clinically diag- nosed abnormals who indicated relatively pure psychiatric syndromes (Dahlstrom and welsh, 1959). The items are arranged in ten clinical and four validity scales of known reliabilities (Hathaway and McKinley, 1951). Reliability studies indicate that profile configurations and rank order positions of any one scale appear quite stable (Layton, 1954, p. 634-660). The authors have reported that test-retest reliability coefficients range from .71 to .83 (Froehlich and qut, 1959). The subject is asked to sort all the items into three categories: True, False and Cannot Say. The Cannot Say score is regarded both as a gross validity measure and a measure of more subtle test-taking attitudes (Ball and Carroll, 1962). Personality characteristics may be assessed on the basis of scores on ten clinical scales. The original psychiatric nomenclature for the clifiical scales has tended to be abandoned in favor of a numerical label for each scale. Although the scales were named according to the abnormal manifestation of the symptomatic complex, they have all been shown to have meaning within the normal range (Hathaway and McKinley, 1951, p. 6). A recent normative group of Michigan State University freshmen is re- ported in Appendix C (Mueller, 1964). Table 3.2 Specifies characteristics of the scales selected for use in this investigation. TABLE 3.2--Characteristics of selected MMPI scales.1 Scale Scale Clinical number name Normal expression 7 L F K 1 H5 Hypochondriasis 2 D Depression 3 Hy Hysteria 4 Pd Psychopatic deviation 5 Mf Masculine- feminine 6 Pa Paranoia '7 Pt Psychasthenia £3 Sc Schiz0phrenia S9 Ma Hypomania 10 Si Social introversion ES Ego strength DY Dependency DO Dominance RE Social reSponsibility PR Prejudice ST Social status CN Control Can not or will not answer Illiteracy, defensiveness, faking good Unusual symptoms, self-depreciation, faking-bad, test-taking attitude Correction factor; self-acceptance Concern about body; tired, lethargic Self-dissatisfied, low morale, unhappy Idealistic, naive, ill under stress Aggressive socially, rebellious, cynical Interest pattern of Opposite sex Perfectionistic, stubborn, hard to know Dependent, anxious, inferiority feelings Negative, odd, difficult, apathetic Expansive, Optimistic, unorthodox Unassertive, self-conscious, shy Spontaneity, personal adequacy, vitality Indecisive, ambivalent, distrusts self Dominates in face-to-face relationships Internalizes social-moral reSponsibility Attitudes toward minority groups Attitudes toward socio-economic status Measure of impulsiveness and planning ‘luhdapted from Hathaway and McKinley, 1951, and Hathaway and Monac heisi ' 19.33. 40 In summary, The Fifth Mental Measurements Yearbook (Euros, 1959) is quoted. 'Warren Norman (Buros, 1959, p. 168) wrote: "This instrument (the MMPI) is probably the most carefully constructed and thoroughly researched inventory available for personality assessment. It is likely to be an increasingly useful clinical tool." In the previously quoted source Albert Ellis stated (p. 166): "... it can be confidently stated that in the whole history of modern psychology there has been no other personality inventory on which so much theoretical and practical work has been done." Ellis added: "In spite of all this research activity, the question of just how valid a clinical instrument the MMPI is has still not been finally settled." Lutheran Youth Research Inventory. The lERI is a psychometric instrument designed to provide a way for'youth to describe themselves and to indicate the kinds of help they desire. Of the 44 available scales, 32 scales were selected for study in relation to Waldorf College freshmen. The norm groups constitute a randomly selected probability sample of approximately 4000 high school youth (ages 15-17) who are members of lutheran congregations in the U. S. and Canada. Two measures of reliability were used in determining the internal consistency of the sub-scales--Kuder-Richardson's formula 20 for items with dichotomous responses and Hoty-Stunkard's analysis of variance for items with 6 responses. The Hoyt-Stunkard reliability coefficients range from .74 to .89 for a mean of .83 on the sub-scales. The re- liability coefficients for the ten major scales range from .84 to .97, a mean of .91. 1+1 Content validity for Lutheran youth has been established by a check on the universe of youth's concerns during a national survey of an exact probability sample. Congruent validity and relative frankness were demonstrated through a high correlation with the MMPI Lie (L) scale. Construct validity has been established through a factor analysis, 30 criterion groups and a cross-validation study. The latter study was a replication in a second national survey of youth in a different church body. Table 3.3 Specifies characteristics of LYRI scales selected for use in this study. 42 TABLE 3.3--Characteristics of selected LIRI scales. ========================================================================= Vari- Name able 1. Family Unity 2. Family Spiritual Growth 3. Parental Authority 4. Intra-Family Relationships 5. Dating and Morals 6, Life Partner 7. Dating and the Church 8. Spiritual Doubts 9. Personal Assurance 10. Inadequacy Feelings 11. Academic Problems 12. Social Relationships 13. Relation to Teachers 1“. Peer Acceptance 15. Judgment of Peers Description Feelings about the degree of love and cohesive- ness in the family. Awareness of signs that the family lacks a common faith. Concern over autocratic parental behavior and parental mistrust. Feelings of hostility toward family members and a troubled awareness of poor interapersonal relationships. Feelings of guilt over present dating behavior. Worry about finding the right marriage partner or being wanted by somebody else. Anxiety over disparity between the ideal of the church and the experiences of real life, Distress over feelings of rebellion against the teachings of the church. Worry over not feeling certain of God's forgiveness. Fears and disappointment which result from a low estimate of self and one's achievements. Fear of failure and concern over academic limitations. Oversensitive awareness of the abilities of others in a group. Feelings of disapproval both from and toward the school, its teachers, and its courses. Lonely awareness of feeling like an outsider and unnecessary to any group, Troubled awareness of personal and peer group behavior that falls below expectations. TABLE 3.3--Continued. 43 W W Vari- Name Description able 16. Personal Feelings of guilt over falling short of Consistency personal standards. 17. Adult Disappointment and critical feelings toward Consistency adults over their failure to live up to high 5 tand ards and ideals . 18. Christian Concern over failing to live as a committed Vocation Christian in daily living and career. 19. Moral Problems Attraction to questionable or unethical activities. 20. Boy-Girl Knowing the Christian concept of sex, dating Relationships and marriage. 21. Spiritual The application of Christian principles to Guidance daily living and thinking. 22. Family Increasing the unity and spiritual vitality Counseling of the family. 23. Self-realization Greater self-understanding and self-realization. 24. Achievement Achievement and the status symbols of success. 25. Service Service and a sense of meaning and beauty in life. 26. Happiness Marital happiness, security and contentment. 27. Justification Recognition and acceptance of the doctrine of justification by faith. 28. Humanitarian Attitudes of mercy or inhumanity toward those of different nationality, race, or creed. 29. Conservatism Agreement with the traditional doctrines of the Lutheran church. 30. Christian Perception of Christian concepts and rejection Concepts of a generalized religion. 31. Forgiveness Personal awareness of God's constant forgiveness. 32. Religious Personal participation in daily emperience of Commitment religious values. I? DEMOGRAPHIC DATA The original design of the investigation being reported in this thesis was confined to the MMPI and LYRI. The research question was: Would either or both of these instruments indicate early in the school year students with acute intrapersonal concerns and with marked drOpout po- tential? To afford an answer based on broader information, three additional sources were added. The first was standard data from the students' files in the Office of Admissions. A biographical, familial, scholastic, and medical history of the student may yield demographic data that could be quantified for statistical manipulation. Extensive study by the investigator prompted the conclusion that the admission forms used were not designed for the kind of collation required. Since such a case study possibility did not prove possible, an alternative was pursued. A second source of demographic data was the construction and ad- ministration of a questionnaire. Based on the survey research model of Kerlinger (1964), 73 forced choice items were constructed. The items included the students' feelings and their intensity about the other students, the faculty, administration, board of regents, alumni and the sponsoring body of Waldorf College, the American Lutheran Church. Selected socio-economic factors, matriculation influences, self-estimate of certain aptitudes, and attendance yielded quantifiable information. Additional items focused on the students' relationships to parents, home congregation, classmates, and close campus friends. All of the items are reproduced in Appendix D. 45 A third source of demographic data was scores on the American College Testing Program (ACT) battery of tests in English, mathematics, social studies, and science. Deve10ped for admissions purposes, the ACT was designed for smaller, less selective colleges. ADMINISTRATION PROCEDURES During the first week of the school year, September, 1963, all en- tering freshmen at Waldorf College completed the MMPI, the LYRI, and the ACT. The instruments of the study were administered by the Dean of In- struction, Richard Moe, and the Dean of Students, Kenneth Johnson. Machine scoring was done by National Computer Systems, Minneapolis, Minnesota. Code numbers were assigned to student responses so that personal associations would not contribute a bias. During the fall of l96b all the freshmen of the previous year re- ceived by mail a questionnaire previously described together with a stamped, self-addressed envelOpe. Anonymity was attempted through coding and the request that names be omitted. Data from completed questionnaires was transferred to code sheets for key punching by the Michigan State University Data Processing Office. Statistical computations were done on the MSU Control Data 3600 Computer. SCORING AND CODING Scores on the LYRI instrument were reported in standard scores. The MMPI scales were standard scores with the exception of the Question (?) scale which was reported in raw scores. The demographic data were reported in raw frequency distributions and percentages. The ACT was reported in standard scores. Coding is reported in Appendix G. 4:- C\ HYPOTHESES TO BE TESTED Two major hypotheses have been formulated and are stated below in research, null, and alternate forms: LHR % % II. HR Selected MMPI and LYRI scales will predict Waldorf College freshmen who withdraw after the freshman year. There will be no difference between a multiple regression coefficient for students who drOp out and zero. Ho: R = 0 There will be a difference between a multiple regression coefficient for students who drop out and zero. HA : R # 0 Selected NMPI and LXRI scales will predict waldorf College freshmen who have acute intrapersonal concerns, measured by faculty judgment. There will be no difference between a multiple regression coefficient for students with acute intrapersonal concerns and zero. Ho: R = 0 There will be a difference between a multiple regression coefficient for students with acute concerns and zero. %:a#o ANALYSIS PROCEDURES MMPI and EYRI A preliminary factor analysis of the MMPI and LYRI instruments was made by the use of the Computer Institute program identified as Fanod 3. 47 Principal axes and analytic orthogonal rotational solutions were per- formed using the MMPI and LIRI scales selected for study. Rotated factor loadings indicated which independent variables correlated highest with the criterion and lowest with each other. A second study was made of the NEPI and LYRI instruments by the use of the Computer laboratory program identified as CORE u. CORE (correla- ion and regression), program a, calculates multiple regressions (least squares fits to arbitrary functions) on the CDC 3600 computer. Correla- tions were calculated to indicate how well the independent variables "explained" the dependent variahle. To restate, the problem was one of .LV giving normalized weights to a series of independent variables in the prediction of the dependent variable. The normalized weights or "beta wei hts" were obtained by adjustin: each variable br 'ts standard devia— tion before estimating the coefficients. Predictor variables were selected from those which correlated hioH Ho. pm pcmoamH:MOm*« .Ho>oa mo. pm pcwoamwcmOmO OOH OH O H OO HH O H OOH OOO OOH O OO ON O N OOHOOOEOO OOH OOH *OOH N NO O ON O OOH OO ON OH OO HN O O mosoaom Hauspmz OOH OH OH N OO O OH N OOH OO ON OH OO NN O O OOOOOOO HmHoom OOH OH OH N OO O OO O OOH OOO .HH O OO NN ON O moapmeonpmz OOH OOH ..OH N OO O OH N OOH mOO *OON OH OO NN O O OOHHOOO mHmswm OOH ON ON O OO OH NH O OOH OO OO OH NO ON HH O OOHOOOEOO OOH «ON ..ON O OO HH ON O OOH OO OO ON OO OH OH O mocmflom Hmpspmz OOH ON OO O OO NH OH O OOH OO OO OH OO ON HH O OOOOOOO Hmfloom OOH ON ON O OO OH HN O OOH OO OO OH OO ON OH O OOHOOEOOOO: OOH OON ..NH O OO OH HN O OOH OO ON HH HH ON OH O OOOHOOO «HO; O z O 2 O 2 O 2 O 2 O 2 O O O z HOOOO OO NO HO HOOOO OO NO HO OOOOO OOO mcpoocoo opso< upsoaonn H 6939.80 358 Em 35090.8 no.“ 3.30m 94m no mdmhwinIONJ 3mg. 87 ANALYSIS OF MMPI AND LYRI RESULTS Factor An sis The male sample (N = 83), using a varimax rotation analysis, yielded ' six factor loadings. Both criterion variables loaded highest on the fifth factor. Seven predictor (independent) variables were selected from among those that correlated highest with the criterion (dependent) variables and lowest with each other for the male multiple regression equations. The female sample (N = 78) yielded nine factor loadings. Both criterion variables loaded highest on the third factor. Eleven.pre- dictor variables were selected for the female multiple regression equations. TABLE b.28--Factor analysis of MMPI-LYRI. J -:__ gale Female Vari- Drop-a Vari- Acuteb Vari- DrOp-c Vari- Acuted able outs able Concerns able outs able Concerns 24 .7516 2a .7516 uz .7652 uz .7652 39 .7416 39 .7nl6 9 -.7784 9 -.7784 27 .6644 27 .66u4 31 -.8588 31 -.8588 36 .6556 36 .6556 25 -.7812 25 -.7812 30 .6317 30 .6317 47 .77h5 47 .7745 23 .4378 23 .4378 19 .7459 19 .7459 #5 .8438 #5 .8438 15 -.7281 15 —.7281 34 .5619 34 .5619 11 -eu977 11 -014'977 ‘59 .4032 b51 .6019 C50 .5493 d51 .4051 The most useful scales were ranked and identified in Table b.29. When all selected variables were combined it became apparent that more LYRI scales were useful than NMPI. The ratio for males was six to two. For females the difference appeared inconsequential, a ratio of six to five. The most discriminating variables for male drapouts were LYRI scales. For females with acute concerns the most discriminating scales were from the MNPI. TABLE b.29-qMost useful R scales (ranked by value of regression coefficient). Drgpouts Agute concerns Vari- Vari- able Scale name able Scale name £22.12 23 Self-realization 23 Self-realization 24 Achievement 35a Hysteria 26 Happiness 27 Justification 27 Justification 39a Psychasthenia 30- Christian concepts 2h Achievement 36d PsychOpatic deviation 26 Happiness 39a Psychasthenia 16 Personal consistency Female 25 Service 47a Prejudice 11 Academic problems 3ua Depression #28 Social introversion 35: Hysteria 9 Personal assurance #2 Social introversion 3&3 Depression 31a Forgiveness l9 Moral problems 9 Personal assurance 35a Hysteria ll Academic problems 473 Prejudice 25 Service 15 Judgment of peers l5 Judgment of peers 45a Dominance #53 Dominance 31 Forgiveness l9 Moral problems L aMMPI scales; unmarked variables are LYRI scales. 89 Multiple Regression The regression equations for drOpouts and acute concerns of both sexes were reported in Table 4.30. The randomly dichotomized samples of 41 males had 33 degrees of freedom. The samples of 39 females had 27 degrees of freedom. TABLE 4.30--Multiple regression equations selected for cross-validation. Qrogouts Sex R 13.1". Vari- Reg. Beta F Partial able coef. wt. cor. coef. Male .4620** 33 a -.O749 24 .0113 .1970 .0058 .1491 39 .0045 .0988 .7500 .0942 27 .0016 .0394 .2956 .0360 26 -.0060 -.1127 .0429 -.0834 36 .0029 .0547 .2312 .0505 30 .0037 .0924 .0844 .0893 23 .0156 .3712 .2649 .3558 Female .4655** 27 a 1.8715 42 .0105 .2199 .8215 .1718 9 -00070 -01-507 0 314’08 -0111? 31 .0009 .0163 .0066 .0157 35 -.oo46 -.0730 .0929 -.0586 25 .0122 .3189 2.3688 .2840 47 -.oo39 -.0628 .0616 -.o477 19 -.0063 -.1061 .1625 -.0773 45 -.0014 -.0276 .0105 -.0198 15 -.0016 -.0354 .0283 -.0324 34 .0086 .1409 .2766 .1007 11 -.0127 -.2672 1.5152 -.2305 44,01 level of confidence. 90 TABLE 4. 30--C ontinued . Acute concerns Sex 11 D.F. Vari- Reg. Beta F Partial able coef. wt. cor. coef. Male .4187* 33 a 2.4237 24 -.0020 -.0461 .0357 -.0329 39 -.0020 -.0564 .1054 -,o564 16 -.0002 -.0067 .0011 -.0059 27 -.0040 -.1342 .6532 -.1393 26 -.0011 -.0274 .0119 -.0190 35 -.0104 -.2053 1.3109 -.1955 23 .0104 .3322 2.8420 .2816 Female .5549** 27 a .2351 42 .0077 .2332 1.0463 .1931 9 .0057 .1763 .5283 .1385 31 .0065 .1769 .8899 .1785 35 .0139 .3171 1.9818 .2615 25 .0020 .0737 .1431 .0726 47 .0187 .4363 3.3621 .3328 19 .0010 .0250 .0102 .0194 45 .0015 .0407 .0260 .0310 15 -.0012 -.0377 .0365 -.0367 34 -.0167 -.3944 2.4544 -.2887 11 -.0044 -.1334 .4274 -.1248 *.05 level of confidence. **.01 level of confidence. 91 gzpotheseg Testing Hypothesis I: Selected MMPI and LYRI scales will predict Waldorf College freshmen who withdraw after the freshman year. The test of the hypothesis (H0 : R.= 0), as reported in Table 4.30, resulted in rejection of the null and acceptance of the alternative (HA : R # 0). Hypothesis II: Selected MMPI and EYRI scales will predict Waldorf College freshmen who have acute intrapersonal concerns, as measured by faculty judgment. The test of the hypothesis (H0 : R.= O), as reported in Table 4.30, resulted in rejection of the null and acceptance of the alternative (HA : a $ 0). It was concluded that the multiple regression coefficients would not have occurred by chance. - a at Table 4.31 will indicate the mean standard deviation, and correlation of the real I and predicted Y for each group. Female dropouts evidence the highest correlation, .31. Since only about 9% of the common variance is included in a correlation of .31, predictions based on such a low correlation could be e::pected to be frequently in error. The results attested the statements of walker and Lev, "A regression equation gives a satisfactory estimation for the group on which it was obtained if there is a high correlation between the observed and estimated scores.... It may give very close estimations for the group on which it was obtained and be much less effective with a new group," (walker and Lev, 1953, p. 321). Guilford has explained, "In a new sample from the same population, the weights would be likely to yield predictive values that correlate with the criterion less than the multiple R'in the sample from which the weights were obtained," (Guilford, 1954, p. 405). 92 When the multiple regression equation (obtained from the first half of the sample) was applied to the second half, the predicted Y identified correctly 22 of 27 drOpouts (81%) and 5 of 14 students (36)) with acute concerns. Incorrectly identified as drOpouts were 5 of 27 students (19%) and as acute concerns 9 of 14 students (64%). Table 4.32 will report the analysis in detail. TABLE 4.31--Mean, standard deviation and correlation of the real Y and the predicted Y for randomly dichotomized samples. S le Male (N = 41) Female (N = 39) Mean Std. dev. Mean Std. dev. Drapouts Y 1.6585 .4801 1.7436 .4427 Y 1.6585 .2218 1.7436 .2059 R .46 .47 Acute concerns ‘§ 1.853? .3578 1.8974 .3074 1.8537 .1498 1.8974 .1706 R .42 .55 §ample II Dropouts Y 1.6098 .4939 1.7180 .4559 Y 1.6550 .2131 1.7013 .2636 R .19 .31 Acute concerns Y 1.7317 .4486 1.8974 .3074 Y 1.8370 .1471 1.9532 .1818 R 93 TABLE 4.32--Analysis of correct and incorrect statistical identification of dropouts and acute concerns. Male Female Dropouts NondrOpouts Dr0pouts Nondropouts N 3% N 73 N g N 5 Y 16 100 25 100 11 100 30 100 Y 12 75 10 4o 10 91 14 47 4! 4 25 15 60 1 9 16 53 Acute Nonacute Acute Nonacute concerns concerns concerns concerns Y 10 100 31 100 4 100 37 100 Y 4 40 24 77 1 25 34 92 4f 6 60 7 33 3 75 3 8 SC cedu An alternative to the addition of variables based on factor analytic loadings was systematic subtraction of variables from a global regression equation that included all 49 selected variables. Because the matrix approached singularity with only 12 of the 49 variables accounted for in the CORE 4 Computer Program, an experimental multiple regression program (SCOPE R) was applied. Systematic reduction of variables will be reported in Table 4.33. The close approximation of number of subjects and number of variables made cross-validation of the resulting extended equations unfeasible. While of statistical interest, the cumbersome equations were not considered economical. TABLE 4.33-unltiple regression coefficients derived by SCOPE R 94 computer program. Drogouts Number of variables 2 used Variable number R R Male 49 1 ... 49 .5829 .7635 39 1,2...16,18,21...24,26,28...31. 330%9369379399u1000u9 057% 07579 34* 1,3...7,9...14,16,18,21...24,26, 280 0 0319339%936037’39’u10u3’u59 47...49 .5629 .7502 32 1 ... 32 (LYRI only) .3895 .6241 30 1.3...7.9...14,16,18,21,24,26, 29...31.33.37.39.41.43.45.47...49 .5276 .7264 25 1,3...7.9...ll,13,14,16,18,21,24, 26.29.31.37.39.41.43.45.47.48 .5224 .7228 21 1.3.7.8,10,ll,12,15 .1891 .4348 17 (MMPI only) 33 ... 49 .0991 .3148 Female 49** l ... 49 .7651 .8747 38** 1.2.3.7.8.10...13.15...19.21...25. 28...37.39...41.43...45.47...49 .6807 .8250 32** 1 ... 32 (LYRI only) .5866 .7659 27"”' l...3,7,8,10...12,15...l7,22,24,25, 29.00%936039000410M9u7 06515 08072 14 (MMPI only) 33 ... 49 .3231 .5684 *Overall F significant at .05 level. **Overall F significant at .01 level. 95 TABLE 4. 33--Continued. L - Acute concerns Number of variables 2 used Variable number R R Male 49 1 ... 49 .6344 .7665 38* 10001691900025'2700033937039. 41,42,43,46,47,49 .6001 .7747 32* 1 ... 32 (LYRI only) .5069 .7120 23** 1.2.7...9.11.14.19...23.25.27.28. 30.32.37.39.41.42.47.49 .5633 .7505 20 1,2,7...9,11,l4,l9...23,25,28,30, 3203794101420“? 05512 0742“ 17 (MMPY only) 33 ... 49 .1853 .4304 16 l,2,7,8,ll,14,19,21.23.25,30, 32.37.41.42.47 .5181 .7198 Female 49 l ... 49 .7220 .8497 32 1 ... 32 (LYRI only) .5018 .7084 26** 1...3,ll,12,14...16,l8...22,27,28, 30.35.35.38.39.41.44...46.48.49 .5401 .7349 14 (MMPI only) 33 ... 49 .2026 .4502 I"Overall F significant at .05 level. **Overall F significant at .01 level. 96 LYRI Release and Nonrelease Samples Compared on the_Mfl£l_ Application was made of the "t" test of the mean difference between students who released LYRI scores and the students who did not sign releases. For each of the 17 MMPI scales reported in Table 4.34 "t" tests indicated differences at the .01 level of significance. Students who did not release LYRI results consistently scored higher on the MMPI. The degree that height on the MMPI scales indicates a tendency toward maladjustment may be related to anxiety about releasing highly personal data. The testing of religious beliefs may have been felt by some as to be an area to be guarded from empirical investigation. Additional research with the total nonrelease sample included may have heightened the results of the current study inasmuch as such deviations from the mean could counteract the central tendency of multiple regression pro- cedures. The data presented graphically in Figure 1 will be documented in Appendix G. mOHSO HOE OOOO NO 8 co .8 OOOO O O O O O O N H ’5'." ‘.’ ', I I u ”mamaom "mama .meamswm ommoamncoc paw ommoaou Hmwu mo aces momspon coconommflona H enemak w: on NO 991009 1 IdMN 98 Mean Differences between Criterion Groups and Normal Groups The mean difference on the MMPI between dropout, acute concerns, and nondropout-nonacute concern students will be reported in graphic form in Figure 2. The profiles followed generally similar patterns. Marked similarity was observed on scale 6 (Paranoia) and St. (Social status). Most divergence was observed on scales Re(Social ReSponsi- bility) and Cn (Control). 8 OO .E em on 8 mm ‘.l1ll'|‘.-'l | I ‘lllluiaw'll II x II x l “Eoocoo I "geomonn I I I I I I "mpmw>mpcoz .mcnmosoo mpsom new .pdomoup moOecOHeOa O O O O O O O O N H I. I- ,0 cl--- ’5’I‘I'E‘Il' ‘. 1'4- mam: .opma>opzoc newspop neoconomme cmozwI m onsmfle ---,-.‘- 0—... 88JOOS L IJNN 100 :0 ‘Ir'...’ pm pm In I... --I...I'-.D“J' 'tl-" -A'Ivl It'll}... '061’1' ‘t’u’ OOHOom HOE em 8 OO 8 O O O O O O O ON H |- I. ’t..- 'l‘l'lllll ‘ ‘IDII.--’- "C'!'-'l‘|"‘ "cnmouoo 6\///\. ”psomonn “mpma>wmcoz ofimemm .OOSQOpmooII N mgowflu C~mmwn.mm~ . fl ‘ m N \O saloos 1 1mm 101 SUMMARY OF THE ANALYSIS OF THE DATA The analysis of the data was detailed in this chapter. The subjects were identified. The results from the instruments used were reported and interpreted. The hypotheses were tested and all statistical procedures eXplained. CHAPTER V SUMMARY, CONCLUSIONS AND IMPLICATIONS SUM/TAM OF THE STUDY The purposes of the study were stated in two dimensions: 1. To investigate factors that relate to waldorf Junior'College freshmen who withdraw after the freshman year and who have acute intrapersonal concerns. 2. To test two major hypotheses: a. Selected scales from the Minnesota Multiphasic_Personali§y Inventogy can be used to predict Waldorf College freshmen who withdraw after the freshman year and who have acute intrapersonal concerns, as measured by faculty judgment. Four instruments were used to gather data derived from the 1963- 1964 freshman class at Waldorf College, Forest City, Iowa. The first instrument was a questionnaire designed to elicit demographic and attitudinal data. The second was the SAT. The third and fourth instruments were the MMPI and LYRI. The design of the study was divided into two dimensions: 1. Analysis of the questionnaire and ACT by the use of the chi square statistic. 2. Analysis of the MMPI and IXRI by the use of factor analysis, multiple regression and cross-validation. 102 103 CONCLUSIONS Demographic and Attitudinal Data Since the chi square statistic is mainly a measure of independence between nominal variables, it can not be assumed that a cause and effect relationship was indicated when certain variables did not exceed the .05 and/or .01 level significance. The presence of a relationship not due to chance was established but not the magnitude or direction of the relationship. The product-moment correlation coefficient provided additional information relative to the magnitude of significant rela- tionships. The correlations that appeared significant at the .05 level or higher for both the x2 and r include the following: 1. The No Information response category was most often responsible for contributing to the chi square cell which accounted for the pre- selected level of significance. Perhaps the limitations of a mailed questionnaire, combined with the nature of the respondents of interest (drOpouts and acutely concerned students), tended to limit responses from approximately one-half of the males and one-third of the females. The questionnaire did not evidence sufficient discrimination when attitudes with their associated intensity were measured in relation to selected campus subgroups. 2. The primary period of withdrawal for both dropouts and acute concerns appeared to be after the first full year in college. waldorf College tended to retain such students during the freshman year. 3. Female drOpouts and.males with.acute concerns have reported most precollege academic mobility. Clinical eXperience indicates that 104 one attraction of a small, residential junior college is its reputation for having a friendly family atmosphere. It appeared that some students who have moved much during their elementary and secondary school years have special needs. 4. Students of both sexes and both groups of interest seemed to want to take full responsibility for the decision to leave Waldorf College. Male drOpouts accepted responsibility for matriculation at Waldorf College, as did the other groups though the latter was without evidence of statistical significance. 5. The person most influential in encouraging male drOpouts and acute concerns to attend Waldorf College has been the parish pastor. For females, parents were the primary influence. Clinical eXperience tends to confirm the result. Earnest pastors and anxious parents seem to look to Waldorf College to assist them in meeting the needs of their students. Possibly parents and pastors were aware of special needs or the potential for withdrawal of their students. 6. The questionnaire did not discriminate adequately the percep- tions of parents on any of the six scales constructed to measure the students' estimate of their parents. Perhaps the complexities of parent-child relationships preclude measurement by even more saphisti- cated contemporary methods. 7. Males with acute concerns apparently have perceived their home congregations as being of a more formal than informal type. Perhaps such students perceive formality in worship to be related to their needs. The kind of relationship is not reported by this investigation. 105 8. Feelings toward or from fellow students were not measured with adequate discrimination by the questionnaire prepared for this study. The single exception to the previous conclusion was the estimate of the closest friends' feelings about the males with acute concerns. A tendency toward warmth was indicated. Perhaps the males with acute concerns not only found friendship they needed at Waldorf College, but perceived such friendship more positively than did the other groups of interests to the study. ACT lationshi s The measurement of college aptitude was not of primary interest to the investigation. Since other admissions data did not evidence sufficient precision for the purposes of the study, only the ACT was reported. The English test appeared most frequently to be useful for the purpose of the study. The Natural Science test ranked second. The Social Science test did not appear to be useful. - YRI Observation of the contribution of the two most useful variables in each multiple regression equation, based on the partial correlation coefficients, permitted the following conclusions: 1. Male Drapouts: Variable 23, LYRI Self-realization (-r) Variable 24, LYRI Achievement (-r) 2. Female Dropouts: Variable 25, LYRI Service (-r) Variable 11, LYRI Academic problems 3. Male Acute Concerns: Variable 23, LYRI Self-realization (-r) Variable 35, MMPI Hypochondriasis 4. Female Acute Concerns: Variable 47, MMPI Prejudice (-r) Variable 34, MMPI Depression 106 Itlappeared that from among the two most useful variatles in each equation the LYRI contributed five variables and the MMPI three variables. When all of the variables in each of the four equations were considered, approximately two-thirds were contributed by the LYRI. Perhaps the LYRI could be expected to measure more precisely a predominantly Lutheran pOpulation than could the NMPI inasmuch as the LYRI was developed for this purpose and was based on Lutheran normative samples and the MMPI was developed on a clinical, not a "normal”, population. The testing of the hypotheses generated for this study indicated that, in fact, the multiple regression coefficient was not equal to zero. There was a difference not due to chance between the deviate groups of interest to this study and their more normal counterparts. The multiple regression coefficient derived from study of the first half of each dichotomized group was highest for female acute concerns (R = .55) and lowest for'male acute concerns (R = .42). Male dropouts (R = .46) closely approximated female dropouts (R = .47). When the multiple regression equation derived from the first half of each group was applied to the second half of the respective group, the multiple R drOpped in each case. Researchers have termed such a phenomena statistical shrinkage, (Guilford, 1954, p. 405). Because of the shrinkage in this experiment the possibilities for accurate pre- diction may be minimal. Correctly identified were 81% of the dropouts and 36% of the students with acute concerns. It can not be concluded that similar results would be obtained with other samples. 107 LYRI Non-release Sample Students who elected not to release LYRI scores were consistently higher on each of the 17 MMPI scales selected for this study. The departure from the mean may indicate tendencies toward pathology if the group is considered as a whole. It should not be concluded that a given student who did not sign a release statement was less well adjusted than the student who did sign. Differences between Criterion Groups and Nondrppouts-nonacute concerns While the profile configuration for criterion groups generally followed that of the "normal“ comparison groups, most divergence was reported on the Social Responsibility (Re) and Control (Cn) scales of the MHPI. It appeared that clinical experience confirmed the pissibi- lity of dissimilarities at the points previously mentioned. General Conclusions Three general conclusions were formulated: 1. Responses from females were generally more frequent than from males. The female sample size was slightly less than the male but the ratio of responses, with some exceptions, indicated relatively more returns from females. Perhaps females found it easier to conform to academic procedures than did the males. Possibly male students were more resistant to revealing inner attitudes. 2. Responses from drOpouts were generally larger than from the students judged to have acute concerns. It was noted that most Waldorf dropouts transferred to another college. They seemed to be interested in academics but not at waldorf College. The acutely concerned students may have felt somewhat defensive about themselves. The nature of their 108 concerns may have mitigated against responding to a study of areas which may be painful. In some cases the study itself may have contri- buted to additional hurt. 0n the other hand, some students with acute concerns may have been encouraged that others were interested in their concerns. 3. 0f the dropouts who had acute concerns the faculty correctly identified 61%. If it can be assumed that students with acute concerns tend to become dr0pouts, faculty judgment of students with acute concerns more correctly than incorrectly identified potential drOpouts. The ratio was six successful identifications to four unsuccessful identifications. It.may be possible to conclude that, in general, the Waldorf faculty was sensitive to students who evidenced acute concerns. It is noted that transfers are included as Waldorf drcpouts. Thus comparison with national drapout studies may need qualification. IMPLICATIONS Implication means here the involvement or close connection of the conclusions with future possibilities. Eight implications will be drawn from.the results of this study: 1. Perhaps personal interviews would serve better than a mailed questionnaire to elicit more discriminating responses. 2. Since it is the function of a college to serve the students it admits, accepting responsibility for>meeting student needs wherever need exists, it seemed that Waldorf College approximated its goals in the area of interest to this study. The college, knowing that students with drOpout potential tend to remain for the full college year, may wish to concentrate especially on the personnel needs of the freshmen. 109 3. Weldorf College personnel workers may wish to observe early in the school year which students have experienced high pro-college academic mobility. How to identify precisely and then meet such needs demands further study. 4. The tendency to want to take full responsibility for with- drawal from Waldorf College may indicate a healthy maturation. In some cases it may also indicate a hesitation to state a projection of responsibility for various defensive reasons. 5. The relatively high degree of involvement of the parish pastor with potential dropouts may tend to encourage close professional rela- tionships between personnel workers and pastors. The relatively high degree of involvement of parents with acutely concerned students may tend to encourage close contacts between personnel workers and parents. Possibly the increasing practice of parent conferences in ungraded elementary schools could be attempted in ways that would not minimize the emerging adult status of the college freshman. 6. The ACT English test may assist Waldorf personnel workers in estimating the prepensity for withdrawal and acute concerns of female students, especially if the ACT is the only standardized instrument available. Caution would need to be exerted in using such limited information for a purpose for which it was not originally intended. 7. If economical data processing were available, the use of an extended multiple regression equation might assist personnel workers in the identification of students with potential for withdrawal and with acute intrapersonal concerns. Workers with time and energy limitations may find such a procedure valuable in permitting statistically 110 supported decisions. Some students could be offered special assistance in meeting their nonacademic, more affective needs. 8. It appeared that females were more responsive to the research procedures than males. Perhaps some females tend to conform too much. Perhaps the nonreSponding students should rate a sincere salute for resisting. It seems easier for some traditional educators to appreciate students who conform to existing patterns. Creative nonconformity tends to be suspect at times. {Igblggs f0; Furthe: §tu§z Along with the further research suggestions previously implied, the following five are suggested: 1. Some students may be encouraged by a college to remain when, for the students' best welfare, they should better become drapouts. Not every student is served well by a given college. Not every high school graduate can profit by a college education. As a guarantor of social acceptability. economic success, and human dignity the value of a college education appears overestimated. The student who withdraws usually does not receive the same continuing attention from the college as the student who returns though both are considered alumni for pur- poses of public relations. Perhaps ways can be found to counteract the notion that withdrawal is equated with failure. The cause and effect relationship is often complex. 2. Some educators feel that the adjustment of freshmen from high school to college is severe enough to warrant placing the very best teachers in freshman classrooms. Such teachers might be sensitive to 111 the teaching methodology contemporary students find comfortable and creative. Little gain is experienced from lamenting the inadequate preparation of the students a college receives. Better would be study of ways to find out who the students are and what motives and study skills they possess. Insight into the student's frame of reference is required. He needs to be understood, not merely classified. 3. A school needs to continually redefine its purposes. The catalog statements of purpose need to be vivified and Operationalized each year. If meeting individual differences is high on the scale of stated purposes, adequate budget would be eXpected to help provide individual instruction in pursuit of the basic academic goals of the college. It is the view of this researcher that the goal of individual instruction (education in the most personal sense) can be approximated by an effective student personnel program that includes competent, available personal counseling. The literature previously cited has suggested that junior colleges could serve uniquely in the complex of educational Opportunities in America if they would budget more adeQuately for their total personnel programs. Thus the problem being stated here is not so much one of research methodology but of over-arching philosOphy of education and curriculum understandings. The best of theories finally require finances that will permit workers to act on the ideas. a. The current study could be replicated to confirm or deny its results. 5. Five related studies might enrich the results of this study: a. The college maladjustment scale (Mt) deve10ped by Klein- muntz for use with the MMPI might discriminate more accurately than any of the scales selected for use in this study. (Kleinmuntz, 1961). 112 b. The college achiever and nonachiever scale (Ac) deve10ped by Altus for the MMPI might discriminate more accurately than any of the scales selected for use in this study, (Altus, 1948). c. The study of institutional or environmental press might indicate factors which contribute to student emotionality, factors for which the college and not the student is reSponsible. Perhaps such an investigation could be incorporated in periodic institutional self- studies. d. The influence of peers seems to be more significant than this study has been able to determine. Since self-image appears related to social acceptance, the dimension of peer relationships could be profitable. It is possible that some students withdraw because they have obtained the academic or personal satisfactions they sought. ReSpect for personhood tends to obviate the imposition of institutional limitations. e. Cross-sectional multi-group and sub-group comparisons could be made. For example, commuters compared with noncommuters. family composition. and high school grade average might be used. MMPI and LYRI profile configurations based on various sub-group samples could be compared. SUMMARY OF THE INVESTIGATION This investigation was interested in the description and possible early identification of Waldorf Junior College freshmen who had a potential for withdrawal and who were judged by the faculty to have acute intrapersonal concerns. 113 The four instruments used were described and statistical procedures detailed. The results of the study have been reported and conclusions stated. Possibilities for further research have been suggested. IEFE IENCES 111+ REFERENCES Altus, w. D. A college achiever and nonachiever scale for the MEPI. JI a221, Pszphol., 1948, 32, 385-397. Anderzon, w. The MMPI: Low 23 scores. J counsel Ps chol., 1956, 3, 22 -228. Ball, J. C. §ocial deviancy and adolescent personality. Lexington: Univers. of Kentucky Press, 19 2. Ball, J. C. & Carroll, Donna. Analysis or MMPI cannot say scores in an adolescent pepulation. JI of clinI ngchol., 1960, 16, 30-31. Bendig, A. w. & Klugh, H. E. A validation of Gough's a; scales in predicting academic achievement. EducII psychol. neasmt., 1956, 16, 516-523. Hillard, R. G. Comparison of teacher ratings of personality with results of an incomplete sentence blank. PersI & GuidI J., 1962, 41, 9, 58. Blaine, G. B. & McArthur, C. C. Emotional problems of the student. New York: Appleton-Century Crofts, 1961. Bogus, Jesse. The community college. New York: McGraw-Hill, 1950. Brown, H. S. Similarities and differences in college papulations on the MMPI. JI a221, zgyghol., 1948, 32, 541-549. Buros, O. K. The fifth mental measurements yearbook. Highland Park, N. J.: Gryphon Press, 1959. . Clark, J. H. Grade achievement of female college students in relation to nonintellective factors: MMPI items. JII socI Egychol., 1953, 37 9 275 -281 e Cronbach, L. E. Essentials of psychological testing. (2nd ed.) New York: Harper & Row, 1960. Dahlstrom, w. G. & welsh, G. S. An figZI handbook; a guide to use in clinical practice and reseaggh. Univers. of Minn. ress, 1960. Drake, L. E. MMPI profiles and interview behavior. counsel. We. 195“! 19 92-950 Drake, L. E. MMPI patterns predictive of underachievement. J, counsel. Eszcholu 1962. 9, 164-167. Drake, L. E. & Getting, E. H. An MMPI pattern and a suppressor variable predictive of academic achievement. g, counsel, Psychol., 1957. 4. 245-247. 115 116 Drake, L. E. & Getting, E. H. An MMPI codebook for counselors. Univers. of Minn. Press, 1959. Drasgow, J. & McKenzie, J. College transcripts, graduation, and the MMPI. J, counsel, Psychol., 1958, 5, 196-199. Edwards, A. L. Statistical methods for the behavioral sciences. New York: Holt, Rinehart & Winston, 1954. Erikson, E. H. Childhood and society. New York: Norton, 1950. Frick, J. w; Improving the prediction of academic achievement by use of the MMPI. JI a221, zgychol., 1955, 39, 49-52. Froehlich, C. P. & Hoyt, K. B. Guidance testing. Chicago: Science Research Associates, 1959. Funkenstein, D. H., et a1. Differences on measures of self-concept. Diseases of the nervous system, 1959, 20, 3-8. Gibbs, D. N. Student failures and social maladjustment. ers & Guid. J., 1965. 45. 580. Glasscock, E. M. An investigation of the value of the MMPI as a prognostic instrument. Unpublished doctoral dissertation, wash. Univers., 1955. Gleazer; E. J. 1264 Junior college diggctogy. washington, D.C.: Amer. Ass. Jr. 001., 19 4. Goodstein, L. D. Regional differences in MMPI responses among male college students. JI cgnggltII Esygho;., 1954, 18, 437-441. Gottesman, I. I. More construct validation of the ego strength scale. J. consultII Psychol., 1959, 23, 342-346. Grace, H. A. Personality factors and college attrition. Peabody We 195?: 35! “6'50. Guilford, J. P. Psychometric Methods. (2nd ed.) New York: McGraw- Hill, 1954. Hackett, H. R. Use of the MMPI to predict college achievement. JI counsI Psychol., 1955, 2, 68-69. Halroyd, R. G. Prediction of defensive paranoid schiZOphrenics using the MMPI. Digs. Abst., 1964, 25, 2048. Hampton, P. J. The MMPI as a psychometric tool for diagnosing personality disorders among college students. 1, soc. Egychol., 1947, 26, 99-108. 11.7 Harmon, H. Modern factor analysis. Chicago: Univers. of Chicago Press, 1960. Hartshorne, H. & May, M. A. Studies in the organization of character. New York: Macmillan, 1930. Hathaway, S. R. 8: McKinley, J. C. MIQI manual. New York: Psychol. Corp.. 1951. Hathaway, S. R. a Meehl, P. E. An atlas for the clinical use of the MMEL. Minneapolis: Univers. of Minn. Press, 1951. Hathaway, 3. R. & Monachesi, E. D. An atlas of juvenile MMPI profiles. Minneapolis: Univers. of Minn. Press, 19 l. Hathaway, S. R. & Monachesi, E. D. Adolescent personality and behavior. Minneapolis: Univers. of Minn. Press, 19 3. Hawkes, G. R. Use of the MMPI in screening college students for counseling purposes. JI educ. £§ychol., 41, 116-121. Hays, w. L. Statistics for S‘ChOlO ists. New York: Holt, Rinehart, & Winston, 1964. Heilbrun, A. B. Personality differences between adjusted and maladjusted college students. JI a221, Egychol., 1960, 44, 341-346. Heilbrun, A. B. Revision of the MMPI K correction procedure for improving detection of maladjustment in a normal population. JI consult, Psychol., 1963, 27, 161-165. Hoyt, D. P. & Norman, w. T. Adjustment and academic predictability. JI counsI ggycho;.. 1954, 1, 96-99. Iffert, R. E. figtention and withdrawal of college students. washington, D.C.: Dept. Health, Education, and welfare, 1958, 1. Jensen, V. H. Influence of personality traits on academic success. Pers, GuidI J., 1958, 36, 497-500. Johnson, B. L. Footnotes on the junior college. [hi Delta Kappan, 1965. #6. 376-380. Johnson, H. & McNeal, B. F. Combined MMPI and demographic data in predicting length of neuropsychiatric hospital stay. JI of consult. ngchol., 1964, 28, 64-70. Jurjevich, R. M. Normative data for the clinical and additional MMPI scales for a papulation of delinquent girls. J en cho ., 1963. 69. 143—1u6. 118 Kerlinger, I. N. Foundations of behavioral research. New York: Holt, Rinehart & Winston, l9 . King, 3. H. Emotional problems of college students. Amer, Ass. Univers. Professors Bulletin, 1964, 50, 327-332. Kleinmuntz, B. Identification of maladjusted college students. J, counsI Psychol., 1960, 7, 209-211. (a) Kleinmuntz, B. An extension of the construct validity of the ego strength scale. J, consult. Psychol., 1960, 24, 463-464. (b) Kleinmuntz, B. The college maladjustment scale (Mt): norms and predictive validity. Educ, psycholI Measmt., 1961, 21, 1027-1033. Kleinmuntz, B. Annotated bibliography of MMPI research among college populations. JI counselI Psychol., 1962, 9, 373-396. Kleinmuntz, B. Personality test interpretation by digital computer. Sciencg, 1963, 139, 416-418. Kraepelin, E. Der psychologische versuch in der psychiatrie. Psychol. Arbeit., 1959, 1, 1-91. Cited by Anne Anastasi, Psychological testing. New York: Macmillan, 1961, 9. Krathwohl, D. R., Bloom, B. 8., & Masia, B. B. Taxonpmy of educational_ objectivesI handbook II; affective domain. New York: David McKay Co., Inc., 19 4. Lang, G., Sterra, A., & Seymour, Marjorie. Psychological needs of college freshmen and their academic achievement. EersII & GuidI J., 1962, 41, 359-360. Layton, w. L. Predicting success in dental school. JI applI Psychol., 195“: 379 634-660, Mann, w. A. Validation of the Cornell index for freshmen at Michigan State College. Unpublished Ph. D. dissertation, Mich. State 0011., 1949. McDaniel, J. w. Essential student pgrsonnel practices for junior colleges. washington, D.C.: Amer. Ass. Junior Colleges, 19 2. McKenzie, J. D. An attempt to develop MMPI scales predictive of academic over-achievement and under-achievement. DissI Abst., 1962, 22, 632. McKenzie, J. D. The dynamics of deviant achievement. zers, & GuidI J., 1964, 42, 683-686. . 119 McNeeley, J. H. College student mortality. U, S, Dept, Interior Bulletin 11, 1937. McQueen, Mildred. College withdrawals. Chicago: Science Research Associates, 1964. Matson, Jane E. Characteristics of students who withdraw from a public junior college. Unpublished doctoral dissertation, Stanford Univers., 1955. Medsker, L. L. The junior college: progress and prOSpect. New York: McGraw-Hill, 19 0. Medsker, L. L. The unior colle e: roaress and 5 act. New York: McGraw-Hill, 1960. Cited 5; B. Schwartz, Is it really higher education? Satupggy Review, Dec. 19, 1960, p. 53. Merlo, F. P. The burgeoning community college. Saturdgy Review, Deco 199 19649 50-5“. Mosier, C. I. Symposium: the need and means of cross-validation. I. Problems and designs of cross-validation. Educ, psychol. Measmt., 1951, 11, 5-11. Parker, C. A. Letter to the editor. J, couns, Psychol., 1961, 8, 88-89. Parker, C. A. The predictive uses of MMPI in a college counseling center. J, couns, Psychol., 1961, 8, 154-158. Patterson, C. H. Selection of rehabilitation counseling students. Pegg, & Guid, J., 1962, 41, 322. Quinn. S. B. Relationship of certain personality characteristics to college achievement. Diss, Abst., 1957, 17, 809. Reader's Guide to Periodical Litepature. New York: The H. W. Wilson, 19 . Ridlor, H. G. Why freshmen fail. Atlantic Monthly, 1961, 208, 56-60. Roessel, F. P. MMPI results for high school dropouts and graduates. Unpublished doctoral dissertation. Minnesota, 1954. Sanders, M. W. The predicting of academic success among university freshmen in a school of education. Unpublished doctoral disserta- tion, NYU, 1950. Sanford, N. (Ed.) The American colle e: a cholo ical and sociolo ical inteppretation of higher education. New York: John Wiley, 19 2. 120 Sarbin, T. R. Clinical inference and cognitive theogy. New York: .Holt, Rinehart, & Winston, 1960. Seashore, H. Academic abilities of junior college students. Junior College J 1958, 75. _______‘.__ov Seegars, J. E. Further investigation of an MMPI scale for predicting college achievement. Pers, & Guid, J., 1962, 41, 251-253. Silver, R. J. MMPI characteristics of a state hospital pOpulation. J, clin, Psychol., 1961, 17, 142-146. Starr, J. M. Guidance practices in selected junior college guidance programs. Junior College J., 1960, 30, 489. Stone, D. R. & Ganung, G. R. A study of school achievement related to personality as measured by the MMPI. J, educ, Res., 1956, 50, 155-156. Strommen, M. P. A comparison of youth and adult reactions to Lutheran youth problems and sources of assistance. Unpublished doctoral dissertation, Univers. Minn., 1960. Strommen, M. P. Profiles of chupch youth. St. Louis: Concordia, 1963. Tead, O. The junior college contribution. Harvard Educational Review, 1954, 14, 124. Time, 85, 22 & 60. Chicago: Time & life, 1965. Thorndike, R. L., & Hagen, Elizabeth. Measurement and evaluation in psychology and education. New York: Wiley, 1955. Torrance, E. P. Changing attitudes concerning academic discipline. Unpublished manuscript. Walker, Helen M., & Lev, J. Statistical inference. New York: Holt, Rinehart, & Winston, 1953. weigand, G. Motivational factors associated with success and failure of probational students. Unpublished doctoral thesis, Univers. of Maryland, 1951. Williamson, E. G. The student personnel point of view. Washington, D.C.: Amer. Council on Education, 1949. Williamson, E. G. An historical perSpective of the vocational guidance movement. Peps, & Guid, J., 1964, 44, 858. 121 Woolf, M. D. & Neolf, Jean A. The student personnel program. New York: McGraw-Hill, 1953. Zeaman, Jean B. Some of the personality attributes related to achievement in college: a comparison of men and women students. Diss, Abst., 1958, 18, 290-291. APPENDIX A DRQPOUT DATA 122 123 DROPOUT DATA Reasons for Withdrawal (Questionnaire Item 76) full-time employment 4 disinterest 2 health 2 dissatisfied 16 moved residence 1 financial 8 transferred schools 22 marriage 3 academic (grades) 7 educ. goals completed 2 Educational Locale of Waldorf Drppouts (Item 75) Allen Memorial School of Nursing Mayville State Col. American Institute of Business Minneapolis School of Art Augsburg Col. Minn., U. of Augustana Col. Moorhead State Col. Baylor U. Muskegon Jr. Col. Capital U. Nettleton Co. (Business) Colorado State Col. Pacific Luth. U. Dana Col. Pharmacy School (unidentified) Ellsworth Jr. Col. St. Cloud State Col. Iowa Luth. Hosp. School of Texas Luth. Col. Nursing Valparaiso U. Iowa State U. wartburg Col. Luther Col. Wis. State U. Mankato State Teachers Col. Summary: Liberal arts colleges: 20 Trade schools: 6 Monthly Income Reported by DrOpouts: Range: $100 - $800 (child care - road construction) Mean: $305 Mode: $250 APPENDIX B DEMOGRAPHIC AND SOCIO-ECONOMIC DATA Geographical Distribution Denomination Rank-in-Class Size of Class ACT Scores Educational Background of Parents Occupational Background of Parents 124 125 Geographical gistribution Iowa 110 40% Montana 2 43% Minnesota 70 24 Texas 1 -1 Wisconsin 18 7 Michigan 1 -1 Illinois 10 4 Colorado 1 -1 Ohio 10 4 Kansas 1 -1 South Dakota 7 3 New Jersey 1 -1 Hawaii 5 2 Nebraska 1 -1 California 4 2 EthiOpia l -1 North Dakota 4 2 Philippines 1 -1 New York 3 1 Norway 1 -l washington 3 1 Tanganyika l -1 Qpnomination mum W American Lutheran Church 210 82 Other Lutheran 18 7 Methodist 12 3 Baptist 3 l Congregational 3 1 Evangelical Free 2 -1 Presbyterian 2 -1 Reformed 2 -1 Christian 1 -1 Coptic 1 -1 Roman Catholic 1 -1 Not given 1 -1 Rank-in-Clasp was ans—her W 91 - 100 22 9 81 - 9o 19 8 71 - 80 28 11 61 - 70 15 6 51 - 60 25 10 41 - 50 14 6 31 - 40 23 9 21 - 30 36 15 11 - 20 25 10 0 - 10 11 4 Unknown 27 12 126 Size of Class Number Percent of Total Less than 25 16 6 26 - 50 55 22 51 - 75 35 14 76 - 100 50 19 101 - 150 19 7 151 - 200 5 2 201 - 300 17 7 Over 301 34 14 No record 24 9 ACT Scopps Frequency distribution of entering freshmen taking the ACT, using standard composite scores and national 12th grade percentiles. National Number of 12th Grade Standa co Students Percentile 31 1 99.9 30 2 99.5 29 5 99.0 28 3 98.3 27 3 97.0 26 5 96.0 25 20 95.0 24 17 92.0 23 12 89.0 22 8 85.0 21 16 80.0 20 13 75.0 19 21 70.0 18 16 64.0 17 16 57.0 16 10 49.0 15 13 42.0 14 9 35.0 13 17 28.0 12 13 22.0 11 5 17.0 10 7 12.0 9 l 8.0 127 Educational Background of Parents Education Fathers Mothers Less than 8th grade 7 2 Grades 8 - ll 84 51 H.S. Graduate 78 105 Some College 38 59 College Graduate 12 17 Post College 26 7 Occupational Background of Parents General Categogy Number Professional 28 Technical 5 Managerial 16 Clerical 3 Sales 32 Service 16 Agricultural 109 Mechanical 11 Manual 23 Military 2 Retired _ 8 APPENDIX C COMPARATIVE MMPI DATA: from MICHIGAN STATE UNIVERSITY FIESHMEN and WALDORF JUNIOR COLIEGE FmSI‘H‘lEN 128 129 A COMPARISON OF MSU FIESHMEN (1964 RANDOM SAMPIE) AND WALDORF FRESHMEN (1963 CLASS) 0N SELECTED MMPI SCALES Male Female MSU Waldorf MSU Waldorf marl Std. Std. Std. Std . scale Mean dev. Mean dev. Mean dev. Mean dev. HS 54 3.71 52 8.81 52 3.18 49 6.88 D 56 4.87 52 11.34 53 4.98 51 7.43 Hy 58 4.71 54 7.07 56 4.07 52 6.89 Pd 57 4. 25 58 9 . 36 57 4. 10 54 8. 25 Mr 61 5.06 57 9.17 49 4.53 49 8.97 Pa 56 2.94 56 7.75 56 2.95 55 7.88 Pt 60 5.34 61 10.99 56 4.64 58 7.23 Sc 61 6.32 60 11.74 58 5.31 56 7.25 Ma 58 4. O9 59 ll . 4O 58 3. 94 56 9 . 56 Es 48 5.59 54 9.75 46 5.14 57 8.92 N = 260 N = 101 N = 205 N = 88 APPENDIX D QUESTIONNAIRE 130 131 MICHIGAN STATE UNIVERSITY East Lansing College of Education 262 November 11, 1964 TO: waldorf College Freshmen, Class of 1963-64 The development Program at Waldorf College comes alive in our awareness as we view the erection of several needed buildings. A similar sense of achievement can be ours as we assist in the study of student feelings as you experienced them last school year. You can help serve the freshmen who follow you at Waldorf by completion of the enclosed questionnaire. It takes about five minutes. Some items may seem quite personal or irrelevant. Be assured that I will keep your responses confidential. Your careful comple- tion of each item will prove valuable in the preparation of my doctoral dissertation. A self-addressed, stamped envelope is enclosed to invite an immediate reply. May I count on your assistance? Sincerely, Dar Roa Directions: 1. Circle the ppg response for each item that most applied to you WHEN YOU WERE A FRESHMAN at Waldorf College. 2. Responses can be clarified and expanded on the margins, if you desire. Please be brief, precise, and complete. 3. Please do not discuss your responses with any one before completing the whole form. 4. This questionnaire does not require your name. 5. The numbering of items corresponds to computer use. 6. Please return immediately. Thank you sincerely! Mail to: Dar Roa (The Rev.) 810 D Cherry Lane Michigan State University East Lansing, Michigan 48822 132 CIRCLE YOUR.CHOICE 1'25 4 My residence before my freshman year at waldorf was: 1 — 99 100 - 999 1000 - 2499 2500 - 24.999 25,000 Up Roral Village Town City Large city 10 11 22 23 24 My family's residence during my freshman year at Waldorf changed: Not at all Once Twice Three times Four or'more times The number of different schools I attended before Waldorf (all grades): One Two Three Four Five or more I place my family in the following social class: Lower Lower middle Middle Upper’middle Upper My family's income, in relation to the family income of my friends, waitich higher Higher Similar Lower Much lower The decision to enter Waldorf College was primarily: My own My father's My mother's Both parents Mutual As py_choice of a college, waldorf rated: First Second Third Fourth Fifth or below I was influenced to consider Waldorf primarily through: Parent(s) Guidance counselor Waldorf admissions counselor Pastor(s) Waldorf alumnus Another college's referral Relative(s) Waldorf teacher Friend Teacher(s) Waldorf administrator My general health during my freshman year at waldorf was: Very good Fair Poor (Please specify: ) My ability to read at the beginning of my freshman year was: Very good Good Fair Poor _ Developmental (remedial) study during my freshman year included: English Math Reading None Eng1.+Math Eng1.+Reading 133 25 My class attendance during my freshman year at Waldorf was: 27 28 29 30 31 Very regular Regular Off and on Irregular Seldom 90% 75% 50% 2572 1575 26 My chapel attendance during my freshman year at Waldorf was: Very regular Regular Off and on Irregular Seldom 90% 75% 50% 25% 15% I characterize my parents as being, in relation to me: Very liberal liberal neither conservative very conservative Very distant distant neither close very close Very warm warm neither cold very cold Very rigid rigid neither flexible very flexible Very unhelpful unhelpful neither helpful very helpful Very religious religious neither nonreligious antireligious 33 35 36 37 38 39 Comments, if any: I characterize my home congregation as being, in relation to me: Very helpful helpful neither unhelpful very unhelpful Very conservative conservative neither liberal very liberal "High church" formal neither informal "low church" Very warm warm neither cold very cold Very distant distant neither close very close Very Haugean Haugean nonHaugean antiHaugean don't know Comments, if any: Very many many a few very few As a freshman, I felt I could consider as my friends on the Waldorf campus: none no 41 42 43 4h 45 46 1+7 48 1+9 50 51 52 53 52+ 55 56 57 58 59 60 134 As a freshman, ;_£gl£_in relation to my_classmates at waldorf: Very comfortable* comfortable neither uncomfortable very uncomfortable Too religious religious neither nonreligious antireligious Very friendly friendly neither unfriendly very unfriendly Very helpful helpful neither unhelpful very unhelpful Very cold cold neither warm very warm As a freshman, 2,22lt.in relation to my.closest gampg§.friends: Very comfortable* comfortable neither uncomfortable very uncomfortable Too religious religious neither nonreligious antireligious Very friendly friendly neither unfriendly very unfriendly Very helpful helpful neither unhelpful very unhelpful Very cold cold neither warm very warm As a freshman, my closest campus friend was my room-mate: Yes No In relation to me, my.classmates during my freshman year 3222: Very comfortable* comfortable neither uncomfortable very uncomfortable Too religious religious neither nonreligious antireligious Very friendly friendly neither unfriendly very unfriendly Very helpful helpful neither unhelpful very unhelpful Very cold cold neither warm very warm In relation to me, my,closest ggmpu§.friends 3322: Very comfortable* comfortable neither uncomfortable very uncomfortable Too religious religious neither nonreligious antireligious Very friendly friendly neither unfriendly very unfriendly Very helpful helpful neither unhelpful very unhelpful Very cold cold neither warm very warm *Comfortable here means at ease, feeling relaxed about another person. 61 52 63 64 65 66 67 68 69 7O 71 72 135 My feelings about waldorf students were, when I was a freshman: very positive positive mixed negative very negative About how strongly do you feel about your answer? not at all strongly not very strongly quite strongly very strongly My feelings about Waldorf faculty were, when I was a freshman: very positive positive mixed negative very negative About how strongly do you feel about your answer? not at all strongly not very strongly quite strongly very strongly My feelings about the waldorf administration were, when I was a fresh- 3:2; positive positive mixed negative very negative About how strongly do you feel about your answer? not at all strongly not very strongly quite strongly very strongly My feelings about the Waldorf Egagd.gf_§§ggnts were, as a freshman: very positive positive mixed negative very negative About how strongly do you feel about your answer? not at all strongly not very strongly' quite strongly very strongly My feelings about the Waldorf glgmg;,were, when I was a freshman: very positive positive mixed negative very negative About how strongly do you feel about your answer? not at all strongly not very strongly quite strongly very strongly my feelings about American.Lgthegen thrgh, waldorf's sponsor, were: very positive positive mixed negative very negative About how strongly do you feel about your answer? not at all strongly not very strongly quite strongly very strongly 136 73 My withdrawal from waldorf College was during my: first week first month first semester second semester 1964 summer first SOphomore week first sephomore month 7h The decision to withdraw from waldorf was primarily: my own father's mother's both parents peers 75 My activity right now is: (please complete one of the following) Study: (name and kind of school) work: (kind) Approx. monthly income $ Study primarily with some work: (please Specify) werk primarily with some study: (please specify) Anned service Marriage 76 Please state here your reason(s) for withdrawing from Waldorf: APPENDIX E MEAN AND STANDARD DEVIATION OF QUESTIONNAIRE ITEMS 137 138 MEAN AND STANDARD DEVIATION OF QUESTIONNAIRE ITEMS FOR DROPOUTS AND NONDROPOUTS Dropouts NondrOpouts DrOpouts NondrOpouts Vari- Mean Std. Mean Std. Mean Std. Mean Std. able dev. dev. dev. dev. 1 2.66 .86 2.39 .64 2.14 .80 2.48 .75 2 3.28 .59 2.92 .55 3.21 .69 3.12 .51 3 2.52 .99 2.60 .73 2.25 .75 2.56 .79 4 3.38 .62 3.08 .44 3.21 .57 3.21 .51 5 2.76 .87 2.58 .71 2.32 .82 2.32 .71 6 3.17 .60 2.97 .50 3.19 .68 3.05 .59 7 2.62 .94 2.83 .74 2.77 .82 2.68 .73 8 2.96 .87 2.53 .86 2.65 .89 2.75 .64 9 2.54 .84 2.45 .81 2.54 .81 2.27 .69 10 2.93 .77 2.83 .72 2.96 .72 2.82 .83 11 2.14 .74 1.89 .76 1.86 .71 1.85 .90 12 3.21 .56 3.35 .63 3.43 .63 3.27 .69 13 4.62 .94 .04 .42 4.13 1.39 .07 .59 14 1.72 1.62 .01 .10 1.54 1.50 .01 .12 15 1.34 .55 .02 .21 2.04 1.84 .03 .23 16 3.96 1.69 .00 .00 3.54 1.86 .10 .82 17 3.00 1.54 3.11 1.47 2.68 1.56 2.47 1.32 18 1.03 .19 1.04 .20 1.04 .19 1.14 .39 19 2.83 1.20 2.80 1.76 2.70 1.81 2.33 1.48 20 3.21 .90 3.01 .67 3.04 .58 3.17 .62 21 2.93 .65 3.08 .65 301“ 059 2.98 .5“ 22 2.61 1.75 1.93 1.65 1.50 1.26 1.85 1.48 23 1.93 1.13 1.75 1.01 1.46 .64 1.38 .78 24 .48 .51 .41 .49 .43 .50 .50 .50 25 .03 .19 .08 .27 .11 .31 .12 .33 26 .03 .19 .09 29 .00 .00 .08 .27 27 .69 .81 .26 .44 .21 42 .15 36 28 24 .44 28 .45 .25 .44 .41 .50 29 .03 .19 08 .27 .ll 31 .00 OO 30 .10 .31 .23 .42 .11 .31 .27 .45 31 .00 .OO .03 .16 .00 OO .08 27 32 10 .31 .18 .51 .29 46 .12 .33 33 03 .19 .03 .16 .04 .19 .05 .21 34 .03 .19 .04 .20 .04 .19 .11 .43 35 1.10 .41 1.16 .44 1.18 .39 1.18 .52 36 2.07 .92 2.07 .92 2.07 .81 1.95 .79 37 3.22 1.55 3.11 1.55 2.78 1.65 3.68 1.05 38 1.21 .49 1.15 .40 1.18 .39 1.06 .24 39 1.55 .78 1.58 .80 1.32 .67 1.21 .60 40 2.44 .87 2.73 1.08 2.69 1.05 2.55 .98 139 MEAN AND STANDARD DEVIATION OF QUESTIONNAIRE ITEMS FOR DROPOUTS AND NONDROPOUTS (Continued) Male Female Dropouts NondrOpouts DrOpouts NondrOpouts Vari- Mean Std. Mean Std. Mean Std. Mean Std. able dev. dev. dev. dev. 41 4.00 .66 3.89 .80 4.19 .83 4.17_ .89 42 1.80 .58 1.94 .63 1.68 .86 1.67 .90 43 3.65 .78 3.55 .75 3.62 .98 3.64 .85 44 4.00 .85 4.21 .74 4.15 .91 4.27 .90 45 1.76 .83 1.99 .74 1.81 .62 1.86 .79 46 1.85 .73 2.31 .78 2.04 .64 2.34 .83 47 2.87 .-2 2.65 .87 2.35 1.02 2.80 .94 '8 2.70 1.11 2.91 . 5 2.88 .99 2.94 1.01 49 2.04 .55 2.27 .85 2.07 .78 2.30 .90 50 3.61 .72 3.31 .8 3.35 .89 3.43 .93 51 2.67 1.03 2.70 .86 3.43 1.13 2.70 1.03 52 2.00 .98 1.82 .73 1.82 .82 1.79 .67 53 2.12 .95 1.84 .71 1.61 .69 1.88 .76 54 2.45 .51 2.63 .64 2.56 .70 2.34 .51 55 2.04 .84 1.88 .55 1.89 .57 1.82 .61 56 2.23 . 7 2.10 .56 2.00 .61 2.06 .58 57 3.55 .91 3.89 .57 3.81 .75 3.98 .65 58 1.50 .65 1.35 .59 1.21 .42 1.23 .46 59 2.45 .51 2.65 .63 2.48 .75 2.32 .56 60 1.71 .69 1.62 .54 1.43 .57 1.55 .83 61 1.91 .73 1.86 .68 1.71 .66 1.66 .57 62 3.91 .53 4.23 .59 4.46 .65 4.42 .63 63 1.52 .51 1.68 .52 1.75 .44 1.75 .47 64 1.96 .81 1.94 .55 1.75 .75 1.91 .55 65 2.41 .59 2.55 .63 2.54 .71 2.37 .63 66 1.89 .58 1.92 .49 1.96 .64 1.97 .50 67 2.23 .61 2.13 .63 2.04 .59 2.00 .56 68 3.73 .70 3.80 .73 4.00 .69 3.92 .76 69 1.36 .49 1.54 .50 1.25 .44 1.46 .53 70 2.50 .60 2.54 .56 2.59 .69 2.29 .46 71 1.52 .59 1.61 .55 1.43 .50 1.57 .66 72 1.73 .63 1.80 .52 1.68 .61 1.69 .58 73 4.09 .61 4.10 .51 4.42 .64 4.45 .61 74 1.00 1.55 2.33 4.04 .00 .00 .oo .00 75 2.00 3.16 4.00 3.16 4.75 2.96 4.50 3.00 76 .80 1.79 2.33 4.04 .00 .00 .00 .00 77 1.40 3.13 1.67 2.89 .00 .00 3.67 3.51 78 1.75 3.50 .00 .00 .00 .00 .00 .00 79 1.00 .00 2.00 .00 1.00 .00 2.00 .00 80 1.64 .48 1.90 .30 1.76 .44 1.93 .25 140 MEAN AND STANDARD DEVIATION OF QUESTIONNAIRE ITEMS FOR.ACUTE CONCERNS AND NONACUTE CONCERNS Male Female Acute Nonacute Acute Nonacute concerns concerns concerns concerns Vari- Mean Std. Mean Std. Mean Std. Mean Std. able dev. ' dev. dev. dev. 1 2.46 .66 2.47 .73 2.63 .52 2.36 .80 2 3.08 .76 3.01 .56 3.25 .71 3.15 .55 3 2.54 .97 2.58 .78 2.63 .52 2.46 .81 4 3.31 .63 3.15 .49 3.13 .64 3.22 .52 5 2.69 1.18 2.63 .68 1.63 .52 2.39 .73 6 3.31 .48 2.99 .54 3.25 .71 3.08 .61 7 2.83 1.19 2.76 .73 2.75 .89 2.70 .75 8 3.33 .89 2.55 .84 2.50 .93 2.77 .67 9 2.42 .90 2.48 .80 2.75 .46 2.30 .74 10 3.00 .85 2.84 .71 2.38 .74 2.93 .77 11 2.00 1.00 1.95 .73 1.75 .46 1.87 .87 12 3.46 .52 3.28 .62 3.38 .52 3.31 .69 13 2.11 2.27 .98 1.97 1.56 1.94 1.00 1.96 14 1.17 1.82 .29 .84 1.33 2.12 .29 .76 15 .72 .83 .27 .60 1.22 1.99 .47 1.19 16 2.17 2.41 .71 1.68 1.67 2.12 .85 1.86 17 3.38 1.50 3.03 1.48 3.50 1.41 2.46 1.36 18 1.00 .00 1.04 .21 1.25 .46 1.09 .33 19 3.62 1.19 2.69 1.64 2.75 1.58 2.42 1.60 20 3.23 .73 3.04 .75 3.00 .76 3.14 .60 21 3.08 .64 3.03 .66 3.00 .53 3.04 .57 22 2.69 1.97 2.03 1.65 1.25 .46 1.80 1.48 23 2.00 1.29 1.77 1.01 1.63 .92 1.38 .72 24 .38 .51 .43 .50 .50 .53 .48 .50 25 .08 .28 .07 .25 .38 .52 .09 .29 26 .08 .28 .08 .27 .25 .46 .04 .19 27 1.00 1.00 .29 .46 .38 .52 .15 .36 28 .15 .38 .29 .46 .50 .53 .35 .48 29 .00 .00 .08 .27 .00 .00 .04 .19 30 .00 .00 .22 .42 .00 .00 .25 .43 31 .00 .00 .02 .15 .00 .00 06 .24 32 .00 .00 .18 .49 .00 .00 .18 .38 33 .08 .28 .02 .15 .00 .00 .05 .21 34 .08 .28 .03 .18 .00 .00 .09 .40 35 1.15 .55 1.14 .41 1.50 .76 1.15 .45 36 2.38 .77 2.02 .93 2.38 1.06 1.96 .76 37 3.00 1.83 3.16 1.51 2.00 1.77 3.55 1.20 38 1.38 .65 1.13 .38 1.75 .46 1.04 .19 9 1.77 .73 1.54 .80 1.50 1.07 1.21 .56 40 2.17 .83 2.73 1.05 2.86 1.46 2.55 .95 141 MEAN AND STANDARD DEVIATION OF QUESTIONNAIRE ITEMS FOR ACUTE CONCERNS AND NONACUTE CONCERNS (Continued) =======================================I=E============================= Male Female Acute Nonacute Acute Nonacute concerns concerns concerns concerns Vari- Mean Std. Mean Std. Mean Std. Mean Std. able dev. dev. dev. dev. 41 3.60 .84 3.95 .75 3.86 1.35 4.19 .82 42 2.00 .63 1.89 .62 2.00 1.60 1.65 .80 43 3.60 .97 3.57 .73 3.14 1.21 3 67 .85 44 4.00 .67 4.18 .78 4.29 .76 4.24 .92 45 1.90 .88 1.94 .75 2.14 1.35 1.82 .68 46 1.75 .75 2.24 .78 2.29 1.50 2.25 .72 47 3.09 .83 2.65 .88 2.67 1.21 2.67 .97 48 1.89 .33 2.96 .90 2.33 .52 2.95 1.02 49 2.40 .84 2.19 .78 2.00 1.15 2.24 .85 50 3.40 .97 3.39 .85 2.83 1.17 3.45 .90 51 2.20 .84 2.81 .87 3.00 .00 2.88 1.13 52 1.67 .78 1.90 .81 2.25 .71 1.75 .71 53 1.85 .69 1.92 .80 2.13 .99 1.76 .72 54 2.60 .70 2.59 .61 2.71 .76 2.37 .55 55 1.80 .79 1.93 .62 2.00 .76 1.82 .58 56 2.20 .63 2.12 .65 2.13 .83 2.04 .57 57 3.80 .63 3.81 .69 3.71 .49 3.96 .69 58 1.46 .52 1.38 .62 1.25 .46 1.23 .45 59 2.50 .53 2.61 .62 2.57 .53 2.36 .63 60 1.60 .52 1.64 .59 1.25 .46 1.55 .78 61 1.80 .63 1.88 .70 1.63 .74 1.67 .59 62 4.00 .50 4.17 .60 4.71 .49 4.41 .64 63 1.77 .44 1.62 .53 1.75 .46 1.75 .46 64 1.92 .79 1.95 .60 2.00 .53 1.85 .63 65 2.40 .84 2.53 .59 2.71 .76 2.40 .64 66 2.00 .63 1.90 .50 2.13 .64 1.95 .54 67 2.20 .42 2.14 .65 2.13 .64 2.00 .56 68 3.60 .52 3.81 .74 3.86 .38 3.95 .76 69 1.46 .52 1.50 .50 1.50 .53 1.39 .52 70 2.50 .71 2.53 .55 2.57 .53 2.37 .55 71 1.60 .52 1.59 .56 1.38 .52 1.54 .63 72 1.90 .57 1.77 .55 1.88 .64 1.67 .59 73 3.78 .67 4.13 .51 4.29 .76 4.46 .61 74 1.50 1.73 1.40 3.13 .00 .00 .00 .00 75 3.00 3.56 3.00 3.25 4.75 3.20 4.63 2.88 76 1.33 2.31 1.40 3.13 .00 .00 .00 .00 77 2.33 4.04 1.00 2.24 3.50 4.95 1.33 2.31 78 3.50 4.95 .00 .00 .00 .00 .00 .00 79 1.36 .49 1.74 .44 1.38 .51 1.73 .45 80 1.00 .00 2.00 .00 1.00 .00 2.00 .00 APWNDIX F CODE BOOK 1. Questionnaire - ACT Study 2. MMPI - LYRI Study 142 CODE BOOK AN INVESTIGATION OF FACTORS LEADING TO THE WITHDRAMAL OF WALDORF JUNIOR COLLEGE F RESHMEN QUESTIONNAIRE - ACT STUDY C . Durward Roa College of Education Michigan State University January, 1965 INSTRUCTIONS FOR THE USE OF THIS CODE BOOK 1. Code 0 or 00 will always mean Not Applicable or Nothing. 2. Code 8 or 88 will always mean reSpondent Didn't Know. 3. Code 9 or 99 will always mean Information Not Available or Respondent Did Not Answer. 4. Coder instructions will always follow a line across the page. 5. In the following pages, the left column contains the variable number; the center column contains the card (deck) and column number; the right columns contain the questionnaire number, a brief description of the response category and the coded responses, pages 2-1 through 4-9, and the LYRI scale name and MMPI scale number and name, pages 1-10 and 1-11. 144 9.4.42.1 Vari- £10.. 2224 99.11.1112 20213 name LYRI scales 1 l 1 - 2 Family unity 2 1 3 - 4 Family Spiritual growth 3 1 5 - 6 Parental authority 4 1 7 - 8 Intrafamily relationships 5 l 9 - 10 Dating and morals 6 1 ll - 12 Life partner 7 1 13 - 14 Dating and the church 8 l 15 - 16 Spiritual doubts -9 1 17 - 18 Personal assurance 10 1 l9 - 20 Inadequacy feelings 11 l 21 - 22 Academic problems 12 1 23 - 24 Social relationships 13 l 25 - 26 Relation to teachers 14 l 27 - 28 Peer acceptance 15 1 29 - 30 Judgment of peers l6 1 31 - 32 Personal consistency 17 1 33 - 34 Adult consistency 18 l 35 - 36 Christian vocation l9 1 37 - 38 Moral problems 20 l 39 - 4O Boy-girl relationships 21 1 41 — 42 Spiritual guidance 22 1 43 — 44 Family counseling 23 l 45 - 46 Self-realization 24 l 47 - 48 Achievement Vari- able 45 46 47 48 49 Deck c- 2' a— c- zo 1:- c- e- e- c- a- a- s~ a— c- c- c— a- Column 49 51 \OVKAWH ll 11 13 15 17 19 21 23 25 27 33 39 41 43 45 47 - 5o - 52 145 CARD 1,2,4 Scale name LYRI scales Service Happiness Justification Humanitarian Conservatism Christian concepts Forgiveness Religious commitment 1 N O \O (I) \1 O\\J'\ (Tho Ms D 47 Pd Mf Pa Pt Sc Ma Si ES DY DO RE PR ST CN MMPI scales Hypochondriasis Depression Hysteria PsychOpatic deviation Masculine-feminine Paranoia Psychasthenia SchiZOphrenia Hypomania Social introversion Ego strength Dependency Dominance Social Responsibility Prejudice Social status Control Vari- able Deck Column 1 2 43 46 48 49 61 62 63 64 66 67 QUESTIONNAIRE STUDY re tem Detai Q'ai Feelings 1 - 2 - 3 - 4- 5 - about Valdorf students very positive positive mixed negative very negative Intensity of feelings about students Fulfil-4 Ill Feelings \JNH Ill 4 - 5.. not at all strongly not very strongly quite strongly very strongly about Waldorf faculty very positive positive mixed negative very negative Intensity of feelings about faculty Ple-J Feelings 1- 2.. 3 - 4.. 5.. not at all strongly not very strongly quite strongly very strongly about waldorf administration very positive positive mixed negative very negative Intensity of feelings about administration 1- 2 - 3 - u- Feelings 1- mtmm not at all strongly not very strongly quite strongly very strongly about board of regents very positive positive mixed negative very negative Vari- able Deck Column 8 10 ll 12 13 l4 2 N 50 51 53 54 55 56 Q'aire Item Detail 68 69 7O 71 72 13 14 147 ARD 2 Intensity of feelings about regents 1.. 2- 3.. L;- Feelings 1 - 2- \n-P‘ko Ill not at all strongly not very strongly quite strongly very strongly about Waldorf alumni very positive positive mixed negative very negative Intensity of feelings about alumni Dumb-4 Feelings 1.. 2 3 4 5 not at all strongly not very strongly quite strongly very strongly about American Lutheran Church very positive positive mixed negative very negative Intensity of feelings about the AIC 1- 2- 3- 4 - not at all strongly not very strongly quite strongly very strongly Time of withdrawal 1.. 2- wanna-u IIIII Decision 1.. 2 - 3- 4 - 5.. first week first month first semester second semester 1964 summer first SOphomore week first sephomore month to withdraw my own father's mother's both parents peers Vari- able Deck Column 15 16 17 18 19 20 21 2 57 58 15 l6 17 18 20 21 148 Q'aire Item Detail Activity 1... 2 - 3 - 4 - 5.. of withdrawal student currently study work study primarily with some work work primarily with some study armed service Reasons for withdrawal (Open ended) CARD 3 Residence before freshman year 1... Fwfil' \n-C’WNH‘Q‘ \n-C‘UN 2: 8 8 w £5 \lmmt-‘WNH Soci Famil #UNH‘d \A'F'UONH 5... rural 1 - 99 village 100 - 999 town 1,000 - 2,499 city 2,500 - 24,999 large city 25,000 up residence before freshman year not at all once twice three times four or more times of schools attended one two three four five six seven or more class of family lower lower'middle middle upper middle upper income much higher higher similar lower much lower Vari- able Deck Column 22 23 24 25 26 28 29 30 31 32 33 35 36 3 wwuuwuuuuuuw 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 149 CARE 3 Q'aire Item Detail 22 23 24 25 27 28 29 30 31 32 33 35 36 Matriculation decision l.- my own 2 - my father's 3 - my mother's 4 - both parents 5 - mutual Waldorf rated as first choice 1 - first 2 - second 3 - third 4 - fourth 5 - fifth or below Matriculation influence: parents Matriculation influence: counselor Matriculation influence: admissions Matriculation influence: pastor Matriculation influence: alumnus Matriculation influence: another college Matriculation influence: relatives Matriculation influence: WC teacher Matriculation influence: friend Matriculation influence: teacher Matriculation influence: WC administrator Health 1 - very good 2 - fair 3 - poor Reading ability 1 - very good 2 - good 3 - fair 4 - poor Vari- able Deck Column 37 3 24 38 3 25 39 3 26 4o 3 27 41 3 28 4° 3 29 43 3 30 Q'aire 37 38 39 4O 41 42 43 150 CARD 3 Item Detail Remedial study 1 - English 2 - Mathematics 3 - Reading 4 - None 5 - English-Mathematics 6 - English-Reading Class attendance 1 - very regular 90% 2 - regular 75% 3 - off and on 50% 4 - irregular 25% 5 - seldom 15% Chapel attendance 1 - very regular 90% 2 - regular 75% 3 - off and on 50% 4 irregular 25% 5 - seldom 15% Parents - very liberal - liberal - neither - conservative very conservative Paren 8 very distant distant neither close very close Paren very warm warm neither cold very cold Paren 3 very rigid rigid neither flexible very flexible mPUNl-‘d' \n-C‘UNl-‘S' \n-LT'UNI-‘d' \ACUNH Vari- able Deck Column Q'aire Item Detail 214 45 47 49 50 3 31 32 33 34 35 36 37 44 45 46 47 48 49 50 151 CARD 3 Parents 1 - very unhelpful 2 - unhelpful 3 - neither 4 - helpful 5 - very helpful Parents 1 - very religious 2 - religious ' 3 - neither 4 - nonreligious 5 - antireligious Congregation l - very helpful 2 - helpful 3 - neither 4 - unhelpful 5 - very unhelpful Congregation 1 - very conservative 2 - conservative 3 - neither 4 - liberal 5 - very liberal Congregation 1 - "high church" 2 - formal 3 - neither 4 - informal 5 - "10w church" Congregation 1 - very warm 2 - warm 3 - neither 4 - cold 5 - very cold Congregation . l - very distant 2 - distant 3 - neither 4 - close 5 - very close 152 QARD 3 Vari- able Deck Column g'aire Item Detail 51 3 38 51 Congregation 1 - very Haugean 2 — Haugean nonHaugean antiHaugean don't know quantity very many many a few very few none 52 3 39 52 Frien 3- 4— 5- ds 1 2 3 4 5 53 3 40 53 Toward classmates l - very comfortable 2 - comfortable 3 - neither 4 - uncomfortable 5 rd 1 2 3 u 5 - very uncomfortable classmates - too religious - religious - neither - nonreligious - antireligious 54 3 p 41 54 Towa classmates - very friendly - friendly neither - unfriendly - very unfriendly 55 3 42 55 Towa classmates - very helpful - helpful - neither - unhelpful - very unhelpful 56 3 43 56 Towa 57 3 44 57 Toward classmates 1 - very cold 2 - cold 3 - neither 4 - warm 5 - very warm Vari- able Deck Column Q'aire Item Detail 58 61 62 63 3 45 46 W W 50 51 58 61 62 63 Ox 1:- 1% CARD 3 Toward friends 1 - very comfortable 2 - comfortable 3 - neither 4 - uncomfortable 5 - very uncomfortable Toward friends - too religious - religious - neither nonreligious 5 - antireligious :me Toward friends 1 - very friendly 2 - friendly 3 - neither 4 - unfriendly 5 - very unfriendly Toward friends 1 - very helpful 2 - helpful 3 - neither 4 - unhelpful 5 - very unhelpful Toward friends 1 - very cold 2 - cold 3 - neither 4 - warm 5 - very warm Closest friend as room-mate 1 - yes 2 - no From classmates 1 - very comfortable 2 - comfortable 3 - neither 4 - uncomfortable .5 - very uncomfortable Vari- able Deck Column Q'aire Item Detail 65 3 52 65 From classmates l - too religious 2 - religious 3 - neither 4 - nonreligious 5 - antireligious 66 3 53 66 From classmates l - very friendly friendly neither unfriendly very unfriendly N I 1:- III 67 3 54 67 From classmates 1 - very helpful 2 - helpful 3 - neither 4 - unhelpful 5 - very unhelpful 68 3 55 68 From classmates 1 - very cold - cold - neither - wann - very warm \ntwm 69 3 .56 69 From friends 1 — very comfortable 2 - comfortable 3 - neither 4 - uncomfortable 5 - very uncomfortable 70 3 57 70 From friends 1 - too religious 2 - religious 3 - neither 4 — nonreligious 5 - antireligious 71 3 58 71 From friends 1 - very friendly 2 - friendly 3 - neither 4 - unfriendly 5 - very unfriendly 155 CARD 3 Vari- able Deck Column 'aire Item Detail 72 3 59 72 From friends 1 - very helpful 2 - helpful 3 - neither 4 - unhelpful 5 - very unhelpful '73 3 60 73 From friends 1 - very cold 2 - cold 3 - neither 4 - warm 5 - very warm CARD 4 T scales 74 4 51-52 ACT English 75 4 53-54 ACT Mathematics 76 4 55-56 ACT Social Studies 77 4 57-58 ACT Natural Sciences 78 4 59-60 ACT Composite Subgroups 79 4 65 Attrition l - drOpout 2 - non-drapout 80 4 66 Faculty rating 1 - acute concerns 2 - non-acute concerns 4 67 Sex 1 - male 2 - female 4 72—74 Subject Number 4 75 Deck Number APPENDIX G COMPARISON OF MMPI MEAN AND STANDARD DEVIATION ETWEEN LYRI RELEASE AND NONREIEASE RESPONSES 156 157 COMPARISON or MMPI MEAN AND STANDARD DEVIATION BETWEEN LYRI RELEASE AND NOchLSASE RESPONSES 1 Male Female LYRI-10371 MMPI only LYRI-WWI MMPI only responses responses reSponses responses Vari- Mean Std . Mean Std . Mean Std . Mean Std . able dev. dev. dev. dev. 33 51.83 8.81 61.42 24.36 48.76 6.88 62.95 22.24 34 51.67 11.34 65.28 23.58 51.49 7.43 63.35 21.96 35 54.08 7.07 63.35 23.32 51.74 6.89 62.55 22.74 36 58.24 9.36 67.51 22.22 54.40 8.25 65.35 21.58 37 57.25 9.17 65.37 23.39 49.47 8.97 63.40 22.80 38 55.46 7.75 67.42 22.37 54.77 7.88 67.50 20.83 39 60.54 10.99 70.42 21.47 57.76 7.23 71.10 17.12 40 60.38 11.74 70.14 21.52 56.38 7.25 71.80 19.25 41 58.60 11.40 66.44 22.84 55.95 9.56 69.45 19.68 42 54.03 9.75 66.56 23.28 56.64 8.92 68.25 19.64 43 53.10 8.32 60.65 25.17 54.33 7.14 62.70 22.26 44 54.20 8.67 66.23 22.90 53.16 8.14 66.55 19.58 45 49.37 8.22 58.98 25.73 48.28 8.13 60.10 24.11 46 47.89 8.73 58.21 26.62 48.42 9.04 56.90 26.43 47 48.50 8.85 61.84 24.89 47.05 8.00 65.05 22.11 48 50.61 8.31 58.93 26.11 49.49 8.26 59.05 24.78 49 52.15 11.00 62.86 24.51 46.94 10.53 60.70 23.98 N = 83 N = 43 N = 78 N = 20 1111 "t" tests were significant at the .01 level. APPENDIX H CROSSAVALIDATION OF VARIABLES SELECTED ON THE BASIS OF REFEATED OCCURRENCE IN TWENTY-FOUR EXPERIMENTS (A CLINICAL BUNCH) 158 159 AN ADDITIONAL EXPERIMENT ON CROSS-VALIDATION BASED ON CLINICAL HUNCHES Various approaches were made to attempt discrimination of the best- predictor variables. Twenty-four "eXperiments" were conducted based on alternative rationales. It appeared that certain variables recurred even though the resons for their use in developing an optimum equation varied. The results of using the recurring variables are reported below as Sample III. (Compare p. 92.) Variables selected Male dropouts: 7.13.14.24.26.37.43.47 Male concerns: 2.11.19.23.25 Female dr0pouts: 1...3,12.22,30...32.36.37.39.43.46 Female concerns: 2,3,11,12,15.18.22.28,30.33.39.44.45.46 Correct identifications Male drOpouts: 89% Male concerns: 64% Female drOpouts: 64% Female concerns: 50% Total drOpouts: 78% Total concerns: 60% Sample III Male (N = 41) Female (N = 39) Mean Std. dev. Mean Std. dev. Drapouts Y 1.6098 .4939 1.7180 .4559 Y 1.6010 . 2708 1.7688 .3460 R 023 039 Acute concerns I 1.7317 .4486 1.8974 .3074 Y 1.7897 .1950 1.9484 .2251 R 03h 025 "'ClTr'lTIflmLflflMfitljfltilllfliififlflflfllfijfimflfi