THE RELATIONSHIP OF ENVIRONMENTAL CHANGE TO STUDENT PERFORMANCE AND ATTRlTl-ON Thesis for the Degree of Ph. D.‘ M-ICI‘HGAN STATE UNIVERSITY Donald H. Voss 1966 7??F$?S ’2 4"‘a—|—kfl"'§flt’ CHEM“? efi , “W .LILLRJ453Y' like?) 131:: $1.33 University h. IIIHIINIHIIH W!!!I!!!NI!Ill/L’IIUIUI/UU!III/HM , 3 12 3 10437 4305 This is to certify that the thesis entitled “The Relationship of Environmental Change to Student Performance end Attrition" presented by Donald H. Vose has been accepted towards fulfillment of the requirements for Eh. D. . degree mm)! Major professor mew. 0-169 ABSTRACT THE RELATIONSHIP OF ENVIRONMENTAL CHANGE TO STUDENT PERFORMANCE AND ATTRITION by Donald H. Voss The Problem Little is known about the impact of the community college upon its students. Even less is known about the effects of the change in environment involved in the tran- sition from high school to a community college. The purpose of this investigation was to study the relationship of environmental change (in the transition from high school to a community college) upon the performance and rates of attrition of community—college students. The following research hypotheses were investigated in the study: 1. High schools will differ from the community college in the amount of demand made upon the students in both intellectual and non- intellectual areas. 2. The demands made upon the students in the high schools will differ in the amount of emphasis placed upon such factors as aca- demic achievement, group life, vocational Donald H. Voss emphases compared to the emphases upon such factors at the community college. Negative change in goodness of fit between environmental demands and student pref- erences in the transition from high school to community college will be related to increased student attrition. Change in goodness of fit between the indi- vidual's preferences and the relative emphases in his new environment (at the community college) will be related to performance: positive change to better performance, negative change to worse per- formance. Goodness of fit at the community college between an individual's preferences and the college's emphases will be related to attrition. That is, good fit will be related to retention; poor fit to attrition. Goodness of fit at the community college between individual's preferences and the college's emphases will be related to per— formance. That is, good fit wlll be related to good performance; poor fit to poor performance. An individual's previous experience with "poor fit" will modify the effects of "poor fit" at the community college: Donald H. Voss a. Subjects who experience poor fit at the community college but who pre- viously experienced good fit at the high school level will leave the college and not continue their education else— where more frequently than poor-fit students who previously experienced poor fit in high school. b. Students who experience poor fit at the community college but who experienced poor fit in high school will perform at a higher level than poor-fit students who experienced good fit in high school. Methods and Procedures One hundred and seventy-two Lansing (Michigan) Com- munity College students were randomly selected from required orientation classes. The subjects were all first-time, full- time (12 quarter hours or more) students in the fall of 1965, Samples were also selected from the senior classes of eight high schools in the service area (an area with a radius of approximately 30 miles) of the community college. All the high-school subjects were volunteers within two months of high school graduation. Each high school sample consisted of 36 subjects. Donald H. Voss The college subjects were given the Activities Index and the Evening College Characteristics Index develOped by George G. Stern.l The community-college subjects were also given a brief (ten-question) questionnaire which paralleled and para- phrased the eleven major emphases of the environmental indexes. The high-school subjects were given the High School Characteristics Index also developed by Stern.2 The three Index test forms consist of 300 parallel items. The Activities Index items refer to common activities for which the individual indicates his "like" or "dislike." The environmental-index items refer to pres- sures, demands, rewards and activities common in an educational setting. The parallel nature of the environ- mentalanuipersonal questionnaires permits assessment of the "fit" between the individual's preferences and his environ- ment's demands. The major analyses of performance used cumulative (fall + winter) grade point averages. The attrition-rate analyses used retention and drOpout data for the entire academic year. 1George G. Stern, Preliminary Manual: Activities Index-College Characteristics Index (Syracuse, New York: Psychological Research Center, 1958); George G. Stern, Scorinngnstructions and College Norms: Activities Index- College Characteristics Index (Syracuse, New York: Psychological Research Center, 1963). 2Ibid. Donald H. Voss Comparisons were made between the "fit" of students' preferred activities (as determined from the Activities Index) and the demands made upon them in high school and at the community college. (The community-college subjects were all graduates of one of the eight high schools studied.) Subjects were assigned to four levels of environmental con- tinuity. Assignment to these categories was made on the basis of major or minor positive or negative change in "fit" between the individual's needs and his environment's demands in the transition from high school to the community college. (Cattell'sl rp pattern analysis statistic was used to express these relationships.) Several methods were used in the computation of "fit" or congruence at the community college. All employed Cattell's rp pattern analysis statistic. Pattern congruence was defined as congruence between an individual's need (preferences) pattern and the pattern of demands in his environment in which all standard score differences were used in the computation of the rp statistic. Deviation pattern congruence used only those standard score differ- ences which exceeded one standard deviation (above or below) the mean of student perceptions of press at the institution. Expressed congruence was computed using the differences between the individual's reported perceptions of his 1J. L. Horn, "Significance Tests for Use with r and Related Profile Statistics," Educational and Psy— _ghological Measurement, XXI, No. 2 (1951), 363—379. Donald H. Voss environment's demands and his stated preferences. Private Beta congruence (in contrast to Adjusted and Pattern con— gruence which used the means of students' perceptions of press or demands) was computed by using the individual's own profile of needs and his own perceptions of environmen— tal demands. Adjusted pattern congruence was computed using only deviations below the means of the samples' perceptions of demands in the intellectual area and above the means in the non-intellectual area. The "objective" measures (Pattern, Deviation pattern, and Adjusted pattern congruence) were more highly related to performance and attrition than the "subjective" measures (EXpressed and Private Beta congruence). Deviation pattern congruence was more highly related to retention and Adjusted pattern congruence was more highly related to per- formance measures. Major Findings The following were the major findings of the study: 1.. The community college's (perceived) demands differed significantly from those at the eight high schools in the intellectual area. In the non-intellectual area the greatest perceived change was a decrease in the emphasis upon play (organized social life) in the community— college setting. 2. The patterns of perceived high school demands did not differ from the patterns of demands at Donald H. Voss the community college (using Cattell's rp pattern analysis statistic) although differences in relative emphases were found when rank—order correlation was used. Negative change in goodness of fit was related to the transfer intentions of high achievement students and to the attrition rates of low achievement students (dropout). Change in goodness of fit between the individual's preferences and the demands of his environment was significantly related to the performance of low achievement students. Positive change in goodness of fit was related to improved performance by high achievement students. Goodness of fit at the community college between an individual's preferences and the college's emphases was related to the transfer intentions of high achievement students and the attrition (drop- out) of low achievement students. Goodness of fit at the community college between an individual's preferences and the college's emphases was related to performance by both high and low achievement subjects. a. Subjects who experienced poor fit at the com- munity college following poor fit in high school drOpped out of the college more fre- quently than those who had previously Donald H. Voss experienced good fit. This relation— ship was the Opposite of the one pre— dicted. b. Subjects who experienced poor fit at the community college following poor fit in high school did not perform at a signifi- cantly higher level than those who had previously experienced good fit. (Improve- ment of fit was related to higher perform- ance. This relationship had not been predicted.) The pattern of demands at the community college fits the "average" student's pattern of pref- erences very well. Any drastic changes initiated in order to "fit" the institution's demands to the individuals who do not "fit" the current environment would result in a less satisfactory "fit" for the majority of the student body. THE RELATIONSHIP OF ENVIRONMENTAL CHANGE TO STUDENT PERFORMANCE AND ATTRITION By 4" Donald Hf‘Voss A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education Guidance and Personnel Services 1966 ACKNOWLEDGEMENTS The investigator would like to acknowledge his gratitude to Dr. William W. Farquhar for his assistance in the preparation of this thesis and for his guidance throughout the doctoral program. Appreciation is also extended to Drs. Bill Kell, Max R. Raines and Richard C. Rank for their assistance and helpful criticism. The investigator is also indebted to the officials, faculty members and students of Lansing Community College and of the eight high schools involved in the study. With- out their permission and actual involvement in the study the research would not have been possible. The investigator is particularly grateful to his wife, Betty, and to his parents for their support and their assistance in a significant portion of the clerical work involved. To paraphrase a statement made by the author of a related research project, ”they seldom knew what it all meant, but they know what it all means." ii TABLE OF CONTENTS Page ACKNOWLEDC MENTS . . . . . . . . . ii LIST OF TABLES . . . . . . . . . V LIST OF APPENDICES . . . . . . . . X Chapter I. THE PROBLEM 1 Purpose A Research Hypotheses A Theory. . 7 Definition of Terms 13 Overview 16 II. REVIEW OF THE LITERATURE . . . . . 17 Environmental Assessment: General. . 18 Environmental Assessment: Pace and Stern Instrumentation . . . . 20 Related Studies . . . . . . 22 Discussion. . . . . . . . 26 Summary . . . . . . . . 29 III. DESIGN . . . . . . . . . 31 The Sample(s) . . . . . . . 31 Instrumentation: The Activities Index. 33 Instrumentation: The Environment Indexes 36 Procedures. . . . . . . 43 Score Transformations . . . . . 46 Definitions . . . . . A7 Statistical Hypotheses. . . . . 52 Ar alysis . . . . . . . 7 Summary . . . . . . . . 63 IV. ANALYSIS OF RESULTS . . . . . . 6? Section I: Press . . . . . . 68 Section II: Continuity . . . . 79 Section III: Congruence . . . . 86 iii Chapter Section IV: Adaptation Level Section V: EXploratory Hypotheses. V. DISCUSSION OF RESULTS. Press . Continuity. Congruence. . Adaptation Level . . . . Exploratory Hypotheses: Continuity Summary . . . . . VI. SUMMARY AND CONCLUSIONS The Problem . The Samples and Methodology The Findings . The Conclusions Discussion. Implications for Future Research BIBLIOGRAPHY APPENDICES iv Page 95 97 110 110 113 116 126 127 128 134 13A 136 139 1244 1A7 153 155 160 LIST OF TABLES Table Page 3.1 Test- retest Reliability Coefficients for the Factor Scales of The Evening College Characteristics Index. . . . 38 3.2 Reliability Coefficients for the Factor Scales of the High School Characteristics Index. . . . . . . . . . HO A.l Total Intellectual Press Comparisons (Each of the Eight High Schools Compared to the Community College) . . . . . 7O A.2 First-order Factor Comparisons for the Six Scales of the Intellectual Press Dimension: Eight High Schools - Community College . 72 A.3 Total Non-Intellectual Press Comparisons (Each of the Eight High Schools Compared to the Community College). . . . . 73 4.A First—order Factor Comparisons for the Six Scales of the Non—Intellectual Press Dimen- sion: (Each of the Eight High Schools Compared to the Community College) . . 75 A.5 Comparisons of Patterns of Press at the Eight High Schools to the Pattern of Press at the Community College-—Using Cattell's r Profile or Pattern Analysis Statistic. p. 77 4.6 Comparisons of the Rank-order of Press Factors at the Eight High Schools and the Community College . . . . . . . . . 79 A.7 Analysis of Variance - Continuity (A levels) and Cumulative (Fall + Winter) Grade Point Average . . . . . . . . . 83 A.8 Duncan's New Multiple Range Test Applied to the Differences Between Four Means (Continu- ity — Fall + Winter G.P.A.) . . . . 8a Table Page 4.9 Analysis of Variance - Continuity (4 Levels) and Achievement (2 Levels) X Performance (Indicated by Cumulative (Fall + Winter) Grade Point Average) . . . . . . 85 4.10 Duncan's New Multiple Range Test (Adapted for Use with Unequal Replications) Applied to Low Achievement Students' Performance in Terms of Cumulative (Fall + Winter) Grade Point Average. Subjects Stratified by 4 Levels of Continuity . . . . . . 86 4.11 Analysis of Variance of Congruence (4 Levels) and Cumulative Grade Point Average for Fall + Winter Quarters . . . . . 91 4.12 Duncan's New Multiple Range Test Applied to the Differences Between 4 Treatment Means—- Congruence and Fall + Winter G.P.A. . . 92 4.13 Analysis of Variance - Congruence (2 Levels) and Educability (2 Levels) and Fall + Winter Cumulative Grade Point Average . . . 93 4.14 Duncan's New Multiple Range Test Applied to 4 Treatment Means: Congruence (2 Levels) and Educability (2 Levels) and Cumulative (Fall + Winter) Grade Point Average . . 94 4.15 Analysis of Variance: Congruence (4 Levels) and Continuity (3 Levels) and Cumulative (Fall + Winter) Grade Point Averages . . 95 4.16 Group Means — Congruence (4 Levels) X Conti- nuity (3 Levels) and Cumulative (Fall + Winter) Grade Point Averages. . . . 95 4.17 Correlations of Cumulative High School Grade Point Averages with Cumulative (First Year‘s) Grade Point Averages at the Community College: (Subjects Separated According to Level of Similarity of Their High School's Pattern of Demands to That at the Community College). 99 4.18 Summary of Results of Major Hypotheses Tested . . . . . . . . . 101 6.1 Summary of the Results of the Investigation of Press in the Eight High Schools and Lansing Community College. . . . . l4O vi Table 6.2 III.2 III.3 III.4 III.5 IV.1 IV.2 IV.3 Summary of the Results of the Investigation Page of Environmental Continuity in the Transition from High School to Lansing Community College Summary of the Results of the Investigation of Environmental Congruence Summary of the Results of the Investigation of Adaptation Level Summary of the Results of the Exploratory Investigation of Environmental Continuity and High School to College Grade Point Correlations . . . . . Sex Composition and Identification of the Samples from the Community College and The Eight High Schools Reliability Coefficients for the Scales of the Activities Index Reliability Coefficients for the College Characteristics Index Scales . Reliability Coefficients for the Factor Scales of the Evening College Character- istics Index . . . . Sample Sizes for the Instrumentation Used in the Study Significant Values of r for Several (K) Degrees of Freedom and the Probability of Exceeding These Through Chance Alone (at .01, .15, and .10) . . . Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshmen) High School #1 (Local Public). Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshmen) High School #2 (Local Public). vii 141 143 208 209 210 211 212 215 219 220 Table Page IV. IV. IV. IV. IV. 4 Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshmen) — High School #3 (Local Public). . . . 221 Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshmen) — High School #4 (Parochial) . . . . 222 U‘l .6 Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index(College Freshmen) - High School #5 (Non—Local Public). . . 223 .7 Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshmen) - High School #6 (Non—Local Public). . . 224 8 Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshmen) - High School #7'(Non-Local Public). . . 225 9 Comparison of Mean Scores on the High School Characteristics Index (High School Seniors) and Mean Scores on the Evening College Characteristics Index (College Freshman) - High School #8 (Non-Local Public). . . 226 .10 Total Intellectual Press Comparisons (Between Each of the Eight High Schools and the Community College) . . . . 227 .11 Total Non—Intellectual Press Comparisons (Between Each of the Eight High Schools and the Community College) . . . . 228 12 Means of Student Performance at Two Levels of Achievement and Four Levels of Continuity for Fall, Winter and Cumulative (Year's) Grade Point Averages . . . . . . 229 viii Table IV.13 IV.14 IV.15 IV.16 IV.17 IV.18 Analysis of Variance: Achievement (Two Levels), Educability (Two Levels), Con- gruence (Two Levels), and Cumulative (Fall +Winter) Grade Point Averages Means of Student Performance at Two Levels of Achievement and Two Levels of Congruence and Two Levels of Educability - Cumulative (Fall + Winter) Grade Point Averages Analysis of Variance: Con-gruence (Two Levels), Achievement (Two levels), Adap— tation Level (Two Levels) and Cumulative (Fall + Winter) Grade Point Averages Means of Student Performance at Two Levels of Achievement; Two Levels of Environ- mental Congruence and Two Levels of Adaptation to Incongruence--Cumu1ative (Fall + Winter) Grade Point Averages Range of Common Beta Press (Means on each of the dimensions 1 One Standard Deviation) Compared to the "Average" Student Profile, "Typical Dropout," and "Typical Intended Transfer". . . . . Profile of Means of 50 Randomly-selected Evening College Characteristics Index test forms (converted to Activities Index equivalents) Compared to the Means of 25 Randomly—selected Male and Female Activities Index test forms ix Page 229 230 230 231 232 233 LIST OF APPENDICES Appendix Page I. DEFINITIONS PERTAINING TO INSTRUMENTATION . . . . . . 161 Need- Press Scale Definitions . . 162 Activities Index Factor Definitions . 165 Environment Index Factor Definitions . 169 II. INSTRUMENTS USED . . . . . . . 173 Personal Data Questionnaires . . . 174 Expressed Preference Questionnaire. . 176 College Characteristics Index . . . 179 Activities Index . . 186 Evening College Characteristics Index . 193 High School Characteristics Index . . 200 111. DESCRIPTION OF SAMPLES AND RELIABILITY COEFFICIENTS . . . . . . . 207 IV. PROCEDURES SUPPLEMENT AND SUPPLEMENTAL ANALYSES TABLES . . . . . . 213 Cattell' s r Correlational Profile Statisticp . . 214 Assignment of Subjects to Continuity Categories . . 215 Assignment of Subjects to Adaptation Level Categories. . . . 217 Development and Use of Expressed Preference Questionnaire. . . 217 Summaries of First— and Second— Order Comparisons (High School — Community College) for each of the Eight Schools Separately . . . . 219 Supplemental Analyses Tables . . . 229 CHAPTER I THE PROBLEM A prominent recent trend in American higher education has been an increase both in numbers and in percentages of students who begin their college education in a community or junior college. The trend is considered by many educa— tional experts to be one which will become even more pronounced. In view of this trend, it seems strange that behavioral scientists concerned with education have done little research in community or junior colleges. If it is true that there is less fundamental research on the opera— tion of colleges and universities than almost any other social or economic institution,l it is markedly true of the community college. The observation was made in a recent report that the book, The American College, which is self— described as a "psychological and social interpretation of the higher learning,' contains only ten references to com- munity or junior colleges in its more than 1,000 pages.2 1Commission on Goals for Higher Education in the South, Within Our Reach (Atlanta, George: Southern Region Education Board, 1961), p. 43. 2James M. Richards, Jr., Leonard P. Rand, and Lorraine M. Rand, "A New Way to Measure Environment," Junior College Journal, 36 (1966), p. 18. __‘-h__._.fi_._.k- Carpenterl noted (in his 1966 presidential address to the American Higher Education Association) the following needs among the many he discovered in a survey of research requirements in the South: ‘1) studies of the "presses" and pressures of college environments upon students, 2) research on the interaction of high schools and colleges, and 3) the interactions of junior or community colleges with other institutions of higher education. Although Carpenter did not specify research into the interaction between high schools and community colleges, such a need seems apparent in view of the increasing importance of the community college in higher education. Community or junior colleges offer programs which include adult education courses, short-term Specialized vocational programs, "terminal" curricula leading to sub- baccalaureate degrees, and the traditional "transfer" cur- ricula designed to prepare the student to continue his education at a four-year college. Representative studies indicate that at least two-thirds of the entering students indicate intention to transfer to four-year institutions. Approximately one-third of those who indicate intention to transfer actually do transfer. Many drop out for strictly academic reasons. Others leave for reasons that are better described as psychological, social, parental or financial. L 1C. Ray Carpenter, "Reflections on Research on Higher Education: Strategies and Tactics," AHE: College and Shiversity Bulletin, 18, No. 11 (1966), p. 5. 3 The attrition rate involves more than academic failure. It may also involve failure to meet the psychological, social or economic demands of the college situation. There is a need for further research on these non—academic factors which may contribute to the current rate of academic attrition.l In Spite of the fact that many of the students who indicate intention to transfer to a four—year institution do not do so, little is known about the comparative charac- teristics of those who transfer and those who do not.2 Still less is known about the effect upon students of the change in environment in the transition from high school to college. Previous research has been centered on students who made the transition from a community college to a four— year institution.3 The community colleges are apparently effective with their transfer and terminal students. The "latent terminal” with transfer intentions does not fall into either category. These students terminate their formal education while in lJohn Summerskill, "DrOpouts from College," The Ameri— can College, ed. N. Sanford (New York: _John Wiley & Sons, 1962). pp. 637-649. 2Dorothy M. Knoell and Leland L. Medsker, Factors Affecting Performance of Transfer Students from Two- to Four- Year Colleges (Berkeley, California: Center for the Study of Higher Education, 1964), p. 32. 3George G. Stern, "Continuity and Contrast in the Transition from High School to College," Orientation to College Learning: A Reappraisal, ed. Nicholas C. Brown (Washington, D. 0.: American Council on Education, 1961), pp. 37—68. college, but they do so while pursuing transfer work. The modal student in the community college is neither the ter— minal nor the transfer student, but the "latent terminal.” It is hoped that this study will serve as a preliminary investigation of the effects of environmental variables upon these ”latent terminal" students in particular. Purpose The purpose of this investigation is to study the relationship of environmental change (in the transition from high school to a community college) to the performance and rates of attrition of community-college students. The objective is to study the effects of the community- college environment upon differing kinds of students with different high school environmental backgrounds to deter- mine whether variable environmental factors are associated with levels of student performance and rates of attrition. The effects of continuity and contrast in environments as the student moves from high school to the community college will be studied. Research Hypotheses I. High schools will differ from the community college in the amount of demand made of students in both intellectual and non-intellectual areas. 11. The demands made upon the students in the high schools will differ in the amount of emphasis placed upon such factors as academic achievement, group life, and IV. vocational emphases compared to the emphases upon such factors at the community college. Negative change in goodness of fit between environmen- tal demands in such areas as academic achievement, group life, vocational emphases and the individual's preferences for such emphases will be related to increased student attrition. That is, if the student's needs (preferences) fit the environmental demands less well at the community college than they did in high school, such change will be associated with increased rates of student attrition. Change in goodness of fit between the individual’s preferences and the relative emphases in his school environment upon such factors as academic achievement, group life, play—work and vocational emphases will be related to levels of performance. That is, a student whose pattern of preferences resembles more closely the demand-pattern at the community college than the pattern in his high school (a positive en- vironmental change) will perform at a higher level than the student whose preferences for environmental demands fit the pattern at the community college less well than at his high school (a negative change). Goodness of fit between the individual's preferences in such areas as academic achievement, group life, play-work and vocational emphases and the VI. VII. environmental demands or relative emphases in such areas will be related to attrition. That is, "good fit" will be related to retention; poor fit to attri- tion. Goodness of fit between the individual's preferences in such areas as academic achievement, group life, play-work and vocational emphases and environmental demands or relative emphases in such areas will be related to performance. That is, "good fit" will be related to better performance; "poor fit" to poorer performance. Previous experience with "poor fit" or lack of correSpondence between the individual's preferences for emphasis upon such factors as academic achieve- ment, group life, play—work, and vocational emphases and the actual levels of demand in such areas made upon him by his environment will modify the effects of ”poor fit" at the community college. 1. Students who experience "poor fit" between their preferences and environmental emphases at the community college but who experienced "good fit" at the high school level will leave the college at a higher rate than students who experience "poor fit" at the community college but who experienced "poor fit" in high school. 2. Students who experience "poor fit" at the com- munity college but who eXperienced "poor fit" in high school will perform at a higher level than students who experience "poor fit" at the community college but who experience "poor fit" at the community college but who experienced ”good fit” in high school. Theory Social scientists are giving increasing attention to some of the more subtle, but highly significant factors in student learning. The physical environment in which the student lives, the kind of college community in which he finds himself and the social structures and processes of which he is a part are examples of the new emphasis. Studies of these factors offer increasing evidence of the importance of the college‘s "climate of learning” as a potent force in determining the outcome of educational pro- grams. The following set of generalizations can be made from college environment studies:1 1. An educational institution has its own distinc- tive atmosphere or climate which appears to remain fairly stable from year to year. 2. Peer group and faculty—student interaction outside of the classroom are important variables affecting student values. These may be stronger 1James G. Rice, "The Campus Climate: A Reminder," Higher Education: Some Newer Developments, ed. Samuel Baskin (New York: McGraw-Hill, 1965), p. 307. determinants of student attitudes and values than what occurs in the classroom. 3. Many ”outside” activities such as extra- curricular programs, academic advising pro- grams, and campus events have effects upon both the mastery of subject matter and the perceived relevance of learning. If the above generalizations are warranted concerning the college student in a residential setting, they are per- haps even more applicable to the community— or junior— college student. Although the community-college student‘s environment typically does not change as dramatically between high school and college because he ordinarily re— mains at home, he is not "typical college material." Collinsl states that the community-college student comes from a lower-middle class home and bears the mark or in- fluence of this background. He has a low level of educa— tional background and academic achievement. Many studies have indicated that he has a more modest academic potential than his four—year—college counterpart. Workers in the community college suSpect, however, that his intellectual energy level, his valuing of education and his time— perSpective may be even more important factors. The distant "pay off” of education may not be sufficient to motivate him 1Charles C. Collins, "Critical Problems of Students," Junior College Journal, 36, No. 7 (1966), pp. 32-36. in his day-to-day academic efforts. It may be that the impact of the community-college student's environment, while not as encompassing as that of his peers at a residential school, may be fully as great upon his performance and his decision to continue his education. Because he takes his education in a "cafeteria" setting, it is even easier for him to express his dissatisfaction than for his residential— college counterpart. He can more readily withdraw psycho— logically, if not physically, from an environment which he finds distasteful. He is more free, in Reisman's termi- nology, to "drop in and drop out." ArgyriS'l theory concerning the problem of integrating the individualand his organization is applicable to the problem of this study. His fundamental postulate is that there is a lack of congruency between the needs of healthy individuals and the demands of formal organizations. He states that, at best, society may hOpe to "satisfiee" the relationships between the individual and the organization. That is, society may simply be able to meet a set of cri- teria which describe minimally satisfactory alternatives rather than trying to work toward some ideally, maximally satisfying arrangement. Although Argyris believes that organizations and indi- viduals cannot be completely congruent, it is possible that incongruence between an individual and his organization may lChris Argyris, Integrating the Individual and the Organization (New York: John Wiley & Sons, 1964). IO provide the basis for a challenge which may enhance growth. Lack of congruence, however, may also cause individual frus— tration at too high a level because the individual's self- expression is blocked. He will then tend to experience "failure" and short time-perspective because he will not be able to define his own goals. Argyris believes that it is impossible to understand the individual without understanding the organization in which he finds himself. It is extremely difficult, however, to describe either human personality or organizational en- vironments with precision. In order to begin work on the primary objective of understanding the individual and the organization, extended discussions will have to be sacri— ficed and selections made of those factors which seem most relevant to the understanding of both the individual and his environment. Although Argyris‘ theory has led to an increasing amount of research and thereby proven its viability, it does not as yet involve the develOpment of instrumentation which permits the direct comparison of individuals and their organizational environments. A significant portion of current environmental research which does involve such com- parison is based upon the theory of Henry A. Murray.1 Murray believes that the individual's environment largely determines his behavior. Because environment changes, often 1Henry A. Murray, Explorations in Personality (New York: Science Editions, 1962). 11 abruptly, one cannot formulate or understand the indivi- dual's behavior without first describing his environmental situation. Murray uses the term ppggp to designate a directional tendency in an object or situation. He distinguishes be- tween alpha press (the environment's demands as they actually exist as far as scientific inquiry can determine them), and beta press--the subject's own interpretation of the environment's demands as he perceives them. .EEEEE: then, are dynamically pertinent attributes of a particular environment in terms of benefits offered to particular needs or the frustrations imposed upon other needs. The process in which the subject at any moment says, "this is good,” or ”that is bad,” Murray calls pressive perception. Pres— sive percgption is an egocentric process which almost always gives rise to adaptive behavior on the individual's part. Actually, the individual does not find most stimulus situ- ations directly effective. Most situations are effective behavior-stimulators because they let the individual know what is coming. The individual becomes more involved with _pressive apperception--tne realization that the object or stimulus may do something if he remains passive or some- thing else if he becomes active. The ppggg of an object or situation, then, is what it 2gp do to or for the subject. The subject may not even be aware of the press being exerted upon him. Murray be- lieves that the process of pressive apperception is largely 12 unconscious—-the organism merely reacts to a stimulus. If an individual apperceives a constellation of stimuli as harmful, he may find his current environment unsatisfying or threatening. He may not be aware of the Specific reasons for his feeling. Corresponding to press in the environment, Murray posits the concept of 222d within the individual. A ppgd is a hypothetical process which is asserted in order to account for certain objective and subjective facts. These ppgdg may be psychogenic or viscerogenic. Their importance varies for the individual according to the ease or diffi- culty he discovers in seeking to satisfy them. Each re- action to an environmental pgpgp has a fortune which may be measured in degrees of gratification or realization. If gratification of a need is defined as success, frustra- tion of a need as a failure; it may be added that success and failure are of major importance in establishing the status of an organism in its community. Murray's conceptualization of need-press interaction has made possible the development of instrumentation which permits comparison of the congruence or ”fit" between an individual's needs and his environment's press. These instruments, developed by George G. Stern, C. Robert Pace and others, will be used in this study of the effects of environmental change upon student behavior. The section which follows includes definitions of terms descriptive of major dimensions of the Stern instruments and terms which 13 pertain to this study alone. (Definitions of the thirty sub-scales of the Stern instruments and the first-order factors derived from them are given in Appendix I.) Definition of Terms The following definitions pertain to the major dimen- sions of the instruments used in this study:1 1. Intellectual Climate.--This dimension represents the more conventional aspects of academic program including staff and facilities. It also includes standards of achieve- ment set by students and faculty, and Opportunities for the development of self-assurance. 2. Non-intellectual Climate.-—This dimension repre- sents the level of organization of academic and student social affairs. It also includes the emphasis upon tech- nical and vocational courses. 3. Emotional Expression.—-This dimension of indi— vidual needs represents high levels of social participation and emotional Spontaneity. It also involves the indi- vidual's level of self—assertiveness. 4. Educability.—-This dimension of individual needs includes elements of both intellectuality and submissiveness Scores on this dimension indicate extent of interest in intellectual activities coupled with orderliness and conform- ity. 1George G. Stern, Scoring Instructions and College Norms: Activities Index - College Characteristics Index (Syracuse, New York: Psychological Research Center, 1963). 14 5. Common Beta Press.--The mean or average of a group's perception of press in a given setting. The Common Beta Press may be expressed in terms of sub-scale—, first- or second—order factor dimensions. (In this study, first- order factors are used.) The following definitions pertain to the terminology indirectly related to the instruments used in the study or to usage peculiar to this study alone. 1. Community College.--A two-year public institution of higher education located in Lansing, Michigan. It offers curricula in the arts and sciences (the traditional "trans— fer” curricula), and certain business, technical, and health sciences curricula. 2. Service Area.--The area served by the college which includes a radius of approximately 30 miles. 3. Area High School.--Any high school located in the above area. Those included in the study were of the following types: a. Local public. A high school located in the same city as the community college and which is publicly governed and tax— supported. b. Non-local public. A high school located outside of the city in which the community college is located and which is publicly governed and tax-supported. C . 15 Parochial. A high school which is privately financed and governed under religious auSpices. 4. Full-time student.--A community-college student enrolled for 12 term credits or more during a given term. 5. Grade-point average.--Numerical average of grades received in academic subjects. a. In high school. English, foreign languages, mathematics, science, and social science. In the community college. Those subjects which ordinarily received transfer credit: those which are acceptable for college credit at a four-year institution and which are accepted by it as credit earned toward gradu- ation. 6. Achievement.-—Academic achievement as indicated by the above numerical designation (grade-point average): a C High achievement. Those students whose high school grade-point average was above the median for high school students beginning work at the community college as full-time freshmen students. Low achievement. Those students whose high scnool grade point average was below the median for high school students beginning work at the community college as full-time freshmen students. 16 7. DrOpout.—-A student who discontinued his education at the community college and who did not continue his edu— cation at another educational institution. 8. Intended Transfer.-—A student who indicated in reply to questionnaire surveys that he or she intended to transfer to another educational institution prior to com— pleting requirements for an Associate-in-Arts degree at the community college. Overview In Chapter II, the review of the literature includes a brief survey of environmental studies and a more detailed description of several studies most similar to and there- fore most relevant to this one. The methodology and procedures, including the statis— tical methods used and the assumptions underlying them, are described in Chapter III. In Chapter IV the results of the study are analyzed. The results of the study are discussed in Chapter V. Chapter II, which follows, begins with a brief state— ment about assessment in educational settings as an intro- duction to the topic of environmental studies in particular. CHAPTER II REVIEW OF THE LITERATURE In thousands of high schools and colleges every year hundreds of thousands of students take batteries of tests designed to measure their personal and academic qualities, and to assess their curricular and vocational choices. In a similar manner studies1 are made of academic organizations. These include investigations of social structure, human and financial input and output, community and regional contexts and many other variables. All of this research is concerned with what might be termed "goodness of fit" or congruence between the categories being assessed. Because the literature about the above concerns is far too broad in sc0pe to be covered adequately, the review which follows will be limited to: 1) a brief survey of research on the impact of different environments upon students, 2) a survey of representative studies using Pace and Stern instruments, and 3) a more detailed review of closely related studies. lC. Robert Pace, "Methods of Describing College Cultures,” Teachers College Record, 63, pp. 267—277. 1? 18 Environmental Assessment: General Farris (in a mimeographed study cited by Argyrisl) investigated Argyris' theory that the "fit" of an individ- ual and the institution in which he finds himself will effect the individual's motivation, affective job experi- ences and his performance. Farris developed quantitative indices for motivation and provision for self-actualization, status and Social congruency. These indices were given to a group of research scientists. Measures of ”objective congruence" (the individual's motive on a given dimension compared to the average per- ception of provision for that motive by members of his department) did not correlate significantly with intensity of motivation. Measures of "perceived congruency" (the individual's motive compared to his perception of the insti- tution's provision for that motive) did correlate signifi- cantly with motivation. Neither measure correlated signi— ficantly with performance. Thistlethwaite2 investigated the effects of college press on the study plans of talented students. The study included 987 men and 513 women at 327 different American colleges and universities. Thistlethwaite found that col— lege environments characterized by faculty affiliation, lArgyris, op. cit., pp. 42-47. 2Donald L. Thistlethwaite, "College Press and Changes in Study Plans of Talented Students," Journal of Educa- tional Psychology, 51, No. 4 (1960), pp. 222-234. l9 enthusiasm, emphasis upon achievement or independence are associated with increased motivation to seek advanced degrees in the arts, humanities and social sciences. Col- lege environments characterized by a lack of faculty em— phasis upon student compliance are associated with increased motivation to seek advanced degrees in the natural or bio— logical sciences. Astinl described 248 colleges and universities and their entering classes in the fall of 1961 using six charac— teristics: intellectualism, estheticism, status, leader- ship, masculinity, and pragmatism. Astin concluded that the characteristics of the entering freshmen classes were highly related to characteristics of the colleges. He felt that the aSpirations of the entering students were well— suited to the curricular offerings of the institutions. Knoell and Medsker2 investigated the performance of junior-college transfer students. They found in their nationwide survey both institutional and individual factors related to differences in the performance of the students who transferred from two—year to four-year institutions. Probability of ”on time” graduation, for example, was lAlexander W. Astin, "Distribution of Students Among Higher Educational Institutions," Journal of Educational Psychology, 55 (1964), pp. 276-87; Alexander W. Astin, ”Some Characteristics of Student Bodies Entering Higher Educational Institutions," Journal of Educational Psy— chology, 55 (1964), pp. 267-275. 2Dorothy M. Knoell and Leland L. Medsker, Articulation Between Two-Year and Four-Year Colleges (Berkeley, Cali— fornia: Center for the Study of Higher Education, 1964). related to choice of major field, choice of four—year college and to sex differences (which were related in part to choice of major). Knoell and Medsker conclude that most junior-college transfer students could be successful in achieving their degree goals if they would choose institu- tions and major fields apprOpriate to their ability levels and previous achievement. Richards, Rand and Randl have developed a relatively simple method of describing junior or community colleges in terms of an institutional profile with six dimensions: cultural affluence, technological Specialization, size, age, transfer emphasis and business orientation. Their goal was the development of a profile which could be used in research studies of the effects of junior colleges on students. They made no suggestions, however, about criteria for the measure- ment of these effects or the relationship between the six variables they described and student characteristics. They conclude that previous work has not sought to develop detailed descriptions of junior—college environments and that no attempt to study the impact of junior colleges on students has been made. Environmental Assessment: Pace and Stern Instrumentation Until the time of Pace and Stern's development of The College Characteristics Index no objective measure of lRichards, loc. cit., pp. 18—20. . . l . enVlronmental press was available. The Index has Since been used extensively by Pace and Stern and others. Stern2 used the College Characteristics Index in a study which involved 1076 students at 23 different colleges. Six factors associated with different types of schools were identified in this investigation. The elite liberal arts colleges were found to score high on the "Intellectual Orientation” and "Social Effectiveness" factors. Several large state and private universities scored high on the "Play” factor. ”Friendliness” (informal social organiza- tion) was characteristic of a mixed group of schools. Denominational colleges were characterized by high scores on the ”Constraint” factor. State teachers colleges were characterized by the ”Dominance—Submission" (custodial— care) factor. The correlation between the "Intellectual Climate” factor at 37 colleges and the per cent of graduates re- ceiving the Ph.D. degree between 1937 and 1956 was .76. The correlation between the ”Intellectual Climate" factor and the National Merit Scholarship Qualifying Test means at 38 schools was .71. 1 IA copy of the index is included in Appendix 11. 2George G. Stern, ”Characteristics of the Intellec— tual Climate in College Environments,” Harvard Educational Review, 33 (1963), pp. 5—41. Pacel described the colleges he studied using three basic patterns of institutional press: high intellectual, high practical and high social. Stern2 analyzed the data from 32 schools in which both the College Characteristics Index and its parallel instrument (The Activities Index) had been administered and found that the differences among the institutional environments were markedly greater than the differences among the student bodies involved. He found that students tended to be enrolled in institutions in which the environmental press was compatible with their inventoried personality needs. Related Studies 3 Schwartz sought to determine the effects of congru- ence and incongruence between individual needs and environ— mental press upon academic performance. His sample included 108 SOphomores at the New York University School of Commerce, Accounts and Finance. Schwartz investigated the relationships between 1 intellectual de endenc and emotional ex ression needs 3 p Y: P and press and (2) congruence between needs and press on lC. Robert Pace, ”Five College Environments,” College Entrance Examination Board Review, 41 (1960), pp. 24—28. 2George G. Stern, ”Congruence and Dissonance in the Ecology of College Students,” Student Medicine, 8 (1960), pp. 304—309. 3Ronald M Schwartz, "Congruence Between Needs of Individuals and Environmental Press as Related to Perfor— mance and Adjustment in a Large Organization" (unpublished Ph.D. Thesis, New York University, New York, 1964). 23 these dimensions. The relationship of these variables to performance and academic adjustment was studied. Intellec- tual and dependency press were related to total academic adjustment but not to performance. Emotional expression press was unrelated to performance or academic adjustment. Neither intellectual needs nor dependency needs were related to academic adjustment or performance. Emotional expres— sion needs were negatively related to performance and total academic adjustment. Congruence of needs and press was not related to performance, but those whose needs were con- gruent with the prevailing press were less well adjusted (adjustment measured by The Inventory of Academic Adjust— _mep£) than those for whom there was less congruence. This finding was the Opposite of what had been predicted. In discussing his results, Schwartz stated that a possible explanation for the negative correlation between congruence and academic adjustment can be found in his per— ception of the environment of the School of Commerce as highly authoritarian, rigid, cold and unfriendly. The lack of positive relationship between congruence and performance might be explained in part, Schwartz felt, by the obser- vation that at the beginning of the SOphomore year the population of students becomes somewhat restricted in range due to dismissal of students for academic failure during the freshman year. 24 In a similar study, Keithl administered the same instruments (The College Characteristics Index and The Activities Index) to a sample of undergraduate students who had completed at least four semesters of work in resi- dence at the University of Alabama. Keith used an average correlation coefficient of need and press factor scores in each of the divisions of the university as a measure of the congruency between expressed needs and perceived presses in each division. No significant relationships were found between the congruency index scores and academic success or reported personal satisfaction with the institution. Keith observed that the congruency index scores were usually low and that variance and range were restricted. The low level of con— gruency and the restriction of range may have caused the lack of significant correlation, Keith felt. The most relevant local study was one done by Campbell2 at Lansing Community College and Michigan State University in 1963-64. Cambell compared need and press scores in samples of 90 ”area” students using the Activities Index 1James A. Keith, "The Relationship of the Congruency of Environmental Press and Student Need Systems to Reported Personal Satisfaction and Academic Success"(unpublished Ph.D. Thesis, University of Alabama, University,Alabama,1964) 2Paul S. Campbell, “Personality Needs of Community College and University Students and Their Perceptions of Their Institutions: An Experimental Investigation"(unpub- lished Ph.D. Thesis, Michigan State University, East Lansing, Michigan, 1964). and the College Characteristics Index. He found significant differences between the two samples both in mean scores of need factors and of perceptions of press at the two insti- tutions. Herrl used the high school fOrm of the indexes (Th3 High School Characteristics Index) to study the relationship between differing levels of academic achievement and extra- curricular participation as these related to different per- ceptions of press in a high school environment. Significant relationships were found between these variables and per- ceptions of press. For example, high achievement students perceived greater press for achievement than did low achievement students. The High School Characteristics Index and the College Characteristics Index have been administered to the same sample of college students. Stern2 administered the two instruments to 2000 freshmen students during orientation week at a major eastern university. Students were instructed to typify their high school experience and their expectations for the university. Stern pulled four sub— groups from this sample: private preparatory (N = 103), parochial (N = 89), localppublic (N ==96) and non—local lEdwin M. Herr, ”An Examination of Differential Per— ceptions of 'Environmental Press' by High School Students as Related to Their Achievement and Participation in Activ— ities” (unpublished Ed.D. Thesis, Columbia University, New York, 1963). 2Stern, ”Continuity and Contrast. . . , op. cit., pp- 33-58. 26 public (N = 29). He found significant differences between the four types of high school environments. Stern felt that the differences would probably become even more sharply delineated if the range of school samples were broadened to include different types of high schools and students who did not intend to go on to college. He also discovered that freshman intellectual press expectancy was unrealistic (compared to college seniors' perceptions of press on this dimension). Stern made the suggestion that student apathy might be the consequence of unfulfilled expectations in the transition from high school to college rather than the cause . Discussion None of the above studies have followed Fishman'sl suggestion to develop research which involves different types of high schools, different types of high school en- vironments and types of college environments. Fishman also suggested that research efforts be concentrated upon seeking to determine whether environmental differences have different effects upon different students. This has not been done except in a general way in junior—college transfer- student studies. lJoshua A. Fishman, "Some Social-psychological Theory for Selecting and Guiding College Students,” The American College, ed. N. Sanford (New York: John Wiley & Sons, 1962), pp. 666-689. 27 The studies cited in the review of the literature which investigated the effects of congruence between needs and press upon performance and adjustment have not con- sidered a factor which may be important. It is possible that students' previous experience with incongruence (lack of fit between their individual needs and the environment's demands) might modify the effects of subsequent similar experience. In addition, the subjects studied (college sophomores and juniors) may have reached a level of adap- tation to incongruency within the college environment itself which might modify the impact of such incongruence. The research designs, in other words, did not permit the study of the possible effects of past experience upon current experiences with incongruency. If, however, psychological stimuli are judged partially on the basis of residuals of past experience with similar stimuli,l it seems reasonable to hypothesize that the effect of disparity might vary according to the individual's level of adaptation (to a given set of stimuli). Although it is difficult to give such previous eXperience precise experimental treatment, it is possible to consider it. For example, descriptions of high school environments could be obtained from students involved in a college study. lH. Helson, "Adaptation Level Theory,’ Sensory, Per- ceptual and Physiological Formulations, Vol. I of Psychol- pgy: A Study of a Science, ed. S. Koch (New York: McGraw- Hill, 1959). It is also likely that the restriction of range discovered in several of the studies was caused in part by student attrition-—both through students leaving the en- vironment to continue their education elsewhere or to leave higher education altogether. It would seem advisable with such a possibility in mind to design a study which would include students early in their college careers. Such research (involving freshmen) would permit examination of a broader range of students and increase the probability of finding significant relationships between environmental demands and student performance and/or attrition. In addition to the above, none of the studies reviewed gave differential treatments to levels or directions of diSparities between individual's needs and environmental press or demands. All units of diSparity were treated as equal. A need score which fell below an environmental press score was treated in the same way as one which was above the press score. It is difficult to weight such differences in amount or direction of diSparity with precision, but certain disparities may be excluded in the treatment of the data. Other differences which exceed a certain range can be emphasized. Such treatment of need-press disparity fol— lows more closely Murray'sl theory that needs differ in importance to the individual according to the ease or dif- ficulty he encounters in fulfilling a given need. lMurray, op. cit., pp. 54—115. Murray clearly insisted that hierarchies of need and relevance of press be considered in press-need studies. None of the above investigations took into account the ef- fects of different levels of disparity. One unit of dis- parity was treated as equal to any other unit. Direction- ality of differences was not considered. Individual need greater than press was treated in the same way as need less than press. Summary The studies included inthe review of the literature are summarized briefly in the chart on page 30. The design of the investigation which is described in the following chapter incorporated ways of treating the data which included broad assessment of the individual's level of adaptation to lack of fit between his needs and his environment's demands. The design also involved con- sideration of directionality and extent of disparity in the assessment of congruency or fit between an individual's needs and his environment's demands. 1:,DH..H.MHH-H3 LCM pm. .o.Ww._-Hm:H.OmJ flmjpomflfiwpflfl H60 mcoHHOHOOQHO OHHOHHOOHCO 1:c6 OH HO HO:-oer Oon 6H63 :6EQaO - .Oc:O.H 6H63 mHCOECOHH>CO HHmmHV -m/HOn 66H; H6H6n26EOH -.:6E£mOHH 666HHOO CH OOOCOHOHHHO OOCHHnHQ 6m6HHOo CH6Hm .coHummHOHHHmd HmHOOHHnsouOHme 6cm pcoE6>6H£O6 AnmaHv OHE6H666 :HOH OH 66H6H6H 6H6; mmmHo Ho mCOHHd6OH6Q HmHHCOHOHHHQ Hoonom cmHm HHOm HHH6H6>HCD .nmCHHO6n OSH OCH CH 6nd no new 6m6H AnomHv anH 6cH6c 6:6 nOHOOc n66: C663H6p 6:306 6H6; 6626H6H66 HHHO HcmO HHHcmHm uHoo HHHCOEEOQ HH6poEmo .OOcmeHOHHOQ OH HHOCOOHHHC Aqerv anm 66H6H6H 63 OH @6366 Hon n63 666Hd new 6666: COOsHOO OOCOOHmcoo OmOHHoo :pHOx .H662Hnsmnm OHE66.Om OH 66H6H6.H HzH6>HH66.6c 61 OH Ocsom 663 666nm new 6666: :663H6n 66:63H om .6.OCOEH OHHOQ OHEOOmOm 6:6 2666c coHnn6HQ AnmmHv uxm HmnnHHus6 cirsmmo OQBOH 665 COHH6H66HH oO 6>HHOM6C HCOOH.HHcmHn < 6w6HHoo mpsmscOm .nH 6H HHOQH COHs 6HOHHOQEOO has anHm 6gp LOHcs :H mCOHH no nsHHooHH O6 ssoH OH OH 666:6O mHC6OOH6 HnOHOoo HC6O$H6 6W6HHOO mcoEO AmeHv . 9) cmcH 6-;6LH3H.:H1 OmOHHoO mcosm OOOSOHOHHHO OOHHOE 6HoE 6H6; OHOQB OmOHHoo CHOHm S 666& .mHC6OOHn com: Awmev OOOHEH in SLH.OHH: O :O _o Hndpm OH Ho nHC6E20HH>C6 Om6HHOO Hochm Unmm a mo OOOHH cHHonOO eHHma- .Q.OH6>6O Op O6HoEOHH6 Own HHOS OOOH>6HQ oz OmOHHoo Ocmm «mOHchHm .a:OHO:OHHc:H H6.Hn;mHH 6:6 6H6mme OHOHHQOHQQO 660:6 HOLH HH qumH V 6H6 m 66Hn6O.H H6LH :666H HHOOO 66:663Hn HOanmnp OmOHHoOIHoHCOH H602 6m6HHoo Hoxm 662 6 HHmocx .eosHo>cH ecoHHsoHHecH nee oH HeomHv B6HHsnnme6s OOH6nHoeoo 6H63.n:oHH6HHQn6 new mOHpmHHOHOOHOSO He6ndpw OmOHHoo CHpm< .m6OO6Hom HOOHmOHOHo 6:6 Hmsdpmc 65H OH 66H69EOO c6O26HO H:HOO 6:6 npsm 63H :H 666Hw66 66OC6>66 x666 Op COHH6>HHoE Homev qum 6H ccH :.H3 OmpmHOonnm 6o OH oczoe 6H6; meomHH6> HHHOOOH HC6H6HHHQ 6m6HHoo OHHOSSHOHumHzB .66COEHOHH6Q OH Ho: HBO OOHH6>HHoE OH 66H6H6H 6H63 6666: 666sp mcHHOOE How COHpmchmmHo Azme anHHwHnHo HOCOszpHHth Hcm 6666c HHOLH mo mcoHHQ6OH6Q .mHmecmHOm :OHOOmOm mHHHmm >6 mCOHHSHOcoo Ho mmCHOCHm H6 H Amvnocpd< HecoHeeoHHnCH CHAPTER III DESIGN The nature of the samples included in the study, the nature of the instruments used and the statistical methods employed in analyzing the data are discussed in this chapter. The Sample(s) The Lansing Community College students selected were first-time, full—time students. They were 1965 graduates of service—area high schools (within a radius of 30 miles of the college) who had not attended college elsewhere. All of the students were enrolled for a minimum of 12 term- credits during the fall term in which they were tested. The students were enrolled in orientation classes required of all full-time freshmen. Assignment to sections in these orientation classes was made entirely on the basis cfl?conwmnencein class scheduling. Because of this "ran— domization,” no further attempt was made to randomize the sample. An inSpection of the curricular and vocational choices of the students was made. The assumption of ran- domization through scheduling appeared tenable. Three hundred and forty students were tested in the initial test period. This group included approximately 31 one-half of the full-time freshmen enrolled in the college in the fall of l965. Those students for whom comparative (high school) background information was available were pulled from the original group. These 172 students (90 male, 82 female) constituted the final community—college sample. The high school students who took part in the study were all seniors in high schools in the service area of Lansing Community College. All of the students volunteered to take part in the study. School staff members (teachers, guidance counselors and principals) who assisted in securing subjects indicated that the samples were broadly repre- sentative. The volunteers were atypical in that they were somewhat above average in ability. In addition, the sub— jects' post high school plans included some form of further education more frequently than was typical at the high school involved. The sample sizes varied from school to school within the range of 36 to #2 subjects. Samples which contained more than 36 subjects were treated as follows: each sub- ject's test was numbered and a table of random numbers was entered to select the necessary (one to six) subjects to be eliminated to equalize the samples. The final sample at each of the eight high school included 36 subjects. The schools in the study were all located within a radius of 30 miles of Lansing. There were three local public high schools, one parochial school, and four non- local public schools. The eight high-school samples 33 included the following ranges of male and female subjects: males: lA-Ql, females: l5—22. (A table listing the exact composition of the community college and high school samples is included in Appendix III as Table III.l. A table is also included in Appendix III, as Table III.5, which gives the final sample sizes for the instrumentation discussed in the following section.) Instrumentation: The Activities Index Stern, Stein and Bloom1 have offered a conceptual scheme for describing the phenomena which are relevant to interaction between individuals and their environmental situations. This conceptualization is based upon Murray's need-press schema. This same source:L describes the development of the Activities Index, a multi—dimensional inventory for meas- uring personality needs. In its present form, the instru— ment consists of 30 ten-item scales correSponding to 18 unidimensional and 12 bipolar needs adapted from Murray. Each item describes a common daily activity or feeling for which the individual indicates his like or dislike. The individual's needs are inferred from his reSponses to these items. This method of measurement and its underlying logic is similar to that involved in many psychological 1George G. Stern, Morris K. Stein, and Benjamin J. Bloom, Methods in Personality Assessment (Glencoe, Illinois: Free Press, l956). (A) .1: instruments.l (A copy of the Activities Index is included in Appendix II.) Reliability Stern's2 initial statements about the reliability of the Activities Index related to profile patterns obtained by means of vector analysis. He stated that inspection of the patterns of test—retest profiles suggested that a cor— relation based on a multi-variate surface corresponding to the profile determinants should be quite high. The most extensive investigation of the reliability of the Activities Index is reported by Stern.3 This in- vestigation was based on the data from 1078 Activities Index test profiles from students at 32 schools. The average sub— scale reliability was .69. (A complete listing of the 3C sub-scales' reliability coefficients as reported in Stern's investigation is included as Table 111.2 in Appendix III.) Validity 3 Stern also cited several research studies in his article which pertain to the validity of the Activities Index. The following generalizations can be made from 1George G. Stern, "Environments for Learning," The American College, ed. N. Sanford (New York: John Wiley & Sons, 1963), p. 703. 2George G. Stern, Preliminary Manual: Activities Index - College Characteristics Indexi(Syracuse, New York: Syracuse University Psychological Research Center, l958), p. 30. N 3Stern, ”Environments. . . , 0p. cit., pp. 690-730. 35 these studies: I) descriptions of behavior which may or might be expected from individual students, psychiatric patients, and industrial personnel derived from need pro- files appear recognizable (by peers, psychiatrists and administrators), 2) similar behavior patterns are related to similar profiles; 3) students or professionals in the same field have been found to have need profiles that differ significantly from those of students or professionals in other fields. Stern and Scanlonl found significant relationships between need scale profiles and career choice. Haring, Stern and Cruckshank2 used the Activities Index to dis— criminate between four groups of teachers in a workshop designed to change teachers' attitudes toward exceptional children. The assessment made by the leeex was confirmed by independent analysis of workshOp transcriptions. 3 Stern also cited several unpublished studies which he claimed supported the validity of the Activities Index. One such study used the Strong Vocational Interest Blank as the independent criterion. Another used the Rorschach, 1George G. Stern and J. S. Scanlon, "Pediatric Lions and Gynecological Lambs," Journal of Medical Education, 33 (1958), pp. 12-18. 2N. G. Haring, G. G. Stern and W. M. Cruckshank, Attitudes of Educators Toward Exceptional Children (Syracuse, New York: Syracuse University Press, 1958). 3Stern, Preliminary Manual. . . , 0p. cit. 36 the Thematic Apperception Test, and Sentence Completion reSponses as the independent criteria. Stern:L implied that the studies which related to the validity of the Activities Index would be discussed fully in a later manual. A revised manual has not been published. When the revised manual is published, it will no doubt include more complete substantiation of the validity of the Activities Index. Instrumentation: The Environment Indexes The environment indexes (The College Characteristics Index, The Evening College Characteristics Index, The High School Characteristics Index, and The Organizational Cli- mate Index) were developed as direct parallels to the Activities Index. The 30 ten-item scales of the two types of instruments are given the same definitions. (Definitions of the sub-scales and factor—scales of the indexes are given in Appendix I.) The environment indexes consist of 300 statements about the demands, pressures and characteristics of organi- zations and their personnel. The three which were deveIOped specifically for use in educational institutions include items which describe characteristics of the students at the institution as well. The individual using the indexes to describe his educational environment is instructed to mark the items "True” or ”False" (characteristic or not lIbid. 37 characteristic of his institution). (Samples of the instru- ments used in the study are included in Appendix II.) Reliability The most extensive treatment of the reliability of the College Characteristics Index is that previously cited in connection with the Activities Index.1 The average reliability reported for the 30 sub~scales was .65. (The reliability for the 30 sub-scales is given in Table III.3 of Appendix III.) Because The Evening College Characteristics Index was used instead of The College Characteristics Index, further studies of the reliability of the latter will not be cited. Fifty students enrolled in introductory psychology courses at the community college were given The Evening College Characteristics Index. The subjects were freshmen or sophomores who had not been part of the original sample. The 50 subjects were re-tested after a period of 30 days. The reliability coefficients were computed using the ll environmental factor scales. The average scale-reliability was .83. The complete results of this reliability check are given in Table 3.l. Because The Evening College Characteristics Index is a new form of the environmental assessment instruments, a second check was made of its reliability. Reliability coefficients were computed on a small sample (N = 50) of ll lStern, ”Environments. . . , Op. cit., pp. 690-730. 38 test protocols selected at random from the original sam- ple's test returns (N = 250). The Split-half (odd—even) technique was used. The formula was supplemented by the Spearman-Brown PrOphecy Formula to give an estimation of the reliability of the test corrected to full length. The average scale reliability computed in this way for the ll first-order factors was .85. (The complete listing of these reliability coefficients is given in Appendix III, Table III.U.) TABLE 3.l-—Test-retest reliability co- efficients for the factor scales of The Evening College Characteristics Index. Subjects: 5O Introductory Psychology Students Interval: 30 Days Scale Reliability Aspiration Level .80 Intellectual Climate .9u Student Dignity .58 Academic Climate .86 Academic Achievement .88 Self-Expression .98 Group Life .83 Academic Organization .90 Social Form .83 Play-Work .80 Vocational Climate .7O MEAN .83 39 Herr:L reported an average sub-scale reliability (for the 30 sub—scales of the High School Characteristics Index) of .5. He used a split-half, odd-even technique. He noted an increasing reliability on many of the scales as the students progressed from the ninth through the twelfth grades. Because the ll environmental factor scales were used in this study, separate reliability estimates were made. It was not possible to obtain enough time in any of the high schools in which the students were tested for a test— retest procedure. All of the High School Characteristics _Ipdex test forms from the eight schools were numbered con- secutively. A table of random numbers was entered to select 50 of these tests for examination. Split—half, odd-even reliability coefficients were computed on the ll factor scales. The correlational formula was supplemented by the Spearman-Brown Prophecy Formula to give an estimation of the test's reliability when corrected to full length. The average scale-reliability computed in this way was .74. The complete results of this reliability check are given in Table 3.2. Validity The difficulty of establishing criteria for testing the validity of personality tests is well—known. The cri— teria themselves are plagued with reliability and validity l . Herr, Op. Cit. MO problems. A frequently-used approach is that of correlating a new instrument with existing, established instruments. In environmental assessment this technique cannot be used. The environmental assessment instruments which have been devel- Oped do not measure directly comparable dimensions. TABLE 3.2.-—Reliability Coefficients for the factor scales of the High School Characteristics Index. (Split-half, odd— even——supplemented Scale by the Spearman- Reliability Brown Prophecy For— mula) Aspiration Level .58 Intellectual Climate .76 Student Dignity .70 Academic Climate .64 Academic Achievement .75 Self-Expression .8A Group Life .9N Academic Organization .6H Social Form .83 Play-Work .81 Vocational Climate .65 MEAN .7A Pace and Stern:L suggested comparing the judgments of different individuals within the same environment to deter- mine whether or not they agree in their perceptions of environmental pressures, rewards and demands. The lC. Robert Pace and George G. Stern, "An Approach to the Measurement of Psychological Characteristics of College Environments,” Journal of Educational Psychology, H9 (October, 1958), pp. 269-277. Ml unpublished studies cited by Sternl related to the validity of the College Characteristics Index used this validity assessment. The studies cited may be summarized as follows: l. Academic participants and observers appear able to recognize and confirm descriptions of college environments based solely on press profiles. 2. The press profiles obtained from student reSponses are consistent with those obtained from faculty and administrators at the same institution. 3. Students describe their own institutions in terms of press in ways that are significantly more alike than the correSponding press descriptions at other institutions. A. The average level of specific needs among students tends to match the average level of corresponding press at the same insti- tutions. In one of the initial studies relating to the validity of the College Characteristics Index, Pace and Stern2 computed a rank-order correlation between the mean scores of student reSponses to the Ipeex and the mean scores of faculty reSponses. For the two colleges with the largest number of faculty reSpondents, the rank order coefficients were .96 and .88. Thirty faculty members at the community college com— pleted the Evening College Characteristics Index. A rank order correlation coefficient was computed between the mean scores of the faculty reSponses and the mean scores of stu- dent reSponses. The correlation coefficient was .90. lStern, ”Environments. . . ," op. cit., pp. 709—716. 2Pace and Stern, "An Approach to the Measurement. . , ice cit., pp. 269-277. 42 Because all of the testing was done during a single year, the analysis of the data involved the assumption of environmental consistency at the high school level for at least one year. Two checks were made on this assumption. A sample of students (N = 30) at the community college were given the College Characteristics Index. A rank order correlation coefficient was computed between the means of the students' perceptions of press at the community college with those found in Campbell'sl 1963 study. The rank order coefficient was .87. Three groups of students at the community college were tested with the High School Characteristics Index. All the students were 1964 or 1965 graduates of the three local public high schools included in the study. The students were instructed to characterize their high school environment as they remembered it. For the first group (High School #l, N = 21), the rank order correlation be— tween the samples' mean perceptions of press and the means of the 1966 seniors' perceptions of press in that school was .99. For the second group (High School #2, N = 15), the coefficient was .94. For the third group (High School #3, N = 22), the coefficient was .96. A rank order correlation coefficient was also com- puted between the means of students' perceptions of press at the community college and their Activities Index scores lCampbell, op. cit. 43 converted to press equivalents. (The process of this con- version of scores is described in the section on procedures below.) The rank order coefficient computed in this way was greater than .99. Procedures The Activities Index was administered to 340 students in six sections of required freshmen orientation classes at Lansing Community College. The sample included approx— imately one-half of the full-time freshmen enrolled during the fall quarter of 1965. At the same session at which the Activities Index was given, the students were given the Evening College Characteristic Index with instructions on how to complete it. They were asked to complete the Ipeex and return it to the next session of the orientation class. (The take— home procedure was used because it was not possible to obtain further time for the administration of the instru- ments during class meetings.) Sternl stated that the environment indexes may be administered in this fashion without loss of reliability or validity. As a check on the above assumption, one-half of the sample of students involved in the test—retest procedure with the Evening College Characteristics Index (Table 3.1 above) were given the test during a regular meeting of a lStern, Preliminary Manual. . .p, Op. cit. 44 Spring term psychology class. The other half of the sample was given the instrument with instructions to complete it and bring it back to the next class session. Visual in- spection of the test—retest results revealed no marked skewing of scores so the two groups were merged for compu- tation of the final coefficients. Because the Evening College form of the indexes was being given, an explanation of its use was given verbally and in writing to enhance the face validity of the test. The students were told that although no form of the environ- mental indexes had been develOped specifically for the community colleges, the questions in the Evening College form of the test were applicable to a community college so this form was being used. (A copy of the written state- ment which was attached to the test booklets is included in Appendix IV.) A personal data form was also attached to the environ- mental index booklets. The form requested information about the student's educational background, status (full- or part- time student), educational goals and plans about transfer- ring to a four-year college or university, and certain family data. (A COpy of this personal data form is in- cluded in Appendix II.) The subjects were also asked to complete a brief (ten- question) questionnaire which included questions which paralleled and paraphrased the emphases of the ll first— order factors of the environmental indexes. Subjects were 45 asked to indicate their perceptions of the levels of the environment's demands and to state their preferences for each of these ll areas. (A COpy of the Expressed Preference questionnaire is included in Appendix II. The use of the Expressed Preference questionnaire is described in Appendix IV.) Late in the spring term the students who took part in the original testing in the fall term orientation classes were sent a brief questionnaire adapted from the personal data form described above. The questionnaire requested in- formation about the student's current educational and occupa- tional status as well as his educational plans. @.copy of this follow-up questionnaire is included in Appendix II.) Background information (below) was obtained for 172 of the 340 original subjects. The subjects became these community-college final sample. A total of 117 of the final— sample subjects returned the personal data form; 101 re- turned the mail questionnaire. Registrar's Office records were examined to secure data on the subjects who did not return either or both forms. During the second semester ten high schools from the service areas of the community college were contacted and given a brief description of the study. Permission was requested to test samples of 30—50 seniors. Permission was granted in eight of the ten schools. The High School Characteristics Index was given to samples of from 36 to 42 volunteer seniors in the eight schools. All of the 46 testing was done within two months of the seniors' gradua- tion dates. In two of the schools members of the guidance services staff administered the tests. Brief conferences were held with these individuals prior to their administration of the instrument. In the other six schools the tests were given with the assistance of teachers, principals or guidance staff members. Score Transformations In order to make the pattern statistics (Cattell's rp) more readily usable in other computations such as the assign- ment to congruence and continuity categories, all were transformed to their equivalent ggscores. The 172 Activities Index forms were first scored in the regular fashion and totals computed for the second- order factors: Intellectual Orientation, Dependency Needs, Emotional Expression and Educabilipy. These were computed in the manner prescribed in Stern'sl scoring instructions. (The definitions of these second-order factors are given in Appendix I.) In order to make direct comparison with each student's environment possible, the Activities Index sub-scale scores were transformed into the corresponding environmental press measure by combining the 30 scale scores into the factors as defined for the environment indexes. This procedure is lStern, Scoring Instructions. . . . op. cit. 47 the same as that used in Keith'sl study. (The definitions of the ll first-order factors are given in Appendix I.) Definitions The following definitions pertain to the analysis and interpretations of the data derived from the Activities Index tests at the community college. They also pertain to the environment indexes used at the community college and the high schools. 1. Common Beta Press.--This press form is the mean of student or faculty reSponses describing the press of an institution (on the ll first-order factor dimensions). The high school samples each included 36 subjects. One hundred and seventeen of the community college sample students com- pleted the environmental index describing the college. These tests were numbered consecutively and a table of ran- dom numbers entered to select 18 male and 18 female "volun— teers” for Common Beta Press comparisons with the high school samples. 2. Private Beta Press.——This press form is the indi- vidual’s own perception of press on the ll first-order factors of the environment indexes. 3. Deviation pattern congruence.--An assessment of the individual's congruence or fit with the institution's press. Individual's scores on the Activities Index were lKeith, op. cit. 48 transformed into press equivalents. These transformed scores were compared to the means of the sample's percep- tions of press on these dimensions (Common Beta Press). The pattern comparisons were expressed mathematically using Cattell's rp pattern analysis statistic. In this measure of congruence only those deviations ip excess 9: one deviation from the Common Beta Press were used in the computation of the rp pattern analysis statistic. 4. Adjustedepattern congruence.—-An assessment of congruence or fit between the individual's needs and the institution's press computed in the same way as (3) above, except that only deviations which were below the mean of the Common Beta Press on the intellectual dimension or above the mean on the non-intellectual dimension were used in computing the rp statistic. 5. EXpressed congruence.——A questionnaire assessment of the individual's stated perceptions of press and his stated preferences. (A c0py of this brief, ten-question questionnaire is included in_Appendix II. The method of use is described in Appendix IV.) 6. Pattern congruence.--An assessment of fit between the individual's needs and the environment's press which was computed using the raw score differences between the individual subject's converted need scores and the press scores of the institution. Direction of deviation was not considered. 49 7. Private Beta Press congruence.--This measure of fit was computed in the same way as Pattern congruence (6 above) but the individual's own perception of the insti- tution's press (Private Beta Press) was used in the com— parison of his need score profile to the institution's press profile. 8. Continuity.--Continuity was defined as the con- tinuation or change in press—need profile congruence. Subjects were assigned to four groups: a. Positive discontinuity: major improvement of fit or congruence; b. Essential Continuity II: minor positive change; c. Essential Continuity I: minor negative change; d. Negative discontinuity: major negative change of fit or congruence. 9. Adaptation Level.—-The subject's level of adap- tation to incongruence or lack of fit between his needs and environmental press was inferred from his congruence score at his high school (Appendix IV). Subjects were assigned to four levels of adaptation (High to Low). The following are Stern'sl definitions of the Intel- lectual or Non-Intellectual dimensions of his environmental indexes and the ll (first-order) factors which make up these two dimensions. Intellectual Climate.—-This dimension includes the more conventional aspects of academic program including lStern, Scoring Instructions. . . , Qp. cit. 50 (a) staff and facilities, (b) standards of achievement set by students as well as faculty, and (c) Opportunities for the development of self-assurance. The intellectual climate is also marked by (d) non-custodial student per- sonnel practices and (e) an absence of vocationalism. The first-order factors included are: l. Aspiration Level. A high score on this factor indicates that the high school or college encourages students to set high standards for themselves in a variety of ways. (Score Sum: Counteraction, Change, Fantasied Achievement, and Understanding.) Intellectual Climate. All of the items con- tributing to this factor reflect the qualities of staff and plant Specifically devoted to scholarly activities in the humanities, arts and social sciences. (Score Sum: Reflective- ness, Humanities-Social Sciences, Sensuality, Understanding, and Fantasied Achievement.) Student Dignity. This factor is associated with institutional attempts to preserve student freedom and maximize personal respon- sibility. (Score Sum: Objectivity, Assurance, Tolerance.) Academic Climate. This factor stresses academic excellence in staff and facilities in the con- ventional areas of natural science, social science and the humanities. (Score Sum: .Humanities-Social Science, Science.) Academic Achievement. Schools high on this factor set high standards of achievement for their students. (Score Sum: Achievement, Energy, Understanding, Counteraction, and Con- junctivity.) Self-Expression. Schools high on this factor offer students Opportunities for the develop- ment of leadership potential and self- assurance through such activities as public discussions and debates, student drama and music. (Score Sum: Ego achievement, Emo- tionality, Exhibitionism and Energy.) 51 Non-Intellectual Climate.—-This dimension shares Self-Expression with the preceding (Intellectual Climate) dimension. The highest loadings, however, are connected with three factors involving a high level of organization of student affairs, both academic and social. The remaining factors are associated with student play and an emphasis on vocational and technical courses. 7. 10. II. Group Life. The four scales on this factor are concerned with various forms of mutually supportive group activities among the student body. (Score Sum: Affiliation, Supplication, Nurturance, and Adaptability.) Academic Organization. The various components of this factor may be regarded as environmental counterparts for an individual's need for orderliness and submissiveness. (Score Sum: Blame Avoidance, Order, Conjunctivity, Delib— eration, Deference and Narcissism.) Social Form. Factor 9 is related to Factor 7 (Group Life) but emphasizes the welfare com— ponents of group life: opportunities for inter- personal assistance. (Score Sum: Narcissism, Nurturance, Adaptability, Dominance and Play.) Play—Work. Schools high on this factor offer Opportunities for participation in a form of social life reminiscent of the pOpular culture of the l920's. (Score Sum: Sexuality, Risk- taking, Play and Impulsiveness.) Vocational Climate. The items of this factor emphasize practical, applied activities; the rejection of aesthetic experiences, and a high level of orderliness and conformity in the student's relationships with the faculty, his peers and his studies. (Score Sum: Practicalness, Puritanism, Deference, Order and Adaptiveness.) (Definitions of each of the 30 sub-scales mentioned in the ”Score Sum" sections above are given in Appendix I.) 52 Statistical Hypotheses The following are the hypotheses of this study stated in testable form: I. Hypotheses pertaining to Press at the High Schools and the Community College. I. No differences will be found between the Intellectual Press dimension at the high schools and the community college. No differences will be found between the Non-Intellectual Press dimension at the high schools and the community college. No differences will be found between the Patterns of press at the high schools and the community college. No differences will be found between the Rank Order of press factors at the high schools and the community college. Hypothesesepertaining to Environmental Continuity. 1. Negative discontinuity students will leave the community college and not continue their education elsewhere at a significantly higher rate than positive discontinuity students. Alternate Hypotheses: a. High achievement, negative discontinuity students will indicate intention to transfer to another institution prior to completing an Associate-in-Arts 2. 53 degree at the community college at a significantly higher rate than high achievement, ppsitive discontinuity students. b. Low achievement, negative discontinuity students will leave the community college and not continue their education elsewhere at a significantly higher rate than low achievement, positive discontinuity students. Positive discontinuity students will perform at a significantly higher level than negative dis— continuity students at the community college. Alternate Hypothesis: a. Low achievement, positive discontinuity students will perform at a significantly higher level than low achievement, negative discontinuity students. III. Hypotheses pertaining to Congruence at the Community College. 1. Students who are low in Deviation pattern con- gruence will leave the community college and not continue their education elsewhere at a significantly higher rate than high Deviation pattern congruence students. 54 Alternate Hypotheses: a. Low Deviation pattern congruence, high achievement students will indicate intention to transfer to another insti- tution prior to completing requirements for an Associate-in-Arts degree at the community college at a significantly higher rate than high Deviation pattern congruence, high achievement students. b. Low Deviation pattern congruence, low achievement students will leave the community college and not continue their education elsewhere at a significantly higher rate than high Deviation pattern congruence, low achievement students. c. High Achievement, low Deviationepattern congruence students who are low in educability will indicate intention to transfer to another institution prior to completing requirements for an Associate—in—Arts degree at the community college at a significantly higher rate than high achievement, low Deviation _pettern congruence students who are high in educability. 55 d. Low achievement, low Deviation pattern congruence students who are low in educability will leave the community college and notcontinue their education elsewhere at a significantly higher rate than low achievement, low Deviation pattern congruence students who are high in educability. 2. High Adjusted pattern congruence students will perform at a significantly higher level than low Adjusted pattern congruence students at the community college. Alternate Hypotheses: a. Low Adjustedpattern congruence students who are high in educability will perform at a significantly higher level than low Adjusted pattern congruence students who are low in educability. b. High Adjusted pattern congruence, positive discontinuity students will perform at a significantly higher level than low Adjusted pattern congruence, negative discontinuity students. IV. Hypotheses pertaining to Adappation Level. 1. Low Deviation pattern congruence students at the community college who are low in Adaptation level (to incongruence) will leave the college and not 56 continue their education elsewhere at a sig- nificantly higher rate than low Deviation pettern congruence students who are high in Adaptation level. 2. Low Adjustedepattern congruence students at the community college who are high in Adapta- tion level will perform at a significantly higher level than low Adjusted pattern con- gruence students who are low in Adaptation level. Alternate Hypothesis: a. Low achievement, low Adjusted pattern congruence students who are high in Adaptation level will perform at a sig- nificantly higher level than low achieve- ment, low Adjusted pattern congruence students who are low in Adaptation level. Exploratory Hypotheses. The two exploratory hypotheses were added to the study to examine more fully the relationship between environ- mental continuity Or discontinuity and academic per- formance. Because these hypotheses were not con- structed prior to the initiation of the study they cannot be considered testable within the framework of the design of the study. They held such promise for further conceptualization, however, that they were 57 added during the course of the analysis. The findings, of course, must be considered tentative and suggestive only. I. There will be significant differences among correlations between high school grade point averages and community college grade peint averages according to the degree of similarity between the two environments. That is, the correlations between high school performance and college performance (as measured by grade point averages) will be significantly different for students from a high school in which environ- mental demands are similar to those at the community college compared to the correlation for students from a high school in which the demands are unlike those at the community college. 2. The correlation between high school performance and college performance (as indicated by grade point averages) will be significantly higher for students whose environmental demands-personal need ”fit" (or congruence) remains relatively constant compared to that for students who experience distinct positive or negative change. Analysis In the following section the assumptions underlying the statistical models used in analyzing the data are presented and discussed. 58 Assumptions: t—test In the evaluation of the difference between two means through the use of the p-test, the implicit assumption is made that the population variances from which the samples are drawn are equal. In rejecting the null hypothesis, the assumption is made that the true difference lies between the means and not in the variances involved.l Although the senior author of the test instruments used in this study uses the petest routinely in his analysis of press-press and press-need differences,2 F-tests will be used to detect any radical departures from homogeneity of variance. In addition, the use of the p-test involves the assump- tion of normal distribution of the ”numerator" variable. Edwards,3 however, states that departures from normality are not crucial so long as the number of observations in each sample is sufficiently large. The smallest samples involved in the basic use of the p—test in this study in- clude 36 subjects. With groups of this size, departures from normality should not invalidate the conclusions drawn. lAllen L. Edwards, Statistical Methods for the Behavioral Sciences (New York: Rinehart and Co., 1954), pp. 270—27l. . 2 . Stern, Scoring Instructions. . . , gp. Cit. 3Allen L. Edwards, Experimental Design in Psycho- logical Research (New York: Holt, Rinehart and Winston, 1960), p. ll2. 59 Assumptions: Cattell's rB Cattell's rp profile or pattern analysis statistic is based upon two restrictions: (l) that the dimensions be independent in the pOpulation, and (2) that the scores be in standardized form. That is, that each variable, regardless of its metric, be given equal weight. Hornl indicated that when the assumption of independ- ence is not warranted rp will be inflated and therefore (using the significance tables provided) appear to be more noteworthy than it is. He added that the psychologist will Often find it difficult to justify the assumption that the K dimensions are, in fact, independent. The fact that the ll first-order factors used in the principal analysis of the data were derived from factor analysis is relevant here. Such factor independence supports the use of the statistic with such data. It does not, however, preclude the possie bility that the dimensions are not independent. The prin- cipal uses of the statistic in this study were: (I) to assign subjects to broad categories of ”conguence" or fit between their expressed needs and environmental demands and (2) to measure dissimilarity between patterns of institu- 2 tional demands in intellectual and "social" areas. In such 1J. L. Horn, "Significance Tests for use with r and Related Profile Statistics," Educational and Psycho-p logical Measurement, XXI, No. 2 (1961), pp. 363-370. 2The above (r ) statistic is described more fully and several levels of Horn's Significance tables reproduced in Appendix IV. 6o broad classification or in expressing dissimilarity, the "inflationary" effect of non-independence between the K-dimensions should not invalidate the use of the statistic. Assumptions: Rank—Correlation Coefficient The rank-correlation coefficient has the advantage that no assumptions are made about the distributions of X or Y. It does involve the implicit assumption that the objects, items or persons ranked be ranked by the same in- dividual or individuals. This implicit assumption was met in the use of the coefficient with data from three high schools and the community college. In the cases in which the assumption was not met, the analysis was made and the exceptions noted. Assumptions: Point—biserial correlation coefficient The chi-square statistic was used in several of the major analyses of the data. The null hypothesis was ac- cepted or rejected in each case on the basis of the out- come of the chi—square test. The relationship was, however, explored further in several cases through the use of the point-biserial correlation coefficient. Use of the point—biserial correlation coefficient involves no assumption as to the distribution of the dichot- omous variable. Generalization is made only to a universe of samples of size N having the same fixed number of cases 61 NO and N1 in the dichotomous categories.l It is also assumed that Y is normally distributed within each X cate— gory and that the two Y distributions have the same variance. The point-biserial correlation coefficient was used in an exploratory and supplemental manner. Because the statistic was not used in basic hypotheses—testing, the assumptions were tested informally through inspection of the data. The inSpection did not reveal any serious depar— tures from these assumptions. Assumptions: Product—moment correlation The use of the product—moment correlation assumes linearity of regression, homoscedasticity, and normal dis- tributions for the variables.2 Again as above, the product- moment correlation was used in an exploratory and supple- mental manner. An informal inSpection of the data did not reveal any radical departures from these assumptions. Assumptions: Analysis of Variance Four basic assumptions are made when the analysis of variance is used:3 1. Observations within groups must be mutually independent° that is each observation is in no way related to other observations. 1 Helen M. Walker and Jospeh Lev, Statistical Inference (New York: Henry Holt and Co., 1953), p. 271. , 2Quinn McNemar, Psychological Statistics (2d ed.; New York: John Wiley and Sons, I955), pp. 122—143. 3E. F. Lindquist, Design and Analysis of EXperimentS in Psychology and Education (Boston: Houghton-Mifflin Co., 1953): PP. 73-78. 62 In this study, each of the sub-groups were from hypothetically different treatment pOpulations, and students were randomly selected to make up the samples for each part of the study. 2. The variance of the criterion measures is the same for each of the treatment pOpulations. This assumption of homogeneous variance can be violated without serious risk, as Shown by the Norton study cited in Lindquist.l This assertion is supported by Hayes,2 under the condition that the number of cases in each sample is the same. In the major analyses of the data (those involving four levels of environmental continuity and four levels of congruence) the subjects were divided into four equal groups. 3. The distribution of the criterion measures for3 each treatment must be normal. Both Lindquist and Hayes2 have stated, however, that the nor- mality assumption is not important if the pDin each sample is large. In this study, the prin— cipal sub-groups contain 43 cases. 4. The mean of the criterion measures must be the same for each of the treatment pOpulations (the null hypothesis). The analysis of variance indicates whether or not dif- ferences exist between the means of groups being studied, but it does not indicate which means are different. Duncan has developed one of the tests that can accomplish this purpose: Duncan's New Multiple Range Test. This test uses the square root of the error mean square of the analysis of variance in the computation of the standard error of the mean. The difference between any given set of means must exceed the Significant studentized range for that pair of lIbid., pp. 78-90. 2William L. Hayes, Statistics for Psychologists (New York: Holt, Rinehart and Winston, 1963), p. 379. 3Lindquist, op. cit. 63 means (the means having been arranged in order of magnitude) multiplied by the standard error of the mean.1 Summary A sample of 340 first—time, full—time freshmen students were tested with Stern's Activities Index in fall orientation classes at the community college. The parallel environmental index (The Evening College Characteristics Ipeex) was given to the students during the same orientation period with instructions to return it to the next class session. A ten-item questionnaire composed of questions which paraphrased the ll first-order factors of the environmental indexes was also administered. Subjects were asked to re- port their perceptions of the level of environmental demands and to express their preferences for emphases in these areas. High school environmental background data was obtained for 172 subjects from the above sample. The information was obtained through testing at the high school level. Samples of 36—42 volunteer seniors at eight service area high schools were administered The High School Character- isticS Index, an instrument develOped to parallel Stern's other environmental indexes. Personal data forms were used to obtain information concerning the students' family background, educational plans and plans to remain at the community college or to lEdwards, Experimental Design. . . ,Op. cit., pp. 136- l40. 64 transfer to another institution prior to completing a full two years' work. A follow—up questionnaire was mailed to the entire sample (N = 340) in the spring term to obtain information about changes in the students' status. Regis- trar's records were used to supplement the mailed-returns. A local validity check was made by comparing faculty and student responses to the Evening College Characteris- tics Index. A rank order correlation was computed between the means of faculty reSponseS (N = 30) on the ll first- order factors of the index and the means of student responses. The resulting coefficient was .90. A sample of 50 students was given the Evening College Characteristics Index in Spring-term psychology classes and retested after a period of 30 days. The test-retest coefficients for the ll first-order factors averaged .83. A small sample (N = 50) of tests from the original sample were selected and checked using a split-half, odd-even procedure. The average first—order factor scale reliability was .85 (when the correlational formula was supplemented with the Spearman-Brown Prophecy Formula). Although sub-scale reliabilities (for the 30 sub- scales) averaging .5 had been reported for the High School Characteristics Index, a separate reliability check was made. The ll first—order factors of the index were used. A Split-half, odd-even technique was used and the formula supplemented with the Spearman-Brown Prophecy Formula. The average first-order factor scale reliability was .74. 65 Thirty students were given the College Character- istics Index. The mean factor scores of this sample were compared to a l963 sample studied with the same instrument at the community college. The rank-order correlation coefficient between the ll factor scale means in the two samples was .87. Three groups of students (N = 21, 15, and 22) at the community college who were 1964 or l965 graduates of three local high schools were given the High School Charac- teristics Index. The samples' mean scores on the ll factors were compared to the mean responses of 1966 seniors tested in the same three schools. The rank-order correlations for the three schools were .99, .94 and .96. Statistical hypotheses concerning differences in press amounts and patterns in the high schools were presented. Predictions of significant difference in intellectual and non-intellectual demands made upon students in the high school compared to the community college environments were made. Predictions of significant differences in performance and rates of attrition were made for varying levels of environmental continuity, congruence and levels of adap- tation. Specifically, differences were predicted in both performance levels and attrition rates for students whose personal preferences fit the demands of the community col- lege better than his high school's demands or worse than his high school's demands. Predictions were also made with reference to the students' "fit" (correspondence of his 66 preferences to the institution's demands) at the community college level both for performance and rates of attrition. Predictions were also made that the student who had ex- perienced demands not to his liking at the high school level would react differently to such lack of fit at the college level (compared to the student who had liked the demands of his high school environment). The statistical methods used in the analysis of the data were presented and the assumptions underlying them discussed. CHAPTER IV ANALYSIS OF RESULTS Section I of this chapter includes the analysis of differences in perceptions of environmental demands in the high schools and at the community college. (Definitions of environmental factors are given in Appendix I.) The data in Section II were obtained through compar— isons made between the congruence ("fit") of the individualks preferences to the demands in his high school environment compared to that at the community college. (The assignment of subjects to environmental continuity categories is dexcribed in Appendix IV.) In Section III the individual's congruence or fit at the community college is analyzed in relationship to perfor- mance and attrition. In the measures of student drOpout (attrition), Deviation pattern congruence (as defined in Chapter III, supra, p. 47) was used. Adjusted pattern con- gruence (as defined in Chapter III, epppe, p. 48) was em- pkwed with the measures of student performance. In both the investigation of performance and attrition the relation— ship between patterns of individual needs and environmental demands was expressed mathematically using Cattell's rp pattern analysis statistic. (This statistic is described in Appendix IV.) 67 68 The results of the investigation of Adaptation level (as defined in Chapter III, supra, p. 49) are presented in Section IV. (See Appendix IV for details of the process used to assign subjects to Adaptation level categories.) Two exploratory hypotheses are presented in Section V relating to environmental continuity and performance. The .05 level is used throughout the analyses as the criterion for the rejection of the null hypothesis. The continuity, congruence and adaptation level meas- ures were related to performance for fall and winter quar- ters. That is, these variables were related to performance as expressed by cumulative (fall + winter) grade point averages. All assessments of attrition rates were based upon the entire academic year. Section I which follows involved the testing of hy- potheses pertaining to press at the eight high schools and the community college. Section I: Press Comparisons were made between perceptions of environ- mental press in eight high schools and in Lansing Community College (as defined in Chapter I, epppe, p. 14). The sub- jects in the high schools were seniors within two months of graduation. The high school subjects were given The High School Characteristics Index. The community college sub- jects were freshmen who were tested late in the fall quarter. The community college subjects were given The Evening 69 College Characteristics Index, an instrument which parallels the high school form. Summaries of the comparisons on the two second-order and the 11 first—order dimensions for each of the eight high schools are presented in Appendix IV (in raw data form). Comparisons were made between perceptions of environ— mental press or demands at each of eight high schools within the service area of the community college and the percep- tions of press or demands at the community college. In each of the eight sets of comparisons, the two seCOnd—order and the 11 first-order factors defined above were compared. The null hypothesis tested for each of the comparisons states symbolically, was: HO: “1 - u2. The alternate hy- pothesis, stated symbolically, was: HA: ”1 # u2. I. Hypotheses pertaining to Press at the High Schools and the Community College. 1. No differences will be found between the Intel- lectual Climate dimension at the high schools and the community college. Significant differences were found between the Intel- lectual Climate dimensions in each of the eight comparisons between the high schools and the community college using the p-test (See Table 4.1). All eight of the differences were significant beyond the .05 level. The null hypothesis in each case was rejected. 70 TABLE 4.1.--Total Intellectual Press comparisons (each of the eight high schools compared to the community college). p values School #1 (Local Public) -4,15** School #2 (Local Public) -5.76** School #3 (Local Public) -4,19** School #4 (Parochial) -3.39** School #5 (Non-local Public) -6.03** School #6 (Non-local Public) -6.69** School #7 (Non-local Public) -6.42** School #8 (Non—local Public) -7.46** Legend: **Significant at 1 Community College press greater) 1 Community College press greater) O *Significant at .O ++Significant at O O ( ( 1 (High School press greater) 5 ( +Significant at High School press greater) a. No differences will be found between the Aspiration Level factor at the high schools and the community college. b-f. Ibid for Intellecutal Climate, Student Dignity, Academic Climate, Academic Achievement, Self-Expression. Significant differences (at the .05 level or beyond) were found between the perceptions of press in the high schools compared to the community college in 32 of the 48 71 comparisons using the p—test. In every case the community college press was perceived as greater than the high school press. The results of the analyses on the 48 first-order factors are summarized in Table 4.2. 2. No differences will be found between the Non-Intellectual Press dimension at the high schools and the community college. Significant differences were found between the Non— Intellectual Press dimension in two of the eight comparisons using the petest. Both of these differences were significant at the .05 level. In one case (School #2), the community college press was greater. In the other comparison (School #4), the high school press was greater. The results of the analysis on this dimension are presented in Table 4.3. a. No differences will be found between the Group Life factor at the high schools and the community college. b-f. Ibid for Academic Organization, Social Form, Play—Work and Vocational Climate. (Self-Expression is shared byihetmo second— order dimensions. It is included in the Non-Intellectual dimension table, but was not re-analyzed.) Significant differences (at the .05 level or beyond) were found between the press at the high schools and the 72 TABLE 4.2.-~First-order factor comparisons for the Six scales of the Intellectual Press dimension: eight high schools - community college. FACTOR ASPIRATION LEVEL School #1** School #3** School #5** School #7** School #2** School #4** School #6** School #8** INTELLECTUAL CLIMATE School #1 = School #3 2 School #5* School #7** School #2 = School #4 = School #6** School #8** STUDENT DIGNITY School #1** School #3** School #5** School #7** School #2** School #4** School #6** School #8** ACADEMIC CLIMATE School #1 2 School #3 = School #5* School #7* School #2 = School #4* School #6** School #8** ACADEMIC ACHIEVEMENT School #1* School #3 = School #5** School #7** School #2** School #4 2 School #6** School #8** SELF EXPRESSION School #1 = School #3 = School #5 = School #7 = School #2 = School #4 2 School #6 = School #8** Legend: (pfvalues) *Significant at **Significant at .05 .01 E Community College greater Community College greater Difference not significant 73 community college on 25 of the 48 comparisons (including the one difference on factor #6, Self—Expression, referred to above) using the p-test. TABLE 4.3.—-Total Non-Intellectual Press comparisons (each of the eight high schools compared to the community college). p values School #1 (Local Public) .12 = School #2 (Local Public) —2.37 * School #3 (Local Public) .99 = School #4 (Parochial) 2.26 + School #5 (Non-local Public) -1.52 = School #6 (Non-local Public) —l.87 = School #7 (Non-local Public) — .57 = School #8 (Non—local Public) —1.21 = Legend: * Significant at .05 (Community College greater) + Significant at .05 (High School Greater) Difference not significant The community college press was perceived as greater than the high school press on the Group Life factor for each of the schools except School #4 (difference not significant). The community college press was greater than the high school press on the Academic Organization factor for Schools #2, 5, 6, 7 and 8. (Differences were not significant for 74 Schools #1, 3 and 4.) The community college press was seen as greater than the high school press on the Social Form factor for Schools #3 and 7. The press at School #4 was greater than the community college press on the Social Form factor. (Differences on the Social Form factor were not significant for Schools #1, 2, 5, 6, and 8.) The Play-Work factor was perceived as significantly greater in each of the high schools than at the community college. Differences on the Vocational Climate factor were not sig- nificant at any of the schools except School #2. The community college press was greater for the Vocational Climate factor than at School #2. The results of the analysis on the 48 first-order factors of the Non-Intellectual dimension are presented in Table 4.4. 3. No differences will be found between the patterns of press at the high schools compared to the community college. The 11 first-order factors were used in these compari- sons. The means of the high school seniors' perceptions of press were used to establish the "pattern" of press at the high school in question. (Data obtained through the use of the High School Characteristics Index.) The means of the community college freshmen sample's perceptions of press at the community college were used to establish the ”pattern” of press at the community college. (Data obtained through the use of the Evening College Characteristics Index.) 75 TABLE 4.4.--First-order factor comparisons for the Six scales of the Non-Intellectual Press dimension: (Each of the eight high schools compared to the community college.) FACTOR SELF—EXPRESSION School #1 2 School #3 2 School #5 = School #7 = School #2 = School #4 2 School #6 = School #8** GROUP LIFE School #1* School #3* School #5** School #7** School #2** School #4 = School #6** School #8** ACADEMIC ORGANIZATION School #1 2 School #3 2 School #5** School #7* School #2** School #4 = School #6** School #8** SOCIAL FORM School #1 2 School #3** School #5 = School #7* School #2 = School #4-- School #6 = School #8 = PLAY-WORK School #l++ School #3++ School #5++ School #7++ School #2++ School #4+ School #6++ School #8++ VOCATIONAL CLIMATE School #1 = School #3 = School #5 = School #7 = School #2** School #4 = School #6 = School #8 = Legend: (p-values) *Significant at .05 Community College greater **Significant at .01 Community College greater = Difference not significant. +Significant at ++Significant at .O5 .01 E High School greater) High School greater) 76 The scores were standardized using the means and standard deviations of perceptions of press on each of the 11 factors. The mean and deviations were computed using the data from all of the high schools and the com- munity college. This standardization was necessary because the use of Cattell's rp statistic requires that all scores be in standard score form. The null hypothesis tested for each of the eight comparisons, stated symbolically, was: HO: rp ; -.311. (This value of rp is required for significant diesimi~ larity by Horn'sl tables of significance at the .05 level (for pattern comparisons involving 11 dimensions).2 The alternate hypothesis, stated symbolically, was: HA: rp < -.311. No significant differences were found in the pre- dicted direction. The rp coefficients for the schools were: School #1 (Local Public): .36; School #2 (Local Public: .18; School #3 (Local Public): .57; School #4 (Parochial): .71; School #5 (Non-local Public): .30; School #6 (Non—local Public): .28; School #7 (Non-local Public): .14; and School #8 (Non-local Public): .19. The correlation for School #3 (Local Public) (.57) was significantly positive at the .02 level. The corre- lation for School #4 (Parochial) (.71) was significantly lHorn, Op. cit., pp. 363-370. 2The r statistic is described more fully in Appendix Iv.p 77 positive at the .01 level. The null hypothesis was ac- cepted in each of the eight comparisons. The results of the analysis of patterns of press at the high schools (compared to the community college) are presented in Table 4.5. TABLE 4.5.-—Comparisons of patterns of press at the eight high schools to the pattern of press at the community col— lege--using Cattell's r profile or pattern analysis statistic. School rp coefficient Significance #1 — Local Public .36 = #2 — Local Public .18 = #3 - Local Public .57 + #4 - Parochial .71 + #5 - Non-local Public .30 : #6 — Non-local Public .28 = #7 - Non-local Public .14 = #8 - Non—local Public .19 = Legend: Difference not Significant Difference Significant at .05, dissimilarity + Correlation Significant at .05, Similarity 4. No differences will be found between the rank order of press factors at the high schools and the community college. Rank order correlations were computed between the order of press factors of the Common Beta Press (means of the samples' perceptions of press on the 11 environmental index factors) at the eight high schools and the community college. The null hypothesis tested for the eight schools, stated symbolically, was: HO: r ;=-.29. The alternate S hypothesis, stated symbolically, was: H rq < —.29. A‘ Significant differences were found between the rank order of press factors in six of the eight high schools when these were compared (individually) to the rank order of the same factors at the community college. The (rs) were significantly negative in these six cases. The null hypotheses, then, were rejected in six cases; accepted in two. The correlations were as follows: School #1, -.39; School #2, -.41; School #3, -.32; School #4, -.37; School #5, .07, School #6, -.39; School #7, .03; and School #8, —.87. The results of the analyses of the rank order of press factors are given in Table 4.6. (All of the rank-order comparisons were made using the means of the eight high school samples' perceptions of press at the high schools and the community—college sample's perceptions of press at the community college.) 79 TABLE 4.6.-—Comparisons of the rank-order of press factors at the eight high schools and the community college. School Significance School #1 - Local Public r8 = —.39 - School #2 - Local Public rS = -.41 - School #3 - Local Public rS = -.32 - School #4 - Parochial r8 = -.37 - School #5 — Non-local Public rS = .07 = School #6 - Non-local Public r8 = -.39 - School #7 — Non—local Public rS = .03 = School #8 — Non—local Public rS = —.87 - Legend: — Difference significant at the .05 level 2 Difference not significant Section II: Continuity Section II which follows deals with the data obtained through environmental comparisons (in the transition from the high school to the community college). Environmental continuity (as defined in Chapter III, supra, p. 49) was related to measures of performance and attrition. The sub- jects were divided into two groups, low and high achievement (as defined in Chapter I, supra, p. 15) for the major 8o analyses. (Details of the method used to assign subjects to continuity categories are given in Appendix IV.) II. Hypotheses pertaining to Environmental Continuity. The following hypotheses pertaining to the effects of environmental continuity and discontinuity upon attrition were tested. 1. Negative discontinuity students will leave the community college and not continue their education elsewhere at a significantly higher rate than Positive discontinuity students. Subjects' need profiles converted to press equivalents were compared to the press at their respective high schools and to the press at the community college. Cattell's rp pattern analysis statistic was used to express this rela— tionship mathematically. Subjects were divided into four groups or levels of ”goodness of fit." Those who remained in the same category of "goodness of fit" at the community college were described as experiencing Essential (environ- mental) continuity. Those who were classified in a higher category of fit at the community college (than in high school) were classified as Positive discontinuity students. Those who were classified in a lower category of fit were described as Negative discontinuity students. The null hypothesis, stated symbolically, was: 2 2 H:XéX.05° O The alternate hypothesis, stated 8l ,2 2 A' X>X.05 chi-square was .884. The null hypothesis was accepted. symbolically, was: H (or 7.80). The value of Alternate Hypotheses: a. High achievement, Negative discontinuity students will indicate intention to transfer to another institution prior to completing requirements for an Associate-in-Arts degree at a Significantly higher rate than pigp achievement, Positive discontinuity students. High achievement was defined as having a cumulative high school grade point average above the median for entering freshmen at the community college; low achievement, having a cumulative grade point average below the median. Only academic subjects were considered: English, foreign languages, mathematics, science and social science. Nega- tive and Positive discontinuity were defined as indicated in (1) above. The null hypothesis, stated symbolically, was: 9 H : x“:; X2 0 05. The alternate hypothesis, stated Symbol- . 2 2 Ically, was: HA: x >x .05 of chi—square was 5.18. The null hypothesis was accepted. (or 7.80). The computed value b. Low achievement, Negative discontinuity students will leave the community college and not continue their education elsewhere at a Significantly higher rate than low —_ achievement, Positive discontinuity students. The null hypothesis, stated symbolically, was: HO: xé é=x2 05. The alternate hypothesis, stated symbol— ically, was: HAzx2>x;205 (or 7.80). The computed value of chi—square was 9.02. The null hypothesis was rejected. 2. Positive discontinuity students will perform at a significantly higher leven than Negative discontinuity students at the community college. Students were divided into four groups on the basis of their congruence scores in high school. These pattern congruence scores were compared to the pattern congruence scores at the community college. Subjects were assigned to four groups on the basis of the relative Similarity of their scores in the two settings (continuity). Those whose scores changed significantly'(posflive1y or negatively) were assigned to separate categories. The four groups were: Positive discontinuity (significant positive change in fit); Essential continuity II (some improvement); Essential continuity I (some deterioration): and Negative discon- _pipeipy (significant negative change). (The details of this assignment process are described in Appendix IV.) The null hypothesis, stated symbolically, was: Hozill = “2. The alternate hypothesis, stated symbol- ically, was: HA: ul>u2. An analysis of variance was computed using the four continuity levels and cumulative (fall + winter) grade point averages. The F ratio of 8.1 was Significant beyond the .05 level. The mean of the Positive discontinuity group 83 was 2.443 of the Negative discontinuity group: 2.36. The Shortest Significant range was .30. The null hypothesis was accepted. The results of the analysis of variance are presented in Table 4.7. The results of Duncan's New Multiple Range Testl are presented in Table 4.8. TABLE 4.7.--Analysis of variance — con- tinuity (4 levels) and cumulative (fall + winter) grade point average. Source Sums of Mean F “ Squares Square Ratio Category Means 9.01 3 3.00 6.02* Within 83.69 168 .4982 *Significant at .001 2. Low achievement, Positive discontinuity students will perform at a Significantly higher level than low achievement, Nega— tive discontinuity students at the com- munity college. The four divisions were continued in this analysis (Positive discontinuity, Essential continuity II, Essential continuity I, and Negative discontinuity) but the subjects were divided further on the basis of achievement. (High and low achievement having been defined as being above or lEdwards, op. cit., pp. 136-140. 84 below the median of entering freshmen in terms of high school grade point average.) TABLE 4.8.——Duncan's New Multiple Range Test applied to the differences between four means (continuity - fall + winter G.P.A.) Group 11 Group I Group IV Group III (Range**) 2.54 2.44 2.36 2.10 .10 ' .30 .18 .31 .44 .32 L_ 1 f _i **Shortest Significant Range Any two means not underscored with the same double line ([1 J ) are Significantly different. Groups are numbered from I (Highest) to IV (Lowest) continuity The null hypothesis, stated symbolically, was: The alternate hypothesis, stated symbolically, 31 Ill 112- S m c) (1) HA: “1:112. An analysis of variance was computed using the four :Levels of continuity and cumulative (fall + winter) grades. The F ratio of 2.58 was significant beyond the .05 level. Tflie Positive discontinuity group mean was 2.25 for the low axzhievement students. The Negative discontinuity group rrlean was 1.86 for the low achievement students. This dif— fEerence (.39) exceeded the difference required at the .05 85 level applying Duncan's New Multiple Range Test adapted for use with unequal numbers of replications.l The null hypothesis, therefore, was rejected. The results of the analysis of variance are presented in Table 4.9. The results of the application of Duncan's New Multiple Range Test adapted for use with groups with unequal numbers of replications are presented in Table 4.10. TABLE 4.9--‘Analysis of variance — con— tinuity (4 levels) and achievement (2 levels) X performance (indicated by cumulative (fall + winter) grade point average). Source Sums of Mean F ‘ Squares Square Ratio Category Means 9.08 7 1.30 2.58* Within 82.62 164 .5038 *Significant at .025 lClyde Young Kramer, "Extension of Multiple Range Tests to Group Means with Unequal Numbers of Replications," Biometrics, 12 (1956), pp. 307—310. 86 MAB E 4 lO.--Duncan's New Multiple Range Test (adapted for use with unequal replications) applied to Low Achievement students' performance in terms of cumulative (fall + winter) g1 rade point average. Subjects stratified by 4 levels of continuity. Shortest Group I Group II Group III Group IV Significant Range 2.25 2.13 1.99 1.86 .12 .31 .26 .33 .39 .34 L_ J L_ 1 Any two treatment means not underscored by the_eame double ([:::::::3 ) line are Significantly different. Section III: Congruence Levels of congruence or fit between subject’s inven- toried activities preferences (Activities Index) and envi— ronmental demands as expressed by the Common Beta Press (means of the samples' perceptions of press on the 11 Evening College Characteristics Index factors) are related to performance and rates of attrition in Section III. III. Hypotheses pertaining to Congruence at the Community College. 1. Students who are low in Deviation pattern congruence will leave the community college and not continue their education elsewhere 87 at a significantly higher rate than high Deviation pattern congruence students. Deviation pattern congruence was computed as follows: congruence between students' converted need scores (con- verted to press equivalents) were compared to the community college environmental demands (Common Beta Press). De- viations which exceeded one standard deviation above or below the perceptions of press were included in the compu— tations. Scores which fell within one deviation of the mean were not computed as deviant. The statistic used was Cattell‘s rp pattern or profile analysis statistic (in the pattern comparisons). The null hypothesis, stated symbolically, was: 2 H : X2 <:x .05. The alternate hypothesis, stated symbol- 0 =: ically, was: HA: X2 > x2 05 (or 7.80). The computed value of chi-square was 6.26. The null hypothesis was accepted. Alternate Hypotheses: a. Low Deviation pattern congruence, high achievement students will indicate intention to transfer to another insti— tution prior to completing requirements for an Associate-in-Arts degree at the community college at a significantly higher rate than high Deviation pattern congruence, high achievement students. 88 The null hypothesis, stated symbolically, was: H : x g x q The alternate hypothesis, stated symbol- ically, was: HA: x2 >X2 05 (or 7.80). The computed value of chi-square was 7.82. The null hypothesis was rejected. b. Low Deviation pattern congruence, low achievement students will leave the community college and not continue their education elsewhere at a significantly higher rate than high Deviation pattern congruence, low achievement students. The null hypothesis, stated symbolically, was: HO: X2 ; x2 05. The alternate hypothesis, stated symbol— ically, was: HA: X2 >X2 05 of chi-square was 5.01. The null hypothesis was accepted. (or 7.80). The computed value c. High achievement, low Deviation pattern congruence students who are low in Educa- bility will indicate intention to transfer to another institution prior to completing requirements for an Associate-in—Arts degree at a significantly higher rate than high achievement, low Deviation pattern congruence students who are high in Educa— bility. Educability as defined by Sternl combines elements of both intellectuality and submissiveness. This dimension l . . . Stern, Scoring Instructions. . . , op. Cit. —————- 89 of the Activities Index reflects a strong interest in in- tellectual activities coupled with a need for orderliness and conformity. The null hypothesis, stated symbolically, was: ; x .05. The alternate hypothesis, stated symbol- ically, was: HA: X2=’X2.O5 (or 7.80). The computed value of chi-square was 1.38. The null hypothesis was accepted. d. Low achievement, low Deviation pattern congruence students who are low in Educa- bility will leave the community college and not continue their education elsewhere at a significantly higher rate than low achievement, low Deviation pattern con— gruence students who are high in Educability. The null hypothesis, stated symbolically, was: 2 2 x g x The alternate hypothesis, stated symbolically, 2 2 X .05 square was 6.13. The null hypothesis was accepted. .05 > was: x (or 7.80). The computed value of chi- 2. High Adjusted pattern congruence students will perform at a significantly higher level than low Adjusted pattern congruence students at the community college. Subjects‘ converted need scores were compared to the means of the environmental press scores of the first—order factors of the environmental index at the community college. Converted press scores above the mean of the intellectual dimension factors or below the mean of the non-intellectual go factors were disregarded. Only those deviations which were below the means on the intellectual dimension factors or .——_.—._._ -— above the means on the non-intellectual dimension factors ——. were included in the computations (using Cattell's r pattern analysis statistic). Subjects were divided into four groups in terms of their Adjusted pattern congruence coefficients. Scores levels of congruence and the cumulative grade point averages for the first two terms' work (fall + winter term). Sub— jects were not separated according to levels of previous academic achievement in this analysis. The null hypothesis, stated symbolically, was: U1 2 U2 : U3 2 Nu. The alternate hypothesis, stated symbol- ically, was: H : pl>u2,p3,u4. The F ratio for the four A congruence groups was 3.34, significant beyond the .05 level Groupilf(highest congruence) mean grade point average i, Group III (third highest) was 2.3M; Group II, 2.26 and Group I (lowest congruence) was 2.10. The differences between the means of Group I and Group II as well as the difference between the means of Group I and Group IV ex- ceeded the shortest significant ranges of Duncan's New Mul- tiple Range test beyond the .05 level. The null hypothesis was rejected. 91 The results of the analysis of variance flour levels of congruence and fall + winter cumulative grade point average) are given in Table b.11. The results of Duncan’s New Mul— tiple Range Test applied to the category means are presented in Table H.12. TABLE A.ll.--Analysis of variance of congruence (4 levels) and cumulative grade point average for fall + winter quarters. Sums of Mean F C source Squares d Square Ratio Category Means 5.20 3 1.73 3.3u* Within 87.50 168 .5208 *Significant at .025. Alternate Hypotheses: a. Low Adjusted pattern congruence students who are high in Educability will perform at a significantly higher level than low Adjusted pattern congruence students who are low in Educability. The null hypothesis, stated symbolically, was: HO: pl = “2. The alternate hypothesis, stated symbolically, was: HA: “1 >u2. Subjects were divided into two groups on the basis of congruence (high and low) and subdivided into two groups of Educability. Achievement was not considered. That is, students were not separated according to levels of previous achievement. 92 TABLE M.l2.—-Duncan's New Multiple Range Test applied to the differences between A treatment means-—congruence and fall + winter G.P.A. Shortest Group I Group III Group II Group IV Significant Ranges 2.60 2.3a 2.26 2.10 .26 .31 .34 .33 .50 .34 f j (Group I: High Congruence, Group IV: Lowest Congruence.) Treatment means not underscored by the same double line ( [2:222:21 ) are significantly different. The performance measure used in the analysis of variance was fall + winter (cumulative) grade point average. The F ratio in the analysis of variance was 2.83, signifi- cant at the .05 level. The category means were: .fligh congruence, high Educability: 2.51; high congruence, low Educability: 2.32, low congruence, high Educability: 2.39; and low congruence, low Educability: 2.10. The shortest significant range1 was .3l for the difference between two means. The difference between the low congruence, high Educability and low congruence, high Educability group means was .29. The null hypothesis, therefore, was accepted. lKramer, loc. cit., pp. 307-310. 93 The results of the analysis of variance are presented in Table 4.13. The results of the analysis applying Duncan‘s New Multiple Range Test adapted for use with un- equal groups are presented in Table 4.14. TABLE 4.l3.--Analysis of variance - con— gruence (2 levels) and Educability (2 levels) and fall + winter cumulative grade point average. Sums of Mean F Source Squares df Square Ratio Category Means 4.47 3 1.49 2.83* Within 88.33 168 .5257 *Significant at .05 b. High Adjusted pattern congruence, Positive discontinuity students will perform at a significantly higher level than low Adjusted pattern congruence, Negative discontinuity students. Subjects were divided into four groups on the basis of Adjusted pattern congruence. These four groups were sub- divided into three groups on the basis of continuity: Posi— tive discontinuity (significant positive change in environ- ment), Essential continuity, and Negative discontinuity (significant negative change). An analysis of variance was computed using the resulting ten groups. (There were no subjects who were classified as Positive discontinuity 94 subjects in the lowest (Group I) category of congruence nor 'any classified as Negative discontinuity in the highest {Group IV) category of congruence.) Cumulative (fall + winter) grade point averages were used in the analysis of variance. Subjects were not, however, subdivided on the basis of previous (high school) levels of achievement. TABLE 4.14.--Duncan's New Multiple Range Test applied to 4 treatment means: congruence ( 2 levels and Educability (2 levels) and cumulative ( fall + winter grade point average. Shortest Group I Group II Group III Group IV Significant Ranges 2.52 2.32 2.39 2.10 .20 .31 .13 .33 .42 -35 [ll J .L A. Group I: High Congruence - High Educability Group II: High Congruence — Low Educability Group III: Low Congruence - High Educability Group IV: Low Congruence — Low Educability Treatment means not underscored by the same double line ([::::::3 ) are significantly different. The analysis of variance F ratio of 2.04 was signifi— cant at the .05 level. The difference between the Positive discontinuity, Congruence IV (Highest) and the Negative i discontinuity, Congruence I (Lowest) groups (.54) was not equal to the required difference for significance at the 95 .05 level using Duncan's New Multiple Range Test (difference required was .60). The null hypothesis was accepted. The results of the analysis of variance are presented in Table 4.15. The category means for the ten gro presented in Table 4.16. TABLE 4.15.—-Analysis of variance: con- gruence (4 levels) and continuity ( 3 levels) and cumulative (fall + winter) grade point averages. Source Sums of Mean F Squares Square Ratio Category Means 9.46 9 1.05 2.04* Within 83.24 162 .5138 *Significant at .05 ups are TABLE 4.16.--Group means - congruence (4 levels X conti- nuity (3 levels) and cumulative (fall + winter G.P.A. Continuity Level Congruence Level Positive Essential Negative 4 (High) 2.49 2.65 3 2.48 2.17 2.48 2 2.24 2.01 2.46 1 (Low) 1.98 1.95 Section IV: Adaptation Level Students were divided into four groups on the basis of congruence or fit between their needs and the 96 environmental press at their high schools. The inference was made that students who had experienced low level of fit between their needs and environmental demands (press) at the high school level would be high in level of adapta- tion to such lack of fit (incongruence). The subject's Adaptation Level assignment thus was the inverse of his congruence level score. For the purposes of the analysis of variance, the subjects were divided into two groups on the basis of Adaptation level (high and low) and subdivided into four groups by using high and low Adjusted pattern congruence. Previous levels of achievement were not used to further subdivide the groups. IV° Hypotheses pertaining to Adaptation Level. 1. Low Deviationgpattern congruence students at the community college who are low in Adaptation level to incongruence will leave the college and not continue their education elsewhere at a significantly higher rate than that for 10w Deviation pattern congruence students who are high in Adaptation level. The null hypothesis, stated symbolically, was: HO: x2 g: x2005 (or 7.80). The alternate hypothesis, stated symbolically, was: HA: X2 > x2.05 (or 7.80). The computed value of chi-square was 2.92. The relationship, however, was in the Opposite direction of the predicted one. That is, high Adaptation level, low congruence students 97 dropped out of the college and did not continue their edu- cation elsewhere more frequently than low Adaptation level, low congruence students. The null hypothesis was accepted. 2. Low Adjusted pattern congruence students at the community college who are high in Adap- tation level will perform at a significantly higher level than low Adjusted pattern con- gruence students who are low in Adaptation 22121. The null hypothesis, stated symbolically, was: HO: pl = 42- The alternate hypothesis, stated symbolically, was: HA: ‘8 >12. The analysis of variance F ratio of 1.79 was not significant. The group means were: High congruence, high Adaptation level: 2.53; High congruencé,’ low Adaptation level: 2.32; Low congruence, high Adap4 tation level: 2.23; Low congruence, low Adaptation level: 2.22. A range test was not computed. The null hypothesis was accepted. Section V: Exploratory Hypotheses Students from the four high schools supplying the greatest number of students to the community college were pulled from the study. These high schools were schools #1, 2, 3, and 4 of the original group. The samples were arranged in order in terms of similarity of high school pattern of demands to that at the community college. Each of the sub—groups'high school grade point averages were 98 correlated with their cumulative (year's) grade point averages at the community college. Secondly, subjects were separated on an individual basis according to their personal continuity scores. That is, Positive discontinuity (significant positive change in ”fit” between individual preferences or needs and en- vironmental demands in the transition from high school to community college), Essential continuity (neither signi- ficant positive or negative change), and Negative discon- tinuity (significant negative change) were separated. The three groups' subjects high school grade point aver— ages were correlated with their community college (year's) cumulative grade point average. V. Exploratory Hypotheses pertaining to Continuity. 1. There will be significant differences among correlations between high school grade point averages and community college grade point averages according to the degree of similarity between the two environments. The null hypothesis, stated symbolically, was: 2 r“. The alternate hypothesis, stated symbolically, was: HA: r1 % ru. Systematic variations were found in the correlation coefficients. The coefficients were: .49, .56, .63, and .88. The two coefficients for the schools most and least similar to the community college in terms of patterns of demands (.88 and .49) were significantly different beyond 99 the .05 level of confidence. The null hypothesis, there— fore, was rejected. The results of this analysis are presented in Table 4.17. TABLE 4.17.—-Correlations of cumulative high school grade point averages with cumulative (first year's) grade point averages at the community college: (Subjects separated according to level of similarity of their high school's pattern of demands to that at the community college.) Number Pattern High School G.P.A. High School of Similarity to College G.P.A. Subjects Coefficient Correlation #2 - Local Public (N=34) .18 .49 #1 - Local Public (N=35) .36 .56 #3 - Local Public (N261) .57 .63 #4 - Parochial (N=11) .71 .88 2. The correlation between high school performance and college performance (as indicated by grade point averages)will be significantly higher for students whose environmental demands--personal needs "fit" remains relatively constant compared to that for students who experience significant positive or negative change. The students were separated into the four continuity groups previously used: Positive discontinuity, Essential continuity II, Essential continuity I and Negative 100 discontinuity. The two Essential continuity groups were combined for this analysis. Correlations were computed between high school (cumulative) grade point averages and college (cumulative year's) grade point averages for each of the sub-groups. The null hypothesis, stated symbolically, was: HO: r1 = r2 = r3. The alternate hypothesis, stated symbol- ically, was H r >r r A‘2 133' . The correlation coefficients for the three groups were: 1. Positive discontinuity (Group 1): .43 2. Essential continuity (Group 2): .50 3. Negative discontinuity (Group 3): .62 The differences between the Essential continuity group's coefficient and those of the other two groups were not significant. The null hypothesis was accepted. Table 4.18 which follows contains a summary of the major hypotheses including the models used and the results found. *mCOmHCmQEoo m HHm CH UCsom mooCoComeQ pmop H proHCHCme HOC q HoCozon Co mo. pm AmCOmHCmQEoo m me mo 3 CH oCsom mooCoCoCCHQ pmop p *mCOmHCmmEOU w HHm CH UCCOH mmoCoComeQ pmmp p *mCOmHCmmEOo m HHm l CH UCCOC mmoCoCommHQ pmop H spHcmHC pcmsspm .m mmeHHo HmSHQmHHmHCH .m Ho>oq COHHCCHmm< .H "mCOHomm wCHonHom on Co mmmsm monHoo mpHCCEEoo UCw .mHoogom zmHm map cmmzpmp oCsom on HHHS mooCmCommHQ .omoHHoo thCCEEoo on 0p pommqeoo mHOOCom Cme me pm mmosm opmeHHo stpooHHopCH CH oCsom on HHHz mooCoCommHQ .< .oonHoo szCCEEoo on ow ooCmQEoo mHOOCom Cme on pm mmoCm 0H wCHCHmpCoQ mommeommm mpHSmmm 46662 mHmonomhm .UopmoH mmmoCCoth COmmE Ho mpHsmoC Co szEECmnn.mH.: mqmonm Csz ooCmCmv ”COHmmmCQxHIMHom .H pCmonHCmHm pOC m mHo>oH mo. pm mCOmHCmQEoo w on mo m CH oCsow mmoCmCoMCHQ pCmonHCmHm pOC mCOmHCmQEoo N nHo. Hm COmHCwQSoo H CH UCsom moCoCommHQ pCmonHCme HOC m McCozmQ Co mo. Hm nmCOmHCmQEoo w me mo m CH oCsom mooCoCmmeQ pCmoHMHCme HOC m MCCOHoC Co mo. pm nmCOmHCmQEoo w on Co m CH 6C30C mmoCoCmHmHQ pmmp p pmmp p pmmp 9 Chan M ”mCOPomw WCHSOHHOH me Co mmoam omoHHoo szCCEEoo on on .maoogom anm map cmmspmp oCsom on HHHz mooCmComeQ .m .oonHoo thCCEEoo me op ooCmQEoo mHOOCom CmHm oCH pm mmowm opwEHHo HmeooHHopCHuCoz me CH oCsom on HHHz mooCmCoCmHQ .o C0Hmmmmmxm mHom .w pCoEo>oHCo< oHSmowo< .m omeHHo 0HE®©wo< .3 mpHsmom H0602 mHmmeomxm .smchpcoo --.ma.s mummH Ho. me pm COmHCmQEoo H CH oCsom ooCoCoCmHQ *mCOmHCmQEoo w HHm CH UCzom mooCoCmHMHQ pCmoHMHCme pOC m MUCOHmQ no mo. pm mCOmHCwmeoo w on mo m CH UCsom mooCoComeQ pCmo :HCHCme HOC m mnCo%mQ Co Ho>mH mo. me pm mCOmHCmQEOo w oCH mo m CH UCsom mooCmCoCMHQ pCmonHCmHm pOC H qUComoC Co mo. pm mCOmHCmQEoo m me Co m CH oCsom mooCoCommHQ Q oHpmempm mHmszCw CCopme C m_HHopHmo pmmp p pmmp p pmmp p pmmp 9 Chop m .monHoo mpHCBEEoo on 0p UmamEOo mHoOCom Cme on pm mmoCQ mo mCCmppmm me Cmozpmn pCsow on HHHz mooCoComeQ .m omeHHo HmCoproo> .m xC63-smHC .m 8Com HmHoom .2 COHpmNHwaCo oHEoomo< .m mCHC Csono .m mpHsmom Hmooz mHmonogxm .UQSCHHCODII.®H.U mqm<5 A 104 mo. pm pcmoHCHcmHm pcmoHCHcmHm poz pcmonCHcmHm poz pcmoHCHcmHm no: N mmo. pm pCmonHCmHm mCOm nHCmeoo w me Ho me ohmsvmlHCo ohmsvmnHCo mamswmnHCo COHpmHoCCoo Coonoumem .psonmoso 0p noHMHoC mpCoozpm pCmEm>oHCow 30H How thsCHpCooch o>Hpmwoz .m .COHHCCHmeH CoCCOCw 0H CoCmCme on COHpCmpCH 0p oopmHmC mpCoospm HCoEo>oHCow CmHC COH mszCHpCooch o>Hpmwoz .m .CoHpHCppm 0p UmpmHmC mpHsCHpCoomHo o>Hpmmmz .H "HCMonHCmHm on 0p oCsom on HHH3 mQHCmCOHpmHoC wCHSOHHOH 0C9 .< .thsCHpCoo prCoECOCH>Cm 0p mCHCHwHCom.mmmmeommm .HH .ommHHoo szCCEEoo oCH on ooCMQEoo mHOOCom Cme me pm mCOpowm mmoCC mo CoUCOIMCwC 0C9 Coospoo CCCOH on HHH3 mooCmComHHm .m mpHsmom Hmooz mHmonomzm .UoCCHHCoonu.®H.: mHmoQ CwHC Cpr mme CmCmHC m pm 0C0CzomHo COHpmoswm CHon mSCHpCoo poC on oonHoo mpHCCEEoo on o>moH HHHz mpCoUSHm ooCmst oCmsvmuHCo :Coo CCmppmm,C0HpmH>oQ 30H .H “who: mCHBOHHom on CH COHCHCppw 0H nopmHoC on HHH3 moCmsproo .< .omeHoo szCCEEoo on pm ooCoCCwCoo 0p wCHCHmHCoQ momonomzm .HHH pmoe mmcmm .mpCmospm mpHsCHpCoomHo o>Hp onHpHsz 302 uwwoz qpCmEm>mHCow 30H Cme m_CwoC:Q + Ho>mH ConHC m up Suommmm oonHCw> HHH3 mpCoUSHm mpHsCHpCoomHo mo mHthmC< o>HpHmom ApCmEo>mHCom 30H .m pmoe omCmm .mpCoospm onHpHsz Bmz NijCHpCoomHo m>prwoz m_CmoCCQ + Cpr Ho>mH CoCmHC m pm ooCmHCw> Egompom HHH3 mHCmozpm mo mHmmHmC< HHHCCHpCoomHo o>HpHmom .H "who: mCHZOHHom on CH ooCmECowCoQ 0p oopmHoC on HHH3 szsCHpCoomHo uthsCHpCoo HmpCoECOCH>Cm .m mpHdmmm Hobo: mHmoCHomhm .UwSCHHCOOII.®H.: mqm onsvmnHCo non 30H anoEo>oHCow Cme .mpCoUSHm pCmEo>oHCom 30H nooCosngoo CCmppwmiCOHpmH>oQ CmHC szp szCmCUmCm mnoE ohmzzomHo COHpmosoo CHon osCHpCoo HOC on omeHoo me m>me HHHz mpCmnspm pCmEo>oHCom 30H nmoCmsCm onswmnHCo uCoo Choppmm,CoprH>mQ 30H .muCoozpm pCmao>oHCom CmHC «ooCoSCwCoo CCoprm COHpmH> qu CwHC Cpr szCoonCm 0CoE Comeme on CoHpCop zCH opmoHoCH HHHz mpCmonm pCoEm>oHCom CwHC AooCost oCmsvmnHCo nCoo CCoppmm.C0HpmH>om 30H mpHdmom H0602 mHmmeommm .U03CHHCODII.®H.: mqmHpmmoz pmme oowm flmwmImBCwCoo Choppmm nopmsmo< oHQHpHdz sz 30H CCCH Ho>mH CmCmHC w pm m.CmoCCQ + EHomCom HHHB mHCoUspm mpHsC mOCmHCm> :HHCOomHU o>HpHmom nmoCmCCw Ho mHthwC< uCoo CCmppwm1nopman< CmHm .m Came mmCmm .CHHHHCCUCCm CH 20H oHQHpHsz Zoz omOCp Cpr Ho>mH CmaHC w pm m_CmoC:Q + ECOMCoQ HHH3 mpHHHCmoCCm oOCCHCw> CH mm x mpCoospm ooCoCCw mo mHmszC< uCoo CCoHpmm oopmsno< 30H .m pmme omam .mpCmnspm oHQHpHsz 302 ooCostCoo CCoppmm copmsmp< m_CmoC:Q + mmw Cme Ho>mH ConHC a pm oonHCw> EHomHoQ HHHz mpCmospm moCmsCm mo mHmszC< 1Coo CCmHHCMIConsmU< Cme .H "who: wCHCOHHom 0C9 CH moCmEHomCmQ 0p ompwHoC on HHHB moCoCCwCoo .m .CHHHHCCCCCC CH CmHC mmozp CMCH szCoCUmCm mHoE oCmszmHo COHpmoCCo HHmCH oCCHpcoo pOC UCm monHoo thCCEEoo 0C9 o>mmH HHHz HCHHHCmoCCm CH 36H mCCmCCCm moCostCoo CCoppmm,C0HHmH> maswmnHCo umm 30H anmEo>mHCom 30H .m mpHsmom Hoooz mHmonoahm .CCCCHHCoo--.wH.C CHCCH 108 Ho>mH Ho. me pm pCmoHHHCme mmoCoHommHQ HCCUHCHCmHm Coz CCCUHCHCme Coz .mpCmECOCH> 1C0 039 onp Cmozpon thCmHHEHm mo oopmmv 0p wCHoHOQom .<.m.w mmeHoo 6H .<.m.o HOOCom CmHm Ho mCOHpmHmHHoo mCoEw mmoCoHom CoHHmHmCCoo -CHC HCCCHCHCmHm 6C HHHz mCmCH .H .HCHCCHCCoo 6H mCHCHmHHmm AHCHCoEonmqu mommeomzm .> .Ho>oH COHpmpmeC CH 30H mmOCp Cpr Hm>mH HmCmHC a pm EHOHCoQ HHH3 oonHCw> Hm>mH COHHCHmCUC CH CwHC mpCmpspm mo mHthwCC moCmusCoo CHoppwm cmpmsmc< 30H .m .Ho>mH C0HHMpmmo< CH CwHC mmOCp CMCH 098C CoCmHC a pm ogoCsomHo mzCHpCoo HOC UCw mmeHoo me m>mmH HHHB Hm>oH COHpmeMU< CH 30H mpCocdpm ooCost oszvmuHCo uCoo ChoppwmlCoprH>oQ 30H .H .Ho>oH COprpmmc< 0p wCHCprComimomonomhm .>H mpHsmmm Honoz mHmmeomzm .U®5CHHCOO¢I.®H.: mqmHpm nwoz Co thsCHpCooch o>HpHmom CmeHo 0p nohmmsoo mpCmnspm mpHCCHHCoo HprCmmmm COH .<.m.w mmeHoo op .<.m.c HOOCom CmHm Co proHCHCme COHpmHoHHoo on Coozpon mmoCoCmm pOC mooCoComCHQ CoHpmHoHHoo uHHU pCmonHCme on HHHB mmoCB .m mHHCmmm Hogoz meoCHommm Ill .UmSCHHCOOII.®H.: mdm<8 CHAPTER V DISCUSSION OF RESULTS The discussion of results below follows the order of presentation of the research hypotheses and results in Chapter IV. 2.1% The differences between the high schools' and the community college's press as perceived by the students in the several samples were significant in all eight com- parisons on the Intellectual Climate (second-order) dimen- sion beyond the .05 level of significance. The differences on this dimension were based primarily upon differences in the first-order factors of ASpiration Level and Student Dignity. That is, students perceived a higher level of demand or press for academic and vocational achievement and increased personal freedom in terms of adult reSponsi- bility. In the non—local high schools, differences on the first—order factor dimensions of Intellectual Climate, Academic Climate and Academic Achievement also contributed to the over-all differences. That is, students from non- local high schools perceived distinct change in the areas 110 111 of academic excellence in staff and facilities and general stress placed upon the academic aspects of school life. The students from all eight high schools encountered significant changes in the Non—Intellectual Climate of their new environment as well. The differences, however, involve both increased and decreased emphases so that the changes are not reflected in the over—all comparisons on this dimension. For the students from all of the high schools except the parochial, the Group Life factor (em- phasis upon mutually supportive activities) received greater emphasis at the community college. For the students from all of the high schools, the community college environment placed a lesser emphasis upon social participation or ' These two changes repre— 0pportunities for student ”play.' sent the major distinctions between their former and their new environment. For the non-local students the new en- vironment also represented an increased emphasis upon Academic Organization or demand for orderliness and sub- missiveness. Because of the questions or items of the High School Characteristics Index and the Evening College Characteris- Ilndex differ considerably in content, direct comparison on the above dimensions is not possible. Informal inSpection of the items of the correSponding scales indicated, how- ever, that the community college environment was seen as placing a greater emphasis upon work in an academic sense, adult levels of responsibility for the conduct of one's 112 own affairs (and for one's fellow students) and a decreased emphasis upon social life in the organized or formal sense. The design of the study did not permit an examination of the impact of these separate factors upon student perform- ance and rates of attrition. A more definitive examination would probably prove to be fruitful. The patterns of press at the eight high schools were not significantly different from that found at the community college. Two of the high schools were, on the contrary, distinctly or significantly similar to the community college in press pattern as expressed by Cattell's rp pattern analysis statistic. This finding was the Opposite of the predicted relationship. Further research will be necessary to determine whether the use of a pattern analysis statistic such as was used in this study represents a more significant mathe- matical statement of relationship than the rank-order co- efficient used in a secondary analysis. Differences between the high school and community college environments were more sharply delineated in the rank-order comparisons. Six of the eight schools rank—order ”patterns" were signi- ficantly different from the 'pattern” at the community college. Because the 11 first-order factor scales of the environment indexes include differing numbers of ten-item sub-scales, the scores were expressed in standard score form and the rank-order correlations re-computed. This new computation avoided the anchoring of the 11 factors on the 113 shortest factor (#4 — 20 items) and the longest factor (#8 — 60 items). The rank-order correlations were signifi- cantly negative in only three cases following the trans- formation of scores into standard score form. Part of the differences between the comparisons expressed in Cattell's rp form versus the rank-order correlations may be due to the requirement of the former that the scores be in stand- ardized form. No attempt was made, however, to determine the relative power of rank-order comparisons between indi— vidual‘s needs and environmental demands in the two settings as a measure of environmental continuity. Continuity The data relating environmental continuity to attri- tion revealed only one significant relationship: that of low achievement, Negative discontinuity to drop out. An inspection of the data and a correlation matrix involving all of the variables indicated that Adjusted pattern con- gruence was more highly related to retention than Pattern or Deviation pattern congruence. It is likely that it would have proved to be a more powerful measure in environ— mental continuity assessment as well. (The correlations with retention were: Deviation pattern congruence: .16, Pattern_congruence: .20, and Adjusted pattern congruence: .25 .) A more significant defect was a basic one in the design. Negative discontinuity was defined as a significant 114 negative change in "fit" between an individual and his environment. (”Significant": sufficient to place him in a lower congruence category at the community college than at the high school level.) Such a definition made possible the assignment of a subject to the Negative discontinuity category whose congruence or "fit" score at the community college was nonetheless in the upper half of the distribu— tion (or at least the upper 75%). For example, a student I could experience 'Negative discontinuity" according to the definition by changing from the highest quartile in high school to the second highest quartile (at the community college). To test the above "defect" in design, the Negative discontinuity, lowest quartile (congruence) students were pulled from the study (without regard to achievement levels though the majority were low achievement students). The resulting chi-square for leaving the college and not con— tinuing elsewhere (drOpout) for this sub—group was signi— ficant beyond the .05 level (chi-square - 15.16). Significant differences were found in levels of per- formance related toenvironmental continuity through the analysis of variance, but the differences between Positive and Negative discontinuity students were not significant. This was probably due in part to the defect in design mentioned above. Because the differences were significant for low achievement students, the data was examined further. The analysis was extended to include the entire year and 115 to describe the effects of continuity and discontinuity upon students at the two levels of achievement. The extension of the analysis to the entire year revealed that there was no apparent effect upon the per- formance of high achievement students. That is, high achievement students' means (in terms of year's grade point average) showed no systematic relationship to con— tinuity. The relationship between continuity and perform- ance did, however, sustain for the entire year for the low achievement students. High achievement students‘ mean performance for Positive discontinuity for the year was 2.663 for Negative discontinuity: 2.69. Positive discon- tinuity, low achievement students' mean was 2.18 for the year. Negative discontinuity, low achievement students' means was 1.82. (Table IV.1l in Appendix IV contains the results of the above analysis in detail.) In this sample, environmental continuity had a significant impact upon students below the median in achievement (in terms of high school grade point averages for entering students), but not upon students above the median. The non-residential community college may have more impact upon its students than the typical residential school due to the fact that there is less environmental change for the community college student. In contrast to his counterpart in a four-year (or any residential) college, the community college student's environmental change typically focuses upon his school environment. 116 His home environment usually remains relatively un- changed. If the above relationships between environmental continuity are typical for low achievement community college students even at the local college, these findings have important guidance implications. Congruence Although significant relationship was found between Deviation pattern congruence and the transfer intentions of the high achievement students, no significant relation- ships were found between this congruence measure and attrition (attrition defined as leaving the college and not continuing elsewhere). The relationships approached significance for the low achievement students and for low achievement students low in Educability. An inSpection of the data and the correlation matrix of all of the variables of the study indicated that Adjusted pattern congruence again would have been a more powerful predictor. Deviation pattern congruence related to retention at the .14 level whereas Adjusted pattern con— gruence was related to retention at the .24 level (the same level of relationship as that of high school grade point average to retention). High school grade point average (cumulative) is relatively independent of congruence between an individual's need or preference pattern and environmental demands. (The correlation in this sample 117 was .17.) High school grade point average and environ- mental ”congruence” might prove to be a more powerful predictor of retention than either used separately. Deviation pattern congruence, on the other hand, was more highly correlated with students' transfer intentions. For all students, Deviation pattern congruence related to a ”No” answer to the question, ”Do you intend to transfer to another institution prior to completing an Associate— in-Arts degree at the community college?” at the .14 level. (Adjusted pattern congruence correlated at .03.) Devia— tion pattern incongruence (the reciprocal of Deviation _pattern congruence) correlated with high achievement students' transfer intentions at .26° Apparently the Adjusted pattern method of measuring congruence which took into account only those deviations which were indicative of less intellectual need than the environment demanded or_more non-intellectual need than the environment provided is a more effective predictor of leaving academic life altogether. The Deviation pattern method, which took into account only those deviations in excess of one standard deviation above or below the mean (of the Common Beta Press), appeared to be a better predictor of intention to transfer or change to another educational setting (partiCularly for high achievement students). Replication with other samples will be necessary before final judgment can be made con— cerning these distinctions. 118 Male and female subjects were separated and a multiple correlation coefficient computed between Emotional ex— pression scores, Deviation pattern incongruence scores, In- tractability (defined simply as the reciprocal of Educabil— ity), and dropout. (The multiple correlation coefficients did not include high school grade point average.) The two coefficients were almost identical. The correlation of the three factors (Emotional expression, Deviation pat— tern incongruence, and Intractability) and dropout for males was .34 (.336). For females the correlation was .34 (.339). The breakdown of the correlations, however, revealed real differences. The simple correlations of the factors for males were as follows (with "dropout" in each case as the dependent variable): Emotional expressions scores: -.l5; Deviation pattern incongruence: .12; and Intractability: .33. For females the simple correlations of these factors with drOpout were: Emotional expression: .22; Deviation pattern incongruence: .16; and Intractability: .13. In this sample, lack of Educability (Intellectuality, Submissiveness, Orderliness and Conformity) was a more important factor with reference to dropout for males than for females. Emotional expression needs, however, were significantly different in their relationship to dropout by the two sexes. For males the relationship was negative (-.l5), for females, positive (.22). This may be a rela- tively ”local" phenomenon, however. Twenty-five male Activities Index test forms and twenty-five female test 119 forms were pulled at random from the original sample. The means of the scores on the 11 dimensions (Activities Index scores) were compared to the Common Beta Press scores at the community college. (Fifty Evening College Character— istics Index tests selected at random were converted to Activities Index equivalents by scoring the 30 sub—scales . . . . . 1 1n the manner required for the Act1V1t1es Index. ) The ”average” male profile means on the Educability dimension were almost identical to the level of the sample's perceptions of press on these factors. This finding was also true for the female subjects' "average" profile scores on the first four factors included in the Educability dimension. The female means on the Submissiveness and Orderliness factors were considerably higher than the en- vironment's demands. Thus a male's lower Educability score (if lower than average) would be less than the environment's demands. The female subject, on the other hand, with a slightly lower (than average) Educabili y score would be more congruent (in terms of needs) with the environment's demands. These "average" profile differences may have accounted in part for the difference between the correla- tion of male and female Educability scores with dropout (for males; .33; for females: .13). Emotional expression needs for the ”average” male were less than the environment's demands on four of the six factors which make up the dimension, and approximately equal Stern, Storing Instructions. . . ,o . cit. 120 to the environment's demands on the other two. Thus a male with higher than average Emotional expression scores fit the environment's demands to a higher degree. The female students ”average" Emotional expression needs exceeded the environment's demands on three of the first-order factors of the dimension (Closeness, Sensuousness, and Friendliness} Thus a female with higher than average Emotional expression needs fit the environment less well. These distinctions may have accounted in part for the difference in the rela— tionship of Emotional expression to dropout by males and females (for males: -.l5); for females: .22). The relationship of each of the congruence measures to retention was examined. Private Beta congruence (the individual's own need pattern related to his own perceptions of environmental press) correlated with retention at -.09. Deviation pattern congruence (deviations considered only if in excess of one standard deviation above or below the mean of the sampleis perceptions of environmental pr ss) correlated with retention at .16. Pattern congruence (all deviations used intjn pattern analysi statistic regardless of amount or direction) correlated with retention at .21. Adjusted pattern congruence (deviations below the mean of Intellectual press and above the mean of Non-Intellectual press used in the pattern analysis statistic)correlated with r,tention at .24. Murrayl believed that the influences of environmental demands could be "apperceived" unconsciously. That is, an individual might react to a constellation of stimuli with- out being aware of his reasons for reacting. A brief ex— ploratory questionnaire was developed to compare subjects' conscious expressions of environmental congruence with ”objective” measures of congruence. (The ”objective” measure based upon comparisons between press perceptions measured by environmental indexes.) The questionnaire used for this purpose was made up of ten questions which paraphrased the major emphases of the environmental indexes. (A copy of the questionnaire is included in Appendix II. The method of its use is described in Appendix IV.) The Expressed congruence scores of the students correlated with retention at .00. The Expressed congruence scores were also correlated with the other measures of congruence. The correlations were: Private Beta: -.17; Adjusted pattern congruence: .03, Pattern congruence and Deviation pattern congruence: .04. Although the differences between the highest correlation of the objective measures (Adjusted pattern congruence, .24) and that of the Expressed congruence scores with retention (.00) are small, they are significant beyond the .05 level. Such differaices lend tentative support to Murray's position. The relationship between congruence and the transfer intentions of high achievement students was significant. lMurray,.9_p_. cit. 122 Low Deviation pattern congruence, low achievement students, however, who dropped out of the college and did not continue their education elsewhere indicated intention to transfer more frequently than non-drOpouts. Sixty-six per cent of these students who dropped out of the college indicated intention to transfer, but only 30% of those who remained for the entire year indicated such intention. The "intended transfer” category was apparently meaningful for high achievement students but was also meaningful in a different way to low achievement students. The relationship of incongruence to transfer inten- tions was not supported when the category of "educability" was added (for high achievement students). It was a mean- ingful addition to the prediction of drOpout by low achieve- ment students. Educability was related negatively to drop- out for all students (correlation coefficient: —.24). 0f the several congruence measures, Adjusted pattern congruence was related to drOpout at the same level (—.24). These relationships were almost identical to the correlation of high school (cumulative) grade point averages to dropout (—.24). A multiple correlation coefficient combining the reciprocals of Deviation pattern congruence and Educability correlated with dropout at .30 (without the inclusion of high school grade point averages as such) for low achieve— ment students. Educability correlated with performance (expressed in terms of cumulative G.P.A. for fall + winter terms) at 123 .21. The analysis of performance levels as expressed by grade point averages was computed using the added sub- groups of high and low achievement students. High con- gruence, high Educability, hhyiachievement students' mean grade point average was 2.79. For Low congruence, low Educability, high achievement students the mean was 2.41. For Low congruence, high Educability, low achievement students the mean was 2.27. For Low congruence, low Edu- cability, low achievement students the mean was 1.96. The analysis of variance F ratio (5.08) was significant beyond the .05 level. (The complete results of the analysis of variance are presented in Table IV.13 in Appendix IV. The means of the eight groups are presented in Table IV.14 in Appendix IV.) The relationship between Expressed congruence scores and performance was also explored. The Expressed congruence scores correlated with cumulative (fall + winter term) grade point averages at .03. The relationship was explored further through computation of cumulative (fall + winter) mean grade point averages. The results were as follows: High achievement, high EXpressed congruence students: 2.59; High achievement, low Expressed congruence students: 2.78; Low achievement, high Expressed congruence students: 2.013 and Low achievement, low Expressed congruence students: 1.81. Here, as in the case of congruence and Adaptation level (joint analysis), the low achievement, low EXpressed 124 congruence students reacted in a different manner than their high achievement counterparts. (Table IV.l6 in Appendix IV.) The correlations of other congruence measures to performance were also examined. Private Beta press con- gruence correlated with cumulative (fall + winter) per- formance at -.15; Deviationppattern congruence: .16; Pattern congruence: .16; and Adjusted pattern congruence at .22. Because of the negative correlation of Private Beta press congruence with performance, this measure was ex- amined further. Private Beta congruence correlated with Expressed congruence at —.l7; with Pattern congruence at .05, with Adjusted pattern congruence at .31 and Deviation pattern congruence at .34. In light of the higher correlations of the "objective" measures of congruence (Pattern, Adjusted and Deviation) with performance than the correlations of "subjective" measures (Private Beta and EXpressed), Murray's assertion that individuals may react to (apperceive) patterns of- stimuli unconsciously would appear again to be tentatively supported. The relationships may be due in part to the "crudity" of the Expressed congruence questionnaire, but this criticism does not apply to the Private Beta measures. A more likely explanation in the latter case appeared to be (from an informal inSpection of the data) the operation of some form of reSponse set. This "set” seemed to be 125 Operating in a similar manner in both the Activities Index and Evening College Characteristics Index test protocols. That is, a negative or positive "set" seemed to be de- pressing or elevating both profiles for a large number of individuals. This ”set" may have accounted in part for the lesser effectiveness of the Private Beta measures as pre- dictors of withdrawal (dropout) or performance levels. Whatever the reason, the objective measures correlated more highly with performance than the subjective ones. The dif- ference between the most effective objective measure and the Private Beta measure (in terms of correlation with per- formance) was significant beyond the .05 level of confidence. The difference between the objective measures' correlation with performance and that of the Expressed congruence measure was significant at .10. The mean difference (.54) between the performance of the high Adjusted pattern congruence, Positive discontinuity students and the low Adjusted pattern congruence, Negative discontinuity students was not significant. The difference would have been significant with a slightly larger sample. Kramer states, "if the number of replications differs I greatly, there will be an increased probability of a signi- ficant difference within a subset of ranked means classified as homogeneous by this test."1 The sub-group inequality may have been a factor in this rather distinct difference lKramer, loc. cit., p. 309. 126 not being significant. Replication with a larger sample will be necessary before judgment can be made concerning the usefulness of these two categories (continuity and con- gruence) in a joint analysis. Adaptation Level The relationship of Adaptation level to drOpout was not significant.. The relationship for low achievement students was the opposite of that predicted. That is, students who had experienced low congruence in high school and who experienced it again at the community college dropped out mppe frequently rather than less. A concept such as ”frustration tolerance" may be more relevant in this case than Adaptation level. The students may have experi- enced all the ipcongruence of which they are capable with- out reacting in a definite fashion. The relationship of Adaptation level to performance by high and low achievement students was analyzed. The F ratio of 11.41 was significant beyond the .05 level. The groups of interest, however, are those students, both high and low achievement, who experienced low Adjusted pat— tern congruence. The high Adjusted pattern congruence' subjects also experienced (in many cases) Positive (environ- mental) discontinuity and all experienced (by definition) high congruence at the community college. The means of the low groups were as follows: Low Adjusted pattern congruence, high Adaptation level, high achievement: 2.49, Low 127 Adjusted pattern congruence, low Adaptation level, high achievement: 2.69; Low Adjusted pattern congruence, high Adaptation level, low achievement: 1.99; and Low Adjusted pattern congruence, low Adaptation level, low achievement: 1.77. (The results of the analysis of variance are pre- sented in Table IV.16 in Appendix IV.) It is not possible within the limitations of the design to separate the effects of congruence, continuity and Adaptation level with precision. The effect of Adap: tation level (to incongruence) appeared to be greater upon low achievement students. The high achievement students who experienced low congruence at the community college after having experienced high congruence in high school did better than those who experienced low congruence previously in high school. The relationship between achievement levels and Adaptation levels needs further exploration, however. High achievement, high Adjusted pattern congruence students who were high in Adaptation level performed very well (Mean 2 2.99) when they experienced high congruence at the community college. Exploratory Hypotheses: Continuity The correlational differences between the two schools most and least similar to the community college (of the four schools supplying significant numbers of students to the college) seemed to warrant investigation in other studies. The level of significance (.01) of the difference and the 128 fact that the variations in the correlations for this sample were systematic (following the variations in profile simi- larities between the high schools and the community college) lends support to this assertion. Although the predicted relationship was not found between Essential continuity subjects' high school grade point averages and community college cumulative (year's) grade point averages, this exploratory investigation would seem to warrant replication also. The difference (.43 versus .62) between the correlation for Positive and Nega- tive discontinuity subjects was not significant for a sample of this size. The difference appears large enough, however, to warrant investigation with a larger sample of subjects in the local setting as well as elsewhere. Summary The comparison of (perceived) press in the high schools to that found in the community college revealed that most of the students encounter increased emphasis upon student dignity, higher levels of aspiration, and mutually supportive activities in the community college setting. Non-local high school students also experience greater emphasis upon intellectual concerns, greater academic ex- cellence in staff and facilities, and increased stress placed upon academic achievement. All of the students also experience decreased emphasis upon social life in the com— munity college. 129 Patterns of press (as expressed mathematically using Cattell's rp pattern or profile analysis statistic) in the high schools do not differ significantly from that found at the community college. Only one significant relationship was found between environmental discontinuity and attrition: that between Negative discontinuity (for low achievement students) and dropout. When the research design was refined by pulling Negative discontinuity, lowest quartile congruence subjects from the study the resulting chi-square was significant beyond the .05 level (Chi-square = 15.6). Although an analysis of variance revealed significant differences between levels of performance related to con- tinuity, the differences between Positive and Negative dis- continuity students were not significant (using Duncan's New Multiple Range Test). An extension of the analysis of the data to the entire academic year revealed no apparent relationship between Negative environmental discontinuity and the performance of high achievement students. A signi- ficant relationship was discovered between Negative dis- continuity and the performance of low achievement students, however. A significant relationship was found between Deviation pattern congruence and the transfer intentions of high achievement students, but not between this congruence measure and attrition (defined as leaving the college and not continuing elsewhere). 130 In this sample, lack of Educability was more highly related to "drOpout" by males than by females. Emotional expression needs were significantly different in their relationship to "drOpout" by males compared to females. For males the relationship was negative (-.15), for females, positive (.22). The differences (in levels of relationship) between both Educability and Emotional expression needs and ”dropout" by the two sexes appeared to be related to differences between the ”average" male and female profiles. An inspection of the two ”average” profiles revealed that low Educability for a male student would indicate a lower pattern similarity to the environment's demands. The ”average" female profile revealed more Educability need than the environment demanded. 0n the other hand, a similar inspection with reference to Emotional expression needs revealed that lower than average need on this dimension for a female student meant that her needs were more in accord with the environment's demands whereas greater (than "average" male) needs on this dimension meant that the male student's need pattern was more similar (than the average) to environmental demands. "Objective" measures of congruence (Deviation pattern, Pattern and Adjusted pattern congruence) were found to be more highly related to retention than were the "subjective" measures (Private Beta and Expressed Preference congruence). For example, Adjusted pattern congruence correlated with retention at .24 while Private Beta congruence correlated 131 at —.09 and Expressed Preference congruence at .00. The possible effect of a "reSponse set" elevating both per— sonal need and environmental press profiles was seen as a possible explanation for the lack of relationship between Private Beta congruence and retention. Similar findings were discovered when the relation- ships of these congruence measures to performance were explored. Adjusted pattern congruence correlated with per— formance at .22; Private Beta congruence at -.15 and Ex- pressed congruence at .03. In view of the above findings, Murray's contention that individuals may react to a constellation of stimuli unconsciously appears to have been given tentative support by the data. Educability was found to be a meaningful addition to the analysis of the data. Used in conjunction with Adjusted pattern congruence and achievement levels, Educability appeared to add to the refinement of the analysis. Cumula- tive fall and winter grade point averages for several of the groups were as follows: high congruence, high Educa- bility, high achievement students' mean: 2.79; low —— congruence, low Educability, high achievement students: 2.41; high congruence, high Educability, low achievement students: 2.27; and low congruence, low Educability, low achievement students: 1.96. (The means of the eight groups are presented in Table IV.14 in Appendix IV.) 132 The relationship of Adaptation level to dropout was not significant. In fact, low achievement students who had experienced low congruence in high school dropped out of the community college more frequently when they experi- enced low congruence in the community college (compared to students who had experienced high congruence in high school). Adaptation level did not appear to be related to levels of performance by high achievement students under conditions of low congruence at the community college. It was apparently related to performance for high achievement students under conditions of high congruence at the com— munity college. That is, continued ”poor fit" did not appear to be related to the performance level of high achievement students, but an improved fit was positively related to performance level (mean for this group was 2.99). (A complete listing of group means is presented in Table IV.16 in Appendix IV.) 0n the other hand, Adaptation level was related to performance by low achievement students under both low and high congruence conditions (at the community college). Further research is needed to clarify the relation— ship of Adaptation level to both performance and attrition for high and low achievement students before the usefulness of this concept in environmental studies can be determined. The systematic variation in correlations between high school and college grades (from .49 to .88) following variations in pattern similarity (high school to college—- 133 .18 to .71) appear to warrant replication in other studies. Similarly, the high school to college grade correlations of the three (personal need--environmental demand) groups appear to warrant replication to determine, for example, why the Negative discontinuity correlation (.62) was higher than that of either the Positive discontinuity group (.43) or the Essential continuity group (.50) or whether this was, in fact, a "local" phenomenon. CHAPTER VI SUMMARY AND CONCLUSIONS In this chapter the study is summarized and the conclusions drawn from the study are discussed. Impli- cations for future research are also presented. The Problem There is little fundamental research done on the Operation of colleges and universities. Still less is done on the community or junior college. The community college, however, is becoming increasingly important as an educa- tional institution. More and more students are beginning their education in such a setting. Little is known about the impact of the community college upon its students. Even less is known about the effects Of the change in environment involved in the tran- sition from high school to the community college. The purpose of this investigation was to study the relationship of environmental change (in the transition from high school to a community college) upon the performance and rates of attrition of community—college students. The following research hypotheses were investigated in the study: 134 135 High schools will differ from the community college in the amount Of demand made upon students in both intellectual and non- intellectual areas. The demand made of the students in the high schools will differ in the amount Of emphasis placed upon such factors as academic achieve— ment, group life and vocational emphases com— pared tO the emphases upon such factors at the community college. Negative change in goodness of fit between environmental demands and student's preferences in the transition from high school to community college will be related to increased student attrition. Change in goodness of fit between the indi— vidual's preferences and the relative emphases in his new environment (at the community colle e) will be related to performance: positive change to better performance, nega— tive change to worse performance. Goodn ss of fit at the community college between an individual's preferences and the college's emphases will be related to attrition. That is, good fit will be related to retention, poor fit to attrition. 136 6. Goodness of fit at the community college between individual's preferences and the college 5 emphases will be related to performance. That is, good fit will be related to good performance; poor fit to poor performance. 7. An individual's previous experience with "poor fit" will modify the effects of ”poor fit” at the community college: a, Subjects who experience poor fit at the community college but who pre— viously experienced good fit at the high school level will leave the college and not continue their edu— cation elsewhere more frequently than poor fit students who previously ex- perienced poor fit in high school. 0‘ Students who experience poor fit at the community college but who experienced poor fit in high school will perform at a higher level than poor fit students who experienced good fit in high school. The Samples and Methodology The Lansing Community College sample consisted of 172 first-time, full—time freshmen students. The subjects were enrolled in required orientation classes. Testing was com- pleted late in the fall quarter. 137 The high school samples consisted of 36 senior sub- jects. These high school subjects from the eight high schools were tested within two months of high school grad- uation. The college subjects were given the Activities Index and the Evening College Characteristics Index. Personal data about family background and educational plans was collected by means of a questionnaire. A brief question- naire was also given which paralleled (and paraphrased) the ll major emphasasof the environmental indexes. A mail questionnaire follow—up was sent to the stu— dents late in the Spring term to determine whether or not there had been any changes in their full-time-student status or in their plans with reference to remaining at the com- munity college. Registrar's Office records were used to supplement the data from these questionnaires. The major analyses of performance used cumulative (fall + winter) grade point averages. The analyses em- ploying student attrition as the dependent variables used attrition statistics for the entire year. Several supple— mental analyses were made which extended the basic analyses to the entire year. Comparisons were made between the "fit" or congruence of students' preferred activities or needs as determined from the Activities Index to environmental demands in high school and at the community college (both the High School Characteristics Index and the Evening College 138 Characteristics Index were used). Subjects were assigned to four levels of environmental continuity. Assignment to these categories was made upon the basis of major or minor positive or negative change in fit (between the indi- vidual's needs and his environment's demands) in the transition from high school to the community college. Cat- tell's rp pattern analysis statistic was used to express the relationship between the individual's needs and his environment's demands at both the high school and the com— munity college levels. Several methods were used in the computation of "fit” or congruence scores at the community college. Pattern congruence was defined as the congruence between an indi- vidual's needs and his environment's demands computed (using Cattell's rp) using_all deviations of students' needs (converted to press equivalents) from the means of students' perceptions of press on the ll first-order factors of the environment indexes. Deviation pattern congruence was com— puted in the same manner except that deviations were included only if they exceeded the mean of students' per- ceptions of press on that dimension by more than one stand- ard deviation. Adjustedgpattern congruence was computed using only deviations below the means of the intellectual press dimension factors and above the means of the non— intellectual press dimension factors. Expressed congruence was computed using the differences between the individual's reported perceptions of press and his stated preferences 139 as deviations (subtracted from a constant positive number). Private Beta congruence was computed through the use of the individual's own profile of needs (derived from his Activ- ities Index test) and his perceptions of the environment's demands (derived from his Evening College Characteristics Index test). Private Beta congruence was computed in the same manner as Pattern congruence, but this measure used the individual's own perceptions of press rather than the means of the sample's perceptions of press. Activities Index scores were transformed into their environmental index equivalents by scoring the 30 sub-scales (which parallel the environmental index's subscales) in the manner prescribed by the instruments' author. The Findings The following findings are presented in the order of the divisions of the study. The significance level set for the rejection of the null hypotheses was .05. The results of the investigation of press are summa- rized in Table 6.1. The findings with reference to environ- mental continuity are summarized in Table 6.2. The results of the investigation of the relationship of congruence to dropout, transfer intentions and performance are summarized in Table 6.3. The results of the tests of the relationships of Adaptation level to drOpout and performance are presented in Table 6.u. The results of the exploratory investigation of environmental continuity in relationship to correlation of high school to college grades are presented in Table 6.5. luO TABLE 6.l.--Summary of the results of the investigation of press in the eight high schools and Lansing Community College. Dimension Model Used Finding Intellectual Climate: .t-test All 8 S. Aspiration Level " All 8 S. Intellectual Climate " A S; 4 NS. Student Dignity " All 8 S. Academic Climate " 5 S; 3 NS. Academic Achievement " 6 S; 2 NS. Self—Expression " l S; 7 NS Non-Intellectual Climate: " 2 S; 6 NS. (Self-Expression--see above) Group Life " l S, 7 NS. Academic Organization ” 5 S; 3 NS. Social Form ” 3 S, 5 NS. Play—Work " All 8 s. Vocational Climate " l S; 7 NS. Press Patterns rp pattern All 8 NS. analysis Rank Order of Press Factors Rank-order 6 S; 2 NS. correlation Legend: S Significant at .05 level or beyond NS Not significant lUl semOAancmnm poz u mz pcoxop no mo. pm pcmofimflcmflm u m mpommosm H mm Homewoq pmme mmCmm mannpasz spnscapcoomfle m 3oz m.cwoczm mocisomnom o>HpHmom mm pCoEo>oH£o< 3oq pmme mwcmm mamwpasz mpflscfipcoomflo mm paws mz 3oz m.cwocsm oocmEnomgmm o>HpHmom um>mflgo< 30g new cmfim mpHBCflpcoomHU mz onwsvmnflno psomosm o>wpmmmz mm pcmEo>mfl£o< Bog nommcwsp mpHBCfipcoomHo m onwswmnfico ow coapcmpCH m>fipmwmz mm pcmEo>mHLo< swam thSCHpcoomflo mm poms mz omwsvmuflno psoaoam m>flpmmoz uo>ofico< 309 new swam manmflnm> mcflccflm Hmcoz pcmccmmmm mpcmEpmoQB mpommnzm .mmoaaoo thCBEEoo wCflmcmq ow Hoonom zwflz Eonw coapflwcmnp opp ca mpflBCHpcoo prcmECOQH>cm mo COHpmepmm>CH one mo mpasmom ozp mo mamEEzmnu.m.w mqm<8 142 pcwondncwfim poz pcomop no Hm>oa mo. map pm pcwoflMflcmHm mocozsmcoo Choppmm popmsmc< mucosswcoo Choppwm coepmfl>mm mpommpsm H mm ”pcowmq pmoB mwcwm thBCflpcoo Co mao>oa mossy mamflpasz nuoocossmcoo mm pCoE mz 3oz mlcwocsa mocmenownom do mHo>oH nsom uo>ofl£o< 30g pom cwflm pmme mwcmm mamwpazz mm poms mz 3oz m_cmoczm mocmEnownom om< Bog uo>mflco< 304 new swam pmme mwcwm mamflpasz mocmsnwcoo mm pcoE m 3oz m.cwo:5m mocmEnomsom mo mao>ma nsom no>oflno< 309 now swflm zpflaflomospm CH 30H mz msmsvmnflco psomosm 0mm 30g uumm pco8m>mflco< 30g mCOHpcmpCH mpflaflnmoSUm CH 30H mz onwdvmufizo nmmmcmse 0mm 30A numm pcm8m>mflzo< swam mz onmsvmnflso psomosm 0mm Sod mm pcoao>ofl£o< Bog mCOHpcmpcfl mm paws m oumsvmuflzo nonmcmse 0mm sod um>mflno< zoq new swam mm paws mz mumsvmnflco psomonm 0mm 30a uo>oH£o< 30g new swam mannaam> mafincfim Hmooz pcmocommo mpcmapmmne mpomnnsm .mocosnmcoo Hmpcoacosfl>co mo coepwwflpmm>CH mzp wo mpasmon ozp do mamassm--.m.o mqmoH mo. on pm prOHMHCme mz m mpoompsm H mm ”UComoq spHschcoo HmseH> mz COHCMHoCCoo IHUCH mo mHo>mH nsom mm HH< szsCHpCoo AprCmE -COCH>cmV HmCOHp5p : s .m .m .H% m COHpmHmCCoo qumCH mo mHm>oH psom mHOOCom Cme anm mm HH< mCHpCHm Hmpoz PCmEpwmaB mpomwpsm .mCOprHmmmoo pCHom mcmnw mmeHoo Op HOOCom CwHC CC prCmEConH>Co mo COprmemm>CH %COHmCOHQXm me mo mpHsmmC on pcmoHCHcmHm poz moCoCCwCoo Cnmppwm cmpmzn©< moCmsanoo Cgmppmm COprH>om Ho>mH COHpmpmmo< m szCCHpCoo Co assessm--.m.© mqm H< CH CmHCunmm pCmE mz mo mHmzHMC< ooCmEComCmm om< moq nm>oHCo< Zoq ccm Cme H< CH onunmm pCmE mz mpmsvmnHCo psomonm 0mm 30g um>mHCo< 30H CCw CmHm mHanam> wCHUCHm Hmpoz HCmoCmme mpCmEpmmCB mpomnnsm .Hm>oH COprpmmn< mo COHpmmemm>CH me mo mpHsmmn me Co sgmsssm--.s.© mqmwws:EEomv mo. pm wcooawecwem e i x. a Pu. _ g usoccob-sor : ”hemmed H mm. zwo.a ma.: am.sm mmwm om em message Hmcoflpmoos t+ om.s =ma a no.2 om.mH ow.m em mm xsoz-smas ” mm.a :mo.a :H.: mo.sm am.: ow mm snob amaoom u mm.a- =H:.H sm.© mo.wm mm.: mm.mm soaommagewso unsweeu< x ©:.m- =H:.H sm.m mo.mm mm.m mo.am mean geese H mm. u :mm.H wfl.m :w.mm OH.: mw.am cosmmonoxm eflmm # mm.mn =NQ.H m:.m :w.mm m®.© wm.em ososm>mflso< ermomo< . ma.a- =mm.a Hw.m mfi.aa mm.m om.m msmsaao assesses .. mo.m- zom.a mm.m om.am mm @ om.ma eoflgmea semesnm x. cm.a- :mo.H mm.m ea m as r ma am momseao Hmjoomsawpew ex mo.w- :@Q.H NH.: mw.qm m m mm.om Hm>oh Hmsoepmnemw< I .o.m not: .c.m :mmz wadmmb w m i - II ..... scoomm mwewaon xwwnséeon a: moonuw Lwhr oefioum amood - Hm Hooco; swam .mnoesuono qwiawto, xwmdM ) ‘ in. 4.1:, .(msmflzo arm. lot L... ewes och pew gmsoeoin Hoogtm Qwflcl gems no sowwsmmsoula.m.>a mumat Ahwomtgw Hoogom cwezw flo. em osmoemetwflw ++ Asmpmmew Hoonom gmamv mo cm ocmuaeegmau . utmoeWecmflm no; wormsmms.a n Agopmogm mwoafiou hoessEEQOV mo. um ecmmamasmem *+ “Lopmohw owmaaoo xpwsseeool m0. is WCModmeCWam x mjoosoonom : Mocomoq mi .e m: :®H.H ma.q mm.sm om.m aw mm mwwseao HQCOeomoo> t.+ mo.m =om.fl no.3 on me we . mm.@m ssoz-gmam H mm. :mH.H :a.: m©.wm mo. fla.mw Egon Hweoom *« mm.m- =mm.a sm.© m©.@m as. e:.mm coapmmflcewso oasmemoa *x mm m- :sm.a em.m mo.mm mo. ms,ma mesa geese u H:.H :mo.a wa.m qm.mm :m. noemmonmxm meow LDMP-LOLOLD—ZI‘ (\l (\J O (\l xx am.m: :mH.H m:.© no.mm mm. om.mm psmsm>mfl£o< oflEmowo< n om.an :m©.H Hw.m ma.HH mo. m@.w momeeao oesoowo< \* mo m: :mm.m mm.m ww.am om. mc.qa zwflcwflm osm©3pm p am a. :wm.a mm.m sa.wm Ho m on em message Hmspomaaence v. gm.m- :oa a sa.q ms.:m s.m Nm.ma Heath achHnmsflmwa .I .Q.m omoz .m.w nwmz 2ngdw> w s 1. gopmmm mmmfiaoo hoessssoo m% doogom swam oewojm mooOA - mm anosom swam AsmECmosw cwmamoov mmmmd. omsmsm mwmfifioo msflzv>m one so monomm smme com Amsoanmm ao escwomsmgo aeo£-m ewes one so mono C) or-‘I ' J ('1 tom some so somegmmEoo:a. 221 « I so... .. . 0%. .~.«“.n mgoometw aoozum :mgmL go. om osmosencmaw +i rMomeLt Hoogom stzL we on Lemoeewowem + osmoeeflsmHm poo mocmgoeeaa n soumosw mwofiaoo thCSEEooL HQ. we womoae4swem ts noomonm mmoflfloo howQSEEooL mo. om w:moeq. .smem s msoocmonom : nose meg H mm. - =mo.a ma.n om.sm 33.: ma.sm message Hmcoflnmuo> ++ ~.w =mm.a H@.s mm.ma ms.m om.sm xsoz-ssam + Hm.m :mo.H :H.: mm.sm mm.: om.flm Egon Hmeoom n @©.fia :mm.a sm.© m©.@m Hw.m ma.mm coepmmflswwso oesoomo< * mm.m- =mm.a so.m mo.mm mm.m mm_am mega geese u mo. :HO.H wa.m q©.mm mm.m me.mm cosmmosmxm maom u 00.: :mo.a m:.© :w.mm ma.© mm.mm pcosm>oe£o< eroUmo< u mm.au =HO.H H@.m @H.HH mm. ®H.m omeHHo ermowo< is mm.@- :mo.H mm.m ©®.Hm 0:.m 00.:H hoeswem unmUSWm H mm a- =mo.a mm.m :H.wm oq.m mm.sm mnmsgao Hmsoemaawocn xx mq.m- :om.a H.: me.:m s:.m sm.am Hm>mq HQCOHpmsHmm< I .Q.@ swoz .Q.w coo: nosam> p m l. scoomm was How >+ -CSEEOn mt Hoocom swam +:Hms& Hmoou - m% Hoogom.£mem "Lnoegmmsw mwéwa o: Awac. moanwasmhomsmcm mmtaaoo sense m who so Waggon ewes etm mm on t: Hoogom snacL mesw moepmwgo «Em: doozow swam one no mm w come no :. oma mud 9nuu.:.>w 1hmLH ..:3EEomL 0 pm osmoaLLEmam .. Ammommnm mmeHoo xlegsEEocL mo Lm LawmaLHCwLm . msoccoonom : ”osmm a u on. - =wo.H NH.z mm.sm mw.m mw.mm mpmsLHo Hmcoepmoo> + :m.m =mo.a H©.: mm mH mm.s mm mm xtoz-smHm ++ m:.m :HH.H qfi.: m©.sm ms.m wm.mm seem Hmauom n Hm.H ::O.H sm.w m©.©m mw.© m©.mm COpranwwno oHEoowo< 4 so. - =mo.a sm.m mo.mm mm.s Ha.sm mead geese u om. - :Ho.H mH.m :©.mm ma.m mm.am scammmsgxm anew H mm. :QH.H m:.© :©.mm ©3.» mm.mm psoEm>mHgo< oHEmowo< x m:.mu :©0.H H®.m ©H.HH H©.m 0H.® opwEHHo oHEmowo< ** mw.mu :mw.H mm.m @w.Hm @m.© mm.wH pocmHm psoUSHm a pq.H- =mo.H mm.@ 3H.wm wm.m no mm mpmsflHo HmsnomHHmogH xx ms.m- =©O.H sH.s ms.:m m:.e me.c Hm>mL Hmcoapesamma l. .Q.m ommz .Q.m smmz mosHm> p E sonomh omoHHoo posjEEoo :% Hoozom 5 Hz II Hmwmtwmum.‘ L% HrosomiswLm H_CcECm iLL ommHHoo. wwwmw. . prHEm omtmco mmeHom onnm>r who so mZoow Come one LWLOHsrm Hoogom QwesL soocw moHHwHLmLcmtmco Hoozom :me one no monoom some Lo costmmEoo-- m >H mHm .* 33 m- =mH.H Hm.3 mm.mH mm.m HH Hm EE03-HQHE - mm. zmH.H 3H.3 m©.sm mm 3 HH.mw EEOC HmHoom .* ms.m- :3m.H smug mo.om m w mm.mm :oHHmNHcemEo 0Hsmem03 .. mm.m- :mH.H Em.m mo.mm HH 3 ms.om mLHH EzoEe - cm.H- :mH.H mH.m 33 mm mm m mm om conm mEExm EHmw .. 3o :wH.H m3.w 30 mm o3 mm.mm Hcmsm>wHEo< OHsmemUE . 3o :mo.H Hw.m mH.HH Hm.m mm.s mHmsHHo OHEQEQUE +. 03.3- :3H H om.m 3w Hm $0.3 mm.wH HHHEmHE Esmespm x Hs.m- :mH H mm.m 3H mm oo.s HH mm etmsH 0 He: .mHHsLEH t. mm.m- :mo.H EH.3 ms 3m mm.3 wo.om He>m3 HHEQHHmEHEwE I .D.m two: G.m Cmoz U(3Lm> H i I Cowoms oonHoo hoeCHEEoc Mk HOOCom Cwnm UHHC3L w 3.UUHUCO HEooH onoHHom wr- 2Coz : mi Ho. Hm» HUQHQQO H . . m HOOCo m CEHE mCH Co m9 HMcHHooL xoprL Chum. AWCHQJLLHWU .1 FLOHHU) HLMWH.£L .Q EJNHEQQEOUIo l .>e EHQLHC3EE0 0L mo. Hm CCmeCHCme L- m3omCmmoEOm : ”HBmeCC. n mm H- :mo H mHu3 mm Cm Hm m on 3m mLmeHHo HmcoHLmUo> ++ om.m :HH H Hob: mmumH NH": 33.nm CC03u>mHm u 3E. :mHLH 3H 3 mqum mm 3 mw mm ECom HwHo om L. Hw.m- Lao.H Cm.m mm.mm mmcm m3 om EoHCmNHEmeo UHEmHm 3 LL H3.3- =30H Cmym mo,mm mHo3 we mH mCHH EsoCL 4 mmLHu :omo+ mHom CQme mmoz quom COHmmmmem CHmm LL Em.MI =mHLH mCLC qmomm 3E.m @momm CCmEm>wHCo< UHEmEmo< .L mH.3 :w@ H kum 0H.HH mem 3®.© mCmEHHo UHEmcmo< LL mm m- LLQ H mmLm om.Hm mmom 3H 3H CLHEmHE Camuspm L. mmomu =mmLH mmow 3H mm #003 33 Hm mCmEHHo H33pomHHmCCH L, Qo.mn :HH H CH.3 mthm mCLm mm.©m HmbmH HmCm4CUCHm1< I. .QLm Cmmz .o,m Cmmz mm3HmC H m I; CUComC mmeHoo HCHC3EEQQ 33 Hoozom ECH: UHIfiOSQ. 1mw.QHUfiAQZ I moHCmHCmHmmCm_3 -WmHHoo HOOCom MCHCCPH I. ('llll. ‘1' ’11:. xmvCH meCUHCmL UCCCCQ HOOCom CmHm mCC Eme CO wwhOUm llll‘ll'll'll 'lllllll’l .HCQEC mCC meH 00. Co mmCOm, Cmm E UCC AmC, CmCE Cc COm C u Hoogmm CwHLL -.mCe o--.x EH H3m3b m Qofl+wmacmmpc qummmo: mwmeflao am>m4 4. -9. mymawao fiGQOHumUo> xpoz-»mam inou Hmanom m%flq msowfi scammmgmxm mama m>mfl£o< owEmbmu< mpmawfio UHEmUmo< hwwcwfihawthwuswm Hmjpmmaampcn L I. ammow+mLMCM< I’lltl'... Axwfimmpw Hoozum Lwfimv HQI Wm HLmQHLaLWam AngmmLQ Huozum Lmev mc_ Hm Pgmoamacwflu qumfiuflcmmw #3; muc¢ympgar «puwwwgw mnmflfior xwazzfisoo¢ do. +m Hinm1%azwau mgmpmwyx Qwufiflom >w~ZsEEGDV mo. pm Icmu»%agmam mjomgmmoEo: - mm. ::m H mfluq om km chm Hm mm ++ mm“: =mo H Hmuq mm ma mm“: mm mm 1 wm m- :om m 2H,: mm km $0.: mfi mm I @m.m- :wm.a kayo mm.mm mqnm movfi ‘* am :. =Mo H wmum mo mm Hwom om mfi I Rm.a- zmo H wdom sm.mm moom wq om xx @d.m- zmo H mq.m qumm :m.m :m.mm pcwe . :@.m- =ofi fl amcm mH.HH mm.m Ha.m I; No.09 zmz.fi mmom mm Hm mm.m M.mfi «I ma m- 2mm H mmho 3% mm ma x mt m x. mm m: :mm H NH 3 wk mm w Im Hm,um I .va xmwz Q m rmmz awjflmb I m ..... I .- I I I mmmflfiom hp~538830 u Hoogmm gnaw awarsm Hmo.ancyz I x% aoouum Lmam ” aweswagw mwkdqmov meHmeopmmmwfinu.wwmfiwcm wrammbm may r2 mr¥fiumm rmmfi w:m Amao ymvcw m wpmwgmwuwhmco moogom mem ac+ co prOmm cmme no ; [1‘ .ll llln. 1‘.VV4. IXQCT /O I/_ 3mm L mar-{GOLD qm—jmw-jmm k \0 my N-Cfi 'w OJ H n\] .3 :3.: K (N '\~ \. O\ H H 0\ra OI.H {T I C J (\J {-.1 ID OJ \C #4 CD CU (‘d r—*| rm I O Q r‘i #1 Cd I‘ j OJ \\3 F—‘I I5 wwIm I.ou:mw suzmv IQ. Im ICIUIIHIme II III.IIII IIOI_I IIIIV mo. II IIIIIIIxmaw I ICmuHMILwII Io; mungaIIHQ “IIIMIIN vmmHHmo H ssEanI HOI Im ICMUHIICmIm «I ~ I II. I ). ) . I I fill). .. IImImmIm pImHHoo HIM CJEI my me Im I;mIImrLuIm I Izompmmoaom : I I «Ia .J .ULQEQI Hm. =mI I II mm mm mII II-IN IIIEIID I.Q oIIIIo: KIQBIHmHm EI0@ HmHuo MHcmeo oHEmcmw: ILHH mmpnxm mHm mm. m IQIII onmw @w om :oII mmI m I:I gIIIIII< IIEIIII< mmeHHo UHEmwmu< .m cm,Hm I RIII IIIchQ IIIcrIm mmeH IIwEHHm Hw:Io@_ HmI:H . . Wm 0H HIIIW ngoHImIHum< I Imhm ;©wz Q m (mm: H p I m IiIII I IIIIiII. II IIIslfilI. Lepomm mmeme awarrfluoo w% Hooznv :mHL .III u.HLHmIIIIm_IHmEIh mm M)JLUI Qwem ”I; at III wIIHIuw HMIHM. mmIp II1IIQIU£< cmeHmo m:_-:1\m me so IIIOom :mme U_% AIIJIswu H0>QUI :wav IUCH moHIw IIIIQMLMCQ Hoogum xme qu :0 wmgom ngE I3 I;IHImII .©.>m mgmqh 227 Agopmme mwmam mmeHoo HIHgsseooI HOI Ia IszHIHCmHm II IsomCQonOI : Ibcmmmq I. I: I- III I m wI m III muam m,IOI IIIIIII IIIOI-con mI II I: I- Imm I mImI m III muqm I OII IIHIIII IIIOI-con I» ., am I- zm: I meI m III Inwm a_moI IIIIIII IIIOI-Iozv II II we I- zm: I mImI m.III meow m III IIIIIII IIIQI-cozv II I. mm_m- Iqm,I mImI m III mImm o mmI IIIIIIOIIIV II I, quq- =II I mva m.I:I meow m.ImI IIIIIII Imuoqv HI I» II I. Iqm.I m.mI m III mumm o NII . IIIIIII IIIOIV mI II IIII- :0: I m.wI m III mumm 0 III IIIIIII IIIOIV II 3.: .l. .ll.l.lul.l|l|l.l .Il 1m... m C. .m WEI 1 li!| I'llli ii I.... - [31!]... |.|I ; uisz> w @ II!!!III- .slla!. .QIm zwmz Hoozom smHI mmeHou HIICSEEOQ .lD-II '1 ‘ I! ll. l‘l'lI||4I-Il‘. '..-.' ‘l..|lll1-Il-"l' .I- . II 'I -III’|OI.I.'I| .‘ill Ill i 'ullulllilcll.'j.lll l- '1‘. .II:‘ II. . III: IIIII .AmmmHHoo HIHcsEEoo wgp csw mHoogom mcp mo Loam cmmspmgv mCOIHIwQEoU mmmga HISIQIHHIICH HmpoE-zuoaI>a mIm<& r b . 24 Ahmwwmpw Haosum gmfimv mo‘ pm pcmoflmwgmflm + pcmoflmwcwflm p0: mucmammmwo a Agmwwmgm mwmmaoo mwfiCSEEoov moo Hm 9cmowmflcwflm a msomcmwoaom : Mbcmwmg a am a: =mH H mama w mma mama m.mma AOfiHnsm amooa-cozv m% u wmo ; =m:.a qua w,mma n.0m m oma Aoflamsm Hmooaqcozv Kw a Na fig =mqofi $.3H m mmfl wowfl o Hmfi AUHHQSQ fifloOflegozv m% 4 mm H- =m: fl ovqfl w.me wuhfi momma AUfifipsm Hmuod-cozv m% + mmrm =Hoofl wvza momma Bond a Nma Aamflgoopmmv 3% a may =mm.a qua w mmfi mowfl 0 mod flowansm Hmooqv m% x km ms =wm a qufl memH Huom :.mqa Aoaassm Hmooqv wk 4 may zmo H chqfl m wmfi mflqfi m mad AUflflgsm Hmoogv aw l. Q . m. CQQWH! ‘ I|-.. l: mmsfim> p a :i- “new gmmz Hoogwm gmflm mommfloo xwaC:&Eoo .AmeHHoo %pacsaaoo $59 cam maoosom Qwfin wxwflm mnp go EDMm Cmmawmnv mQOWHammEOU mmmgm HmSQQmHHmpcancoz prOEsguHHV>H mqm<5 R) R) \Q The means of student performance (fall + winter and cumulative year s grade point average) at two levels of achievement and four levels of continuity are given in w H" :able I” l2 TABLE IV.l2o-—Mears of student perform nce at two levels of achievement and four levels of continu11y for fall, winter and cumulative (year's) grade po oi.t aver ages tall W‘rter Ta”1lative df‘f'ffl‘ 1. '~ -h g..- 4 'rf 1n! . 'V' EUi y Le 61 Mean Mean K SEQVS/ Low Achievement 2:25 2 l2 3’) 4 i O F itive 211 Essential TI (High) 2013 2 25 211 Essential r (Low) 1099 1065 1380 Negative l°86 1069 1182 High Achievement Positive 2:53 2159 2956 Essential 7: (High) 2.:u 2057 2169 Essential 7 \Low) 2164 2049 2053 Negative 2,69 2053 2.67 Table 7Vcl3 inclu des the data from the analysis of var1an3e of two levels of achievement; two levels of Egg; 9 it r-J }_._J cab} r) two levels of congruence ard cumulative ifall r l ,4 winter)gr ade point averageso TABLE TV l3 --Analysis of variance: achime\ ;e:t -two levels.)J educability {two levels‘) congruence (tw'o levels) and cumulative {fall + winter) grade point averages Source Sums of Squares df Mean Square F Ratio Sategory Means l5152 T 2 3? 5308‘ l1thin 761l8 162 béLB * Significant at 00003 230 Means of student performance at two levels of achieve- Q) ment, congruence and Educ b H lity for cumulativ gfall + winterl grade point averages are included in Table T7114 TABLE TVJlML--Means of student performance at two levels of achievement and two levels of congruenc and two levels of educability-- cumilative ;fall + winter) grade point averages; Achievement Con ruence Educatilitv Mean Low High High 212’ Low' H.gh LOW‘ l193 Low Low H‘gn 2 C? l w Low Low l./6 *v: ‘ v.2 .9 ‘7 high high High 21.g High High Low 21oo High Low High 2-bl High Low L~w 2‘4; Sne results of the analysis of variance§two levels of congruence, achievement and Adaptation levey are presented in Table TV.151 TABLE Iv,l5.--Analysis of variance; congruence "wo levelsz, achievement {two levels}) Adaptatior le el utwo levels) and cumulative {fall + winter; gr de pCirt averages Source sums of Squares df Mean Square 2 Ra‘io Category Means 30031. llt3b 4: LA) Du (7\ Jr-J (DU) 0 .1:— Within 62039 1 * Si nificant at 00005 R) (.0 FJ The means of student performance at two levels of achievement, congruence and Adaptation level are listed in Table 77-101 (Performance means based Upon cumulative :fall + winter) grade point averagesa) TABLE IV.lE --Means of student perform nce at two levels of achievement; two levels of en- vironmental congruence and two levels of Adar~ tation to incongruence--cumulative {fall $.- _ ‘__ winter) grade pOint averagesc Achievement Congruence Adaptation Level Mear Low High High 2.18 Low High Low 1L9? Low LOW‘ High 1199 Low Low Low 1177 High High High 2.99 High High Low 2159 High Low :gh 2_59 High Low Low 2‘56 Student Profile, ”Typical Dropout," and "Typical Intended—Transfer." TABLE IV.17.—-Range of Common Beta Press (Means on each of the dimensions + One Standard Deviation) Compared to the "Average n) anewtra teuotqeooA xqu-Ketd wuod {etoog uotieztusfiao egtq dnoug uotssaudxg '3198 iuawenatqov otwapeov BQEWTIO otmepeov BQBWIIO IBUQOSIIQQUI IeAeq uotqedtdsv £4 \2 d) r. f. ‘H Cl) :1) (I) L H C, Ch 4.) «U m L «5 LL 1 c p l (U C = 'U 2 (I) L 4—3 (1) C a) 3 'U L g: u 0 E a) O u {L m+) E m 0 u¢s E L L C M O t 3 H (L i) 11; r4 H as ho LU «I! 0 d! 0 0 L H H a; Q) Q L). ho > 7a in C ¢ 9 E‘ M : : : n; )3 / / / 0 4 U ..- C m m ‘\* G.) q \‘ ‘ \ D C) C) O O m :r m N ZIX :x 1XI OIWBDPOV 3IIIA :IIA 3IA 1A 3A1 :II 5I aoqoeg Joioeg Joqoeg Joqoeg 103023 Joqoeg Joioeg Joqoed Joqoeg JOQOEd 233 coamCmEHo conmmpaxm Hmcofiposm coamcmEHQ mafiafiomosom d d yd d .d 7d .d .d .d .d H E B E B E B B B B B B O O O O O O O O O O O 1 1 1 1 1 1 1 1 1 1 1 O O O O O O O O O O O J J J J J J J J J J J VA K V). I A I A A A I I I I no X I to I I I. A I I as a. I I I. I on I a. 00 rd .. S uv 3 J S 8 S 0 d W 3 X I. 8 TL n J .0 O I on d e u 3 J Q D. I 1 I u 0 J U S ..L . m 8 T. Tr U 1 I. a D. n O V T. J 8 A 1 8 S S .L O S S S T. p E 8 TL m S I. n 8 S S T. 1 J T. T: U S U 8 T... U I .L. 8 8 A 8 U 8 J A 8 U 0 S O 8 S 8 S 1 8 S 1 U 1 1 U S S S T. U S 8 S n 8 S O 8 J E S U S 8 .L 0.. S S 1 F OH 3. 2' 3H I I 1 .. ‘ \ mmmpw>q mamsmm o u mmmpm>< mama 4‘ ed mmmnm>¢euHoom nIIIIIIIIIIu uncommq man ‘14 .mEhou uwmu xmvcH mmfiufi>fiuo< Emu ocm vans pmuomammumHEopcmm mm mo memo: ecu ou omeQEoo Amucmam>asom meCH mmHuH>Huo< ou omupm>coov Snow pmmp xmccH moaumfipmuomsmco mmmHHootmoHcm>m omuomammnzasoocmm om ho memo: wo mHHuomeI.mH.>H mqm