A. TYPOLOGICAL APPROACH TO THE STUDY OF MEN'S RESIDENCE GROUPS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY GEORGE ROBERT STANDING 1968 Ins... This is to certify that the thesis entitled A TYPOLOGICAL APPROACH TO THE STUDY OF MEN'S RESIDENCE GROUPS presented by George Robert Standing has been accepted towards fulfillment of the requirements for Ph.D. . Guidance — degree in _— and Personnel Services /5/%»U M on] professor Due September 6% 1968 0-169 The quality and falls at Mic within the 315115 of 40‘ «is: on adj ‘- 1 an§\~ctual “.‘e types C Nation, k tie « larerm 3‘ . ‘ resldn ABSTRACT A TYPOLOGICAL APPROACH TO THE STUDY OF MEN'S RESIDENCE GROUPS by George Robert Standing The purpose of this study was to investigatethe quality and character of group life within mens' residence halls at Michigan State University, and more explicitly, within the residence hall house—-a subdivision within the halls of 40 to 70 students. The problem was trifold: First, an attempt was made to develop a multivariate description or typology of 27 house groups, three randomly selected in each of nine men's halls. Multiple discriminant analysis was used to extract linear combinations of house characteristics from which the types of houses were developed. Second, the im- pact on adjusted grade-point-average (gpa) and on the in- tellectual disposition of freshmen grouped according to the types of houses in which they lived was studied. In addition, by using a 2 x 4 analysis of covariance design, the interaction on the dependent variables between types of residences and four subculture orientations of freshmen The relation :‘ezic perfcr exarined. Refe dynamics of of the natu: grcup influe which the s: freshmen, 1: house groups axieties c: ience. Var; 10 create iI A n»: SiSting of a: definitiI George Robert Standing was also considered. Third, the character and extent of the "climate of learning" within the house were studied. The relationship of climate of learning to freshman aca- demic performance and intellectual disposition was also examined. Reference group theory, an understanding of the dynamics of small groups, and a theoretical consideration of the nature and origins of student subcultures and peer group influence provided the theoretical framework within which the study was developed. It was suggested that freshmen, in particular, would tend to identify with their house groups in order to cope with the ambiguities and anxieties created by the demands of the college exper- ience. Various characteristics of group life were thought to create inter-house variations in their environments. A new instrument, the House Analysis Survey, con— sisting of 128 items, including measures and an operation- al definition of house climate of learning, was develoPed to assess house characteristics. Four scales from the Omnibus Personality Inventory were used to measure intel- lectual disposition. Students also provided self-descrip- tions of their subculture orientation. Usable data on the House Analysis Survey was ob- tained from 884 (60%) of the residents of the 27 houses during the latter part of the Winter quarter. Responses were generally internally consistent in describing house characterist :1 resident of intellect 1:; freshmen .‘n ..‘| ".8 {ES-lit Six: nant an lvs“ atien in the resulting d; accounted fc The tie variance LN “vase IEC‘C“ t"q ‘QA on ace George Robert Standing characteristics. Sixty-one per cent (N = 669) of fresh- man residents completed both the pre- and post-measures of intellectual disposition. The mean GPA of non-respond- ing freshmen was lower than the respondents, tempering the results somewhat. Sixteen of the 26 possible roots of the discrimi- nant analysis were significant suggesting extensive vari- ation in the characteristics of house life. Five of the resulting discriminant functions were interpreted. These accounted for 66.7 per cent of inter-house variation across the variables. The first function, accounting for 28 per cent of the variance, differentiated among the houses primarily on the basis of residents' ratings of house academic per- formance. The second function seemed to differentiate between the houses primarily on the basis of residents' general ratings of their residence hall. The third was thought to differentiate among the groups on the basis of house reputation which seemed to be based more on social than on academic performance. The forth function seemed to separate the groups along a continuum of residents' ratings of compliance within the houses with residence hall and University regulations. The fifth function was interpreted as reflecting a general overall performance rating of the houses. No differences were found in the mean gpa's wasted by ferentiated which they 1 disposition teases. The: characteris: :‘e;endir.g c: :rientation. oriented 5:; :3 be more terized by . George Robert Standing adjusted by measures of academic ability, of freshmen dif- ferentiated according to any of the types of houses in which they lived. Nor were the measures of intellectual disposition apparently influenced by any of the types of houses. There were limited indications, however, that characteristics of houses may have influenced students depending on what they described to be their subculture orientation. Within certain types of houses "vocationally" and "collegiately" oriented freshmen tended to perform less adequately than "non-conforming" or "collegiately" oriented students and vice versa. The latter pair seemed to be more often positively influenced by houses charac- terized by academic or intellectual variables, while the former by social and non-intellectual variables. The climate of learning did vary between the houses. A positive climate was inversely related to the proportion of freshmen in the house and directly to house academic performance and satisfaction. Freshmen and older students tended to perceive their house climate similarly. No evidence was found indicating that the climate of learning did in fact influence freshman adjusted gpa or intellectual disposition. A TYPOLOGICAL APPROACH TO THE STUDY OF MEN'S RESIDENCE GROUPS by George Robert Standing A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Education 1968 Dar; eel and assi for which I support and :3" Guidanc tine and wig tended to D: in the Writi Cf CO‘Chain Dr. Eldon N: the Guidanm were perSOn can Dr. ."aluatim : :31: ACKNOWLEDGMENTS During the course of the doctoral program, coun- sel and assistance has been extended from many sources for which I am deeply grateful. May I acknowledge the support and guidance of Dr. Buford Stefflre, chairman of my Guidance Committee, who has given generously of his time and wisdom. A special note of appreciation is ex- tended to Dr. Irvin Lehmann, who, during a critical period in the writing of the thesis, assumed the responsibilities of Co—Chairman in Dr. Stefflre's absence from the campus. Dr. Eldon Nonnamaker and Dr. Sheldon Lowry, as members of the Guidance Committee both encouraged the project and were personally helpful in many ways throughout my pro- gram. Dr. Willard Warrington, Director of the Office of Evaluation Services, in many ways made the study possible through his very generous allocation of resources of the Office of Evaluation Services. His support and encourage- ment are deeply appreciated. A special measure of appreciation is extended to Dr. Donald Adams, Director of Residence Hall Programs, and to his many colleagues serving as staff in the residence ii halls at Mic collection c For :ent over t}. indicate my ako for our halls at Michigan State University who facilitated the collection of the data in so very many ways. For her love, patience, and sustained encourage- ment over the several years of the program, I wish to indicate my deep appreciation for my wife, Lawrin and also for our two daughters, Shauna and Lisa. iii Chapter T, R3713. ‘7 , [—4 O x TABLE OF CONTENTS Chapter I. II. III. NATURE OF THE PROBLEM . . . . . . Introduction . . . . . . . . . . Background of the Study . . . . . Residence Hall Life and the Climate of Learning . . . . . . . . . Statement of the Problem . . . . Operational Definitions . . . . . Theoretical Development . . . . . Introduction . . . . . . . . Reference Group Theory . . . Group Dynamics . . . . . . . Hodgkins' Theory . . . . . The Residence Hall House as a Reference Group . . . . . Cohesion As a Pr0perty of House Life The Ambiguous New Situation . Organization of the Study . . . . REVIEW OF THE LITERATURE . . . . Introduction . . . . . . . . . . Student Subcultures and Peer Groups . . . Contextual Analysis: Two Studies Residence Groups . . . . . . Research on Groups . . . . . . . Research on Residence Halls . . . Normative Systems . . . . . . Impacts of Special Programs . Residence Halls at Michigan State Summary and Conclusions . . . . . INSTRUMENTATION . . . . . . . . . College Qualification Test (CQT) Michigan State University Reading iv of Univer51ty Test . . Page 10 12 15 15 17 20 22 24 26 30 33 36 36 39 43 52 60 60 67 69 73 76 76 77 ‘- AF. Chili O V. HOUSE I "1m" V“ .5”... o 4 DESIGN Popula Pr Pc .ypoth RESULT Genera The M“ ¢ e—J :x (D I O '— p QC) The Summer Chapter IV. Attitude Inventory (AI) Student Subcultures . . . House Analysis Survey (HAS) . . . . . . . . . sumary I O O O O O C 0 DESIGN OF THE STUDY . . . Population and Sample . Pre—Test . . . . . . Post-Test . . . . . . Hypotheses and Methods of Multiple Discriminant summary 0 O O O O O O O O “SULTS O I O O O O I O C General Observations from Analysis Survey . . . The Multiple Discriminant House Differences . . Interpretation of the Function . . . . Interpretation of the Function . . . . Interpretation of the Function . . . . Interpretation of the Function . . . . Interpretation of the Function . . . . The Analysis . . . . . Analysis . . . . . the House Analysis of First Discriminant Second.Discriminant Third Discriminant. éoérih'nisériminéné Fifth Discriminant Impact of Types of Houses On the Academic Performance of Freshmen . . . . . . . .,. Hypothesis IIal . . . Hypothesis IIa2 . . Discussion . . The Impact of Types of Houses On the Intellectual DiSposition of Freshmen . . Hypothesis IIb . . . Hypothesis IIb2 . . . Discussion . . . . . The Climate of Learning . Discussion . . . . . Summary . . . . . . . . . Page 78 84 88 95 97 97 103 104 113 115 127 131 131 140 152 156 158 159 160 167 168 171 177 182 182 190 191 195 206 214 "taster V04 {77 Cl'V‘n' :- C '9. UV.--‘. Backg: Instr: Method Limits Conclu Ge Pe arerrocn: )3 w- . I‘MPLli ICE Chapter VI. SUMMARY AND CONCLUSIONS . . . . . . . . Background and Theory . . . . . . . . . Instrumentation and Sample . . . . . Methodology and Results . . . . . . . . Hypothesis I--Results . . . . . . . Hypotheses IIa and IIb--Results . . Hypotheses IIIa - III , The Climate of Learning--Resu ts . . . . . Limitations of the Study . . . . . . . Conclusions . . . . . . . . . . . . General Observations . . . . . Peer Group Norms and Influences How Do Houses Differ? . . . . . Why Do the Houses Differ? . . . . . What is the Impact of These Differences? The Climate of Learning . . . . . . Houses as Reference Groups . . . . Suggestions for Future Research . . . . BIBLIOGRAPHY . . . . . . . . . . . . . . . APPENDICES C O O O O O O O O O O O C . vi Page 221 221 '224 228 228 232 235 237 240 240 241 243 246 249 250 251 254 256 267 pt?— 5—.- 4.2 4.3 4.4 4.5 5.1 :U m H 0} N! 'U '1 O ' .‘T n 'I’?"(). LI ST OF TABLES Table Page 3.1 Reliability coefficients (internal consistency) for four Omnibus Personality Inventory scales . . . . . . 81 3.2 Product-moment correlation coefficients between four scales from the Omnibus Personality Inventory, MSU Reading Test, CQT total, and accumulative fall and winter quarter grade-point-average, 1964-65, for 637 male Michigan State University freshmen . . . . . . . . . . 82 4.1 Characteristics of nine Michigan State University residence halls comprising the research sample, Winter Quarter, 1965 . . . . . . . . . . . . . . . . . . 100 4.2 Description of total and freshman samples in 27 houses within nine Michigan State University residence halls, including number of residents, and number and per cent tested during post-test, Winter Quarter, 1965 . . . . . . . . . . . . . 105 4.3 Comparisons between the means of pre- tested freshmen who did and those who did not participate in the post-testing on the four AI scales, MSU Reading, CQT total, and accumulative fall and winter grade-point-average, 1964-65 . . . . . . 111 4.4 A comparison of pre-test subculture orientation of freshmen who did and did not participate in the post-testing . . 112 4.5 The operational definition of the climate of learning and related HAS items . . . 124 5.1 Mean scores, standard deviations and range of house means from 20 "aspects of life" within 27 men's residence hall houses at Michigan State University . . 132 vii 5.3 5.4 5.7 L11 o to Him (f (Y —- )4 ( f Lat m m (D (3 — F-I C. Table Page 5.2 Mean ranks assigned to 10 house activities or problems by residents. Residents were responding to the request to rank the 10 statements in the order (1) of the concern the respondent's house had shown for the item, and (2) of the respondents' personal preference of what should b§_the most important ac- tivities of the house. . . . .‘. . . . . . 138 5.3 Latent roots, (A), explained variance, chi- square values, degrees of freedom, and statistical significance levels for each of the 26 discriminant functions . . . . . 142 5.4 Fifty HAS variables included in the dis- criminant_analysis and the standardized weights (a) of each on the first five discriminant functions . . . . . . . . . . 146 5.5 Mean discriminant scores of the 27 houses on the first five discriminant functions . 154 5.6 The number of freshman residents grouped by subculture orientation in houses (1) with high mean scores and (2) with low mean scores on five discriminant functions . . 169 5.7 Analysis of covariance of two-quarter ac— cumulative grade-point-average of fresh- man residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function I . . . . 172 5.8 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of fresh- man residents classified according to subculture orientation and type of house on Discriminant Function I . . . . . . . . 172 5.9 Analysis of covariance of two-quarter ac- cumulative grade-point-average of fresh- man residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function II . 173 viii 5.13 5.14 :3‘ (n Anal :1‘ DJ 71 4 b j 1 Me c. 1) r1 (3 (i; I ‘I Ana‘ 71{\p (h K (D D) 1 a __-— ' LL‘ If ("3 n: .91" Q) (N 1 ’3 :f‘ f‘ MEa: AJ ( \ (\v "——"- Table Page 5.10 Mean twquuarter accumulative grade—point- averages, adjusted by MSU Reading and COT total scores, for groups of fresh- man residents classified according to subculture orientation and type of house on Discriminant Function II . . . . 173 5.11 Analysis of covariance of two-quarter ac- cumulative grade-point—average of fresh- man residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function III . . . . . . . . . . . . . . . . . . . 174 5.12 Mean two-quarter accumulative grade-point averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classified according to sub- culture orientation and type of house on Discriminant Function III . . . . . . . 174 5.13 Analysis of covariance of two-quarter ac- cumulative grade-point-average of fresh— man residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function IV . . . . . . . . . . . . . . . . . . . . 175 5.14 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classified according to sub- culture orientation and type of house on Discriminant Function IV . . . . . . . . . 175 5.15 Analysis of covariance of two-quarter ac- cumulative grade-point-average of fresh- man residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function V . . . . 176 5.16 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classified according to sub- culture orientation and type of house on Discriminant Function V . . . . . . . . . 176 ix 5.22 5.23 4.24 Anal ~ 0- Anal “- I‘d-c. Ana An Table Page 5.17 The number of freshman residents who com- pleted pre- and post-measures of intel- lectual disposition (four OPI scales on the Attitude Inventory). The residents are grouped by subculture orientation in houses (1) with high mean scores and (2) with low mean scores on five-dis- criminant functions . . . . . . . . . . 183 5.18 Analyses of covariance of freshman post— test scores on the four OPI scales, ad- justed by pre-test scores on the same scales, for high and low house groups on Discriminant Function I . . . . . . . 185 5.19 Analyses of covariance of freshman post- test scores on the four OPI scales, ad! justed by pre-test scores on the same scales, for high and low house groups on Discriminant Function II . . . . . . 186 5.20 Analyses of covariance of freshman post- test scores on the four OPI scales, ad- justed by pre-test scores on the same scales, for high and low house groups on Discriminant Function III . . . . . . 187 5.21 Analyses of covariance of freshman post- test scores on the four OPI scales, ad— justed by pre-test scores on the same scales, for high and low house groups on Discriminant Function IV . . . . . . 188 5.22 Analyses of covariance of freshman post- test scores on the four OPI scales, ad- justed by pre-test scores on the same scales, for high and low house groups on Discriminant Function V . . . . . . . 189 5.23 Mean Theoretical Orientation post-test scores, adjusted by pre-test scores, for groups of freshman residents classified according to subculture orientation and type of house on Discriminant Functions I and V . . . . . . . . . . . . . . . . 192 5.24 An analysis of variance of 27 houses on a measure of the climate of learning of the houses (HAS Part I, item 55) . . . . 196 5.28 5.29 ()1 w O 15.1 Inte The he A FEIa tea Table Page 5.25 Intercorrelations of house mean scores on a measure of the climate of learning and on four measures of house cohesion . 202 5.26 The number of freshman residents grouped by subculture orientation in (1) houses with a high academic climate of learning and (2) houses with a low climate of learning . . . . . . . . . . . . . . . . 205 5.27 The number of freshman residents who com- pleted pre- and post-measures of intel- lectual disposition (four OPI scales on the Attitude Inventory) grouped by sub- culture orientation in (1) high climate of learning houses and (2) low climate of learning houses . . . . . . . . . . . 205 5.28 Analysis of covariance of two-quarter ac- cumulative grade-point-average of fresh- man residents, adjusted by MSU Reading and CQT total scores, for high and low rated houses on the climate of learning . . . . . . . . . . . . . . . . 207 5.29 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of fresh- man residents classified according to subculture orientation and the climate of learning of their houses . . . . . . 207 5.30 Three analyses of variance of responses to the measure of house climate of learning between freshmen and older residents . . . . . . . . . . . . . . . 209 5.31 Analyses of covariance of freshman post— test scores on the four OPI scales, ad- justed by pre-test scores on the same scales, for high and low rated houses on the climate of learning . . . . . . . 213 A.l Means and standard deviations of responses to items from Part I of the House Analysis Survey for each of the 27 Houses . . . . . . . . . . . . . . . . . 267 xi Table A.2 Mean: 4 J R3 Mean; A.4 Rea. A.5 Valc.~ Table Page A.2 Means and standard deviations of responses to items from Part II of the House Analysis Survey for each of the 27 houses . . . . . . . . . . . . . . . . . 272 A.3 Means and standard deviations of ranks assigned by residents of each of the 27 houses to items 35 to 44 of Part I of the House Analysis Survey . . . . . . 291 A.4 Means and standard deviations of ranks assigned by residents of each of the 27 houses to items 45 to 54 of Part I of the House Analysis Survey . . . . . . 292 A.5 Values of Horst's r, a generalized measure of reliability? for House Analysis Survey Items . . . . . . . . . . . . . . 293 xii Figure LIST OF FIGURES Page 5.1 Location of mean discriminant scores of 5.2 The the 27 houses on the first five discriminant functions . . . . . . . . . 163 location of mean scores of 27 houses on a measure of the climate of learn- ing of residence hall houses (HAS Part 1, item 55) . . . . . . . . . . . . 198 xiii Appendix .’ T Pl. PALM-d m 10 Hr "'9": I'll. I... TO HOUSE "AI-3' ' ‘ his. \._"‘ h‘ A“? .“~ "\va— . l _ Anni._ O.\' LIST OF APPENDICES Appendix Page A. FALL QUARTER LETTER OF INFORMATION SENT TO THE SAMPLE . . . . . . . . . . . . . . . 267 B. WINTER QUARTER LETTER OF INFORMATION SENT TO THE SAMPLE . . . . . . . . . . . . . . . 270 C. HOUSE ANALYSIS SURVEY . . . . . . . . . . . . 273 D. TABLES OF HOUSE AND TOTAL MEAN SCORES AND STANDARD DEVIATIONS ON HOUSE ANALYSIS SURVEY ITEMS . . . . . . . . . . . 267 E. TABLE OF HORST'S MEASURES OF RELIABILITY ON HOUSE ANALYSIS SURVEY ITEMS . . . . . . 293 xiv cu Drlnginc 3.?"le ' «.suatlon 01 . J:cn s -d' and O: L 'c“s “use ~11 Stur. CHAPTER I NATURE OF THE PROBLEM Introduction The urgency with which our society has looked to higher education in the wake of world wide competition in technology and ideology, in addition to the sheer weight of numbers of students seeking advanced learning, has pro- voked a broad introspection into the adequacy and philos- Ophies of our colleges and universities. Even more dramatic in bringing about not only internal but also external ex- amination of institutions of higher learning have been the Open, and often hostile confrontations of recent years, between students, faculty, administrators and the community. It is not sufficient for colleges and universities to be satisfied with only a pedagogic approach to prepar- ing a student to view critically, humanely, and discern- ingly the external world. These institutions must, in addition, themselves serve as models through their willing- ness to be subjected to the closest possible examination in order to fully rise to the challenges of this age. One who has engaged most critically in examining higher education beneath the microscope has written 1 environ: atreor; gecifi' terrela: a 4 Hr! ['4’ act“ unu»- incact : m AAA -. ‘ ‘ "a n ,. :UIIEon-S a;__ ea. “5 Q -..ociy a, "1 - :rcec canZ If the development of the individual as a whole is the primary aim (of higher education), then col- leges should organize all their resources in efforts to achieve it. Such planning of a total educational environment must be guided by a theory of personality-- a theory in the terms of which it is possible to state specific goals for the individual, describe the in- terrelations of his various psychological processes, and understand the ways in which he changes under the impact of environmental influences. The ability within higher education to analyze students and their college environments has been enhanced not only as a result of increased interest by research- minded academicians, but also as a result of the develOp- ment of sophisticated instruments and methodologies, rel- evant theories and high-speed digital computers capable of handling complex statistical analyses. Ralph Tyler, summarizing an early conference on "college influences on personality," remarked on the advanced state of then cur- rent descriptions of students and their colleges and noted that changes in their behavior "are now characterized in anthropological, psychological, and sociological terms."2 A few years later Robert Pace summarized several ways by which college environments had been assessed,3 and most recently Newcomb and Feldman have authored a thorough 1Nevitt Sanford, Where Colleges Fail (San Francisco: Jossey-Bass Inc., 1967), P. xv. 2Lloyd Morrisett, Jr., "Research on College Influ- ences on Personality," Social Science Research Council Items, 13: No. 3, 1959. 3C. Robert Pace, "Methods of Describing College Cultures," Teachers College Record, 63, 276, 1962. review of t‘: leges Upon J is large n; Lat ten ye Of j in assessicl generally 3- L"=51 app-ea: I9 ,. S c review of the literature relevant to "The Impacts of Col- leges Upon Their Students, citing several hundred studies, the large majority of which had been completed within the last ten years.4 Of particular interest and apparent significance in assessing the impact upon students of higher education is what has been loosely referred to as the "climate of learning" or "environment for learning" (the terms would generally seem to be synonymous in most contexts in which they appear).5 The phrase seems to relfect a broad, often vaguely defined (if defined at all) set of variables sug- gesting the degree to which students' behavior, values, and/or attitudes are directed toward somewhat intangible intellectual concerns, as Opposed to more traditionally collegiate, vocational, social or anti-intellectual orien- tations. Stereotypes, traditions, the quality and nature of students admitted, faculty, physical facilities, the 4Theodore M. Newcomb and Kenneth A. Feldman, The Impacts of Colleges Upon Their Students, A Report to the Carnegie Foundation for the Advancement of Teaching, 1968. 5Ordway Tead, The Climate of Learnin (New York: Harper & Bros., 1956); Melvene D. Hardee, Personnel Ser- vices for Improving the Campus Climate of Learning," Jour- nal of the National Association of Women Deans and CounseIors, 24 (1961), 122-7; Lewis B. Mayhew, "The Intellectual Tone at Any University: Its Progress and Measurement," Journal of the National Association of WomenDeans and Counselors, 25 (1962), 156-60; John J. Prior, "Peer-Grofip Influence on the College Climate for Learning," Journal of College Student Personnel, 5 (1964), 163-7. community setting and the interaction of there variables all tend to define an atmosphere which, to the degree that it seems to be conducive to learning, represents a gauge against which institutions are subjectively judged. A helpful definition and example is offered by McCracken: A climate for learning exists on a campus when all members of the community perceive the reality of the college as a place to learn and where wisdom is valued. On such a campus, "adventure of the mind" would mean the kind of academic freedom that advocates, supports, and defends inquiry, criticism, exploration, and action. A college whose climate furthers learning would be sensitive to and responsive to individuality; it would be unimpressed by needless conformity. There would be a relationship of teacher and student marked by a sense of mutual responsibility for freedom of judgment and responsible action. Above all, such a climate would be perceived by student and teacher as one permitting the individual to reveal his feelings, to act out his ideas, and think as he wishes. Such a freedom to be unique or just to "be" without re- prisal--or even fear of reprisale-from peer or tgacher is a pre-requisite for "a climate for learning." Though many, if not most, educators would accept McCracken's description to be a highly desireable state, its attainment is seldom achieved and then, perhaps, only for certain mem- bers of the college community under certain circumstances. Several groups of students subjected to a typical yet unique set of environmental circumstances which were thought to define or relate to a climate of learning were the subjects of this study. The construct, "climate of 6C. W. McCracken, "Student Personnel Work and the Climate of Learning," Personnel and Guidance Journal, 42 learning" did in fact lend itself to statistical descrip- tion and, as a result, the impact of the climate on these groups of students was assessed. Background of the Study Contrary to many stereotypes, recent research in— dicates that the climate of learning of an institution is not exclusively a product of high standards of academic performance, nor of the faculty, nor of age,nor of tradi- tion. Rather, these variables interact with what have been shown to be other very pervasive determinants--the characteristics and backgrounds of the students admitted, the nature of the relationship they have with one another and with the general environment of the institution, and other characteristics of the student culture (for general sum- maries of relevant research see Sanford, Newcomb and Wil- son, Yamamoto and Newcomb and Feldman).7 Let it suffice to say at this point that these variables tend to modify or reinforce whatever impact the college experience might otherwise have upon students. 7Nevitt Sanford (ed.), The American College_(New York: John Wiley & Sons, Inc., 1962); Theodore M. Newcomb and Everett K. Wilson (eds.), College Peer Groups (Chicago: Aldine Publishing Co., 1966); Kaoru Yamamoto (ed.), The_ College Student and His Culture: An Analysis (Boston: Houghton Mifflin Co., 1968); Newcomb and Feldman, loc cit. Indeed, one might say that a student's integration of and exposure to these variables i§_the college experience. Newcomb and Feldman are more optimistic than Jacob in considering whether or not the college life has an impact in restructuring students' values and attitudes though they acknowledge that they were led "to pose questions that were at once more specific and more complex" than Jacob's somewhat simpler probe of "the impact of the col- lege experience."8 Newcomb and Feldman further indicate:- In spite of the limitations of data on net changes (in students as a result of the college experience), it seems altogether likely that some students in some colleges experience some changes that are attributable to the fact of being in college. And so our Inquiry shifts to precisely such questions--from the demon- strations of preponderant trends to the analysis of particular conditions under which particular kinds of impacts can be demonstrated. This shift does not imply an abandoning of our search for generality, but rather the espousal of a different kind of general question: under what conditions--regardless of where those conditions are found, and regardless of prepon- derant trends in contemporary American colleges in general--areparticular kinds of impacts likely to occur?5 Thus, it will be assumed for the purposes of this study that a fruitful climate of learning is attributable to and a characteristic of not only the institution as a whole but, within larger institutions at least, differen- tially to its component parts--individual departments and 81bid., pp. 3, 297-308; Philip E. Jacob, Changing Values in College: An Exploratory Study of the Impact of College Teaching TNew York: Harper Bros., 1957). 9 Newcomb and Feldman, Op. cit., p. 299. to the non-academic environs of a student's life. Residence Hall Life and the Climate of Learning Of interest in examining areas of the college ex- perience which may directly or indirectly influence change and learning is the student's living situation while at- tending--whether he commutes, resides off campus or in a residence hall.10 Of particular concern to this study is the quality of residence hall life and the degree to which the climate fostered by the extensive and intensive inter- personal relationships of students within various living units is supportive of the goals of the institution and contributasto the learning process and individual develOp- ment. In a broad sense, the study may have fiscal im- plications when one considers the staggering investment in college related residential facilities across the na- tion and the monetary needs in order to provide housing to accommodate swelling enrollments. About 1.5 million student spaces valued at $7.5 billion in 1966 (were they 10Marjorie M. Lozoff, "Personality Differences and Residential Choice," Growth and Constraint in College Stu- dents, Joseph Katz (ed.), U. S. Department of Health, Edu- cation, and Welfare Project No. 5-0799 (Stanford, Calif.: Institute for the Study of Human Problems, Stanford Univer- sity), pp. 294-372; Theodore M. Newcomb, "Student Peer- Group Influence," The American College, Nevitt Sanford (ed.), (New York: John Wiley and Sons, 1962), 469-488; Newcomb and Feldman, op. cit., pp. 197-226. institUtiC: teasing nee spaces at i for $11 t‘ has at th‘ inning in: growth and :at're mai‘ ing financ- rected as priate 90:: Ge: tiniof a 1 itose whic? beisolate' :knstrati< better posJ tn classrr If pro: housind realitz to be replaced at current prices) were available in all institutions of higher education in 1965. Projected housing needs for the decade were 1.5 to 1.8 million spaces at an estimated cost, including related facilities and rehabilitation and replacement of existing units, of from $11 to $16 billion.11 This study includes a modest look at the impact of certain architectural types of housing in an attempt to assess their relevance to the growth and development of the students. Research of this nature may shed some light in insuring that the overwhelm- ing financial investment in students' residences is di- rected as accurately as possible toward achieving appro— priate goals. Generally, if conditions conducive to the produc- tion of a beneficial climate of learning and, conversely, those which Operate in opposition to such a climate could be isolated, those charged with the development and ad- ministration of college housing programs might be in a better position to promote growth and learning beyond the classroom. Several years ago Strozier suggested: If proper recognition of the importance of student housing to higher education ever becomes a universal reality, it will mark not only the greatest change 11Israel Rafkind, The Federal Government's Col- lege Housing Loan Program (Washington, D.C.: American Council on EducatIOn, 1966), p. 2. l in stc Of hig: change thougn he ' sectrality indirect gration of colleges a Paytond Cc Justin Mo: .. the prg LELOVatiOn faculty of Th shiversity in the mat and innOva Stituuons. :an State \ in student personnel administration in the history of higher education, but also will represent a basic change in American educational philosophy as well. Though he may have overstated his case, nevertheless, the centrality of residential life to many campuses today is in direct harmony with his proposition. Note the inte- gration of "living" and "learning" at such experimental colleges as the University of California at Santa Cruz, Raymond College at the University of the Pacific, or the Justin Morrill College at Michigan State University. And on a broader scale note the increasing number of insti- tutions deeply concerned with promoting the best possible environmental circumstances in their residence halls and in the process making major committments to educational innovation such as the incorporation of classrooms and faculty offices into residential centers. The residence hall program at Michigan State University, perhaps as a result of its size--the largest in the nation--but more importantly because of its programs and innovations, has served as a model to many other in- stitutions. The residence hall program for men at Michi- gan State University provided the locus for this study. 12R. M. Strozier, The Housing of Students (Washing- ton, D. C.: American Council on Education, 1950), P. 1. learning, an istics of '. fornance a: residing i: (l iescriptio; their Char. to the res. 5iscrimina. for the en} 5:31:35 aCr: .3?ng a ty: (2 attitUdes 5°: StUdY tl-eSe were residEHCe 10 Statement of the Problem The purpose of this study was to investigate the quality and character of group life within mens' residence halls with particular regard to the perceived climate of learning,and to examine the impact of various character- istics of residence hall and group life on academic per- formance and certain attitudes and ideologies of freshmen residing in the halls. The problem was trifold. (1) An attempt was made to develop a multivariate description or typology of student groups according to their characteristics. The groups were defined according to the residence hall house in which they reside. Multiple discriminant analysis, a statistical technique which allows for the empirical study of "the configuration of social groups across multiple criteria..." was employed in devel- oping a typology.13 (2) Academic performance and change in certain attitudes of freshmen residing in the houses selected for study was investigated to determine whether or not these were related to residence in different "types" of residence hall houses. (3) Academic performance and change in certain 13Salomon Rettig, "Multiple Discriminant Analysis: An Illustration," American Sociological Review, 29 (1964), 398-4020 attitudes descripti: $818 inve S so classii in the cha For insta: tions in ; u. - 'V .-L.IlS m 2 tion. diff and admin reSalter-J; or not Cla hall in wh vale mx, Or E any 5‘89ch TL' OCDA 11 attitudes of freshmen classified according to their self- descriptions of their orientations toward college life were investigated to determine whether or not students so classified were differentially affected by variations in the characteristics of the houses in which they re- sided. Several related questions were also considered. For instance, it was assumed that because of the varia- tions in individual characteristics of residents, varia- tions in quality and patterns of interpersonal interac- tion, differences in house leadership and in physical and administrative features of the residence halls, the resultant group properties of house residents would vary. Of particular importance was the determination of whether or not classification of houses according to the residence hall in which they were found,and/or residence hall com- plex, or physical features of the hall would parallel to any degree the empirical classification of the houses based on group characteristics. 2 The multiple-discriminant analysis provided an opportunity to investigate relations between certain di- mensions of residence hall group life that were of par- ticular concern. FOr instance, consideration of theory discussed in the following section suggested that group cohesion or satisfaction with life in the house would ngt necessarily be related to the climate of learning as 12 perceived by house residents. On the other hand, it was hypothesized that the perceived climate could be related to academic performance within the respective houses. Specific research hypotheses will be stated in Chapter IV. Operational Definitions Several terms used frequently in the study require explanation of the context in which they occur as follows: Hgggg will be used in the study to refer to the administrative and/or physical subdivision within resi- dence halls at Michigan State University under the juris- diction of a resident assistant. It will also refer to the students housed in such an area when considered as a group. In several of the Michigan State halls, the term "precinct" is used to designate a house. Only the term house will be used here. Resident Assistant refers to a paid part-time student employee of the University, one of whom lives with the residents of each house. The resident assistant is an official representative of the University and is charged with promoting a positive academic, social and cultural environment in his house, with advising house government, with certain responsibilities related to the promotion of good order within the house and other various duties as may be assigned by his head resident. The resident assistant I selected ': dence Hall cradnate s ' ,I residence 0—3 :5 the Un; in? in st'.‘ field. Ai sible to h State Ulli‘y of these i l3 assistant is typically an upper division undergraduate selected by the head resident and/or the Office of Resi- dence Halls Programs, though he may be a sophomore or a graduate student. The principle student personnel officer in each residence hall is the Head Resident, a full time employee of the University and frequently a doctoral student major- ing in student personnel administration or a related field. All resident assistants in the hall are respon- sible to him. Residence Hall Complex or simply Complex refers to a geographic grouping of residence halls at Michigan State University. Several such complexes exist. Many of these have certain physical facilities in common such as recreation and dining. Some have consolidated levels of administration and/or associated academic programs. Architectural design in certain complexes is similar. Any or all of these features may be combined in a given complex. A more specific delination of the halls involved in the study and the complexes to which they belong is given in Chapter IV. Typology will refer to the empirical classifica- tion of the residence hall groups based on the statistical interpretation of and placement on combinations of group variables resulting from the multiple discriminant analy- sis. One of the main purposes of the Study was to 1‘ ‘ N l A 51511113111: parties ar an aningi' literatur: “taxonony' with a non ccio-culi tiHCtilre é rowed fro." 15 Signif : . LET-lentil": tiClllay: 10 ind'vidual C16filled bi. ad 3 14 distinguish between groups according to their group pro- perties and interpret the nature of these differences in a meaningful fashion. In Chapter II under the review of literature, the terms "typology," "classification," and "taxonomy" as these have been used in studies relating to the characteristics of social groups will be discussed and compared. Subculture is defined as "a group of individuals with a normative system, within the context of a larger socio-cultural system, which distinguishes it as a dis- tinctive segment of the total culture,“ a definition bor- rowed from Hodgkins whose treatise of student subcultures 14 The term will is significantly related to this study. frequently relate to students' self-report of that "par- ticular ideology toward higher education" to which they individually ascribe.15 Four such subcultures originally defined by Clark and Trow will be considered.16 Environment is a rather encompassing term which has taken on added significance through pioneering efforts l4Benjamin Joseph Hodgkins, "Student Subcultures-- An Analysis of Their Origins and Affects on Student At- titude and Value Change in Higher Education," (Unpublished Ph.D. dissertation, Michigan State University, 1964), p. 73. lsIbid., p. 88. 16Burton R. Clark and Martin Trow, ”The Organiza- tional Context," College Peer Groups, Theodore M. Newcomb andeverett K. Wilson (eds.), (Chicago: Aldine Publishing Co., 1966), pp. l7-70. 15 in the social and psychological study of higher education of Pace and Stern, Astin and Holland, and more recently 17 In general the environment will include any Astin. characteristic of the college community which may affect the student, particularly those external conditions and characteristics which may impinge upon the student in the context of the residence hall. Climate of Learning is considered to be a specif- jr: condition or set of conditions and influences within the college environment, the nature of which mediates students' predispositions toward academic, intellectual and cultural growth. An idealized definition of the term was presented on page four. An empirical definition of the term will be presented in Chapter IV. Theoretical Development Introduction The intent of this section is to bring to bear the contributions of theory to the develOpment and inter- pretation of the problems under consideration. Validat- ing evidence and empirical research suggesting refinements l7C. Robert Pace and George G. Stern, "An Approach to the Measurement of Psychological Characteristics of Col- lege Environments," Journal of Educational Psychology, 49 (1958), 269-277; Alexander W. Astin and John L. Holland, "The Environmental Assessment Technique: A Way to Measure College Environments," Journal of Educational Psychology, 52 (1961), 308-316; Alexander W. Astin, The College En- vironment (Washington, D. C.: American Council on Education, 196—877‘— 16 in the development of the problem are presented in Chap- ter II. The influence of groups on attitudes and behavior is not a new concept. Indeed, "That men act in a social frame of reference yielded by the groups of which they are a part is a notion undoubtedly ancient and probably sound."18 Yet as Hyman and Singer point out, awareness of group influence is hardly sufficient in and of itself to explain either deviations from expected behavior pat- terns within the membership of a given group, or to pre- dict within any degree of assurance the direction of be- havior.19 Thus, in singling out the residence hall house as the object of the study, several theoretical dimensions of group behavior must be probed. The basic foundation for the study will be laid within the theory of reference groups, a concept first suggested by Hyman20 and given prominence by Merton and 21 Rossi. The likelihood of, and the conditions under 18Robert K. Merton and Alice Kitt Rossi, "Contri- butions to the Theory of Reference Group Behavior," Read- ings in Reference Group Theory and Research, Herbert H. Hyman and Eleanor Singer (eds.), (New York: Free Press, 1968), P. 35. 19Herbert H. Hyman and Eleanor Singer (eds.), Readings in Reference GroupgTheory and Research (New York: Free Press, 1968), p. 3. 20Ibid., p. 5. Hyman and Singer discuss the his- tory and development of reference group theory. The original reference to which they refer is, Herbert H. Hyman, "The Psychology of Status," Archives of Psychology, No. 269, 1942. 21Merton and Rossi, Op. cit., pp. 28-68. 17 which a residence hall house might serve as a reference group having significant impact on residents' behavior will be explored. This will be done through analysis of certain characteristics of residents, particularly fresh- men and of reference groups themselves. This latter as- pect will be reviewed primarily through a brief syn0psis of the theoretical literature on small groups. Lastly, some of the characteristics of general student peer groups and student subcultures and their interaction with the residence hall environment will be considered. Reference Group Theory Hyman initially used the concept to describe how an individual develops his conception of his status in relationship to others.22 A person's perception of his status depends upon the group or groups of people with whom he compares himself, that is, his reference groups. As the utility of the concept has grown it has generally come to refer to any group to which an indiv- idual's behavior and attitudes are related. A significant addition to the theory was elabor- ated by Kelley who distinquished between "comparative" 22Herbert H. Hyman, "The Psychology of Status," Archives of Psychology, No. 269 (1942). Excerpts from. fhe study are cited in Hyman and Singer, op. cit., pp. 47-165. sources of .23 reference behavior, tics of it paints wh; Elewise: I A a In 1 gr pe re a tain s extent is co: F» U I I~ U m the group :4 - .5; his I; § .; 18 and "normative" reference groups, describing two different sources of influence which groups may have on the individ- ual.23 He states: "A group functions as a comparison reference group for an individual to the extent that the behavior, attitudes, circumstances, or other characteris- tics of its members represent standards or comparison points which he uses in making judgments and evaluations."24 Likewise: A group functions as a normative reference group for a person to the extent that its evaluations of him are based upon the degree of his conformity to cer- tain standards of behavior or attitude and to the extent that the delivery of rewards or punishments is conditional upon these evaluations.2 One group may, though not necessarily, serve both func- tions. Such a group is most often a membership group or a group in which membership is desired.26 For example, a student might aspire to membership in a certain frater- nity. He would modify his behavior both in compliance with his perceptions of the attitudes and characteristics of the members which are attractive to him and in antici- pation that his acceptance into the group is conditional upon his compliance to fraternity norms. 23Harold H. Kelley, "Two Functions of Reference Groups," Readings in Reference Group Theory, Herbert H. Hyman and Eleanor Singer (eds.), (New York: Free Press, 1968). Pp. 77-83. 24 Ibid., p. 81 251bid., pp. 81-82. 26Ibid., p. 81. 0: rent to t': 'L Wren a d A cf the re: Eran time lie format c.e’v’iates f CG‘l‘Ited f5 3th” than itSelf Ra's La Ellshing b 19 Other dimensions of reference group theory rele- vant to the problems confronted in this study are reviewed by Hyman and Singer. Though somewhat forgotten in much of the recent use of the theory, stress has been placed from time to time on reference "individuals" a concept analogous to that of "role-model." Hyman and Singer state: The parenthetical remark by Newcomb that a membership group may be a potent normative reference group "(particularly as symbolized by leaders...)" strongly suggests the role of the reference individual as the carrier of the reference group's norms, but it ap- pears to have been lost inside the parenthesis. It would be greatly to our advantage to reinstate the concept.2 Central to reference group theory is its contri- bution in eXplaining the role of non-membership groups in determining or influencing behavior, self-evaluation and 28 Thus, when member behavior the formation of attitudes. deviates from the expected direction it can often be ac- counted for in terms of allegiance to a reference group other than the membership group (the new membership group itself may also be a new membership group). Lastly, Hyman and Singer cite Newcomb as distin- guishing between "positive" and "negative" reference groups. 27Hyman and Singer, op. cit., p. 9. 28Merton and Rossi, op. cit., p. 35. In the la: norms and, between h; 8 (All whether or| inividual a referen: or protect lith a IQ; etil/or va; 3:31.13 1' r. 0; v. 51311 91‘0; X is Pres :epts dew MIPS as N t g‘) 20 In the latter case an individual may reject a group's norms and, as a result, attempt to maximize differences between himself and the group. Several conditions are cited which may determine whether or not a given group serves as a referent for an individual.29 For instance, an individual may choose as a reference group one which will enhance his self-regard or protect his ego. An individual is not likely to choose as a comparison group one whose standing is either so high or so low as to not be meaningful for the indiv- idual. Rather, an individual will tend to choose groups with a roughly comparable level of ability, attitudes, and/or values. GroupgDynamics Operating from a slightly different context, theory flowing from early research in the dynamics of small groups has been well summarized by Golembiewski. It is presented here in an attempt to draw together con- cepts describing the internal mechanisms of reference groups as they might Operate within residence groups. 1. A common motive(s) conducive to interaction among individuals is the basis of the forma- tion of small groups. a) N.P. Gist and L. A. Halbert make this point, 29Hyman and Singer, op. cit., pp. 13-16. 21 for example, in commenting on the widely prevalent adolescent gangs. "The gang is a form of adjustment," they note, "that boys, and even girls, make whenever their family or neighborhood do not satisfy their major wishes in a conventional way." 2. Group situations, especially when they are of sufficient duration to permit the development of a status hierarchy and role structure, may significantly affect behavior. a) In the pioneering experiments of Lewin, Lippitt, and White, for example, sharp changes in the behavior of children were induced when they were abruptly transferred from groups characterized by "authoritarian leadership" to those characterized by "dem- ocratic leadership," and vice versa. The investigators were apparently successful in creating different "social climates," or "atmospheres".... b) The individual as a group member, in short, becomes a member of a functioning system and is influenced by that system. The in- dividual is not necessarily aware of that influence.... 3. In time, a group structure of hierarchical status and clear-cut in-group demarcations develops. a) On the group level, for example, Toki dem- onstrated the disintegrative effects of re- moving a small-group leader during most phases of play. b) Also of interest here is the function of such status in influencing the individual's behavior in and beyond the group. Thrasher aptly indicated the importance of an in- dividual's status to himself in this way. "Any standing in the group is better than none, and there is always the possibility of improving one's status. Participation in gang activities means everything to the boy. It not only defines for him his pos- ition in the society he is greatly concerned with, but it becomes the basis of his con- ception of himself.30 30This concept is central to reference theory. Mer- ton and Rossi (0 . cit., p. 35) phrase the issue in a ques- tion: "...under w ich conditions are associates within one's own groups taken as a frame of reference for self-evaluation 22 4. In time, norms and role expectations which regulate intragroup activities are standard- ized. a) Such norms or role expectations need not b) C) d) be anti-social or criminal. But they are prescriptions of behavior meant to stabi- lize the group's internal environment and to some extent to control the conditions of its existence in a broader environment. Thus conflict with that broader environ- ment is quite possible. In fact, just such conflict with some broader environment--a neighborhood, a formal organization like an army, or "society"—-forced the develOpment of the small-group concept to explain theoretical- 1y behavior which was both clearly con- trary to that normally expected in the broader environment and which was also clearly organized rather than individual behavior.... ....The small-group approach...conceives of the group as a collection of individ- uals and a "plus." As Merei concluded from HIE study of children's groups: "Thus the group 'plus' is not some sub- stance hovering above the group--it is the hold their customs and habits have on members; it is tradition, the carrier of which is the individual, who, in turn, is strengthened by it.... Because the function of norm and role is the control of behavior, the small group is socially relevant.... Hodgkins' Theory In his excellent thesis on the origins and effects and attitude-formation, and under which conditions do out- groups or non-membership groups provide the significant frame of reference?" 31 Robert T. Golembiewski, The Small Group (Chic- ago: University of Chicago Press, 1962), pp. 19-22. 5f student Eodgkins r :cdel upor upon the e ...rnoc' Aneri to the no is left 1; strive f0 He then d 23 of student subcultures on attitude and value change, Hodgkins rejects reference group theory as an adequate model upon which to base his study.32 His concern‘was upon the emergence of broad general subcultures within the context of the university as a whole. He stated: ...modern American colleges and universities, like American society, no longer require such a specific commitment on the part of their undergraduate student body... Though he would allow that within the "sociocultural" system of the university a minimal level of compliance to the norms of the system must be met, much latitude is left to the student "as to whether he desires to strive for the other goals supported by the system."33 He then develops a theory where subcultures emerge as students strain for self-consistency when their educa- tional goals conflict with institutional goals.34 32Hodgkins, Op. cit., p. 70. 33Ibid., p. 72. 34"Self-consistency" is used in the sense of that construct central to Lecky's theory of personality. He states: "We conceive of the personality as an organ- ization of values which are felt to be consistent with one another. Behavior expresses the effort to maintain the integrity and unity of the organization." (Pres- cott Lecky, Self-Consistency, The Shoe String Press, Inc., 1961. p. 152.) [fir-3' ' '5' vxjno ‘. Ihus, student or look else.- 35 tion. " Ho 1' his theory. avenue of su: The Residenc. Not . lenentary t titudes and tion may be 95 signiric Hr . I :9“ for the A be Feet of hous iduals who S larly if the fOintly Cree bywhich eac 131129 the e r‘ 24 Thus, students could "conform, minimize the inconsistency, or look elsewhere for support for their own goal orienta- tion."35 Hodgkins provided empirical evidence to support his theory. However, he purports to investigate only one avenue of subculture development and peer influence. The Residence Hall House as a Reference Group Not inconsistent with Hodgkin's theory but com- plementary to it is the prOposition that students' at- titudes and conformity to academic goals of the institu- tion may be modified as a result of the highly personalized and significant interaction occurring in living situations, particularly for freshmen confronting the college environ— ment for the first time. A basic principle involved in describing the im- pact of house life is suggested by Newcomb "that indiv- iduals who spend a good deal of time together--particu- larly if they do so without a sense of constraint—- jointly create norms, concerning their common interests, 36 by which each of them is influenced." In order to max— imize the educative outcomes he proposes three applications 35Hodgkins, op. cit., p. 2, Abstract. 36Theodore M. Newcomb, "Student Peer-Group In- fluence," The American College, Nevitt Sanford (ed.), (New York: Wiley & Sons, 1962), p. 485. 25 of the principle: (1) promotion of a reference group of such a size that some selectivity of associations is allowed, (2) awareness of the fact that living arrange- ments provide the greatest single source of interaction for most students, and (3) the overlap of classroom ex- periences with living-group membership in order to in- crease the possibility of shared "intellectual excite- ment." The theoretical analysis in the classic study by Festinger, Schachter and Back using living situations of married college students supports Newcomb's propositions and provides additional indications as to when a house might function as a significant reference group.37 A cohesive group tends to develOp when the group is at- tractive for any of several reasons. But within an in- formal social group such as the house, its attractive- ness "will be mostly affected by the extent to which one has satisfactory relationships and friendships with other "38 To the extent that the house members of the group. and its residents can satisfy such interpersonal needs as status, acceptance, and goal fulfillment, identifi- cation with the house will be more pronounced. This 37Leon Festinger, Stanley Schachter and Kurt Back, Social Pressures in Informal Groups (Stanford, Calif.: Stanford—Ufiiversity Press, 1950). 38 Ibid., p. 164. 26 process is described as "means control" of the group.39 Festinger gp_gl also suggest that in addition to under- standing how much change a group can bring about, we should know over what activities of individuals the group has influence, which these authors describe as the "power field."40 Cohesion As a PrOperty of House Life Cohesion and the concept of reference groups are closely related. A reference group depicts the relation- §pip_between a group and an individual for whom the group has some degree of attraction, regardless of whether or not the individual is a member of the group. Cohesion, on the other hand, will be used in the study to describe a property of a group. A highly cohesive group would normally serve as a reference group for its individual members, but all reference groups are not necessarily cohesive. The term "cohesion" has been defined in many ways. Most authors agree, however, that it generally refers "to the degree to which the members of a group 41 desire to remain in the group." Cartwright offers a 39Ibid., p. 165. 4orbid., p. 166. lDorwin Cartwright, "The Nature of Group Cohe- siveness," Gropp Dynamics, Dorwin Cartwright and Alvin Zander eds. TNew York: Harper and Row, 1968), p. 91. 27 more elaborate definition to which he subscribes in his extensive description of its properties. He attributes the definition to Festinger: ...group cohesiveness is "the resultant of all forces acting on members to remain in the group." These forces are determined jointly by certain properties of the group and by certain characteristics of the members which, in conjunction, can be conceived as the immediate determinants of cohesiveness. One might be tempted to place a value judgment prematurely on the desirability of creating a highly co- hesive house or residence hall. Several significant studies suggest the tenuousness of such a judgment for the norms and other points of attraction of a cohesive group are not necessarily productive or consistent with the goals of some larger social system of which the group may be a part. "This power that groups have," notes Newcomb, "can be applied to educational advantage, to educational detriment, or to neither. Very often in my own university I have seen that the norms of student "43 groups are contra-educational. Both Stogdill and Et- zioni comment on the ambivalent characteristic of cohesion in their major theoretical treatises of group-dimension- ality. Stogdill views both cohesion and productivity as 421bid., p. 91. 43Theodore M. Newcomb, "The Contribution of the Inter-Personal Environment of Students Learning," NASPA, Proceedings of 49th Annual Conference. 5: 2 (October I667). p. I76. 28 outputs of groups rather than positing a causal relation- 44 . . . . Et210nl, who organizes his concep- ship between them. tion of complex organizations around the concept of "compliance“ with various levels of group induced forces, develops cohesion more or less outside his formal theory noting that cohesion does not determine the orientation of a group to the larger organization of which it may be a part.45 Consistent with the possibility of "contra- educational" functioning of group norms, Lozoff concludes that fraternity life for at least some students (those lacking in academic aptitudes and abilities) may have provided them with sufficient security and self-esteem 46 The to allow them to survive in the academic milieu. extent to which what might normally be considered anti- intellectual behavior might actually serve an adaptive function is suggested in her elaboration of a thought from Deutsch: 44R. M. Stogdill, Individual Behavior and Group Achievemepp (New York: Oxford University Press, 1959), pp. 13, 271—272. 45Amitai Etzioni, A Comparative Analysis of Com- plex Organizations (New York: Free Press of Glencoe, 1961), p. xviii. 6Marjorie M. Lozoff, "Personality Differences and Residential Choice," Growth and Constraint in College Students, ed. by Joesph Katz. Quote taken from chapter draft received from the author through personal corres- pondence. 29 Dr. Helene Deutsch speaks of group participation among younger adolescents as providing an Opportun- ity for peer-approved regressive behavior in the service of slowing growth so that disintegration can be avoided and progress eventually abetted. Thus, even some of the regressive aspects of fra- ternity living may have had functional value for students who needed relief from the strains of moving too rapidly toward independence, hetero- sexual mutality and confrontation of the differences in the values, ideas and behavior of pe0ple.47 Thus, highly cohesive student groups may not be too un- like cohesive industrial employee groups which have been found to be effective in maintaining group standards, but these standards may be either high or low regarding productivity.48 One could conclude that even though a house gives evidence of being highly attractive to its resi- dents, in and of itself this will not produce an excit- ing intellectual environment. Rather, it may well help to insulate residents from the rigors of the academic community. Hence it would seem that Chickering may be slightly "off target" in what would seem to be his over emphasis on residences becoming reference groups per se, 49 for their occupants. There is abundant evidence that 47Ibid., p. 28 (of draft copy). 488. E. Seashore, "Group Cohesiveness in the In- dustrial Group," University of Michigan, Survey Research Center, Pub. No. 14, 1954. 49Arthur W. Chickering, "College Residences and Student Development," Educational Record, 48 (1967), 179-186. 30 fraternities have served this function well for years. Yet several studies likewise indicate the fraternities' ability to insulate their members from the influence of academia.‘ Chickering seems to Operate from the premise that within residence halls, student groups seldom serve as reference units, which may in fact be true (though he offers no substantiating evidence). It would seem that his interpretation would be more precise if, in addition, his emphasis were placed on how to work with residence groups toward understanding and possibly modifying existing norms. He does put forth some im- aginative proposals toward using reference groups to serve educational ends but these seem secondary to his basic premise. Chickering's article seems to underscore the fact that as of yet, we really have little empirical evidence demonstrating successful experimental modifi- cation of group norms in the college context. The Ambiguous New Situation The complexity of the university environment as a social system has been pointed out earlier. The stu- dent new to the environment, though in many ways having been conditioned to know what to expect, must neverthe- less confront many new and perplexing situations, 31 particularly if he must live away from home.50 The ex- tent of his:"UQrootedness" will, of course, depend upon many variables. He may already have a highly developed set of interpersonal relationships in his new environ- ment through well established friendships, or he may have none. His mental and emotional equipment may be well adapted to cope with the ambiguities and anxieties of his new situation or they may be lacking. Several studies suggest that though students tend to have "some valid idea of the relative strength of various pressures in their new environment...(they) also have a general, stereotyped, and perhaps idealized image of college life which only imperfectly relates to what they are 51 about to find..." Eisenstadt theorizes that a complex, ambiguous situation may give rise to an individual an- choring himself within a reference group and/or to a set of what he describes as "reference norms." He states: ...there exists a multiplicity of reference norms and groups to which an individual may direct him- self and that his choice between them is very large- ly determined by the kind of social situation he is in. These different reference norms are evoked when the impact of the institutional structure on the individual puts him in a somewhat problematic situation from the point of view of his status and 50Ernest A. Andrews, "The Residential College Student: A Study in Identity Crisis" (Paper read at the Annual Meeting of the American Orth0psychiatric Associa- tion, Washington, D.C., March, 1967). 51Newcomb and Feldman, op. cit., pp. 74-83. 32 collectivity aspirations. Thus it may be suggested that the kinds of reference orientations and norms that will be evoked in a given situation...will de- pend on the interplay between the particular social situation in which an individual finds himself and his perception of this situation in terms of his status-image or levels of aspiration. By way of application, freshmen entering the college environment confront several conflicts both sub- tle and direct. As they strive for consistency and goal fulfillment they are inclined toward various groups and subcultures which meet these ends. Within their residence hall and in particular within their house, they tend to make an interpersonal investment in one another through their awareness of their shared predica- ment and of the interaction that will necessarily per- sist over several months. If returning students are housed in the hall, new students may be attracted to them for the "old hands" can introduce the new students to the subtleties of the system in terms of the "mini- mal level of compliance" to the broader system described 53 but also to other instruments of goal by Hodgkins, fulfillment in terms of needs for acceptance, social status and prestige. 528. N. Eisenstadt, "Studies in Reference Group Behavior," Readings in Reference Group Theory, ed. by Herbert H. Hyman and Eleanor Singer (New York: Free Press, 1968), p. 425. 53See pages 22-23 of this thesis. 33‘ If the relationships within the house for what- ever reasons are positive and goal fulfilling then the likelihood of the house becoming a significant reference group is substantially enhanced. The above should in no way be construed to sug- gest that the house will invariably function as a ref- erence group for any or all of its members. What is at— tempted here is to suggest that the potential is there. A priori the specific conditions that give rise to ref- erence group status of a house in the context of many competing groups and norms within gpd_beyond the insti— tution is speculative. Indeed, rephrasing the purpose of the study would suggest that it is an attempt to determine what conditions do result in a group attaining reference group status and what obtains from such sta- tus for the residents who ascribe to it such a role. Organization of the Study In the following chapter research relevant to the problems considered in this study will be presented. It will include an empirical develOpment of variables which are thought to be significant in the understand— ing of residence groups and to describe conditions under which a residence hall house might serve as a reference group. 35 The major instrument used in this study, the House Analysis Survey, and others used in the research design are presented in Chapter III. The sample, state- ments of the problem in appropriate research form and the statistical analysis are presented in Chapter IV. The results are presented in Chapter V. In Chapter VI the study is summarized, its limitations considered and the conclusions presented. CHAPTER II REVIEW OF THE LITERATURE Introduction During the past ten years our awareness and understanding of institutional differences between colleges and universities in terms of characteristics of both student bodies and the college environments has been greatly enhanced. Yonge,in his review of re- search on the college student, glowingly reports that research by Astin, Pace and Stern relating to the col- lege environment "have provided an inestimable contri- bution to the literature dealing with the student in higher education. Their pioneering studies are truly break-throughs; they have shifted the research emphasis J’ Excellent sum- from a descriptive to a dynamic model.’ maries of various works dealing with inter-institutional differences can be found in the chapters by Yonge, Michael and Boyer, and Boyer and Michael respectively in the "Higher Education" edition of the October, 1965 1George D. Yonge, "Students," Review of Edu- ggtional Research, 35 1965, 259. 36 37 Review of Educational Research.2 Other than to note the increasing emphasis on the inclusion of non-intel- lective predictors in the college selection process3 and the extensive variation in college environments,4 no general consideration will be given this important area of research. The reader might, in addition, be referred to four general reference works relating to the interaction of college environments, the psycho- social development of the college student, and peer group influence. Two of these,edited by Yamamoto5 and by Newcomb and Wilson,6 are collections of both widely quoted theoretical and empirical writings which have generally been printed elsewhere. The American College edited by Nevitt Sanford is a highly important work as a result of Sanford's intensive effort to draw together into a unified whole the divergent strains of research and theory related to higher education. Last, Newcomb 2Ibid., pp. 253-263; William B. Michael and Er- nest L. Boyer, "Campus Environments," Chapter II, pp. 264-276; Ernest L. Boyer and William B. Michael, "Out- comes of College," Chapter III, pp. 277-291. 3Yonge, Op. cit., pp. 254, 261. 4Michael and Boyer, op. cit., pp. 264-276. 5Kaoru Yamamota (ed.), The College Student and His Culture: An Analysis (Boston: Houghton Mifflin Co., 1968). 6Theodore M. Newcomb and Everett K. Wilson (eds. College Peer Groups (Chicago: Aldine Publishing Co., 1966). 38 and Feldman have put together an admirable review of research in their report , The Impacts of Colleges Upon Their Students.7 Much of the research reported has been accomplished since The American College was published. Several avenues of research have relevance for the study at hand, though not all are of equal signif- icance. In the pages that follow, the concept of stu- dent subcultures will be further develOped with a con- sideration of relevant research. Next, two highly rel- evant studies by Nasatir and by Selvin and Hagstrom will be considered in depth. General research on living groups will be considered in two parts: the first will treat reports that provide insights into relevant di- mensions of group life having an impact on behavior and attitudes, and hence are to be considered in assessing group differences. The second part considers several studies conducted on college housing groups which, in general, attempt to assess group differences. Housing research at Michigan State University will then be con— sidered and will be followed by a general summary of the chapter. 7Theodore M. Newcomb and Kenneth A. Feldman, The Impacts of Colleges Upon Their Students, A Report to the-Carnegie Foundation for the AdvancEment of Teaching, 1968. 39 Student Subcultures and Peer Groups Clark and Trow have develOped a fruitful method of classifying the orientations of students toward higher education and of studying student subcultures.8 They postulate four hypothetical subcultures which are characterized according to students' positive or nega- tive tendency to identify on the one hand with ideas and on the other with their college. The resulting orientations are portrayed below.9 Involved with ideas Much Little Identify with Much Academic Collegiate their college . Non- thtle conformist Vocational The characteristics of each are described as follows: Collegiate Culture The most widely held stereotype of college life pictures the "collegiate culture," a world of foot- ball, fraternities and sororities, dates, cars, 8Burton R. Clark and Martin Trow, "The Organiza- tional Context," In Theodore M. Newcomb and Everett K. Wilson, College Peer Groups, Chicago: Aldine Publish- ing Co., 1966, Pp. 17-70. 9 Ibid., p. 24. 40 drinking, and campus fun....In content, this system of values and activities is not hostile to the col- lege, to which in fact it generates strong loyalties and attachments. It is, however, indifferent and resistant to serious demands emanating from the fac- ulty for an involvement with ideas and issues over and above that required to gain the diploma. Vocational Culture To these students, many of them married, most of them working...college is largely off-the-job training, an organization of courses and credits leading to a di- ploma and a better job than they could otherwise com- mand....These students have little attachment to the college....these students are also resistant to in- tellectual demands on them beyond what is required to pass the courses...ideas and scholarship are as much a luxury and distraction as are sports and fra- ternities. Academic Culture The essence of this system of values is its identifi- cation with the intellectual concerns of the serious faculty members. The students involved work hard, get the best grades, talk about their cousework out- side of class, and let the world of ideas and know- ledge reach them....For these students, the attach- ment to the college...is to the institution which supports intellectual values and Opportunities for learning; the emotional tie to the college is through the faculty and through campus friends of similar mind and temper. Nonconformist Culture These students are often deeply involved with ideas, both the ideas they encounter in their classrooms and those that are current in the wider society of adult art, literature, and politics. To a much greater degree than their academically oriented class- mates, these students use off-campus groups and cur- rents of thought as points of reference, instead of the official college culture....The distinctive qual- ity of this student style is a rather aggressive nonconformism, a critical detachment from the col- lege they attend and from its faculty...and a gener- alized hostility to the college administration.... 41 these students pursue an identity, not as a by-product, but as the primary and often self-conscious aim of their education.1 In several studies these four subcultures have been described in brief paragraphs and students asked to characterize themselves according to the one best like himself. In a validation study of a new instrument, the College Student Questionnaires, 13,000 entering freshmen representing 23 different institutions were asked to in- dicate which of the four best described their attitudes for going to college.11 Fifty-one per cent selected the "collegiate" or- ientation while only 18.5% subscribed to the "academic." The "vocational" was picked by 26.5% and the non-conform- ist by only 4%. Institutions varied widely in the pro- portion of students holding each of the four. 12 Gottlieb and Hodgkins and Hodgkins13 have l°Ibid., pp. 20-24. 11Henry Chauncey, "Excerpt of Remarks," Annual Meeting of the College Entrance Examination Board, October 28, 1964, New York City, Unpublished report, Educational Testing Service. 12David Gottlieb and Benjamin Hodgkins, "College Student Subcultures: Their Structure and Characteristics in Relation to Student Attitude Change," The School Re- view, 71: 1963, 266-289. 13Benjamin J. Hodgkins, Student Subcultures--An Anal sis of their Ori ins and Effects on Student Attitude gpd VaIue Change in Higher Education, TUnpuinshedDOOtoral thesis, Michigan State University, 1964). subjected t‘. validity an change. Re sncultures standing tn It studies usi cited above- :enbers of they are 5: herance to xeh'; that limite Several fur. ces WhiCh I: (1) As is 1e to. fr (2) T1: to t‘: I I I __ -— 42 subjected the constructs to several tests of empirical validity and have noted their relationship to attitude change. Results generally supported the validity of the subcultures and pointed out their relevance in under- standing the influence of higher education on the student. It is important to note that the majority of studies using the Clark-Trow typology, including those cited above, do not classify students as interacting members of a group sharing a similar orientation. Rather, they are simply typed according to selection of or ad- herance to similar orientations.l4 Newcomb and Feldman, in a broader context than that limited to the Clark-Trow subcultures, have listed several functions extrapolated from many different sour- ces which peer groups are thought to serve. (1) As part of the intermediate stage between the family and larger post-college world, the col- lege peer group may help the individual student through the crisis of achieving independence from home. (2) The peer group offers general emotional support to the students; it fulfills needs not met by the curriculum, the classroom, or the faculty. (3) The college peer group can provide for the stu- dent an occasion for and practice in getting along with peOple whose background, interests and orientations are different from his own. (4) Through value reinforcement, the peer group can provide support for not changing....Yet, it can l4Newcomb and Feldman, Op. cit., p. 234. 43 also challenge old values, provide intellectual stimulation and act as a sounding board for new points of view, present new information and new experiences to the student, help to clarify new self-definitions, suggest new career possibil- ities, and provide emotional support for students who are changing. (5) The peer group can offer an alternative source of gratification and of positive self-image, along with rewarding a variety of nonacademic interests, for students who are disappointed or not completely successful academically....Friends and social ties may also serve to discourage voluntary withdrawal from college for other than academic reasons. (6) College peer-group relations can be significant to students in their post-college careers--not only because they provide general social train- ing but also because of the develOpment of per- sonal ties that may reappear later in the career of the former student.1 Contextual Analysis: Two Studies of Residence Groupp The genesis of this thesis lay in part in two research reports which considered the interaction of be- havior and the social context in which the subjects lived. Using data from a larger study, Nasatir noted that the failure rate of a group of male students who had begun school at the same time varied between the six dormitor- 16 ies in which they lived from zero to 56%. The residents 15Ibid., pp. 236-237. For readers desiring to probe these dimensions more deeply, note that several references are included in the original report support- ing each of the above items. 16David Nasatir, "A Contextual Analysis of Aca- demic Failure," The School Review, 71 (1963), 290-298. 44 of these ostensibly homogenious living units were select- ed to avoid what he had described as "pronounced dif- ferences in the social backgrounds of students living in 17 Though he cites different types of residence groups." no evidence he indicates that the backgrounds of hall residents were "roughly comparable." Suggesting that each dormitory had achieved somewhat of a distinctive character, be classified the nature of each hall. This was accomplished simply according to "the proportion of members expressing agreement with a statement that the most important reason for attending college is to obtain a basic general education and appreciation of ideas."18 Though his report did not include the number per hall responding positively to the statement, Nasatir further dichotomized the six halls according to "those above and those below the mean of the distribution pr0portions of group members choosing the academic response as the pur- pose of a college education." Nasatir reported that the failure rate was greater for students whose personal academic orientation was out of harmony with the academic context of the dormi- tory in which they lived (personal academic orientation was also determined by the individual's single response 17Ibid., p. 292. 18Ibid., p. 293. 45 to the same questionnaire item). Specifically, failure rate for "non-academic" students living in "academic" halls was somewhat higher (26%) than for "non-academic" students in "non-academic" halls (14%). Likewise, "aca- demic" students living in "non-academic" halls had a higher dr0pout rate (12%) than their counterparts in "academic" halls (7%). Thus, though academically orien- ted students generally had a lower drop-out rate than non-academically oriented students, both did poorest when housed "out of context." Nasatir also investigated the failure rate of students who were considered as not having been "inte- grated" into the dormitory. These students were defined as those spending little of their time in company with other members of their residence groups--another single item response from a questionnaire.19 The failure rate of the non-integrated was twice that of the integrated regardless of academic orientation. The "non-integrated," "non-academic" student residing in an "academic" hall experienced the greatest failure rate (29%) while no "integrated" "academic" student in the "academic context" failed. This lead Nasatir to conclude for the high fail- ing group: "These men are not only out of joint with 191bid., p. 295. 46 their larger surroundings, but also are denied many of the supports that group membership can provide."20 Several methodological problems make Nasatir's data highly suspect. Chi squares were not presented and his classification system, particulary that of the halls where we were not quoted the actual proportions used, makes it difficult to have confidence in his data. Nevertheless, his findings potentially are highly signif- icant if they can be replicated, in demonstrating an im- portant interaction between the individual and his resi- dence. Later in his dissertation he reported "a pronoun- ced effect upon the failure rate of students already pre- disposed to failure" in the context of residence halls.21 Students predisposed to fail living in highly cohesive groups had a lower attrition rate than students with a similar disposition in less cohesive halls. Little in- teraction was noted between students pep predisposed to failure and their living situation. He does not report a replication of the earlier study. Nevertheless, his results do add to the credibility of the former report. The second study, by Selvin and Hagstrom, can 2orbid., p. 296. 21David Nasatir, "Social Sources of Academic Failure: A Contextual Analysis," (unpublished disser- tation, University of California, Berkeley, 1966), P. 86. 47 best be described as a report of a research methodology, though its relevance to the study herein reported is based equally, if not more, on the research reported to demonstrate the method.22 The authors are primarily concerned with theories of group structure. After not- ing the immature state of such theories and the gross inconsistency in lists of group prOperties put forth by theorists, they suggest "the strategic direction for work on group structure is the empirical reduction of the theoretically derived lists of prOperties to a few underlying dimensions, which can then be the ingredients 23 of different theories of group structure." To demon- strate their point they offer a description of a method which they refer to as a "multivariate typological pro- cedure" for use in field studies involving a large num- ber of groups. Summarizing their procedure they state: ...standardized data on a large number of variables are gathered from a sample of the members in 15 or more groups. These data are aggregated into means, proportions, and other statistical parameters to describe each group. Factor analysis then reduces to a manageable number the dimensions along which each group is classified. This combination of "aggregative group characteristics" and factor analysis results in factors of group structure and 22Hanan C. Selvin and Warren 0. Hagstrom, "The Empirical Classification of Formal Groups," American Sociological Review, 28: 399-411, 1963. 231bid., p. 402. 48 types of groups that are statistically powerful and easy to interpret. Sixty-one characteristics of each of 20 residence groups in their illustrative study were measured and factor analyzed. Five interpretable factors emerged from the analysis, which were labeled in order: "social satisfaction," "sociometric cohesion," "political con- servatism," "economic status and lack of achievement orientation," and "school spirit sentiment," the labels being derived from the variables with high loadings on the respective factors. The factors were viewed as "dimensions in a 25 Each of the 20 groups 'space' of group properties." was then "roughly" plotted in the two-dimensional space determined by scores of the groups on the first two fac- tors, resulting in a small number of broad types on the factors. Though certain of the sorority groups in the sample tended to have similar characteristics, the authors were quick to point out that none of the common labels such as sorority, dormitory, or cooperative house corresponded precisely to any one of their empirically determined classes. Questionnaire item responses of students in the 24Ibid., p. 399. 251bid., p. 406. 49 sample, whose orientation to higher education had been classified in a manner similar to that used by Nasatir, were compared (though in this case "vocational" respon- ses were used rather than "academic"). This was done according to the empirically determined type of housing in which the students lived. The scanty results reported by Selvin and Hagstrom indicate an interaction between students' orientations, the type of housing in which they live and their questionnaire responses. Selvin and Hagstrom have published a paper, "Cri— ticisms and Afterthoughts," as an appendix to their re- 26 Two of the principle criticisms cently reprinted study. were directed toward what was felt to be the authors' inaccurate interpretation of, in one case, previous re- search and, in the other, of their results. Both issues are of only passing interest here. Other criticisms were, however, directed toward an important procedural inadequacy in their prOposed computation of factor scores for groups under study. The problem resulted from their use of a smaller number of groups than variables. This does no violence to the factor structure but does pre- vent the "inversion of the matrix of correlations between 26Hanan C. Selvin and Warren 0. Hagstrom, "Criti- cisms and Afterthoughts," College Peer Groups, ed. by Theodore M. Newcomb and Everett K. Wilson TChicago: Aldine Publishing Co., 1968), pp. 185-189. 50 the variables" necessary for the computation of group factor scores.27 The interpretation of the first two factors which seem to suggest two different properties of cohesion was 28 also questioned. But the authors have quite adequately defended their findings in a more extensive treatment of the issue.29 Nevertheless, these problems coupled with the proposal by Rettig that multiple disciminant analysis could serve similar ends as those described by Selvin and Hagstrom led to the rejection for the study at hand of their "multivariate typological procedure."30 Selvin and Hagstrom take Rettig to task for failing to clearly distinguish between "classification" 27Hanan C. Selvin and Warren 0. Hagstrom, "Reply to Davis," American Sociological Review, 28: 814, 1963. 28David Gold, "Some Comments on 'The Empirical Classification of Formal Groups,'" and Selvin and Hag- strom, "Reply to Gold," American Sociological Review, 29: 736-739, 1964. 29Warren 0. Hagstrom and Hanan C. Selvin, "Two Dimensions of Cohesiveness in Small Groups," Sociometpy, 28: 30-43, 1965. 30Salomon Rettig, "Multiple Discriminant Analysis: An Illustration," American Sociological Review, 29: 398- 402, 1964; A note of appreciation is extended to Dr. Joseph Saupe, Associate Director of Institutional Re- search, Michigan State University, for having initially suggested the multiple discriminant analysis as an alter- nate method of analysis. 51 and "discrimination."31 But the former authors are per- haps somewhat overly wedded to the use of the term in its numerical taxonomic context in biology. They make ref- erence to Sokal and Sneath's work on numerical taxonomy.32 But these latter authors do not reject the use of mul- tiple discriminant analysis; rather, they note its lim- itations particularly in regard to its early uses by Rao33 who clearly was involved in attempts to discrimin- ate between existing classifications. But this does not exhaust possible uses of the statistic and seems to be the point Rettig attempts to make in the context of re- search on human groups. Perhaps theirs ought not to be an argument between "classification" and "discrimination," but more appropriately an issue over ppgp is to be class- ified. Selvin and Hagstrom seem to be concerned really with the classification of group properties,34 while Rettig looks to the classification of "configuration" of groups.35 In any event the particular statistical 31 p. 186. 32Robert R. Sokal and Peter H. A. Sneath, Prin— ciples of Numerical Taxonomy (San Francisco: W. H. Free- man and Co., 1963). 33 Selvin and Hagstrom, "Criticisms...., op. cit., Ibid., p. 38. 4Selvin and Hagstrom, "The Empirical Classifi- cation...," op. cit., pp. 400-404. 35Rettig, op. cit., p. 398. 52 tool selected should be that one best able to treat the problem at hand. Research on Groups Selvin's interest, at least in the classifica- tion of groups began with earlier research which pointed out methodological problems and inconsistencies regard- ing the subject. He had introduced his concept of using "aggregative variables" in a study of the leadership 36 Selvin suggested: climate in military groups. There is no reason why the method developed for in- ferring measures of perceived leadership from the followers individual ratings cannot be extended to other characteristics of organizations as well. Systematic comparative studies of bureaucracies and other formal groups are rare; it is difficult to summarize the characteristics of a complex organi- zation or even to select appropriate variables for describing them.3 Several other reviewers have also called attention to the problems associated with describing groups.38 In the context of reference groups Hyman and Singer have pointed out the necessity of "a large amount of empiricism 36Hanan C. Selvin, The Effects of Leadership (Glencoe, Ill.: The Free Press of Glencoe, 1960), p. 28. 37 Ibid., p. 164. 38Ivan D. Steiner, "Group Dynamics," Annual Review of Psychology, 15: 422, 1964; Selvin and Hagstrom, pp; cit., p. 402; and John K. Hemphill and Charles M. Westie, "THe Measurement of Group Dimensions," Journal of Psy: chology, 29: 325-42, 1950. 53 and "the development of simple instruments to measure 39 Forehand and Gilmer a person's reference groups..." have extensively discussed the value of studying vari- ation in what they term "organizational" or "environ- 40 They selected five variables from mental climate." thirty group properties extensively subjected to re- search as good examples for the study of organizational 41 I I I I These five are "Size, organization struc- variation. ture, systems complexity, leadership pattern, and goal directions," none of which have been consistent deter- 42 More importantly, the first three minants of behavior. are inappropriate for inter-group comparison for this thesis because of the general homogeniety of the house groups along these dimensions. And, in the general con- text in which they have normally been studied, the final two variables have only limited usefulness, for essentially the same reason. Forehand and Gilmer note that the anal- ogy between individual personality and organizational climate has frequently been used, but suggest "more 39Herbert H. Hyman and Eleanor Singer, Readings in Reference Group Theory and Research (New York: The Free Press, 1968), p. 13. 40Garlie A. Forehand and B. von Haller Gilmer, "Environmental Variation in Studies of Organizational Be- havior," Psychological Bulletin, 62: 361-382, 1964. 41 Ibid., p. 373. 421bid., p. 373. 54 literally, that climate may be treated as a construct, and the 'personality of an organization' identified and dealt with."43 An excellent example of both uses of organiza- tional climate in this fashion and of the study of in- teraction of the individual in his environmental con- text was reported by Halpin and Croft.44 Using the Organizational Climate Description Questionnaire, they delineated six "climates" of elementary schools. These are seen as falling along "a continuum defined at one end as gpgp, at the other, as Closed."45 Turning from the classification problem, several studies suggest Specific variables related to attitude change and behavioral influence within face to face groups. These are considered below. An oft quoted study and the basis for the de- velOpment of important segments of theory of small groups is that of Festinger, Schachter and Back.46 Sociometric 43ibid., p. 377. 44Andrew W. Halpin and Don B. Croft, The Organiza- tional Climate of Schools, (Chicago: Midwest AdminiStra- tion Center, The University of Chicago, 1963). 45 Ibid., p. 2. 46Leon Festinger, Stanley Schachter, and Kurt Back, Social Pressures in Informal Groups (Stanford, Cal.: Stanford University Press, 1950). Refer also to the "theory" section in Chapter I of this thesis. 55 interaction and attitudes within two different married housing projects on the same campus were measured. Their conclusions follow: In a community of people who are homogeneous with respect to many of the factors which determine the development of friendships, the physical factors arising from the arrangement of houses are major determinants of what friendships will develop and what social groupings will be formed. These social groupings create channels of communication for the flow of information and Opinions. Standards for attitudes and behavior relevant to the functioning of the social group develop, with resulting uniform- ity among the members of the group. Pressures to- ward conformity to these standards may result in the exclusion of deviates from the social group. Other people deviate because they were never in communication with the group.47 Physical distance between housing units in the above study was basic to the pattern of friendship devel- Opment. A similar finding in both vertical and hori- zontal types of residence halls was reported by Yarosz and Bradley.48 Newcomb, in another classic study, found that inter personal relationships tended to stabilize within a relatively short period of time.49 It is generally necessary, however, in order for this to occur that ac- quisition of new information about one another cease. 47Ibid., p. 151. 48Edward J. Yarosz and Harry Bradley, "The Rela- tionship Between Physical Distance and Sociometric Choices in Two Residence Halls," International Journal of Socio- metry and Sociatry, 3: 42-45, 1963. 49Theodore M. Newcomb, The Acquaintance Process, (New York: Holt, Rinehart and Winston, 1961). 56 Stabilization tends to lead to an increased relationship between interpersonal attraction and agreement about other residents.50 Newcomb found that relationships were generally well established within four months in two different small groups with members having had no previous acquaintance.51 Newcomb's findings conflict in one respect with an interpretation Jacobson places on his cross-sectional study of 100 committees similarly structured working for a voluntary organization.52 These had been divided into four equal groups: very active committees, medium active, low active and inactive. Many differences across several variables led the author to conclude that the groups were "in various stages of development." For in- stance, "the potential member's self percept" and "his anticipations of the expectations and projected program of the group" accounted for differences between the groups functioning at lower levels of activity. At higher levels "operating group goals, group reward and sanction systems, and group identification" differen- tiated levels of activity. This research would seem to SOIbid., p. 207. 51Theodore M. Newcomb, "The Prediction of Interé personal Attraction," American Psychologist, 11: 582, 1956. 52Eugene Jacobson, "The Growth of Groups in a Voluntary Organization," Journal of Social Issues, 12: 18-23, 1956. S7 imply an orderly sequence of developmental steps which does not seem justified by his data. An equally plaus- ible explanation would be that the groups differed in their internal make-up and consequently proceeded along different avenues, a proposition more in accord with Newcomb's findings. Darley, Gross and Martin conducted an extensive year long study of female college students housed in 13 53 Though initially residents similar rooming houses. seemed heterogeniously mixed, by the end of the year they seemed to have changed in such a way "as to produce homogeniety of variance among the 13 houses." The de- gree of satisfaction experienced was related to the pro- portion of membership returning from the previous year; to a lower prOportion of students leaving the house during the year; to a greater number of paired or recip- rocated sociometric choices; and a higher number of friendship choices within the house compared to outside the house. Group productivity seemed to be related to the quality of leadership and cohesiveness of the group. Rasmussen and Zander found that when a person 53John G. Darley, Neal Gross, and William E. Martin, "Studies of Group Behavior: the Stability, Change, and Interrelations of Psychometric and Socio- metric Variables," Journal of Abnormal and Social Psy- chology, 46: 565-596, 1946. 58 is attracted to an organization and certain issues are important to the group, his level of aspiration conforms to standards attributed to the group.54 Feelings of failure may occur from non-conformity to these standards in proportion to his attraction to the group and his per- ception of the importance of the issues to the organi- zation. These results may be related to Nasatir's55 findings reported earlier, and to Kidd's finding that students rejected in sociometric ratings in residence halls among other things "rated themselves lower on scholastic effort" than other residents.56 In an important study of reference groups using college housing groups Siegel and Siegel found: When divergent membership groups with disparate at- titude norms were socially imposed on the basis of a random event (on subjects who initially shared a common reference group), attitude change in the sub- jects over time was a function of the normative at— titudes of both imposed membership groups and the individuals' reference groups. The greatest attitude change occurred in subjects who came to take the im- posed, initially nonpreferred, membership group as their reference group. 54Glen Rasmussen and Alvin Zander, "Group Membership and Self-Evaluation," Human Relations, 7: 239-251, 1954. 55 Nasatir, op. cit., pp. 290-298. 56John W. Kidd, "The Students Live," Evaluation in the Basic College at Michigan State University, Paul L. Dressel, (edJ7 (New York: Harper, 19587: p. 174. 57Alberta Engvall Siegel and Sidney Siegel, Ref- erence Groups, Membership Groups, and Attitude Change," Journal of Abnormal and Social Psychology, 55: 360-364, I957. 59 Leadership as a variable in the influence of groups on behavior has been studied from many perspec- tives. Given the roles of the resident advisor, head residents and hall officers, Etzioni's consideration of "charisma" in the context of leadership has relevance here.58 Charisma "is the ability of an actor," he sug- gests, "to exercise diffuse and intense influence over the normative orientation of other actors." Kidd noted that resident assistants who were relatively high in leadership prestige and in friendship ratings seemed to have high morale and c00peration among 59 their residents. Golembiewski pointed out that "mem- bers of small groups...can identify a rank order of pow- erful individuals.60 These individuals in turn are gen- erally aware of their power position within the group. They behave accordingly," And Taylor reported, in his studies at Sarah Lawrence, "that the quality of life within a given residence depended most of all on the qualifications of the house president for holding office."61 58Amitai Etzioni, A Comparative Analysis of Com- plex Organizations (New York: The Free Press of Glencoe, I961), p. 203. 59John W. Kidd, "An Analysis of Social Rejection in A College Men's Residence Hall" (unpublished disserta- tion, Michigan State University, 1951). 60Robert T. Golembiewski, The Small Group (Chicago: University of Chicago Press, 1962), pp. 100-101. 61Harold Taylor, "Freedom and Authority on the Campus," The American College, ed. by Nevitt Sanford (New York: Wiley & Sons, 1962), p. 791. ‘ 60 Research on Residmce Halls A few years ago Anderson made the interesting observation that of ten schools studied only one had formalized in writing a philosophy of residence hall living.62 Though more is now known of the dynamics of college life, the finding hints at the frequent gulf be- tween what is known and its application in residence hall administration. Several areas will be considered in this section; normative systems in residence groups, com- parisons with other living styles, and the impact of special programs. Normative Systems Lidzey and Urdan observed that "even within the homogeneous college community it appears that individual dorms develop their own standards and preferences in terms of which choice is mediated and oriented," an ef- fect which could "conceal relationships" in studies of 63 the larger community. Sinnett found several differ- ences “in diverse variables" between randomly assigned, 62Gary Robert Anderson, "An Examination of Resi- dence Hall Counseling Programs for Men in Ten Selected Midwestern Colleges and Universities," (unpublished dis- sertation, Northwestern University, 1959). 63Gardner Lindzey and James A. Urdan, "Person- ality and Social Class," Sociometry, 17: 60, 1954. 61 freshman coed residents of two dormitories.64 Clingan was less successful and noted only one minor difference between residents of two halls differing in architectural style in evaluating the effects of an experimental pro- gram applied to both halls.65 Boyer chronicled the de- velopment of "micro-social systems" among groups of students living together in residence hall suites and noted higher academic performance in groups with high mutual acceptance and an emphasis on studying.66 The Harvard University Houses represent one of the early major attempts to create a unique educational residential environment. Jencks and Riesman67 charac- terized them according to students' perceptions of the differing "personalities" of the houses. Undergraduates are generally required to live in the houses after their 64E. Robert Sinnett, "Socio-Economic Status and the Use of Campus Facilities: A Tale of Two Dormitories," Personnel and Guidance Journal, 43: 993-996, 1965. 65Wallace Eugene Clingan, "The Educational Effects of an Accelerated Personnel Program in the Men's Residence Halls at the University of Missouri" (unpublished disser- tation, University of Missouri, 1959), Dissertation Abstracts, 20: 3627. 66Ronal K. Boyer, "Student Peer Group: Its Ef- fect on College Performance" (unpublished dissertation, Case Institute of Technology, 1965). 67Christopher S. Jencks and David Riesman, "Pat- terns of Residential Education: A Case Study of Harvard," The American College, ed. by Nevitt Sanford (New York: Wiley & Sons, 1962): pp. 731-773. 62 freshmen year. A senior professor administers each house assisted by several younger faculty members who also live in and act as tutors. Students may request a certain house but selection is somewhat controlled to provide each house with a cross-section of the student body. Vreeland and Bidwell studied values and attitudes of res- idents, tutors and the faculty house master.68 The houses were found to differ along two broad categories of house goals, one centering on an "individual-orienta- tion," that is, centering around respect for individual variations in attitudes and behavior styles. The other, the "collectivity-orientation," emphasized group cohesive- ness and loyalty. The extent of change in residents' values was "strongly" related to "peer involvement," a measure of mutual attraction among residents of a given house. Where peer involvement was high, student change was toward peer values and somewhat away from those of the Master's. Where peer involvement was low, however, student change was more toward that of the Master's. Based on the patterns of interaction between stu- dents at Sarah Lawrence, Taylor found some confirmation that "the relationships among students in the residences are the greatest single factor in their general attitude 68Rebecca Vreeland and Charles Bidwell, "Organi- zational Effects on Student Attitudes: A Study of the Harvard Houses," Sociology of Education, 38: 233-250, 1964-65. 63 toward the college and toward themselves."69 Lozoff came to a similar conclusion and offers a partial interpreta- tion: ...the relationship between the social environment of the students and the intellectual receptivity and emotional well-being was frequently noted by students and interviewers over the four years (of the study). In the Interviews, most of the students indicated that much of their mental and psychic energies were involved in developing a sense of their own separatedness and uniqueness and modify- ing their behavior so that they could relate to others in a gratifying and meaningful way. The res- idence groups to which they belonged played an im- 70 portant part in defining the patterns of adaptation. Lehmann and Dressel concur and add another dimension to our understanding of interpersonal relationships-in the halls. The most significant experience in the collegiate lives of these students has been their association with differing personalities in their living unit. Although the peer group, comprised of the total body of students, did not have too much impact upon the behavior of these students, the analysis of inter- view and questionnaire data strongly suggested that discussions and "bull-sessions" were a potent factor in shaping the attitudes and values of these college students. 1 69Taylor, op. cit., pp. 789-790. 7OMarjorie M. Lozoff, "Personality Differences and Residential Choice," Growth and Constraint in College Students, ed. by Joseph Katz. Quote taken from chapter draft received from the author through personal corres- pondence. 71Irvin J. Lehmann and Paul L. Dressel, Critical Thinking, Attitudes, Values in Higher Education (Coopera- tive Research Project No. 590, Department of Health, Edu- cation and Welfare, Michigan State University, 1962), p. 268. 64 Several studies demonstrate selective differences between the residents of different styles of living ac- commodations such as Greek units, residence halls, board- ing houses, and residences of commuters. Residence hall groups tended to fall between fraternities at the high end and boarding units and commuters at the low on mea- sures of social relationships.72 Residence hall students tended to be higher in independent thought than frater- nities73 and had a higher proportion of over-achieving males while fraternities had a higher proportion of under— 74 achievers. Hartnett at Michigan State University found that residence hall living tended to be associated with positive changes in academic achievement in contrast to Greek living which was associated with negative change.75 72Robert Dollar, "A Study of Certain Psychological Differences Among Dormitory, Fraternity, and Off-Campus Freshman Men at Oklahoma State University" (unpublished dissertation, Oklahoma State University, 1963): and Roger Jay Fritz, "A Comparison of Attitude Differences and Changes of College Freshman Men Living in Various Types of Housing" (unpublished dissertation, University of Wis- consin, 1956). 73Ibid. 74Char1es L. Diener, "Similarities and Differences Between Over-Achieving and Under-Achieving Students," Personnel and Guidance Journal, 38: 396-400, 1960. 75Rodney T. Hartnett, "Place of Residence as a Factor in Academic Performance Patterns of College Stu- dents" (unpublished dissertation, Michigan State Univer- sity, 1963). 65 However, the several studies that have investigated dif- ferences in academic performance between housing types are generally inconsistent, perhaps in part because in- ternal variations within types have generally not been 76 In this latter in- reported except for fraternities. stance regarding fraternities and sororities, several studies indicate important differences in status, influ- ence and academic performance of Greek units.77 Paren- thetically, it should be pointed out that these groups have been well established in the literature as highly attractive reference groups influencing student behavior along several dimensions. Apparently, with infrequent but important exceptions such as the Harvard Houses, the relative attraction of residence halls is not nearly as great as that of the Greek units. This is not meant to exclude the possibility of a given hall or subgroup with- in the hall playing a role similar to that of the typical Greek unit. Indeed, the exploration of this possibility is central to this study. 76For example see R. E. Prusok and W. B. Walsh, "College Students' Residence and Academic Achievement," Journal of College Student Personnel, 5: 180-184, 1964. 77E.1F.Ga.rdher and G. G. Thompson, Social Rela- tions and Morale in Small Groups (New York: Appleton- Century-Crofts, 1956): Robert E. Matson, "A Study of the Influence of Fraternity, Residence Hall and Off- Campus Living on Students of High, Average and Low College Potential," Journal of the National Association of Women Deans and Counselors, 26: 24-29, 1963: and William A. Scott, Values and Organizationg: The Study of Fraterni- ties and Sororities (Chicago: Rand McNally Co., 1965) . 66 Two other important studies that reflect the in- teraction between personality, residential choice, and the influence of various types of broad living categories 78 and Lozoff.79 In a very insight- are those of Selvin ful, though somewhat impressionistic analysis of residence hall life at Berkeley by noted architects, the point is forcefully made that the traditional "dormitory conditions tend to filter out students whose presence adds diversity and a sense of intellectual dialogue to the (university) community."80 They are convinced that the halls, at least at Berkeley, cater to the "collegiate" type of student to a large extent because of the failure in the halls to adequately provide for "independence, (and) a diversity of activities and friends...characteristics of successful student living. And yet, it is the search for these conditions that drives many students 925 of the dormitory."81 78Hanan C. Selvin, "The Impact of University Ex- periences on Occupational Plans," School Review, 71: 317-329, 1963. 79Lozoff, loc. cit. 80Sim Van der Ryn and Murray Silverstein, Dorms at_Berkeley (Berkeley, Calif.: Center for Planning and Development Research, University of California, 1967), p. 27. BlIbido, pp. 27-28. 67 Impacts of Special Programs Neither Chesin82 nor Herbert83 were able to dis- cern any significant difference in attitude change, aca- demic performance or satisfaction between freshmen living in halls with older students and those living in units without upper-classmen or with only a small proportion. Chesin, however, did find that more freshmen Withdrew from units housing upper-classmen than from "freshmen units."84 Though Chesin does not consider the possibil- ity, perhaps, were more information available, this find- ing could be related to Nasatir's provocative results. A few studies have indicated some positive ef- fect of roommates enrolled together in common courses or curricula and correlation between their academic perform- ance, but the results are not consistent.85 Two final studies are included in this section because of their provocative implications as portraying 82Sorrell E. Chesin, "The Differential Effects of Housing on College Freshmen" (unpublished dissertation, Michigan State University, 1967). 83David J. Herbert, "The Relationship Between the Percentage of Freshmen on a Residence Hall Corridor and the Grade Point Averages of the Occupants," College and University, 41: 348—352, 1966. 84Chesin, op. cit., p. 101. 85Newcomb and Feldman, op. cit., p. 213-216. 68 significant dimensions of house life. Lozoff, noting that the "self-esteem of most freshman and s0phomore stu- dents is not very sturdy," described the role frequently played by fraternities in arranging social Opportunities and thus facilitating social growth.86 To the extent a residence hall group could facilitate the same inter- action, one could speculate that its attractiveness to the residents would be enhanced. In the second study Davis87 reported "the pro- portion of seniors who endorse intellectual values ('true climate of intellectualism') and the prOportion who see their classmates as having intellectual values ('per- ceived climate of intellectualism'LH' Across 135 colleges and universities on a given campus the prOportion of students perceived as having intellectual values consis- tently laged behind the proportion endorsing intellectual values.88 Though the study does not directly relate to residence halls, it does call into question why this finding occurred. An alternate interpretation to those offered by Davis, would be that contra-intellectual peer group norms impinging on students create a condition 86Lozoff, loc. cit. 87James A. Davis, "Intellectual Climates in 135 American Colleges and Universities: A Study in 'Social Psychophysics," Sociology of Education, 37: 110-128, 1963. 88Ibid., pp. 117-119. 69 where students frequently deny their intellectual in- clinations before their peers. Hence, students only infrequently would see other students reflecting similar values. For a positive intellectual environment to emerge these postulated norms would have to be reduced. Residence Halls At Michigan State University Several studies in recent years have been direct- ed. toward a partial assessment of the Michigan State University housing program. Centra, using Pace's Col- lege and University Environment Scales (CUES), compared perceptions of residence hall environment separately for freshmen and upper-classmen grouped according to the de- 89 Differ- sign and location of their residence halls. ences were noted on what Centra termed "an intellectual- propriety dimension;" that is, certain residence groups "tended to be more academically competitive and estheti- cally aware, as well as less convention-flouting, than other residence groups." For both men and women the Brody group of halls, fairly modern structures each build- ing housing a single sex somewhat far removed from campus, were perceived as having the least intellectual environment. 89John A. Centra, "Student Perceptions of Resi- dence Hall Environments," Office of Institutional Re- search, Michigan State University, June 13, 1966 (unpub- lished office report). 70 At the Opposite end of campus and on the Opposite end of the ratings were a small group of men's halls (Abbot and Snyder). Women rated the West Circle group of halls highest, somewhat removed from Abbot and Snyder but of the small general physical design. Between these ex— tremes were the “living-learning units" (Case-Wilson- Wonders, Akers-Fee-McDonel) and in addition, for women the East Campus Group (Mason and Phillips). The living- learning units are large structures housing men in one high-rise wing and women in another separated by a com- mons facility housing residence hall staff, dining and recreation areas, and faculty Offices and classrooms. Certain of the units have, in addition, auditoria or branch libraries and laboratories. Some evidence was also found suggesting that students' perceptions of their residence hall environment may color their views of the total university environment. Adams, in an evaluation of the first of the living-learning units working from the framework of the Clark-Trow subcultures, found that at the beginning of the third year at Michigan State "subculture identity" differentiated a student's residence.90 "Collegiate" 90Donald Van Adams, "An Analysis of Student Sub- cultures at Michigan State University" (unpublished dis- sertation, Michigan State University, 1965), p. 128. students dents to ale" to n number of dence hal 1 all catep dicated t twp to career. tional ar To what 6 the Hawt] ment of Where, 0 three YE these la fOlmd b: both St \ (D UIU‘CIJL‘JS (D ‘<;' "1‘ 71 students tended to move to fraternities, "academic" stu- dents to supervised housing and COOperatives, "vocation- als" to married or unsupervised housing while the small number of non-conformists tended to remain in the resi- dence halls. In the forth year the general tendency in all categories was toward unsupervised housing. Responses Of the sample of former residents in- dicated that 54% found their experiences in their living group to have been the most profitable Of their college career. In retrospect they highly favored the coeduca- tional and all-freshmen aspects Of the living unit.91 To what extent their perceptions were distorted through the Hawthorne effect as a result of the special treat- ment Of the hall, or in their not having living else- where, or by their having to resort to memory recall some three years after the experience is not known. HOwever, these latter findings to a large extent parallel those found by Olson.92 Over a period of years Olson has assessed the somewhat unique characteristics of the Michigan State University living-learning units and their impact on both students and faculty. Beginning with an open-ended 911bid., pp. 123-125. 92LeRoy A. Olson, "Methods and Results of Research on Living-Learning Residence Halls" (paper read at the an- nual Forum of the Association for Institutional Research, Boston, Massachusetts, May 3, 1966): several other reports by Olson are available through the Office of Evaluation Services, Michigan State University. questionn ing-learn ions were was admin period of pests of versity. : instructc tions, st residence residence C the IESpc bEtweeI-l . 72 questionnaire study of residents of Case hall, the liv- ing-learning unit studied by Adams, Olson develOped var- ious versions of a fixed response questionnaire which was administered to residents of various halls over a period of years. Included were items about several as- pects of the intellectual and cultural life of the Uni- versity. "Attitudes toward class atmosphere, student- instructor contact, advising, study methods and condi- tions, student inter-relationships, social activities, residence hall conduct, regulations, accommodations and residence hall life were also considered."93 General results indicated differences between the reSponse patterns of male and female residents and between the various types Of halls grouped somewhat as 94 Faculty responses to the they were in Centrafs study. living-learning units like those of the students were generally favorable, among other things, indicating more out of class student contacts and a more informal atmos- phere.95 However, only "inconsistent" results were noted in comparing performance on common final examinations between residents of living-learning halls and other halls.96 93Ibid., p. 5. 94Centra, loc. cit. 95Olson, op. cit., p. 7. 96Ibid., pp. 7-8. 73 Summary and Conclusions In this chapter several relevant themes have been considered in order to lay a foundation for the further study Of the characteristics of small group life in residence halls and the influence of these character- istics on behavior of residents. Students were able to characterize themselves according to four hypothetical constructs relating to different orientations to college life. These constructs were useful in assessing the manner in which student sub- cultures mediate the influence of the college environment They apparently possess a fair amount of construct validity. There were strong indications Of an interaction between the characteristics of various residence groups and of the residents along several dimensions. The methodology for describing the "personality of groups" was not consistent in part because of lack of concur- rence as to what are the significant dimensions and in part because of the different methodologies employed. The specific questions asked varied and hence would dic- tate different statistical approaches and different the- oretical or empirical frameworks. However, new applica- tions and multivariate techniques were indicated as new and potentially promising approaches. 5 several | denonstra in other tween the nal. Cer dent that character sufficier 74 Nevertheless, within the research on small groups several broad categories of variables have consistently demonstrated a relationship to attitude change or have in other ways influenced or mediated the interaction be- tween the larger social system, the group, and the individ- ual. Certainly, under certain circumstances it was evi- dent that a housing unit had the capacity to take on the characteristics of a reference group. Propinquity, a sufficient length of time for a normative system to de— velOp, a capacity within the group to satisfy various needs of residents (particularly those associated with Status, acceptance and survival in the academic commun— ity) , leadership, and group cohesion have each been shown to be related to the attractiveness of housing groups. And they are related to the degree to which housing groups serve more than simply a membership group function. There were some empirical indications of the development Of different normative systems which dif- ferentially influenced residents within living units. This was most pronounced in but not limited to frater- nities and sororities which do develop strong norms and de"land a high degree Of compliance. The review of the literature nevertheless revealed no studies, with the PoSSible exception of that of Selvin and Hagstrom, having had as their primary purpose the systematic Study of relatively homogenious residence hall groups. 75 Of the several studies assessing the impact of college living the majority considered differences between, rather than within types of housing which may well have masked the very differences they attempted to uncover. Studies at Michigan State University with its highly articulated living-learning residence halls, yet With a mixture Of traditional units, indicated inter- hall differences in students' perceptions of the Univer- Sity and of their living conditions. In conclusion, the review underscored the pos- sibility of discerning significant and relevant varia- tions in the "personality" of groups within the Michigan Stai1ze residence hall program. If, indeed, such differ— ences can be empirically assessed, the findings may en- riCh our understanding of those conditions best calcu- lated to create an enriching educational environment. The instruments and methodology employed in this attempt at such an assessment are considered in the following th> chapters . CHAPTER I I I INSTRUMENTATION The description of methodology Of the study has been subdivided into two chapters. The first of these is devoted to a presentation and discussion of the five instruments used in collecting the data. In the follow- ing chapter the population and sample, the hypotheses, and methodology are discussed. The five instruments are, in the order in which they are presented, (1) the College Qualifications Tests, (2) the Michigan State University Reading Test, (3) the Attitude Inventory consisting Of four scales from the Omnibus Personality Inventory, (4) a measure of the four Clark-Trow subcultures, and (5) the Honse Analysis Survey. The latter is an instrument de- signed for this study to assess characteristics of life Within residence hall houses (copies of instruments 4 and 5 appear in the Appendix). College Qualification Tests (CQT) The CQT consists of three tests measuring verbal 76 77 ability, numerical ability and general information. Their three scores contribute to a total score which the authors indicate to be "highly predictive of first sem- ester grade point average." Validity coefficients range from .34 to .73.3 The authors also report correc- ted Odd-even score reliability coefficients of .97 for male college freshmen and .96 for female freshmen. Lehmann and Dressel have reported a split-halves coef- ficient of .93 on a Michigan State University sample. Only CQT total scores were used in this study. Michigan State University Reading Test This is a 45 item test developed at Michigan State University as a measure of students' ability to comprehend reading material.6 Lehmann and Dressel re— Port a .79 reliability coefficient from a group of 1George Bennett, Marjorie G. Bennett, Winburn L. Wallace, and Alexander G. Wesman, College Qualification Tests, Manual (New York: The Psychological Corporation, Revised, 1961) . 2 Ibid., p. 45. 31bid., p. 46. 41bid., p. 53. . 5Irvin J. Lehmann and Paul L. Dressel, Critical IElli-\fnking, Attitudes, Values in Higher Education—(Coopera- tlVe Research Project NO. 590, Office of Education, U. S. Department of Health, Education and Welfare, Michigan State University, 1962), p. 24. f. 6"MSU Reading Test" (East Lansing, Michigan: Of- 1°e of Evaluation Services). 78 Michigan State Freshmen7 and correlations with the CQT Totzil Score of .70 for male freshmen and .71 for females.8 Attitude Inventory (AI) Four scales from the Omnibus Personality Inven- tcxrjy (OPI) were combined for the study and administered t<> 'the sample under the title, Attitude Inventory. The f<>tLI'included were "Thinking Introversion" (TI). "The- oretical Orientation" (TO), "Estheticism" (Es), and "Autonomy" (Au) and are described below. Thinking Introversion (TI) - 60 items Persons scoring high on this measure are charac- terized by a liking for reflective thought and academic activities. They express interests in a broad range of ideas and in a variety Of areas, such as literature, art and philosophy. Their thinking is less dominated by objective condi- tions and generally accepted ideas than that of thinking extroverts (low scorers). Most extro- verts show a preference for overt action and tend to evaluate ideas on the basis of their practical, immediate application. Theoretical Orientation (TO) - 32 items This scale measures an interest in, or orienta- tion to, a more restricted range of ideas than is true Of TI. High scorers are interested in science and in some scientific activities, in- cluding a preference for using the scientific method in thinking. They are generally logical, analytical, and critical in their approach to problems. 7Lehmann and Dressel, loc. cit. 8Ibid., p. 321. 79 Estheticism (Es) - 24 items High scorers endorse statements indicating di- verse interests in, as well as an appreciation of, artistic matters and activities. The focus of their interests tends to extend beyond paint- ing, sculpture and music and includes interests in literature and dramatics. .Autonomy (Au) - 40 items The characteristic measured is composed of non- authoritarian attitudes and a need for indepen- dence. High scorers are sufficiently indepen- dent Of authority, as traditonally imposed through social institutions, that they Oppose infringe- ments on the rights Of individuals. They are tolerant of vieWpOints other than their own, and they are nonjudgmental, realistic, and intellec- tually liberal. The OPI was originally designed by McConnell and Heist for use in studies of growth and development Of Stildients using scales derived from several sources which, £01? the most part, were thought to measure characteris- tics of self-actualizing peOple.lo Six of the fourteen SGales in the current version of the OPI have been used as a. "composite index of the degree of intellectual 9"Omnibus Personality Inventory (Form F) --— Brief Scale Descriptions," Center for the Study of Higher Edu- cation, Berkeley, California, no date (mimeographed). 10Omnibus Personality Inventory, Research Manual (gfiilfikeley, Calif.: Center for the Study of Higher Educa- ¥<>r1, 1962); Appreciation is extended to Dr. Paul Heist, li?€2ctor,for granting permission to use portions of the OPI - The OPI was distributed for research purposes only anti .is copyrighted by the Center for the Study of Higher EduCation, 1963. 80 orientation" of students.11 These six include the four scales described above plus two scales, "Complexity" and "Religious Orientation." Brown, using four of the six, Thinking Introver- sion, Theoretical Orientation, Estheticism and Complexity, studied relationships between first year freshman grades, self-reports of intellectual activities and the scale scores.12 Correlation coefficients between the OPI scales and the intellectual activities ranged from .10 to .49, with a median of .24. Only the TO scale differed appre- ciably from zero (.16) among the OPI scales and the ac- tivity measures when correlated with grade point average. In another report Brown also noted that scale scores in- creased more for students living in a residence hall sub- jected to an enrichment program of informal contacts with faculty than for a control group not involved in the special program.13 llPersonal correspondance from Paul Heist, Director, Center for the Study of Higher Education, Berkeley, Calif., Sept. 17, 1964. 12Robert D. Brown, "How Compatible are the Intel- lectual and the Academic Aspects of College Life? An In- vestigation of the Relationship Between the Intellectual and the Academic Aspects of College Life" (paper read at the American Personnel and Guidance Association Conven- tion, Dallas, Texas, March, 1967). 13Robert D. Brown, "Manipulation Of the Environ- mental Press in a College Residence Hall," Personnel and Guidance Journal, 46 (1968), 550—560. Table 3.1 Reliability coefficients (internal consisten ) for four Omnibus Personality Inventory scales Scales Sample TI TO Es Au Normative Group* (N=2,390) .85 .74 .80 .80 *Kuder-Richardson: Formula #21 aCenter for the Study of Higher Education, Berke- ley, Calif., Omnibus Personality Inventory Research Manual, 1962, p. 11 (Since the collection of data for this thesis a later version of the OPI has been published. The scales of the more recent version are very similar to those used in this thesis. The most significant change is the reduc- tion of the TI scale from 60 to 43 items). In Table 3.1, reliability information concerning the OPI scales included in the Attitude Inventory is pre- sented. The data were Obtained from a normative sample representing several colleges and universities and sug- gest a reasonably high level of internal consistency. Intercorrelations of the four scales, MSU Reading test, CQT, and fall and winter accumulative grade-point-average for the 637 male freshmen investigated in this report are presented in Table 3.2. These four OPI scales were included in this study for both theoretical and pragmatic reasons. If it were possible to meaningfully discriminate between the 81 Table 3.2 Product—moment correlation coefficients between four scales from the Omnibus Personality Inven- tory, MSU Reading Test, CQT total, and accumu- lative fall and winter quarter grade-point- average, 1964—65, for 637 male Michigan State University freshmen Variables l 2 3 4 5 6 1. Thinking Introversion 2. Theoretical Orientation .67 3. Estheticism .59 .28 4. Autonomy .31 .33 .13 5. MSU Reading Test .32 .32 .20 .39 6. CQT total .29 .34 .20 .40 .77 7. Two Qtr. G.P.A. .19 .17 .06 .22 .49 .51 characteristics Of the 27 houses included in this study, one might expect a differential impact on freshman atti- tudes as a result. Particularly, if within a given house or group of houses, peer-norms seemed more in support of academic-intellectual variables than in other houses, one might expect the impact of such an environment to be positively reflected in the attitudes of freshman resi- dents. Likewise, if the environment were anti-intellec- tual, the reverse might result. The six scales purported to measure intellectual disposition seemed apropos to the question (The reader may wish to review in this regard 82 83 the discussion of a "Climate of Learning" presented in Chapter I). It was felt, however, that the average respon- dent should be able to complete the test battery within a certain length of time in order to insure maximum student cooperation for both pre- and post-test sessions. It was therefore decided that the total length of the Attitude Inventory should not exceed 150 items, somewhat under the 196 of the six scales. The selection Of which four of the six scales to be finally included was not totally arbitrary. The autonomy scale was of interest over and above its inclusion as one of the six measures of intellectual disposition. The allegation has been made that tradition- al residence hall life, despite the best of intentions, may inhibit independence and individual growth rather than foster these traits.14 Though a comparison between types of housing (e.g. residence halls, fraternities, Off- campus housing) rather than within types would be necessary to adequately test this proposition, variations in house characteristics might nevertheless differentially influ- ence growth toward autonomy. l4Sim Van der Ryn and Murray Silverstein, Dorms at Berkeley (Berkeley, California: The Center for Plan- ning and Development Research, University of California), pp. 27-28, 65-68. 84 The Thinking Introversion, Theoretical Orienta— tion, and Estheticism scales were felt (somewhat arbi- trarily) to assess attitudes that might best be subject to change in a positive academic climate. These three scales comprised all but one of those used by Brown in his assessment of intellectual disposition.15 Heist had also indicated that these were three of the four that were under consideration in other studies of intellec- tual disposition (the forth being Complexity which was also used by Brown).16 Student Subcultures In order to obtain an indication of students' orientations toward higher education and hence of the subculture with which they might individually identify, four descriptive paragraphs paralleling the Clark-Trow postulated subcultures were used in the study.17 Stu— dents were requested tO indicate which of the four para- graphs best described the kind of person they considered themselves to be and to respond to other related questions. Thus, the paragraphs constitute Operational definitions 15Brown, loc. cit. 16Paul Heist, Director, Center for the Study Of Higher Education, Berkeley, Calif., Telephone inquiry, August, 1964. 17See pp. 39-41. 85 of the subcultures. The paragraphs follow. Type 1 (Vocational) This kind of person views education principally as a means of preparing for his professional future. He is not particularly disinterested in the social or purely intellectual phases of campus life, though his participation compared to some may be limited. This person does his homework but tends to do little outside reading or restricts his reading to the light, general entertainment variety. All things considered, this person's primary reason for being In college is to obtain professional training. Type 2 (Nonconformist) This person is interested in learning about life in general, but in a manner of his own choosing. He is very interested in the world of ideas and books, and eagerly seeks out these things. Outside of the classroom, this person would attend such activities as the lecture-concert series, Provost lectures, foreign films, etc. This person Often pursues his own interests in place of or in addition to mere course requirements and will frequently do extra readings in order to obtain a more complete under- standing of the world in which he lives. From a social point-of—view, this person tends to reject activities such as fraternities, sororities, and the social events that many consider a part of cam- pus life. When this person does join, it will usually be one of the more intellectual, academic or political campus organizations. For the most part, this person would consider himself to be some- one who is primarily motivated by intellectual curIOsity. TYpe 3 (Academic) This person is in some respects like Type 2 noted above. He is concerned with books and the pursuit of knowledge, bat is also the kind Of person who leads an active social life on campus. He is inter- ested in getting high grades and tries to maintain a high grade-point average. He is the kind of per- son who will eagerly work with student or hall govern- ment, fraternities, committees, and activities Of this type. He would feel that both the social side of college life and the academic are important for His general development. 86 Type 4 (Collegiate) This is the kind of person who is more concerned with the social phases of college life and learning to get along with individuals. He identifies close- ly with the college and enjoys attending as many campus social and athletic events as possible. This person may be interested in intellectual kinds of things but will, for the most part, find greater satisfaction in student government, parties, activi- ties, etc. He is concerned about his education but feels that the develOpment of his social and leader- ship skills are certainly important. Much of his college life will be centered around non-academic type activities such as committees, fraternities or sororities, or resident hall type activities. This person will try to maintain his grades but does not feel that he must necessarily make the highest grades or go out of his way to do extra or non-assigned readings in order to be a success in college. Gottlieb and Hodgkins originally developed the descriptive paragraphs of the Clark-Trow subcultures from which the above were adapted.18 After pre-testing the statements they concluded that they "were reasonably valid." The paragraphs were subsequently used in the "Senior-Year Experience Inventory," as a part of an ex- tensive four year study of attitudes and value change at Michigan State University.19 The "College Student Questionnaire," a relatively new instrument published by the Educational Testing Ser- vice also incorporated descriptive paragraphs Of the 18David Gottlieg and Benjamin Hodgkins, "College Student Subcultures: Their Structure and Characteristics in Relation to Student Attitude Change," The School Review, 71 (1963), 266-289. 19 Lehmann and Dressel, Op. cit., p. 300. 87 four subcultures though these were not identical to those authored by Gottlieb and Hodgkins. Adams used the ver- sion from ETS in his study of subcultures at Michigan State University.20 The paragraphs used in this study,and quoted pre- viously in this chapter,were modified slightly from those developed by Gottlieb and Hodgkins in order to incorpor- ate part of the phraseology used in the ETS version. The latter seemed more in harmony with the original conceptions Of the subcultures postulated by Clark and Trow. The typology in any of the forms in which it has been used thusfar must be cautiously interpreted for, as Hodgkins indicated, the paragraphs can serve only as 21 Nevertheless it was thought that "crude" indicators. the paragraph approach, having a sound basis in theory, would be more appropriate for the purposes of this study than the single questionnaire item approach used by Nasa- 22 23 tir and Selvin and Hagstrom in describing students' orientations to college. Certain salient points have been 20Donald Van Adams, "An Analysis of Student Sub- cultures at Michigan State University" (unpublished dis- sertation, MSU, 1965), pp. 46-49. 21Benjamin Joseph Hodgkins, "Student Subcultures-- An Analysis of their Origins and Affects on Student At- titude and Value Change in Higher Education" (unpublished dissertation, Michigan State University, 1964), p. 113. 22David Nasatir, "A Contextual Analysis of Academic Failure," The School Review, 71 (1963), 290 23Hanan C. Selvin and William O. Hagstrom, "The Empirical Classification of Formal Groups," American Sociological Review, 28 (1963), 399-411. 88 made by Hodgkins concerning the use of the paragraphs and are quoted here. They (the subcultures) are pat mutually exclusive. This is in keeping with the point...that differences between students in educational goal orientation were differences of degree not kind. The intent in develOping these statements was not to force the subject into a category which may have distorted his response, but to make the scope of the statement sufficiently broad to allow relatively easy recog- nition Of similarity on the part of the respondent. House Analysis Survey (HAS) Central in the collection of the data was the House Analysis Survey (HAS), a new instrument construc- ted for this study to assess (1) characteristics of house life in residence halls and (2) related attitudes of residents. It consists of 128 questionnaire items. It also includes the four paragraphs describing the Student Subcultures which were used to Obtain a post-test mea- sure Of students' orientations to college. The items were divided into two sections to facilitate administra- tion and were designed to be efficiently scored on the IBM 1230 Optical Scanner. A large number of questionnaire items were cre- ated originally for this study having their roots in re- search and theory related to group life. A review of both the theoretical and empirical literature had sug- gested several possible avenues for exploration of 24Hodgkins, loc. cit. 89 relevant dimensions of house life. These have been re- viewed in Chapters I and II and include such molar di- mensions as group morale and cohesion, individual and group status and prestige, general attractiveness and satisfaction with the group. Others are academic and intellectual climate and leadership, the capacity of group life to satisfy social needs, physical character- istics of the buildings and resulting patterns of inter- actions, program differences, length Of association among residents, group goals and individual character- istics Of residents. In addition, head residents, res- ident assistants, students, and other housing officers were invited to suggest other salient characteristics of house life. Next several research instruments were studied in search of ideas for specific items to measure the above dimensions. These included both the College Characteristics Index and its later version, the College and University Environment Scales, the College Student Questionnaires, and the local instruments developed by Olson for use in assessing attitudes of students in 25 Michigan State University residence halls. From these 25C. Robert Pace and George G. Stern, "An Approach to the Measurement of Psychological Characteristics of College Environments," Journal Educational Psychology, 49 (1958), 269-277; C. Robert Pace, College and University Environment Scales (Princeton, N. J.: Educational Testing Service, 1963); Richard E. Peterson, College Student Ques- tionnaires (Princeton, N. J.: Institutional Research Pro- gram for Higher Education, Educational Testing Service, 1965); LeRoy A. Olson, "Student Attitudes Inventory" (East 90 instruments several additional items were adapted for use in the specific context of residence hall and house assessment. In particular, several items were incorpor- ated from Olson's studies which had been shown to dis- criminate between halls and/or were otherwise of direct interest. To reduce the number of items, several factors were considered. It was felt that the majority of the items should reflect house or hall characteristics rather than students' general attitudes about the university or themselves. In this sense the questionnaire was pat- terned after the CCI and CUES where students are asked to serve as "reporters" whose "aggregate judgment pro- vides an Opinion poll approach" to the characteristics 26 This is not meant to infer, of a college environment. however, that the HAS has the sophistication Of the sur- veys developed by Pace. It was also felt that the items should reflect variables that might logically or the- oretically be expected to vary from house to house. And, in general, items sampling as many pertinent character- istics as possible were to be included within the practical Lansing, Mich.: Office of Evaluation Services, Michigan State University, 1964). 26C. Robert Pace, Preliminary Technical Manual: College and University EnvirOnment Scales (Princeton, N. J.: Educational Testing Service, 1963). 91 limits of the statistical design and time alotted within testing sessions. .As was pointed out in Chapter I, theories of group characteristics are in an early stage of develop- ment. Hence no a priori assumptions were made that all or even the most significant measures of group life were finally included. Several items were eliminated on the basis of apparent irrelevancy, redundancy or inadaptibility to the scoring procedures or general format of the instru- ment. The remaining items were then submitted to col- leages in the Offices of Residence Hall Programs and Evaluation Services and to the author's thesis committee for reactions. Additional items were deleted or modi- fied. Finally, the resulting preliminary form Of the instrument was administered to a group of approximately thirty students, including two resident assistants, re- presenting two different houses which were not to be in- cluded in this study. Their impressions and suggestions were received and final modifications were made. By way Of overview the HAS may be subdivided into several parts. The first 14 items provide demo- graphic information about the respondents including length of residence in their house. On items 15 through 34 a nine-point scale is used by the respondents to rate ‘. their house or hall on several dimensions of house life 92 such as "participation in intramural sports" and "Abil- ity to study in the house." Respondents are requested to rank order items 35 through 44, each of which describes a house activity,in the order of the degree of concern which the house has shown for each activity. On items 45 through 54 the same ten activities are reranked in the order of what the respondent would most prefer to be the most important activities of the group. Questions 55 to 58 treat the climate of learn- ing of the house and residence hall. These questions are preceded by an operational definition of the climate of learning which might exist within a house. Respon- dents rate their house climate and respond to other re- lated questions. In Part II of the HAS, students respond on a five-point scale to 63 questions which assess many di- nensions of house or hall life. On a few questions, respondents indicate their attitudes toward either per- sonal items or characteristics of the University commun- .ity. The student subculture descriptive paragraphs fol- low and students are asked to respond to four questions about these. Two open-ended questions provide socio- Ruetric data. These are followed by a final question on which students can react to the questionnaire itself and/ (>1: discuss other dimensions of house or hall life inade- cIllately treated in the HAS. 93 No attempt was made to apply traditional relia- bility formulas to the questionnaire or to obtain a re- test measure Of reliability. It is questionnable whether or not appropriate or adaptable reliability measures are available that could be applied to the HAS in toto. A general measure of reliability is implied in the analysis of the data. As was indicated earlier, respon- dents serve as reporters of house characteristics for the majority of the items, in the questionnaire. In this regard each item is independent; to the extent that residents agree on the item one can then infer that the house characteristic assessed has been reliably measured. To the extent that residents disagree the reliability of their aggregated responses diminishes. Thus reliability is an inverse function of the standard deviations of house means on the items. Horst has developed a somewhat infrequently used "generalized expression for the reliability of measures."27 It was initially conceived for use in situations where :several measures Of the same function are obtained for each of a group of persons such as when individuals are rated by several judges. Horst states in his summary: 27Paul Horst, "A Generalized Expression for the geliability of Measures," Psychometrika, 14 (1949), 21- l O 94 In general the number and source Of the measures may vary from one member of the group to another. We take the mean of the measures for each person as the best estimate of the function for that per- son. The conventional formulas can not be used to determine the reliability of a set of means so ob- tained. A formula is develOped which provides a unique estimate of the reliability of such a set of means. The formula is more general than some of the well-known reliability formulas, so that these formulas are shown to be special cases of the more general formula.28 His formula was applied to the HAS items where appro- priate (demographic and attitude measures were omitted). In justifying the application the analogy is made be- tween Obtaining several measures of the same variable for an individual and obtaining several measures of a variable for a house. The formula is based on the well-known general- ized formula for the reliability coefficient 2 ‘nhere oezis the error variance and Go is the observed 29 ‘variance of the measures." The observed variance is ‘the squared standard deviation of the means for a given item. The resulting formula is 281bid., p. 21. 29Ibid., p. 24. 95 where N = the number of houses (27) = the number of measures (reporters) for house i, = the mean of these measures for house i, 1 Ci = the standard deviation of these measures for house i, and o = the standard deviation of the means for M the N houses.30 Resulting Horst r's are reported in Appendix E Items assessing house characteristics with low r's were excluded from the multiple discriminant analysis de- scribed in the following chapter. Summary In this chapter the five instruments used in the <:ollection of data have been reviewed. These include 'the College Qualifications Tests and the Michigan State Chliversity Reading Test, measures of academic aptitude administered to students new to the University when they enroll. The Attitude Inventory used in the study con- sists of four scales from the Omnibus Personality \ 3°Ibid., p. 23. 96 Inventory, "Thinking Introversion," "Theoretical Orien- tation," "Estheticism," and "Autonomy." A forth measure consists of descriptive paragraphs of the four student subcultures postulated by Clark and Trow and is used by students to indicate their orientation to college life. The last instrument is the House Analysis Survey (HAS) develOped to assess characteristics of house and resi- dence hall life across several dimensions. Available reliability and validity information for each instrument was presented. CHAPTER IV DESIGN OF THE STUDY Population and Sample During the 1964-65 academic year (when the data were collected for this study), Michigan State University Operated fourteen relatively large undergraduate residence halls for men grouped in five locations or complexes on campus.1 These halls differed somewhat from one another in age, architectural style and program (e.g., "living- learningf'coeducational). The residents of these four- teen halls (7,370 men, winter quarter, 1965) constituted the pOpulation from which the sample was drawn. Each Of 'the halls was subdivided into eight to twelve houses de- pending on its size and physical characteristics, for a total Of 142 houses. From 40 to 80 students lived in a lmouse. The mean number of students in a house during the Exariod of the study was 52. Owen Hall, a large residence for graduate stu- dénts is not included in these figures nor was it con- sidered for inclusion in the study. 97 98 Residents returning for successive quarters had priority in selecting the room, house, and hall in which they desired to live. Freshmen and other new students generally were randomly assigned to hall and house, though they may have specified a preference with no assurance that it would be honored. Roommate requests were honored as far as possible, though preferences had to be mutual between parties. Unmarried freshmen living beyond com- muting distance were required tO live in University resi- dence halls or in one of the limited number of spaces available in approved off-campus housing. Freshmen were commonly assigned to live with other freshmen, though with some frequency they were also assigned to live with SOpho- mores and upper classmen, depending on roommate preferences and the pattern and number of specific room requests from returning students. It was also a common practice to liouse, temporarily at least, three students in many of 1:he larger rooms generally designed for two. This was a :Eumction both of lack of available space for swelling en— znollments and an attempt to maintain a high level of occu- Pancy throughout the year. With normal attrition during the year students so housed could request to move into Vacancies occurring elsewhere. As indicated earlier a resident assistant was assigned to each house and was re8ponsible to the head resident advisor of the hall. 99 Student government organizations within both the house and the hall generally were similar in all houses and halls on campus. Of the fourteen on-campus halls, nine were selected to be included in the study. ‘The selection was not arbi- trary; rather, representative halls from each of the five complexes were included. Care was also taken to pair halls within a given complex, with the exception of one hall where another hall similar in those characteristics used for pairing did not exist. Thus, in studying inter-house differences the design of the research also allowed com- parisons to be made between (1) halls differing in archi- tectural and program characteristics, and (2) the complexes in-which they were located. Pairing the halls facilitated double-checking correlates of any inter-hall similarities noted. Descriptions of each hall selected for the study :Eollow in Table 4.1. It may be noted from the table that lialls one to four represented the relatively new living- Jxearning, coeducational concept, halls five and six were traditional men's halls, and halls seven through nine rep- resented two additional architectural hall concepts. It would have been prohibitive to study every holise within each of these nine halls. Thus, three houses wa-‘i-vtzhin each hall were randomly selected by lot prior to “a .IUI 100 condos muasomm Ommaaoo wuflmu0>aco Hmuo>om “mEoou no name come no mucmoflmmu he UOHMbm such mcflcH0m©m nufl3 mEOOH cmfilo3u,cfl O>HH xmameoo mucoowmmu unmoxm m can H made: on HMHHEHm mmma momemo adamowmmbm madmuocmm .Hmcowpmosomoo.mmflnlflm Hamm .3em mmm m .Hamn can ca O>HH Op ownflswmu no: nab wuwHOflHm cm>flm mumuumq ocm muud ca muoflmfi «mono pub» EOHM conflmmo huasomm ammo u>mm new unosmu mommmao «mumuumq one mUH4 mo xmameoo ommaaou co mandamEm ummoxm muommmmu umoE cw H vmma momfimu .02 same on umaflsam «coaumummo no new» amuse Hams .m.m mam N .Hamc may ca om>HH mHOnmE mocmflom HMfiOOm How mnemflbom OHEmomom mm mcfl>nmm mucmooum mumspmum «mmocmflom Hafioom mo mmoaaou can Eoum oOOflmmo avasomm can an ubmsmu mommmao Hmum>mm «Hams cw m>wa on omnwswmu no: usn moaoummoum co>wm mocmflom HMHOOm ca mcHHOnmE mucoosum “such ocm Soon mosum .msoonomn N no ocmomeoo mmuflsm cmE w "cmflmoo Econ «Human Ham: cocooflmmu ocm .mEooummmao 0cm movemmo muasomw .mmfluflafiomu coaumcuown ocm mcflcao mcflmcon comm mmomusmlwuase muoumam momma an xmameoo m.cmeo3 Eoum omumummmm mcfl3 m.cwE .Hmcowu vmma mcmfimo Imocoooo .mmfluufln «coaumnmmo no name umuflm Hash .m.m mam H omumameoo mucooflmmm .oz coaumflnommn Hmmw coaumooq mo HOQEOZ Ham: mama .nmuumso gauges .mHQEmm -500 maflmg mocwcflmma earmamehgs spasm cassava: one noummmmu mnu mzwmahm a mo mOflumflumuomnmco H.v manna 101 .omson sumo CH topmooa numb mcmm omHmH oco “mEOOH cmEIooHfi oEom I03“ tsunami: buxoamEoo onp ca Hoc Ham; ogy ca unosmu uoc mommmao “moooo omson muflcn xflm on» no OBu momamo nmsoau pmoo pop me Ham: on» «mean: xam Ham .z.z mo>Hom wuflaflomm cowumouoou cam mcwcwo Hmuucoo vmma mdouw omHmH.m «onmEoo m CH mmcfloafisb xfim mo oco Hamm moonm New > .mOHumeouom momemu mo numso Hmoammsm Ham c« m .02 Hams on umHHeHm mama noucmo .m .02 Hamm ca oobauomoo mmcwz 03¢ mo ocooom HHmm Hmoz ohm m .Houom Inmno CH Hmcowuflomuu ounuoouflbonm «Ham: on» ad omson nomo mcfl>uom snob =mcmm= omhma oco bufl3 mumEm o mEoon cmE oonnu oEOm .o3u madame “Ham: osu ca 0 m unmsmu no: mommmao «Hmcoaumosoooo no: “Hams mvma Houcou mafiabfimu um£3oEOm omuma m mo mosa3 03» mo oco Hamm Hmoz Hem m onQEOU .GOflumHomo mo Hmom ocooom one moma msmEmU ca mane “s .02 dawn on Hmomucmpa mflamuocmo Hams .z.m emm v m.mnmEmo co momnooo omonu oxmu Ho onmEoo onu ca onob3>cm mGOHuoom omnsoo omoaaoo muflmno> lac: ca adouco mmE xoameoo oz» mo mucoowmou “xoamEoo on» ca maams mons» Ham mo moauwaflomm omomnsmuwuade on» cw unmsmu mommmao pom wouoamsoo mucopfimom .oz cowumauomoo “mow cowumooq mo Hobfidz Hamm pascaucoo H.¢ passe 102 .momodaoo oocooflmou Hmucoamuomxo m.muflmuo>flco oumum cmmHsOHz mo umuflm on» oEmoob ocm oouuo>coo mm3 m .02 Ham: .Hmom mcH3OHHom on» moanso omafio .xoamfioo msmEmU .m.m onu CH umnu on HmHHEfim ooumfluflcfi mm3 Emnmoum omoaaou huflmuo>flco m ocm ooaoooeoH mms huflawomm Hmuucoo on“ lawsum mean How coauooaaoo mvmo onu mcflsoaaom Hmoh anoomom on» ocausan .oocowom Hmfloom cam oocoflom amusumz .mofluflcmesm .ommsmcmq ocm usmcosa cmOHHoE< ono3 momusoo moons .moHOEO£QOm 0cm coE Inmonm Ham How confisvon oHo3 .mosum onu mo oEHu onu um .30Hn3 oomnooo oHoo Room mo umamcoo omons .unmomu oum momucoo coaumosoo Hmuocom :OHcB ca muflmuo>flco oumum cmm Iwnoaz mo cOHmH>Hob5m OHEoomom ocm o>flumuumwcHEUm on» ma omoaaoo muwmuo>flcom .omaos nomo ca ooumooa numb ocmm omHmH oso “mEOOH cmEIooHnu oEOm Iozu hacfime «mcfl upaasa umaufim ca mmapmaaomm mascap an new mes mucoocum concrp m .02 Hamm Eonm ooumummom adamowmmnm mofluaaflomm comumouoou ocm ocflcflo xmameoo use ucoomflom Ham: m.coeo3 oxaamn cm unmsmu $MMM 5.Mwm mmv m uoc mommmao «ousuoouflnoum :HmGOAuaomHB= momamu pmma .3.z msouo .5 .oz aamm Mom mm oEmm Hams mo cowumfluomoo Hamm wooum How m coflumflnomoo Uouoameoo cowumooq mucoofimom .oz Hmow mo Honfisz Ham: pascaueoo H.s magma 103 time beginning of the 1964-65 academic year. The resulting sample thus consisted of 27 houses.2 P re-Test In order to secure OOOperation for the project the author, at the invitation of the Director of Resident Hall Programs, met with head residents to explain the project during their workshop prior to the beginning of the aca- demic year. Resident assistants of the 27 houses in- volved were personally contacted, and the project ex- plained to them as an attempt to study general character- istics of hall life (the explanation was quite general in order to minimize contamination); their cooperation was requested. An explanatory letter was sent to all Jflesidents of the 27 houses during the first week of school inviting them to participate and giving them the date of testing to be held the following week. Though the primary concern was to obtain freshman responses on the four OPI scales of the Attitude Inventory (AI) and 0n the Subculture descriptions, all residents were asked to participate. This was done (1) in order for the pur- POses of the test to be less Obvious and (2) in 2Two houses in as many halls were deleted from the study and two others from the same halls put in their Place to avoid conflicting with the data collection for another research project. 104 consideration of the possibility of using the additional data in an expanded analysis as part of the Office of Iaxraluation Services' series of residence hall research projects. The Al was administered the same evening in a central location in each of the nine halls. After the session resident assistants were given testing materials for distribution to residents who were not in attendance. In the follow-up residents returned the materials by mail. Table 4.2 describes, in part, the resulting sam- pie.3 Though not indicated in the table, of the 1573 residents of the 27 houses, 1456 (92.5%) completed the AI and subculture descriptions. As may be seen from the table, 733 or 47% of the residents of the 27 houses were first quarter freshmen. Six hundred and ninety-seven of these freshmen (95%) participated. Bbst-Test The purpose of the post-test was two-fold: (1) to obtain a measure of house life from the total sample as assessed by the House Analysis Survey (HAS) and (2) to 3Table 4.2 introduces a two-digit coding system uSed in designating the 27 houses in the study. The first Eligit (1 through 9) indicates in which of the nine halls 1n the study a given house is located. The second digit (1 through 3) differentiates between the three houses within a given hall. 105 .me mm3 Ummfi mflmmfl w¢m MO HQQEUC HMSfiUM 058 .UOUHMUmHU mHm3 mUmMH M>H03B¥ MHM Moe MMM vM wNv MMN MOM eMMM HMeH MNMH mHmuou 0M e M H MH M MM 0N MM MM MM oM NH oN M NV MN «M MN vM MM NM MM M NH N MM MH NM MM NM NM HM NM «H MM M «M Me He MN HM NM MM NM MH HN M MM eN NM NM NM MM NM MM 0N NM v HM MM MM Nv NM HN HM ov NH Me M MM Me Nv om vM HN MN MN M MN N Mv NM MN vH oM NM NN MM 9H MN N NM OM HM HM HM MM HN MN MN NM H Me MM HN Mv MM MM MM oM N MN 0 MM MN MM «M HM vM NM MM M MH H MM MH NM NN Nv Nv HM 0N wH oN M MM MN wN MV HM MM MM HM MH MH H NN NH MM Mv MM MM NM MN M HM M Mv NM HM vM NM MN HM MM MH NH 0 NM NH NN MM HM «M Me NM NH om o OM OM NM MN om om Nv NM M M H 0N OH HM MM Mw Mv He HM vH MN M NM MN MM MN Mv 0M MM HN NN HM 0 NM HM 0M HM NM QM NM NN HN MN M oM NM HN MM HM MM HM NM «H HN M ev VN vM MN mv em MN NN MH MH H NM MH NN MM Mv HM NN NM M vH H MN MH NM HM 0M HM HN NN MM Mv N 0M Me MM NM VM MM MH MN MN MM N NM MM MN Nv MM NM NH MMM MN MN v MHM MM MNN MM 0M vM HH omm .Huo .Huz omnos cH amounumom muqopHmoH .Huo HHmm poumop .Huo.uuz .Huo .mp0 nucoouom umouuuoom mchHmEou onomob Hmuou omm omsom cH ommucoo poumou .Huz HHmm GH .02 Honasz po>OE .oz lucoouom smoum.oz Inom Honfisz omnom onEmm ucoon m cofinmoum mucoonom mo Honfidz Hmuoa .MMMH .Hounmdo HoucHB .uoouluoom mcHHsp ooumou ucoo Mom Mam HOQESG Dam .mucmvflmmn MO HOQESG vGHUDHOGH emHHmn mocmfiflmmh kuHmH®>HGD mumum cmmH£OHS QGHG GHfiUH3 momfion NN CH mmHmEmm Gmfinmwuu Ucm Hmuou MO GOHHQHHOmOQ N.v OHQMB 106 obtain a post-test of intellectual attitudes or disposi- tion of freshmen in the houses as measured by the four omnibus Personality Inventory (OPI) scales of the Atti- tude Inventory (AI). The instruments were administered during a two-week period beginning in late February through early March. The assumption was made in admin- is tering the HAS at this time that patterns of interac- tion between residents as well as group norms were well established and would vary but little between then and the end of the academic year. This was in accordance with observations expressed by Newcomb.4 By this time tflne residents had lived together for over five months, apparently a sufficient time for the "acquaintance process" to have progressed adequately. It may be argued that several factors could have intervened either just prior to or after the post-testing Within any of the houses which could have dramatically affected residents' perceptions of house life. For in- Stance, in the small but possibly significant turnover of residents between the fall and winter quarters, highly in- fluential residents could have been lost or gained which, 4Theodore M. Newcomb, "The Prediction of Inter- Personal Attraction," American Psycholo ist, 11: 582, 1956; See also Eugene Jacobson, TThe Growth O Groups in a Voluntary Organization," Journal of Social Issues, 12: 18-23, 1956. Jacobson argues the merits of cross-sec— tlional studies of large numbers of groups as a means of studying their patterns of growth and characteristics. 107 theoretically, at least, could have shifted the focus of house life substantially. Or a house could have been very successful or unsuccessful in a particular endeavor, the influence Of which would not be directly available for observation through the HAS. Hence, there is no pretense in the study that such contingencies were controlled. All other things being equal the assumption was made as indicated that house life had generally stabilized. Prior to the second testing a letter was again sent to all residents soliciting their interest and par- ticipation. Individual meetings were held with head resi- dents and resident assistants. On this occasion meetings were also held with as many house Officers as possible in addition to the resident assistants. Residents were ad- vised that their own AI scores would be made available to them after the data were collected and, likewise, that group means on the HAS items would be given to the resi- dent advisor and/or house Officers early spring quarter. The testing session was held on two consecutive evenings in central locations in each of the nine halls. Initial response was comparatively poor and as a result resident assistants and house Officers were asked to as- sist in securing more returns during the two week period following. Additional letters were mailed to those who did not participate and personal contacts were made. Referring again to Table 4.2, Of the 1481 winter quarter 108 residents in the 27 houses only 60% finally completed 1:11e questionnaires. Twelve of the 896 HAS's were unusable leaving 884. The participants completed both the AI and HAS, though on the AI only scores of freshmen were of immediate interest. 3 Prior to the post-testing session but subsequent to the initial test, 64 or slightly under 9% of the fresh- men living in the houses moved elsewhere or withdrew from tflne University. Thus, 669 remained and of these, 406 com- pleted the AI a second time,in addition to the HAS. This represented 61% Of the potential freshman sample.5 Grade point data for both fall and winter quarters for all 669 freshmen residents remaining in the houses both quarters was available for testing several of the hypotheses. Of these 32 (slightly less than 5%) had not completed the Al or the measure of College Types fall quarter. Thus, data for testing hypotheses relating to the interaction Of academic performance, place of resi- dence, and college type was available for 637 freshmen. 5It is interesting to note that participation of freshmen in the post-test was at about the same level as that of older students. Perhaps this is in part of re- fILectioneof the Operation of peer group influences. 109 The samples may be summarized as follows: Number of residence halls studied . . . . . . 9 Number of houses studied .'. . . . . . . . . . 27 Total number of residents completing usable House Analysis Surveys winter quarter 0 O O I I O O O O O O O O O O O O O 884 Number of freshman residents living in the 27 houses both fall and winter quarters O .0 O O O O O O O O I O O O O I O O 669 Number of freshman residents who completed the College Types measure falquuarter (MSU Reading, CQT, grade point data available) I O I O O I O O I O O O I O O O I 637 Number of fall and winter quarter fresh- man residents who completed the AI both quarters and the College Types measure fall quarter . . . . . . . . . . . . . . . . 406 One may speculate as to reasons behind the com- };aratively poor participation in the second test period. £3tudents were relatively free of distractions and alle- §;iances during the pre-testing early fall quarter. Sev- eeral conflicts were evident winter quarter that had not tween nearly as pronounced earlier, such as intra-mural auid other social activities and academic pressures. The ghost-testing period was held somewhat late in the quarter 1J3 order for the groups to stabilize as much as possible from resident-turnover between quarters. But this may have placed the session unfortunately too close to finals. NC31: to be overlooked is the fact that participation was VOluntary; hence, many students undoubtedly simply ignored the requests . 110 Of concern was the fact that in several houses the percentage completing the HAS was disappointingly small. After consultation with the thesis committee and members of the Evaluation Services staff, it was decided to proceed with the analysis using data from all 27 groups. It was felt that the study was exploratory in nature and that since students serve as reporters of group life in com- pleting the HAS, their responses would nevertheless give some picture (albeit possibly biased) of house life. The very fact that participation did vary rather dramatically was in and of itself of interest in the study. It raised several questions as to what characteristics of house life might give rise to 86% participation from one house and only 23% in another. In order to determine the nature of any differences between pre-tested freshmen who participated in the post- test and those who did not participatF in the post-test, several t-test comparisons were made.’ Mean scores of the two groups were compared on the fouriscales of the AI, on the MSU Reading Test, and on the CQT total. Accumulative mean grade-point-averages for the fall and winter quarters of the two groups were also compared. Results are pre- sented in Table 4.3. As may be noted in Table 4.3 freshmen who particie pated in both the pre- and post-testing sessions had a significantly higher CQT total mean score and a 111 Table 4.3 Comparisons between the means of pre-tested freshmen who did and those who did not par- ticipate in the post-testing on the four AI scales, MSU Reading, CQT total, and accumula- tive fall and winter grade-point-average, 1964-65 (N=406) (N=23l) Pre- and Post-test Pre-test Only Variable Part1c1pants Part1c1pants E Mean S.D. Mean S.D. TI 31.6 9.84 30.5 8.68 1.33 TO 18.0 5.18 18.0 5.08 .01 Es 9.6 4.62 9.0 4.27 1.63 Au 22.9 6.36 22.8 6.21 .33 MSU Reading 31.7 7.85 30.7 6.61 1.70 CQT total 142.7 27.23 135.0 23.10 3.64* 2 Qtr. GPA 2.43 .73 2.26 .66 2.93** *Significant beyond the .001 level. **Significant between the .01 and .001 level. significantly higher mean grade-point-average over the period Of the study. NO differences were noted on the four AI scales or the MSU Reading test. Nevertheless, several findings of this study (presented in Chapter V) had to be cautiously interpreted in light of the dif- ferences between the two groups. 112 The distributions of college subculture orienta- tions in the two groups were also compared. Table 4.4 A comparison of pre-test subculture orientation of freshmen who did and did not participate in the post-testing Subculture Type* Group Total I II III IV Pre- and Post-test N 132 71 166 37 406 Participants % (33) (17) (41) (9) (100) Pre-test only N 82 22 91 36 231 Participants % (36) (9) (39) (16) (100) Total 214 93 257 73 637 % (34) (15) (40) (11) (100) 2 *Subculture Type I, Vocational;]1,'Nonconformist; III, Academic; IV, Collegiate. It is evident from Table 4.4 that students who failed to participate in the winter quarter testing dif- fered from those participating during both sessions in the distribution of college subculture orientations. The differences were significant beyond the .01 level in a chi—square analysis. Comparing the proportion of students in each subculture in the total sample (N=637) with those participating only in the pre-test (N=23l), it would ap- pear that the pre-test only group was more heavily weighted 113 toward the collegiate and vocational orientations and had proportionately fewer non-conformists and academics. The group participating in both test sessions was more weighted toward the non—conformist and academic with fewer vocationals and collegiates.6 Hypotheses and Methods Of Analysis The hypotheses under consideration and the statis- tical treatments which were employed in their.analyses are presented_jointly in the following paragraphs. Thus, some confusion may be avoided in relating each hypothesis to its appropriate method of analysis. Hypothesis I: It will not be possible to discriminate among the several residence hall houses on the basis of linear combinations of variables describing their group char- acteristics. If, however, it were possible to discriminate be- tween the houses, several attendant questions were to be considered as follows: Question A: What are the interpretations of the linear combinations of variables (dis- criminant functions) which may result? 61n terms of the Clark-Trow theory underlying the four subcultures the pre-test only group contained more students tending away from intellectual activities while the post—test group contained more tending toward such an identification. 114 Question B: Is it possible to develop a meaningful typology of the residence hall groups according to the description of the discriminant functions and the location of the groups in multi-dimensional space? Question C: Does the location Of houses in the em- pirically derived multi-dimensional space relate in a meaningful fashion to the classification of houses according to program and/or physical characteris- tics (such as common location in the same hall and/or complex or construction similarities such as living-learning features)? As was discussed in Chapter II, the factor analytic approach described by Selvin and Hagstrom for classifying groups played an important role in the development of this study.7 The HAS was initially designed as a means to ob-- tain "aggregated" measures of group characteristics. Sub- sequently the Selvin-Hagstrom method was discarded in light of criticisms directed toward their procedures and the suggestion that multiple discriminant analysis could be appropriately incorporated into the design as a means of differentiating across multiplevariables among the several groups in the study. Multiple discriminant analy- sis was employed to test Hypothesis I and to provide the data directed toward answering Questions A, B, and C. 7See pp. 46-52 in Chapter II. 115 Multiple Discriminant Analysis Multiple Discriminant Analysis is a statistical procedure for maximizing the ratio of the variability be- "tween groups to the variability within groups across sev- eral variables simultaneously. Fisher first developed the concept Of the discriminant function as a means of classi- fying an observation into one of two or more groups.8 Rao furthered the technique in his studies of twelve Indian castes. Among other contributions he developed a method for representing the centroids of groups in a space having fewer dimensions than the original number of variables.9 Subsequently Bryan published a computational routine for obtaining the latent roots and vectors of the matrices involved.10 Rettig has succinctly summarized the mathematical properties of multiple discriminant analysis as follows: 8Ronald A. Fisher, "The Use of Multiple Measure- ments in Taxonomic Problems," Annals of Eugenics, 7:179- 188, 1936; Ronald A. Fisher, "The Statistical Utilization of Multiple Measurements," Annals of Eugenics, 8:376-386, 1938. 9Radhakrishna C. Rao, "The Utilization of Multiple Measurements in Problems of Biological Classification," Journal of the Royal Statistical Society, Series B, 10: , Radhakrishna C. Rao, Advanced Statistical Methods in BiOmetric Research (New York: John Wiley and Sons, 1952). 10Joseph G. Bryan, "The Generalized Discriminant Function: Mathematical Foundation and Computational Routine," Harvard Educational Review, 21:90-95, 1951. 116 The multiple discriminant function is based on two different covariation matrices. One matrix B consists of the covariation among the group judgments across the variates. That is, the mean judgments on a given variate by each group are subtracted from the overall mean judgments of the combined groups. This deviation score is obtained for each variate and the cross-products of these deviation scores constitute the B matrix. The second or error matrix W consists Of the cross-products of the deviations of the indi- vidual judgments from the group means across all vari- ates. The discriminant function seeks to determine the latent root A and characteristic vector 3 which maximizes the ratio of the two matrices BW‘l. The latent root is found by subtracting a constant A from the main diagonals of the matrix BW"l so that the determinant of the resulting matrix equals zero. We then solve the equation 3 (BW‘1 - AI) = 0, where I is an identity matrix and E is a vector which, when mul- tiplied by the root 1, equals the product of the vec- tor with the matrix. This characteristic vector 3 is the canonical variate and consists of the weights for the variables which linearly maximize the differ- ences between the groups. Each root extracted is directly proportionate to the amount of between-group variance accounted for by the canonical variate. The proportion of explained variance among the groups is obtained by the ratio 11/21, where A is the root of interest and 21 is the sum of all the non-zero roots. Following the extraction of a root, it is possible to determine the amount of residual variance still to be accounted for. The latter is approximately distrib- uted as a chi-square and can thus be used to determine whether further dimensions should be Obtained. The degrees of freedom for the chi-square are Obtained by the equation df = (pen)(k-n-1), where p = the number of variates; n = the number of dimensions extracted, and k = the number of groups. Each dimension thus extracted corresponds to a distinct root: all are un- correlated or perpendicular to each other. The dimensions thus obtained, however, do not re- flect the relative contribution of each to the vari- ance among the groups. To Obtain the latter, it is necessary to scale each vector by dividing each of its weights by the within-group standard deviation of the vector. This standard deviation is obtained by multiplying each weight by the corresponding element in the W matrix. The result of this scaling process is that the between-group variance of the vector is equal to its latent root 1. Thus, the scaled dimension 117 is now in accord with its relative contribution to the total between-group variance, and its weights reflect the relative contribution of each variate to the di- mension. Following the scaling of each dimension, a com- posite mean score for each group may be obtained by computing the cumulative product of the weights and original means of the group, across the p variates. These composite means are computed for each group on each dimension. Summing the squared differences be- tween these means for any two groups across all di- mensions produces the intergroup distance, or the Mahalanobis D2, the square root of which represents a measure of Euclidean distance in n-dimensional space.11 The number of discriminant functions or resulting roots from the matrix solution is the lesser of p and k-l, though as Rettig indicated not all will necessarily be significant. It is important to note that in obtain- ing the roots, each successive discriminate function maxi- mizes the variance among groups after the influence of the previous function(s) has been removed...Each solution. is orthogonal (perpendicular) to previously extracted solu- tions in the multi—dimensional space. This may have a cer— tain confounding effect in interpreting the functions, how- ever, particularly in an analysis with many variables of 11Salomon Rettig, "Multiple Discriminant Analysis: An Illustration," American Sociological Review, 29:398-402, 1964; The reader may also wish to refer to Maurice M. Tatsuoka and David V. Tiedeman, "Discriminant Analysis," Review of Educational Research, 24:402-420, 1954, for a comprehensive overview of the historical development of the statistic. 118 relatively low reliability and no a priori distinction be- tween groups, such as the one herein reported. As has been pointed out by Jones and Bock, the resulting func- tions are not necessarily "pure" clusters of variables. 12 Rather "in terms of within-sample factor structure." they are likely to be "complex." Important constella- tions of variates describing significant dimensions Of group life in reality may not necessarily be orthogonal to one another. Hence the condition of perpendicularity in discriminant analysis contributes to a somewhat arti- ficial character of the results.13 From its conception, discriminant analysis has been used for classifying observations into groups as is well illustrated in the important study by Tiedeman, Bryan, and Rulon.l4 Several authors have, however, de- scribed the utility Of the statistic as trifold: "(a) the establishment Of significant group-differences, (b) the study and 'explanation' of these differences, and 12Lyle V. Jones and R. Darrell Bock, "Multiple Discriminant Analysis Applied to 'Ways to Live' Ratings from Six Cultural Groups," Sociometry, 23:162-176, 1960, p. 172. l31bid., p. 175. 14David V. Tiedeman, Joseph G. Bryan and Phillip J. Rulon, The Utility of the Airman Classification Bat- tery_fpr Assignment of Airmen to E1 ht Air Forge Special- ties (Cambridge, Mass.: EducatIOna ResearEh Corporation, 119 finally (c) the utilization of multi-variate information from the samples studied in classifying a future individ- ual known to belong to one of the groups represented."15 Despite the emphasis on the "classification" or "place- ment" function, at least one author has indicated his feeling that the statistic"is more useful in understand- ipg the major differences between groups than it is for 16 The focus of this placing individuals in groups." study was on the understanding and explanation of any group differences found to exist. Thomas17 has developed a program utilizing the CDC 3600 computer to extract the latent roots in multiple discriminant analysis following the general procedures outlined by Bryan.18 In addition to the latent roots of the equation, the program also provides group means on the variables, an intercorrelation matrix for both indi- vidual groups and the total sample, the per cent of vari- ance accounted for by each function, a chi-square value (as developed by Rao) for testing the significance of 15Tatsuoka and Tiedeman, o . cit., p. 414; Jum C. Nunnally, Psychometric Theory (New York: McGraw-Hill, 1967), p. 388. 16 Ibid., p. 400. 17Computer Institute for Social Science Research, Michigan State University, "DISCRIMZ," Technical Report No. 33, by Stuart Thomas, July 5, 1968. 18Bryan, loc. cit. 120 0 solutions, degrees of freedom, centroid scores for each group on the functions, vector weights,scaled (standard- ized) vector weights, and an intercentroid distance matrix. The program was used to Obtain the linear com- binations of variates and other statistics from the HAS data necessary in order to test Hypothesis I and provide answers to Questions A-C. Hypothesis IIa: Hypothesis IIb: There will be no differences in the (adjusted) academic performance of freshmen differentiated according to the types of houses in which they live (types defined according to clustering of houses along the discriminant func- tions and/or in the multi-dimensional function space). Nor will there be an interaction between types of houses and the pre-test sub- culture orientation Of the residents. There will be no differences on any of the four post-test (adjusted) measures of intellectual disposition of freshmen differentiated according to the types of houses in which they live. Nor will there be an interaction between types of houses and the pre-test subcul- ture orientation of the residents. Before this series of hypotheses could be tested it was necessary to determine the outcome of the test of Hypothesis I. If in fact differences among the groups were ascertained in the discriminant analysis, the group centroid scores (group means on the linear combinations of variables) could then be located on each significant 121 function and/or in multi-dimensional function space. Dif- ferences between groups of houses located according to their clustering on the discriminant functions could then be tested. Of concern was the differential influence of house life on freshmen. In order to assess differences in aca- demic performance, a series of two-factor analyses of co- variance were used. Accumulative grade-point-average of freshmen for the fall and winter quarters, 1964-65, ad- justed by CQT total and MSU Reading Test scores, was used as the measure of academic performance. The four OPI scale scores, used as the dependent variables in testing Hypothesis IIb, were each adjusted in the analyses of co- variance by pre-test scores on the same scale. The multi-factor analysis of covariance allows one to test simultaneously for differences on more than one adjusted main effect (independent variable) and for inter- actions that may be present between the main effects. In testing Hypothesis 11a, and IIb type of house was defined as one main effect and subculture orientation the other. Differences between subcultures were not of major concern in this study (Hodgkins and Adams had already reported differences in academic performance between subcultures).19 19Benjamin J. Hodgkins, "Student Subcultures--An Analysis of their Origins and Affects on Student Attitude and Value Change in Higher Education," (unpublished dis- sertation, Michigan State University, 1964); Donald V. Adams, "An Analysis of Student Subcultures at Michigan State University," (unpublished dissertation, Michigan State University, 1965). 122 However inclusion of subculture orientation as a main ef- fect provided the Opportunity to test for interactions be- tween orientations and types of houses. Thus, if signifi— cant interactions resulted, it would suggest that different types of houses affect one or more types of students one way and other types a different way. Hypothesis IIIa: Hypothesis IIIb: Hypothesis IIIc: Hypothesis IIId: Hypothesis IIIe: There will be no differences among the houses in residents' perceptions of their house climate of learning. There will be no relationship between the cohesion of the houses and the per- ceived climate of learning. There will be no relationship between mean grade-point-averages of the houses and the climate of learning. There will no differences in (adjusted) academic performance between freshmen residing in houses having a "high" cli- mate of learning and those living in houses with a "low" climate of learning. Nor will there be an interaction between the level of the climate of learning and students' subculture orientation. There will be no differences on any of the four (adjusted) post-test measures of intellectual disposition between freshmen residing in houses having a "high" climate Of learning and those living in houses with a "low" climate of learning. Nor will there be an interaction between the level Of the climate Of learning and students' subculture orientation. Before this series of hypotheses could be tested, it was necessary to determine the outcome of the test of 123 Hypothesis IIIa, analogous to the testing of the previous series Of hypotheses. A measure of residents' perceptions of their house climate of learning was included in Part I of the HAS (item 55). After having read the operational description Of a climate of learning, students were asked to respond on the nine-point scale to the following question, "Where would you rate the general 'climate of learning' of your house?" Hypothesis IIIa was tested by simple analysis of variance of group-mean scores on the item. The Operational defini- tion of the climate and the related questionnaire items are reproduced in Table 4.5. The measure of climate of learning was also one of the several HAS variables included in the multiple dis- criminant analysis. It was theoretically possible that on one or more of the linear combinations of HAS variables, the climate item would have high loadings, perhaps to the extent, in combination with other variates, of describing multi-dimensionally the climate of learning. A priori, however, this judgment could not be made and because of the intrinsic interest and significance of the item, it was subjected to independent analysis.20 20The point can be made that differences between groups on any or all of the HAS items may have existed, and, hence, each item could have been tested by analysis Of variance. However, the very real advantage of multiple discriminant analysis lies in the fact that the variables Table 4.5 The Operational definition of the climate of learning and related HAS items Questions 55-58 ‘ Clisate of Learning The following paragraphs describe what we will refer to as the "clinats of learning" of a house. Read the section carefully and then answer the questions at the and according to your appraisal of your house. . a a a O 0 0 0 a ‘ Sousa activities and attitudes on canpus vary in the degree which these support or conplsnent the nissIon of {He Univeriiiy of preparing students to understand and deal with the problens an n a of the world in which they live. Think of this degree of support as lying along a line, at one end groups of residents, perhaps entire houses, whose activities stron 1 en rt a climate of learning: at the other and, houses or subgroups of residents who are no? only unInvoIved in such a clients but who also strongly resist its influence. The descriptions to follow are not neant to inply that social life, athletics, and other activities conflict with a "clients of learniniT' Such progress say or say not operate effectively regardless of the clients. Also, students say legitinatsly feel that their life within the residence hall is their own to lead as they see fit and that "learning" is properly confined to the classroon and library. Here are descriptions: ”Sign" Clinnts of Learnigg . Visualise a group of residents or an entire house where the excite-ant of learning, experiencing and growing literall abounds. Here exists an hlnost continual excfiinge of ideas, attitudes, dis- cussions of arf for-s, new discoveries in science, political controversy, confrontation and~discusnion of values. "Bull sessions” are often deep and stisulating. Cultural activities, such as the Lecture- Conosrt Series and Provost Lectures, are strongly supported. Fresh-en in the house rapidly have their intellectual borisons broadened and stinulnted. Discussions of classroon topics continues sell beyond the walls of the cinssroon. - . 'Low' Clients of Learning At the other extrene, learning is generally left to the classroos. It is not that residents don't study outside of class or work for their grades. It is Just that little, if any, of the intellectual life of the university carries over into the life of the house. ”liTI‘iessions" seldoa have intellec- tual depth or substance. Attespts to stisulato sore enlightening activities are seldon supported, and one who does night be regarded as a "highbrow“ and out of touch with his house-ates. Such a house say be a satisfying place to live because other characteristics of the house or subgroup possess great value for the residents. Social, fraternal or athletic activities say be proninent. but it is alnost as though a social nors existed against too such involve-ant in acadsaic learning. Selection of classes is often based on the ease with which one can get by. Fresh-en soon learn the ways of the group and confors. Though they indicate concern over their studies, they are readily distracted fron than. Tbs '1nbstwssn” Cass Between these two extra-es one can visualize a third group or house whose activities and attitudes neither stron 1 support such learning experiences and intellectual excite-ant nor 0 as then with any coniIifincy. or suéb a group house life any seen to be inde endent of the "sission o; Ebe University.” however, our feeling is that subgroups or entire houses tend Io Iean sore one way than the other, though slsnents of both sides say exist in any given group at any given nosent. CLIIATI OI LEARNING ”Sign” ”Inbetween" "Low" 1 I 2 I 3 I 4 S S I 7 I S I 9 Strong Fairly lodsrats Tendency Inbetween Tendency loderats Fairly Strong and open Strong toward toward strong and open support for and/or involve-sat opposition to and/or lack of .in tbs Clinats of Learning involve-ant in the Clients of Learning ‘1, *9 notion. SS. Shore would you rate the general ”clients of learning' of your house? so. Ibat has been the level of the ”clients" which YOU’have personally experienced through those with when you associate the soot in the house, regardless of the general clients of the house? S7."bere would you personally like the level of the ”clients" to be in your house? SS. Ibors would you rate the general ”clients of learning” of the residence hall in which you live? 124 125 Hypothesis IIIb was tested by product-moment correlation analysis of the relationship between group mean scores on the HAS climate of learning item and on each of four HAS items operationally defined to be rough measures of cohesion. The distribution of the group mean scores of the items was assumed to be normal and the mean scores continuous over the population from which the sam— ple was drawn. The four items used were: Item 32, Pt. I: "Your level of satisfaction with living in this house" (rated on a nine-point scale) Item 9, Pt. II: "When it comes right down to it, I really have little allegiance to either my residence hall or my house" (response on a five-point scale). Item 47, Pt. II: "I would prefer to move to a dif- ferent house" (five-point scale). Item 59, Pt. II: "There are 8 to 12 houses in your residence hall. Where would you rate your house generally in con- trast to the other houses in the hall?" (five-point scale). An early intent of the study was to measure cohesion from sociometric data obtained in the final questions on the are not independent of one another, many of which are likely to be highly correlated. Thus, the discriminant analysis technique allows for the covariation of variables. And, in addition, as Tiedeman points out, "It may well be that only a small number of the variables with significant differences in means are contributing to discrimination among the groups while other variables which by themselves provide no means of discrimination may aid considerably when taken in conjunction with the rest" (David V. Tiedeman, "The Utility of the Discriminant Function in Psychological and Guidance Investigations," Harvard Educational Review, 21:74, 1951. 126 HAS. However, disappointing returns of the survey pre- cluded this possibility. Thus, in substituting these survey items no tight definition of cohesion was pretended. Rather the items were thought to provide an approximation of "the degree to which the members of a group desire to "21 The hypothesis was included in remain in the group. order to explore the theoretical conclusion of the dis- cussion in Chapter I that group cohesion and the existence of a positive climate of learning are not necessarily re- lated.22 Hypothesis IIIc was also tested by a product- moment correlation analysis of the relationship between group means on the measure of house-climate of learning and the mean grade-point-average of the houses. This hypothesis tested the assumption that groups of students who in fact were functioning well academically would per- ceive a positive climate of learning and, likewise, groups of students performing less adequately would perceive a less positive climate. Hypotheses IIId and IIIe were tested using the two—factor analysis of covariance model described pre— viously in this Chapter. Freshman accumulative grade- 21Dorwin Cartwright, "The Nature of Group Co- hesiveness,"Group Dynamics,h1Dorwin Cartwright and Alvin Zander, (eds.) INew York: Harper & Row, 1968), p. 91. 22See Chapter I, "Cohesion as a Property of House Life," pp. 26-29. 127 point-average for the fall and winter quarters, adjusted by their MSU Reading and CQT total scores, was the dependent variable in Hypothesis IIId. Post-test scores on each of the four OPI scales, adjusted by pre—test scores, were* the dependent variables for testing Hypothesis IIIe. Level of climate of learning in groups of houses was one main effect and college subculture orientation the other. The latter was again used in order to test for interaction between the two main effects. Of primary concern was whether or not a positive climate of learning within a~ house or group of houses would have a salutory affect on the academic performance of freshmen over and above what one would expect from their academic ability test scores. Likewise, would a negative climate adversely affect fresh- man performance? Testing the interaction between the level of the climate of learning and student subculture orientation provided an opportunity to determine if dif- ferent types of students were differentially affected by the climate in which they lived. The 5% (.05) level of confidence was specified in testing for the significance of results throughout the study. Summary Three houses in each of nine residence halls were randomly selected for study. The nine halls represented 128 the various types of accommodations and programs for men. at Michigan State University in 1964-65. Freshmen were generally randomly assigned to houses, though returning students could express room and hall preference. During the first weeks of the academic year an Attitude Inventory consisting of four Omnibus Personality Inventory scales and a measure of four student subcultures were administered to residents of the 27 houses. The four Attitude Inventory scales were used as measures of intel- lectual disposition. Well into the winter quarter resi- dents were again tested on the Attitude Inventory and on the House Analysis Survey, a measure of the characteris- tics of house life. In order to more fully understand the sample, com- parisons, using the t-test, were made between two groups of freshmen. Though both groups had resided in the houses during the two academic quarters between pre- and post- testing, one group had participated in both pre- and post- tests while the second group participated in only the pre- testing. No differences were found between the groups on the four pre-test scores of the measures of intellectual disposition nor on the MSU reading test. The group par- ticipating in both test sessions, however, had a signifi- cantly higher mean CQT total score and mean accumulative fall and winter grade-point-average. Differences were also noted between the groups, by a chi-square analysis, 129 in their distributions of subculture membership of the residents. Three sets of hypotheses were presented; the methodology used in testing each was discussed. The first hypothesis concerned whether or not the several houses could be differentiated on the basis of their group charac- teristics. Several questions were raised concerning the possible nature of any differences that might be disclosed. Multiple discriminant analysis, a statistical technique which maximizes the ratio of the variability between groups to the variability within groups was discussed at length as the method employed in testing the first hypothesis. The second set of hypotheses considered the pos- sible existence of differences in academic performance and in post-test measures of intellectual disposition be- tween freshmen living in different types of houses. Types of houses were defined according to the location of house centroid scores on statistically significant discriminant functions and/or in multi-dimensional function space. Through a two-factor analysis of covariance it was also possible to test for interactions between residence in different types of houses and the subculture orientation of the residents. The measure of academic performance ‘was adjusted by MSU Reading and CQT Test scores; the :measures of intellectual disposition were adjusted by pre-test.measures on the same scales. 130 The final set of hypotheses were used to study the climate of learning of the houses. Climate of learning was measured by residents' responses to a questionnaire item in which they were asked to rate the climate of their house. Simple analysis of variance of group mean scores was used to test for differences in climate between the houses. Product-moment correlation was used to study the relationship between climate of learning and (1) house- cohesion and (2) house grade-point—average. Differences in the adjusted academic performance and intellectual dis- position of freshmen living in houses with different levels of the climate of learning were considered. Two-factor analyses of covariance were again employed in order that the possible interaction between the level of climate of learning and student subculture orientation could be studied. The .05 level of confidence was adopted in testing the results. CHAPTER V RESULTS In this chapter the results of the statistical analyses of the data are presented. Each hypothesis is considered in turn and is followed by a discussion of the results of the statistical test of the hypothesis. Gen- eral observations from a consideration of responses to the House Analysis Survey (HAS) are also presented. Though these are only indirectly related to the hypoth- eses, they are of general interest in gaining a better understanding of the characteristics of residence hall life for men at Michigan State University. General Observations from the House Analysis Survey The mean scores of the 884 respondents on several of the HAS items provide a rough indication of overall charac- teristics of residence hall and house life. Results from two sections of Part I of the HAS are reported in Tables 5.1 and 5.2 respectively. In Table 5.1 total group means from the 884 residents responding to 20 HAS items are re- ported. Based on their "observations and opinions of con— ditions within" their houses, students were asked to rate 131 d”.i . Fl 132 ....n......|l.. Amhmv Emumoum HcflUOm may on.” ms.m mm.a mm.m an acaummaofluumm can now unommum .NN _ Amuse mmson we» vm.m ma.m mm.H mm.m mo Eonmonm HMHOOm can omaa anacom .Hm .opo .mowsmomoaflnm .mmmUH .mcSmmH Amhmv mo mcfitcmumumpcs Moo» on mmsoc mm.m mm.¢ mo.~ av.m on» cflnufl3 mmaa mo coflsunflnucoo .om Ammmv Ham: occupancy one oo.m mm.m mn.a Hm.v cacuHB omsoz can no coaucusmmm .ma 8mm: mm.m ma.m oo.m mm.v Hocuomou m>c£ m3 moefiu coom one .ma Amman mmson man an mermnmaonom H~.n mv.a oo.m HN.¢ no oOGmEHOMHom anmtmom mo Ho>mq .ha Ammmv na.m mm.H ma.m mm.v munomm Hmuofimnucfl ca nmooosm .wa Ammmv manomm amusecnucfl va.m H¢.m mm.a mm.v ca coflummHOHuncm can now uuommsm .ma . . «efizv coaumwnomoo EouH .oz «««mcco2 omdom mo omcmm a m ecmmz AH uummv am>nnm mflmzaqu mmsom EouH muHmHo>HcD cumum :cmflnowz um momsoa Hams mocmpwmmu n.coE hm canuw3 :wmaa mo mundane: om Eonm mucus mmson mo mouse was .wnoHMMfl>op unoccuum .mmHOOm new: H.m magma 133 Rummy AmvmumEEoou ma.s ss.m ma.~ He.m use» nuns :oHuommmHumm “sow .Hm Aommv mGOHumasmoH Hams mocotwmmn mH.m me.m ma.H ma.s guns muqmnnmwu no mocmaamsoo .om Aommv omson Hcasofluncm mm.m Hm.~ mm.H mm.e when an man>na no msflm> .am Amhmv mHHHm om.m oo.m oa.~ mm.m poms ou cmcfl>oum mofluficsunommo .mm Ammmv Hm.v Hm.m mn.a mm.m mmson on» canvas mmocflaocoflnm .SN Amway omsoc ecu mo mo.n mm.v mm.a en.m omea Hmnsuaso can HmouomaaoucH .mm Aammv mn.o om.m va.m HH.m mmsor one ca apoum ou muHHHnd .mm Amhmv mucoflmmo mmson va.m mm.m mm.H om.v ompooao on» no mflcmnocmoa one .vm . loamy unnumammn mm m ms.~ mH.N m~.s ucmcnmmm man no marmumomma was .mm more: under 0 case . . arazv coaumflnommo EouH .02 are m m o m «new: AH Dummy mo>n5m mammacad omsom EmuH Acmscflucoov H.m manna 134 .mEmuH can so momma omsos mm on» no women on» ucomonoH rump unocaaea .Eoufl or“ mom 2 one mcHHmBOH Sous cm>fim m ou Ucommou ou coafimw maachHmmmooo mucoocommwu Hcsofl>flocfl .Ho>m3oc nvmm mmz m4: mcfiuoamaoo z Hmuoaer .Auoom >Ho> u m can .Hoom u m .xcoz n h .xcoa cauufla m on o» mono» u m .auouomwmaumm u m .woom on ou women n v .6000 n m .coom aum> u m .ucoaamoxm u Hy mamom and o no conoom mus Emuw scone “ammo Hams mocmonmu was» an mcn>na nm.v u om.~ mm.a mm.m gufi3 coauommmflumm mo Ho>oa Hoom .wm Amway mmsoa was» an mca>aa om.m I mm.~ mm.a Hm.m cuss cofluocmmflumm mo Ho>oa H50» .mm Amhmv mGOHDM©OEEooom Ham: mocooflmon em.m I mm.~ mm.a ma.v nufl3 cofluommmflDMm Homecom snow .Nm . . «aazv coaumflnomoo EouH .oz «rename: omsom mo omcmm o m «can: AH Dummy mo>u5m mwmhamcm omsom EouH Aomscnucoov H.m magma 135 their houses on each of these items. Houses were rated from "Excellent" (1) through "Satisfactory" (5) to "Very Poor" (9). A comparative analysis of individual house mean scores on several of the items is the topic of a later section of this chapter. It is sufficient at this point to say that house means on many of the items varied ex- tensively. Items 27 and 31 received the highest ratings (in terms of being the most positive) of this set of items. Both are concerned with interpersonal relationships and WA would suggest that these tend to be positive within the houses. The next most positively rated items, numbers 33 and 34, suggest an overall general satisfaction with life in both the hall and in the house within the hall. The degree of satisfaction however does vary between the houses as can be noted from the range of house mean scores. The item receiving the lowest rating (in the sense of tending to be weak or poor) was item 26, relating to the "intellectual and cultural life of the house." Judging from the overall ratings it would appear that, in general, the intellectual dimension of life in the house is not a prime factor in determining general satisfaction. Item 20 also received a comparatively low overall rating, an item which also relates to the intellectual life of the house. In contrast, academic performance as rated on item 17 was generally considered "to be good." It would seem that within the house academic performance and the intel- lectual and cultural life of the house differ. Ediflgyu .3 ._ 136 The means of three items relating to the social life of the house also fell slightly below the "Satisfactory" level. Item number 25, "ability to study in the house," also received a relatively low rating, though like most of the others the range of house means on the item was broad. The leadership within the house from the resident assistant rfi and house officers was found to be better than "Satisfactory." It is significant to note that of all the activities of the houses rated in this set, the one receiving the high— est rating was 15, "Support for and participation in intra- mural sports." In reflecting on these results the fact that only 60 per cent of the residents responded to the HAS should be kept in mind. It should also be remembered that the fresh— men who did not complete the questionnaire tended to have lower CQT total scores and to have lower grade-point-averages ~than those who did complete the test. Though one can only speculate, it is reasonable to assume that the non- respondents would have tended to agree with the general ordering of the items, particularly noting the placement of the intramural item. Questions 35-44 of Part I of the HAS listed ten activities or problems which were thought to "require or invite the concern of the house group as a whole." Para- phrasing from the instructions from the section, the resi- dent was asked to rank the statements in the order of the 137 concern which his group had shown for each of the activi- ties. The rankings were to have been based on the obser- vations of the respondents; the activity considered to have been the most important was to have been assigned rank number one; the next most important, rank number two and so on. Respondents were then asked to re-rank the F1 items in the order or what they would most prefer to be the most important activities of the group. Unfortunately many of the respondents failed to rank one or more of the items. Their responses were consequently omitted in tabu- lating the data for Table 5.2. The items are presented in the table in the order of the mean rank initially assigned each item. Item 38, "Sport, intramurals" was considered to be the most important activity concerning the houses during the year, according to the mean rank assigned. Twenty- four of the 27 houses rated the item first or second (ac- cording to the mean ranks of the individuals houses). In contrast, of least concern to the houses was item 39, re- lating to the role of the house life in broadening the cultural and intellectual perspectives of the residents. These findings would seem to be consistent with the ratings assigned to similar items in the section previously dis- cussed in this chapter. Study conditions within the houses were considered to be relatively important (item 37), but only slightly more so than "arranging and participating in 138 Table 5.2 Mean ranks assigned to 10 house activities or prob- lems by residents. Residents were responding to the request to rank the 10 statements in the order (1) of the concern the respondent's house had shown for the item, and (2) of the respondent's personal re- ference of what should be the most important activi- ties of the house.* "House "Personal Item Activities and Problems of Concern" Preference" No.** Concern to the House Mean Mean Ranking*** Ranking*** 38. Sport, intramurals 2.71 4.37 37. Study conditions of the house 4.06 2.27 35. Arranging and participating in 4.16 4.27 social activities 40. Providing assistance for individ- 5.30 4.35 uals' problems (study, social, personal) 44. Discussion, enforcement and debate 6.10 6.79 of rules and regulations of the house, hall and university 36. Participation in and/or discussion 6.20 6.38 of student government (hall, AUSG, etc.) 42. Participation as a group or with 6.22 6.48 the Hall in special events, e.g. blood drives, sitting together at games, projects, etc. 41. Except for items ranked higher, the 6.26 8.48 men regard the house as little more than a place to sleep and eat. Activities ranked lower gen- erally do not concern the men. 43. Keeping the house clean and tidy 6.38 6.16 39. Arranging and participating in ac- 7.56 5.46 tivities to deepen residents' understanding of issues, phi- losophies, the arts, etc. If an in- *The mean rankings were based on an N = dividual responding to the HAS failed to rank one or more of the above items, his responses were not included in the tabulation of the means. **The items are listed in the order of the total group mean rank assigned the items. ***Rho between the two rankings equalled .59. 139 social activities" (item 35). It is significant to note that for many residents, item 41 (". . . the men regard the house as little more than a place to sleep. . .") re- flected the situation in their houses more than group con- cerns for either house sponsored intellectual activities or house neatness. There is a degree of incongruity evident between the two different rankings, one according to observations and the other according to personal preference. A corre- lation between the rank orders of the mean rankings was computed to determine the extent of the incongruity. The The resultant rho equalled .59, suggesting a moderate cor- relation between the ratings. Possibly the comparative rankings indicate the Operation of an underlying anti-intellectual peer-group norm, a result of which is a general inhibition of resi- dents' involvement in intellectual activities. This may occur even though the personal preferences of many would be to elevate intellectual involvement in the house. Peer norms would seem to be strongly supportive of intramural and social activities. Attainment of good grades may rep— resent a minimal compliance with the imposed norms of the larger social system--those of the University. House norms do seem to support adequate study conditions, perhaps in the pursuit of "grades" necessary to meet the minimal level of compliance demanded by the institution. But the norms 140 do not support, and perhaps even oppose, group behavior directed toward "intellectual" activities for the sheer sake of learning. Overall means, house means, and standard devia- tions for all items in the HAS appear in Appendix D. The Multiple Discriminant Analysis 3 of House Differences 5 Hypothesis I: It will not be possible to discriminate among the several residence hall houses 3 on the basis of linear combinations of variables describing their group char- LJ acteristics (as measured by selected HAS items). Rt! - The available computer routine for solving dis- criminant analyses was capable of handling no more than 50 variables, or in this case, 50 HAS items and/or other measures of house characteristics. Of the 121 HAS items which had been scored on what were assumed to be linear scales, 50 were selected. These represented what was thought to be an optimal combination of the items. The judgment was made on the basis of theoretical and/or in- trinsic interest of the items and the consistency of house ratings on the items as measured by Horst's i (an internal consistency indicator--see discussion of the HAS in Chapter III). What were thought to be the weaker of pairs or groups of items assessing essentially the same characteristics were also deleted. The multivariate 141 linear combinations of these items, or discriminant func- tions, were then obtained. An item analysis revealed that three of the HAS items selected had distributions of responses that departed substantially from normal. Consequently certain response categories on the items were collapsed into lower cate- ‘3 gories in order to provide a more normal distribution. } Normality of the variables is an assumption of the dis- criminant analysis. Items 2, 3, and 13 of Part I of the Fl“'_“"&.-VA' 4. HAS were the variables which were collapsed. This re- sulted in three categories for items 2 and 3 and only two for item 13. Item 14 (HAS Part I) was also dichotomous. Maxwell has demonstrated that it is acceptable to include dichotomous variables in discriminant analyses.1 Results of the discriminant analysis are listed in Table 5.3. The maximum number of latent roots (1) of the discriminant equation was 26 (number of groups minus one). Each of the roots, the amount of total variance accounted for by each root, Rao's chi-square value for testing the significance of each root, degrees of freedom and the level of significance are presented. 1A. E. Maxwell, "Canonical Variate Analysis When the Variables are Dichotomous," Educational and Psycho- logical Measurement, 21:259-271, 1961. 142 Table 5.3 Latent roots (1), explained variance, chi-square values, degrees of freedom, and statistical significance levels for each of the 26 discrimi- nant functions . Explained 2 . . . 0180.. A Variance x df Significance Function (per cent) Level 1 1.8709 28.140 891.70 75 <.001 2 .9335 14.040 557.46 73 <.001 3 .7313 10.999 464.08 71 <.001 4 .4660 7.009 323.45 69 <.001 5 .4340 6.527 304.76 67 <.001 6 .3332 5.012 243.17 65 <.001 7 .2721 4.093 203.51 63 <.001 8 .2646 3.979 198.47 61 <.001 9 .2152 3.237 164.81 59 <.001 10 .1590 2.391 124.74 57 <.001 11 .1304 1.961 103.64 55 <.001. 12 .1288 1.938 102.47 53 <.001 13 .1069 1.609 85.91 51 <.Ol>.001 14 .1043 1.569 83.92 49 <.01>.001 15 .0796 1.19é 64.78 47 <.os 16 .0756 1.137 61.61 45 <.05 17 .0597 .898 49.04 43 n.s.* 18 .0583 .877 47.93 41 n.s.* 19 .0528 .794 43.48 39 n.s.* 20 .0359 .541 29.86 37 n.s.* Table 5.3 (continued) 143 2322:... . 3253313? >8 df “92:32?“ (per cent) 21 .0343 .516 28.53 35 n.s.* 22 .0305 .459 25.41 33 n.s.* 23 .0258 .387 21.50 31 n.s.* 24 .0232 .349 19.42 29 n.s.* 25 .0136 .204 11.40 27 n.s.* 26 .0090 .136 7.58 25 n.s.* *Not significant 144 Sixteen of the 26 latent roots of the system were significant, 14 of these beyond the .01 level of confi- dence. The sum of the latent roots provided an estimate of the total variance of dispersion among the 27 houses as defined by the HAS items analyzed. The accumulative variance accounted for by the 16 significant roots amounted I“ to 94.8 per cent of the total. The remaining ten roots accounted for the balance of only 4.3 per cent and this amount apparently represented only chance variation among the houses. 4 On the basis of these findings Hypothesis I was rejected. It was evident that several linear combinations of variables from the HAS did discriminate among the 27 houses included in the study. Question A: What are the interpretations of the linear combinations of variables (discriminant functions) which may result? The discriminant functions may be interpreted by examining the relative contribution of each variable on each of the significant functions. For this purpose the standardized weights of each variable were used rather than the "normed weights" (the original vector coeffi- cients). Tiedeman and Bryan offer the following explana- tion. It can be shown that the individual values of the dis- criminant function are independent of the units of measurement, and origin of coordinates of the initial variates, since the coefficients automatically adjust 145 themselves (linearly) to the scales employed. On the other hand, the interpretation of separate coefficients does depend on the units of the initial variates.2 Had all the variables had equal units of measurement and approximately equal standard deviations the vector coeffi- cients could have been used. Since this was not the case use of the standardized weights accurately reflected the relative contribution of each variate to the discriminant functions. The 50 HAS items used in the discriminant analysis and their standardized weights on each of the first five significant discriminant functions are presented in Table 5.4. As can be determined from Table 5.3, these five functions accounted for 66.7 per cent of the total vari- ance among the houses. In general the interpretation of the significant discriminant functions was difficult; the results are thus speculative. With the possible exception of the first function complex relationships between the variables were abundantly evident. It would appear that the large quan- tity of non-independent variables represented by the HAS items had a confounding effect on attempts at interpreta- tion at least beyond the fifth function. As may be noted 2David V. Tiedeman and Joseph C. Bryan, "Predic- tions of College Field of Concentration," Harvard Educa— tional Review, 24, 1954, p. 132. I146 Table 5.4 Fifty HAS variables included in the discriminant analysis and the standardized weights GU of each on the first five discriminant functions Item 1 “4 No. Item al “I n: “I u "I w PART I HOUSE ANALYSIS SURVEY 2. Class in college -.48 .13 .42 -1.25 -.17 First quarter freshman Second quarter freshman Third quarter freshman Low Sophomore (40 to 62 hrs.) High Sophomore (63 to 84 hrs.) Junior (85-129 hrs.) Low Senior High Senior (will graduate this academic year) Graduate Student Othe : special, temporary, etc. QQQU‘hWNH O O C I om e a IT ..- 3. How many quarters have you lived in this house, including this quarter? -1.04 -l.45 1.63 -.57 -.ll 1. This quarter only 2. Two quarters 3. Three quarters 4. Four quarters 5. Five quarters 6. Six quarters 7. Seven quarters 8. Eight quarters 9. Nine quarters 10. Ten or more 13. Which of the following is correct concerning your present place of residence? ("On-campus" refers to University housing) -.30 -1.37 .45 -.49 -l.67 l. The Housing Office made both my current room and hall as- signment this year. I have not lived elsewhere on campus this year. 2. The Housing Office made my current room assignment, but I re uested to live in this half. I have not lived else- where on campus this year. 3. I requested both my current room and hall assignment this year. I have not lived else- where on campus this year. 4. I requested to move to this hall after livi g elsewhere on campus t 15 year. HouSing assigned my current house and room. 1147 Table 5.4 (continued) Item No. Item “I H O N 14. 15. 17. 18. 19. 20. 21. 23. 24. 25. 26. 29. 30. S. I requested to move to this hall and house after having lived elsewhere on campus this year. But Housing as- signed my room. 6. I requested to move to this hall and room after having lived elsewhere on campus this year. 7. I moved to my current room from a different house in this same hall earlier this year. 8. I moved to my current room from another room in this same house this year. 9. 0t er Did you request to live with an of your present roommates, ra er than being assigned together by Housing? 1. Yes 2. No Support for the participation in intramural sports Level of academic performance or scholarship in the housec The good times we have togetherC Reputation of the house within the residence hallc Contribution of life within the house to your understanding of issues, ideas, philosophies, etc.C Social life and social program of the house The lsadeaship of the Resident Assistant The leadership of the elected house officersc Ability to study in the housec Intellectual and cultural life of the house Value of living in this par- ticular house Compliance of residents with resident hall regulationsC .25 -1.51 9.04 -.95 .73 .54 .10 .11 .81 -.03 ".16 -.59 .52 .86 -.93 .45 -3.39 -.23 -s12 .04 -1.50 1.55 .41 .98 “1.57 2.75 1.03 -3.43 .40 -2s18 -1.53 .70 1.34 -s21 1.15 .42 -.37 -.70 -2.52 -2.94 1.40 -2.60 .50 1.14 -2.84 .14 -4.23 1.58 .28 -1s73 1.75 -s4° Ju48 Table 5 . 4 (continued) Item No. I tem 31. 32. 33. 34. 55. 56. 57. 58. Your satisfaction with your room- mate(s)c Your general satisfaction with residence hall accomodationsc Your level of satisfaction with living in this housec Your level of satisfaction wish living in this residence hall Where would you rate the general ”climate of learning' of your house? What has been the level of the 'climate" which YOU have person- ally experienced through those with whom you associate the most in the house, regardless of the general climate of the house?c Where would you personally like the level of the "Climate" to be in your house? Where would you rate the general 'climate of learning" of the residence hall in which you live?c II HOUSE ANALYSIS SURVEY) (PART 5. 10. 14. I think I would have done better academically so far this year had I lived in a different house 9 Residents of the house keep their rooms clean and neat.c I feel that fellows in the house are too involved in cliques.c When it comes right down to it, I really have little allegiance to either my residence hall or my house.c A number of campus leaders live in the house.c I would enjoy having faculty mem- bers visit informally with the house occasionally in order to discuss ideas, issues, their in- terests and work, etc. .11 -s‘3 -s9‘ .39 .43 ‘s‘l .45 ".40 ‘e67 -057 -.31 .24 1.18 -e26 -2e1‘ ‘sOI 1.13 -.03 -.52 .40 1.53 .16 1.18 “.35 1.20 's97 ”.03 .00 1.4: 1.59 “1.35 .27 .48 .22 -s01 .34 's71 .69 .46 1.57 -eoz .07 -1s51 2.76 -.09 .04 -1.35 1.12 -.40 ‘e‘3 -1.98 ’s77 .61 1.85 -s5‘ .57 1.24 -2.85 1.06 1.20 '.80 .34 -3s 35 -s91 -s38 .63 .82 ‘egg ‘s92 ..__..-.'r-'V'-'-‘.Il r ‘ p .149 Table 5.4 (continued) Item No. .-._, -3.“ _ .:,_:. .... . W... Item “I H 0| a: U! u: 0! u: 15. 16. 18. 20.. 22. 24. '28. 29. 37. 38. 41. 42. Comparatively speaking, our house is known for some of its original, novel or creative (though perhaps somewhat questionable) ideas and activities.c -.23 I feel that I am generally accepted and appreciated by those who live in the house.c —.73 Students in the house exhibit a high degree of concern for the rights of others.c .02 Living in my house is a major factor in making me feel a part of this university.c -.30 Many in the house tend to be more concerned about the amount of work required in a course or how easy it is to get a grade rather than the quality of the instructor or the contribution of the course to the individual.C .71 There are a number of traditions in the house.C -.04 The "intellectual" enjoys little status in the house in which I live.c -.56 Residents in the house have been involved in an above-average num- ber of disciplinary problems.C -.12 Topics of "bull-sessions" in the house are superficial rather than of depth or substance.c .55 The men in the house would be more likely to compliment some- one on a nonacademic (social, athletic, etc.) achievement than on an academic or inte1~ 1ectua1 achievement.c -.04 Students in the house have high ethical standards with respect to cheating, etc.c .16 There's quite a bit of pressure (subtle or otherwise) in the house to participate in house and university activities.C .13 -.11 -.11 .59 -.70 .33 -1s49 .37 -1.80 -.78 -1.22 -.75 *.50 ~1.07 -.81 -s26 -2s18 .44 -s76 -.21 .38 -.92 -.73 -.15 -1.58 3.78 -.93 .06 -1.46 -082 .34 .17 1.20 -1002 .15 -1.65 2150 Table 5.4 (continued) Item _ NO. Item 81 e “.l 44. 46. 47. 48. 58. 59. 60. There really isn't much interest in international affairs, social issues, or scientific discovery expressed among the residents of the house.c -.69 My house has effective means of dealing with residents whose be- havior isn't acceptable to the group.c .06 I would prefer to move to a different house.c .33 I would prefer to move to a different residence hall.0 .81 What proportion of your closest male friends at MSU live or have lived this year in your residence hall (including your house)? .48 1. Almost all 2. Meet 3. About half 4. A few 5. Almost none There are 8 to 12 houses in your residence hall. Where would you rate your house generally in con- trast to the other ouses in the hall? 1. One of the best 2. Better than average 3. About average 4. Below average 5. One of the worst 1.93 Where would you rate your residence hall compared to the other men a 5311s? .43 l. One of the best 2. Better than average 3. About average 4. Below average 5. One of the worst -.30 .24 .35 -.60 43.73 5.64 .34 .59 “.20 -.10 -2.14 -1e" .77 .87 1.38 .13 -012 .90 1.26 .58 .91 -.6‘ “1.15 '1.25 .41 151 Table 5.4 (continued) Item — - . -— No. Item a1 a2 a3 a4 a5 63. What degree of influence have the residents of your house had on YOU? e50 -s58 e83 s12 'e13 r l. A very positive influence ' 2. Some positive influence 3. Little or no influence 4. Some negative influence 5. A strong negative influence ~ | 8All 3 entries have been multiplied by 10. i bThe distribution of responses to these items in the discriminant analysis were reduced to two (item 13) and three (items 2 and 3) categories to correct their score distributions which departed excessively from normal. For items 2 and 3, responses 3-10 were combined into one category. For item 13, respon- ses 2-9 were combined into one category. cHAS Part I items 15-34 and 55-58 are rated on a nine point scale. A low score indicates a positive rating, a high score a negative rating. HAS Part II items 5-48 are scored on a five point scale. A low score indicates strong agreement with the item and a high score strong disagreement. These scoring procedures should not be confused with the magnitude of the stand- ardized weights (3). 152 from Table 5.3, discriminant function six accounted for only 5.01 per cent of the total variance; subsequent significant functions individually accounted for even less. Conse— quently, no attempt was made to interpret the functions beyond number five. One general interpretation of this dilemma would be that very real differences do exist be- “A tween the houses on important dimensions of house charac- teristics. But the variables employed in the measurement of the underlying dimensions by themselves are too micro- scopic and interrelated to sharply define the differences. :2 Also, and perhaps more importantly, each successive dis- criminant function extracted in the solution of the latent roots of the system is orthogonal to the preceding func- tion. This results in a somewhat artificial combination of variables when compared to the real world. The under- lying dimensions of house structure are not necessarily orthogonal to one another.3 These considerations should be born in mind in the interpretations of the first five discriminant functions which follow. Interpretation of the Rirst Discriminant Function The first function accounted for 28 per cent of the total variance (Table 5.3), somewhat more than any of 3These points have been considered more extensively in the section "Multiple Discriminant Analysis," in Chapter IV. 153 the other functions. Its interpretation would seem to be relatively straight forward. HAS Part I, item 17 (Table 5.4) was the prime operant among the 50 variables in dif- ferentiating among the houses on this function. Thus, this linear combination of variables is predominantly one of academic performance and scholarship in the houses (as re- r1 ported by the residents). Houses with high academic per- formance tend to have low scores on the function and houses with low ratings have high scores (Table 5.5).4 Four other variables seem to have some slight re- r” lationship to academic performance in differentiating be- tween the houses. There would seem to be a positive inter- action between the rating of academic performance and resi- dents' general ratings of their houses in comparison to other houses in their halls (HAS Pt.II,#59). There is an indication of a negative relationship to the degree of support for.mx1participation in intramural sports (HAS PT.I, #15). There may be a slight tendency for the number of 4In attempting to interpret the functions one must take into account not only the sign of the standardized weight of the HAS items on the function, but also the direction of the scoring of the items. Thus on the first function, houses with high scores are those whose residents report poor academic performance. The standardized item weight is positive indicating that a house with a high score on the HAS item (a poor academic performance rating) would have a high’house mean score on the function. Item 15,Pt.I, however, has a negative standardized weight; the item is scored the same direction as item 17 (the academic performance item), thus suggesting a negative relationship between the two items in differentiating between houses on the function. 154 Table 5.5 Mean discriminant scores of the 27 houses on the first five discriminant functions House First Second Third Forth Fifth No. Function Function Function Function Function 11 2.431 —1.124 -2.976 .867 -.303 12 3.524 .201 -.393 -.725 .559 13 1.109 -.494 -3.744 -.691 1.008 21 1.580 -.935 -.512 -.157 .790 22 1.565 -l.349 -2.723 .245 .625 23 5.179 -l.976 -2.168 .008 -.756 31 3.841 -2.092 -l.675 -.243 .767 32 2.409 -l.436 -2.120 .830 .362 33 3.433 -2.190 -1.808 -.352 -.136 41 -.246 -l.423 -1.979 -2.156 -l.108 42 1.740 -l.479 -.587 .993 .223 43 .724 -2.072 -.608 —.073 .486 51 2.575 -2.668 -l.986 -l.048 .390 52 2.851 —2.535 -2.225 .172 -.533 53 -.l99 -l.842 -l.116 -.619 .084 61 1.613 -2.806 -l.614 .779 -.204 62 1.762 -3.332 -.617 —.085 -.437 63 2.091 -2.613 -l.222 -.674 1.324 71 3.769 -2.019 -.001 -.216 -.403 72 1.776 -l.352 -1.905 .418 -l.088 73 4.806 -l.677 -1.340 —l.470 .199 81 3.884 -1.199 -.825 -.446 .875 82 1.452 -1.130 .115 .316 -1.091 83 2.671 -l.639 -3.164 -.009 .354 91 -.190 -2.268 -1.305 .429 .900 92 2.128 -2.901 -1.996 -1.677 .315 93 3.199 —2.948 -2.392 -.083 .035 T",— 155 campus leaders in the house to be associated with the level of academic performance (HAS Pt.II,#lO). And the average length of residence would seem to have some small positive association on the first function with perceived academic performance in the house. The Michigan State University Office of Residence ta Hall Programs provided information each quarter to each house as to house standing within the hall in mean grade- point-average (gpa). Of the six houses with the lowest mean scores on function one (Table 5.5), five had the L: highest mean gpa fall quarter, 1964 within their respec- tive halls. The sixth house was second in its hall. At the other end of the continuum the two houses with the highest mean scores on the function had the lowest mean gpa within their halls. Summarizing the function, it would seem that it differentiated among the houses on the basis of residents' ratings of house academic performance. Their ratings of academic performance, in turn, seem to have been based on data reported to them by the Residence Hall Programs Of- fice. A high level of academic performance in a house would seem to have some slight relationship to (1) a posi- tive comparative rating of the house within the hall, (2) a tendency toward poorer support for intramural activities, (3) the number of campus leaders in the house, and (4) a proportionately larger number of returning students. 156 (In this last regard the relationship is probably a func- tion of the tendency for mean gpa's of older students to be higher than those of freshmen.) Interpretation of thg Second Discriminant Function The second function accounted for 14 per cent of F1 the variance and represents that combination of the vari- ables best discriminating among the houses after the ef- fects of the first function have been removed. This func- lfi' ~ 2 tion seemed to separate the houses along a complex measure of satisfaction with both the residence hall and house. The function has a relatively heavy but not exclusive em- phasis on the social life of the house. The principle contribution comes from a rating of the residence hall compared to other men's halls (HAS Pt.II,#60). Paired with this item is a rating of the house in comparison to other houses in the hall (HAS Pt.II,#59). It would seem that the two variables in combination differentiate be- tween certain poorly rated houses located in relatively highly rated halls and better rated houses located in less satisfactory halls. The higher rated houses would seem to be those with a more satisfactory social life and program (Pt.I, #21), even though satisfaction with the hall was low. There would seem to be few traditions in the houses of 157 the halls receiving low ratings (Pt.II,#24). It is inter- esting to note the inverse relationship of satisfaction with residence hall accomodations (Pt.I,#32) and ratings of the hall. Apparently a negative hall rating does not necessarily reflect dissatisfaction with the physical ac- comodations. Suggestions of additional subtle characteristics of houses differentiated by this function can be noted by observing the other variables with slightly elevated weights. Mean length of residence in houses located in poorly rated halls tended to be shorter (Pt.I,#3); resi- dents were more often assigned their rooms by chance as opposed to being granted a preference (Pt.I,#l3). The house intellectual and cultural life tended to be positive however (Pt.I,#26), though residents were apparently in- clined to question the value of living within their par- ticular house (Pt.I,#29). The poorly rated halls tended to have correspondingly low ratings on the measure of the climate of learning of the hall (Pt.I,#58). Residents also seemed to feel that their houses in these halls had an above average number of disciplinary problems (Pt.II, #29). They also indicated that their fellow residents would be more inclined to compliment one another for a pppacademic achievement as opposed to an academic achieve- ment (Pt.II,#38). I. f- I” us nfiv ~"_I 158 Interpretation of thg Third Discriminant Function Eleven per cent of the variance was explained by the third function. Variables receiving high weightings on this function suggest that it discriminates among the groups according to their reputation in the hall (Pt.I, T7 #19). It also explains some of the elements that may 1 comprise the reputation of a house. Houses with poor reputations had low mean scores on the function; those tending toward good reputations had high mean scores. Further analysis suggests that reputation (as defined by the function) is not based on academic per- formance (Pt.I,#l7), which may even be inversely related to a good reputation in some of the houses. On the other hand, houses with good reputations tend to have a well regarded social life and program (Pt.I,#21) and have a number of traditions (Pt.II,#22). Both the halls in which houses with high scores on the function are located, and houses with high scores received favorable ratings when compared to other halls and houses (Pt.II,#59 and #60). To a lesser extent, houses with good reputations tended to give support to intramural sports (Pt.I,#15); the lead- ership exhibited by the resident assistant was positive (Pt.I,#23); and there were campus leaders living in the house (Pt.II,#10). Curiously though the level of satis- faction would seem to be inversely related to reputation 159 in at least some of the houses (Pt.I,#33). This may have been a.reflection of the significant interaction between college type and houses classified according to their group mean score (high vs. low means) on the function. As will be seen later in this chapter, non-conforming stu- dents seemed to perform more adequately in at least one house having a poor "social" reputation as defined by the function. The same house was well regarded academically. Interpretation of thg Fourth Discriminant Function Though function four accounted for only seven per cent of the remaining variance, it did differentiate be- tween the groups in an interesting fashion. It seems to be primarily characterized by variations in the residents' conduct and degree of compliance with hall and university regulations. The most heavily weighted item on the func- tion differentiated between the groups according to whether or not residents tended to be involved in an above average number of disciplinary problems (Pt.II,#29). A high mean score on the function indicated a tendency toward few dis- cipline problems, and relatively high compliance with regu- lations; a low mean score indicated the reverse. The lead- ership of the resident assistant (Pt.I,#23) seemed to be positively related to a lower incidence of disciplinary problems and to residents' compliance with hall regulations Ffi “ FA” :1!'.ia‘. ‘..~L V I 160 (Pt.I,#30). Nevertheless, a greater number of disciplinary problems was curiously associated to some extent with sat- isfaction in some of the houses (Pt.I,#33). The ability to study within the house (Pt.I,#25) and good social pro- grams (Pt.I,#21) were characteristic of houses with fewer disciplinary problems. Residents were more likely to keep their rooms clean and neat (Pt.II,#6). In the other direction more campus leaders tended to live in what might be termed the low compliance houses (Pt.II,#10). Such houses were more likely to be known " for their original, novel or creative (though perhaps some- what questionable) ideas and activities (Pt.II,#15), and were likely to have a number of traditions (Pt.II,#24). The "intellectual" more often had status in these houses (Pt.II,#28). Likewise, residents in their selection of courses were perhaps more interested in the quality of the course rather than how easy it was to get a grade (Pt.II,#22). From these last results there would seem to be a hint of intellectual rebellion in at least some of the houses characterized by a relatively low order of compli- ance with regulations. This of course, would not neces- sarily be the case in all such houses. Interpgetation of the Fifth Discriminant Function Only 6.5 per cent of the variance was accounted for by the fifth function. The underlying characteristic 161 of this function is perhaps a little more confusing. The houses are primarily differentiated according to their level of support for and participation in intramural sports (Pt.I,#15). Houses supporting intramurals also perceive a more positive climate of learning within the hall (Pt.I,#58) and tend to be satisfied with the house “1 (Pt.I,#33). The social life of the house, however, is 4 weighted in the other direction suggesting that the func- tion differentiated between some houses supporting intra- murals as Opposed to those having a strong social program. This curious relationship may to some extent be a function of the degree of variance already accounted for by previous discriminant functions. There would also seem to be an inverse relationship between the support given intramurals and certain houses being known for their novel or creative ideas (Pt.II,#15). This function may reflect a general performance factor primarily weighted toward intramural sports, but also including the discriminating ability of remaining variance of variables reflecting other house programs. Question B: Is it possible to develop a meaningful typology of the residence hall groups according to the description of the dis- criminant functions and the location of the groups in multi-dimensional space? To a large extent an answer to Question B has been provided by the presentation and interpretation of the 162 five discriminant functions. In Table 5.5 the mean (cen- troid) scores of the 27 houses on the five discriminant functions previously described are listed. These same mean scores have been graphically portrayed in Figure 5.1. In that figure the position of each house on each of the five functions can be seen. Houses with a low score on a given function are located to the left and those with high scores to the right. The house mean scores on the 50 HAS items included in the discriminant analysis are listed with mean scores on the other HAS items in Appendix D. These mean scores should not be confused with the house mean (or centroid) scores on the discriminant functions. An examination of house means on the HAS variables aids in the interpretation of the functions. An interpretation of the nature of the differences between the houses becomes most lucid at the extremes of the continua. Toward the center of the distributions of houses on the five functions, the relationship between the variables is clouded. This is in part a function of the discriminant analysis itself in that scores tend to be normally distributed on each function. Thus they do tend to be more heavily distributed toward the center. Typically, in interpreting a discriminant function a priori distinctions exist between the groups compared, from which one could pre— dict differences or at least interpret them when they are found. The a priori homogeniety of the houses as a type .soauocsu new so msoflumooa Hams» com momsoc hm ecu mo some “comonmon nosed one scams mHoQEsz«« .sofluossm usmsHEflHOch ecu Mo mosam> one ounceccfl nosed on» o>onm mncnesz « Nb «inn 1% A _ _ _ _ _ N T m- on \m. 3 Rewind“ mm. mmv > cosuocsm “855383 I u.» t is «A ..ng a. s ch 4 .0 ms um. , e _ _ q 341% d— 2: _ . d . _ a W H 0 HI NI .. as NO >H sofluossm ucmsafifluomflo R w\ nu «we 6 an R. 3 - um. \Jquse \m. - nu mu \m \n spun. Ne mm .uw \\ Mm mx . . . «J4. a. .. q a— . q . _ l 4 _ . 4|. ._. H 0 HI NI MI vl HHH GOHHOGDW UGmGHEflHUmHQ NNNN Mn. ”N “l 0n. 00 .Nm , N\ n\ 0. exodus Ni m nu \n in «580 on Me _ — I — % H— j: —- = . aa_ a Ifi — H o m- T HH GOHUOGSW UGwGfiEHHUHmQ NR sh xx Gun 3.3 R Bax A. Mme mm -u... an. ...n Meme .3 .m an ex, A.» 0.3» q . _ _ a .4. . — A a 1 4. — Mil—4.11 . A a _ J31 «m v m N a o H sowuossm usmcwafiuomao mcowuossm usmcflfiwnomfic c>flm umuam or» so monsoc hm can no mmuoom ucmnflefluomao come no coaumooq H.m onsmflm 163 164 of living unit did not provide that lever in this study. Nevertheless, houses at the extremes of the five functions do tend to differ from one another along the lines described in the descriptions of the functions. Hence the answer to Question B is a qualifed yes. The direction of differences between the houses along the five continua will now be briefly considered. Houses with low scores on function one (e.g., houses 41, 91, and 53) tended to have high ratings of their level of academic performance. All three had the highest mean gpa of their respective halls fall quarter, l964--a fact of which the residents were undoubtedly aware when they completed the HAS. The two houses with the highest scores on function one (73 and 23) had both had the lowest mean gpa for their respective halls. Houses falling at the low end of function two tended to have high ratings of their halls compared to other campus halls, though the houses did not necessarily receive such a high rating. At the other end, the houses tended to be rated high but the halls in which they were located tended to receive poor ratings. At the low end of function three residents' re- sponses suggested that their house reputation left some- thing to be desired. At the opposite end, the reputation of houses so located tended to be more positive. 165 Compliance with hall and institutional regulations tended to be poorer in houses located at the lower end of function four in contrast to houses receiving high scores on the function. Houses with low scores on function five were those whose support for intramural sports tended to be compara- “ tively weak, while those at the opposite end were likely to have stronger programs. From Figure 5.1 the close cluster- ing of houses on function five can be observed. This may in F'fl“ part account for the ambiguity of this function, where relationships between highly weighted variables were not distinct. Question C: Does the location of houses in the empiri- cally derived multi-dimensional space relate in a meaningful fashion to the classifica- tion of houses according to program and/or physical characteristics (such as common location in the same hall and/or complex or construction similarities such as living- learning features)? In light of the large number of significant func- tions and the relatively low percentage of the total vari- ance accounted for by any single function, no attempt was made to depict the houses in multi-dimensional space. A three-dimensional portrayal is visually possible. However, the first three functions accounted for only 53 per cent of the total variance, and thus left 47 per cent of inter- house variation unexplained. Consequently, no attempt was :nade to portray the distribution of houses in more than one dimension as represented in Figure 5.1. 166 Only on the distribution of house means on func- tion Unawould there appear to be any consistent patterning of houses according to physical characteristics of the halls. As was indicated in the preceding section, houses with low scores on the function tended to be rated as better than average or best among the men's residence halls at Michigan State University. At the other end of the con- tinuum, houses so located tended to be rated by residents below average among the halls. A more complete interpre- tation of the function was given in a previous section of this chapter. It would appear that Hall 1 was consistently given a low rating by its residents. Hall 2, designed similarly, also tended toward the same end of the continuum though the mean score of one of its three houses fell in the middle of the distribution. These two halls were both in their first year of operation which would explain, at least,why houses with high scores on the function tended not to have well established traditions. Also, residents were more often randomly assigned to the hall. Additional explanations were suggested from some of the responses from residents of these halls to an Open- ended item on the HAS (Pt.II,#70). Several residents had indicated a level of dissatisfaction with the four man suite concept employed in these new halls. The poor rat- ings may simply have been a function of the newness of the halls. 167 With the exception of one house in Hall 5, all the houses of Halls 5, 6 and 9 were located at the low end of discriminant function two. Halls 5 and 6 are the East and West wings of a large, older, somewhat traditional men's residence unit. Hall 9 is also an older facility having a traditional character. It would appear that residents n of these halls tend to regard them quite highly. It is also significant to note that none of the houses of the living—learning units (Halls 1-4) were included among the eight houses having the lowest mean scores on the function (though Hall 4 did receive comparable good ratings on the item asking residents to rate their hall--HAS, Pt.II,#60). In general, it is important to note that differ- ences between houses seem to predominate much more than any observable inter-hall differences on the five functions, with the single exception noted above. This would in part at least be a function of the variables considered, most of which were designed to assess inter-house differences. Nevertheless, it would seem that the houses do depict an important level of interpersonal interaction within the university milieu. Likewise it is obvious that the dif- ferences between the groups are extensive. The Impact of Types of Houses on the Academic Performance of Freshmen Hypothesis IIal There will be no differences in the (adjusted) academic performance of freshmen differentiated according to 168 the types of houses in which they live (types defined according to clustering of houses along the discriminant func- tions and/or in the multi-dimensional function space). Hypothesis IIa2 Nor will there be an interaction between types of houses and the pre-test sub- culture orientation of the residents. In order to test Hypothesis IIa houses with mean r- 1! scores located at each of the two extremes of each of the five discriminant functions were grouped together. Two groups of houses were thus created depending on the loca- tion of the mean scores of the houses on each of the five discriminant functions previously analyzed. The distribu- tion of houses on each of the functions was portrayed in Figure 5.1. On each of the five functions an attempt was made to select those houses at the extremes that were some- what separated from houses clustering toward the center of the continuum. The numbers of students in each of the five high and five low groups are presented in Table 5.6. Each of the "high" and "low" groups is subdivided accord- ing to the numbers of freshmen in each of the four sub- culture orientations. Hypothesis IIal Using the groups so created five two-factor (2 x 4) analyses of covariance were computed, one for each of the five discriminant functions. Type of house (high vs. low 169 Table 5.6 The number of freshman residents grouped by subculture orientation in houses (1) with high mean scores and (2) with low mean scores on five discriminant functions* Subculture High Mean LOW Mean Totals Orientation Score Houses Score Houses Discriminant Function I Vocational 72 58 130 Non-conformist 32 25 57 Academic 100 70 170 Collegiate 33 16 49 Totals 237 169 406 (Houses (73,23,81,3l, (61,21,22,82, Included) 71,51,12,33,93) 13,43,91,53,41) Discriminant Function II Vocational 42 55 97 Non-conformist 26 18 44 Academic 83 41 124 Collegiate 20 19 39 (Houses (12,13,21, (52,63,51, Included) 11,82,81) 61,92,93,62) Discriminant Function III Vocational 50 63 113 Non-conformist 34 29 63 Academic 79 80 159 Collegiate 30 20 50 Totals 113 192 385 (Houses (71,82,12,21, (32,23,52,93, Included) 42,43,62,81) 22,11,83,13) l 170 Table 5.6 (continued) Subculture High Mean Low Mean Totals Orientation Score Houses Score Houses Discriminant Function IV Vocational 64 80 144 Non-conformist 37 31 68 Academic 75 88 163 Collegiate 16 25 41 Totals 192 224 416 (Houses (42,11,32,61, (53,63,13,12, Included) 91,72,82,22,52) 51,73,92,41) Discriminant Function V Vocational 56 44 100 Non-conformist 14 33 47 Academic 76 73 149 Collegiate 18 16 34 Totals 164 166 330 (Houses (63,13,91, (33,11,71,52, Included) 81,21,31) 72,23,82,4l) *All 27 houses are plotted on each discriminant function in Figure 5.1. 171 houses) constituted one main effect; subculture orientation the other; the dependent variable was adjusted freshman gpa. The results of the analyses are presented in Tables 5.7-5.16. As may be noted from the five covariance tables (Tables 5.7, 9, 11, 13, and 15), the main effect differ- ence on gpa's between the four college types (subculture orientations) was consistently significant as was expected. This study did not directly concern itself with subculture differences (these have been extensively covered else- where).5 But it was necessary to include them in order to study the interaction between subculture orientation and type of house. No significant differences in adjusted freshman gpa were noted between types of houses (the other main effect) on any of the five analyses of covariance. Con- sequently the null hypothesis was not rejected for Hypothe-> Sis IIal. Hypothesis IIa2 In testing Hypothesis IIaZ, the interaction between type of house and subculture orientation on the third 5Benjamin Joseph Hodgkins, "Student Subcultures-- An Analysis of their origins and affects on student atti- tude and value change in higher education" (unpublished dissertation, Michigan State University, 1964). 172 Table 5.7 Analysis of covariance of two-quarter accumula- tive grade-point-average of freshman residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function I Source of Variance SS df V F Between House Groups .1158 l .1158 .3266 E Between College i Types 5.5443 3 1.8481 5.2131* I House Groups x ' College Types 1.0852 3 .3617 1.0203 Error 140.3866 396 .3545 l *Significance level <.Ol>.OOl Table 5.8 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classi- fied according to subculture orientation and type of house on Discriminant Function I Subculture High Mean Low Mean Totals Orientation Score Houses Score Houses Vocational 2.41 2.48 2.45 Non-conformist 2.38 2.36 2.37 .Academic 2.29 2.50 2.39 Collegiate 2.08 1.99 2.04 Totals 2.29 2.33 173 Table 5.9 Analysis of covariance of two-quarter accumula- tive grade-point-average of freshman residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function II Source of Variance SS df V F 1“} Between House Groups 1.0200 1 1.0200 2.9198 . Between College Types 3.6418 3 1.2139 3.4750* House Groups x College Types 2.3710 3 .7903 2.2624 _ Error 102.7038 294 .3493 W *Significance level <.05 Table 5.10 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classi- fied according to subculture orientation and type of house on Discriminant Function II Subculture High Mean Low Mean Totals Orientation Score Houses Score Houses Vocational 2.40 2.46 2.43 Non-conformist 2.22 2.75 2.48 Academic 2.28 2.35 2.32 Collegiate 2.15 2.05 2.10 Totals 2.27 2.40 174 Table 5.11 Analysis of covariance of two-quarter accumula- tive grade-point-average of freshman residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function III Source of Variance SS df V F Between House Groups .0036 l .0036 .0100 run Between College Types 4.8025 3 1.6012 4.4127* ' House Groups x College Types 3.0071 3 1.0024 2.7624** 9 Error 136.0698 375 .3629 i4; * Significance level <.01>.001 ** Significance level <.05 Table 5.12 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classi- fied according to subculture orientation and type of house on Discriminant Function III Subculture High Mean Low Mean Totals Orientation Score Houses Score Houses ‘Vocational 2.53 2.34 2.43 Non-conformist 2.35 2.69 2.52 Academic 2.35 2.33 2.34 Collegiate 2.18 2.02 2.10 Totals 2.35 2.35 175 Table 5.13 Analysis of covariance of two-quarter accumula- tive grade-point-average of freshman residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function IV Source of Variance SS df V F Between House Groups .1388 l .1388 .3994 n Between College Types 4.3069 3 1.4356 4.1317* House Groups x College Types .3387 3 .1129 .3249 Error 141.0745 406 .3475 L‘” *Significance level <.01 Table 5.14 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classi- fied according to subculture orientation and type of house on Discriminant Function IV If - Iubculture High Mean Low Mean . . Totals Irientation Score Houses Score Houses ocational 2.39 2.39 2.39 >n-conformist 2.49 2.34 2.42 :ademic 2.46 2.47 2.46 rllegiate 2.11 2.07 2.09 Totals 2.36 2.32 176 Fable 5.15 Analysis of covariance of two-quarter accumula- tive grade-point—average of freshman residents, adjusted by MSU Reading and CQT total scores, for high and low house groups on Discriminant Function V Source of Variance SS df V F Between House Groups .9039 l .9039 2.4987 Ea}, Between College Types 4.1203 3 1.3734 3.7968* | iouse Groups x College Types 1.0079 3 .3360 .9288 E Error 115.7541 320 .3617 L’” *Significance level <.05 Table 5.16 Mean two-quarter accumulative grade—point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classi- fied according to subculture orientation and type of house on Discriminant Function V Subculture High Mean Low Mean Totals )rientation Score Houses Score Houses locational 2.41 2.45 2.43 Ion-conformist 2.49 2.40 2.44 Icademic 2.34 2.17 2.25 lollegiate 2.23 1.94 2.08 Totals 2.37 2.24. 177 iminant function was significant beyond the .01 level. quently the null hypothesisfor this interaction on as rejected. The null hypothesis was not rejected 1ny of the other four interactions. The five tables of adjusted category meangpa's les 5.8, 10, 12, 14, and 16) indicate the predicted es of the dependent variables when the effects of the Reading Test and CQT total scores have been removed. 5e adjusted gpa's are of interest in noting the direc- n of differences between the types of houses and sub- ture orientations. The direction and magnitude of the ferences in adjusted gpa's between the categories listed the tables also help one interpret the nature of sig- Eicant interactions between the two main effects. scussion A problem encountered in testing all of the hy- atheses requiring the analysis of covariance test was ample size. A sufficient number of houses had to be elected at both ends of the distribution of mean scores :0 insure a large enough sample such that if differences lid exist, the test would be sufficiently sensitive to be significant. This was particularly crucial in testing the hypotheses concerning the measures of intellectual disposition (Hypothese IIb and IIe) . For these hypotheses 178 the sample was already diminished because a large number of freshmen had not completed the post-test intellectual disposition measure. The power of the statistical test is usually increased by making the sample size larger, and the likelihood of committing a type II error is decreased. With smaller samples power is decreased. Yet in this study, given the available houses from which to draw, to increase sample size by adding more houses to the high and low groups would -a_l_sp_ tend to decrease >ower. This paradox lies in the fact that power is inversely elated to the heterogeniety of the population.6 If the umber of freshman residents of each of the 27 houses lready included in the research sample could have been Icreased, power of the tests would have risen accordingly. t that was not an Open alternative. The only Option ailable was to add additional house groups from the ailable 27 with mean scores nearer the center of the dis- butions on the functions. As has already been noted, racteristics of houses near the center of the distribu- 'l were not clearly differentiated. Perhaps the best :lution of the problem lies in a future replication of study with tighter controls and a larger sample. ing these problems in mind the results of the analy- >f covariance tests are considered next. 6William S. Ray, An Introduction to Experimental a (New York: Macmillan Co., 1960), p. 68. 179 Despite the fact that the first discriminant func- tion differentiated between houses primarily on the basis of residents' ratings of house academic performance, no differences in gpa's were noted between the groups of houses when the effects of academic ability were removed. It is a comon practice of housing administrators to compare liv- ing groups on the basis of their unadjusted mean gpa's. It would seem, however, that in this case mean gpa did 10t reflect any differences in, at least, freshmanresi- lents when academic ability was controlled. Frequently, hen housing types are compared on the basis of academic erformance by analysis of covariance, observed differences 1rn out to be a function of academic ability, rather than ' the type of housing.7 With reference to Table 5.10, the pattern of ad- sted mean gpa's is interesting. Though the null hypothe- was not rejected, differences between the two groups ~ 3 houses approached significance (.085 level). Function had discriminated between groups primarily on the basis residents' ratings of their halls. The low meanscore ses had relatively high mean ratings; houses at the ar end of the continuum were less well regarded. And was the only function upon which one could observe a erning of houses according to the halls in which they 7For example see Ralph E. Prusck and Bruce Walsh, lege Students Residence and Academic Achievement," Lal of College Student Personnel, 5:180-184, 1964. 180 were located. Though the results are clearly inconclusive, the adjusted mean gpa's of three of the four subtypes are in the direction of the more positively regarded halls. The interaction term of the third analysis of co— variance was significant (in the test of Hypothesis IIaZ) . Group mean scores high on the third function tended to re- flect houses with good reputations though apparently not with regard to academic performance. Residents of houses with low mean scores on the function rated their house reputation somewhat lower than residents of houses with iigh mean scores. Through observation of adjusted group eans in Table 5.12, the significant interaction seems to e a function of higher mean scores of vocationally and allegiately oriented freshmen in houses with good reputa- ons (and/or lower mean scores in houses with poorer putations) . The reverse would seem to be true for non— nforming students. An explanation may be that the vo- :ionally and collegiately oriented students derive more er-group support and reinforcement for their orientations hin their apparently well-regarded houses with relatively ong social programs and traditions. On the other hand -conforming students seemed to perform better in houses 9 the house reputation, based on non—academic values, not as good. One could speculate that there were other 181 elements in these houses supporting the value systems of the non-conformists. And conversely perhaps non-conforming students in the mileau of the houses with the positive (but apparently non-academically oriented) reputation did not receive support for their system of values and needs. Three considerations should be kept in mind. First, no cause and effect relationship should be assumed in these analyses. Secondly, the interpretations are speculative. Thirdly, there is the possibility that the significant interaction discussed above was itself a product of chance. Inasmuch as a series of analyses were computed, the pos— sibility of committing a type I error is increased through :he laws of probability. Given the fact that the con— idence level of the one significant interaction was well eyond the .01 level, the likelihood of this being a chance Lfference is diminished . To recapitulate, the test of Hypothesis IIa2 pro- ced no significant differences in adjusted gpa's between pes of houses. Only one of the five interactions be- een types of houses and subculture orientations was sig- icant. Collegiate and vocational students seemed to form better academically in houses with good social itations. Non—conformists apparently did better in :es with poorer reputations, but with a better academic sphere. 182 The Impact of Types of Houses on the Intellectual Disposition of Freshmen Efimwthesis IIbl: There will be no differences on any of the four post-test (adjusted) measures of intellectual disposition (as measured by the four OPI scales) of freshmen differentiated according to the types of houses in which they live. IIb : Nor will there be an interaction between types of houses and the pre-test sub- culture orientation of the residents. Hypothesis IIb was tested in a manner similar to that used in testing Hypothesis IIa, with the exception that the dependent variables were the four post—test mea— sures of the same variables. The results are reported in Fables 5.18 to 5.22. Table 5.17 lists the number of res- Ldents in each category. Though the houses were grouped .n a fashion identical to the groupings employed in testing ;he previous hypothesis, the category n's were slightly Inaller. Usable data from the pre- and post-test measures .f intellectual disposition were available for only 60 per ent of freshman residents. ypothesis IIbl A review of Tables 5.18 through 5.22 provides the allrnming information. No significant differences were >t£xi between any of the five groups of high and low houses iLrst main effect) for any of the four intellectual dis- >siJzion measures. The null hypothesis was therefore not ejected for Hypothesis IIbJ . 183 Table 5.17 The number of freshman residents who completed pre- and post-measures of intellectual dispo- sition (four OPI scales on the Attitude Inven- tory). The residents are grouped by subculture orientation in houses (1) with high mean scores and (2) with low mean scores on five discrimi- nant functions* Subculture High Mean Low Mean Totals Orientation Score Houses Score Houses Discriminant Function I Vocational 45 42 87 Non-conformist 22 22 44 Academic 59 50 109 Collegiate 20 5 25 Totals 146 119 265 (Houses (73,23,8l,31, (61,21,22,82, Included) 71,51,12,33,93) 13,43,91,53,4l) Discriminant Function II ‘Vocational 30 33 63 Non-conformist 22 ll 33 .Academic 65 25 9O Collegiate 10 9 19 Totals 127 78 205 (Houses (12,13,21, (52,63,51,61, :anluded) 11,82,81) 92,93,62) Discriminant Function III Vocational 32 41 73 Non-conformist 26 24 50 Academic 54 63 117 Collegiate 15 10 25 Totals 127 138 265 (fflouses (71,82,12,21, (32,23,52,93, ' Irm:luded) 42,43,62,81) 22,11,83,13) 184 Table 5.17 (continued) Subculture High Mean Low Mean T t 1 Orientation Score Houses Score Houses 0 a S Discriminant Function IV Vocational 35 52 87 Non-conformist 32 22 54 Academic 52 55 107 Collegiate 7 15 22 Totals 126 144 270 (Houses (42,11,32,61, (53,63,13,12, Included) 91,72,82,22,52) 51,73,92,41) Discriminant Function V Vocational 42 27 69 Non-conformist 13 25 38 Academic 52 48 100 Collegiate 9 10 19 Totals 116 110 226 (Houses (63,13,91, (33,11,71,52, Included) 81,21,31) 72,23,82,41) *All 27 houses are plotted on each discriminant unction in Figure 5.1. Table 5.18 185 Analyses of covariance of freshman post-test scores on the four OPI scales, adjusted by pre-test scores on the same scales, for high and low house groups on Discriminant Function I I If Source of Variance SS df V F Thinking Introversion Between House Groups 2.992 1 2.992 .0821 Between College Types 351.856 3 117.285 3.2174* House Groups x College Types 68.480 3 22.827 .6262 Error 9332.173 256 36.454 Theoretical Orientation Between House Groups 2.859 1 2.859 .2539 Between College Types 71.243 3 23.747 2.1091 House Groups x College Types 115.356 3 38.452 3.4150* Error 2882.498 256 11.260 Estheticism Between House Groups 2.909 1 2.909 .3381 Between College Types 34.218 3 11.406 1.3258 iouse Groups x Zollege Types 12.498 3 4.166 .4842 3rror 2202.394 256 8.603 Autonomy Between House Groups 9.637 1 9.637 .5154 Between College Types 96.058 3 32.019 1.7125 iouse Groups x Zollege Types 28.473 3 9.491 .5076 Error 4786.641 256 18.698 '7 *Significance level <.05 133 . . i. , my.» 5.1" 186 Table 5.19 Analyses of covariance of freshman post-test scores on the four OPI scales, adjusted by pre— test scores on the same scales, for high and low house groups on Discriminant Function II Source of Variance SS df V F Thinking Introversion Between House Groups .246 1 .246 .0074 Between College Types 93.238 3 31.079 .9317 House Groups x College Types 39.181 3 13.060 .3915 Error 6537.866 196 33.356 Theoretical Orientation _ ‘1‘.- —.w I'.l 1.» 1 Between House Groups 2.141 1 2.141 .1818 Between College Types 43.721 3 14.574 1.2374 House Groups x College Types 23.732 3 7.911 .6716 Error 2308.439 196 11.778 Estheticism Between House Groups 1.729 1 1.729 .1848 Between College Types 23.153 3 7.718 .8249 House Groups x College Types 46.106 3 15.369 1.6426 Error 1833.868 196 9.356 Autonomy Between House Groups 6.419 1 6.419 .3861 Between College Types 61.624 3 20.541 1.2354 House Groups x College Types 28.379 3 9.460 .5689 Error 3258.889 196 16.627 Table 5. 20 187 Analyses of covariance of freshman post—test scores on the four OPI scales, adjusted by pre— test scores on the same scales, for high and low house groups on Discriminant Function III Source of Variance SS df V F Thinking Introversion Between House Groups .063 1 .063 .0018 Between College Types 240.864 3 80.288 2.2556 House Groups x College Types 9.345 3 3.115 .0875 Error‘ 9112.166 256 35.594 Theoretical Orientation Between House Groups .049 l .049 .0041 Between College Types 35.301 3 11.767 .9807 House Groups x » College Types 5.036 3 1.679 .1399 Error 3071.493 256 11.998 Estheticism Between House Groups .831 1 .831 .0981 Between College Types 51.714 3 17.238 2.0334 House Groups x College Types 20.154 3 6.718 .7924 Furor 2170.226 256 8.477 Autonomyi Between House Groups .296 .296 .0184 Between College Types 113.429 3 37.810 2.3405 House Groups x Cdflege Types 27.864 3 9.288 .5750 Error 4135.554 256 16.155 .Ifll‘ri!) In!“ 188 frauolxe 5.21 Analyses of covariance of freshman post-test scores on the four OPI scales, adjusted by pre- test scores on the same scales, for high and low house groups on Discriminant Function IV Source of Variance SS df V F Thinkingilntroversion Between House Groups 13.077 1 13.077 .3893 Between College Types 272.792 3 90.931 2.7073* House Groups x College Types 129.377 3 43.126 1.2840 Error 8766.122 261 33.587 Theoretical Orientation Between House Groups 9.116 1 9.116 .8096 Between College Types 62.328 3 20.776 1.8453 House Groups x College Types 3.236 3 1.079 .0958 Error 2938.537 261 11.259 Estheticism Between House Groups 25.405 1 25.405 2.8443 Between College Types 37.896 3 12.632 1.4143 House Groups x College Types 25.400 3 8.467 .9479 Emror 2331.211 261 8.932 Autonomy Between House Groups 39.988 1 39.988 2.3139 Between College Types 82.855 3 27.618 1.5981 ' House Groups x CdUege Types 35.614 3 11.871 .6869 Error 4510.482 261 17.282 *Significance level <.05 189 Table 5.22 Analyses of covariance of freshman post-test scores on the four OPI scales, adjusted by pre— test scores on the same scales, for high and low house groups on Discriminant Function V I I; Source of Variance SS df V F Thinking Introversion Between House Groups 79.289 1 79.289 1.9643 Between College Types 204.794 3 68.265 1.6912 Iouse Groups x Zollege Types 232.030 3 77.343 1.9161 Brror 8759.018 217 40.364 Theoretical Orientation Between House Groups 20.434 1 20.434 1.8569 Between College Types 95.694 3 31.898 2.8987* iouse Groups x College Types 102.553 3 34.184 3.1065* 3rror 2387.895 217 11.004 Estheticism Between House Groups 2.173 1 2.173 .2484 Between College Types 41.202 3 13.734 1.5695 House Groups x College Types 11.268 3 3.756 .4292 Error 1898.850 217 8.750 Autonomy Between House Groups 13.496 13.496 .9252 Between College Types 121.154 3 40.385 2.7687* House Groups x College Types 45.138 3 15.046 1.0315 Error 3165.193 217 14.586 *Significance level <.05 190 Significant differences in Thinking Introversion res were noted between college types (the second main fect) on two of the five Thinking Introversion measures ables 5.18 and 5.21). One other significant F between 11ege types was noted (Table 5.22) on the Theoretical ientation scale. But it is most likely that all of the -gnificant F's were the result of chance sampling fluc- aations, given the large number of comparisons made in esting Hypothesis IIb. And again differences between :ollege types are not of immediate significance to this study. 1 ‘53!" if.) 4. _ Hypothesis IIb2 The interactions between type of house and sub- culture orientation were significant on two of the five tests for differences on the Theoretical Orientation scale (Tables 5.18 and 5.22) . The types of houses involved were those differentiated by Discriminant functions I and V. These differences may well have been a function of a chance variation in the samples drawn from the population. Nev- ertheless, because the observed differences were consistent with theory the null hypothesis for Hypothesis IIb2 is qualifiedly rejected for the interaction term in the two instances cited . 191 iscussion In Table 5.23 adjusted Theoretical Orientation core means of freshmen are presented for the interaction ;ests which were significant (test of the Hypothesis IIbZ). Fhe students are classified by college type and by types >f houses differentiated on discriminant functions I and V. It would appear from the table that in both cases lescribed, the non-conforming and academic residents are Lnfluenced in one direction according to type of house, and the vocational and collegiate students in the other direction. According to the theory upon which the subcul- tures are derived, the vocational and collegiate orienta- tions differ from the non-conforming and academic in the degree of affinity of each pair for ideas and the academic life. The latter pair, of course, are more interested in the intellectual domain. Though the observed differences in the means between the groups are not profound,tfluadirec- tion of differences does seem to coincide somewhat with those observed by Nasatir.8 He had noted that "vocationally" oriented students were more likely to succeed when they resided in halls with a less intense academic climate. The group of houses with high mean scores on func- tion I were generally those whose ratings of academic 8David Nasatir, "A Contextual Analysis of Academic Failure," The School Review, 71:290-298, 1963. 192 able: 5.23 Mean Theoretical Orientation post-test scores, adjusted by pre-test scores, for groups of freshman residents classified according to subculture orientation and type of house on Discriminant Functions I and V* ._. I Subculture High Mean Low Mean Totals )rientation Score Houses Score Houses Discriminant Function I w 1 3" Vocational 19.20 17.66 18.43 5 Non-conformist 19.26 20.03 19.64 1 Academic 17 . 75 19 .08 18. 41 , :J‘ Collegiate 18.11 16.45 17.28 Totals 18.58 18.30 Discriminant Function V Vocational 17.86 18.50 18.18 Non-conformist 19.91 19.67 19.79 Academic 18.34 17.23 17.79 Collegiate 16.16 19.85 18.01 Totals 18.07 18.81 *These two tests of interaction between subcul- tures and types of houses were significant beyond.the .05 levelcfifconfidence in the analyses of covariance (see Tables 5.18 and 5.22). 193 erformance of the house were somewhat negative. The .ouses with low mean scores on the function enjoyed more 'avorable academic performance ratings. Thus, one explana- ;ion of the results would be that students valuing intel- .ectual activities found more support in houses with a high .evel of academic performance resulting in higher Theoretical >rientation scores. Or they may have found less support in iouses with poor academic performance. Either or both prOp- | nsitions could be true. The situation would be reversed 1“ Eor students not identifying with the intellectual life. 1 A parallel situation may pertain in considerating :he character of houses dichotomized by discriminant func- tion V. The variable contributing the most to the function related to support for intramural activities. It is hard to integrate that dimension into the interpretation. But, in addition, the rating of the climate of learning of the hall also had a high weight on the function, as did an item relating to the intellectual and cultural life of the house. Houses with high mean scores on these three items tended to have high mean scores on the function. Likewise houses with low scores on the items tended to have low scores on the function. Thus, one could again postulate that voca- tionally and collegiately oriented students may have found more reinforcement for their value system within houses having a less intense academic academic orientation. Con- versely, residence in more academically oriented houses 194 lay have been detrimental to vocationally and collegiately »riented students. One could postulate the reverse effects ?or non-conforming and academically oriented students. The Theoretical Orientation and Thinking Introver- sion scales are quite highly correlated. One would thus expect the results of the analyses of covariance to be P“? ;imilar for both scales. But where the Theoretical Orien- :ation interaction terms had been significant on tests on :he first and fifth discriminant functions, the Thinking 1‘17 -- ‘- ( Introversion was significant on neither. Thus, the plau- l” sibility of the above interpretations is weakened. Further exploration of the significant differences noted is clearly in order. The point should be made that in general, the four measures of intellectual disposition failed to differentiate with any consistency between groups of students regardless of how they were classified. The types of houses as defined by the discriminant functions made little difference. Nor did students' orientations to higher education, except as noted previously. Perhaps more time would be required for differences, should they exist, to emerge. The fact that respondents completing the four scales in both the pre- and post-testing differed from those who did not complete the instrument the second time may also have been a factor. Perhaps any existing differences would have been more sharply defined had the non-responding students, who tended 195 to have lower gpa's and CQT total scores, been included in the sample. Summarizing, there were no significant differences between the types of houses (described by the five discrim- inant functions) on any of the four measures of intellectual disposition (Hypothesis IIbl). Two of the interaction terms (between college types and types of houses) were significant--both on the tests of adjusted Theoretical Orientation scores (Hypothesis IIb2). Though the direction of differences seemed consistant with theory, the differ- ences may nevertheless have been a product of chance sam- pling fluctuations. The null hypothesis was not rejected for any other interaction in testing Hypothesis IIbZ. The Climate of Learning Iypothesis IIIa: There will be no differences among the houses in residents' perceptions of their house climate of learning. Hypothesis IIIa was tested by a simple analysis of ariance for differences between the mean scores of the 27 ouses on the measure of the climate of Learning. On item 5 of Part I of the HAS, residents were asked to "rate the aneral 'climate of learning'" of their houses on a nine >int scale. Low scores on the item indicated that there LS "support for and/or involvement in the Climate of arning" in the house. High scores suggested "opposition and/or lack of involvement in the Climate of Learning." 196 Group mean scores on the item are found in Table A.1 in Appendix D and are graphically portrayed in Figure 5.2. Results of the test appear in Table 5.24. An analysis of variance of 27 houses on a measure of the climate of learning of the houses (HAS Part I, item 55) Table 5.24. Source of Variance SS df V F Between House Groups 385.764 26 14.837 5.198* Within House Groups 2446.236 857 2.854 2832.000 883 . Total *Significance level <.01 The test was significant well beyond the .01 level. Thus the null hypothesis was rejected and it was concluded :hat there were differences between the 27 houses in res- .dents' ratings of house climate of learning. Differences ranged between the groups from a low ean of 3.86 in house 91 to a high mean score of 6.18 in Duse 23. House 91's score falls between "Moderate'I and Tendency toward" "support for and/or involvement in the .imate of Learning" on the nine-point scale. House 23's ore indicates that there was a "Tendency toward . position to and/or lack of involvement in the Climate Learning" within the house. The mean, 5.007, of the :al sample of 884 residents of the 27 houses fell at the l-point of the nine—point scale. 197 Though not directly related to the hypotheses con- sidered in this section, the distributions of the 27 houses on the measure of the climate of learning reveal certain interesting relationships. These may be observed in Fig- ure 5.2. All three of the houses in hall 4 were rated by their residents as having high climates of learning. Con- versely, the mean ratings of all three houses of hall 3, the sister hall to hall 4, indicated a low climate of learning. The two halls are of very similar design, are located within the same complex, and host similar academic programs. Hall 4 at the time of the study was only in its second year of operation while hall 3 was in its third. Though data are not available, it would have been of inter- est to know whether or not these differences tended to >ersist had more houses been sampled from the halls. Per- laps the differences reflected the influence of a staff .ember(s), the class level composition of residents, or imply the effect of one additional year of operation. House 91 received the most positive climate of earning rating. The other two houses within hall 9 fell ‘- or above the median of houses on the distribution. >use 23 had the poorest rating, though the other two »uses sampled in hall 2 were given relatively positive .tings. The ratings of hall 1 tended to be negative, ough one of its three houses fell at the median of the 198 momcome Amm Emufi .H puma mcmv momson Hams mococflmmu mo mcflcumma mo mumaflao mcu mo madmmma m :0 mmmcon hm mo mmuoom smog mo coflumooa one m.m mucmflm .R on a» us A . $.19 c, 1.0 a mm mm m. w. a. .6“ a... a} m... \m ._.. a...» ._.. o.o m.m o.m m.v o.v oumnocoz pum3ou wocopcme cumBou mocmocme opmumooz mcflcumoq mo mumEHHO onu ca ucoam>ao>ce mcflcumoq mo oumeflau mnu CH mo xoma u0\ocm ou coflufimommo ucmEo>Ho>cH u0\ocm How unommsm =zoq= :ammsumnaH. games: UZHZMdmA m0 madEHAU Y 199 distribution of houses (which incidentally coincided with the mean of the total sample). The houses of hall 1, and to a lesser extent of hall 2, had high mean scores on dis- criminant function II. A high score on that function was interpreted to indicate a low level of satisfaction with the hall compared to other halls. The ratings of houses within the other halls in the sample on the climate of learning measure fell on both the high and low sides of the continuum» It would appear that house climate of learning is E primarily a function of the house rather than a function of a more pervasive climate permeating the hall (though in some halls, such as 2 and 4, the reverse may have been true). It is also significant to note that the climate- of-learning halls (halls 1-4) did not emerge as a group having more positive climates of learning than the other halls. In fact the most sharply defined difference was noted between two climate-of—learning halls (3 and 4) lo- cated within the same complex. The interpretation must of course be qualified in :hat the ratings were based on residents' responses to a :ingle questionnaire item of unproven validity (though 'esponses were generally consistent within a given house). ikewise within each hall a limited number of houses was ampled, though the selection was random. 200 Certain other observations should be interjected at this point though they do not directly relate to the stated hypotheses. Residents were asked to respond to the question (HAS Pt. I, #56): "What has been the level of the 'climate' which YOU have personally experienced through those with whom you associate the most in the house, re- r} gardless of the general climate of the house?" The total group mean on the item was 4.114, indicating that their IFSLW. .- ‘ 1mg _ _l"'. experience tended toward some "involvement in the Climate of Learning." It would also suggest that their personal experiences were more positively directed toward a climate of learning than their group experiences, as was indicated by their ratings of the house climate (M=5.007). One interpretation is that the discrepancy in the responses to the two items was a function of peer group norms. As has been suggested, these norms tend to be anti-intellectual and would dictate that group behavior coincide with the lOI'II‘lS. But such an atmosphere would not necessarily govern Ln individual's posture in more intimate relationships. n alternate explanation would be that residents tended not 0 reveal themselves to others as being in opposition to ntellectual involvement. To do so would probably be in— >nsistent with their self-concept. Residents were also asked where they would “per- lnally like the level of the 'climate' to be in" their use (HAS Pt. I, #57). The mean of the responses was 201 2.752 indicating a "tendency toward" or "moderate" "support for and/or involvement in the Climate of Learning." Unless there were other factors operating, it would seem that residents were generally willing to support a climate of learning somewhat more intense than that which they indi- cated to exist in their houses. But apparently their will- raq ingness did not reflect a level of motivation sufficiently -i strong to overcome the situation, or to lead them to move from the house during the year. The disparity between the ratings would suggest however one possible reason why the if? annual attrition from the halls is relatively high. Hypothesis IIIb: There will be no relationship between the cohesion of the houses and the per- ceived climate of learning. Product-moment correlations between the four mea- sures of cohesion and the climate of learning are listed in Table 5.25. Climate of learning was significantly correlated with each item. Positive ratings of the climate were likely to be associated with more cohesive houses (negative correlations in Table 5.25 are a function of the directionality of scoring of the items). The null hypoth- esis was therefore rejected. The four items which by definition were considered to be measures of house cohesion were all moderately inter- correlated in the expected direction. No causal relation- ship between climate and cohesion can be assumed however. Both could be a function of another set of conditions 202 Table 5.25 Intercorrelations of house mean scores on a measure of the climate of learning and on four measures of house cohesion* Variable 2 3 4 5 1. Where would you rate the general "climate '. of learning" of your i} house? .46 -.54 -.73 .75 2. Your level of satis- faction with living in this house. -.50 -.48 .51 ‘33-'er ( 3. When it comes right down to it, I really have little allegiance to either my residence hall or my house. .85 -.75 4. I would prefer to move to a different house. 5. There are 8 to 12 houses in your residence hall. Where would you rate your house generally in con- trast to the other houses in the hall? *All of the product-moment correlations are signifi- cant beyond the .05 level. 203 which determined the magnitude of each. Correlations of the magnitude indicated in the table would suggest that though a clear relationship existed between climate of learning and cohesion, there was still much latitude for Apparently certain cohesive houses may not have variation. Nor were all houses with had a relatively high climate. a relatively positive climate of learning necessarily highly cohesive. Perhaps the level of cohesion of a house could best be described by examining houses' responses on items three, four and five as listed in the Table. All are highly inter- correlated. It would be interesting to know the attrition rate in low cohesion houses at the end of the academic year. Hypothesis IIIc: There will be no relationship between mean grade-point-averages of the houses and the climate of learning. The mean accumulative gpa's for the fall and winter quarters for each house were correlated with house mean The resulting scores on the measure of climate of learning. product-moment correlation coefficient was .69 suggesting There- a fairly high relationship between the two variables. the null hypothesis was rejected. fore, {Fhe relationship would suggest that to some degree perceptions of the climate of learning may have residents' been a reflection of the level of academic performance in the house. As was indicated earlier residents were gen- Whether or erally aware of the mean gpa's of their houses. 204 not Hus knowledge biased their responses to the climate questunais unknown. Neither can one ascertain whether or not the relationship would have persisted had affects of academic ability been removed. N‘o test was made of this contingency for the total sample. A test was made, how- ever,<fifthe nature of the relationship for freshmen res- idents. The results are presented in the discussion of Hypothesis IIId. Hypothesis IIIdl: There will be no differences in (ad- justed) academic performance between freshmen residing in houses having a "high" climate of learning and those living in houses with a "low" climate of learning. Hypothesis IIIdZ: Nor will there be an interaction between the level of the climate of learning and students' subculture orientation. The ten houses with the highest ratings on the climate of learning measure and the eight houses with the lowest ratings on the measure were formed into two separate groups for the purpose of testing the hypothesis. Houses mnitted from the test were those with mean scores on the unan located near the center of the distribution of houses. ‘actors entering into the selection of houses were con- xidered in the "Discussion" section following the presen- ation of the results of the test of Hypothesis IIa. able 5.26 describes the resulting sample broken down by ubculture membership. lMean gpa's of freshman residents of high and low Limate houses were compared in an analysis of covariance Q 205 Table 5.26 The number of freshman residents grouped by subculture orientation in (1) houses with a high academic climate of learning and (2) houses with a low climate of learning Subculture High Climate Low Climate Total Orientation Houses* Houses** Vocational 61 83 144 Non-conformist 27 21 48 Academic 70 84 154 Collegiate 20 31 51 Totals 178 219 397 *High climate houses included houses 41,43,53,61,91,21, 22,42,62,72. **Low climate houses included houses 23,31,33,51,73, 81,52, and 83. Table 5.27 The number of freshman residents who completed pre- and post-measures of intellectual dispo- sition (four OPI scales on the Attitude Inven- tory)grouped by subculture orientation in (1) high climate of learning houses and (2) low climate of learning houses Subculture High Climate Low Climate Total Orientation Houses* Houses** Vocational 33 48 81 Non-conformist 22 12 34 Academic 42 44 86 Collegiate 6 18 24 Totals 103 122 225 *High climate houses included houses 41,42,53,6l,91,21, 22,42,62,72. **Low climate houses included houses 23,31,33,51,73, 81,52, and 83. 206 in order to test Hypothesis IIId. MSU Reading Test and CQT total scores were again used as covariates. The re- sults of the test(§ppear in Table 5.28. The adjusted cat- egory mean gpa's are listed in Table 5.29. As was expected differences between mean gpa's of college types were significant. Mean gpa's between the group of high climate houses and the low climate houses did not differ significantly ,(Hypothesis IIIdl) nor was the interaction between the two groups of houses and the four college types significant (Hypothesis IIIdZ). Con- sequently, the null hypothesis was not rejected for either Hypothesis IIIdl or IIIdZ. Discussion Residents' responses to the measure of house climate of learning were not totally independent of the measure of freshman academic performance. These same freshmen whose gpa's were compared in testing Hypothesis IIId constituted a large proportion of the respondants on the HAS. Thus, in testing the hypothesis the possible ramifications of this contingency had to be considered. For instance it was possible that freshmen generally perceived the climate of learning across the houses differ- ently than older students. If this were the case the mean report of the climate of learning would be biased because of the disprOportionately large number of freshmen in 207 TableELZB Analysis of covariance of two-quarter accumula- tive grade-point-averages of freshman residents, adjusted by MSU Reading and CQT total scores, for high and low rated houses on the climate of learning Source of Variance SS df V’ F Between House Groups 1.0637 1 1.0637 3.0306 Between College Types 7.6212 3 2.5404 7.2378* House Groups x College Types .7164 3 .2388 .6804 Error 135.8328 387 .3510 *Significance level <.001 Table 5.29 Mean two-quarter accumulative grade-point- averages, adjusted by MSU Reading and CQT total scores, for groups of freshman residents classi- fied according to subculture orientation and the climate of learning of their houses Subculture High Climate Low Climate Totals Orientation Houses Houses Vocational 2.50 2.35 2.43 Non-conformist 2.50 2.51 2.51 Academic 2.55 2.30 2.42 Collegiate 2.06 1.96 2.01 Totals 2.40 2.28 208 several of the houses. Other hypothetical biasing influ- ences could also be postulated. The product-moment correlation coefficient between house mean responses on the climate of learning measure and the proportion of freshman residents in each of the houses was computed. The resultant -.52 £_indicated a moderate inverse relationship between the two variables. It was therefore necessary to explore whether or not the relationship was a function of the differential perceptions of freshmen compared to older students or of something else. Three comparisons were made. The responses of freshmen on the climate of learning item were compared to those of older students through a simple analysis of vari- ance test. Likewise responses of freshmen in the houses with high climates of learning were compared to those of older students in the same houses. And a similar test was made between freshmen and older students in the low climate of learning houses. Results are presented in Table 5.30. Results of the analyses indicated that in none of the three cases did significant differences exist. It would therefore seem that the responses of freshmen on the measure of climate of learning were at least consistent with the responses of older students. This would be a further indication of the reliability of the climate of learning construct in noting the agreement between the observations of freshmen and older students. 209 Table 5.30 Three analyses of variance of responses to the measure of house climate of learning between freshmen and older residents Source of Variance SS df V F Comparison Using the Total Sample Between Groups .215 l .216 .067 WEE Within Groups 2825.744 882 3.204 1 Total 2825.959 883 __________________________________________________________ l = = 0 = = * ' (Nfrosh 427' Mfrosh 4'99' Nolder grp. 457’ Molder grp. 5'02) I I 1; Residents of High Climate Houses Between Groups 2.008 1 2.008 .749 Within Groups 836.438 312 2.681 Total 838.446 313 = = o = = * (Nfrosh 113' Mfrosh 4'38’ NOlder grp. 201’ Molder grp. 4'21) Residents of Low Climate Houses Between Groups 8.005 1 8.005 2.778 Within Groups 763.598 265 2.882 Total 771.603 266 = = 9 = = * (Nfrosh 133’ Mfrosh 5'59' Nolder grp. 134' Molder grp. 5'94) *"M" represents the mean of the group, "N" the number in the group. 210 Most importantly these results would suggest that in fact where there was a relatively higher proportion of freshman residents in a house, mean ratings of the climate of learning of the house tended to be lower. These find- ings suggest that the nature of the climate of learning ‘within a house is in part dependent on the number of fresh- if? man residents. One could surmise that in this respect the influence of older students in the house tends to be pos- itive. 1 The product-moment correlation between the mean of E’; freshmen by house and the proportion of freshmen in the house was not significant (-.09). Thus there was no indi- cation that a larger number of older students in the house influenced the level of freshman academic performance, even though the climate may have been enhanced. In the test of Hypothesis IIIdl the resultant F between house groups in the analysis of covariance (Table 5.28) reached a confidence level of .079. Though this did not achieve the specified level (.05) which would have lead to rejection of the null hypothesis, its closeness warranted further analysis of the results. The differences in the adjusted gpa's of the groups, even though not significant, were in the expected direction (Table 5.29). High climate houses had an adjusted gpa of 2.40 and low climate houses, 2.28. Likewise, the gpa's of three of the four subculture groups were in the 211 expected direction. Little difference was noted in the non-conformist group adjusted gpa (2.50 vs. 2.51). Per- haps this could be interpreted in light of the hypothesized independence of the institution of this group. Inasmuch as these differences were not significant even though the direction of differences fits a theoretical interpretation F} it may well be that even their direction was a chance sampling fluctuation. There is an intriguing comparison between these ‘mu‘ results and those of the earlier analysis of covariance £’ test of gpa differences between houses differentiated on the first discriminant function (Hypothesis IIal). In the former test the two compared groups of houses reflected residents' ratings of the academic performance of their houses. But in the analysis of covariance test the dif- ferences were not apparent and the null hypothesis was not rejected. A comparison of category adjusted group means revealed no consistent direction in the mean gpa's. In contrast, in this last test, though again the differences were not significant, the direction of differ- ences was as expected and was consistent with theory. This may suggest that had the latter test been more power- ful significance would have been achieved; and the climate of learning found to influence academic performance. The failure of the interaction term to obtain sig- nificance in the test of the effects of college types and I) 212 house climate of learning on gpa does not lend much support to Nasatir's findings.9 His results indicated that the more vocationally oriented students performed less ade- quately in what he had defined as "academically" oriented housing groups. Hypothesis IIIelz There will be no differences on any of ran the four (adjusted) post-test measures of intellectual disposition between freshmen residing in houses having a "high" climate of learning and those living in houses with a "low" climate of learning. Hypothesis IIIeZ: Nor will there by an interaction between 3‘5 the level of the climate of learning and students' subculture orientation. In Table 5.31are presented the results of the four analyses of covariance computed to test Hypothesis IIIe. The four analyses test for differences on the two main effects and on the interaction term for each of the mea- sures of intellectual disposition. Only one significant difference, unrelated to the stated hypotheses, was noted. The subcultures differed on their Thinking Introversion scores.10 However, even this difference was probably a product of chance fluctuation in the sampling distribution. 9Nasatir, Op. cit., pp. 292-293. 10It may be of interest to the reader to note that similar differences were noted on two of the five tests of Hypothesis IIbl on the Thinking Introversion scale (Tables 5.18 and 5.21). 213 Table 5.31 Analyses of covariance of freshman post-test scores on the four OPI scales, adjusted by pre- test scores on the same scales, for high and low rated houses on the climate of learning Source of Variance SS df V F Thinking Introversion Between House Groups .2813 l .2813 .0083 Between College Types 394.9858 3 131.6619 3.8764* House Groups x College Types 47.3860 3 15.7953 .4650 Error 7336.5248 216 33.9654 Theoretical Orientation Between House Groups 6.1857 1 6.1857 .5274 Between College Types 74.0348 3 24.6783 2.1042 House Groups x College Types 61.5253 3 20.5085 1.7487 Error 2533.2645 216 11.7281 Estheticism Between House Groups 1.2289 1 1.2289 .1404 Between College Types 63.5789 3 21.1930 2.4206 House Groups x College Types 2.0864 3 .6955 .0794 Error 1891.0978 216 8.7551 Autonomy Between House Groups 13.1128 1 13.1128 .7115 Between College Types 120.0463 3 40.0154 2.1711 House Groups x . College Types 11.0877 3 3.6959 .2005 Error 3981.1134 216 18.4311 *Significance level = .01 214 The null hypothesis for both parts of Hypothesis IIIe was not rejected. Apparently residence in houses with positive climates of learning as Opposed to somewhat negative climates had no discernible affect on the intellectual dis- position of freshmen. Nor was there any apparent inter- action between the level of the climate of learning of the houses and subculture orientation of residents. This finding is generally consistent with what was noted in the tests of Hypothesis IIb. Summary The results of the statistical analyses have been presented and discussed in this chapter. Each of the hypotheses and related findings are restated below. In addition general observations of the data are summarized. Residents' responses on selected items from the House Analysis Survey indicated that most experienced satisfying interpersonal relationships within their res- idence hall houses. The level of satisfaction with both residents' houses and halls was favorable. But the intel- lectual and cultural life of the houses "tended to be a little weak" as perceived by the residents. And house social programs received relatively poor ratings. The principle concern of house life was most often the intra- nmral program. The next most important concern.was study conditions within the house. Of ten items ranked the 215 Jeastimportant concern noted was most often house-arranged intellectual and cultural activities. Residents did tend to:uwicate a willingness to alter the concerns of their houses if they could realize their personal preferences. Ifijus were the case, study conditions would most often becflfprinciple concern. Arranging cultural and intellec- tual activities would rise to fifth place in.the order of preference. It will not be possible to discriminate among the several residence hall houses on the basis of linear combinations of vari- ables describing their group characteristics. Hypothesis I: The null hypothesis was rejected. Sixteen of the 26 possible roots of the discriminant analysis were sig- nificant. These accounted for 94.8 per cent of the total variance between the 27 house groups. Fifty HAS items were entered into the discriminant analysis. Five of the resulting discriminant functions were interpreted through a study of the variables with high standardized weights. rho attempt was made to interpret more of the functions because of their complex and apparently somewhat artificial character, and because of the relatively small amount of the total variance for which they accounted. The first interpreted function differentiated among the groups primarily on the basis of residents' ratings of :he academic performance of their house. The second was lepicted as a function of satisfaction differentiating 216 between highly rated houses in poorly rated halls and houses in highly rated halls. Residents' ratings of house social life seemed to be related to the level of satisfac- tion along with several other variables. The third function differentiated primarily on the basis of the general rep- utation of the houses. Reputation seemed to have been based on many factors. The forth function separated the groups along a continuum of compliance with residence hall and University regulations. At least some houses tending to be less compliant seemed to have a distinct intellectual focus in certain aspects of their house life. The fifth function may have reflected a general performance rating of the houses, primarily weighted toward intramural activ- ities. Only a limited typology of the houses could be constructed. The relationships between the variables on the discriminant functions interpreted were not always :lear. In addition the functions differentiated primarily Between the houses at the extremes of the continua. Little indication was found suggesting that the ouses differed from one another on the five functions on me basis of some distinguishing physical or program char- There :teristic of the halls in which they were located. lS one exception; on the second function two new living- arning halls tended to have mean scores at one end of e continuum and three traditional mens' halls at the other. 217 The two living-learning units were not as highly rated by residents as were the traditional halls. Hypothesis IIal: There will be no differences in the (ad- justed) academic performance of freshmen differentiated according to the types of houses in which they live (types defined according to clustering of houses along the discriminant functions and/or in the multi-dimensional function space). Hypothesis IIaz: Nor will there by an interaction between types of houses and the pre-test subcul- ture orientation of the residents. The null hypothesis was accepted for Hypothesis IIal. There were no significant grade-point-average dif- ferences between groups of freshmen classified according to the position of their houses on each of the five dis- criminant functions. Even though the first function had differentiated between the houses on the basis of house academic performance, the differences were not evident for freshmen when the effects of ability were removed. The null hypothesis was rejected on only one of the five tests of Hypothesis IIaz. When houses were dif- ferentiated on the basis of their mean scores on the third discriminant function, a significant interaction between the subculture orientations of residents and type of house was noted. Vocationally and collegiately oriented students seemed to perform better in houses with positive social reputations and/or more poorly in houses with somewhat negative reputations. The reverse seemed to be true for non-conforming and academic students. Though the significant 218 interaction may have been the result of a chance sampling fluctuation, the direction of differences seemed consistent with previous research and theory. Hypothesis IIbl: There will be no differences on any of the four post-test (adjusted) measures of intellectual disposition of freshmen differentiated according to the types of houses in which they live. Hypothesis IIb2: Nor will there be an interaction between types of houses and the pre-test subcul- ture orientation of the residents. The null hypothesis was not rejected for Hypothesis IIbl. There were no differences in the intellectual dis- position of freshman residents of houses grouped according to the house mean scores on each of the five functions. Neither was the null hypothesis rejected for Hy- pothesis 11b2, with two conditional exceptions. Out of twenty analyses of the interaction between type of house and residents' subculture orientation, only two were sig- nificant. Both were tests for interaction on the Theoreti- cal Orientation scale, one when houses were differentiated on the first function (reflecting house academic perfor- mance) and the other on the fifth function (general house performance). Hypothesis IIIa: There will be no differences among the houses in residents' perceptions of their house climate of learning. {the null hypothesis was rejected. The houses did differ in residents' perceptions of the climate of their houses. There was some indication that the climate of ' ‘n—mg 219 learning was consistent throughout certain of the halls. But generally, the climate of the houses seemed to be de- pendent upon conditions within the house rather than the hall. The climate of learning within houses in the living- 1earning residence halls considered as a group did not seem to differ from other types of halls. Residents indicated F} that they personally tended to experience a more positive climate of learning than that of their houses. They also seemed willing to support a more dynamic learning climate ; j w; in their houses. There will be no relationship between the cohesion of the houses and the per- ceived climate of learning. Hypothesis IIIb: The null hypothesis was rejected. Climate of learning and house cohesion were moderately correlated. There will be no relationship between mean grade-point-averages of the houses and the climate of learning. Hypothesis IIIc: The null hypothesis was rejected. Climate of learning and mean grade-point-averages of the houses were lmoderately correlated. There will be no differences in (adjusted) academic performance between freshmen residing in houses having a "high" climate of learning and those living in houses with a "low" climate of learning. Hypothesis IIIdl: {ypothesis IIId2: Nor will there by an interaction between the level of the climate of learning and students' subculture orientation. There will be no differences on any of ypothesis IIIel: the four (adjusted) post-test measures 1.33%?!“ Baum 220 of intellectual disposition between fresh- men residing in houses having a high cli- mate of learning and those living in houses with a low climate of learning. Nor will there be an interaction between the level of the climate of learning and students' subculture orientation. Hypothesis IIIez: The null hypothesis was not rejected for both parts ci'Hypothesis IIId and IIIe. There were no significant differences in either grade-point-average or the measures of intellectual disposition between freshmen residents of high and low climate houses. Freshman residents' perceptions of the climate did not seem to differ from the perceptions of older students. This was the case for not only the total sample, but also for the two groups of residents living in high climate houses and in low climate houses. The climate of learning was inversely related to the proportion of freshmen living in a house. But the mean grade-point-average of freshmen grouped by house was not related to the proportion of freshmen in the house. Freshman grade-point-average was :orrelated with the climate of learning. But as noted above the differences seemed to disapear when the influence »f academic ability was removed. CHAPTER VI SUMMARY AND CONCLUSIONS Background and Theory That the influence of the college experience emanates from more than just the classroom and library has become a well recognized, but not completely under- stood fact. More and more have we come to realize that many variables operate in determining not only the con- tent of what the student learns, but also the degree and direction of his personal growth and development. This study sought to examine one such facet of the learning environment to which college students are fre- quently exposed. The general purpose of the study was to inves- tigate the quality and character of group life within mens' residence halls at Michigan State University, and more explicitly, within the residence hall house. A house is the major administrative subdivision within a hall, normally housing about 50 students. The house was viewed as a basic context in which students, par- ticularly freshmen, interact and in which they are intro- duced to peer group expectations and influence. 221 ~wm 222 The theoretical framework indicated that the houses would vary in their characteristics including their climates of learning, despite, what, on the sur- face, would appear to be their homogeniety. Evidence was presented indicating that the nature of the houses may differentially influence students when classified. “3 according to their orientation to college life. These 1, orientations were thought to reflect the subcultures with which students identify. ,2 L 3 The research problem was trifold: First, an attempt was made to deve10p a multivariate description or typology of 27 houses, three in each of nine men's halls at Michigan State University. Second, grade- point-averages and measures of intellectual disposition of freshmen residing in different types of houses were compared. The intent was to determine whether or not the types of houses defined in the typology influenced academic performance and/or attitudes toward the academic experience. Also of concern was whether or not different types of houses and subculture orientations interacted .Ln differentially influencing students. Thirdly, the character and extent of the climate of learning within the houses was studied. Its influence on the academic performance and attitudes of freshmen was examined. Reference group theory, an understanding of the 223 dynamics of small groups, and a theoretical development of the nature and origins of student subcultures and peer group influence provided the theoretical framework within which the study was conducted. Conditions determining whether or not a house might function as a reference group were proposed. Propinquity, positive interpersonal ffli relationships,a sufficient length of time for a normative I system to develOp, a capacity within the group to satisfy various needs of residents (particularly those associated J _ with status, acceptance and survival in the academic com- E munity), leadership, and group cohesion, among others were indicated. Each was shown to be theoretically re- lated to the attractiveness of informal groups and to the degree to which houses might in fact function as reference groups. An integration of the theories suggested that freshmen, in particular, would tend to identify with their houses in order to cope with the ambiguities and anxieties created by the demands of the college experience. Fresh- men's need to quickly adapt to their new circumstance was seen as reinforcing the normal emergence of group struc- ture. However, there was no theoretical basis for assum- ing that the outcomes of interaction within a house, even over extended periods of time, would necessarily produce a dynamic climate of learning. On the contrary, there were indications that the result could be contra-educational, 224 thatlunme life might serve an adaptive function in in- sulating younger students from a too rapid advance into the rigors of academic life and from.advancing too rapid- ly into maturity. The review of the literature suggested several specific dimensions of small groups that were to be con- sidered in order to understand the nature of house life. Research was also cited suggesting that college housing groups could in fact develOp a group normative structure which could differentially influence residents. Research at Michigan State University indicated that differences existed between residents of different halls in their attitudes and perceptions of university and resident hall life. Instrumentation and Sample Five different instruments were employed in the collection of the data for the study. These included the Michigan State University Reading Test and the College Qualifications Test, both of which had been administered to new students prior to the beginning of the academic year. Both are measures of academic aptitude. Pre- and post-test measures were obtained from the research sample on the Attitude Inventory which con- sisted of four scales from the Omnibus Personality Inven- The OPI was originally designed to assess tory (OPI). 225 the growth and development of college students. The four scales selected for inclusion in the Attitude Inventory were thought to measure the intellectual disposition of students. These scales were Thinking Introversion, Theoretical Orientation, Estheticism, and Autonomy. Their significance lay in the consideration of the in- fluence house life might have in altering intellectual attitudes of freshmen. The forth measure used consisted of descriptive paragraphs of four hypothetical student subcultures pos- tulated by Clark and Trow.1 Each of the subcultures re- flected a different orientation to college. These were the vocational, non-conformist, academic, and the col- legiate. Freshmen's pre-test self-descriptions on the paragraphs were used in testing several hypotheses in the study. The final and most important instrument, the House Analysis Survey (HAS), was locally developed as a means of assessing the salient characteristics of house and, to a lesser extent, residence hall life. It con- tained 128 items of various types, including a post-test measure of students' subculture orientations. The lBurton R. Clark and Martin Trow, "The Organiza- tional Context," In Theodore M. Newcomb and Everett K. ‘Wilson, College Peer Groups, Chicago: Aldine Publishing Co.,‘1966. 226 majority of items were used to obtain residents' ratings of various dimensions of house life. As part of the in- strument, "climate of learning" was operationally defined and several questions were directed toward its assessment. Available reliability and validity for each in- strument was reported. Reliability coefficients for the p CQT were in the .90's and validity coefficients in the 11 .70's. Reliability of the MSU Reading Test was in the .70's and was correlated with the CQT at .70. Reliabil- ity coefficients in the .70's and .80's were reported for the four OPI scales, though little validity inform- ation was available. Indications of construct validity of the four subculture orientations were in evidence. The HAS items were found generally to be internally con- sistent measures of house characteristics. The nine resident halls included in the study represented the various types of accommodations and pro- grams for men at Michigan State University in the 1964- 65 academic year. The three houses included in the study from each of the nine halls had been randomly selected. Freshmen were generally randomly assigned to houses, though returning students could express a room, hall, and roommate preference. During the first weeks of the academic year the Attitude Inventory and the measure of student subcultures were administered to the residents of the 27 houses. 227 Ninety-five per cent of the freshman residents partici- pated. Well into the winter quarter residents were again tested on the Attitude Inventory and also on the House Analysis Survey. Unfortunately usable HAS responses were obtained from only 60% (N=884) of the total number of residents (1481) then living in the 27 houses. Sixty-one per cent of the total number of freshmen (669) living in the halls participated, providing 406 usable pre- and post-test measures on the Attitude Inventory. Freshmen who failed to participate in the post-test were compared to those who completed the Attitude Inventory in the post-test session through a series of tftests.- No dif- ferences were found between the means of the two groups on the four pre-test measures of intellectual disposition, nor on the MSU Reading Test. The group participating in both test sessions, however, scored significantly higher on the CQT total and in their mean accumulative fall and ‘winter grade-point-average. Differences were also noted between the groups, in a chi-square analysis, in the distribution of subculture orientations of the residents. It was evident from the findings that results, particu- larly those employing the measures of intellectual dis- position where the sample was greatly reduced, would have to be cautiously interpreted. 228 Methodology and Results Three sets of hypotheses were develOped in order to probe the nature and impact on freshmen of differences between residence hall houses. Each hypothesis was an extension of theory and/or previous research on the nature of groups and the college student and environment. Hypothesis I--Resu1ts According to the first hypothesis, it would not be possible to discriminate among the houses on the basis of linear combinations of variables describing house charac- teristics. Multiple discriminate analysis, a statistical technique which maximizes the ratio of the variability between groups to the variability within groups was used in testing the hypothesis. Rettig had suggested the po- tentially fruitful use of the technique in studying the covariation among social groups on several characteristics. The technique produces multiple solutions or roots equal in number to the lesser of the number of variables included cur the number of groups compared minus one. Each root is prOportional to the total amount of between-group variance accounted for in the analysis.3 2Salomon Rettig, "Multiple Discriminant Analysis: In) Illustration," American Sociological Review, 29:398-402, 1964. 3 Ibid., p. 399. 229 Sixteen of the 26 possible roots of the discrimi- nant analysis were significant. These accounted for 94.8 per cent of the total variance between the 27 house groups. Fifty items from the House Analysis Survey (HAS) were en- tered into the discriminant analysis. These items were thought to represent an optimal combination of the vari- ables assessing group characteristics in the HAS (the capacity of the computer program treating multiple dis- criminant analysis was limited to 50 variables). Question A.--Three questions were raised as exten- sions of the first hypothesis. The first (Question A) was an inquiry into the interpretations of the linear combina- tions of variables (discriminant functions) resulting from the discriminant analysis. Each discriminant function cor- responds to one of the roots of the solution of the analy- sis; each successively extracted function is perpendicular to the others in the multi—dimensional function space. The standardized weights indicate the relative contribution of each variable on each function. These weights were thus used in interpreting the nature of five of the sixteen significant functions. These five accounted for 66.7 per cent of the variance between the 27 houses. No attempt was made to interpret more of the functions because of their complex and apparently somewhat artificial character. 230 In addition, the uninterpreted functions individually accounted for relatively small amounts of the total variance. The first function accounted for 28 per cent of the variance. It differentiated among the houses primarily on the basis of residents' ratings of the academic perfor- mance of their houses. Residents' perceptions of their house academic performance seemed to be based on the com- parative quarterly standing of the house within the hall in mean grade-point-average. House standings were reported to each group quarterly by the Residence Hall Programs Office. The second function was depicted as an indication of satisfaction with one's residence hall and house. It accounted for 14 per cent of the variance. Relatively highly rated houses in poorly rated halls were differen— tiated from houses in highly rated halls. Satisfaction with the social life and program of the house was associ- ated with hall satisfaction. The third function differentiated between the groups primarily on the basis of general house reputation. Rep- utation was apparently based on many factors. A good rep- utation particularly seemed more a product of non-academic variables, such as the social program of the house and traditions, than of academic performance. In some houses having a poor reputation the rating of academic performance ‘was high. The latter finding may have been a manifestation 231 of an anti-intellectual normative tendency of the student culture. The function accounted for 11 per cent of the variance. The fourth function seemed to separate the groups along a continuum of compliance with residence hall and University regulations. At least some houses tending to comply less seemed to have a distinct intellectual focus within certain aspects of their house life. The function accounted for seven per cent of the variance. The fifth function was more ambiguous and accounted for only 6.5 per cent of the variance. It was interpreted as reflecting a general performance rating of the houses, primarily, though not exclusively, weighted toward intra- mural activities. Question B.--The question raised was whether or not it was possible to develop a meaningful typology of the residence hall groups according to the results described above. Only a limited typology of the houses could be con- structed following the lines of the above descriptions of (the functions. The relationships between the variables on the interpreted functions were not always clear. In addition the functions differentiated primarily between the houses at the extremes of the continua. Question C.--The question was raised as to whether or not the position of houses on the discriminant functions I! Evil! ...IM. (.5 E 232 would relate in a meaningful fashion to any distinguishable physical or program characteristic Of the halls in which they were located. The mean (centroid) scores Of each house on the five functions determined the relative location Of the house on each function. The results were generally negative with the excep- tion Of the distribution Of houses on the second function. Houses within two new living-learning halls tended tO have mean scores on the continuum indicating a relatively poor residents' rating. Houses within three traditional mens' halls tended to cluster at the Opposite end of the second function. These halls received more positive ratings by their resi- dents. Hypotheses IIa and IIb--Resu1ts The second set Of hypotheses considered the possi- ble existence Of differences in academic performance (Hy- pothesis IIa) and in post-test measures Of intellectual disposition (Hypothesis IIb) between freshmen living in houses Of different types. House types were defined ac- cording tO the location Of house means on each Of the five significant discriminant functions. Five pairs Of houses .had thus been created consisting Of those houses falling, respectively, at the polar ends Of the continua. 233 Using a 2 x 4 analysis Of covariance design the effects Of types Of houses and subculture orientations on the dependent variables could be studied. The design also provided for an analysis Of the interaction between house types and orientations. The dependent variable in Hypothesis IIb was freshman gpa adjusted by MSU Reading Test and CQT total scores. The dependent variables in Hypothesis IIb were freshman post—test scores on the four measures Of intellectual disposition, each considered separately. These measures were each adjusted by pre-test scores of the same scale. Hypothesis IIal.--There were no significant gpa differences between the groups Of freshmen classified ac- cording tO the position Of their houses on each Of the five discriminant functions. Even though the first function had differentiated between houses on the basis Of their academic performance, the differences were not evident for freshmen when the effects Of ability were removed. Hypothesis IIa2.--Only one Of the five tests for interaction on freshman gpa between subculture orientation and type Of house was significant. In the one significant interaction found, houses, differentiated primarily on the basis Of their social reputation (Function III), apparently influenced residents in different ways according to their subculture orientation. Vocationally and collegiately 234 oriented students had higher mean adjusted gpa's in houses with gOOd reputations and/or lower mean gpa's in houses with poorer reputations. The results were reversed for academically oriented and non-conforming freshmen, perhaps related tO the fact that in some houses residents' rating Of house academic performance was inversely related to house reputation. Hypothesis IIbl.--There were no significant differ- ences in the intellectual disposition Of freshman residents Of houses grouped according tO types Of houses. Hypothesis IIb2.--There were no differences in the tests for interactions between types Of houses and subcul- ture orientations on the measures Of intellectual disposi- tion--with two conditional exceptions. Of the twenty anal- yses Of the interaction, only two were significant. Both were tests Of differences on the Theoretical Orientation scale. One difference was noted in the test between houses which were differentiated on function one (primarily a function Of academic performance); the other in the test between houses differing in their scores on the general performance function (V). In both cases adjusted mean Thebretical Orientation scores Of "collegiates" and "voca- tionals" were higher in houses which tended tO have low academic performance ratings. Their scores tended to be lower in houses which residents rated as having a higher IES‘E 3:- 235 level Of academic performance. The situation was reversed in considering the scores Of academically oriented and non- conforming students. H otheses IIIa-IIIe, The Climate O Learning--Results Hypothesis IIIa.--Residents' perceptions Of the climate Of learning Of their houses were significantly dif- ferent. The results were based on a simple analysis Of variance between house mean scores on an HAS item rating the climate. Generally, the climate Of learning within the houses was apparently a function Of conditions within the house rather than within the hall, though there was some indication that the climate Of houses in certain halls was consistent throughout those halls. Responses to another HAS item in- dicated that residents tended tO personally experience a more positive climate Of learning through their personal associations in their houses than would be indicated by their general rating Of house climate. In addition, res- idents seemed willing to support a more dynamic learning climate in their houses, though apparently their level Of motivation was not sufficient to bring about change. Hypothesis IIIb.--House mean scores on the measure Of climate Of learning were moderately correlated with mea- sures Of house cohesion. 236 Hypothesis IIIc.--Climate Of learning and the mean gpa's Of the houses were moderately correlated. Hypothesis IIId and IIIe.--Hypotheses IIId and IIIe were similar to Hypotheses IIa and IIb in that the same statistical model (analysis Of covariance) was employed in the tests for differences. But in IIId and IIIe the house groups (first main effect) were delineated according to whether their climate Of learning received high or low ratings. The possibility Of differences in freshman aca- demic performance and in post-test measures Of intellectual disposition, respectively, between houses with high climates Of learning as Opposed to houses with low climates was treated in Hypotheses IIId and IIIe. The interaction between house climate and subculture orientation was investigated. NO significant differences were found in either adjusted gpa or in any Of the four measures Of intellectual disposition between the two groups. Nor were any Of the interactions between house climate and subculture orienta- tion_significant. Freshmen gpa's were found tO correlate with house mean scores on the climate Of learning. Also, the climate was inversely related to the prOportion Of freshmen in the house. In order to determine whether responses to the climate Of learning were a reflection Of perceptual dif- ferences between freshmen and Older students several ‘h-A' .‘ 237 comparisons were made with results as follows. Considering the total sample, freshman perceptions were not significantly different from those Of Older students. Neither did the perceptions of freshmen and Older students living only in the high climate houses differ. Nor did the two groups differ when only residents Of low climate houses were con- sidered. It was concluded that the climate Of learning Of a house was in part a function Of the proportion Of fresh- men in the house; but that the perceptions Of freshmen and Older students were consistent with one another. ‘Finally, though freshman gpa was correlated with the climate, as was noted in the test Of Hypothesis IIId, those differences seemed tO disappear when the influence Of academic ability was removed. Limitations Of the Study Several limitations Of the research design have been implied or explicitely stated in this report. The most important Of these are here recapitulated. 1. Results should not be generalized beyond the men's residence halls, their residents and programs at Michigan State University for the 1964-65 academic year. The gen- eral environment, student bodies, and residence hall pro- grams vary extensively among institutions of higher edu- cation. Even at Michigan State University since the period when the data were collected, significant changes have 238 been introduced in the residence program, including the establishment Of residential colleges and modifications in the student government structure within the halls. 2. With the exceptions Of the College Qualifications Test and the Michigan State University Reading Test, all the instruments employed in the study were experimental in nature. Their validity in assessing that for which they were designed has yet tO be fully substantiated in each case. For example, the descriptive paragraphs Of student subcultures are thought tO describe hypothetical constructs which have been shown to have a degree Of construct validity. But they call only for a single response from the student whereby he classifies himself. In addition, the specific paragraphs used in this study were modifications Of two earlier versions Of the test. There were some indications that students' responses on the version used in this study would not be completely consistent with the versions used in previous research. Thus results from the assessment Of college types reported herein must be caustiously compared to other research. The four scales from the Omnibus Personality In- ‘ventory have been revised since the collection Of data. The House Analysis Survey, though data Obtained from its incorporation in this study proved highly useful, is in its infancy. 239 The question may also be raised as tO whether or not the instruments in general measured with a sufficient degree Of sensitivity the variables under consideration. 3. The interpretations of the five significant dis- criminant functions are tentative. As was discussed pre- viously there is a certain artificial element in considering the functions as descriptions Of the real world. Likewise, the large number Of interrelated variables and several groups included added tO the difficulty in interpretation. 4. Freshman responses on the measure Of house climate of learning may not have been totally independent Of their grade-point-averages or Of their responses on the measures Of intellectual disposition. Several comparisons suggested that they were indepent, but the possibility of some con- tamination can not be discounted. In testing the hypotheses related to the discriminant functions or to the house types differentiated by the discriminant functions, the question of independence should not be at issue. Class level was included as one Of the variables analyzed in the discrim- inant analysis. 5. The interpretation Of all findings in the study must be qualified to the extent that they were based wholely or in part on data gathered directly from house residents. Only 60 per cent Of the residents completed the House Anal- ysis Survey. And though the items included in the discrim— inant analysis generally seemed tO have a relatively high 3 I 240 level Of internal consistency, it is not known what differ- ences would have been recorded if a more complete response were available. Similarly, only 61 per cent Of the fresh- man sample completed the pre- and post-test measures Of intellectual disposition. Conclusions The results Of specific tests Of the hypotheses plus insights gained in considering the data as a whole have led the author to several conclusions. These are pre- sented in the following paragraphs and, where appropriate, are related to theory and the research discussed in Chapters I and II. General Observations 1. Residence hall houses differ extensively in their group characteristics along several dimensions. 2. Residence halls also differ extensively, though not necessarily along the same dimensions as the houses. 3. The nature and extent Of differences in residence halls and houses strongly argues for a multi-variate stat- istical approach as was proposed by Selvin and Hagstrom and by Rettig and such as that incorporated in this study.4 4Hanan C. Selvin and Warren 0. Hagstrom, "The Empir- ical Classification Of Formal Groups," American Sociological Review, 28:399-411, 1963; Rettig, Op. cit., p. 398. 241 4. The House Analysis Survey and/or similar approaches tO the assessment of inter-house and hall differences ap- pear tO be highly useful technique for studying the char— acteristics and impact Of residence hall life. The many limitations Of the instrument and its need for refinement are recognized. In its current form the variables measured among other things are too molecular. Items can be refined and scales develOped in order tO more accurately assess molar characteristics of house life. Parsimony in-the results derived from the instrument as it now stands is lacking. Peer Group Norms and Influences 5. Obvious undertones Of a pervasive anti- or at least non-intellectual behavioral norm impinging on students gen- erally were Observed in the study. When students were directed to rank several concerns or house activities, participation in intra-mural sports was ranked first. Sponsorship Of intellectual activities within the house was tenth and last.' In contrast, the second place ranking, given tO a concern for study conditions within the house, was probably a reflection Of what Hodgkins described as a necessity in order to meet the minimal level Of compli- ance demanded by the institution.5 Thus, whether he liked 5Benjamin H. Hodgkins, "Student Subcultures--An Analysis Of Their Origins and Affects on Student Attitude and Value Change in Higher Education," (unpublished dis- sertation, Michigan State University, 1964), p. 72. 242 it or not, in order for a student tO remain in the insti- tution, he had to achieve at least to a specified level or be withdrawn from the environment. The environment itself Offered many more enticing rewards (at least for some) than the attainment Of grades. In the description Of the third discriminant function there were indications that house reputation is a function Of social and other non-intellectual variables. Individuals and even significant prOportions Of the residents Of a given house could elect not tO comply with the norm with various outcomes as will be reviewed later. The degree Of compliance with the norm was probably consistent with one's subculture orientation. In fact compliance with such a norm may be a tangential way of viewing the subcultures. Consistent with the theory postu- lated by Clark and Trow, vocationally and collegiately oriented students are those inclined toward non-intellectual values.6 Non-conforming and academically oriented students are more likely tO violate the norm through their acceptance Of academic values and participation in intellectual en- deavors. The degree Of departure would be more pronounced in the non-conforming subculture. The academics, who in addition identify with the institution, would tend to depart from the norm to a smaller degree. 6Clark and Trow, loc. cit. 243 How Do Houses Differ? 6. Indications from the location of houses on the discriminant functions suggest that house differences are primarily products of forces within, and characteristics of, the house groups rather than of the halls. Otherwise, relationships between houses within a given hall would have been more consistent. As it was, on several of the measures in the study two houses within the same hall seemed to be polar opposites. Thus, it is concluded that house life represents an important level of interaction to be consid- ered in the study of college environments. 7. Houses differ from one another as described in the interpretations of the discriminant functions. Distin- guishing characteristics Of several houses were portrayed. But like factor analysis, the differences that were evident could reflect no more than the relationships between the specific variables analyzed. Nevertheless, on an a priori basis it was concluded that the variables investigated represented significant dimensions of house life (with certain limitations heretofore noted). 8. Houses differ along a function related to the aca- demic performance of their residents. But the differences in academic performance primarily reflected the mean input of academic ability into a house; not the influence per se of house environment. Lila. . c 244 9. A tentative conclusion is reached that some houses with relatively high levels of academic performance de- emphasize intra-mural sports. Likewise certain houses with lower academic performance ratings seem to place a heavier emphasis on intra—muralactivities. At least some houses with a more positive academic performance rating have a comparatively poor reputation. Both conclusions are con- sistent with that reached in conclusion 5 above. 10. Students are more satisfied with some types of halls than with others. Those halls receiving more favor- able ratings tend to have more established traditions and positive social programs. They also tend to be similar in their physical characteristics, being older and somewhat traditional structures. 11. Halls receiving a less favorable rating were both of the same type and were in their first year of operation. The reasons for the dissatisfaction are obscure, though it may relate to a physical characteristic such as the partic- ular suite arrangement in the low rated halls; it may relate to the programs of the halls which differed from other halls studied (the programs centered around course offerings in two different academic divisions of the University); it may be a function of the hall staffs and their philosophies; and/or it may simply be a function of the newness of the halls. In the event that the latteris a factor, student personnel administrators may wish to consider the impact 245 of dissatisfaction on residents during the first year of operation. 12. Houses differ in the degree of residents' compli- ' ance with institutionally and hall imposed regulations. Negative compliance seems to be related, in part only, to a more predominant intellectual influence in one or more houses. Deviant behavior within the halls probably takes many forms. At one end of the continuum might be behavior characterized as intellectual rebellion and would represent a considered infraction of regulations distantly analogous to civil disobedience. Students ascribing to intellectual values, particularly the non-conformist, theoretically would be more inclined to disregard rules viewed as in- fringing upon his personal freedom. At the Opposite end of the hypothetical continuum would be behavior motivated by immaturity and would tend more toward prankishness and occasionally maliciousness. Less mature students and par- ticularly those tending to reject the intellectual life would be more inclined in this direction. If this hypoth- esis is valid it may explain why certain houses (where less compliant behavior is tolerated) have a strong intellectual orientation. The leadership of the resident advisor is seen to be a factor in more compliant houses, though no indication of the nature of his leadership is available for analysis. 246 13. Houses differ in their climates of learning as rated by residents. This point will be discussed more extensively later in the chapter. 14. Houses differ in their level of cohesiveness. Why Do the Houses Differ? The underlying reasons behind differences observed between the houses are more obscure than the fact that the houses do differ. The study was meant to be descriptive; and conclusions reached as to why they differ are made cautiously and tentatively. 15. Houses, with a few notable exceptions, do not differ as a function of hall characteristics on the variables as- sessed in the study. 16. There are only inconsistent indications that the programs of the living-learning residence halls produce differences on the variables studied. This is not totally consistent with Olson's findings in his comparative studies 7 The incon- of Michigan State University residence halls. sistency may be a function of the level of analysis and the variables considered. 7LeRoy A. Olson, "Methods and Results of Research on Living-Learning Residence Halls," (paper read at the annual forum of the Association for Institutional Research, Boston, Mass., May 3, 1966). 247 17. Houses differ, at least in their levels of aca- demic performance, from one another as a function of the academic ability of residents. However as was pointed out academic ability must be regarded as an input variable and generally does not reflect student growth and development as a product of house life. 18. The leadership of the resident assistant is a positive factor in determining the level of compliance with institutional regulations within a house. 19. Physical characteristics, age and tradition seem to result in variations in some of the variables assessed, though the results are inconsistent. It is quite possible that what is reflected represents to some degree a general hall gestalt developed over the years. Though the indica- tions of a pervasive hall "personality" are slight, the possibility is consistent with the situation in the Harvard houses, as described by Vreeland and Bidwell, which were characterized as each having a "personality."8 20. One finding leads to the conclusion that the pro— portion of freshman (or conversely upper-division) students is one determinant of the climate of learning as perceived by residents. The higher the proportion of freshmen, the lower the climate of learning. This would argue against 8Rebecca Vreeland and Charles Bidwell, "Organiza- tional Effects on Student Attitudes: A Study of the Harvard Houses," Sociology of Education, 38:233-250, l964-65. 248 all-freshman houses or halls to the extent that climate of learning, as assessed in this study, represents a desirable condition. Chesin reported no differences in freshman attitudes and performance related to the proportion of freshmen in a house.9 His results do not necessarily contradict the findings of this study. The current research represents a very different approach to the issue. 21. The data are insufficient to thoroughly probe underlying bases for differences. Theory would however indicate that many other variables are in part determinants of house characteristics. Since no evidence to the con- trary is available, one would assume that houses are influ- enced by variables such as the informal leadership emerging - within a house (whether or note that leadership coincides with the formal leadership such as house officers and the resident assistant). Houses are influenced by other input characteristics of the residents themselves, such as the fortuitous distribution of residents' subculture orientations, or of specific skills and abilities (e.g., athletic prowess or academic ability, which has already been considered). 9Sorrell E. Chesin, "The Differential Effects of Housing on College Freshmen," (unpublished dissertation, Michigan State University, 1967). 249 The patterns of interaction between residents may vary as a result of both residents' characteristics and the somewhat fortuitous pattern of room assignments, and so on. What Is the Impact of These Differences? 22. There is no evidence that differences in the types of houses defined by the discriminant functions influence the academic performance of freshmen when the effects of ability are removed. 23. 'There is no evidence that differences between types of houses influence the intellectual disposition of freshmen. 24. The adjusted academic performance of academically oriented freshmen may be better in houses with a compara- tively poor social reputation but a higher level of aca- demic performance. Their performance may be worse in houses with a lower academic performance but strong social programs. The reverse may be true for collegiately and vocationally oriented students. The conclusion lends some support to the findings of Nasatir and of Selvin and Hagstrom that living units may differentially influence residents.10 A 10David Nasatir, "A Contextual Analysis of Academic Failure," The Social Review, 71:290-298, 1963; Selvin and Hagstrom, lac. cit. 250 The Climate of Learning 25. The climate of learning does vary significantly among houses, but the differences to some extent reflect the ability level of the residents. Nevertheless, resi- dents did agree with some consistency as to the level of the climate within their houses. 26. The climate of learning has no demonstrable ef- fect on freshman academic performance when the influence of ability is removed. 27. The climate of learning has no demonstrable ef- fect on freshman intellectual disposition. 28. Nor is any interaction between subculture orienta- tion and the level of the climate indicated. The differ-3 ences in the climate of learning between the houses, though significant, were not large. It may well be that the dif- Eferences are not sufficiently pronounced to produce change in the variables studied. It may also be that the impact of the climate of learning lies in dimensions other than those investigated. 29. The climate of learning is related to house co- hesion though, as suggested by Stogdill, both cohesion and the climate of learning (as a measure of group productivity --in Stogdill's terms) are products of the input-character- istics of the residents, rather than functions of one another.11 11R. M. Stogdill, Individual Behavior and Group Achievement (New York: Oxford University Press, 1939), Pp. 13' 271-2720 251 Houses as Reference Groups 30. The variation in residents' levels of satisfaction and their degree of identification with their houses sug- gest that the houses do function as reference groups for some but not for all residents. Within a house conditions may be such that a general level of cohesion, satisfaction, and group iden- tification may be very high, indicating that the house has become a reference group for at least the majority of its residents. In other houses conditions may be such that the house is literally little more than a place where residents sleep. 31. The houses generally seem to foster positive interpersonal relationships between residents. As indi- cated by Festinger, Schachter and Back, this is a pre- requisite in order for an informal group to influence members' behavior and, thus, function as a reference group.12 But need satisfaction was not complete within the houses. Residents indicated dissatisfaction with the social programs and life of their houses (though a given house may have received a high rating). There were indi- cations that many residents were interested in a better 12Leon Festinger, Stanley Schachter and Kurt Back, Social Pressures in Informal Groups (Stanford, Calif.: Stanford University Press, 1950), p. 164. 252 intellectual climate within the houses, though they appar- ently did not Oppose the more prevalent non-intellectual norms of the student subculture in any overt fashion. An.inability to fully meet the social needs of residents may explain why fraternities seem to succeed in demanding and getting a higher level of compliance with their standards of behavior than is the case in the resi- dence halls (at least for the collegiately inclined stu- dent who by definition values social norms to a higher degree than many of his peers). Likewise an inability to fully satisfy intellectual needs, even though residents acquiesce to the social norms of the house, may account for part of the yearly turnover of residents. It would also echo the conclusions of Van der Ryn and Silverstein who felt that too frequently conditions within residence halls alienate the very stu- dents whose presence in the hall (were they to remain) would positively influence the climate.13 32. The above conclusions argue for diversity in pro- grams and opportunities within the houses and residence hall program generally. Rigidity, forced conformity and l3Sim Van der Ryn and Murray Silverstein, Dorms at Berkeley (Berkeley, Calif.: Center for Planning and Devel- opment Research, University of California, 1967), p. 27. 253 narrowness of program may result in the exclusion of many students from an identification with the hall. And the impact of positive house and hall programs would be deluded. 33. The results do not demonstrate any profound out- comes of the residence hall experience in terms of dis- cernible academic and intellectual influence. There were some suggestions that need satisfaction within the house is related to personal growth and development, in addition to general satisfaction with residence hall life. The preceding is not in any way intended to disparage other outcomes of residence hall and house life. But it does underscore the need to more carefully evaluate both the assumed relationship of various hall and house programs to desired outcomes and the philosophical and empirical bases of residence hall Operations. Perhaps the secret of improving the residence hall experience and capitalizing on its strengths lies in four areas: (1) emphasis on existing strengths in terms of promoting and improving programs found to satisfy needs of residents, (2) providing relatively unstructured di- versity of Opportunity for varied experiences as part of the hall programs, (3) legitimizing intellectual behavior to the students as an alternative model to the often in— sulating influence of peer groups, and (4) making explicit to the residents, particularly the freshmen, the subtle norms that do seem to influence their behavior. 254 Suggestions for Future Research In this final section of the study, several avenues for future research are collated. These may help refine the inadequacies of the present study and extend and/or challenge its conclusions. 1. The House Analysis Survey merits further develop- ment and refinement. The items may be combined into scales : through a factor analytic approach. 2. Similar studies could be extended to women's resi- dence halls and to programs on other campuses. A replica- tion at Michigan State using refined instruments would help establish the characteristics of house life. 3. The reasons behind differences between houses should be carefully probed in order to both gain a better understanding of the differences and to learn how charac- teristics such as house norms might be altered or reinforced. 4. The dimensions along which houses and halls differ should be considered more extensively. The techniques employed in this study were gross and lacked the level of sensitivity probably called for. 5. The determinants of the climate of learning in the houses should be explored more carefully and completely. 6. It would be of interest to know how residents who adopt their house as a reference group differ from those who have little identity with it. A corollary question 255 would be whether or not the impact on residents who identify with the hall is more pronounced than was generally noted in this study. 7. Sociometric measures could be employed in future studies as a refinement of the assessment of several of the variables considered to be important in understanding house differences. 8. Lastly, an instrument such as the House Analysis Survey could perhaps be fruitfully employed in action re- search with house residents. The explication of house differences, norms, and values reflected in inter-house and hall comparisons may assist residents in better under- standing the influence of subtle environmental forces on their behavior. Such awareness may be the forerunner of change. 37% BIBLIOGRAPHY BIBLIOGRAPHY Adams, Donald Van. "An Analysis of Student Subcultures at Michigan State University." Unpublished dissertation, Michigan State University, 1965. Anderson, Gary Robert. "An Examination of Residence Hall Counseling Programs for Men in Ten Selected Midwestern Colleges and Universities." Unpub- lished Ph.D. dissertation, Northwestern Univer- sity, 1959. Andrews, Ernest A. "The Residential College Student: A Study in Identity Crisis." Paper read at the Annual Meeting of the American Orthopsychiatric Association, Washington, D.C., March, 1967. Astin, Alexander W. The College Environment. Washing- ton, D.C.: American Council of Education, 1968. Astin, Alexander W. and Holland, John L. "The Environ- mental Assessment Technique: A Way to Measure College Environments." Journal of Educational Psychology, 52 (1961), 308-16. Bennett, George; Bennett, Marjorie G.; Wallace, Wilburn L,; and Wesman, Alexander G. College Qualification Tests, Manual. New York: The Psychological Corporation, 1961. (Revised edition). Boyer, Ernest L., and Michael, William B. "Outcomes of College." Review of Educational Research, 35 (1965), 277-91. Boyer, Ronald K. "Student Peer Group: Its Effect on College Performance." Unpublished dissertation, Case Institute of Technology, 1965. Brown, Robert D. "How Compatible are the Intellectual and the Academic Aspects of College Life? An Investigation of the Relationship Between the Intellectual and the Academic Aspects of College Life." Paper read at the American Personnel and Guidance Association Convention, Dallas, Texas, March, 1967. 256 257 Brown, Robert D. "Manipulation of the Environmental Press in a College Residence Hall." Personnel and Guidance Journal, 46 (1968), 550-60. Bryan, Joseph G. "The Generalized Discriminant Function: Mathematical Foundation and Computation Routine." Harvard Educational Review, 21 (1951), 90-95. Cartwright, Dorwin. "The Nature of Group Cohesiveness." Group Dynamics. 3rd ed. Edited by Dorwin Cart- wfight and Alvin Zander. New York: Harper and Sn, Row, 1968. ' Centra, John A. "Student Perceptions of Residence Hall Environments." East Lansing, Michigan: Office , of Institutional Research, Michigan State Univer- sity, June 13, 1966 (mimeographed). Chauncey, Henry. "Excerpt of Remarks." Annual Meeting ‘ of the College Entrance Examination Board, t” October 28, 1964. Unpublished report, Educational Testing Service. Chesin, Sorrell E. "The Differential Effects of Housing on College Freshmen." Unpublished dissertation, Michigan State University, 1967. Chickering, Arthur W. "College Residences and Student Development." Educational Record, 48 (1967), 179-86. Clark, Burton R., and Trow, Martin. "The Organizational Context." College Peer Groups. Edited by Theo- dore M. Newcomb and Everett K. Wilson. Chicago: Aldine Publishing Co., 1966. Clingan, Wallace Eugene. "The Educational Effects of an Accelerated Personnel Program in the Men's Residence Halls at the University of Missouri." Unpublished dissertation, University of Missouri, 1959. Computer Institute for Social Science Research, Michigan State University, "DISCRMZ," Technical Report No. 33, by Stuart Thomas, July 5, 1968. Darley, John G.; Gross, Neal; and Martin, William E. "Studies of Group Behavior: the Stability, Change, and Interrelations of Psychometric and Sociometric Variables," Journal of Abnormal and Social Psychology, 46 (1946), 565-96. 258 Davis, James A. "Intellectual Climates in 135 American Colleges and Universities: A Study in 'Social Psychophysics.'" Sociology of Education. 37 (1963), 110-28. Diener, Charles L. "Similarities and Differences Between Over-Achieving and Under-Achieving Students." Personnel and Guidance Journal, 38 (1960), 396-400. Dollar, Robert. "A Study of Certain Psychological Dif- Fr“ ferences Among Dormitory, Fraternity, and Off- ” Campus Freshman Men at Oklahoma State Univer- sity." Unpublished dissertation, Oklahoma State University, 1963. Eisenstadt, S.N. "Studies in Reference Group Behavior." Readings in Reference Group Thesry and Research. Edited by Herbert H. Hyman and Eleanor Singer. New York: Free Press, 1968. r"' Etzioni, Amitai. A Comparative Analysis of Complex Organizations. New York: Free Press of Glen- coe, 1961. Festinger, Leon; Schachter, Stanley; and Back, Kurt. Social Pressures in Informal Groups. Stanford, California: Stanford University Press, 1950. Fisher, Ronald A. "The Use of Multiple Measurements in Taxonomic Problems." Annals of Eugenics, 7 (1936), 179—88. . "The Statistical Utilization of Multiple Measurements." Annals of Eugneics, 8 (1938), 376-86. Forehand, Garlie A. and Gilmer, B. von Haller. "Environ- mental Variations in Studies of Organizational Behavior." Psychological Bulletin, 62 (1964), 361-82. Fritz, Roger Jay. "A Comparison of Attitude Differences and Changes of College Freshman Men Living in Various Types of Housing." Unpublished disser- tation, University of Wisconsin, 1956. Gardner, E. F., and Thompson, G. G. SocialLRelations and Morale in Small Groups. New York: Appleton- Century-Crofts, 1956. 259 Gold, David. "Some Comments on 'The Empirical Classifi- cation of Formal Groups.'" American Sociological Review, 29 (1964), 736-39. Golembiewski, Robert T. The Small Group. Chicago: University of Chicago Press, 1962. Gottlieb, David, and Hodgkins, Benjamin. "College Student Subcultures: Their Structure and Characteristics in Relation to Student Attitude Change." The School Review, 71 (1963), 266-89. Hagstrom, Warren 0., and Selvin, Hanan C. "Two Dimen- sions of Cohesiveness in Small Groups." Socio- metry, 28 (1965), 30-43. . Halpin, Andrew W. and Croft, Don B. The Organizational Climate of Schools. Chicago: Midwest Adminis- tration Center, University of Chicago, 1963. Hardee, Melvene D. "Personnel Services for Improving the Campus Climate of Learning." Journal of the National Association of_Women Deans and Counselors, 24 (1961), 122-7. Hartnett, Rodney, T. "Place of Residence as a Factor in Academic Performance Patterns of College Students." Unpublished dissertation, Michigan State University, 1963. Hemphill, John K., and westie, Charles M. "The Measure- ment of Group Dimensions." Journal of Psychology, 29 (1950), 325-42. Herbert, David J. "The Relationship Between the Per- centage of Freshmen on a Residence Hall Corridor and the Grade Point Averages of the Occupants." College and University, 41 (1966), 348-52. Hodgkins, Benjamin Joseph. "Student Subcultures--An Analysis of Their Origins and Affects on Stu- dent Attitude and Value Change in Higher Edu— cation." Unpublished dissertation, Michigan State University, 1964. Horst, Paul. ”A Generalized Expression for the Relia- bility of Measures." Psychometrika, 14 (1949), 21-31. Hyman, Herbert H. "The Psychology of Status." Archives of Psychology, No. 269, 1942. 260 Hyman, Herbert H., and Singer, Eleanor, eds. Readings in Reference Group Theory and Research. New York: Free Press, 1968. Jacob, Philip E. Changisg Values in College: An Ex- Study of the Impact of Coliege Teaching. New York: Harper Bros., 1957. Jacobson, Eugene. "The Growth of Groups in a Voluntary Organization." Journal of Social Issues, 12 (1956), 18-23. Jencks, ChristOpher S. and Riesman, David. "Patterns of Residential Education: A Case Study of Har- vard." The American Collsge. Edited by Nevitt Sanford. New Yrok: Wiley & Sons, 1962. Jones, Lyle V., and Bock, R. Darrell. "Multiple Dis— criminant Analysis Applied to 'Ways to Live' Ratings from Six Cultural Groups." Sociometry, 23 (1960), 162-76. Kelley, Harold H. "Two Functions of Reference Groups," Readings in Reference Group Theory. Edited by Herbert H. Hyman and Eleanor Singer. New York: Free Press, 1968. Kidd, John W. "An Analysis of Social Rejection in a College Men's Residence Hall." Unpublished dissertation, Michigan State University, 1951. . "The Students Live." Evaluation in the Basic College at Michigan State University. Edited by Paul L. Dressel. New York: Harper, 1958. Lecky, Prescott. Self-Consistensy. The Shoe String Press, Inc., 1961. Lehmann, Irvin J. Dressel, Paul L. Critical Thinking, Attitudes, Values in Higher Education. Coopera- tive Research Project No. 590, Department of Health, Education and Welfare, Michigan State University, 1962. Lindzey, Gardner, and Urdan, James A. "Personality and Social Class." Sociometry, 17 (1954), 47—63. 261 Lozoff, Marjorie M. "Personality Differences and Resi- dential Choice." Growth and Constraint in Col— 1ege Students. Edited by Joseph Katz. Stanford, California: U.S. Department of Health, Educa- tion, and Welfare Project No. 5-0799. Institute for the Study of Human Problems, Stanford Uni- versity. .Matson, Robert E. "A Study of the Influence of Fraternity, Residence Hall, and Off-Campus Living on Students of High, Average and Low College Potential." Journal of the National Association of Women Deans and Counselors, 26 (1963), 24-29. f Maxwell, A. R. "Canonical Variate Analysis When the Variables are Dichotomous." Educational and Psychological Measurement, 21 (1961), 259-71. Mayhew, Lewis B. "The Intellectual Tone at Any Univer- sity: Its Progress and Measurement." Journal “ " of the National Association of Women Deans and Counselors, 25 (1962), 156-60. McCracken, C. W. "Student Personnel Work and the Cli- mate of Learning." Personnel and Guidance Journal, 42 (1964), 904-7. Merton, Robert K., and Rossi, Alice Kitt. "Contribu- tions to the Theory of Reference Group Behavior." Readisgs in Reference Grosp Theory and Research. Edited by Herbert H. Hyman and Eleanor Singer. New York: Free Press, 1968. Michael, William B., and Boyer, Ernest L. "Campus En- vironments." Review of Educational Research, 35 (1965), 264-76. "Michigan State University Reading Test." East Lansing, Michigan: Office of Evaluation Services, Michi- gan State University. Morrisett, Lloyd Jr. "Research on College Influences on Personality," Social Science Research Coun- cil Items, 13, No. 3, 1959. Nasatir, David. "A Contextual Analysis of Academic Failure." The School Review, 71 (1963), 290- 98. 262 Nasatir, David. "Social Sources of Academic Failure: A Contextual Analysis." Unpublished disserta- tion, University of California, Berkeley, 1966. Newcomb, Theodore M. "The Prediction of Interpersonal Attraction." American Psychologist, 11 (1956), 575-86. . The Acquaintance Process. New York: Holt, Rinehart and Winston, 1961. . "Student Peer-Group Influence." The American College. Edited by Nevitt Sanford. New York: John Wiley & Sons, 1962. ., and Wilson, Everett K., eds. College Peer Groups. Chicago: Aldine Publishing Co., 1966. . "The Contribution of the Inter-Personal En- _, vironment of Students Learning." NASPA, Pro- *“ ceedings of the 49th Annual Conferenceyq 5 (1967), 175-78. ., and Feldman, Kenneth A. The Impacts of Col- leges Upon Their Students. A Report to the Carnegie Foundation for the Advancement of Teaching, 1968. Nunnally, Jum C. Psychometric Theory. New York: Mc- Graw-Hill, 1967. Olson, LeRoy. "Student Attitudes Inventory." East Lansing, Michigan: Office of Evaluation Ser- vices, Michigan State University, 1964. (Mimeo- graphed.) "Methods and Results of Research on Living- Learning Residence Halls." Paper read at the annual Forum of the Association for Institu- tional Research, Boston, Mass., May 3, 1966. Omnibus Personality Inventory, Research Manual. Berke- ley, California: Center for the Study of Higher Education, 1962. "Omnibus Personality Inventory (Form F) -- Brief Scale Descriptions." Berkeley, California: Center for the Study of Higher Education (mimeographed, no date). 263 Pace, C. Robert. "Methods of Describing College Cul- tures." Teachers College Record, 63 (1962), 267-77. II . College and University Environment Scales Princeton, N. J.: EducatiOnal Testing Ser— vice, 1963. . Preliminary Technical Manual: College and University Environment Scales. Princeton, N. J.: Educational Testing Service, 1963. ., and Stern, George G. "An Approach to the i Measurement of Psychological Characteristics of College Environments." Journal of Educational Psychology, 49 (1958), 269-77. Peterson, Richard E. College Student Questionnaires. Princeton, N. J.: Educational Testing Service, Institutional Research Program for Higher Education, 1965. Prior, John J. "Peer-Group Influence on the College Climate for Learning." Journal of College Stu- dent Personnel, 5 (1964), 163-7. Prusok, R. E., and Walsh, W. B. "College Students' Residence and Academic Achievement." Journal of College Student Personnel, 5 (1964), 180-84. Rafkind, Israel. The Federal Government's College Housing Loan Program. Washington, D.C.: Ameri- can Council on Education, 1966. Rao, Radhakrishna C. "The Utilization of Multiple Measurements in Problems of Biological Classi- fication." Journal of the Royal Statistical Society, Series B, 10 (1948), 159-93. Rasmussen, Glen, and Zander, Alvin. "Group Membership and Self-Evaluation." Human Relations, 7 (1954), 239-51. Ray, William S. An Introduction to Experimental Design. New York: Macmillan Co., I960. Rettig, Salomon. "Multiple Discriminant Analysis: An Illustration." American Sociolegical Review, 29 (1964), 398-402. 264 Ryn, Sim Van der, and Silverstein, Murray. Dorms at Berkeley. Berkeley, California: Center for Planning and Development Research, University of California, Berkeley, 1967. Sanford, Nevitt, ed. The American College. New York: John Wiley & Sons, Inc., 1962. . Where Colleges Fail. San Francisco: Jossey- Bass, Inc., 1967. F Scott, William A. Values and Osganizations: The Stqu_ E} of Fraternities and Sororities. Chicago: Rand . McNally Co., 1965. Seashore, S.E. "Group Cohesiveness in the Industrial Group." Ann Arbor, Michigan: University of Michigan, Survey Research Center, Publication Number 14, 1954. Selvin, Hanan C. The Effects of Leadership. Glencoe, Illinois: The Free Press of Glencoe, 1960. "The Impact of University Experiences on Oc- cupational Plans." School Review, 71 (1963), 317-29. ., and Hagstrom, Warren 0. "The Empirical Clas- sification of Formal Groups." American Socio- logical Review, 28 (1963), 399-411. ., and Hagstrom, Warren 0. "Reply to Davis." American Sociological Review, 28 (1963), 814. ., and Hagstrom, Warren 0. "Reply to Davis." American Sociological Review, 29 (1964), 736-39. ., and Hagstrom, Warren 0. "Criticisms and Afterthoughts." College Peer Groups. Edited by Theodore M. Newcomb and Everett K. Wilson. Chicago: Aldine Publishing Co., 1968. Siegel, Alberta Engvall and Wiegel, Sidney. "Reference Groups, Membership Groups, and Attitude Change." Journal of Abnormal and Social Psychology, 55 (1957), 360-64. Sinnett, E. Robert. "Socio-Economic Status and the Use of Campus Facilities: A Tale of Two Dormitories." Personnel and Guidance Journal, 43 (1965), 933-96. 265 Sokal, Robert R. and Sneath, Peter H. A. Principles of Numerical Taxono_y. San Francisco: W. H. Freeman and Co., 1963. Steiner, Ivan D. "Group Dynamics." Annual Review of Psychology, 15 (1964), 422. Stogdill, R. M. Individual Behavior and Group Achieve- ment. New York: Oxford University Press, 1959. Strozier, R. M. The Housing of Students. Washington, D.C.: American Council on Education, 1950. Tatsuoka, Maurine M., and Tiedeman, David V. "Discrimi- nant Analysis." Review of Educational Research, 24 (1954), 402-20. Taylor, Harold. "Freedom and Authority on the Campus." The American College. Edited by Nevitt Sanford. New York: Wiley & Sons, 1962. Tead, Ordway. The Climate of Learning. New York: Har- per and Bros., 1956. Tiedeman, David V. "The Utility of the Discriminant Function in Psychological and Guidance Inves- tigations." Harvard Educational Review, 21 (1951), 71-80. ., and Bryan, Joseph C. "Predictors of College Field of Concentration." Harvard Educational Review, 24 (1954), 122-39. .; Bryan, Joseph G.; and Rulon, Phillip J. The Utility of the Airman Classification Bettery for Assignment of Airmen to Eight Air Force Specialties. Campbridge, Mass.: Educational Research Corporation, 1951. Vreeland, Rebecca, and Bidwell, Charles. "Organizational Effects on Student Attitudes: A Study of the Harvard Houses. Sociology of Education, 38 (1964-65), 233-250. Yamamoto, Kaoru, ed. The College Student and His Cul- ture: An Analysis. Boston: Houghton Mifflin Co., 1968. 266 Yarosz, Edward J. and Bradley, Harry. "The Relation- ship Between Physical Distance and Sociometric Choices in Two Residence Halls." International Journal of Sociometry and Sociatry, 3 (1963), 42-45. Yonge, George D. "Students." Review of Educational Research, 35 (1965), 253463. "R APPENDIX A FALL QUARTER LETTER OF INFORMATION SENT TO THE SAMPLE APPENDIX A MICHIGAN STATE UNIVERSITY - East Lansing University College - Office of Evaluation Services We .~.' October 1, 1964 Dear Student: . '1' ..r’.‘ ._ l 1 1 The Resident Assistant in your house or precinct will soon be contacting you, if he has not already done so, regarding a special research project in which your house or precinct and two others in your residence hall have been asked to participate. Because it is impossible to contact you individually I am writing this letter to give you information regarding what we are asking of you with respect to your time and involve- ment and, likewise, to explain generally the aims and nature of the experiment. Considering the many millions of dollars expended in con— struction of residence halls at MSU we have very little information, other than "hunches", as to whether one type of hall has any real advantages or disadvantages over an- other. There are many possible variables which could enter into giving a hall unique characteristics that really make a difference as far as students are concerned. Also, it is very likely that what may be beneficial for one student may be detrimental for another. It is factors such as , these that we hope to consider this year. Thus the main purpose of the study is to explore the variations in liv- ing patterns, in likes and dislikes of residents, and in other house or precinct characteristics that may evolve as a product of living in one type of residence hall at MSU as opposed to another. This will be possible only with your assistance. Here is what we ask of you: All of the men in your house will be asked to give about three hours of their time dur- ing the year divided into two separate sessions. The first 267 268 October 1, 1964 Page 2 will last less than one hour and will be held this coming Monday, October 5, at 7:00 p. m., unless your R. A. has made arrangements with you for a different time. He will inform you of the location which will either be in your own house or precinct or in one of the meeting rooms in your resi- dence hall. The second session will be held Winter Quarter at a time yet to be set. This second session will last a little longer than the first, but no longer than two hours. We will meet with your house officers in order to keep you informed of arrangements. Let me stress a few points: 1) The houses selected for the study were picked randomly within 9 MSU residence halls. These 9 represent each of the major types of halls and special programs at MSU. Perhaps this will explain why your house was selected and not another. 2) The special analytic technique we are using and the fact that our interest is with your house as a group make it vital that every resident of your house including the R. A. participate. Otherwise the time, effort, and expense entering into the study may be in vain. The responses and attitudes of a minority could dictate the future programs of a majority. 3) Though you will be asked to put your name on your questionnaire, your responses will be guaranteed complete confidentiality. No one on the housing staff will be al- lowed to see your responses. The research staff will treat them as part of the house group they represent. However, any individual who may wish to discuss the re- sults is cordially invited to contact me Spring Quarter at the conclusion of the study. 4) We urge you to take the matter seriously and give your honest Opinions. If you are one of those who will complete the questionnaire in his own house or precinct rather than in a group, you must do your own work and not discuss the questions or your responses w1th anyone. 5) If special circumstances ab- solutely prevent you from completing the questionnaire at the time set by your R. A., please contact him to make special arrangements. We are trying to take the attitude that just as the Univer- sity tries to teach us objectivity, it must also lead the 269 October 1, 1964 Page 3 way by not being afraid to look critically at itself and its programs. As far as this study is concerned this can be done only with your OOOperation. I sincerely look for- war to your participation. We have attempted to make this as painless for you as possible. Please accept my thanks. Sincerely yours, Robert Standing “3 Project Director ! RS:sja APPENDIX B WINTER QUARTER LETTER OF INFORMATION SENT TO THE SAMPLE Mr APPENDIX B MICHIGAN STATE UNIVERSITY - East Lansing University College - Office of Evaluation Services February 20, 1965 Dear Student: Last week a group of your house or precinct officers and as your Resident Assistant participated in a brief discussion f? of the second phase of the MSU residence hall research pro- 1‘ gram in which your house was involved earlier this year. * I am now writing to request your participation in this second and last phase of the study in order that the pro- ject may be successfully completed. 4.: You may recall that your house plus two others in your hall were randomly selected, along with three houses in each of eight other men's halls for participation in the study. The general purpose of the research is to explore the nature and effects of variations in living patterns, in likes, dislikes and preferences of residents, and in other house or precinct characterists that may result as a product of living in one type of residence hall or house at MSU as opposed to another. We have three goals in mind for the study: (1) to provide informatiOn as to the value of a wide variety of aspects of hall living as a basis for change and improvement of the programs; (2) the gen- eral advancement of the understanding of the dynamics of group living; and (3) immediate feedback of information to the houses and individual participants in the study for their consideration and benefit. Let me clarify the last goal as to how it pertains to you and your house. First, the most important questionnaire you will be asked to complete this week related almost exclusively to your observations and Opinions about the nature e£_house life as you have experienced it this year-- inciuding everything from the athletic program to study conditions. About the first week of Spring Quarter we will provide a tabulation of the responses of your house to your house officers. This hopefully will be used as a basis of consideration of strengths and weaknesses of your house programs. Ne information will be made available which could in anyway be used to evaluate, embarrass or identify any individual. Secondly, an explanation of the questionnaire used last Fall and again this week, plus one's own results, will be available to those who wish to check with me sometime 270 271 February 20, 1965 Page 2 Spring Quarter. I will be available most Monday through Thursday afternoons in 294 Bessey Hall. Though some of you may not find this second questionnaire as interesting as the housing questionnaire, I believe at the completion of the study you will see its value. The two questionnaires will take you about an hour and a half to complete. Though we recognize that this is some- what demanding of your time, we were at least successful in cutting it down from the two hours we indicated to you last Fall. The session will begin 22 7:00 p.m. Please pe_prompt. Below is a list of the times and locations of each testing center for each hall involved in the study. Check the time and place for your hall. Feb. 24 Beb. 25 E. Wilson Wed. Wilson Aud. W. Fee Thur. 137 Fee E. Shaw " East Lower E. Akers " 137 Akers Lounge Emmons " 151 Brody W. Shaw " West Lower N. Wonders " Won. Kiva Lounge Snyder " Dining Ro: Bryan " 151 Brody Snyder With the exception of Snyder, test materials will be available from 6:30 p.m. to 8:30 p.m. in case you would need to begin either earlier or later than 7:00. In Snyder, materials will be available from 7:00 p.m. to 8:30 p.m. Check with your house president or R.A. if you are unable to attend your scheduled session. You could possibly attend one of the other sessions on a different night. There will also be a make-up session during the day Saturday, though we strongly hope you can attend with your own group es scheduled. If for some reason you did not participate in the first part of the study last Fall, we urge you to nevertheless participate this time. In order for the results to be meaningful and of full value to your house, as well as to the study, ge_need virtually 100 per cent participation. 272 February 20, 1965 Page 3 We sincerely solicit your interest and participation. We urge you to take the matter seriously and respond accurately and honestly. Again, we will try to make it as painless yet as profitable to you and your house as possible. Please accept our thanks. Sincerely yours, Robert Standing Project Director RS:gs APPENDIX C HOUSE ANALYSIS SURVEY ph- APPENDIX C IDUBI LIALTBII IUIVIT Office of lveluetion Services lichigan State university, 1203 This survey is designed to provide a hroad description of the characteristics of life within residence hall houses and ecincts. It asks for your oheervations, opinions, and preferences with regard to house and residence EaII living. The inforsetion obtained sill he used to provide an isage of house life--how it differs fros house to house and hell to hell--end insight into sose of the factors which produce the particular characteristics of each house. hopefully, the research will provide clues to the strengths and-weaknesses of the liehigan State University housing progren, whet students like and dislike, and the type of influence one type of residence hall exerts on student life cospered to another. _ In no case will the data he used in anyeay to evaluate any individual or group of individuals. The .tuay is ends for research purposes only. All oases provided will he coded hy the project director and will be known only to his. All data will he treated on e grou hesis. lo inforsstion sill be revealed to any individual which sill indicate his, or anyone else a, status I n s group. There are no right or srong anseers. however, the usefulness of the survey is entirely dependent upon the truthfulness with which the questions are answered. Ie urge you to sake eaoh enseer an accurate reflection of your reel feelings. - ' ' ' . Throughout the questionnaire the word "house“ is used to refer to hoth houses end pm uincts within the residence hells. more»: 1. There are tvo parts to the Survey requiring tso separate enseer sheets ehich are labeled ”Part I” an a "Part II” on the top. In addition, an Attitude Questionnaire sill also he given. leny of you cospleted this test earlier this year. ’If’Is repeated as an isportent part of the study in order to provide indications of your current attitudes after having cospleted part of the school year. 2. Use only the special pencil provided. If you erase do so cospletely. Cosplete the inferna- tion called for at the top of each answer sheet: lane, date, student nusher, ness of your residence hall, and race nusher. - 3. Also write our student nusher in the vertical colusn of blank hoses under the heavy arrow. Then III! 6‘! EPIC! In eacE o! the six roes of ten spaces that corresponds to each digit of your student nusher. 4. SASS can ANS!!! ONLY for each ites, except where indicated. he sure to answer each ites. The responses listed say not coincide exactly with your point of view. In such cases, choose the alternative that is nearest your point of vise. Iork rapidly. 5. Ihen you have finished hoth parts of the louse Analysis Survey go on to the Attitude Inventory. PAST I 3. See seny quarters have you lived in this house, including this quarter? 1. This quarter only 0. Six quarters 1. Age at last birthday: 2. Two quarters 7. Seven quarters l. 18 or under 6. 21 3. Three quarters I. light quarters 2. l7 7. 22 4. Ibur quarters 9. line quarters 3. 18 S. 23 l 8. live quarters 10. Ten or acre 4. 19 2. 24 ’ S. 20 10. 25 or older 4. See seny quarters have you lived in this residence hall? 2. Class in college: 1. This quarter only S. Sis quarters l. First quarter fresh-en 3. Two quarters 7. Seven quarters 2. Second quarter fresh-an 3. Three quarters 8. light quarters 3. Third quarter fresh-an 4. Tour quarters 9. line quarters 4. Low Sophosore (40 to 62 hrs.) 5. Five quarters 10. Ten quarters 5. High Sophomore (63 to S4 hrs.) 6. Junior (SS-129 hrs.) 5. In which College are you currently enrolled? 7. Lee Senior 1. University College 6. Education S. Sigh Senior (sill graduate this acadesic 2. Arts a Letters 7. Engineering year) 3. Agriculture 3. Natural Science or 9. Graduate Student 4. Business lose Iconoeics 10. Other: special, tesporery, etc. 5. Con-unicetion 9. Social Science Arts l0. Veterinary ledicine 273 10. See such education do you tentatively plan to obtain? One year of college The years of college Three years of college bachelor's degree Iaster's degr Professional Sd.D or Fb.D Other ee degree (S.D., Meyer, etc.) Ihich of the follceing categories cones closest to your father's occupation? If your father is retired, deceased, or unesployed, indicate his forser or custosary occupation. Iark only one. . unskilled sorker, laborer, fare worker Sesiskilled worker (e.g.,sechine operator) Service sorker (police-en, fire-an, barber, silitary noncossissioned officer, etc.) Skilled eorker or crafts-en (carpenter, electrician, plusber , etc . ) Sales-an, bookkeeper, secretary, office worker, etc. Seeiprofessional or technician (laboratory or sedical technician, draftssan, etc.) Oener, sanager, partner ssall business, far- or loeer level governsental official; also silitary cossissioned officer Profession requiring a bachelor's degree (engineer, elesentary or secondary teacher, etc. ' Oener, high level executive—-large business or high level governsent agency Professional requiring an advanced college degree (doctor, lawyer, college professor, etc. So. such forsel education does (did) your father have? Indicate onl the hi best level (I.e., sark only one 0! {Es fen aIIernaEIves.) . lo foreal scEEEling Sose grade school Finished grade school Scse high (secondary) school Finished high school business or trade school Sose college Finished college (four years) Attended graduate or professional school (e.g., law or sedical school) but did not attain a graduate or professional degree Attained a graduate or professional degree (e.g.. IA. run, In) 9. 10. lo. Indicate the extent of your sother's forsal education. See the alternatives In the pre— ceding question. Iark only one. Shich best describes your position in your fesily? . l. I as an only child. I. I as the oldest of the children in the- falily. S. I as the y t of the children. 4. I have both o der and younger brothers and/or sisters. ll. l3. 14. Did you request to live with a -2- Ihich of the following best describes the cos-unity which you think of as your boss teen during your high school days? l. Suburb in a setropolitan area of sore than 2,000,000 population 2. Suburb in a setropoliten area of 500,000 to 2,000,000 - 3. Suburb in a eetropolitae area of 100,000 to 500,000 In a city (not a suburb) of sore than ten eillion In a city of 500,000 to 2 000 000 In a city of 100,000 to 500,000 In a city of 50,000 to 100,000 City or two of 10,000 to 50,000 0. Cos-unity of less than 10,000 10. Fare, reach or other open country Ice seny different positions of leadershi , elective or appointive, do you held In caepus organisations (e.g. house, dore, social, religious, etc.)? 1. Sons 4. Three 2. One 5. Feur or sore 3. Tho Ihich of the folloving is correct concerning your present place of residence? ("On-caspus" refers to University housing) 1. The housing Office sede both sy current roos and bell assignsent'ffiis year. I have not lived elsewhere on caspus this year. T53 Housing Office eade sy current rcos assignseut, but I requested to live in this hall. I have no? IIved elsewhere on caspus this year. I requested both ny current roos and hell assign-en! tEIs year. I have not lived elseehere on caspus this year. I requested to sove to this hell after livi eleeehere on caspus this year. Housing ass! and up current house and recs. I requesfea go sove to this hell and house after having lived elseehere on caspus EEIs year. But Sousing assigned sy roos. I requested to sove to this hall and rooe after having lived elsewhere on caspus this year. 7. I scved to sy current roos free a different house In this sass hall earlier this year. S. I soved to sy currenf rocs fros another roos In this sane house this year. 9. Other of your present roe-sates, rather than ing assigned together by Scusing? Tee 1. 2. So 2774 Quest ions l5-34 Questions 15 through S4d list aspects_c1_lifs within the house or precinct is which you live. Sesed on your ohaezzaiinna deninisnso Hsondiiisna_siihia.iha_hauaa_durias the roar..rais_lsan.hauaa on each of the itess, using the 0-point scale described holes: 1 J a 1 a l 4 [L 5 I e I 1 I a l a Incellent Very Good Tends to Satie- Tends to Seek Seer 'Very Foor Good be good factory be a little weak Then, for question 15, if you feel that ”support for and participation in intrasural sports“ "Tends to he a little week," you would darken space 00 on your sewer sheet. by to thin of point ”9' on the scale as the nidpgint between conditions tending to be satisfactory and those tending to be unsatis- factory. 15. Support for and participation in intrasural 25. Ability to study in the house . sports 20. Intellectual and cultural life of the house ”on, lS. Success in intrasural sports ‘"“ 27. Friendlinees eitbin the house fi l7. Level of acadenic perfornance or scholarship N in the house , 2S. Opportunities provided to neet girls i lS. The good tines we have together 20. value of living in this particular house E 10. Isputation of the house eithin the residence :0. Conplience of residents with resident hell 1 hell regulations ; 20. Contribution of life within the house to your 31. Thur satisfaction with your roossete(e)t understanding of issues, ideas, philosophies,etc. - ' 32. Tour general satisfaction with residence hall 2l. Social life and social progras of the house accosodetioas L 22. Support for and participation in the social 33. Tour level of satisfaction with living in progras this house 23. The leadership of the Sesident Assistant 34. Teur level of satisfaction with living in V this residence hall 24. The_leedership of the elected house officers " O 0 O 0 0 O 0 0 O 0 O O O 0 ~ 0 0 Questions 35.44 Salon are listed 10 activities which generally require or invite the concern of the house OIOII AS A IIOLI. Based on whatsyou have observed in your house this year, rank the 10 stateeents in the order of the concern which the group has Shawn, for-ally or infer-ally, for each of the activIIIu Indicate the rank assigned to each of the activities on the anseer sheet. The activity considered to have been IOST IIDOSTAIT is assigned rank #1, and you should darken the first space opposite the nusber of that activity on the answer sheet. The activity which you feel has been LIAST IIIOITAIT to ::e house should be assigned rank 010 and the tenth space opposite the nusber of that activity could darkened. he sure to assign one rank to each of the 10 activities, even though you say find it difficult. You say vIsh In use the nergin OIIyour test booklet to deternine your rankings before you transfer then to the answer sheet. lone-ber, rank according to your observations of the concerns of the 13322. Activities and Problens Ibich Save Concerned the house 35. Arranging and participating in social (45) 40. Providing assistance for individuals' (50) activities problens (study, social, personal) as. participation in and/or discussion of (4s) ‘1- 'Ic'P* '0' *t°" ""°¢ 5185"- ‘5' "S (51’ student govern-eat (hall, AUSS, etc.) ""'d th‘ h°“" " little '9’. than ' place to sleep and eat. Activities ranked 37. Study conditions of the house - (47) loser generally do not concern the sea. 42. Participation as a group or with the Sell (52) 3" "°'tv 1""'“"1' (“) in special events, e.g. blood drives, sitting together at gases, prodecte,etc. 43. Keeping the house clean and tidy (SS) 44. Discussion, enforce-eat and debate of (54) rules and regulations of the house, ball and university SS. Arranging and participating in activities (40) to deepen residents' understanding of issues, philosophies, the arts, etc. Qpeetions 45-54 lbw, re-rank the ten activities above, this ties in the order of what TOU’IOULS IOST ill!!! to be the nest isportant activities of the group. Folloe the ease instructions as above. censider the activities to be nunbered free 45 to 54, as indicated to the right of the itess. use these spaces on your aeseer sheet and assign ranks accordingly. e e e e e e e e e e e e e Questions 55-58 Clinete of Learnin The following paragraphs describe what we will refer to as the "clients of learning" of a house. Read the section carefully and then answer the questions at the end according to your appraisal of your house. 0 e S 0 0 0 0 0 House activities and attitudes on caspus var in the degree which thesem rt or con le:ent the mission of {he University of preparing studen s o understand and deal with e pro lens a the world in which they live. Think of this degree of support as l i alo a line, at one end of residents, perhaps entire houses, whose activities stron 1 en rga ciigiie of learning; at the other end, houses or subgroups of residents who are no? oniy uninvoived‘c in such a clients but who also strongly resist its influence. The descriptions to follow are not neant to inply that social life, athletics, and other activities conflict with a "clients of learninET' Such progress say or say not operate effectively regardless of the clinate. Also, students say legitinately feel that their life within the residence hall is their own to lead as they see fit and that ”learning" is properly confined to the classrooe and library. Here are descriptions: "High” Clinete of Learnisg Visualize a group of residents or an entire house where the excitenent of learning, experiencing and growing literally abounds. Here exists an lnost continual exchinge of ideas, attitudes, dis- cussions of arf iorss, new discoveries in science, political controversy, confrontation and discussion of values. "Bull sessions" are often deep and stinulating. Cultural activities, such as the Lecture- Concert Series and Provost Lectures, are strongly supported. Fresh-en in the house rapidly have their intellectual horizons broadened and stinulated. Discussions of classrooe topics continues well beyond the walls of the classrooe. “Low" Clinete of Learnisg At the other extra-e, learning is generally left to the classrooe. It is not that residents don't study outside of class or work for their grades. It is Just that little, if any, of the intellectual life of the University carries over into the life of the house. “3311 sessions" seldon have intellec- tual depth or substance. Attenpts to etinulate nore enlightening activities are seldon supported, and one who does night be regarded as a 'bighbrow’ and out of touch sith his house-ates. Such a house say he a satisfying place to live because other characteristics of the house or subgroup possess great value for the residents. Social, fraternal or athletic activities say he proninent. But it is alsoet as though a social care existed against too such involvesent in acadesic learning. Selection of classes is often based on the ease with which one can get by. Freshnen soon learn the ways of the group and confers. Though they indicate concern over their studies, they are readily distracted free then. The 'Inbetveen“ Case Between these two extrenes one can visualize a third group or house whose activities and attitudes neither stron 1y support such learning experiences and intellectual excitenent nor 0 pose thee with any consisfency. For such a group house life nay seen to be independent of the "nission of {he University.” However, our feeling is that subgroups or entire houses tend {o iean sore one way than the other, though elements of both sides nay exist in any given group at any given nonent. CLIMATE OI LEARNING "Sigh" . "Inhetween" "Low" ‘1 l 2 3 4 5 5 7 S 2 " “ . 41> Strong Fairly loderete Tendency Inbetween Tendency Ioderate Fairly Strong and open Strong tovard toward strong and open support for and/or involvesent opposition to and/or lack of 1’ in the Clinete of Learning involvement in the Clinete of Learning ~_ \ I Questions 55. There would you rate the general 'clinatc of learning' of your house? 55. That has been the level of the "clients" which YOU have personally experienced through those with whoa you associate the scat in the house, regardless of the general clients of the house? 57. There would you personally like the level of the "clinete" to be in your house?. There would you rate the general "clients of learning” of the residence hall in which you live? -4- 275 PART II Now begin use of the second answer sheet for the House Analysis Survey. Be sure you have filled in all the information called for at the top of the sheet: name, date, student number, name of your residence hall, and room number. lxcept where indicated, answer the questions according to the following scale. 1. Agree strongly 2. Agree, but not strongly 3. Uncertain 4. Disagree, but not strongly 5. Disagree strongly l. I feel like I've gotten to know the sen in the house fairly well. 2. Life in the residence hall or residence hall complex detracts from my interest in the total university. I find it difficult to feel a part of such a large university. Residents’ behavior in the house is not a 4. problem. 5. I think I would have done better academically so far this year had I lived in a different house. 6. Residents of the house keep their rooms clean and neat. I feel that fellows in the house are too involved in cliques. 8. lore efforts should be made to increase school spirit at Michigan State. 9. Ihen it comes right down to it, I really have little allegiance to either my residence hall or my house. 10. A number of campus leaders live in the house. 11. Ihen I‘go to an activity such as a show, concert, ball game, etc., I usually go with or doubledate with guys from the house. 12. Iy room is generally quiet enough for effective study. 13. I as active in one or more extra-curricular activities. 14. I would enjoy having faculty members visit informally with the house occasionally in order to discuss ideas, issues, their interests and work, etc. . 15. Comparatively speaking, our house is known for some of its original, novel or creative (though perhaps somewhat questionable) ideas and activities. 16. I feel that I as generally accepted and appreciated by those who live in the house. 17. I often get together with other students in the house to discuss issues raised in classes. 18. Students in the house exhibit a high degree of concern for the rights of others. 19. There isn't anyone in the house with whom I would particularly want to discuss a personal problem. 20. 21. 27. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. Living in my house is a major factor in making me feel a part of this university. It is important for the house as a group to participate in and support all-campus competition and activities such as homecoming, water carnival, blood drive, etc. Iany in the house tend to be more concerned about the amount of work required in a course or how easy it is to get a grade rather than the quality of the instructor or the contri- bution of the course to the individual. I have learned to enjoy one or sore cultural activities this year that I had not really appreciated before. IFih There are a number of traditions in the house. Students are entitled to have a stronger voice in the detersination of university policy than they now have. Facilities such as grills and lounges are major sources of distraction fros my studying. laintenance of college traditions is an important aspect of college life. The I’intellectual" enjoys little status in the house in which I live. .l, ‘3‘ ‘ Residents in the house have been involved in an above-average number of disciplinary problems. I think of myself more as a member or resident of the residence hall, than as a member or resident of the house. Every one living in the house would be welcome to all house activities. The most important think I can get out of college is training for a vocation. The most isportant think I can get out of college is an exposure to ideas, people, social problems, philosophies, etc. Residents in the house are expected to parti- cipate in all house sponsored aciivities. There aren't many fellows in the house who would be willing and interested to talk about issues, ideas, etc. The university should be concerned about the moral behavior of its students. Topics of “bull-sessions" in the house are superficial rather than of depth or substance. The sen in the house would be more likely to compliment someone on a nonacadesic (social, athletic, etc.) achievement than on an academic or intellectual achievement. House meetings are of little value. The social program of the house provides good opportunities to meet some of the ”sharper" girls on campus. Students in the house have high ethical standards with respect to cheating, etc. There's quite a bit of pressure (subtle or otherwise) in the house to participate in house and university activities. REV: 43. 44. 45. 46. 47 48. 49. 50. 51. 52. 55. 1. Agree strongly 2. Agree, but not strongly 3. Uncertain 4. Disagree, but not strongly 5. Disagree strongly Residence hall regulations are generally reasonable. There really isn't such interest in inter- national affairs, social issues, or scientific discovery expressed among the residents of the house. Residence hall regulations allow students ample freedom. ly house has effective means of dealing with residents whose behavior isn't acceptable to the group. I would prefer to move to a different house. I would prefer to move to a different resi- dence hall. I would prefer to move off campus. O O O t t . How many ISU football games did you attend this fall? 3. Two 5. Four 1. lone 2. One 4 . Three I vote or intend to vote in various campus elections. 1. Very often 2. Often 3. Occasionally 4. Seldom 5. Almost never Do you hope to Join or are you now a member or pledge of a fraternity? 1. Yes 2. Not sure 3. No . Iy free time is spent with fellows from the house. 1. Almost always 2. About 3/4 of the time 3. About 1/2 of the time 4. About 1/4 of the time 5. Almost never How often have you attended special lectures and seminars this year, such as the Provost Lecture series, programs of different political groups, departmental seminars and colloquia, etc.? 1. Very often 2. Often 3. Occasionally 4. Seldom 5. Almost never Row often have you attended concerts, plays, series, travelogues, etc., this year? . Very often . Often . Occasionally . Seldom . Almost never GQUNH 57. 59. 60. 61. 62. -6- I participate in "bull-sessions" in the house. 1. Very often 2. Often 3. Occasionally 4. Seldom 5. Almost never what proportion of your closest male friends at ISU live or have lived this year is your house? 1. Almost all 2. lost 3. About half 4. A few 5. Almost none what proportion of your closest male friends at '80 live or have lived this year in your residence hall (including your house)? 1. Almost all 2. lost 3 . About half 4. A few 5. Almost none There are S to 12 houses in your residence hall. where would you rate your house eral in contrast to the other houses in 1. One of the best 2. Better than average 3. About average 4. Below average 5. One of the worst where would you rate your residence hall compared to the other 14 men s 1. One of the best 2. Better than average 3. About average 4. Below average 5. One of the worst what degree of influence has the Resident Advisor had on the men of your house? 1. A very positive influence . Some positive influence . Little or no influence . Some negative influence . A strong negative influence cubes” what degree of influence has the Resident Advisor had on you? 1. A very positive influence Some positive influence 3. Little or no influence 4. Some negative influence 5. A strong negative influence what degree of influence have the residents of your house had on you? 1. A very positive influence 2. Some positive influence 3. Little or no influence 4. Some negative influence 5. A strong negative influence PLEASE GO ON '10 m WIN PM}! 276 Questions 64-67 kinds of things. find in any American college. In every college that we know of. there are different kinds of students who enjoy doing different Listed below are some comments or descriptions about the kinds of students you might Read each of these over and then answer the questions which follow as best you can. we know that it is difficult to "peg" yourself in some slot but please make a choice for each of the four questions. TYPE 1: This kind of person views education principally as a means of preparing for his professional future. He is not particularly disinterested in the social or purely intellectual phases of campus life, though his participation compared to some may be limited. This person does his homework but tends to do little outside reading or restricts his reading to the light, general entertainment variety. All things considered, this person‘s primary reason for being in college is to obtain professional training. TYPE 2: This person is interested in learning about life in general, but in a manner of his own choosing. He is very interested in the world of ideas and books, and eagerly seeks out these things. Outside of the classroom, this person would attend such activities as the lecture-concert series, Provost lectures, foreign films, etc. This person often pursues his own interests in place of or in addition to mere course requirements and will frequently do extra readings in order to obtain a more complete understanding of the world in which he lives. From a social point-of—view, this person tends to reject activities such as fraternities, sororities, and the social events that many consider a part of campus life. when this person does join, it will usually be one of the more intellectual, academic or political campus organizations. For the most part, this person would consider himself to be someone who is primarily motivated by intellectual curiosity. TYPE 3: This person is in some respects like Type 2 noted above. He is concerned with books and the pursuit of knowledge, but is also the kind of person who leads an active social life on campus. He is interestEH'in getting high grades and tries to maintain a high grade- point average. He is the kind of person who will eagerly work with student or hall govern- ment, fraternities, committees, and activities of this type. He would feel that both the social side of college life and the academic are important for his general development. TYPE 4: This is the kind of person who is more concerned with the social phases of college life and learning to get along with individuals. He identifies closely with the college and enjoys attending as many campus social and athletic events as possible. This person may be inter- ested in intellectual kinds of things but will, for the most part, find greater satisfaction in student government, parties, activities, etc. Be is concerned about his education but feels that the development of his social and leadership skills are certainly important. luch of his college life will be centered around non-academic type activities such as committees, fraternities or sororities, or resident hall type activities. This person will try to maintain his grades but does not feel that he must necessarily make the highest grades or go out of his way to do extra or non-assigned readings in order to be a success in college. Now that you have read each of the four descriptions, answer the following questions. Indicate your answers by darkening the space by the number of the type on your answer sheet. 64. 65. 66. 67. which of the above types comes closest to describing the kind of person you consider yourself to be? '~ which of the above is least descriptive of the kind of person you consider yourself to be? which of the above types comes closest to describing the kind of person you would like to be if you had a choice? which of the types is most descriptive of the majority of men in your house? i O I t a t t a a a List the residents in the house when you would most enjoy having as your roommate. Name as few or as many as you like. (This information will be kept in strictest confidence by the director of the project and will be coded for purposes of analysis.) Please use first and last names. PLEASE GO ON TO THE NEXT PAGE 69. whom do you consider to be the real leaders of the house? Name as many as necessary. 70. Please comment briefly on any other aspects of house and/or residence hall life which you feel have been inadequately treated in the questionnaire. 60 on to the Attitude Inventory. -8- APPENDIX D TABLES OF HOUSE AND TOTAL MEAN SCORES AND STANDARD DEVIATIONS ON HOUSE ANALYSIS SURVEY ITEMS coo; om... MK.— wh; ow; m6.— Nmop mm." mm; X...— mm; 2.; SJ 8: so; Fe; NM." mm; 5.9 5: Ne; on; 8... mm; 9.; to; ms; 2.; mm m2...“ mm.m mm.m 56 em...“ mag 3.... 5...." em; 33. mm.m :6 No.6 volt. {Hm v~.m $6 56 5...” :.n 3.0 mo.m mm.m 8.9 2.0 «Tm mo.» mnfi 3.. m mn...~ sew twow oo.~ anon ow.~ vw.~ Noom : m~.w mo.~ ms; S.~ mm." 8.9 mm; 8.9 S: o~.~ cw.~ oo.w 2.; ~m.~ 3.~ $.N ov.w n~.~ aw gain. 36 pmom Con 36 56 mm.m 8.0 mm F m .p m mp.m mm..." nm.m mmé cm...” 8..» $6 33. 8.m 85 3.0 56 35 oo.o mmé «NJ 3.m man 3.... 3.. a 8.3 :.~ 8S 8.... SJ 8.... 8;" new 22.. Ed 8.~ z.~ ~..~ 2S m3 8.... 12 BS and 33 SJ «.3 3..“ mg 23 new 8.~ m.,: 3 8w.” 8.6 86 :6 23 N...“ 8; E6 86 26 8.... 5.... 8.... 8... 3.... New 3.0 8.9 a.» 2.... as $6 :5 «to :6 3.“ :3 2...“ f a 0. mm; mm.— mm; 3;; 2V.— 9: mo; co; 2; mp; om. me; mm. mm.— mm. 5." :3 mm. mo; no.9 3.9 3. NF; mm; p 3.»... 8.? $.v mm.m :6 moo.m 3...... pm.m 5.... mm.v min 3: 8... m~.¢ 13.3...» wm.v mm.v mm.¢ mgr mo.m mw.m mm.n 5: mm... 80.~ we; 8; mm; 3." ~v.~ No.9 2. mm.— mm.~ o¢.~ S.N mm.w moon om.~ om.— 3... ...m... on; $4.. we; so.~ mm; no; mm. 8. mm. mm. .5 oo~.m .m~.~ eo.m gum vo.m 55 Fw.~ oWN $.m mm...” 3...” ..N.m wo.m I; mo.v mm.m :5 $.N :.v Chm mm.~ 23m mm.~ mm.w S.N.. oo.~ mo.~ mm.— f v m8.- em. 3. — on; B; an; 09— 5. mm; m¢.~ afim NN.~ m~.~ mm.N :.N om.— a." no; mm.~ cw; co; 8. mm. mm. 00. 5.0 00; am mmm.~ mN.N mm.~ P $.— m.m vv.m 8.~ oQN mo.m wm.~ min oqm oo.v wm.m «We Nm.m mo.m Nm.m mm.m 55 we.” mv.~ v~.~ oo.N mm.— mo.~ oo.~ mm; S.N.. as. m $o.~ mo.~ oo.~ mo.~ mm.~ vm.~ 8.9 mm.— mv.~ wm.~ mm; mm.~ FN.N mute. om.— mm.— 2.. ~55 mm.m 8.m mm; mw.m 83 $2.. ~N.m 88 mm.m mv.m ~—.m nw.v 3; 33¢ ~¢.v NW... NO) 0) e emim wo.~ 51o. :J I; 2., .5 m mo v 3... Sum wmom mo.m .2 N m e me- (.0 as e mam... Fm; mm; em; 5., om... mm. mm; pm; 2; 3; mm; mm.— Km; mm." on.— mm. mm. mm. mm... 8., mo; we." 3; or; P... 2.— 2: am mt... om.m om.m 3%... NF; 2.... mm.m 53¢ 00;. we; em...” N«.m mm.¢ ow... we... on... 2%.. mv.m 38 mv.v {an 84.. 3; mm; 2; mm.m mm.m Son 3.. — town: own: mm... mm»: mm... mm». we... on»... 3... pm»: we... rm... mm... no... we... mm... mm»: mm... mm... mm... 5.... mm... mm... mm... mm... mm... we": mm»: «.02 a8 #3. mm mm 5 8 we 5 2 N~ 2. mm mm 8 mm mm 3 no. Nv 3 mm mm 5 8 mm Pm mp NF Z 5: wwaor momsom mm one no some now mo>usm mammamsd owner one we H puma Eoum mauve ou momsommou mo mcoflumw>o© pumpsmum pom memo: H.¢ cases 0 XHDmemd 2778 Fmo.~ 3.9 mw.~ S.~ ofim mm.~ 2..— ms; 2; we; 3: o~.N :.~ 8; 2.... ooou v—.~ oo.~ ~o.~ mwo— 1.; 8; I; m—.N 8; Na; 8; av.N 5 NS...“ o~.o gen 5;. ~o.m 3.... 5.... -.m I.» 9.0 ~o.m No.0 sqm 2; 07m 36 mm; 8: an; mo.m me... New 2.0 m—.m 8.v 86 moot oN.m g cm 2L; 3; vim em; an; cm; 2.; Nu; noon 2....— moop Eon 8." on; on." 5; me." 3: mm." 9..— ~m.— on; 2..— m—.« E..— mv; 5; 3.9 gm moqv oo.m mm; ~m.w m—.m mN.N om.v $.v oo.r N¢.m mN.¢ mo.w qu mm.~ Fm...“ mm.m o~.m ..o.m $.m mm; pw.¢ 31¢ 00.0 mo; 26 S..." mm.m Z.m .2 mp Nmo.N mm; mm.~ mm; ~N.N o~.~ oo.~ no; mm; mm; ~N.N flop em; cm; 3: mm; oo.~ mm; S.~ B; an." mo.~ n~.~ B.~ mm; mm; 8.” m¢.~ aw mmm.v Bum mm.v mo.v «mi. mm.m 23¢ ~p.¢ mqv om.m mm.m pm; we; 8.? min vwom nw.m 8.0 mflm nmov ~v.v mm.v $.m mm.m 3.0 writ 85 omoe .2 mp mmm.— mm; mm; mm. mm; 3.9 m~.—. no; NNJ we; 5.9 pm; mo; 2.: mm. 3." to; co; mo.— «m.— mm; mo.— on; 2.39 3; 5... en; 3.? 3 SN; mN.m :3. me." mm.¢ om.m mm.m so.» mm...” ES 85 mm.m 5N4” ~7~ mo.¢ m~.m mw.~ we.» 8.— Neon mN.v mm.m pm; 2...... 3....” mm.~ afim I; .2 : m9.~ mm; am." No; 8; 2.; on; 3.; mm; mm; B; ~v.— 5; we; we; mm." mm.— 8;. mm; mm; mm; mm; ~¢.r mm. mm; 2.... or; 5; .5 m5; o_..m 3.0 33m mN.m mm.m Fm.~ no; omit. I..m No; Z..m 3.: man to..." 23¢ 3.? on.n 8.5 smon mwov mm)“ mn.m mv.v EHN 26 mm; :6 .2 2. gm; mm; km; mm; mm.— om; m—.~ mm.— o~.~ pm; 2.; smow Np." wo.w av; 3..— mm; on; ow; mm.— we; 2; em; to; m...— on; $3 8... am 89v 3.... ES om.m om.m Omsm :3.” o}. 0min we; 5.m 3.0 mnor vm.¢ moi». co...“ oqm 5.N pan mm.v min mm.v mm.m swim :.~ om.m NQN :6 x). m— oon. me. on. on. 0*. on. Ne. Ne. Fe. on. we. me. mm. we. mv. me. me. me. we. av. we. we. av. on. on. me. me. on. on own; co... me; me... mm; mm; mm; 3; 2..— 3: S; on; m"; cm." mm... on; co; om; 5.9 cm.— Z..— 8; 5.9 we; we." 5;. moor an; .2 1 N3. ms. mm. 8. —o.— om. ms. 8. No. mm. S. 3. 2.. 5. No. No. S. 2.. 8. mm. 3. we. em. mm; 5. mm. we; 8.— on own." we; mm.— mm.— mm; 3.9 mm; vmop so; mm; :o— mm; 2.; pm... me... ~v.— am... we; SJ on; we; mm; mm; 8...... mm." mm; C..— va 8.... NF com-z owuz mNaz mmaz mmaz Nmaz Neuz 09.2 3.2 5.2 mvaz emu-2 -a2 nvaz ovaz nmaz mmaz mmuz mmuz mNaz 3.2 mmuz emu: mmaz mun: mmaz ~12 one: .02. a5 ._.op mm mm 5 mm 3 S Q. NN 2. mm mm 5 mm mm 5 me we pt mm mm ..m 8 N «N 9 Np up 5:. muse: Acmssflucouv H.< manmh 279 000.0 00.0 00.0 00.0 00.0 8.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0.0.0 00.0 00.0 00.0. 8.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 8.0 0.0.0 .0 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 2 00 000.0 8.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 8.0 00.0 00.0 00.0 00.0 08. 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 .0 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0. 00.0 00.0 00.0 00.0 00.0 00.0 8. 0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 f 00 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0200.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 90 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00. 0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 .2 00 000.0 00.0 00.0 00.0 0 0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00. 0 0. .0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 o0 000.0 00.0 00.0 00.0 0 .0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 8. 0 00. .0 00. 0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 f 00 000.0 00.0 00.0 00.0 00. 0 0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 .5 000.0 00.0 00.0 00.0 00.0 0 0 .0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 8.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 f 00 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00. 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0 o 00 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00 0 .00 0 000.000.0 . 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0.00 . 00. 0 00. 0 00. 0 00.0 00.0 00.0 00.0 00. 0 8. 0 .2 00 000.0 00.0 00.0 00.000.000.000.000.000.0 8.000; 00.0 00.000.000.00 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 n.0 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 8.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 8.0 00.0 00.0 00.0 00.0 00.0 x: 00 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 f 00 000.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 a. .02 ._5 .80 00 00 00 00 8 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00. 5.: 0000: Awmscflucoov H.« magma 280 8a.. 3.. m... 3.. o... E... .N.~ I... a... o... 3.... F... F... .o.~ 8.. .QN N... #0.. 8.. 8.. 8.. «a; 8.. Bo..o..~ no.~ o..~ c..~ .8.n on...” 39.0 3...... $.v-.n.n on; at; on.» um.» 36 8.... ...N 3.» n... no; 3.» om.~ on.» 3.» o..v 3; SJ 5: 8.} on; an.v e..v £8 ..m.. 3.. 8.. 3... 3.. s... m..~ o... ~v.. SJ 8.. 3.. 3.. 5.. 8.. No.~ o... 8.. 3.. a... 8.. 8.. 3.. F... 2... 3.. no.~ 3.. am . .5...” mm; min ww.~ 8; .m.~ n+4 .0; m...” 5.» 3..” 2..” m~.n .m.m an; 2...... 0..” co.» S.N. on... 8: $3.6m; o..v 3.» on.v no.0 ...v 1,. mm 2%. m..~ .o.~ a... oo.~ 3.. on.~.no.~ .n~.. .¢.. 8.. 8.. 2... on.“ 8.. o..~ 2... tn... 2... 3.. 2.. 8.. 8.. mo... mo...u&.~... cw.“ a... on 3.; 8...” us; m¢.n 3.. m..m av; .v.v $.n a...“ 8: 2.; No.” 3: ha; 36 Vt.” swim so...” 33 no.v ...mv o..v «no. as.” ...v ...t .m.n 8.. N... 2... .........~..8~ ...m 8.. .8. ...~ 8.. ...~ 8.. .... .... o... 8.. 8.. o... 8.. 8.. .8. 8.. .8. .8. 8.. 8.. 3... 8.. 3.. .. 8... 2.. .8. .8. ...~ 8.. 8.. 8... 8.. 8.. .8. ...w 8.. 3.. .... N... .... .8. .8. .... 8.. 3.. 2.. 8.. m... 8.. 2.. 8.. a. .. .2... 3. om.. 3.. 3.. m... N... 3.. 2.. 3.. 8.. m... 3.. 3.. on.. 3.. 3.. .... ..m.. .wm.. mo.~ v... 3.. .9. 0a.. 5.. 3.. 2... am 8.; mm.¢ N...“ a..." 3.... mo; 3.... 3.0 mp... «v.v mm.» on; we...” so; 3.? n..m 8..” .o.m 36 8: 2.; 8.6 8..» 3...” owom 2..." n..v ~N.v f on .8: 8.. .8. 8.. 8.. .8. o... .... 3. o... .8. 8.. 8.. 8.. o... 8.. 8.. 8.. .8. o... 2... 8n. 8.. 8.. 8.. .... 8.. 8w. .. 8... 8.. .8. .8. 8.. ...N 8.. .8. .8. 8.. 8.. 2.. .... 8.. m... 8.. o... o... 8.. 8.. 8.. .... o... a... 8.. 8.. .8. 8.. t .« too-z own: mN-z man: mwuz «an: NV... can: 3.: .muz 3:: tan: wwuz men: at... «no: man: mun: an... oNaz.S-z mm... as: mum-gown: can: ~34. own... .5 .oz. .8.o.8~¢58~¢ 82E 2.885 ”www.mnvwc $8.33 muww.~a..u. .. 5... an; .umsc.ucoov ..< 0.... 1'11?“ .EL’M I...“ 281 r I! .4114Lv. 11.....H..I.ICPJ .n’fiht .mcoHumw>ww numucmum 0cm mammfi nflmnu musmfioo ou mumfiumoummmcfi cmmn m>mn wasoz u. .mocmm .mamom msoncflucoo m :0 cmuoom ammn m>mn ou UmEdmmm uoc mnmz mam». mmmna .cmuuweo mum ma can .HH .0H .0 mama. H uumm 04m i 00... 0~.. .0.. .m.. .... .~.. 0v.. 00.. 00.. .m.. ¢~.. 0... -.. 00.. 00.. s... 00.. N0.. .0.. .0. 0... 00.. ouo. 00.. 00. .0.. 00.. ~0.. 90 000.v 00.v v0.¢ ...0 00.. m..0 00.0 .0.0 00.0 s... 00.0 .0.0 00.0 .v.¢ 0..v .0.v 0¢.¢ 00.? vv.0 00.v 00.0 0n.¢ 00.v 0v.v 0..¢ 00.0 ...v .N.0 3L 00 000.. .... 00. 00.. .0.. 0 .v.. .¢.. -.. 2.... .N.. 00. 00.. 3.. 00. N0.. 00.. 0... 0N: 0... 0.0.. .~.. 00.. 0... 0~.. 0... .~.. mm... .0 N0..N 0~.N ...N 00.N 00.0 0 . .0 00.~ 00.~ ...N 00.0 00.~ n..~ 0..~ 0..~ v..~ 00.0 ~0.~ +0.~ .0.~ ...N 00.0 ...N 00.N 00.~ m..~ 0..~ 0..~ 00.~ x: .0 0N0.. V0.. 00.~ 00.. v... N..N .0.. .0.N vm.. 0... .0.. m... .0.. ~0.. 00.. 00.. .... .0.. 00.. 07.. N0.. ~0.. .0.~ 00.. .0.. .0.~ 0... .0.. .0 ....v 00.0 0~.0 ¢~.0 00.0 00.. m¢.v .0.v 00.0 00.v 0... ~0.0 00.0 00.0 0N.» .0.¢ 00.0 v0.0 00.. 00.v .0.0 .0.v 00.0 00.. .0.0 N~.¢ 00.0 00. z: 00 .0... «0.. v0.. 0v.. .0.. .0.. .0.N 00...00.. 00.. 00.. 00.. 00.. .0.. .0.. 0... .... 00.. 0... 0n. . .~.. 00. 0.. .0.. 00.. «0.. 0... m... .0 p00.0 00.0 00.0 00.0 00.0 .m.v ...0 ...0 ~0.v 0~.0 v0.0 00.. .¢.v 00.0 00.0 00.0 0..¢ 00.v .0.0 «0.0 m~.0 00 . ~ .0 0..0 v0.v 00.v 0m.0 ...0 00.0 n: 00 0.2 0Nu2 mwuz 00.2 0m»: mmuz Nvuz 00.2 ¢.uz .muz 0vuz ¢mu2 Nun: nvuz 00:2 00.: 00.2 0~uz 00.2 0~uz .muz 00.2 our: 00.: mmuz 00.2 «v.2 00.2 .oz .00 .0. 00 N0 .0 00 N0 .0 n. N. .. 00 N0 .0 00 N0 .0 0* N¢ .v 00 «0 .0 mm NN .N n. N. .. 20.. 00:0: .vmdcflucoov H.4 magma 282 0.0.0 0v.. 00.. v0.. 00.. 00.. 0... .0. 00.. 0... 00.. 00.. 00.. 00.. 00. 0... 00.. 00. 00.. .0. 00.. 00.. 00.. 00.. 00. 00.. 0... 00.. 00 000.0 00.0 v0.0 00.0 v0.0 00.0 .0.0 00.0 .0.0 00.0 .0.0 00.0 .0.0 .0.0 0v.0 v0.0 00.0 .0.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 ...0 0000 00.0 1; .0 .0... .... 00.. 00.. 00.. 0... .0.. 0... mo. 00. .0.. 0... 00. 0... 00.. 00.. 00.. 0... 00.. 00.. 00.. 00.. .0.. 00.. 00.. 0... 00.. 0... .0 .00.0 00.0 00.0 0v.0 00.0 00.0 00.0 00.0 .0.0 00.0 00.0 .0.0 00.0 00.0 0080 .0.0 00.0 00.0 00.0 0..0 0..0 00.0 00.0 00.0 00.0 00.0 0v.0 ...0 x: 0 .0... 00. 00.. 00.. 00. 00.. 00. 00.. 0... 0... 00. 00. 00. 00.. 00.. 0... 00.. 00.. 0.. .0.. 0... 0... .... .0.. 00.. 00. 00.. 00. 90 .e..0 0..0 0..0 .0.0 00.0 00.0 0..0 00.0 00.0 .0.0 00.0 00.0 00.0 .0.0 00.0 0¢.0 00.0 00.0 00.0 00.0 00.0 0..0 00.0 00.0 .v.0 00.0 .0.0 00.0 I; 0 ..0.. v0.. 00.. .0. 00.. 0... 00.. 00.. 0... .0.. 0....0N. 00. 00. 00.. 00.. ¢... 00. 00.. 0... 00. 00.. 0... 0... 00.. .0.. 0... 00.. .0 .v..0 00.0 0..0 .0.0 00.0 0... 0..0 0..0 00.0 v0.0 00.0 00.v 00.0 00.¢ 0..0 00.0 00.0 0..0 0..v 00.0 .0.0 0v.0 0v.0 0..0 00.0 00.0 0.00 .0.0 10 0 0.0.. 00.. .0.. 00.. 00.. 00.. 00.. 0... 0... 00.. .0.. 00.. 0... 00.. 00.. 00.. 0... 00. 0... 0... 0... 00.. v0.. .0.. +0.. 00.. .0.. 00.. 90 00..0 00.0 00.0 .0.0 .0.0 .0.0 00.0 00.0 00.0 00.0 00.0 ...0 00.0 00.0 00.0 0v.0 00.0 .0.0 00.0 00.0 00.0 0..0 00.0 00.0 00.0 00.0 00.0 00. 0 30 v 000.. .0.. 00.. 00.. 00.. .0.. 0... 00.. 00.. 00.. 00.. 0... .0.. 00.. 0. . 00.. v0.. 00. . 00. . m... 00.. 0... 00.. #0.. 00. v... 00.. 0... .0 .00.0 00.0 00.0 00.0 0v.0 0v.0 0..0 .0.0 00.0 .0.0 00.0 ...0 00.0 00.0 00.0 00.0 00.0 00. 0 0. .0 00.0 00.0 00.0 0v.0 0..0 0..0 00.0 .v.0 00.0 30 0 0.... 00. 00. 00. 00.. v... 00.. ~... 0... 0... 00.. .0. 00. 0... 00. 0... 00. . 0... 00 0... 00.. v0.. 00.. .0.. ¢0.. 00.. .0.. .0. .0 000.0 00.0 v... .0.0 0..0 00.0 00.0 00.0 0... 0..0 00.0 00.0 004v .0.0 v0.0 .0.0 .0. 0 00. 0 00.0 00.0 00.0 V0.0 v0.0 00.0 00.0 t0.0 v0.0 00.. z. 0 000.. 00.. 0... v0.. 0... 00. 00.. 0... 00. 0... 00.. .... 00. v09. 0... 00. .0.. 00. . v.. . 00.. 00.. 00.. .0.. .0.. 00.. v0. 00. 00.. .0 .00.0 00.0 00.0 v0.0 00.0 .0.. .0.0 00.0 00.0 .0.. 00.0 00.0 00.0 00.0 00.0 0v.0 0... 00. . 00. . ...0 00.0 ...0 .0.0 0..0 .0.. ...0 00.0 .0.0 a: . 000a: 00uz 00uz 00uz 00uz 00uz 00.2 00.2 v.uz .0uz 00.2 00.2 00.2 0v-z 00.2 00.2 00.2 00.0 00.3 00.2 .0.2 00.2 00.2 00.2 00.2 00.2 00.2 00.2 .02 .00 .0. 00 00 .0 00 00 .0 0. 00 .0 00 00 .0 00 00 .0 00 00 .v 00 00 .0 00 00 .0 0. 0. .. ru.+ 0000: mmmson 00 may 00 nomm How >0>Hsm mammamcd mmsom map 00 HH puma Eouw mEm.H on mmmcommmu mo mcofluMH>m© cumocmum 0cm mammz 0.0 magma 283 000. 0.. .0.. .0. 0... 78. 0.. 00. 8. .0. .0. 00. Q. 00. ... 00. 0,0. 00. .0. 0.. 00. 0.. 00. 00. 0.. 00. 0.. 00. .0 00..0 0.0 ...0 00.0 3.0 .0.. 0..0 0040 .0.0 .0.0 0..0 00.0 .0.0 .0.0 00.0 0000 .0.. 0..0 8.. 00.0 0..0 00.0 .0.0 .0.0 00.0 00.0 0..0 ....0 8. 0. 000.. ._.... 00.. 00.. 00.. .0. .0.. .0.. 00.. 0.. 00... 0... 00.. 00.. .0.. 00.. $0.. 00.. .0. 00.. 00.. 00.. #0.. 00.. 00.. 0... 0... 0... .0 08.0 0..0 .0.0 00.0 00.0 0... 0.0 00.0 .0.0 00.0 .0..0 00.0 3.0 00.0 .0.0 0.00 00.0 0.0 .0.. .0.0 00.0 0.10 00.0 0..0 00.0 0..0 .0.0 ...0 I. 0. .00.. 00.. 0... 0.... 00.. 00.. 0... 00. 00.. 00.. 00. .0.. 8. 0... 00.. 0... 0.. 00.. 00.. .0. 00.. .0. 811100. 00.. .0.. 0... 00. 0 .0 2 80.0 8.. .0.0 00.0 3.0 .0.. 00.0 0..0 .0.0 00.. 00.0 .0.0 00.0 00.0 00.0 0..0 00.. 00.. 00.0 .... .0.. 3... .0.0 3.. .0.0 00.0 0.0 00.. v. 000.. $.. 00.. vmm. 00.. .0.. 00.. 00.. 00.. 00.. 0.... 00.. 00.. 00.. 00.. 3... 00.. 00.. 0... 1.... 0... .0.. 00.. .v.. .0.. .0.. 00.. .v.. .0 .00.0 00.0 00.0 ...0 0..0 .0.0 00.0 .0.0 3.0 .0.0 0..0 00.0 00.0 .0.0 ...0 00.0 0..0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 .0.0 00.0 ...0 00.0 8.. 0. 000.. .0. .v.. 0v.. 00.. 0.. 0.... 00.. 00.. 3.. .0.. 00.. 00.. 00.. 00.. 00.. 00.. 00. 0.... 00.. 00.. 00.. 00.. 3... 0... 00.. 00.. 00.. .0 00..0 8.. .0.0 ...0 00.0 00.0 ...0 0..0 00:0 .0.0 00.0 00.0.3.0 0..0 .v.0 00.0 00.0 00.0 00.0 0..0 00.0 00.0 00.0 0..0 00.0 00.0 0‘0 .0.0 .2 0. 0.0.. 0... 00.. .0.. 00.. 3... .0.. 0v.. 3.. 00.. .v.. 00.. 00.. 00.. .0.. +0.. 0‘. 00.. 00.. 00.. .0.. .0.. 00.. .0.. 00.. .0.. 00.. 00.. .0 000.0 00.0 00.0 00.0 00.0 00.0 0..0 ...0 00.0 ...0 00.0 0..0 00.0 0..0 00.0 00.0 3.0 00.0 v..0 00.0 0..0 .0.0 00.0 00.0 00.0 00.0 .0.0 00.0 f .. .00.. 3.. 0... .... 0... 00.. v0. 00.. 0... 00.. 0... 0... .0.. 00.. 00.. 0... .0.. 00.. 00.. 00.. 0.. 00.. 0.. 00.. 8. 00.. .0.. 00.. .0 000.0 0..0 00.0 00.0 00.0 .¢.0 00.. 0040 00.0 0..v v0.0 0000 00.v 00.0 0..0 0... 00.0 ...0 0..0 00.0 0... ...0 0'; 00.0 00.0 00.0 00.0 .0.0 .t 0. 00... +0.. 0... 00.. .0.. .0.. 00.. .0.. 00.. 00.. 00.. 00.. .0.. 00.. 00.. .0.. 00.. 00.. 00.. .0.. .0.. .0.. 0v.. 00.. 00.. v... 0‘. .m... .0 000.0 00.0 00.0 V0.0 0..0 0..0 00.0 .0.0 .0.0 00.0 ...0 00.0 00.. .0.0 0..0 0..0 v..0 ...0 0..0 00.0 00.0 00.0 0v.0 00.0 0..v 00.0 .0.0 .0.0. 8.. 0 .00.. 00a. 00.0 00... 00a. 00.. 07. 00.. v... .0.. 0:0 v0..- 00} 0..0 00... 00.. 00.. 00.. 00.3 00.. .0.. 00.. 00.. 00.. 00.. 00.. 0:. 00..- .2. .53. 00 00 .0 00 00 .0 0. 0. .. 00. 00 .0 00 00 .0 0v 0. .v 00 00 .0 00 00 .0 0. 0. .. 5: um 00: Aomscfiucoov N.« magma 2E34 .., rrdud. ...... ..v ..s .E ammo, S.» a; o~.— coop mNo— moo— -op 3.. mm; oNo— soon 3; 2; mp; N; 3; co; no; mm. Sop coop or; 8.— me. 3.. «No— 3; a» mom.» ov.» mm.~ 3..» 3: 3cm 8» ON.» an.» :.~ on» Sou an.» on» Na.» 2..» CHN 5.» Cow 5...» cm.» 8.» on; m»: 2..» 3; go» now? £ vw mo»; on; t»; :3 :o— N»; F»: $3 a»; mu; Nu; nu; ma: owo— ow.» «Now o»; :3 a—g— mpo— cw; nu; 33 N»; 38’ vuop 5: 5N; .» 583.. mmow to.» Bow ow.» Eow Bow 8.~ omwm m».» »»o» co.» omww oo.» 8.~ 5N.» on.» 3.» tom 9.» mwou 2.» Row $.w 8o~ mmou 2..~ voou f »N 2;.— mmo crow Np; mo; 8. 2..— ¢~op cw." 2..— 3. up; 8. 3.— 8; 2; mp." 8. oh.— »so ovop 8. ago— 3. w»; 8. 5., 3.— »o¢.~ mNoN Bow Cow mmow v».~ low o».~ 8; otoN m»o~ «wow 3» SJ n-ow a...» $.N top Bow oo.~ noou avow »¢.~ 9o" mv.~ 5.“ mm.» $.N 1!:5 «N :«J NNop on; NN... em." mm; op." 33 mm; :o— o»; oNop 8. mm." mm; o»: »No— to; on.— n—.— no; no; on; .5.— w—.— as.— »N; NV; on camaw as.» Sow »~.~ Eow o~.~ »¢.~ »».~ :.~ mp.~ »~.~ {ow ~N.~ Nv.~ msom $.N c».~ SJ no.» Nmow «wow SJ 82o. co.» mmom 3.» qu m»- 8.. E or»; am; a»; hm; am. «Now m»... C; 3; v»; hm; mm; mm. owo— wwop SJ 2..— mNop 3.— mwop a»; F»; m»; m»; an; N».— anop 9.. on So.» ow.» to» low 8; mo.~ no.» »».» 3..» :.~ sw.» 8.» mm.~ mmoN mm.» 2..» via cm.» :.~ 3.» 2.» 3.» 2.» So» mmow moo» 34w cc.» 5 cu 2v; NV; m~.— mm.” on; N»; o»; mfip a»; »¢.— 2..— mN.’ a; pmop $.— mmop :3 $3 av." «v.» 3,... coop mu; coop N»... »~o~ $.- 3: on m3.» 2..» an.» 9.» mmow on.» 2..» mm.» m».» on.» 3.» mm.» 9..» on.» ooom Np.» pv.» 3..» 5.» 8.» cc.» 5.» on.» 3...» 2.» 3.» on.» «N.» f 2 gm; 2.... mm; a». no; 3.» up; 5.— mmJ Nu." mm; :3 2.— F»; co; no; g.— c»; 3. S.— »—.w a; gap on; 2.» mm; mm.» :0» a» mac.» 8cm »m.~ 8o~ v0.» ..moN 3.» »».» tom low 8.» Sam »~.o~ 5.~ «co» 3.» we.» mm.~ meow on.» 2.» «no» I...» an.» 3.» on.» as.» a».~ I. or 0mm.» N»; ac... 3.— mm.« $3 or; cw; o»; o»; »—.r Bop ~No— N»; :3 ~».— «v.9 Sow o~.— on; X».— o~.— o~.— 2.— ~»o— mo; 0»; sNop a» mac» mt.» two» »N.» m».» mm.» :o» 2.» 3o» 3.» mm.» Sow sow »w.~ vm.» m».» Sow cm.» Sou 3.» mm.» or.» 5.» 2.» mm.» m».» maow »»o» f : $9.2 ow-z mun: mmuz mN-z ~»uz 232 o»... In: 3.2 ov-z 3oz -uz 932 332 »»uz m».z mwuz mmuz mN-z p»... m»... mun: »»nz mun: o»... ~72 m»... .02. .55.. »m 8 5 no 8 5 2. Q. C 8 3 5 n» Nm 3. »v wt 3 »» N» 9» »~ an an »— Np z. 5:. uaaoz. Aomscflucoov ~.¢ manna 285 FN».~ m¢.— ~¢.— »N.— mN.P m».p tN.F o».~ Np.— »N.— pt.» P».F KN." o».— mN.p 0N.F m».p m».— No." mp.— mm.» N».w N».p at.» m—.p mp.~ mN.— Nv.p a» 9...» ov.» Nm.N m¢.» mm.» 3.» mm.» o—.» E.» 3.» o».» NN.» mm.» 0N.» or.» 3.» Nh.» 3.» 3.» on.» mt.» 8.» z.» 3.» cm.» NN.» «N.» z.» .2 N» Fmp.p mF.F o».P .N.p NN.~ mp.p »P.p »m. nN.P ms. v».— mp.p m». mo.— cp.p NN.— NN.F N¢. um. mp.p mp.P mm. cN.p »F.— m«.— N».« »».p mN.— a» opr.N om.N »O.N oo.N —~.N mo.N NP.N oo.N oo.N mm.p mN.N mp.N Na.» mm.— om.N c».N »N.N »N.p »~.p »N.N »o.N NN.— vm.N m¢.N om.p Nv.N No.N mo.N n: w» o~».~ o».« N».— mN.p ~N.~ m—.p ~».w o¢.— m».p p¢.p —».~ 0N.» cm.” mN.p FN.— mo. No.p m».F tN.— s».p Na.— n».p ~».— v».p mp.» ov.— oo.p en.» am —wp.» co.» wN.» »~.» ow.N «a.» mN.» »o.» c—.» av.» mo.N »o.» ~N.N o».N om.N mo.N as.» ea.» Na.» mo.N n¢.» »N.» pN.» pN.N am.» m~.N msw» Nm.N 1: o» —»p.w om. N».— »N. oo.F mm. N».F oo.— N». Po.p mm. N». Na. om. F». o—.— mm. N». oo.— mp.p m». w». mo.— p». »p.w »—.— vo.» vo. — a» non.» o~.N.p».» ~v.¢ mo.» »m.v mm.N No.» on.» am.» "a.» m».¢ N».¢ mm.» «m.» vN.» ow.v m—.v ~m.~ on.» wp.v wm.» tn.» m».¢ Nm.» pm.N m». N Na. » 2: mN N»N.— oo.p mm. mo.— cc." NN.F ¢».F ON.P NF." mN.p m». F NP. F 00 P vo.— ~—.p or. — Np. p mp.r oo.F ON.F p».F »N.p vp.— »w.p pp.~ N».p ——.— m — a» m3.» oo.» 2.» Q...» N~.N mm.» 5N mm.N mm.» mo.» 3. » a». N »N.» mm.» $.N mo. » m». » to.» no; mw.N mp.» 8.N mm.N mm.» 8.» oo.» om.N 3.N f »N F—N.— FN.P FN.w o~.— 0N.— N».F NN.F Nm. ro.— No.9 mN. mp. — FF. P op.p vo.— m».— N». p at .p o». oo.« m».— »p.~ »N.— m».» »N.F hp.— ~».— ¢N.w a» m~m.N o».N m~.N Nm.N ov.N mm.N F5.N Nm.N FN.N »N .N NN. N rs. N vw. N »m.N .m.N NN.N mm. N » .N m».N mm. N mm. N »v.N »~.N Nw.N mm.N mm.N mN.N om.N 2: KN NPF.— mm. NN. »m. wo.p vp.p mm. F».— mo.p mm .9 wo.F NN. a». mo.— Np.p mm. No.p tp.— m—.— —N.— so.p —».p F».— »N.p mm. oN.p p».p o—.» a» Nmo.¢ om.¢ o¢1v t».v No.» wo.¢ —N.¢ Km.» mN.v :w » nw.v p¢.¢ o~.v mo.v oo.¢ mv.v mo.¢ pm.» Nm.» we. v »—.v »m.» vm.» m~w» o—.¢ v0.» om.» ,—.¢ 1; mN N»p.w 0N.p Nu. »—.p »N.F op.p m..— mF.p .0., mm. mN.P ~N.w mm. mm. NN.- pp.p »F.— mN.p mN. m«.— no." »».p Nv.— m».p pN.— wN.p pN.p ON.» a» m——.N om.N pt.» pm.N oo.N Nm.« vN.N NN.N m».N oo.N oN.N pN.N Na.» N5.» N—.N mo.N oo.N wv.N N~.— NN.N vm.— om.N pN.N m—.N p».N Na.» mo.N m—.N 3: mN vomoz QNuz mNuz m»uz nNuz Nmuz quz omuz t—uz pnoz 0..: vmuz NNuz nruz ecu: n».z mmnz mNuz mmuz mN.z p».z mnuz mmuz »».z mNuz o».z Nvoz mmuz .02 .co hop »m N» pm »» N» —m »N N» «s »m N» F» »m No «a »v N¢ Fv »» N» p» »N NN wN »p N» pr rm». manor, Acmscflucoov N.< magma 286 ooN.N mm. mm. NN. a». NN.N m». NN.N a». N». No.N Na. »o.N »N.N v». an. oN.N No.N ma. NN.N vo.N NN.N N». m». No.N mm. N».N vN.N mNo.» cN.¢ m».v N... »N.. mm.N »m.» cm.» »N.» oo.v No.¢ oc.v mm.» Nv.» »o.v mv.v »N.v N».» NN.» NN.« NN.» »o.v .c.¢ vN.¢ an.» we.» »¢.» NV.» £3 at no».N NN.N mv.N NN.N »N.N »N.N »N.N v».N mm.N NN.N »N.N »N.N vo.N NN.N »N.N »N.N o».N o».N »N.N No.N »¢.N v».N m».N N¢.N t».N ¢N.N o».N 0N.N -» NNc.N NN.N Nm.N »m.N w».N Nm.N Nm.N »N.N Nn.N »N.» »N.N »N.N mo.» No.» N».N Nv.N co.» NM.N NN.N »v.N »va mo.N cm.N »N.N mo. N Nv.N om.N »N. N x: m» moN.N mm. mm. om. m». »N.N »N.N Na. »N.N »N.N »N.N N». mm. to.N No.N »o.N NN.N N».N NN.N v». »N.N mo.N om. ¢N.N NN.N mo.N mm. N». a» oNo.N »N.» »N.» »v.» o¢.N »o.» »v.N o¢.N »».N .m.N om.N vN.» »N.» »N.» »».N oN.N NN.N mo.N oo.» vo.N »o.» o¢.N vw.N NN.N mm.N Nm.N »v.N on.N n: o» me.N No.N »N.N No.N mo.N .N.N No.N mm. N». N». »N.N No.N »N. N »N. N »N. N »N.N »N.N v».N NN.N »N.N vN.N NN.N NN.N o». N .o. N NN.N mN .N »N.N a» »NN.N mm.N »N.» ».N m».N om.N »m.N N¢.N Nm.N »N.N Nm.N N».N mo. » Nm m.N »N. N Nv.N NN.N NN.N Nm.» »N.N vN.N om.N on.N mN N.N mN .N »N.N mN .N »N.N I; N» Nm».N »v.N ¢m.N NN.N »N.N m¢.N NN.N »».N m».N NN.N m».N »N.N ov.N N».N N».N »N.N »N.N nv.N »N.N »N. N o».NN N». N mv. N »N.N »N.N o».N N¢.N Nv.N a» mac.» no.» NN.» »o.N 0..» mo.» NN.» ow.N »m.N mo.» oN.» NN.N mm.N vN.» om.N Nv.» »N.» »N.N Ne.» m» ».N om.N v». N v». » co.» Nt.» NN. » No. » NN. N I; m» ovN.N »o.N »N.N mo.N mN.N »N.N mo.N no.N 0N. mo.N »N.N mm. NN.N NN.N Nm. m7 N NN .N NN.N »o.N »N.N »N.N mo.N mm. oN.N No.N m N m7 N «N. N a» va.» 0N.» »t.» NV.» NN.N vm.N »N.» »o.» »N.» »N.» o».» N».» vm.» mm.» No.N »N. N NN. » on.» N».» m».» Nv.» NN.» »N.» on.» NN.» em.N N». » NN. N 1L m» NNN.N NN.N mm. »N.N oN.N NN.N vN.N m». .o.N No.N »o.N N». mm. vo.N mo.N »N. »N.N NN.N »N.N Na. oN.N NN.N .N.N NN.N »N.N NN.N »N.N co. N a» wm¢.» or.» Nm.» NN.» v..» »N.» m¢.» »N.» vw.» No.» »N.» NN.» on.» Nt.» m».» 00.. »».» mw.» vm.» »N.» »N.» at.» .w.» »n.» NN.» mm.» »N.N 7 » . x: v» Nmo.N rm. »N.N o».N vm. »N. mm. mm. ON. wo.N »N.N oo.N »N.N ON.N 0N.N »N. vo.N mm. No.N vN.N »m. mo.N N». No.N vN. m». No.N vo.N n» mmo.N 0N.N ¢».N Nm.N oo.N »N.N oo.N 0N.N No.N NN.N om.N oo.N n¢.N m».N ova NN.N oo.N mo.N N».N »N.N om.N mo.N »N.N »N.N No.N mw.N mm.N »N.N -: »» comm: ONIz »Nuz m»uz »N.: N».2 Nvuz o».2 anz N».z m¢az v».z NN.: »vuz mtuz »»uz m».z mNuz mmuz »N.: N»uz n».z »N.: »».z »N.: m».z Nvuz w».z .oz wa NON »m N» N» »m N» N» »N NN NN »m N» N» »m Nm N» »v N. No . n» N» N» »N NN NN »N NN NN In». ”use: Awmscflucouv ~.« magma 287 In... mVN.N 8. 8. NN. on. »N.N NN.N »N.N NN.N NN.N NN.N »N. V». »N.N no: »».N N».N »N. NN.N VN.N Va. NN.N N°.N »V.N ON.N »».N »N.N »».N o3.» OV.V »».V a»; Q.» »o.V »N.» »N.» NN.» o7V 3.» 97V »N.V mqtgoV Va.» NN.V an; »o.V 07V »N.V 3.» 21V 8.» NN.» 90.» N7» »N.» nu »NN.N »N.N 8. N». »N.N N..N. »».N »».N »N.N A...N.N VN.N NN.N on. Vo.N No.N N».N no; VN.N »N.N »N.N »o.N NN.N N».N VN.N 3. N».N NN.N NN.N »N.N.» 8.» VN.V ao.V bu.» »m..V .8.» Nu.» oo.V »o.V No.» »N.V mn.V »N.V NN.» Na.» »N.N. m7..V »».V on.» »o.V NN.» on.» NN.» NV.V mu.» No.V we.» 2... 8.. 8.. 8. 8.. x... 8....3: 8. .~.. 2.. 8. 3.. m... u... 3.. 2.. 8.. 8.. 8. x... 8.. 8.. 8.. 8. w... a... 3.. «ad 2.... «..u SJ a... 3.3.59?» 3... 3.... .3 .3 :4. ...~ ...~ .3 3.... 8.... ...N .3. a... a...» .3. 2.»... ...n ...n 8.~ 8.... NN».N N».N »».N »N.N oV.N »».N 0V.N »N.N NN.N NN.N »».N VN.N »».N NV.N »N.N 8.N »».N »V.N »».N NN.N NN.N o».N »N.N NN.N »».N NN.N »N.N »N.N $7N 8.N No.» V».N N».N 3..N..»».N »V.N »N.N N».N om.N NN.N »».N »».N »».N VN.N »N.» »».N NN.» mV.N NN.N VN.N »».N »N.N N».N 3.» N7» VV.N »»7N »N.N 3. »N.N »N.N NN.N NN.N »».N Vo.N NN.N VN.N NN.N »N.N .Vo.N »».N »N.N VN.N »».N »».N o7N »N.N »N.N NN.N no; VN.N Va; 07.. NN.N. m3.» 8.» 3.» 0V.» VV.N VN.N NN.N NV.N V».N N».N NN.» »N.» 3.» 5.03.... AN».N No.» 8.» V».» »».N 0V.» 8.N N».N »N.N V».» NN.N 8.N VN.N NNN.N NN.N »».N o». N».N »N..» N».N oo.NVNN.N up...» NN.N VN.N »N.N NN.N »N.N N».N »N.N »N.N »».N 3.N Vo.N No.— N».N »».N N».N N».N »N.N »N.N Nom.N 8.N »».N »».N NN.N NV.N »».N »N.N »N.N »».N »N.N o».N NV.N »».N »».N »».N »N.N »N.N 8.... »».N »».N N».N Q.N Vn.N »N.N »N.N N».N NN.N »mo.N »N. a». a». NN.N. NN.N NN.N »o.N »N.N NN.N no. N». No.N m». N». »N. mo.N NN.N »».N a». No.N N».N o». NN. »N.N »N. VN.N »».N Na.» 8.» »N.V no; »VU» co.» Na.» NV.» Nu.» 0V.» 3.» 3.» co.» »V.» »N.» VN..V »N.» 3.» VV.» 3.» No.» 8.» 8.» o».V V7» oo.V »N.» NV.» VNo.N VN.N No.N N». »N. 3. co... »N.N NN.N No.N Na. 8. NN. 8. Vo.N NN.N 8. »».N 3. 8. »».N 8. 8. V7 no; a». »m. 8. 08.N »N.» »o.» »V.N N».» NV.N VN.N »N.» »N.» NN.N N».N »N.N »».N 8.N NN.» we.» VV.N »».N no.» NV.» N».N 8.N 3.N »N.N »»wN »».N VN.N or..N NV 0V '£8 9 £»V NV 3».- ON.3 »N.. a?» »N.. N»... NV..- o».u 3.: N».- 0V.» V».- NNon »V.- »V.- »».n »».n »N.. a»... »N.. N».I m».- »N.. »».a »N.: »».u NVnn o»..- .GSNQNNNNQN»N»»NNNNN»mNoN»»mNon»VNVNV»»N»N»»NNNNN»NNNNN unacx on: tuN. Nvmncwucoov N.¢ manna 288 v»... m». .0.. .».. »... »N.. .0.. .».. .t.. m... 3.. 3.. 2... »... .».. 8. n... .... I... 3. 3.. 3.. .».. .».. a». 8. 8.. m».. .»..» o..» 0..» 3.» »».» .».N »..» »».» 3.» I.» 3.» »».» on.» 2.» m...» »».N »».» »».» .».N »e.» 3.» mm.» to.» no.» »».» on.» .0.» v0.» 9.»; .... 8.. »v.. 3.. »».. .».. »».. m... m... »».. v».. »N.. N... m... 3.. .».. 2... 8. »N.. .».. .... »N.. mm.. »o.. »»....o.. »N.. »5.» o...» m...» »».» om.» »».» Nm.» 3.» .».» o..v I...» I.» »».» E.» .».» a...» »N.» t...» »..» 2..» 0'.» 2..» m..» »o.» .».» ¢¢.» at.» m..» 8.... a». 3.. o».. »... 8. m». .».. .... 5. 3.. .o. .... .».. N... 8. o... .... 2. 8.. »b.. .o. »o.. o... 9. o».. 3.. m... 08.» m...» »s.» we.» »».» 3.» 3.. »s.» »».» »».v 8.v 8... »..v .m.» 8.» mo... 3.» a.» o... 3: I.» »».v a...» »».» om.» »».» »v.» Vt.» »»».. m... m».. »N.. »¢.. .... »N.. »».. »N.. 3... 3.. o... »o.. a... m».. »».. o... .»... o».. »v.. o»... a... .».. 3.. »N.. .».. mo.. 8.. .m..» on.» um.» I.» »».» on.» an.» .m.» 3.» ...N an.» ac.» »».» co.» um.» »..» »s.» on.» em.» 2..» 9..» 2..» »».» »o.» »».» .v.» ow.» «m.» »N». N». ».. 2. o». 2.. m». m». cm. a». 5. v». o». 2. 2. N». r... 8. 2. n... 2. .o. a». a». Z. 2.. v». 5. .»v.» »...» »v.» vm.» 3.» mm.» 0..» I...» »».» 3.» »».» »».» .».» vt.» »».» no.» .v.» »N.» mm.» on.» »».» »».» »».» »».» »».» ...» 8.. 3.. .m... c... 8.. o».. m». n»... »N.. »».. .».. »». .».. mm. »... »o.. 8.. N... »N.. 8.. o».. »m. .... 3.. v».. »N.. 5. 8. m... m». 8..» co.» .N.» mo.» 0...» om.» .».» .».» co.» 8.. »».N m... we.» no.» .».» .t.» »».» »».» »..» »».N »m.. »v.» 3.» o».» .o.» »».. 5.. 3.. av... .».. »».. 3.. .¢.. n... .».. .... o... .o. »N.. v». .0.. »N.. »o.. .o. o... 3.. .».. 2. C. o». m... as. .».. .o. n... 8. »»..v on.» em.» 8.» »».» 2..» mm.» o..v 8... mm... »..¢ »».v I... 8.» to; 38v »».v n».¢ m... »v.v »».v o».v mm.» at; »».» ...v ...v .¢.v 93.. .».. $.. 8.. 3.. m».. o... m».. .».. 00.. 3.. »m.. »o.. »v.. »c.. m... or... on.. m».. 3.. m... 3.. .».. 2.. a... 8.. »N.. mv.. own.» 2..» »».» I.» 3.» co.» »¢.» »».» .o.» .o.» »».» 8.» »».» mm.» »».» »o.» .m.» 3.» 8.» me.» C.» .m.» »v.» o».» 9..» »».N »».» »..» £6 en a» £8 v» 3.»... £3 1...» for. Emu: o»ua »».n mm... »N.. »»u: »v.3 o».2 3.: .».. at: t»; »».u »vi 0... »».a »».a m»... a»; m».3 .».. »».x »».c »».x »».a m».- »¢... m».- .5 .0.. »m N» .m »» »m .m m. N. .. »m »m .u »m »m .n »v »v .v »» »» .» »» N» .» ». ». .. uaao: .0: tub— ...sc.ucoo. ~.4 ..nma 289 awe." ~9— 8; mm. 8. no.— ppJ no; 8. No; no.9 1..— mo.— 8.9 8.— to; vo.— SJ oo.— 3." 3. no: no. 8. no; 8. 8. 8... ‘vmuom mo.~ 5; mo.~ ma; mp.~ m~.~ 8:. mm; n¢.~ o~.~ mN.N 8... mm.~ 3.N on.~ oqw mn.~ 8.~ m~.~ mp.“ mN.~ on.“ um.~ on.~ and m¢.N m‘w 30. on. we. 8. 5. 2.. no. I. 3. no; 5. 2. .2. 8. 3.. no. as. we. 2.. 8. no. 8. cm. 2.. so. no. 2.. 8. cm¢.~ mo.~ mm.m $.N $3.. m—.~ pv.~ 2..~ on.~ noqm mm.~ $.N ~N.~ sqw 33.. 2..~ »N.N »N.N MP.~ mo.~ m~.~ $.N gum vm.~ $.N ma.~ on.~ pn.~ :m. mm. 8. 2.. mo.— 2.. 8. ma. 2. 3. co; 2.. g. 5. 8. c». 3. oo.— cm. 3. 3. 8. 3.9 no... 2. 3; co; 8. ~$.N wm.~ moaw o~.~ 8.~ Fn.~ E..." 8.~ 2..— ~—.~ m~.~ op.N SJ cm.~ pm.“ mWN $.~ cm; $.N -.~ 3.... :.N $.N ov.~ :J on.“ on.~ an.“ mum. an. cm. 5. No; «h. 3. 5. mm. mm. we; 2. mo. 3. 3. S. om. mm. 8. 8. mo. 3. 8. mm. 2. 8. no; 2.. mon.~ o~.~ vm.~ to; oo.n 5.9 to..." 3.~ mm... mm... 36 mo.~ 8; $.N mm.~ S.N.. mm.~ mm; —N.~ mm." om.~ prim INN ~v.~ 8.9 5...“. mn.~ mn.~ 0N0; 8. wm.‘ E. 8. 2.. mo; mo; no.9 no; 3. 5. pm. mm. 2.. op; 3. No. g. 2. co.» 8. 8. mm. mm. 3.9 mm. mm. m3; cm.» at: 5.. Nm.~ 8.9 p~.~ :.~ awow 3: an; mo.— mo.— an; 3.9 to; to; mm; ”$.— 33 m—.~ cm.— Nm.~ ~v.~ —~.~ 3.” oo.n flaw one; ~o.— mo.— 3. No; mm. mm. No.9 om. Np; 3.? 3. S. on. 3. NF; 3.. wm. No. or; 1.. mo. 2: 5. 5. oo.— so; 3. www.m‘. mm.~ «Wm mm; o~.m mm; oo.~ mm.~ mm; o—.~ ow..." 3: -.~ an; 2.“... -.m No.9 mm; on; $.w m—.~ v~.~ -.m —~.N mm; mam mm." $.w mN... vm.— SJ 2; -.p mm.» mm; mm; mm.— NNJ co... op." mo; 3: on.— Np... 5; mm. 8.. mo." mm. on; 3.. mm; mm. mm; m?— on; amQN BUN pm.~ wqw mm.~ mm.~ tm.~ mo.~ m~.n mm.~ oo.~ afiw mN.~ wo.~ :.N 3.~ o—.~ 2.; mo.~ 3.~ m~.~ I.N mm.~ vm.~ mo.~ m~.~ oo.~ om.~ mmm; 3.9 3: 5.. 9.: 2..." mm; tn; mm; v~.— 3.. NF; m~.— on; ma." ~—.— v~.— mNJ 5.9 «N; am. mm... 3.9 st; 3; 3.. pm; an... ov~.~ 9.6m mm.~ mv.~ Ns.~ mm.~ mm.~ 8.~ m~.m 3.~ om.~ om.~ vo.~ can 3.». ~—.m :.~ “Wm mn.~ om.~ w‘w can 8.~ ~m.~ Pm...” -.~ mm.~ 8.0 £8 £3 23 £3 5 {mm {mm 5m... own: mwun mm... mm»... mm... wt... on: 1... 2w.- mw... van. NN.. mt.- ovns mmuu an... 3... man: mm... «man on; can: m6.- m~.- can. Ntua an.- .5 5h mm mm 8 B No 8 2 3. Z. 8 No 5 mm mm. 5 av wv 3 mm Nm 3 mm «w 8 2. N— : unao: .02 in: Acmscfiucoov ~.« manna 290 .JJ.. 2.. 2.. 8.. SW. S: 1.. SJ... 8.. NN.. S... 8.. 3.. 3.. SN. S: v... S... 8.. 3.. :1. 8.. 2.. SJ... 8.. SJ: 8.. 8.. 8 SJ SJ .SJ 8.. SJ :J SJ SJ SJJ EJ SSJ ..J S... :J J... aJ SJ SSJ SJ SSJ SSJ S... ..J SSJ ...S ..J SJ .SJ .2 S NN.. 2. 3. SS. S. 2. SS. 3. .S. S.. S. S. J.. S.. S... ... z. 2. S... 8. SS. S. SS. 2. S. SS. S. 2. ..S SSJ SJJ SSJ ..J SJ .SJ SSJ SJ ..SJ SJ .SJ SSJ SJ SJ SJ mJ SJ SJ S..J SJ SJ SJ SJ .SJ SJ SSJ .SJ ..J 2 SS S.N.. S... NN.. .J.. SW. S... S.N.. 8.. 2.... NN.. S... S... S... 3... SN. .~.. 2... S... S... S; S... SN. 8.. SN. 3... 3.. NN.. 8.. 8 SJ SJ ...S ...S ~.J S..S SJ SJ SJ .SJ SSJ SJ 2.... SJ 8.... .JJ SJ SSJ S..S SJ ...S 3.... SJ SJ SJ SSJ -J ...J .2 S .SS.. S~.. SJ.. S... SJ.. ~S.. J... SS.. .... .S.. S... SS.. NN.. S... S... SS.. SS.. S~.. SS.. SJ.. .S.. SS.. S~.. SS.. S~.. SS.. N... SS.. .02 ...S s. S S .S S S .S S. J. .. S S .S S S .S S J. .v S S .S S ..J .J S. J. .. 5: ~26: .vmsaflucoo. ~.< manna 291 SSSSSSSS.SS.J.SSS..SS.S.SSS.SSJSSS.S.==S S..J SJJ SJ SJJ .SJ SJ .SJ SSJ .JJ .JJ SJ ..J SSJ ...J SSJ SSJ ..J SJ SSJ ..J SSJ SJ SSJ SJ SJ .S.J SJ SJ .S 2...... ...S .S... ..J S.J SJ SSJ ..J S... SJ .SJ SJJ ..J SSJ SJ SJJ J..S SSJ SJ SSJ SJJ SS... 2... SJ ..J SJ JSJ SJ .2 .. SS..S.S,S.SSS.SSSSSSJJSSSJSSSSSSSS.SSS..:S S..J SJ SJ 5.. SJ SJ SJ SJ SSJ SJ .SJ .JJ ..J SJJ SJ SJ SJ 8.. SJ S.J SJ SJ SJ SJ 8.. SJ ..J ...J 2.. 33. .JJ SSJ SJ SJ SSJ SSJ ..J .3 SJ SSJ .S.S ..J SJ SJ JSJ J.J SSJ S... S... S... SJ SJ SJ SJ SJ ...S ..J .2 S. ..SSS.SS...SSS..SS..SJS.SSJS.SSSS:.S SSJ SJ SJ ..J S.J SJ 3.. SJ .3 SJ SJ SJ 8.... SJ SJ SJ ..J SJ ...J SJ 8.. SJ .SJ S... JSJ S.J ..J SJ 8. SJ SJ .SJ 2... SJ .J...SSJ SSJ SJ S... ..J SJ SJ SJ 8.... 8.. SJ .SJ 8... 3.... 8... SJ SJ SJ SJ SJ 8.. ..J .2 S SS.SSS.SS.S.S.SS.JS.S.SSJ..S...S.S:J SJ .3 SJ ...J SJ SJ .S.J SSJ SJJ J.J SSJ SJ S.J SSJ 8J SSJ SJ SJ 8.... .JJ SSJ SJ 8.... SJ SJ JJJ SJ SJ 8 SJJ SJ SJ 8.. 2.... J... ..J SJJ SJ 3.. SJ SJ .0.... SJ 8.... SJ SJ SJ SSJ SSJ SJ .3 SJ SJ SJ 5.. SJ 2... .2 .. .JS.S..S.SSSS..SSS..SSSSSSJ.S..:.S ...J .SJ SSJ SJ ...J S .. S... SSJ SJ S.J SSJ .JJ S.J SJ JSJ SJ J..J SJ SJ SSJ SJ SSJ SJ .S.J SJ SJ 5.. 8J 8 SJ J... 8.... SJ .S.. SS.. 2.. SJ SJ .3 .SJ 2.... S.. 8.... SJ SJJ SJ 3.. 3.... SJ SJJ ...S SJ S..S S... .J.. SJ .... .2 S S. S.S.S.S S S.S.S.S.S.S.S.S.S.S.S S S S... S.S.S S S S.S 2...... SSSJ ..J .SJ SJ .S.J SJ SJ SJ SSJ S.J ..J S.J .JJ SJ :2 SJ SJ SSJ .JJ SJ SJ SJ ..J SJ .3 5.. SJ .JJ 5 .SJ SJJ SJ 3.. .3 SJ SJ SJ 3.. SJ 8.. SJ 5... SJ... SJJ SSJ SJ JSJ SJ 3.. SJ S... SJ SJ 8.. S... .J.. ..J .2 SS .J..JJ..S....S....S......J.J.::S ..SJ ..J SJ 1.. SJ S... 3.. SJ SJ 8.. 3.. SJ 8.. ..J SJ 8.. SJ SS. 8.. SJ SJ ..J J... S... S. SJ 5.. SJ 8 .J.J SJ .3 ..J SSJ SJ SJ JSJ .JJ SJ S.J SSJ ...J SJ.. JSJ SJ .JJ .S... SS .JJ SJ SSJ SSJ SSJ .JJ 8.... 5.. .SJ .2 SS J.JJ..SS.S.JS.J.SS.SSJSJJS.SSES SSJ ...J SJ 2.. ..J SSJ .S.J SJ SSJ SJ ..J SJJ SJ SJ JSJ SJ ..J 8.. SJ SJ 8.. .JJ .S.J .JJ 5.. 8.. SJ SJ 8 S... .3 SJ SJ SSJ .J.. J... 22 SJ JSJ SJ SJ ..J S.J SJ 5.. .J.. 8.. SJ 8.. SJ .... 5.. SJ 2.... SJ .... S... .2 .S SSS.S..SSSS.SSSSSS.SSS.S.SS.S.:J .SJ J..J .SJ .SJ ...J JJJ ..J SJ SJ SJ SJ SSJ SJ SJ SJ SJ SJ SJJ .SJ SJJ JJJ .JJ 3.. SJ 8.. SJ .SJ ..J ._S SS..S SJ SJ SSJ SJ S... SJ SJ .3 .JJ 8.. SJ SJ SJ JJJ SJ 8.. SJ .3. SJ SSJ SJ .3 S..S SJ SJ 2.... .SJ .2 SS SS..SSS.JJSJJJSJJSSJJJJJSJSJSJ...:S SSJ SJJ SJ .SJ SJJ S... SJ .3 ...S .JJ JSJ SJ 3.. SJ SJ SJ :3 SJ SJ .3 5.. SJ SSJ SJ S... .JJ S... 8.. .S S... 1...... SJ SJ 8.. SJ 8.. J... S.. .SJ SJ 2.... SJ SJ S... S... 8.. SJ SJ S... SJ S... J... SJ .SJ SSJ ..J SJ .2 S S... t... SJ... .S... .J... SJ... SS... SJ... .... Sm“. S... S... S... S... S... .S... mS... SJ... SS... SJ... SJ... SLJ... SJ... JS... SM... S... S... S... 38.88.SSSJS.SS.J.cSSSSS.SS.J...SSJS.SSJJJ.JS.J....025: was: >m>u5m mHmAHMQJ mmsom wsu mo H uumm mo Jv 0» mm Emufi ou mmmson mm may .0 20mm mo mucmcwmmu xn Umcmflmmm mxcmu mo mcowumw>m© vumvcmum cam mammz m.< manna m m J m m m m m m o a m m m J m m mom a mi. J m m a o :3. JmmoN VmoN muoN NN.N oo.N mo.N wJoN om.N omoN No.N 9N oooN JmoN oWN JnoN ¢NoN rNoN JmoN mcoN amoJ omN NNnN. SoJ SoJ NJ.N mooN NmoN JvoN am 8J6 tNoJ 3%... SJ. 2..» JJoJ anoJ omen 36 June Jtoo Jame oJoJ oJoJ NJoo omoc mnoJ NnJ Nvom Nmoo 8.0 86 JJN. Jan oJ JJoJ oooJ and $6 3. 3. o o m. m J noJ u o J o m J J c n m a mom a m o o o a J o m m J.:: JanoN :oN 8; JooN mmoN meoN oWN Nm.N JmoN :N onN o‘N goJ NnoN ovoN omoN JnoN JooJ JJ..N JOSN oNoN 8; BoN mJaN otoN nJoN oNoN mJ.N 8 8J3 mNom goJ. mmom «mom 8cm Nmon omoo meow 00$ Jmom 2.0 «now Jmom Jmom coon Jmom Nmom 56 31mm Jeom an; 8.... 26 $6 $3 :6 Jmov 8.. mm a J o m m m J J o o m o o m a J o m J m.J a m ...sJ J 0 mom a a 53. owNoN omoN $.N 0N.N mm; mmoN mNoN mJoN EoJ JnoN NN.N NJ.N JJoN JooN mmoJ mooN JmoN tch o¢oN tmoN mN.N mm; 8; rmoN mooJ Jn.N 8; vooJ am Novoo 35 86 Noon mNoJ nNom mJ.m JJom Nano 8..." SJ. 2.0 cam 33 26 9.0 JNJ Nada onto Nmom JN.J oo.J meoJ OJoo Jooc 36 Son NmoJ f Nm 3 3 OJ 3 OJ OJ 3 OJ OJ 3 OJ OJ 3 3 3 3 OJ OJ 3 3 OJ 3 OJ OJ 2 OJ OJ 3 =4: SNoN JJoN NmoN mmoN vN.N tmoN mmoN mJoN cmoJ nNoN wmoN ONoN NooJ ONoN 8oJ mch SoN nooN mJoN omoJ NJ 8; mmoN omoN SJ 8; meoN mJoN am mJ¢om mmoJ room 36 3.x atom ¢moo wmuJ meow 00.9 no.» 85 vJ.o Jmoc Soc 33 JmoJ @Joo omom 3.6 anon 21¢ «Now Jeom mwom JJ.o mJom JNom 3.. Jo m N m m n v v m noN N m N N N m N v N m N ¢ ¢ N n v N + N 52¢ mo...N mm.N m I..N mm.N mJoN mNoN mm.N mv.N mm.N I N ao.N mwoN Jn.N mmoN JmoN mJoN NmoJ JmoN Nm.N mo.N mv.N vo.N mQN JeoN mo.N omoN mmoN on Jmmor 5...” mm; 33. qu mm.v we; Jo.v mJ..m on...” mmoo 9.; Na...” 9; 3: OJ; 2: Eon mm...“ mNov mmov No; :3. JJ.¢ 3; m¢.m met 55 .t on m m N m m. o m. n J J J o o J u m m m o J n n m N m V n m 52.. mmo.N EoJ wmoN Neon ov.m mmoN Jm.N mo...” NJ.N EN EN oJom oJ.m mch owoN voom JJ.N mNom oJoN 8oN oN.N 8.N Jm.N m¢.N JJ.N JJ..N omoN goo am mmtov NJov JJo¢ mmow o‘m Jeom 3.... or; 2.6 mmoo meow $.o N¢oo meow atom 85 «a; moon 3.... mm.v 2.; econ Joor no; tmom om; meow Eon .2 at 2 v n r N v m N ¢ n m N m r v c v N m m n m m ¢ J. N m N n 52: 93m.N mvoN «NoN mNoN mmoN Jm.J mmoN mm.N mm.N om.N JN.N JN.N Jo.N vmoN JmoN JJ.N mmoN Jm.N ow; Jm.N mmoN VN.N Jm.N mm.N vJ.N «moN o¢.N oooN or. 95mm... mm...” mecm mm.m moon Jm.m Jmom coon mmom Nm.v oNov rm; mn.¢ Jmov Jmov $.¢ Jm.m owot Nvov o¢o¢ :3 Ron Ntov QJoe JJom Joom mmon 39¢ .2 o¢ J J J J J J J J J J J J J J J J J J J J J J J J J J J J 52¢ J8; NQJ mmoJ Jv.J JmoJ Jo.N Jo.N awoN mvo $.J JmoJ mmoJ mm. OJ.J omoJ J¢oJ n¢.N mm.N mJ.N mo...“ mooJ mm; Nm.J mJ.J 5... SJ mooJ mJoJ on VJNoN JJ.J mJ.N mmoJ NN.N omoN omiN wwow mm; JJoN mmoN meoN $.J vJoJ Jm.N JmoJ mmoN mo.m anon NmoN oJ.N $.N vooN mm; mJoN J J ovoN NN.N a... Jr J m m J m m.J m m m m m n m o o m J m J. a J m moJ o 0 new J J 224: mnWN mm.N mmoN mN.N om; JNoN mo.N SN 2; NJ.N 3N oJ. J.N NN.N NQN Nm.N Jm.J 0N.N mm; am.N 8.J Jo.N Nn.N oo.N JJ.N no; mmoN vm.N X...N aw Jami Nmom N¢.m meow 2.3m om. m J96 onm 86 JJ. 0 Jam 25 vJon Jmon mNom mNoJ Nv.m mm.w Nmo¢ 83 0J6 :6 no; «New mNoJ mmom mJow Jeow 8.. mt N v m. J. N N n N moN v v v. m m N N m e N m N N m m n n m ¢ 5.4: NJNoN JmoJ mv.N mJ.N NN.N NvoN mNoN mNoN Jm.J mo.N mmoN mo.N 8; SJ JvoN mNoN mmoN Nm.N mv.N wN.N Jm.J mJ.J NJ.N JN.N nmoJ ¢o.N JmoJ JnoN ow mJNoJ. 85 JJom NN..“ om.m mvom mmov Nmov «Jon 8:. Joov 3.; OJ... Jmov ooov 8..” NJov ovov mm.m Nmom mJ.n NN.m mJov omoc JJom nNov Jmom JJot f at 3?... JJ... mNu: tn... NN.: on: mm... No: :1. mN-z mm... Nmuz man an... NV... Jmu- mm-a mNuz Jm... mNuz mNuz JNu. mNn: on... mN-z Jm-z Jm-x vmuz ...oJOJBNm588SnJNJJJnonJmmmNmJmmvN¢J¢mmNmeMNNNJNnJNJJJ .025: SS8: Jm>usm mHmMHmcm mmsom may no H unmm mo cm on mJ mEmuJ on mmmso: JN man mo comm mo mucmwwmmu an cwcmwmmm mxcmu mo m:0JHMH>mU vumccmpm can mammz v.¢ magma APPENDIX E TABLE OF HORST'S MEASURES OF RELIABILITY ON HOUSE ANALYSIS SURVEY ITEMS I?" APPENDIX E Table A.5 Values of Horst's E! a generalized.measure of reliability, for House Analysis Survey items* Item** Item Item (HAS Pt. I) rH (HAS Pt. II) rH (HAS Pt. II) rH 15 .92 1 .90 33 .38 16 .97 2 .46 34 .40 17 .97 3 -.44 35 .58 18 .72 4 .68 36 .16 19 .91 5 .67 37 .55 20 .68 6 .74 38 .60 21 .91 7 .67 39 .40 22 .88 8 .34 40 .79 23 .89 9 .72 41 .68 24 .79 10 .81 42 .71 25 .83 11 .33 43 .27 26 .81 12 .63 44 .66 27 .77 13 .09 45 .15 28 .83 14 .38 46 .67 29 .86 15 .82 47 .70 30 .84 16 .29 48 .81 31 .30 17 .19 49 .38 32 .67 18 .69 50 .46 33 .78 19 —.09 51 .56 34 .81 20 .72 52 .42 55 .79 21 .26 53 .19 56 .40 22 .40 54 .28 57 -.04 23 .00 55 .15 58 .79 24 .83 56 .31 25 .41 57 .43 26 .22 58 .54 27 —.48 59 .92 28 .72 60 .92 29 .87 61 .83 30 .73 62 .71 31 .54 63 .40 32 -.22 *An explanation of Horst's g is presented on pp. 93-95- **Horst's Efs were not computed for items not scored on a continuous scale. 293 175 3332 mllII. A" H Y." " SI” III E” H N”! " E“, "Ill " H "I "ll 3 1293 03