1‘ ' .I‘, . - - "134 - L";'I;II1§l-Y‘n.‘h I [1.40" ,1, 2:: IQ‘ ' §}'}I' ”ii: WI {lg-2,11% ‘4}:sz : 1 I 1 31L 5 1ng II.I.'"I 1315 'ijfh E '11:; :3; ”'thl‘¥"J-?‘ Mn‘é? 1‘1: L‘hé t} €433", "11’le I III? “A!“ “,5: 2;? 3L ii] "5 1 WI F' ”I. r1133: :Jf‘j' {Ah ”I ‘5qu- .1J' 1":in {PI}: 1'3“. I- 5mg- - .."".‘.’I‘.:, ~.I' ’11II'1IQHTI1I‘é—1..i-Ihinj‘flrhI-I" “11qu II-I'IIr' . ESE“, 113%: . I "4111 H '11..;1I|‘1;1I:I ffiII' 11111 I‘,fihih I 'T’lfrfatgfii , ‘l , ,1.II .r'. 'I [(1 II ,-,I-. -. 1! '1'11’I13-1g1111m1 % '11‘JI-Er £11151: III? 1 ,L 11:, ‘ L ... ‘ g- , . MEL”. ”‘I’; £1143; ‘ I ”IV" ' ' ' ' MI" P’ 3 . I ~ .‘ , r}. Id . , I . I I fund"? IIII’" - VIEIIKII 5:11 ‘I - I ; ' 2 , 73g; .581, ' 3'1? ’IPIIII‘III v. “LII $3.1 1' . ,1 1k. I’If‘ .= IR g.“ E333 I -'1I ‘3'“; .1!" L1“? . 5.2) :5" ‘" I.” W. a . :“L‘I‘I W! O; '9" 15 I 1,4' 2'."- Il . L'fi" 1% I «1; mi :1: ':- a; III‘IIII .III 1 If” -" Iv ' ~' =34 "-5'...-I I Ih' :51. I ‘ 'I '1 'i- I a! «I 1 . In 51'] I a??? I" 3"? 3:35.353? I . ‘ UIWayfimffirara fififfiI- .21.; .. .1. “1“". C "'1 $11.3“? '1?qu I v.5 . 1If*gnr%fk . 31,-“ I. i . Iii-III - IT. -I 3' RN? .I. 5 I” , 1 1 '11 “I; “rig ‘EF‘V‘: [It Laffi‘ui‘I 1' 5L1r‘11‘1'v111l’ ~11“? ‘ . rim @f :. “Fl I. : Ii': ['51 41}; fi'I ‘Ifl fIau .' ’ Ill Il"'.«.IL_.I .. ISL - . ' - ' ' .. 1 J11 'g‘II Ix': 3?: I I . 'I ' . I v 7,1I 'f 1‘ ; '1' “‘0‘ .3-L‘1 : ".‘ " . .:-‘ ' I 1 ‘- I"); I “ {:15 ’ .l I. "I ‘0‘”. . " f, ' I‘.:5 '4 I . -. «.5 "g1" -. V - _' '1 ' ~ _ -I ‘ .°- ' - ' k' . "a .. '| "arr.I I I .l ‘ IIII .- ‘ " ' 'Jiz6;" "‘I ‘ II. .I- ' ' .. -s'?‘ ”‘:L.‘ 'LI 1 L!“ I ’ 1 . ' . I ‘ " ' - " . I‘lxi. 'li ' 31"" « I E;;I.II‘I:‘1I""I , I- 1 1J3! "- 'lt‘H 1t”: 3‘ ‘0 I .‘ ' _I “I. 1 1 5:1,. 'I 1“-‘fI’._I. :III1. I - ' 5v ‘ 1:131"th 11,":I in“; ' | . . ‘TI 0... I..-'I 31‘ - .I. I III. I .z- _. I .I-- :I, .‘ "‘ ;, 0:1: .‘7' ::!I,:,:',I.'II $7.11!,“ ,fI. . .I' .. «I Irvimmflfl"mux“‘ .. IL. I -I II I ‘1 1 l| I lu,“ um:- «I In. .- . ..,.,I- .- :w-II‘I - .- - :13“; 'II'f I :_,:.I EI'lfl‘I‘lzwu 37"13' I? II ICII. "‘ ‘ ' ie'1,,..I. I-~.II--I I.I .; I. r I 11 -;I" . . 'I II IH.‘IJI.‘IIiI ' I I"I’Il _' . II' I 1 II 1.‘ I" .0. “:1II I (1] . I'M”. 1r ' .. ~.I . ”H1...” .. I1... I .1 ‘ M .V I . ‘1 ”I111..,1. NHL“, “5., . I‘. L 1 1. 1 (A. ' - .1 I: . 1 7:: .I . II‘TI'iI “.I'i'.’ I 5 7'4"?” I"? ' ' .'..I1[l"’1‘I"I:|' II In...“ ‘I'III 11.011 . ,3. .I. ,l :1) .‘II. Y:..."'I ", 1' II M. ,' .. I .4.‘ .I' " ‘ .'. .I- I II .I 31': ,1. ‘ ..'II,'1":III . - I, 1‘ '1 ';. I rig-1'1. , 5w. I" I I7- I . 'H I u' [FINN 'II'I- I H’!‘ 1-". "g; :12“ I .I:' :3 V g‘II'II'I”I‘”4III"' I- I '- w? I- I . "—4 ‘SII . - W .I"III :~_.I:1.'.- ..,. 13;?" II-‘I 2 - «I’ :aI'I‘I' III 1qu‘ I. «"1 ' III: . .. «III ”W 5'3 III-EIMIII-M" 4% I?“ "III“? iI‘ Jim I“ I ' I [b14114] :- I'II' .I L ’I I I Ii.“ "I ' :HIEI;"1‘IXI""I-I.:r 'L‘ f ‘ u ' ‘ .. - I I , I i-‘IIII‘IIII .t -II. 'I'.‘.. III . . If ' II I . - , “ . Is,I|'!Ii‘II “I“ . II,-I-‘I'I;.-I-. ..I , . I .I': ... (Im- II '1'}: .‘- " ’ *‘I-I' «I' I'I'H "‘II' 1'I‘I...I',.I‘II«.1 ‘.I “1.21131“. “LI“; ARI ' I .. ~'I .. I I... .I II .. . 43...»! «Ix-«a I m. 131., I . III: '7 ‘ Mfg.” . H ”I ' II" I :‘I'.1I_.| ‘5 ". .1I'1Yg" I J IIY H “.4.” nII .3 I I I , .II.II ; 'o.l"_'.l.‘I.I 1 "1. I...» r3": . ’1! 'I quid“: {3g 1?? - i‘:-I “I. . -I' LI:- bi I\‘ h I' ‘ «I- I". A' "-30 " ‘J‘ Sigh; v‘ I ' ‘ ”if. '1" ' ‘1' I- ' “.3 I I't‘ 1:.1:’:I7'I:j if“. ' ..III. IIII'IIo ' “1'1“ h {IIIJIN‘I' .I’P 5.21 I III I’r'1 In“? .‘I 4.51.1.2: XIII" I'III. "113E935 131131342“ .5155? ‘ it -I~}..I‘I. 1 II (If #JgpifiI-Ivomngffimt i-‘gfififidht' I A I? I --I -I«¢ “nhfi II ' ' " . Vi inc‘figrfifi11n";i&"£w:¢f Pam. "9] $111,411 J{ 1%- I _I 2 ., J IE: I. 17315154,»: ’"Cj ' z"?- ‘2’: :[Iv'i‘fl- 34.4 {1'1 Ifii Pk 5: WI "I“ .. 1H1 .. II; ‘1'“; If .I \ I1 ”I'ffaI'jruz‘AI'I 11410 £5th I-f-I I3 "1'th M" II? I“ .3. ,m, . 391.? IMI I :LI‘NI‘VQ‘F' If”; 3' .‘I . swab ISIEIII I I... 1‘ “I‘ifiizfi‘fjg THESlS lllllllllllllllll”IlllllllllllllllIllllllllllllllll 3 1293 01050 35 This is to certify that the dissertation entitled A Study of the Relationship Between the Perception of Selected Variables Within the Residence Hall Environment and College Student Drinking presented by Julia Ann Hower has been accepted towards fulfillment of the requirements for PHD degree in College and University Administration Uofessor Date August 21, 1996 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 v__.‘_i 4 ‘- '- W‘ 'W v'uwr—"v‘V—‘T'V ' or v V‘"" LIBRARY Michigan State University ' mes III RETURN BOX to remove this checkout from your mom. TO AVOID FINES mum on or before date duo. DATE DUE DATE DUE DATE DUE --A——————'- A STUDY OF THE RELATIONSHIP BETWEEN THE PERCEPTION OF SELECTED VARIABLES WITHIN THE RESIDENCE HALL ENVIRONMENT AND COLLEGE STUDENT DRINKING By Julia Ann Hower A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Educational Administration 1 996 ABSTRACT A STUDY OF THE RELATIONSHIP BETWEEN THE PERCEPTION OF SELECTED VARIABLES WITHIN THE RESIDENCE HALL ENVIRONMENT AND COLLEGE STUDENT DRINKING By Julia Ann Hower This study employs variables from the social-psychological framework, Problem Behavior Theory, developed by Richard lessor. The purpose of the study was examine the relationship between college student drinking and social influence variables within the residence hall environment. A random sample Of college freshmen were surveyed regarding their perception of their close friends who lived on the same floor, their roommate and Resident Assistant with regard to those individuals modeling for drinking, pressure for drinking, approval for drinking and controls against drinking. Respondents were also asked about their motivations to consume alcohol, quantity and fi'equency Of alcohol consumption and negative consequences experienced as a result of alcohol use. Respondents were classified as either non-drinkers, non-problem drinkers, or problem drinkers based upon their frequency of intoxication and number of negative consequences experienced within the first six of school. Sixteen statistical hypotheses were tested using t-tests, analysis of variance, multiple regression and discriminant analysis. Significant differences were found for fourteen of the sixteen hypotheses tested. Students residing in alcohol free rooms perceived fewer models for drinking, less pressure for drinking, less approval for drinking and more controls against alcohol related problem behavior . The findings supported the directional hypotheses suggested by Problem Behavior Theory. Problem drinkers perceived more approval for drinking, modeling for drinking, pressure for drinking and less controls against alcohol related problem behavior from their close friends on the floor and their roommate than either non-drinkers or non-problem drinkers. Discriminant analyses revealed that the perceived environment variables were more useful in classifying non-drinkers and problem drinkers (88% accuracy rate) than non-problem drinkers and drinkers (76% accuracy rate). TO MAH. Without whose support, guidance and love this dissertation would never have been completed. iv ACKNOWLEDGEMENT S I would like to acknowledge and thank my advisor, Dr. Kathryn Moore, for the valuable assistance she gave me in completing this dissertation. Her support academically, financially, and emotionally were indispensable. I would also like to thank my committee members, Drs. Anna Neumann, Anne Austin and Linda Forrest for their help and guidance with this dissertation. I wish to thank Dr. Richard Brandenburg and the stafi‘ of the Office of Academic and Student Afl‘airs within the College of Agriculture and Natural Resources for their support over the past two years. My graduate assistant position was instrumental in allowing me to focus on and complete my dissertation. I want to thank my friend Cindy Helrnan for without her encouragement I would not have taken my first doctoral level class as well as Kay Ketzenberger for her contimrous support and especially her assistance with discriminant analysis. lastly, I wish to acknowledge the faculty, stafl‘ and students of Lyman Briggs School for their understanding of my need to complete my degree and that in order to do so, I had to leave my position as Director of Student Affairs. My experience in Briggs was very rewarding and will always hold many fond memories for me. Special thanks to my former colleagues and now dear fiiends, Ed Ingraham and Sandy Conner, for their on-going support, guidance, and friendship. TABLE OF CONTENTS LIST OF TABLES ....................................................................................................... viii LIST OF FIGURES ......................................................................................................... x CHAPTER 1 Introduction and Statement of the Problem ...................................................................... 1 Purpose of the Study ....................................................................................................... 9 Research Questions .......................................................................................................... 9 Significance of the Study ............................................................................................... 11 Limitations of the Study ................................................................................................. 12 Organization of the Study .............................................................................................. 12 CHAPTER 2 REVIEW OF THE LITERATURE Introduction ............................................................................................................... 14 College Student Drinking Patterns ................................................................................. 15 Demograme Differences in Alcohol Use ....................................................................... 18 Individual and Personality Correlates of College Student Alcohol Use .......................................................................................... 21 Social/Environmental Correlates of College Student Alcohol Use .............................................................................. 23 The Residence hall Environment and College Student Drinking ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, 27 Problem Behavior Theory and the Perceived Environment ............................................. 31 Summary ............................................................................................................... 37 CHAPTER 3 METHODOLOGY Population and SamPle .................................................................................................. 38 Data Collection .............................................................................................................. 40 Timing of Data Collection .............................................................................................. 41 Instrumentation .............................................................................................................. 42 Statistical Analyses ........................................................................................................ 46 Research Hypotheses ..................................................................................................... 47 Definition of Terms 48 oooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo vi CHAPTER 4 FINDINGS Description of Sarnplc .................................................................................................... 51 Establishment of Drinking Classification ........................................................................ 52 Drinking Classification and Demographic Variables ....................................................... 55 Perceived Environment Variables and Demographic Variables ....................................... 62 Perceived Environment Variables and Drinking Classification ........................................ 73 Prediction of Drinking Classification .............................................................................. 76 Summary of Statistical Hypotheses ................................................................................ 88 CHAPTER 5 CONCLUSIONS Summary of Findings ..................................................................................................... 93 Irnplications for Practice .............................................................................................. 105 Limitations ............................................................................................................. 108 Recommendations for Future Research ........................................................................ 111 APPENDICES Appendix A Residence Hall Survey .......................................................................... 113 Appendix BI Introductory Letter ............................................................................... 120 Appendix C First Letter of Transmittal ..................................................................... 121 Appendix D: Second Letter of Transmittal ................................................................ 122 Appendix E: Third Letter of Transmittal ................................................................... 123 Appendix F: Return Postcard .................................................................................... 124 Appendix G: Letter of Approval for Study ................................................................ 125 Appendix H: Young Adult Survey Request ............................................................... 126 Appendix 1: Questionnaire-Variable Code ................................................................ 127 BIBLIOGRAPHY 133 ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo vii 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18 4.19 LIST OF TABLES Frequency Distribution of Demographic Variables .................................. 54 Frequency Distribution of Drinking Classification ................................... 55 Frequency Distribution of Drinking Classification by Gender .................. 56 Frequency Distribution of Drinking Classification by Size of Residence Hall ............................................................................ 57 Frequency Distribution of Drinking Classification by Type of Residence Hall .......................................................................... 59 Frequency Distribution of Drinking Classification by Alcohol Free Floor Residency ................................................................ 59 Percentage of Non-drinkers and Drinkers Residing on Alcohol Free Floors ................................................................................ 60 Frequency Distribution of Drinking Classification by by Residential Academic Program .......................................................... 62 One-way MANOVA and Univariate F-tests for Perceived Environment Variables by Gender, Type of Residence Hall, Size of Residence Hall, Specialty Housing Option, and Residential Academic Program ............................................................... 65 T-tests of Means of Perceived Environment Variables by Gender ........... 66 Analysis of Variance Of Perceived Environment Variables by Size Of Residence Hall ............................................................................ 68 Analysis of Variance of Perceived Environment Variables by Type of Residence Hall .......................................................................... 69 T-tests of Means of Perceived Environment Variables by Alcohol Free Room ................................................................................ 70 T-tests of Means of Perceived Environment Variables by Alcohol Free Floor ................................................................................. 72 T-tests of Means of Perceived Environment Variables by Residential Academic Program Enrollment ............................................. 73 One-way MANOVA for Drinking Classification and Perceived Environment Variables ..................................................... 74 Mean Scores of Perceived Environment Variables for Non-drinkers, Non-problem, and Problem Drinkers ...................................................... 75 Discriminant Functions for Predicting Non-drinkers, Non-problem drinkers and Problem Drinkers ......................................... 77 Discriminant Analysis of Non-drinkers, Non-problem drinkers, and Problem Drinkers ............................................................................. 78 viii 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28 Classification Results for Non-drinkers, Non-problem drinkers and Problem Drinkers ................................................................................... 81 Discriminant Functions for Non-problem Drinkers and Problem Drinkers ................................................................................................. 82 Discriminant Analysis of Non-problem Drinkers and Problem Drinkers ................................................................................................. 83 Classification Results for Non-problem Drinkers and Problem Drinkers ................................................................................................. 84 Discriminant Functions for Non-drinkers and Problem Drinkers ................................................................................................. 84 Discriminant Analysis of Non-drinkers and Problem Drinkers ................. 85 Classification Results for Non-drinkers and Problem Drinkers ................ 86 Multiple Regression Analysis Predicting Frequency of Negative Consequences .......................................................................... 87 Multiple Regression Analysis Predicting Frequency Of Times Drunk .......................................................................................... 88 ix Figure 2.1 Figure 4.1 Figure 4.2 Figure 4.3 LIST OF FIGURES The Conceptual Structure of Problem Behavior Theory .............. 32 Territorial Map for Discriminant Analysis of Non-drinkers, Non-problem Drinkers, and Problem Drinkers ...... 8O All-group Stacked Histogram for Canonical Discriminant Functions for Non-problem Drinkers and Problem Drinkers ....... 82 All-groups Stacked Histogram for Canonical Discriminant Functions for Non-drinkers and Problem Drinkers ...................... 85 Chapter 1 INTRODUCTION AND STATEMENT OF THE PROBLEM Alcohol abuse is the major substance abuse problem on college campuses today (Gonzalez, 1993). Across the nation, college and university presidents cite alcohol abuse as their greatest concern within the campus environment (Carnegie Report, 1990). Students experience extremely negative consequences from alcohol abuse including health problems, property damage, poor academic performance, legal or institutional disciplinary problems, violence and death (Berkowitz and Perkins, 1986; Burrell, 1992; Shore & River, 1985; Strange & Schmidt, 1979). Institutions face adverse publicity, expanded tort liability laws, increased social host and property owner responsibilities, and student attrition (American Council on Education, 1988; Clay, 1989). A comparison of a 1982 and a 1991 nationwide survey of student drinking patterns found more students in 1991 reporting problems with the law, getting into fights, engaging in acts of vandalism, missing class, earning poor grades and losing jobs as a result of their drinking (Engs & Hanson, 1992). Students who drink heavily report lower GPAs than students who drink moderately or abstain (Lall & Schandler, 1991; Maney, 1990; Presley, Meilman & Lyerla, 1993). Over seven percent Of the 1991 freshman class will leave school for alcohol-related reasons (Eigen, 1991). The financial costs of alcohol abuse are staggering for both students and institutions. Students spend $5.5 billion dollars on alcohol, more than they spend on soft drinks, tea, milk, juice, coffee or books combined (CASA, 1994). As the New York Times reported (1993): "Though intoxicated students routinely smash toilets, yank out sinks, punch through ceilings, head butt street lamps, uproot ornamental trees and body- slam vending machines, few are apprehended. Repairs become a hidden cost underwritten by the taxpayer or parent or squeezed from the library or faculty salary budget” (Matthews, 1993, p. 38). One in three college students drinks primarily to get drunk (W echsler & Issac, 1992). In a 1991 survey of 56,000 students, 42% reported engaging in binge drinking (consuming five or more drinks at one occasion) at least once during a period of two weeks (New York Times, 1992). Despite the fact that 80% of all college students are underage and cannot drink legally, 96% report that they have drunk in the past year (Johnston, O‘Malley & Bachman, 1992b). The US. Department Of Health and Human Services Oflice of Substance Abuse Prevention (1991) describes the alcohol problem on college campuses as essentially one of culture and environment and suggests that the solution to this problems lies in changing the social norms and behaviors of college student life. In Alcoholism/Chem Demendency and the College Student (in Rivinus, 1988, p. xi), Ernest Boyer expressed this concern: All human communities have their dark side, and college communities are no exception. As the papers in this issue of the Joum_al Of College Student Psychotherapy remind us, the same environment that fosters personal growth, learning and commitment to others, may also intensify the dangers of drug and alcohol abuse. . . .While the pleasures students derive fiom drinldng and drugs are public, the pain has been privately suffered in lost days, lost selves and tragically, lost lives. Colleges long have been reluctant to take responsibility for these causalities of campus life. The Collgg'gtte Environment and Alcohol Use College alcohol use and misuse is a complex function of both person and social environmental variables (Brennan, et al, 1986; Goodwin, 1989; Maney, 1990). The investigation of social influences within the collegiate environment (e. g. peer influence, modeling, norms, social context) and its affect on student alcohol use has become a major interest to researchers in both higher education and social psychology (Berkowitz and Perkins, 1986b; Goodwin, 1989; Jessor, 1981; Martin & Hoffman, 1993; Moos, 1979; Perkins, 1991; Prentice & Miller, 1983). College campuses provide an environment in which peer interaction and socialization are intensive and drinking among college students is strongly influenced by peer norms (Baer et al. 1991; Perkins, 1991). Alcohol use is not only encouraged in the college social environment, but negative social sanctions are imposed upon those who do not drink (Cronin & Ballenger, 1991', Shore, 1983). Students' perception of the university as permissive along with inconsistent policies and double standards regarding alcohol use may contribute to increased consumption (Guyton 1989). In discussing the importance Of the campus environment on student alcohol use, Robert DuPont (Rivinus, 1988, p. 41) states: "If I were going to create an environment that encouraged chemical dependence 1 could hardly improve on the contemporary American College." One of the most fiequent criticisms Of the research in the area of college student alcohol use is the lack Of consideration of multiple factors predicting alcohol misuse (Barnes, Welte, & Dintcheff, 1992) as well as the lack of theoretically based studies (Saltz & Elandt, 1986). Few studies have examined more than one or two variables at a time (Brennan, Walfish, & Aubuchon, 1986), and there is a continual call for studies to explore relationships among a variety of demographic, personality, environmental factors and multiple measures of college alcohol use and abuse to increase our understanding of the complex nature of student alcohol use (Brennan et al, 1986', Saltz & Elandt, 1986). In short, a good deal is known about how students abuse alcohol in the collegiate setting, but far less is known about the institutionally induced forms Of exposure to alcohol misuse. The Res_idence Hall Environment agd Alcohol Use One of the most intensive peer environments on campus is the residence hall. Many colleges and universities require students to live on campus for a portion of their college careers. Residency requirements are designed to assist students in integrating into the academic and social realms of the institution. On-campus residency has been found to exert a positive influence on persistence and academic achievement (Pascarella & Terenzini, 1991). Residing on campus, however, may not only expose students to the positive aspects of institutional community and culture but to the negative aspects as well as alcohol abuse. Several studies have established a relationship between alcohol use and living environment (Bachman et al, 1984; Barnes et al. 1992; Jones et al., 1992; Martin & Hoffman, 1993', Saltz & Elandt, 1986; Schall et al., 1992). Both residence hall context and type of housing have been identified as variables associated with drinking behavior among college students (Igra & Moos, 1979; Perkins & Berkowitz, 1986). Shore and Rivers (1985) found that an individual will be a drinker to the extent that the living arrangement he or she has entered can be characterized as having a strong social component. Igra and Moos (1979) found that the drinking context of the residence hall, informal social involvement in college and the lack of commitment to religious and academic values were all related to drinking. In a study Of factors associated with alcohol use among university students, Schall et al, (1992. P. 134) states: Students living in a university environment where there is social pressure to drink, where heavy drinking is approved and where alcoholic beverages and the places to consume them are readily available will drink relatively heavily on the average. . Dormitories and especially fraternities and sororities are living environments within a university fostering the above characteristics and therefore promoting relatively heavy drinking. Martin and Hoffman (1993) conclude that it is difficult to examine college student alcohol use without examining the living environment. There has been a limited amount of research on specific variables within the residence hall setting and their relationship with college student alcohol use. Past research on the effect of college residence on alcohol use has identified differences between fraternities but have ignored variations among dormitory living groups (Igra & Moos, 1979). Most research on college residence halls and alcohol use has categorized living environment as a dichotomous variable, eg. either a student lives in a dormitory or does not live in a dormitory, rather than examining specific variables Kim the environment of dormitory life. An examination of social behavior in a social context, such as college student alcohol use in a university residence hall, requires the study of both the person and his or her environment. The investigation of the specific personal and interpersonal mechanisms that transmit and mediate the influence of the collegiate environment requires a theoretical framework which encompasses a broad spectrum of variables documented to be associated with problem drinking. The theory base that will be utilized in this study is Jessor and Jessor's Problem Behavior Theory which was developed to account for variation in adolescent and young adult involvement in a variety of problem behaviors including problem drinking. Problem BeLavior TheOLV Problem Behavior Theory (J essor & Jessor, 1977) represents a usefirl social- psychological framework to study problem alcohol use among college students. The Jessors and their colleagues have presented impressive support for Problem Behavior Theory in cross-sectional and longitudinal studies of high school youth and college students (Costa, Jessor, & Donovan, 1989; Donovan, Jessor & Costa, 1988; 1991; Donovan & Jessor, 1985; Donovan, Jessor, & Jessor, 1983; Donovan, Jessor & Costa, 1991; Donovan, Jessor & Costa, 1988; Donovan & Jessor, 1985; Donovan, Jessor & Jessor, 1983; Jessor 1991; Jessor, Chase & Donovan, 1980; Jessor & Jessor, 1977; Jessor & Jessor, 1975; Jessor et al, 1968; Sadava, 1984). The concepts and measures Of Problem-Behavior Theory have been used in a large number of studies and the theory has been used to examine a variety of adolescent problem behaviors (Chassin, 1981; Chassin, Presson, & Sherman, 1984; DiTecco & Schlegal, 1982; Rooney & Wright, 1982; Sadava; 1985; Sadava & Forsyth, 1977; Smith, Canter & Room, 1989). Problem behaviors are "behaviors that have been defined socially as a problem, as a source of concern or as undesirable by the norms of conventional society and their occurrence usually elicits some kind of social control response” (Donovan, et al, 1991). Problem Behavior Theory suggests that problem behavior among adolescents is the result Of interaction of demographic, social-psychological, and behavioral factors (Robinson, 1987). The primary foci of Problem-Behavior Theory is three systems of psychosocial influence: 1) the Personality System which includes attitudes, values, beliefs, knowledge and expectations of adolescents regarding alcohol use, 2) the Perceived Environment System which includes the influence of parents and fiiends attitudes and behavior on adolescent use, and 3) the Behavior System which represents the degree of involvement in both problem behavior and conventional behavior by the adolescent. The variables within each of the three systems reflect both instigations for problem behavior and controls against it, and in combination suggest a state of proneness which is the likelihood of the occurrence of problem behavior; including problem drinking. Of primary interest to this researcher was the Perceived Environment system and its applicability to exploring facets Of the residence hall environment that are associated with alcohol misuse. Jessor (1981; 1973) maintains that a "logical analysis of the nature of environmental influence on behavior makes clear that it is the perceived, or meaningful, or phenomenal environment with which behavior is most invariant." Within the Perceived Environment system, problem behavior is related to perceptions of low controls from significant others, and Of approval for and models for engaging in problem behavior. Greater approval, pressure and models for alcohol use are variables in the perceived environment system that appear to increase the likelihood of problem drinking. Jessor utilizes both proximal and distal variables in analyses (J essor, 1973). Proximal variables are directly associated with problem behavior while distal variables are linked to problem behavior only "indirectly and by way of theory (Jessor, 1991)”. To measure specific attributes of the residence hall environment that may influence drinking, the author modified the variables to capture the salient social relationships of the residence hall. This modification is supported by Jessor's definition of the Perceived Environment as the one that ”has the most invariant relation with the behavior since it is the environment of immediate meaning and the one to which the actor is responding" (Jessor, 1977). Therefore, the Distal Structure variables include controls through roommate, fiiends on the floor and their Resident Assistant. Within the Proximal Structure, the variables include approval for drinking by roommate, fiiends on floor and Resident Assistant; models for drinking by roommate and fiiends on floor; and pressure for drinking by roommate and friends on the floor. PURPOSE OF THE STUDY This study had three primary purposes: To examine the drinking behavior of new full-time freshmen students who live on campus during their first six weeks at Michigan State University. The alcohol consumption variables were examined in relation to gender, type and size of residence hall, specialty housing Option (i.e. alcohol fi'ee room and alcohol free floor residency) and enrollment in a residential academic program. To investigate the relationship between factors within the perceived residence hall environment and drinking behavior among freshman students. More specifically, this study investigated the balance between instigations for and controls against problem alcohol use within both the distal and proximal structures of the perceived residence hall environment. To test the usefulness of Problem Behavior Theory in accounting for variation in drinking behavior among students and across gender, size and type of residence hall, specialty housing options and residential academic program. RESEARCH QUESTIONS This study addressed the question Of whether a relationship exists between students’ problem drinking and specific environmental factors within the residence hall. This study attempted to answer the following questions: 1. Are there significant differences in the self-reported drinking behaviors Of Michigan State University freshman when categorized by gender, size and type of residence hall, specialty housing option or residential academic program? Are there significant differences in the perception Of the perceived environment variables (approval of drinking, models of drinking, pressure for drinking, and controls) among Michigan State University Freshman when categorized by gender, size and type of residence hall, specialty housing option or residential academic program? Are there significant differences between Non-drinkers, Non-problem drinkers, and Problem drinkers in their perception of the perceived environment variables of approval for drinking, models for drinking, pressure for drinking and controls 10 against drinking? 4. Do the perceived environment variables of controls, approval of drinking, models for drinking and pressure for drinking combine to predict drinking group membership? METHODOLOGY Collection of Data TO obtain the necessary data for the study, the Young Adult Follow Up Study 1981 Annual Questionnaire (Jessor, 1981) was administered to a random sample of new fieshman students residing in a campus residence halls on the Michigan State Campus during the fall semester 1995. A random sample of 1800 students was generated from the MSU registrar oflice. The self-administered questionnaire was mailed directly to each individual, preceded by an invitation describing the study and inviting the student to participate in the study. A follow-up letter and second copy of the questionnaire was sent two weeks later to all non-respondents. Two weeks after the second letter, another letter and survey were sent asking the respondents to return the completed questionnaire. Subjects were asked to return a separate post card at the same time they returned their questionnaire. They were informed that the post card received by the deadline would be entered into a drawing for a gift certificate from the MSU bookstore. Instrumentation The questionnaire used in this study contained selected measures from the Young Adult Questionnaire (Jessor and Jessor, 1981), which consists of psychometrically developed measures designed to assess the personality, perceived environment and ll behavior system variables of Problem-Behavior Theory. The perceived environment system and behavior system sections of the questionnaire were used in this study. The perceived environment system consists of the distal structure and proximal structure. Three variables form the distal structure including: roommate, fiiends on floor and Resident Assistant controls. The proximal structure includes roommate, fiiends on floor, and Resident Assistant approval of drinking; roommate, and fiiends on floor models for drinking ; roommate and fiiends on floor pressure for drinking. Many psychometric tests of validity and reliability have been conducted on the Young Adult Questionnaire. Several studies have supported its use as a valid and reliable research tool with college students (Donovan & Jessor, 1978; Jessor, 1987; Jessor, Donovan & Costa, 1991; Jessor & Jessor, 1978; Jessor & Jessor, 1973;; Robinson et al., 1987; Sadava, 1985). mam The research questions were investigated using an instrument developed by Jessor (1981) that explores the Perceived Environment System and Behavior System in relation to problem drinking. Survey responses were examined at the .05 level of significance, using analysis of variance (AN OVA), t-tests, multivariate analysis of variance (MANOVA), multiple regression and discriminant analysis. SIGNIFICANCE OF THE STUDY Higher education spends millions of dollars each year on alcohol-related costs in the areas of education, prevention and policy enforcement. Programs and policies must be 12 based upon a thorough understanding of factors associated with substance abuse to be effective. Studies such as this one may enable student affairs professionals to better identify the kinds of students who, because of their personal characteristics or the characteristics of the environments in which they live, are at risk for problem drinking. Additionally, programmatic efforts should target those factors that are amenable to change in a way that could result in a subsequent reduction in the incidence of substance abuse. In the case of this study, this may include hall, floor and/or room assignment, roommate assignment, specialty housing, support for student involvement programs, and staff training in the areas of education and policy enforcement. L'mitatigns 1. The instrument includes measures from only the perceived environment system and the behavior system. Personality measures were not included. 2. The survey items used were modified from Jessor’s Young Adult Questionnaire (1981). Respondents were asked to answer the questions in relation to specific relationships within the residence hall or to their experience at Michigan State University. The validity of the original items may have been affected by these modifications. 3. The population included only freshman students from Michigan State University and the results are generalizable only to this population. Organization of the Study The study is organized into five chapters in addition to the appendices. The second chapter includes a review of the literature relevant to college student drinking including the social-psychological framework of Problem Behavior Theory. The design 13 and methodology used in collecting and analyzing the data are presented in Chapter III. The fourth chapter contains an analysis of the data. The fifth chapter includes a summary of the study, findings, conclusions, implications for the practitioner in student affairs and recommendations for future research. Chapter 2 REVIEW OF THE LITERATURE Introduction In this chapter, a review of the literature related to the following areas of college student alcohol use is presented: (1) college student drinking patterns, (2) the individual and personality correlates related to college student alcohol use, (3) the social and environmental correlates related to college student alcohol use, (4) the residence hall environment and its relationship to college student drinking patterns, (5) Problem Behavior Theory and the Perceived Environment System (J essor, 197 7) and (6) Perceived Environment Variables within the Residence Hall Environment. The first section includes a review of the literature on the nature of college student drinking behavior with a description of the fi'equency and amount of consumption, the phenomenon of binge drinking, and the negative consequences fi'equently experienced by students who misuse alcohol. The second section focuses upon individual and personality correlates related to alcohol use among college students including attitudes, beliefs, and motivations, as well as demographic differences associated with alcohol use. Section three includes a review of research documenting the relationship between alcohol use and peer influence, peer modeling and the social context of drinking. The fourth section contains an examination of research that includes the residence hall environment as a contributing 14 15 factor to college student drinking. Section five is an overview Of Jessor's Problem Behavior Theory as a social-psychological fiamework through which to examine the relationship between the residence hall environment and alcohol consumption. The last section provides a description of the factors within the Perceived Environment System of the residence hall that are of primary interest in this study. College Student Drinking Patterns The literature on alcohol use among college students is extensive. A substantial amount Of the research has focused on documenting the frequency and pattern of college student drinking. Due to inconsistencies in the quality, nature and methodology of the research, however, many large-scale regional and national studies have produced widely variant results (Berg, 1970; Berkowitz & Perkins, 1986). The use of college students as convenient samples and different operational and theoretical definitions of drinking have resulted in reports of campus alcohol misuse ranging from 6% to 72% in various studies (Berkowitz & Perkins, 1986; Liljestrand, 1993; Meilman et al., 1990). In response to this criticism, several researchers have conducted studies using sophisticated sampling and analysis procedures and have found that the percentage of American college students who have consumed alcohol "in the last 12 months” has remained stable at approximately 90% (Johnson et al., 1989; Meilman, 1990), with rates of alcohol misuse averaging between 20 and 25% depending upon the criteria used (Barnes, et al, 1992). There is little question that alcohol is the drug Of choice on college campuses. Over 90% of students report at least occasional use of alcohol and this figure has remained constant over several decades (Blane & Hewitt, 1977; Hughes & Dodder, 1983; 16 Johnston, 1991; Wechsler & McFadden, 1979). Johnson et al., (1992) assert that ninety- three percent of college students report having consumed alcohol in their lifetime and 96% of these students drank in the last year and of these, four out of five drank in the last month. In a large scale study Of over 7000 students at 34 New England colleges, Wechsler & McFadden (1977) found that 30% of the men and 13% Of the women were "fiequent-heavy" drinkers. Binge Drinking Binge drinking is defined as consuming more than five drinks in a row and is associated with unplanned and unsafe sexual activity, physical and sexual assault, unintentional injuries, criminal violations, interpersonal problems, physical or cognitive impairment and poor academic performance (Hanson & Engs, 1992; Johnson, 1993; Wechsler & Issac, 1992; Wechsler, et al., 1994). In a study of over 17,000 students at 140 schools, Wechsler (1994) found that almost half (44%) of college students were binge drinkers, including almost one fifth (19%) who were frequent binge drinkers. One fifth (22%) of all college students reported they had been binge drinkers in high school whereas another fifth (22%) became binge drinkers in college (W echsler et al., 1994). Utilizing the CORE Alcohol and Drug survey, the Department of Education has surveyed over 58,000 students at 78 institutions to gather baseline and trend data regarding alcohol use among college students. Data from this survey revealed binge drinking is related to age as 47% of students under 21 reported binge drinking as compared to 35% of legal drinkers over 21 (Presley et al., 1993). Negative Consequences 17 The prevalence of students who engage in problem drinking ranges from 6% to 72%. Men and women report different types of negative consequences as a result of problem drinking. Men are almost ten times more likely to get into physical fights after drinking and three to four more times likely to have an automobile accident than women. One recent study that assessed a broad range of alcohol-related problems among undergraduates of one eastern university found somewhat greater incidences for men on several specific items, yet noted no significant gender difl‘erencc in overall reporting of problems (O'Hare, 1990). Wechsler and Issac’s (1991) recent study of first-year students at Massachusetts colleges found an almost three to one male/female ratio of frequent heavy drinkers and four to ten times greater incidences of first-year men than women getting into physical fights, trouble with police, or property damage after drinking. Less Of a gender gap (1.5 of women to 1.0 men) was found for consequences such as a hangover, doing something regrettable, memory loss, missed classes, and getting behind academically. Problems associated with alcohol use by college students appear to be increasing except in the area of drinking and driving. In a national study of college student drinking problems from 1982 to 1991, Engs and Hanson (1992) report that there was no significant change in the proportion of heavy drinkers (i.e. those who consumed more than five drinks on any one occasion at least once per week) across the nine years of research (1982- 24.4%; l985=24.6%; l988=25.7%, 1991=26.8%). Increases in the frequency of the following ten problems were observed: hangovers, cutting classes, being criticized for their drinking by someone the person is dating, trouble with school administration, l8 damaged property, pulling a false fire alarm or other such behavior, vomiting, missing a class because of a hangover, legal trouble, fighting after drinking. Significant decreases in the frequency of the following problems were noted: coming to class after having several drinks, driving a car after having several drinks, driving a car when they knew they had drunk too much, and driving a car while drinking. Pre-College Drinking Patterns Several studies have focused on the nature of drinking patterns students bring with them to college. Between 41 and 61% of college students report that they had their first drink between the ages of 14 and 17 (Banks & Smith, 1980; Gonzalez, 1979). Banks and Smith (1980) report that a small percentage of college students (6%) had their first drink at 18 or Older. In review of several studies, Saltz and Elandt (1986) found that the research on change in drinking patterns from high school to college was mixed. Banks and Smith (1980) and Barnes and Welte (1983) both found that students are likely to decrease their alcohol use over the course of time fi'om high school to their senior year in college. Three other studies (Brown & Gunn, 1977; Girdano & Girdano, 1979; Seay, 1984) found, however, that alcohol consumption increases linearly with the amount of time spent in college. Demoggafiphi c Differences in Alcohol Use Cadet The literature on differences between male and female alcohol use among college students is mixed. Several studies have documented differences between men and women 19 undergraduates’ alcohol consumption (Saltz & Elandt, 1986) whereas others have found none (Banks & Smith, 1980; Hill, 1979). The majority of research findings suggest that men consistently drink more frequently, drink greater amounts of alcohol and experience more negative consequences fiom drinking than women (Barnes & Welte, 1983; Engs, 1977a; Friend & Koushki, 1984). Saltz and Elandt (1986) reviewed 19 studies and found that the range for drinking among male college students was 81-91% with an average of 91%. For females, the range was 78-98% with an average of 88%. Recent research on gender differences in the area of college student drinking has indicated a possible convergence of gender differences in recent years (Engs & Hansen, 1990). Drinking rates among women have increased although men still tend to drink more fi'equently and to consume larger amounts of alcohol (Berkowitz & Perkins, 1986; Engs & Hansen, 1990; Maney, 1990) and are more likely to become problem drinkers from high school to college (Donovan, et al., 1983). Robinson et al, (1993) examined 472 female and male college Students' knowledge, attitudes and personal and peer use behaviors regarding alcohol and other drugs. Both short-term and long-term alcohol use behaviors were measured and significant gender differences were found for both. Of the respondents, 73.0% of males and 67.1% of females reported short-term alcohol use (in the past 30 days). A significantly higher number of males (83.5%) than females (77.8%) reported long-term use (use over 6 months). Brennan, et al. (1986) reviewed studies that included a multivariate analysis to identify relationships among a variety of demographic and personality factors. Most researchers agree that male college students drink more than females and may drink more 20 often than females (Engs, 1977; Kaplan, 1979; Orford et al, 1974, Wechsler & McFadden, 1979). Gross (1993) reported a significant interaction between age and gender. Men, whether of legal drinking age or not, consumed significantly more alcohol than did women, regardless of legal status. Women under the legal drinking age had higher rates of consumption than women of legal drinking age or older, whereas the opposite pattern was found for men. The findings suggest a pattern whereby the underage men appear to start drinking at a high level of risk and this level increases as they attain legal drinking age. Although underage women begin drinking at a higher level of risk, they decrease their consumption, so their risk decreases after they reach the legal drinking age (Gross, 1993). The results confirm that underage drinking by men and women occurs at a high rate. Ethnieinr With regard to racial differences and alcohol use, white students drink more frequently and in greater amounts than non-white students (Presely et al. 1992). Nonwhite students are more likely to abstain, drink infrequently or be moderate drinkers (Barnes & Welte, 1983; Humphrey, 1983). One exception was the finding of Greenfield et al. (1980) of a modest correlation between ethnicity and alcohol consumption for average daily volume of alcohol consumed but not for frequency of intoxication. Overwhelrningly, white males drink more than any other group, averaging over 9 drinks per week. \Vrth respect to weekly drinking, Hispanic males average 5.8, white females 4.1, black males 3.6 and black females average 1 drink per week (Presley, et al., 1992). In a study of Native Americans, Hughes and Dodder (1984) found that the beer consumption 21 of whites and Native Americans was similar, but whites consumed more wine and liquor and reported fewer drinking problems (Saltz & Elandt, 1986). Class standing Class standing also appears to be associated with drinking levels as consumption of alcohol declines each year a student is in school. Students are more likely to drink and drink in greater quantities as freshmen than they are as seniors. In the senior year, seniors are likely to drink fi'equently but less per occasion (Presley et al., 1992; Saltz & Elandt, 1986). Hartford et al., (1983) found that students described their own drinking behavior as excessive during the freshman and sophomore years, decreasing their use over time as they developed a sense of maturity and responsibility. Friend and Koushki (1984) compared first and second semester drinking patterns of freshmen and found that the percentage of students who drank increased from 81.5% to 97.9% during the fieshman year. They found no differences between second-semester fi'eshmen and later classes. They concluded that 16% of entering college fi'eshmen began to drink upon coming to college but their drinking stabilized over the course of the four college years. Individual and Personglity Correlates of Collfiege Student Alcohol Use A variety of attitudinal and personality characteristics have been empirically associated with problem drinking (Berkowitz & Perkins, 1986). The diversity in individual characteristics that are associated with drinking suggests that students do not use alcohol for the same reasons (Brennan et al., 1986). AttitudeL Beliefs and Motivation Students list many reasons to drink, including tension reduction (Gonzalez, 1979), 22 boredom (Farber et al., 1980) and escapism (Beckman & Bardsley, 1981). The most fiequently cited motivation for drinking among college students, however, is the desire increase sociability (Hanson, 1984; Looney, 1976; Shore et al., 1983). A positive attitude toward drinking is associated with more frequent alcohol use (Saltz & Elandt, 1986). Individuals who hold tolerant or positive attitudes toward a substance are more likely to use it than individuals who hold negative attitudes toward it (Akers et al., 1979). In a study of the relationship between attitudinal measures and alcohol problems, Klein (1992) found a strong association between students' beliefs about drinking and the number of alcohol-related problems they experienced. Additionally, he found gender differences in alcohol-related attitudes in that females were more likely to advocate responsible drinking practices whereas the males were more apt to adhere to maladaptive and potentially damaging beliefs about drinking. In a comparison of the attitudes and beliefs regarding alcohol use between heavy, moderate and light drinkers, McCarty et al. (1983) found that heavy drinkers believed more strongly that drinking resulted in enjoyable experiences and did not produce unpleasant experiences. Persorgrlity Characteristics Personality characteristics such as lowered impulse control, greater proneness to deviant behavior, lowered expectations for academic success and greater value placed on independence are also associated with problem drinking (Jessor, 1977; Moos et al., 1979). Utilizing several cognitive and personality inventories, Schall found that the total amount of alcohol consumed during a four week period was significantly correlated with 23 several personality measures including the following: disinhibition or lack of inhibitory control, boredom susceptibility, acceptance of heavy alcohol use, use of alcohol to solve personal problems, lack of responsibility as a guest where alcohol is served, expression of concern over one’s own drinking. Extroversion, impulsivity, lack of social conformity and socialization are additional personality characteristics that were significantly correlated with total alcohol consumed. In a study designed to identify personality variables that differentiated fieshman college abstainers who became drinkers fiom abstainers who remained abstinent during the freshman year, Moos et al. (1977) found that drinking was associated with more extroverted behaviors and impulsive, rebellious personality characteristics. Social/Environmental Correlates of College Studem Alcohol Use The influences of parents and peer modeling on adolescent and young adult alcohol use have been the focus of a substantial body of empirical research (Akers et al., 1979; Clayton & Lacy, 1982; Huba, \Vrngard & Bentler, 1980; Jessor & Jessor, 1978; Kandel, 1973; Kandel, Kessler, & Margulies, 1978; Needle, et al., 1986; Perkins, 1985). Social influence on adolescent drug use is attributed to two social learning processes: active social peer pressure and the modeling of behavior (Graham, Marks & Hansen, 1991). Several cross-sectional and longitudinal studies have shown that peer drug use is predictive of an individual's drug use (Jessor & Jessor, 1978; Kandel, Kessler, & Margulies, 197 8), and there is substantial evidence that an individual's fiiendship and peer 24 group environment is an important determinant of alcohol use in late adolescence (Kandel, 1980; Perkins, 1985). Several studies have shown that 1) adolescents with substance using peers are more likely to use substances and to use more fiequently than adolescents with non-using peers; 2) adolescents from families in which one or more members use alcohol and other drugs are more likely to use substances than adolescents with family members who are non-users; and 3) parents have less influence than peers on adolescents' drug use (Kandel, Kessler, & Margulies, 1978; Needle, et al., 1986). In a study designed to examine both social and psychological predictors of alcohol consumption, Goodwin (1990) surveyed a stratified random sample of 300 college students in the Northeast. Peer pressure positively predicted alcohol consumption and satisfaction on the item, "I attend parties to be with my fiiends." The results also indicated a negative relationship between alcohol consumption and feeling that the party was an obligation. There is some evidence within the literature that best friends are more influential than other fiiends on drinking (Akers, Krohn, Lanza-Kaduce & Radosevich, 1979; Johnston, 1973; Kandel, 1973; 1978; Urberg, 1993). In a study that examined the role of parents, peers and religious traditions as determinants of alcohol and drug use among college students, Perkins (1985) surveyed the entire student body of a small, liberal arts college. Perkins found that the attitudes of fiiends and social fi’aternity membership are the most significant influences on alcohol consumption whereas parental attitudes have little influence on student behavior. 25 Modeling Modeling is another form of social influence that increases consumption. Modeling can be described as a passive form of social pressure (Graham et al., 1991) and involves the social modeling of substance use to one's peers and fiiends. According to Graham (1991, p. 292): The student may simply gain the information directly through undetected observation of behavior or indirectly through casual conversation with fiiends who use substances themselves, or through conversations with other friends regarding those who are users. Importantly, unlike active pressure, which calls for an immediate response, response to passive pressure may be delayed; experimentation may occur weeks after exposure to a model. Several studies have found a positive correlation between adolescents' ratings (perceptions) of their fiiends' use of a substance and their own current and future use of that substance (Bergen & Olsen, 1963; Castro et al., 1987; Chassin et al., 1981; Chassin, Presson, Sherman, Corty & Olshavsky, 1984; Collins, et al., 1987; Elliott et al., 1985; Huba & Bentler, 1982; Stein, Newcomb & Bentler, 1987; Sussman, 1988). Age was a significant factor and by their mid-teens or later, adolescents were more influenced by their peers than their parents. The finding that peers have more influence than parents on adolescent drinking was confirmed by Jessor and Jessor (1975) and Biddle et al., (1980). Urberg (1993) found, however, that the influence from peers may be at its highest in early adolescence and peers may become less influential in later adolescence. Caudill and Marlatt (1975) examined the effect of modeling of drinking on other drinkers. A member of the research team posed as a confederate college student and 26 either drank heavily or moderately and also behaved in either a fiiendly or aloof manner. Male college students participated in a 15 minute wine tasting study either alone or with the experimental confederate who drank heavily (700 ml of wine) or lightly (100 ml of wine). Men who participated with the heavy drinking confederate drank significantly more wine (M=364.1 ml) than either men who did not have interact with a confederateM=l80.8 ml) or men who drank with the light drinking confederate (M=141.9 ml). The light drinking confederate and no confederate conditions did not differ significantly. The attempt to influence the strength of the modeling had no effect. They found that heavy drinkers help to define a situation as an appropriate setting for increased alcohol use. Caudill and Marlatt (1975) concluded that the confederate provided cues about appropriate behavior. Subsequent studies demonstrated the strength of the modeling effect in a variety of settings and investigated potential parameters. Another study (Cooper, 197 8) suggested that both social setting and simultaneous drinking produce strong modeling effects. Studies indicate that neither the race (Watson & Sobell, 1982) nor the social status (Collins & Marlatt, 1981) of the model alters the modeling effect. Lied and Marlatt (1979) found a strong interaction between the drinkers sex and drinking habits. Men and women categorized as heavy drinkers drank similar amounts (45 drinks or more per month) and were more responsive to differential modeling than light drinkers (l 7 or less per month). The nature of the interaction between drinkers has a substantial impact on whether or not modeling alters drinking behavior. Modeling seems to be one mechanism that contributes to increased alcohol consumption among group drinkers. Strength of the influence is a firnction of the usual drinking level. Heavy drinkers exert 27 more modeling influence and are more easily influenced (McCarty, 1985). Lau et al. (1990) found support for both peer and parental modeling in a study of first year college students and their parents. Parental modeling was related to present use of alcohol while peer modeling was related to the firture use of alcohol. Bank et al. (1985) reported that peer modeling was significant whereas parent modeling was not. They concluded that peers are more influential than parents and that peer influence occurs predominately through modeling, whereas parents have more influence through their attitudes and beliefs rather than their drinking behavior. (Ary, 1993). Social Context Research on college student drinking has recently begun to include the social context of peers, family and environment as potential factors in problem alcohol use (Goodwin, 1990; Berkowitz, 1986). Extensive research with pre-adolescents has consistently demonstrated that the "social context among youth is more powerful than personality correlates in predicting the initiation and involvement in problem drinking behavior patterns (Kandel, 1980 in Berkowitz, 1986).” Social influence has a direct impact on initial use and experimentation with alcohol (Graham et al., 1991; Halevsky, 1987; Perkins, 1985) and is a central component of all models of adolescent substance use (Castro et al., 1987; Elliot, et al., 1985; Flay et al., 1983; Huba & Bentler, 1982; Huba et al., 1979; Leventhal & Cleary, 1980). The Residence Hall EnvironmenL and College Student Drinking The power of peer group influence on adolescent alcohol use is well established. Yet the social context of the residence hall environment as a primary source of peer group 28 influence on college student drinking has received limited attention even though some researchers view residence halls as a factor deserving of more research (Martin & Hofi‘man, 1992). The association between residence hall living and campus substance abuse has been documented in several studies (Martin and Hoffman, 1993; Moos, 1979; Schall et al, 1992). Barnes et al. (1992) states "...there is something unique about living in a dorm situation with same age peers that contributes to alcohol misuse." Indeed, the social context of the residence hall contributes to the degree of exposure to drinking models (Igra & Moos, 1979; Jessor & Jessor, 1977), social pressure to drink (Schall et al., 1992), nrisperception of peer drinking norms (more excessive than they are) (Perkins & Berkowitz, 1986) and the increased availability of alcohol and opportunities to socialize (Barnes & Welte, 1992). Drinking is primarily a social activity among college students (Barnes et al., 1992) and very few students report drinking alone (Harford, 1983). The social context of the residence hall provides students with increased opportunities for heavy social drinking although the drinking itself may not occur in the residence hall. Barnes et al. (1992) makes a distinction between "living in a residence hall" and "drinking in a residence hall". Residence halls may provide "more of a natural setting for parties, as reflected in higher frequencies of drinking in mixed-sex groups, than off campus dwellings." (Harford et al., 1983). Additionally, Harford et al. (1983) found that students with three or more roommates reported higher frequencies of drinking in mixed groups, night clubs and bars. An increased number of roommates is likely to represent an increase probability of 29 exposure to a peer drinking group. College students drink more than their non-college going peers and living on campus may be a factor in this pattern. In a study that compared the two groups, college attendance was not predictive of alcohol use. Living in a residence hall, however, predicted both heavy drinking and alcohol related problems after other predictor variables were controlled (age at onset of use regularly, biological father's drinking, father living in home) (Barnes et al., 1992). Moos (1979), in Evaluating Educational Environments, conducted a longitudinal study of the drinking context of 59 residence hall living groups and found alcohol use to be moderately related to dorm drinking context, students' involvement in informal activities and to students’ lack of commitment to conventional values. Students increased their average alcohol consumption from fall to spring of the fieshman year. Demographic variables of gender and religion had independent effects on alcohol use. Females were more likely to drink at the time of the second measure (spring of freshman year) suggesting that females may be more influenced by their dorm peers to increase their alcohol consumption over the course of their fieshman year than males students. The relationship between living unit and alcohol use was demonstrated in the model of college drinking developed by Martin and Hoffman (1993). The model examined the relationships among alcohol use, expectancies, living environment, peer influence and gender to drinking behavior. Hierarchical multiple regression analyses revealed that alcohol expectancy variables accounted for 36% of the variance in alcohol use, living unit accounted for an additional 8%, and peer influence, 5%. Alcohol use was significantly 30 higher among students living in fraternities, group houses or co-ed residence halls than among students who lived at home with their parents. In a longitudinal study of 17,000 young adults, Bachman and Johnston (1984) found that while senior year high school alcohol use was the most predictive of post-high school drug use, use of alcohol and other drugs was also predicted by post-high school living arrangements. Bachman and Johnston (1984) found that full time students, both men and women, reported the highest percentage increase in alcohol and other drug use. The authors (Bachman and Johnston, 1984) described this increase in alcohol consumption as "catching up" because those students who became firll-time college students after graduation fi'om high school reported relatively low drug use during high school. The authors also examined the role of student status versus employment status of young adults and found that the effects on alcohol consumption attributable to living arrangements were generally at least 3 time larger than those attributable to either students status or employment status. The residence hall offers students an "intermediate level of autonomy. . .in which students are subject to fewer controls of adult authorities that inhibit unhealthy behaviors such as heavy alcohol use " (Jones, et al., 1992). One historical study of college life suggests that living on campus provides students an opportunity to become immersed in a "predominantly hedonistic culture" which promotes alcohol use (Horowitz, 1987). In comparison to students who commute to college, residential students are "much more likely [than commuters] to become less religious and more hedonistic, hedonistic referring to drinking, smoking, sexual activity..." (Astin (1985). Students living on campus drink significantly more alcohol than commuter students (Cooney & Nonnarnaker, 1992). 31 Problem Behavior Theory aad the Perceived Environment Problem-Behavior Theory will be used as the theoretical framework for this study. Figure 2.1 presents the conceptual structure of the fi'amework. The theory is designed to be a comprehensive social psychological framework and the fi'arnework is described by Jessor as “neither a grand nor overarching theory but as a network of concepts of modest scope oriented toward a delimited concem—problem behavior in you ” (Jessor, 197 7, p. 21). This psychosocial model maintains that human behavior is the result of person- environment interaction. The theory consists of three interdependent systems of variables: The personality system, the perceived environment system and the behavior system. The systems are considered to be causal or explanatory nature in that they are “presumed to be most directly influential to the occurrence or non-occurrence of any particular behavior” (J essor, 1990, p. 18). In problem behavior theory, the variables from each system represent either instigations or controls that, in combination, generate “proneness,” or the probability of resultant problem behavior. Although proneness can exist in one, two or all three of the systems, overall psychosocial proneness is the central concept of the theory and is used to predict and explain variation in problem behavior. Jessor also employs the concepts of “proximal” and “distal” variables. Proximal variables are those that are directly related to problem behavior while distal variables are related to behavior through theory. According to Jessor (1977), proximal variables may exhibit a stronger relationship with problem behavior but distal variables, as they are less obvious in their connection to problem behavior, may be more theoretically interesting. The personality system includes such variables that are cognitive in nature such as 32 Se. .83.... can... Bream 8.32.. .o 822.5 3388 2.... .N 2a.... E .55.... .28... @358 .859: 85.13.... 5.3.0 on; .3338 1:363:00 5v... 3.5.3. 63...... 9......3. .35.... .53.... 1.2.8 9.33.... 8.2.90 8: as... as... .25 05 5.5.9.3! 9.35.5 88......— eg beryl—on 5059..— ...sflm .2223. \ 83......» 8.22.8. 168 6 $83.2. 8. .132 8...... 8.5.... 53...... .9. 1.3... 3.3.... 50.5.0: 1.039:— 3 1.5.5.: .3601.- 2=§Sm 1.5.6... 898 8.46.: 338.». 852.... 6...... .9 2.8... 1......8 6...... 3.5.30 3:8... 93.3.5 .5»... E298 aeggcm 33080.. E . J. £83.... 8.32:. 1.3! 8:3»... .o 8.3.38... 153...... gnu-Sm 35:00 «3.3.85 .838... .83 1883-182... Susana... 53.5... 95.3.0 38... 8.5.5... 3...... 3.30.. 8.3.5.8... .5. 5388.... «6980.523. .8 3.5909...— §5 60 81> .§£u< co 31> 835m :03 b1 . . . :5... 893.8 3.3.3.0.. .A— 5...»... na... 95302928.. 3.33.... to? £93.92 canes... co...8..=.8 2.8.... 2.350 38... air—£502 2.03:3.— E8:— 38... .33....890 2.05552 Econ-6:3 832...... 3.2.36.8... Scream acoggcm 390m 83.5.» EURO—0:0.»mménmucm 833...; 2.8250 was Eczemwxunm 138m 33 values, expectations, beliefs, and attitudes. The personality systems combine to reflect “social meaning and social experience rather than deep-lying drives” (Jessor, 1990, p. 21). The 11 variables in the Personality System are organized into three structures which are considered as either instigations for problem behavior or controls against it. Personality system proneness to engage in problem behavior is characterized by low values on achievement, self-esteem, attitudinal intolerance of deviance, moral attitude and religiosity and high values on independence, social criticism, and external control. The perceived environment system, which is of primary interest in this study, includes “perceived controls and instigations fi'om significant others in the individual life, particularly parents and friends” (Jessor, 1977, p. 29). Jessor (1977) maintains that the perceived environment is the most meaningful to the individual and behavior is most closely connected to the individual’s perceptions of their environment. The variables within the perceived environment system reflect the supports, influence, models of significant others in the environment. Proneness to problem behavior in the perceived environment system is characterized by lower amounts of controls against problem behavior and greater approval, pressure and models for problem behavior. The behavior system contains variables which meet Jessor’s (1977) definition of problem behavior which is behavior that is “socially defined as a problem, as a source of concern, or as undesirable by the social and or legal norms of conventional society” and ‘irsually elicits some form of social control response” (Jessor, 1990, p. 23). The variables within this system includes both conventional behavior as well as problem behavior. Conventional behavior is non-problem behavior such as church attendance. As in the 34 other systems, proneness to problem behavior is characterized by the balance between instigations and controls. Proneness within the behavior system is characterized by such behaviors as problem drinking or drug use and low involvement in conventional behaviors such as church attendance. As noted earlier, Figure 2.] presents a schematic drawing of the interrelated systems posited by Problem Behavior. Jessor (1990) emphasizes that while eachof the systems is connected to all of the other systems through bi-directional arrows, there is nevertheless an “explanatory directionality” that is implied in the framework. The primary goal of the framework is to explain problem behavior and such an explanation relies upon the combination of all of the instigations and controls in the various systems operating theoretically in a prescribed direction, that is, to predict and explain problem behavior. Perceived Environment VMes with_in the Res_idence 11111 J essor's Problem-Behavior Theory, and in particular the Perceived Environment System, provides a theoretical framework to examine the perceived environment of the residence hall in relation to alcohol use. For the purpose of this research, Jessor's conceptualization of the Perceived Environment variables Controls, Approval of Drinldng, Models for Drinking and Pressure for Drinking were used. The peer reference groups and agent of social controls, however, were adapted to reflect the social context of the residence hall environment. J essor used parents and fiiends as representatives of the major reference groups in the respondent's social environment. \Vrthin the residence hall environment, the primary 35 reference groups were defined as the student's roommate and fiiends on floor. Hays and Oxley (1986) examined the social networks of college freshman living in a residence hall and found that dorm students had networks with greater densities than students who did not live in a residence hall. Moos (1979) suggests that a measure of friend use or approval within the dormitory context is an important "mediating variable in a chain from dormitory context to alcohol use" (p. 123). In this research, the construct of roommate, was conceptualized as a source of influence but the relationship was not characterized as fiiendship in that a student's roommate could or could not be considered to be the student's fiiend. The role of the Resident Assistant was conceptualized as the agent of social control within the residence hall environment who responds to problem drinking which is " socially defined as problematic, usually eliciting social control responses” (J essor & Jessor, 1977, p. 31)." The role of residence hall staff as agents of social control is described in the following section. fiogial Con_trol within the Residence Haj Environment Despite the rejection of the doctrine of in loco parentis by the courts in the 1960's, changes in federal laws impose upon institutions an obligation to regulate student drinking behavior. The passage of The National Minimum Drinking Age of 1984 made consumption of alcoholic beverages illegal for almost 75% of traditional-age college students and required institutions to formulate policies against underage drinking on campus (Califano, et al., 1994). Additionally, the 1989 Amendment to the Drug Free School and Communities Act, requires colleges and universities to establish and publicly 36 disseminate campus policies regarding substance abuse. This amendment requires institutions to prohibit the unlawful possession, use or distribution of drugs and alcohol and to impose sanctions on those who violate the regulations (Gehring, 1992). The overall effect of such legislation is to place universities in the role of "caretaker and regulator with increased liability" if campus rules are not firlly enforced (Califano, et al, 1994) Housing administrators are heavily involved in this process since many sanctions are imposed on residential students (Gehring, 1992). Concern over alcohol-related incidents, has led many institutions to enact strict alcohol policies. Several college and universities have banned beer kegs, alcoholic punches, and the delivery of alcohol to campus (Thomas, 1991). Some have banned alcohol completely from dormitories. The primary agent of enforcement in campus residence halls is an undergraduate staff member or Resident Assistant. In conjunction with other university housing staff, the resident assistant participates in the ”definition, interpretation, observance and enforcement of the college drinking rules" (Rubington, 1991, p. 375). In a study of drinking sanctions in fi'eshman residence halls, Rubington (1991) explored the impact of policy enforcement on the number of student alcohol violations. Utilizing the fi'amework of ”sanctions theory" (e. g. the greater the risk of apprehension, the lower the chances of deviant behavior), he found that alcohol violations were significantly reduced after fall term. He cautions, however, that strict enforcement of university alcohol policies may not result in students’ drinking less but in their drinking in locations other than the residence hall. As many RAs noted in the interviews, once 37 freshmen learn they cannot drink in the dorms, they either give up the idea of drinking, or much more likely, find places outside the residence hall where they can drink. Nonetheless, Rubington concludes that the marked drop in alcohol violations as reported by the RAs attests to the fairly eflicient working of the system of formal control (Rubington, 1991). Summm In this chapter, an efl‘ort has been made to examine the nature of college student drinking within the collegiate environment. A review of the research documents and describes the drinking patterns of students including the phenomenon of binge drinking, negative consequences associated with drinking, pre-college drinking patterns and demographic differences in use. Alcohol consumption is also related to a variety of individual and environmental factors including attitudes, beliefs, and motivations; personality characteristics; peer group influence; social context. In reviewing the literature, however, it is evident that college student alcohol use is a complex issue which requires the consideration of multiple factors. The role of the residential living environment is a promising area of research. A better understanding of the role of the peer group within the residence hall environment and its association with problem drinking is potentially usefirl for both the institution and students. Many researchers on the college environment and alcohol advocate a theoretical approach to firrther research. The inclusion of a theoretical framework, Problem-Behavior Theory, serves as a comprehensive approach to further our understanding of the nature of interaction between the college environment and the student. Chapter 3 METHODOLOGY This study examined the relationship between perceived residence hall environment variables and problem drinking. Selected measures from The Young Adult Survey Questionnaire (Jessor & Jessor, 1981) were used to measure the variables. It was hypothesized that the subscales of the Young Adult Survey would provide data appropriate for statistical testing of the hypotheses. For the purpose of this study, measures from two of the three Problem Behavior Theory systems were be employed: (1) Perceived Environment System and (2) Behavior System (Problem Drinking Measure). The primary focus of this study was the relationship between the Perceived Environment System and the Problem Drinking Measure. This chapter describes the methodology and procedures used in the study. The subject pool is described, along with the data collection procedure and study design. The research hypotheses are stated and the analysis of the data described. The Population and Sample This study was limited to a sample of the population which consists of first time freshmen enrolled at Michigan State University for the 1995 fall semester and who live in a campus residence hall. To estimate the size of the population, data from the Michigan 38 39 State University Registrar and Student Life Department data were reviewed to determine the number of male and female freshman students living on campus during fall term or semester for the past five years. It was determined that 97% of freshman live on campus each fall semester. The MSU Registrar's office provided data for the freshman class on the 18th day of the fall term or semester for the past five years. The number of students in the fieshrnan class was estimated to be approximately 6,000 with 55% females and 44% males. According to the drinking definition to be utilized in this study (Jessor, 1977), it was estimated that the percentages of students in each drinking classification for both men and women would be approximately 20% Problem Drinkers, 70% Non-Problem Drinkers, and 10% Non-drinkers. To conduct meaningful statistical comparisons, Balian (1988) recommends at least 30 cases for each group classification studied. To obtain this number of cases per classification the researcher determined that a sample size of at least of 600 students; 300 men and 300 women was necessary. This sample return would provide the minimum of 30 cases for both males and females per drinking classification needed for meaningful statistical comparisons (Balian, 1988). Balian (1988) recommends increasing the sample size 70-300% when using mailed surveys to compensate for attrition, respondent refusal to participate or other circumstances depending upon the follow up measures to be employed (Balian, 1988). The researcher predicted a low return rate due to l) the length of the survey at eight pages, and 2) the questions about alcohol use may be considered by some respondents as intrusive. Additionally, some responses may feel reluctant to answer questions that may indicate that they have violated university 4O regulations or state and federal laws. The researcher estimated a 30 percent return rate and set the sample size at 1800 students or 30% of the freshman class as the necessary sample size to reach a return rate of 600 respondents. The fieshman subjects in this study may or may not be randomly assigned to a particular housing unit. Freshmen are generally assigned to a particular residence hall only if they indicated such a preference at the time of application for housing. If no preference was indicated, each student was then randomly assigned to a residence hall where vacancies exist. Each student was assigned to a two or four person suite on a floor section containing 47 other students and an undergraduate staff member. The number of freshmen or upperclassmen living on each floor section varied with some floors being mostly freshmen and others being mostly upperclassmen. PM Freshman students were encouraged by the university to arrive on campus on August 23, 1995. A letter informing respondents of the study and inviting their participation was mailed to the random sample of freshman students living on campus at Michigan State University on September 18, 1995. On September 25, 1995, a letter introducing the study was sent to each respondent. On October 2, 1995, a cover letter plus a copy of the Young Adult Survey was mailed to each respondent. The return date for the questionnaire was October 9, 1995. Each respondent received a stamped post card to be returned to the researcher to indicate the return of their survey. Return envelopes were designed to be returned to the researcher via the Michigan State University campus mail 41 system. Each respondent was offered a copy of the results of the study. To increase the return rate, incentives were offered by the researcher. A drawing was held from the postcards returned by October 9, 1995 for a $100.00 gift certificate to the MSU bookstore. A follow-up transmittal, a copy of the questionnaire with a return date of October 23, 1995, a pre-addressed return envelope and a results-requested envelope was sent to non-respondents via campus mail on October 16, 1995. A second drawing for a $50.00 gift certificate to the MSU bookstore was held for those who returned their questionnaire by October 23, 1995. finingof Dati Collection The data collection for this study occurred at the end of the first six weeks of the academic school year. The timing of the data collection served as a control measure by surveying students in the living environments to which they had been assigned by the university. The researcher wanted to avoid the phenomenon of ”self-selection” (Kandel, 1980) whereby individuals select a peer group based upon their drinking preferences. Additionally, the researcher was interested in examining students’ perceptions of their original living environments created by random assignment by the university. Data regarding students’ perceptions of university-created living environments assist housing administrators in understanding better the nature of the living environments that are created by the random assignment process. The researcher estimated that six weeks was adequate time for the new student to experience the most proximal aspects of the college environment including their 42 roommate, other residents on the floor, and the Resident Assistant. I_nfistrumentation The Young Adult Survey (J essor, 1981) was selected as the instrument for this study to measure the environmental variables and their relationship to problem drinking among college students. The Young Adult Survey, in its original form, is a 450 item, paper and pencil instrument, which includes several subscales. Psychometric investigations have supported the utilization of the instrument as a valid and reliable research tool. It contains a subset of the variables fiom the larger framework of problem- behavior theory as well as items designed to specifically measure variables within the residence hall environment. The Young Adult Survey is comprised of items which pertain to three systems: Personality System, Perceived Environment System and the Behavior System. For the purposes of this research, selected items fi'om the Perceived Environment System and Behavior System were used. The survey items were modified for use in this study. The Perceived Environment questions were modified to focus on peer relationships in the residence hall. For example, a question that was originally worded “Do you have any fiiends who do not drink at all?” was changed to “Do you have any friends on the floor who do not drink at all.” The questions were repeated for roommates and Resident Assistant when appropriate. According to Light, Singer and erlet (1990), validity in an instrument describes how well a measure actually assesses what it is intended to assess. The items within the Young Adult Questionnaire were modified only in terms of specificity of relationship and not in any other way. The Behavior system items were modified in only that respondents were asked to answer the question in terms of the time 43 they had spent at MSU. Otherwise the content of the items were identical to the J essor’s Young Adult Questionnaire. The majority of these variables were measured by single and multiple item scales derived from problem behavior theory and abbreviated from an early version developed to test the theory in a longitudinal study of adolescent psychosocial development (Jessor, 1977; 1991). Jessor (1977) describes the Young Adult Survey as a ”theoretical instrument, comprised of multi-item scales, formulated from a construct validity perspective about test construction and psychometrically developed." Each of the summative, multiple-item measures of the Perceived Environment System were evaluated for its psychometric adequacy by calculating two difference Table 3.1 Scale Properties for the Perceived Environment System Measures Men Womar (5:84) =100 Number Scale 1981 Measures of items range Alpha HR Alpha Hr Perceived Environment System Distal Structure Parental Controls 1 1-4 - - - - Friends Controls 2 2-8 .73 .59 .68 .53 Proximal Structure F riends' Approval of 3 -9 to +9 .26 .ll .43 .20 Problem Behavior Friends Models for 3 -9 to +9 .52 .27 .53 .28 Problem Behavior indexes: Cronbach's (1951) alpha, an estimate of reliability based on the interim consistency of responses; and Scott's (1960, 1968) Homogeneity Ratio (HR), a measure of the weighted-average inter-item correlation of each scale. These scale properties are 44 presented in Table 3.1 for each of the Perceived Environment System measures and for the 1981 College Study subsample. Jessor (1977) describes the homogeneities of the scales as "quite adequate but, as expected for such short scales, their alpha reliabilities are low. " Table 3 .2 presents the number of items, scoring, and Cronbach’s alpha of the perceived environment measures that were used in this study. Several items are combined into subscales or can be examined as single item measures. The first measure, Residence Hall Modeling is a four-item measure assessing the amount of drinking models a student perceives in the overall residence hall environment. The Approval for Drinking measure is a three-item subscale that includes Friends’ on Floor, Roommate, and Resident Assistant Approval for drinking. The Modeling variables assess the frequency with which the respondents had been encouraged or influenced by their fiiends or roommate to engage in drinking. Higher scores reflect higher than average degrees of modeling for drinking. Table 3 .2 Perceived Residence Hall Environment Measures Residence Hall Perceived # of Scale Cronbach Environment Measures Items Range alpha Distal Structure Residence hall Models for Drinking 4 16-0 .81 Resident Assistant Control 2 8-1 .71 Proximal Structure Approval for Drinking 3 9-3 Models for Drinking 4 13-3 .78 Friends 2 9-2 .71 Roommate 2 4-1 .68 Pressure for Drinking 2 8-2 .63 Controls Against Problem Behavior 4 16-1 .78 Friend 2 8-1 .65 Roommate 2 8-1 .77 45 Pressure for drinking by either the respondent’s roommate or fiiends on floor includes a measure asking about the frequency with which their roommate or fiiends on the floor put pressure on them to drink. The Controls variables consist of multiple items asking respondents if their fiiends on the floor, roommate or Resident Assistant would try to influence their behavior if the respondent "were going to do something illegal or that most people would think of as wrong.” Responses options range from “definitely would not” to ‘definitely would.” Higher scores indicate a greater perceived amount of disapproval by the roommate, Resident Assistant or Friends on the floor for socially disapproved behavior. Problem drinking proneness in the perceived residence hall environment involves greater roommate and fiiends on floor approval of drinking, greater pressure and models for drinking and less perceived controls against problem behavior within the environment. With regard to the Behavior System, Jessor (1977; 1991) utilizes a concept of "problem" drinking that focuses on alcohol consumption to the point of drunkenness and on the negative social and interpersonal consequences associated with drinking. Respondents were classified as problem drinkers if they reported being drunk four or more in the first six weeks of school or if they reported more than four (out of ten) negative consequences due to their drinking. The possible negative consequences included being criticized by fiiends for drinking, missing class, being in trouble with the police, getting into a fight, damaging public property, engaging in loud or disorderly conduct, experiencing roommate conflicts over drinking, having an accident on campus or off campus, driving after drinking, and having resident assistant roommate, fiiends expressing 46 concern over the respondent’s drinking since arriving at school. Demographic variables were placed at the end of the questionnaire as . recommended by Babbie (1991). The demographic variables included gender, size and type of residence hall, specialty housing option, and residential academic program. Anonm'ty The respondents were assured of complete anonymity. Each respondent was instructed to return their completed survey separately from their return postcard. The return postcards were only used to remove respondents’ names from the mailing list so they were not sent a second or third survey and to select the recipients of the gifi certificates. No coding system for identification was used on any of the questionnaires. Self-reports of alcohol use The accuracy of self-reports of alcohol use is well documented. Self reports have been found to be satisfactory with respect to reliability and validity as long as the researcher assures the respondents that their responses will be confidential (Dielrnan, 1991) Statistical Analysis The majority of questions on the survey are considered continuous, interval data. Various statistical analyses were used to interpret the data including one-way AN OVA, one-way MANOVA, t-tests, post hoc multiple comparisons tests and discriminant analysis. The .05 alpha level of significance was used. The hypotheses were stated in Chapter 1. They are restated here as operational, directional hypotheses: 47 Hypotheses 1. There will be no significant differences in the self-reported drinking behaviors of Michigan State University freshmen when categorized by gender, size and type of residence hall, specialty housing option or residential academic program. There will be no significant differences in the scores on perceived environment variables (Approval for drinking, Models for drinking, Pressure for Drinking and Controls) among Michigan State University freshmen when categorized by gender, size and type of residence hall, specialty housing option or residential academic program. . Problem drinkers will perceive more Roommate, Friends’ and Resident Assistant Approval for Drinking than Non-drinkers or Non-problem drinkers. Problem drinkers will perceive more Roommate and Friends’ Modeling for drinking than Non—drinkers and Non-problem drinkers. . Problem drinkers will perceive more Roommate and Friends’ Pressure for Drinking than Non-drinkers and Non-problem drinkers. Problem Drinkers will perceive less Roommate, Friends’ and Resident Assistant Controls than Non-drinkers and Non-Problem Drinkers. The perceived environment variables of Approval for Drinking, Modeling of Drinking, and Pressure for Drinking and Controls will combine to predict drinking group membership. 48 Definition of Terms In this dissertation research, several terms will be used that require definition: Non-drinker. A person who reports never having consumed alcoholic beverages. Alcohol Free Floor. A housing option chosen by students in which each floor resident signs an agreement stating that he or she will not possess or consume alcoholic beverages within the confines of the floor. Alcohol Free Room. A housing option chosen by students in which each resident of a residence hall room signs an agreement stating that he or she will not possess or consume alcoholic beverages within the confines of the room. Approval for Drinking. The perception that engaging in drinking can be a means of gaining approval from and establishing connectedness with either a roommate, fiiends on the floor or the Resident Assistant. Controls. The perception that others hold relatively strict standards for behavior and would exercise sanctions against behavior of which they disapprove. Freshman student. A male or female firll-time undergraduate student who is attending college for the first time and is living in a campus residence hall at Michigan State University. Models for Drinlgrg, The perception that roommate or fiiends on floor engage in drinking and potentially provide an opportunity to learn how to engage in drinking; access to what may be needed to engage in drinking; evidence that drinking can be accomplished and is not completely unthinkable and the social controls against drinking are not very efi‘ective or are not implemented. 49 Mive Consequences Scale. A scale created by summing the frequency of negative consequences experienced in ten different life areas including being criticized by fiiends for drinking, missing class, experiencing roommate difficulties, having been told alcohol was creating problems at school, experiencing trouble with the police or being arrested, having an accident, driving a car while under the influence of alcohol, or having fiiends, roommate, or Resident Assistant express concern regarding drinking. Non-Problem Drinker. Person who consumes alcohol but has been drunk less than four times since coming to MSU and scoring less than 5 on the Negative Consequences Scale. Pressure for Drinking. The perception that roommate or fiiends on the floor exert pressure to encourage the consumption of alcohol when the consumption of alcohol is undesired. Problem Drinker. Person who has been drunk four or more times since coming to MSU and scored 5 or more on the Negative Consequences Scale. Resident Assistant. A male or female upper class undergraduate student who is employed by Michigan State University on a half-time basis as a live in stafi’ member on the residence hall floor. Residence hall. A building of eight to twenty-four floors totaling 500-1200 students on the campus of Michigan State University. Male and female residents may live on the same floor in the residence hall or separately on adjoining floors in the residence hall. R_es‘»idence hall floor. A group of approximately 50 students living on a wing of a residence hall. 50 Res_idence Hall Modeling. The perception that other students living in the residence hall engage in drinking. Other students are defined as students living in the same hall, students living on the same floor, students with whom the respondent spends some time and close fiiends of the respondent. Size of Res_idence lia_ll_. Residence halls categorized as either small, medium or large; housing approximately 400, 800 and 1100 students respectively. Type of Res_idence Hall. Residence halls categorized as Freshman Intensive, Core, or Upper Class. Freshmen Intensive halls house almost 75% freshmen, a Core hall houses a mix of all student classes but predominately freshmen and sophomores, and Upper Class halls house mostly upper class students with a small percentage of freshmen. Chapter 4 FINDINGS Introduction The purpose of this study was to examine the relationships between perceived environmental variables within the residence hall environment and college student drinking behavior. Three primary research questions with corresponding research and statistical hypotheses were identified. The data analysis and interpretation for the research questions are presented in this chapter. The chapter is divided into three sections: 1) description of the sample and establishment of drinking classification, 2) research questions with statistical hypotheses and 3) a summary of the findings. Deschtion of theSample Research instruments were mailed to 1800 new freshman students at Michigan State University who were residing in campus residence halls and who had not requested their name and address be restricted. A total of 1222 total surveys were returned; representing a 68% response rate. Of those 1222 surveys, 1197 were complete and used in the statistical analysis. Seventy-nine percent of the respondents were 18 years of age and only .6 % were 21 years of age or over; 87 % reported that they consume alcohol. Approximately 13% of 51 52 the respondents were classified as Non-drinkers, 57% as Non-problem drinkers, and 30% at Problem drinkers. Thus, this sample had 10% more Problem drinkers than originally estimated. Almost 33% of the respondents indicated that their drinking had increased since coming to MSU, 11% said it had decreased and 56% said it had stayed the same. Table 4.1 presents the frequency data for the demographic variables of gender, size of residence halls, type of residence hall, specialty housing option and residential academic program. The male/female ratio of respondents, 42% males and 58% females closely parallels the male/female ratio of the freshman class at Michigan State University of 44% males and 56% females. The majority of respondents or 61% reside primarily in large halls with 17% residing in medium size halls and 19% residing in small halls. Sixty percent of the respondents were living in CORE halls while 26.1% were living in Freshman Intensive Halls and 11.3% percent were living in Upper class halls. Establishment of Drinking Classification Donovan and Jessor (1978) utilize a combination of frequency of drunkenness and negative consequences experienced to classify respondents into the categories of Non- drinker, Non-problem drinker and Problem drinker. Donovan and Jessor (1978) define problem drinking in a variety of ways using either fiequency of drunkenness or negative consequences experienced over a specified length of time. The authors state that the use of different definitions of problem drinking “. . .provide a sense of the robustness of the explanatory fiamework and of its generality across criteria of different stringency” (pp. 1512). One definition of problem drinking used by Donovan and Jessor (1978) used was 53 drunkenness two or more times in one month or two or more than two negative consequences in any of at least three areas (equal to a score of six or higher on the Negative Consequences Scale). For the purposes of this study, Respondents were divided into the categories of Non-Drinker, Non-problem drinker and Problem drinker based upon the same two criteria, frequency of drunkenness (four or more times) during the first six weeks of school at MSU and a score of five or more on the Negative Consequences Scale. The fi'equency of drunkenness was set at four since the respondent’s were asked about their behavior over six weeks to two months depending upon when they returned their questionnaire. To be classified as a Problem drinker in this study, a respondent had to report five or more negative experiences on the Negative Consequences Scale during the time period of six weeks. The Negative Consequences Scale was created by summing each respondent’s scores for the frequency of which he or she experienced negative consequences in ten different life areas. These life areas include having being criticized by fiiends for drinking, missing class, experiencing roommate difficulties, being told alcohol was creating problems at school, getting into trouble with the police or being arrested, having an accident, driving a car under the influence, or having fiiends, roommate, or Resident Assistant express concern regarding drinking. Responses to each life area item ranged from 0 to 4 with values of O for never, 1 for once, 2 for 2 or 3 times, 3 for 4 or 5 times and 4 for 6 or more times and were added across all ten items, yielding a score range of 0 to 40. Thus, any single area, even if it is an area of chronic problems, can contribute only 4 points to the total. A cutoff point of 5 or more is used to define the problem drinking 54 group, and such scores automatically indicated alcohol-related negative consequences in more than a single area. Non-Drinkers were defined as respondents who have never consumed alcohol or had not consumed alcohol while enrolled as a student at Michigan State University. The Table 4.1 - Frequency Distribution of Demographic Variables Demoggaphic Variable f % Gender Male 499 41.6 Female 693 57.8 Missing 5 .4 Total 1197 100.0 Size of Residence Hall Small 226 18.9 Medium 201 16.7 Large 736 61.4 Missing 34 2.8 Total 1 197 100.0 Type of Residence Hall Freshman Intensive 312 26.1 Core 716 59.8 Upper Class 133 l 1.1 Missing 36 3 .0 Total 1197 100.0 Specialty Housing Option Alcohol Free Floor 102 8.5 Alcohol Free Room 250 21.0 Non-Option 833 69.5 Missing 12 1.0 Total 1197 100.0 Residential Academic Option None 1013 84.6 James Madison College 49 4.1 Lyman Briggs School 77 6.4 RISE 1 l .9 ROSES 29 2.4 STAR 18 1.5 Total 1197 100.0 55 Non-problem Drinker category includes respondents who reported having been drunk less than four times while enrolled at MSU and scoring less than five on the Negative Consequences Scale. Table 4.2 shows the frequency distribution for the drinking classification. 4.2 Frequency Distribution of Drinking Classifications Drinking Status f % Non-Drinker l 54 12. 8 Non-problem drinker 676 56.4 Problem Drinker 367 30.6 Total 1197 100.0 Drinking Classification and DemogLaphic Va_ri_ables Descriptive and inferential statistical data will be presented for each combination of drinking classification and demographic variable. Chi-square measures of association and eta were used to test the research hypothesis at the .05 level to determine if significant differences existed in self-reported alcohol consumption of Michigan State University freshman in relationship to the selected demographic variables of gender, type and size of residence hall, specialty housing option and residential academic program. Eta squared can be interpreted as the proportion of variance in the dependent variable explained by the difference between groups. For clarification, research question 1 is restated here: Research Question 1: Are there significant differences in the self reported drinking behaviors of Michigan State University freshmen when categorized by gender, type and size of residence hall, specialty housing option, and residential academic program? 56 For the purpose of analysis, each variable within research question one was tested separately and statistical hypotheses were specified for the variables gender, type and size of residence hall, specialty housing option and residential academic program. genre- The following statistical hypothesis was identified to answer research question 1 and is presented in the null form: _S_tAatistical Hypothesis 1: There are no significant differences in the self-reported drinking behavior of Michigan State University freshmen when categorized by gender. Table 4.3 presents the fi'equency distribution for drinking classification by gender. Similar percentages of males and females, 86% of males and 88% of females, report drinking while at MSU. The percentage of non-drinking females is almost equal Table 4.3 Frequency Distribution of Drinking Classification by Gender Drinking Classification Male Female n=499 n=693 f % f % Non-drinker 68 14 86 12 Non-problem drinker 254 51 419 61 Problem drinker 175 35 187 27 Total ' 497 100.0 692 100.0 got; Eight cases missing. x1 (2, n=1189) = 11.27, p< .01. Eta‘ = .002 to the percentage of non-drinking males. Proportionally, there are more female non- problem drinkers (61%) than males (51%) but more male problem drinkers (35 %) than female problem drinkers (27%). A chi-square measure of association indicates that 57 drinking classification is independent of gender (X2 = 11.27, p < .01). Therefore, statistical hypothesis was rejected. It must be noted, however, that eta == .053 indicating that the less than 1 % of the variance between the drinking classifications two groups is explained by the differences between males and females. Size of Res_idence ILall The following statistical hypothesis was identified to answer Research Question 1 and is presented in the null form: Statistical Hypothesis 2: There are no significant differences in the self-reported drinking behaviors of Michigan State University fi'eshmen when categorized by size of residence hall. The first of two residence hall variables that were examined was size of residence hall. MSU residence halls can be divided into three categories by size, small, medium and large with each housing approximately 400, 800, and 1100 students, respectively. Table 4.4 Frequency Distribution of Drinking Classification by Size of Residence Hall Drinking Small Medium Large Classification n=226 n=202 n=739 j % f % f % Non-Drinker 34 15 30 15 89 12 Non-Problem Drinker 13 5 60 1 14 56 412 56 Problem Drinker 57 25 58 29 238 32 Total 226 100.0 202 100.0 739 100.0 Note. Thirty cases missing. X2 (4, 1167 ) =5.22, p > .05. Etaz = .004 According to this classification, there are 5 small halls, 7 medium halls and 10 large halls on campus. Of the respondents, 226 or 19% live in small halls, 202 or 17% live in 58 medium halls and 739 or 62 % live in large halls with 31 missing cases. Table 4.4 presents the frequency distribution of drinking classification by size of residence hall. A chi-square test of association revealed no statistically significant measure of association between drinking classification and size of residence hall. Therefore, Hypothesis 2 was not rejected. Type of Residence Hall S_t_atistical Hypothesis 3: There are no significant differences in the self-reported drinking behavior of Michigan State University freshmen when categorized by type of residence hall. The second residence hall variable examined was type of residence hall. At MSU, each residence hall houses different percentages of freshmen, sophomores, junior and senior students. To describe the different student population of each hall, the residence hall is classified as either Freshman Intensive, Core or Upper Class. A Freshman Intensive building houses at least 75% freshman students, a Core building houses all classes of students but predominately freshmen and sophomores, and Upper Class halls house mostly upper class students with a small percentage of freshmen students. The frequency distribution of drinking classification by type of residence hall is presented'in Table 4.5. Similar percentages of respondents within each drinking classification were represented in each type of residence hall. A chi-square measure of association revealed no significant measure of association between drinking classification and type of residence hall. Therefore, Hypothesis 3 was not rejected. 59 Table 4.5 Frequency Distribution of Drinking Classification by Type of Residence Hall Drinking Classification Freshman Core Upper Class Intensive n=719 n=136 n=312 f % f % f % Non-Drinker 46 14 86 12 21 1 5 Non-problem Drinker 186 60 396 55 79 58 Problem Drinker 80 26 237 33 36 27 Total 312 100.0 719 100.0 136 100.0 Note. Thirty cases missing. X1 (4, 1167) =7.3, p > .05. Etal = .005 Specialty Housing Option - Alcohol Free Room Statistical Hypothesis 4: There are no significant differences in the self-reported drinking behavior of Michigan State University freshmen when categorized by alcohol fiee room residency. The relationship between alcohol consumption and two specialty housing options, Alcohol Free Rooms and Alcohol Free Floors, were examined. Table 4.6 presents the fi'equency distribution of drinking classification by Alcohol Free Room designation. As Table 4.6 Frequency Distribution of Drinking Classification by Alcohol Free Room Residency Drinking Classification Alcohol Free Room Non Alcohol Free n=252 Rooms n=833 f % f % Non-Drinker 8 l 3 2 54 7 Non-problem drinker 147 58 468 56 Problem Drinker 24 10 3 1 1 3 7 Total 252 100.0 833 100.0 Note. Ten cases missing. Alcohol Free Floor respondents not incuded (n=102). X’(2, 1085)=151.0,p<.001.Eta2 = .127 shown in the table, 32% of the Alcohol Free Room respondents are Non-drinkers versus 60 only 7% of those respondents living in Non-alcohol Free Rooms. Only 10% of Alcohol Free Room respondents are Problem drinkers compared with 37% of the Non-alcohol Free Room respondents. As shown in Table 4.6, there is a significant association between drinking classification and the alcohol fiee room option. Hypothesis 4 was rejected. Specialty Hou,si_ng Option - Alcohol Free Floor The relationship between alcohol free floor residence and drinking classification was examined and the results are presented in Table 4.7. The statistical hypothesis is restated here: Statistical Hypothesis 5: There are no significant differences in the self-reported drinking behavior of Michigan State University freshmen when categorized by alcohol free floon Table 4.7 Percentage of Non-Drinkers and Drinkers Residing on Alcohol Free Floor Drinking Classification Alcohol Free Floor Non-alcohol Free n=102 Floor Option n=833 f % f % Non-Drinker 17 17 54 6 Non-problem drinker 55 54 468 56 Problem Drinker 3O 29 31 1 37 Total 102 100.0 833 100.0 Note. Ten cases missing. Alcohol Free Room respondents not included (n=252). X2 (2, 935) =14.0 p <.001.1«:m2 = .008 As shown in Table 4.7, seventeen percent of respondents living on Alcohol Free Floors are Non-drinkers compared to 6% living on Non-alcohol fiee floors. Twenty-nine percent of Alcohol Free Floor respondents are Problem drinkers as compared to 37% of 61 respondents living on Non-alcohol Free Floors. A chi-square test revealed that the association between drinking classification and residency on an Alcohol Free Floor is significant at the .001 level. Therefore, Hypothesis 5 was rejected. Residential Academic Pram Statistical Hypothesis 6: There are no significant differences in the self-reported drinking behavior of Michigan State University freshmen when categorized by residential academic program. There are five residential academic programs at Michigan State University. The fiequency distribution of respondents per program is presented in Table 4.8. For data analysis, the number of respondents per program were aggregated into one total sum for participants in Residential Academic Programs. Table 4.8 presents the frequency distribution of drinking classification by residential academic program. There was a higher percentage of Non-drinkers and a lower percentage of Problem drinkers enrolled in residential academic programs than in those enrolled in non-residential academic program. A chi-square test was significant at the p < .05 level of significance. Eta squared indicates, however, that less than 1 % of the variance in drinking classification is explained by the difl‘erences between enrollment in a residential academic program and not being enrolled in such a program. Therefore, Hypothesis 6 was rejected. 62 Table 4.8 Frequency Distribution of Drinking Classification by Residential Academic Program Drinking Classification Residential Academic Non Residential Program Academic Program n=184 n=1013 f % % Non-Drinker 32 18 122 12 Non-problem drinker 1 11 60 565 56 Problem Drinker 41 22 326 32 Total 184 100.0 1013 100.0 Note. x1 (2, 1197) =90, p < .01. Eta: = .007 Perceived Environment Va_riables aad Demographic Variables The purpose of the second research question is to examine the relationship between the perceived environment variables and the demographic variables of gender, size and type of residence hall, specialty housing option and residential academic program. The perceived environment variables include two distal measures, Residence Hall Models for Drinking and Resident Assistant Control and four proximal measures including Approval for Drinking, Models for Drinking, Pressure for Drinking and Control. The means and standard deviations, multivariate analysis of variance (MANOVA), t-tests, one- way analysis of variance (ANOVA) and multiple comparison statistics will be utilized for the data analysis for this research question. For clarification, Research Question 2 and Research Hypothesis 2 are restated here: Rmch Question 2: To investigate the relationship between factors within the perceived residence hall environment and drinking behavior among freshmen students. More specifically, this study will investigate the balance between 63 instigations for and controls against problem alcohol use within both the distal and proximal structures of the residence hall environment. Resegch Hypotheses 2: There will be no significant differences in the scores on the perceived environment variables of Approval for Drinking, Models for Drinking, Pressure for Drinking and Controls among MSU freshmen when categorized by gender, type of residence hall, size of residence hall, specialty housing option or residential academic program. The first statistical hypothesis set forth in response to Research Question 2 is stated below: Statistical Hypothea_is__7_: There are no significant differences among MSU freshmen scores on perceived environmental variables when categorized by gender, type of residence hall, size of residence hall, alcohol free room, alcohol fiee floor and residential academic program. Table 4.9 presents the results of the one-way MANOVA of the perceived environment variables when categorized by gender, type of residence hall, size of residence hall, alcohol free room, alcohol free floor and residential academic program. An analysis of the table indicates that there were significant differences in the scores on the perceived environmental variables with respect to all of the six categorization variables. To provide information regarding which variables contributed to the differences between the categorical groupings, the univariate F—tests for each perceived environmental variable are also presented in the table. Each categorization variable will be discussed separately. A t-test or a one-way AN OVA was conducted on each subscale item of the variable that contributed 64 significantly within the multivariate analysis to provide some insight into where the differences may exist. The univariate statistics are not adjusted for the fact that several comparisons were made and thus should be used with a certain amount of caution (Norusis, 1994). To avoid reporting significant results in the presence of mild violations of homogeneity of covariance matrices or multivariate normality, Pillais’ Trace was used for the multivariate criterion as it offers the most in terms of power and robustness (Norusis, 1994). 65 59V a... .5. V moo .mc. vde m8. 2... :88. 4.2 88. 8..“ :88. 8... :88. 8.8 :88. S... 2:32.. .88. 8.8 :88. 2.. :88. 8.8. 8.8. 8.... .88. 2.8 :88. SM 8882 2.. 28 3. SN :88. 2... R... :2 N... a: 88. .8... .928... 8.. 8.... 88. men :88.. 3a 2.. 8.~ .8. NR. :88. 68 3:80 .288 E. 88. .8. 8m. 3... 8.... 8.8. 8... .8. 8.~ :88. 8.... 5. 8.. EN _..:.8. 8.: :88. 2: NE. SN 8.... 8... 88. 3. .25. .191. m 4.1... m an m 4.1... m an H... new m .3. .8... "2 .83. "2 85 25 .880 .8. RN m. .8. 58mm... 0.588... .8828... .8. 8... 88. .8... 8.... .282 .8. N. . .. no... .83. 8.... .882 .8. m. .m 8.8. .3. 8.8.8.. .8 8a .8. 8.~ 88. .3. 8828.. .0 2:... .8. 8.2 8. .. .880 H4502...: 1.9.. .3; 8. -.=8.i.....fil_m or... Inesmln. 8.131..Im_~..__._.1.. :8.on 28239.4 3:528... E... .830 mamas: 538% £2. 8528,.— ..o cum in: 8528.3. mo 09C. .8250 .3 $3855 EoE:o._>=m 338......— ..o.. 38.8-”. 8233.5 can <>OZ .05). Statistical Hypothesis 3: There are no significant differences in the self-reported drinking behavior of Michigan State University Freshmen when categorized by type of residence hall. The null hypothesis was not rejected (X2 = (4, 1197) 7.35, p > .05). Statistical Hypothesis 4: There are no significant differences in the self-reported drinking behavior of Michigan State University Freshmen when categorized by alcohol fiee room residency. The null hypothesis was rejected at the .001 level of significance (X2 = (2, 1197) 151.0, p < .001). Eta = .357 with 12.7% ofthe variance in drinking classification explained by differences between alcohol fi'ee room residency and non-alcohol fi'ee room residency. S_t_atistica1 Hypothesis 5: There are no significant differences in the self-reported drinking behavior of Michigan State University Freshmen when categorized by alcohol free floor residency. The null hypothesis was rejected at the .001 level of significance (X2 = (2, 1197) 14.0, p < .001). Eta = .094 with less than 1% of the variance in drinking classification explained by differences between alcohol free floor residency and non- alcohol free floor residency. Statistical Hypothesfip; There are no significant difi’erences in the self-reported drinking behavior of Michigan State University Freshmen when categorized by enrollment in a residential academic program. The null hypothesis was rejected at the .05 level of significance (X2 = (2, 1197) 9.0, p < .05). Eta = .086 with less than 1% of the variance in drinking classification explained by differences between enrollment in a residential academic program and non-enrollment in a residential academic program. Statistical Hypothesis 7: There are no significant differences on the scores of perceived environment variables of Michigan State University freshmen when categorized by gender. The null hypothesis was rejected at the .05 level of significance. A one-way MANOVA revealed significant differences between males and females on the Modeling for drinking, Approval for drinking, Pressure for drinking and Control measures. 90 Statistical Hypothesis 8: There are no significant differences on the scores of perceived environment variables of Michigan State University freshmen when categorized by size of residence hall. The null hypothesis was rejected at the .05 level of significance. A one-way MANOVA revealed significant differences between respondents on the Modeling for drinking, Approval for drinking, Pressure for drinking and Control measures when categorized by size of residence hall. Statistical Hypothesis 9: There are no significant differences on the scores of perceived environment variables of Michigan State University freshmen when categorized by type of residence hall. The null hypothesis was rejected at the .05 level of significance. A one-way MANOVA revealed significant difi’erences between respondents on the Modeling for drinking, and Pressure for drinking measures when categorized by type of residence hall. Sgtistical Hypother 10: There are no significant differences on the scores of perceived environment variables of Michigan State University freshmen when categorized by alcohol free room residency. The null hypothesis was rejected at the .05 level of significance. A one-way MANOVA revealed significant differences between respondents on the Modeling for drinking, Approval for drinking, Pressure for drinking and Controls measures when categorized by alcohol fi'ee room residency. Statistical Hypothesis 11: There are no significant difl’erences on the scores of perceived environment variables of Michigan State University freshmen when categorized by alcohol free floor residency. The null hypothesis was rejected at the .05 level of significance. A one-way MANOVA revealed significant differences between respondents on the Modeling for drinking, Approval for drinking, Pressure for drinking and Controls measures when categorized by alcohol fiee room residency. Statistical Hypothesis 12: There are no significant difl’erences on the scores of perceived environment variables of Michigan State University freshmen when categorized by residential academic program. The null hypothesis was rejected at the .05 level of significance. A one-way MANOVA revealed significant differences between respondents on the Modeling for drinking measure when categorized by enrollment in a residential academic program. St_atisticfial Hypotheses 13-16: 13. Problem drinkers will perceive more Roommate, Friends’ and Resident Assistant Approval for Drinking than Non-drinkers and Non-problem drinkers. 14. Problem drinkers will perceive more Roommate, Friends’ Modeling for Drinking than Non—drinkers and Non-problem drinkers. 91 15. Problem drinkers will perceive more Roommate and Friends’ Pressure for Drinking than Non-drinkers and Non-problem drinkers. 16. Problem drinkers will perceive less Roommate, Friends and Resident Assistant Controls than Non-drinkers and Non-problem drinkers. Statistical hypotheses 13-16 were accepted at the .05 level of significance. A one- way MANOVA and post hoc Scheffe‘ multiple comparison tests supported the directional hypotheses. The third question tested the usefirlness of Problem Behavior Theory in predicting Problem drinking behavior. Three analyses were conducted using discriminant analysis and multiple regression techniques. Each analysis used a different combination of drinking classification groupings. In each analysis, using the perceived environment measures resulted in a statistically significant improvement of predicting drinking classification accuracy over chance. The results are summarized below: Analysis 1: Predicting Non-drinker, Non-problem drinker and Problem drinker status. The overall classification rate was 65%, which was a statistically significant (z=6.00, p < .001) improvement of the 56% accuracy rate predicted by chance alone. Fifty-one percent of Problem drinkers were accurately classified. Analysis 2: Predicting Non-problem drinker and Problem drinkers status. Non-drinkers and Non-problem drinkers were combined into one group. The overall classification rate was 76%, a statistically significant (2 = 5.6, p_< .001) improvement over the 62% accuracy rate predicted by chance alone. Forty-eight percent of Problem drinkers were accurately classified. Analysis 3: Predicting Non-drinker and Problem drinker. A test of the theory to discriminate between the groups with most extreme scores (high versus low) on the perceived environment measures. The overall classification rate was 88.8 %, a statistically significant (z=17.8, p < .001) improvement over the 70% accuracy rate predicted by chance alone. Ninety-five percent of Problem drinkers were accurately classified. Analysis 4: Predicting Negative Consequences. A test of a multiple regression model to predict the continuous measure of Negative Consequences using the perceived environment measures. Four measures contributed significantly to the model (Residence 92 Hall Modeling, Friends’ and Roommate Models for drinking and Friends’ Pressure for drinking) and 33 % of the variance in negative consequences was accounted for using the model. Analysis 5: Predicting Times Drunk. A test of the multiple regression model to predict the continuous measure of Times Drunk using the perceived environment measures. Four measures contributed significantly to the model (F riends’ Pressure for Drinking, Friends’ Models for Drinking, Residence Hall Modeling, and Roommate Modeling for Drinking) and 10 % of the variance in negative consequences was accounted for using the model. Chapter 5 SUMMARY, FINDINGS, CONCLUSIONS AND RECOMMENDATIONS Summm The problem of this study was to examine the relationship between the perceived residence hall environment and problem drinking behavior among Michigan State University fieshmen. Participants for this study were randomly selected fi'om the 1996 fieshman class who lived on campus and had not requested their name and address be restricted. Given the length and nature of the questionnaire, the sample size was set at 1800 to ensure an adequate numbers of respondents within the each drinking classifications. An introductory letter was sent to all respondents during the fourth week of school informing them of the study and encouraging their participation. The eight page survey was sent to each respondent with a cover letter, a return postcard, a pencil and return envelope. After the initial mailing, respondents who had not returned their postcard were sent a second survey and cover letter again requesting their participation. Two weeks after the second mailing, a third mailing was sent to those respondents who had not returned their post card. A total of 1222 surveys were returned, representing a 68% return rate. Of those 1222 surveys returned, 1197 were complete and used for statistical purposes, with the data analyzed using the SPSSX Statistical Package for the Social 93 94 Sciences. Three primary research questions were formulated with 16 corresponding statistical hypotheses which were tested at the .05 level of significance. Statistical analyses included both descriptive and inferential statistics, utilizing t-tests, one-way ANOVA, one-way MANOVA, Scheffe‘ ’ post hoc multiple comparison tests, and discriminant analysis. The findings presented for each research question will be summarized in this section. Drinking Classification fll Demogiaphic Variables The purpose of Research Question 1 was to investigate the relationship between drinking classification and the demographic variables of gender, size of residence hall, type of residence hall, specialty housing option and residential academic program. Men With regard to gender, there was a statistically significant difi’erence in the percentage of males and females when categorized by drinking classification. Upon further inspection of the data, however, it is clear that the differences are quite small. This finding is similar to many others mentioned in the review of the literature (Engs & Hansen, 1990) suggesting that high percentages of both male and female freshman undergraduates drink and that a significant percentage of both genders engage in problem drinking. Size and Type of Residence Hall_. No significant differences were found for either size of residence hall or type of residence when respondents were categorized by size or type of residence hall. This finding suggest that these two variables were not associated with the drinking behavior of 95 the respondents. Alcohol Free Room. Thirty-two percent of the respondents living in Alcohol Free Rooms were classified as Non-drinkers and only 10% were classified as Problem drinkers. Ninety percent of the respondents living in the Alcohol Free Room option fall under the umbrella of either Non-drinker or Non-problem drinker. Clearly, students who select the Alcohol Free Room option are much more likely to be Non-problem drinkers and less likely to be Problem drinkers. This finding may be useful to housing staff in determining the percentage of rooms occupied by students within each drinking classification. When viewed from the perspective of drinking classification, 60% of the Non- drinkers and 7% of Problem drinkers live in Alcohol Free Rooms. The majority of the Non-drinkers included in this study chose the Alcohol Free Room option. This finding provides substantial evidence of the appeal of the Alcohol Free Room option for fieshman students who do not consume alcohol. In this regard, the Alcohol Free Room option is an important element in the residence hall environment as part of the alcohol and other drug prevention program. Alcohol Free Floors. With regard to Alcohol Free Floors, 17% of the respondents living on Alcohol Free Floors were Non-drinkers compared to only 6% of respondents identifying themselves as Non-drinkers on the Non-alcohol Free floors. Twenty-nine percent of the respondents living on Alcohol Free Floors were Problem Drinkers compared to the 37 % living on Non-Alcohol Free Floors. Seventy-one percent of the respondents living on 96 Alcohol Free Floors fall within the category of Non-drinker or Non-problem drinker compared to 62% living on Non-alcohol free floors. The drinking behavior of the student population residing on Alcohol Free Floors differs significantly fiom the drinking behavior of students living on Non-alcohol free floors. A student living in on an Alcohol Free Floor is more likely to be living with Non-drinkers and Non-problem drinkers and less likely to be living with Problem drinkers than a student living on a regular residence hall floor. The Alcohol Free Floor appears to be a less appealing choice for Non-drinkers than the Alcohol Free Room Option as only 11% of the Non-drinkers chose to live on an Alcohol Free Floor. Drinking behavior may be less a factor for a student who chooses an Alcohol Free Floor than the student who chooses an Alcohol Free Room. Alcohol Free Floors are reputed to be quieter and cleaner so the choice of this option may be due to increased housing satisfaction more than drinking behavior. Regardless if the option is chosen by students due to their own drinking preference or for a quieter, cleaner living environment, they will be less exposed to fewer problem drinking models on the floor. Aeaidential Acafdemic Prm Eighteen percent of the respondents enrolled in residential academic programs identified themselves as Non—drinkers compared to 12% of respondents who identified themselves as Non-drinkers and were not enrolled in such a program. Twenty-two percent of the respondents enrolled in residential academic programs identified themselves as Problem drinkers compared to 32% of respondents not enrolled in such a program. There appears to be an association between drinking behavior and enrollment within a residential academic program albeit a minor one. All residential academic 97 programs at MSU have an admissions policy allowing any student admitted to MSU to enroll in the programs on a first come, first serve basis. Thus, the drinking behavior of students enrolled in residential programs might well reflect the same characteristics as students not enrolled in such programs. However, two of the programs, James Madison College and Lyman Briggs School, have historically attracted students with superior academic credentials. The higher number of Non-drinkers and the fewer number of Problem drinkers within residential academic programs is a finding that is consistent with J essor’s premise that students who place a higher value on academic achievement are less likely to engage in problem drinking. Enrollment in a residential academic program is a variable Jessor (1973) would describe as part of the distal social environment and lacks a substantial relationship with variation in drinking behavior. Perceived Environment Vaflbles and Demogiaphic VariAblea There were two primary research hypotheses addressed by Research Question 2. The first research hypothesis examined the relationship between the perceived environment variables, Approval for drinking, Modeling for drinking, Pressure for drinking and Controls, and the demographic variables, gender, size and type of residence hall, specialty housing option and residential academic program. A one-way MANOVA revealed that there are significant differences between the measures on the perceived environment variables when respondents were categorized according to each of the demographic variables. Univariate tests were conducted for each combination of perceived environmental variable and demographic category to further examine the differences between the demographic grouping variables. 98 M91 With regard to gender, male respondents perceived more Residence Hall Modeling for drinking, Friends’ Approval for drinking, Friends’ and Roommate Models for drinking, Friends’ and Roommate Pressure for drinking than female respondents. Female respondents, however, perceived more Resident Assistant Approval for drinking, and Resident Assistant, Friends’ and Roommate Controls. There were no statistically significant differences between males and females with regard to Roommate Approval for drinking. This finding suggests that the majority of male perceive more support for drinking within their living units than female respondents. Male residence hall floors can be characterized as more supportive of drinking through more approval, modeling and pressure to drink and less controls within the environment against problem drinking. Size of Res_idence Hag Respondents living in large halls perceived more Roommate modeling and Roommate pressure for drinking than respondents living in small halls. Respondents living in medium size halls perceive more Resident Assistant Control than respondents living in large halls. No significant differences were found for Residence Hall Modeling, any of the Approval for Drinking variables or Friends’ and Roommate Controls. The size of the residence hall seems to make little difference in students’ overall perceptions of the residence hall environment. The size of the residence hall building may be too “distal” a variable in relation to the respondent to be of much practical significance with regard to either drinking behavior or perception of the environment. With regard to the differences in perception of Roommate Modeling and Pressure, the roommate relationship may be of 99 more importance to students living in the large halls. Type of Res_idence Hflli When respondents were categorized by type of residence hall, a one-way AN OVA revealed that freshman students perceive less modeling and less pressure for drinking within all freshman living units than fieshman who are assigned with upper class students. No other differences were found on the other measures between the difi’erent types of halls. Thus, there does not seem to be much difference in the way students’ perceive the residence hall environment within the different types of balls. Alcohol Free Room. Statistically significant difi’erences were found between respondents living in Alcohol Free Rooms and Non-alcohol Free rooms on nine of the eleven perceived environment variables including all of the Modeling for drinking variables, Approval for drinking variables, Roommate Pressure for drinking and both Controls measures. This finding provides strong support that students residing in Alcohol Free Rooms perceive less approval, modeling, and pressure for drinking and more controls against problem behavior within their living unit. Respondents living in Alcohol Free Rooms also perceive less overall residence hall modeling for drinking suggesting that the drinking environment within a student’s room may be linked to his or her perception of the total environment. Alcohol Free Floor. Statistically significant differences were found on five of the eleven variables when respondents were categorized by Alcohol Free Floor residency. Again, there were significant differences between the two groups with regard to all of the Modeling and 100 Pressure variables. Respondents living on an Alcohol Free Floor perceived less overall residence hall modeling for drinking suggesting that floor drinking environment may also be linked to the perception of the drinking climate within the entire hall. There were no differences between the groups with regard to the Approval variables. This finding suggests that the floor environment is different from other residence hall floors in terms of the amounts of drinking models present on the floor and the amount of pressure for drinking exerted on the floor. The lack of significant differences on the Approval for drinking and the Controls variables between Alcohol Free Floor residents and those not living on such a floor do not differ in their attitudes towards drinking or their attitudes toward problem behavior. Many students may choose this option to avoid an environment in which there is more drinking and pressure for drinking and not because of their personal beliefs about alcohol use. RasifdenflAcanleniic Prom Significant differences were found between respondents enrolled in a residential academic program and those who were not enrolled in such a program on only one variable, Models for Drinking. This finding suggests that students enrolled in residential academic programs perceive fewer models for drinking within their floor environment. Perceived Environment Vambles a_nd Drinking Classification The second research hypothesis explored the differences between the scores on the perceived environment variables when respondents were categorized by drinking classification of Non-drinker, Non-problem drinker and Problem drinker. The four corresponding statistical hypotheses were directional and postulated that Problem drinkers 101 would score higher than Non-problem drinkers and Non-drinkers on the Modeling, Approval, and Pressure variables and they would score less on the Controls variables. A one-way AN OVA supported all of the directional hypotheses except for Resident Assistant Approval for drinking with Problem drinkers perceiving less Resident Assistant Approval for drinking than either Non-drinker or Non-problem drinkers. These results support the relationship between the perceived environment and drinking behavior postulated by Jessor in Problem Behavior Theory. The connection of the social environment- especially the social environment within the residence hall-with problem drinking among college students is supported. Jessor (1973) describes “environmental proneness” for problem drinking as an environment characterized by greater perceived approval for drinking, greater perceived modeling for drinking, greater perceived pressure for drinking and less perceived controls against problem behavior. The Problem drinkers included in this study clearly perceive their residence hall environment as possessing the attributes listed above. There were no significant differences on two of the perceived environment variables tested, Resident Assistant Control and Friends’ Pressure for drinking. No variation on scores was found on the perception of Resident Assistant Control for any of the analyses. The researcher hypothesizes that disapproval or discouragement of problem behavior from the RA is not perceived in the same manner as disapproval or discouragement of problem behavior fiom either fiiends or a roommate. Perhaps the opinion of the RA is not as meaningful to the perceiver and therefore does not serve as a control against problem behavior. The enforcement role of RAs may preclude the 102 perceiver from attaching meaning to the RA’s opinion in the same way they would to a fiiend or roommate. The lack of association between perception of RA control and drinking behavior within the residence hall suggests that drinking behavior is more associated with the social influence of fiiends on the floor and roommate than with the institutional relationship of staff to student. The timing of this study may be factor in this finding. Students’ relationship with the Resident Assistant may change over time. It is possible that the relationship was not as well established within the first six weeks of school as the fiiends’ on the floor chosen by the respondent. Additionally, within the first six weeks of school, disciplinary consequences for problem behavior may not be clear to students and the social control firnction of the RA not yet established. Friends’ Pressure for drinking was also a variable for which no significant difi’erences were found between drinking groups. The pressure variable must be interpreted cautiously. The mean scores on this variable are quite low indicating a very small amount of pressure was perceived. Overwhelrningly, students responded that they almost never or very infrequently experienced pressure to drink from fiiends or roommate. The researcher hypothesizes most students do not experience overt pressure for drinking; that is, no one forces them to drink. Overall, these findings support the hypothesis that peer influence is associated with drinking behavior within the perceived residence hall environment. More specifically, two sources of peer influence in the residence hall environment can be more clearly identified; fiiends on the floor and roommate. Thus, the perception of friends on the floor and roommate approval, modeling, and pressure for drinking are significantly related to 103 drinking behavior. The purpose of Research Question 3 was to test the usefulness of Problem Behavior Theory to predict drinking classification utilizing discriminant analysis. Three separate analyses were conducted using three different drinking classifications. The first analysis was conducted to predict membership in the three original drinking classifications of Non-drinker, Non-problem drinker and Problem drinker discumed at the beginning of this chapter. The classification rate of 65% for this analysis was significantly better than chance although only 51.5 % of Problem drinkers were accurately predicted. Using the second analysis, which combined Non-drinkers and Non-problem drinkers into one single group, the drinking classification accuracy was slightly reduced to 62% with 89.8% of Non-problem and 48% of Problem drinkers accurately classified. The third analysis tested the ability of the perceived environment variables to discriminate between the two extremes groups, Non-drinkers and Non-problem drinkers. The classification rate was significantly improved to a 88.8% with 95 .4% of Problem drinkers accurately classified. For all three analyses, the Modeling variables made the most significant contribution to the prediction of problem drinking. Similarly, Jessor’s identified Friends’ Models for problem behavior as the most significant perceived environment variable for both college men and women in the prediction of problem drinking (Jessor, 1991). Modeling as a form of social influence that increases alcohol consumption is well supported by these findings. The perception that an individual’s fiiends use alcohol is significantly related to current and filture alcohol use (Bergen & Olsen, 1963, Castro, et al., 1987; Chassin et al., 1981). For college students, the perception of the overall number 104 of students using alcohol within the hall as well as both fiiends on the floor and roommate’s use of alcohol is a factor in predicting alcohol use. The prevalence of models provides the opportunity to learn how to engage in the behavior, access to alcohol, evidence that the drinking can occur and that social control against the drinking is minimal. One purpose of this study was to test the usefillness of Problem Behavior Theory in predicting problem drinking using perceived environment variables within the residence hall environment. The results of the predictive analyses confirmed the usefirlness of utilizing the perceived environment measures within the residence hall context to predict drinking behavior. Modeling for drinking within the hall was the primary factor in discriminating between drinking groups. The variables of Approval for drinking, Pressure for drinking and Controls were also significant factors in discriminating between the drinking groups although to a lesser extent than the modeling variables. It may be concluded from these analyses that the more modeling for drinking a student perceives in his or her living environment the more likely he or she is to engage in problem drinking. Controls against problem behavior, however, were also a significant factor within the environment. The perception that others, either fiiends on the floor or roommate, hold relatively strict standards for behavior and would disapprove of problem behavior, seems to mitigate against the engagement in problem behavior. As suggested by Jessor (1977), the perceived environment consists of “patterns” that are either more or less conducive to problem behavior. A student living on a residence hall floor characterized by high approval for drinking, fi'equent modeling of drinking, and less controls against 105 problem behavior may be residing in an environment that provides greater proneness to problem behavior. Two multiple regression analyses revealed that certain perceived environment variables usefirl in predicting two continuous measures of Negative Consequences and Times Drunk. The perceived environment variables of Residence Hall Modeling, Friends’ and Roommate Modeling for drinking and Friends’ Pressure for Drinking entered the regression model to predict Negative Consequences and accounted for almost 33% of the variance among students in the fiequency of negative consequences they experienced. This is strong support that peer modeling within the residence hall environment is related to students’ engaging in alcohol-related problem behavior. A second analysis was conducted to predict the number of times students’ reported they had been drunk since the beginning of school. The model generated from this analysis revealed that the variables of Residence Hall Modeling, Friends’ Approval and Modeling for drinking and Roommate Pressure drinking may be related to fi'equency of drunkenness. The model accounted, however, for only 10 % of the variance among students reported frequency for drunkenness. The researcher speculates that the Times Drunk measure was not as reliable a measure as the Negative Consequences Scale because it was generated fiom rank order data rather than continuous data thereby affecting the regression equation. Implications fogractice The most important implication of this study is that students’ perceptions of the residence hall environment appear to be related to their drinking behavior. The residence hall environment seems to present a source of social influence that affects drinking 106 behavior. Ideally, practitioners responsible for either housing programs or management of alcohol and other drug programs will find the findings of this study useful in their design of programmatic interventions. The vast majority of students included in this study consume alcohol on a regular basis. Thus, there exists strong evidence that the norms of this sample of college freshmen support at least some alcohol use. It is essential that administrators acknowledge the reality of alcohol use among fi'eshman students entering the institution; the likelihood that many students will increase their drinking after arriving on campus and that student alcohol use is related to social influences within the collegiate living environment. Clearly, one of the most important interventions is the offering of both alcohol free room and alcohol free floor options for students. Ifthese options continue to be chosen by future students in the same or increasing proportions, it is very probable students will perceive much less approval, modeling, and pressure for drinking in their immediate living environment. The degree to which a student is located in a living context where problem behavior is prevalent and where there is social support for its occurrence will continue to present administrators with a complex challenge. According to Problem Behavior Theory, to ofi’set a perceived environment supportive of problem alcohol use an individual would require strong “personality controls” or the perception of strong social controls within the environment. Given the findings of this study, it is diflicult to say whether the Resident Assistant could be the source of strong social control. Given the pervasive nature of student drinking, messages of abstinence (“Just Say No”) may not be usefirl. Rather, students may benefit from a well-defined definition of 107 responsible drinking that focuses upon the avoidance of negative consequences. Framing alcohol use within the broader context of healthy behavior may be a promising approach. Martin and Hoffman (1993) promote the use of the Health Belief Model which includes five predictors of health protective behaviors including a). perceived susceptibility; b) perceived severity of consequences; c) perceived benefits of change; d) perceived barriers to change; and e) perceived self-efficacy. The Health Belief Model provides an explanation as to the reasons individuals engage in health-risking activities. Students may not be firlly knowledgeable of the potential negative consequences of abusive drinking nor appreciate the benefits of responsible alcohol use; they may not understand their own motivations for abusing alcohol or realistically assess their ability to change their behavior. For students, peer influence and acceptance within the college culture may be very difficult barriers to changing drinking behavior. Utilizing the Health Belief Model, programmatic interventions could assist students in understanding the potential negative consequences of alcohol abuse as well as identifying ways to resist environmental pressures. The lack of significance of the Resident Assistant in influencing drinking has implications for RA training in the area of alcohol education. The fieshman students in this study were not significantly influenced by their RA’s approval-disapproval of problem behavior with regard to their alcohol use. It is possible that the RA’s approval- disapproval of problem behavior is important in other ways such as in discouraging rule violations. Nevertheless, it clear that the RA does not serve as a control agent for students in terms of drinking behavior. In addressing student drinking behavior, the role of the RA will have to be conceptualized differently than as a source of disapproval 108 against problem drinking. Research into how normative behavior can be influenced by an external agent of control may be usefill. The role of the roommate upon drinking behavior must be considered. While it is neither feasible nor ethical to assign students based upon drinking classification, facilitating discussions about roommate alcohol use could occur with students upon their arrival to campus. This study focused on the perceived environment and asked students to report their perceptions about the drinking behavior of their peers. It did not include external measures of the environment to check the accuracy of students’ perceptions of their peers’ behavior. Other researchers have attempted to measure students’ peer environments more objectively (Kandel, et al., 1980). For example, a researcher can check the accuracy of a respondent’s perceptions of his or her fiiends’ alcohol use by gathering data directly fiom the respondent’s fiiends and comparing it to the data provided by the respondent. Although these measures may provide a more reliable estimate of the actual amount of peer alcohol use to which the student is exposed, this objective reality may not be what the student experiences in their environment. Berkowitz (1990) states that it may be more important for practitioners to understand the nature of peer influence as witnessed through the eyes of the student. That is, what the student perceives to be true may be more important than the reality of peer norms. In that a student’s perceptions of peer alcohol use appear to be related to his or her own use, practitioners may seek ways in which to influence students’ perceptions of peer alcohol use within the collegiate environment. Berkowitz (1990) suggests that 109 educational experiences can play an important role in affecting student’s perceptions and that it may be possible to “resocialize” students through educational efforts to examine more closely the social contexts in which their peers talk about and use alcohol and other drugs. The results of this research clearly indicate that peer modeling and approval of use are salient predictors of college student alcohol use. This finding confirms the need to continue to focus prevention efl’orts on not only the ways in which students’ perceive their peers but also the ways to reduce conformity to peer norms of irresponsible alcohol use. Limitations Problem Behavior Theory utilizes three systems: 1) personality system, 2) perceived environment system and 3) behavior system to explain problem behavior. The primary limitation of this research is that only the measures of the perceived environment system and behavior system measures fi'om Problem Behavior were tested. A study employing the personality system as well as the other measures would account for more variation among the different drinking groups and perhaps offer an improved classification rate for problem drinkers. According to Jessor (1991, p. 601), an account of problem drinking must include “multiple explanatory domains as well as their interactions.” J essor’s 1981 Young Adult Survey had several limitations. Many of the measures were not continuous, and therefore, were difficult to use in the statistical analyses. Additionally, the inclusion of the pressure variable within the modeling scale actually decreased the reliability of the measure. It is the opinion of this researcher that modeling for drinking and pressure for drinking are two distinct concepts and should not be 110 combined into an overall modeling measure. In addition, the negative consequences included in the survey could be more complete. While the list is quite inclusive, additional items more reflective of the consequences experienced by women (unwanted pregnancy or sexual assault) would make it more comprehensive. Additionally, the items used from the Young Adult Questionnaire were modified to ask specifically about relationships within the residence hall context. The researcher maintains that increasing specificity would not decrease the validity of the items because the questions may be more clearly understood by the respondent. The study was conducted in the first six of weeks of the school year in order to control for students’ self selection of living environments based upon drinking preferences. It is possible that at the beginning of the school year students are more likely to experiment with new behaviors to help them cope with feelings of anxiety, respond to pressure to fit in with new acquaintances, or to combat feelings of loneliness or isolation. Data collected at a different time of year may reveal a decrease in frequency of drunkenness or negative consequences as students learn to avoid either becoming intoxicated or experiencing negative consequences when they drink. In using discriminant analysis, it is recommended to use an external analysis on a different sample to test the classification results generated in this study. This study was limited to an internal analysis in which the classification rule was derived from the same data used to generate the classification table. This may be problematic in that an internal analysis may rely on variations specific to the sample and may not be generalizable to other samples. According to Huberty et al. (1989, p. 161), “an internal analysis may be 111 acceptable if the total number of cases is large; large defined by the current authors as a data set in which the smallest group size is five times the number of predictor variables.” “nth the use of six predictor variables, the minimum number of respondents per group needed was 30; which this sample clearly exceeded. Therefore, while an external analysis would provide additional information, the results of this internal analysis should be considered satisfactory. Recommendations for firture reseam Further research utilizing the personality system as well as the perceived environment system is needed. Drinking behavior among college students is clearly a filnction of the complex interplay of both individual personality traits as well characteristics of the perceived residence hall environment. Additional research in this area would assist in identifying which personality variables are implicated in college student drinking and which personality and perceived environment variables combine to best predict college student problem drinking. One of the few factors within the residence hall that can be controlled by the institution is the role of the undergraduate staff member. Resident Assistants spend an enormous amount of time enforcing alcohol policies on their floors. This finding provided minimal support for the role of the RA in affecting drinking behavior through either of the two RA variables; RA Approval for drinking or Control. Research specifically designed to determine the relationship between RA enforcement of alcohol policies with drinking behavior may be of interest. While J essor has used Problem Behavior Theory to examine cross-sectional data, 112 he recommends longitudinal studies to measure change in drinking patterns over time. Collecting data from students who remain in the same living units through out an academic year could provide valuable insight into how drinlo'ng behavior changes over time and as well as allow for comparisons across living units at different points in time. Problem Behavior Theory was usefirl in not only predicting problem drinking but also in predicting Non-drinking and Non-problem drinking as well. Research might focus on the factors predicting these drinking classifications as opposed to problem drinking. The theory seemed most usefill in predicting the extremes of drinking behavior, either non-drinking or problem drinking. The perceived environment variables were not as useful in predicting students who fall in the middle group of non-problem drinking. Future research is needed to more accurately understand the differences between all three groups. By incorporating additional variables from both the personality and behavioral system of Jessor’s theory it may be possible to better predict the non-drinkers and non-problem drinkers. Both the personality system and behavior system contain measures that reflect controls against a student’s engagement in problem behavior. Identifying these controls may be beneficial to administrators in understanding why some students engage in problem drinking while others do not. APPENDIX A % APPENDIX A RESIDENCE HALL SURVEY IT ONLY TAKES 10 MINUTES TO COMPLETE! Aerr completing the survey, place it in the return envelope. Fill out the postcard. Take the survey and the postcard to your ball reception desk and ask them to place both in campus mail. Please return the survey 11th away so I can enter you into Ihe drawing for the gift certificate to the MSU bookstore. Thank you very much! E YOUR RESIDENCE HALL This section has to do with the students you know in your residence hall. FILL IN THE CIRCLE All of Them Most of Them Several 1-2 None I. About how many of the students 0 O O O 0 you know in your residence holl drink alcohol at least sometimes? 2. As for as you know, about how 0 O 0 O O mony students on YOUR Iloor drink alcohol of least sometimes? 3. About how many of the students 0 O O O 0 you hang around with drink alcohol 01 least sometimes? 4. About how many of your close friends 0 O O O 0 drink alcohol of least sometimes? [13. FRIENDS ON YOUR FLOOR This section has to do with your FRIENDS who live on your residence hall floor. 5. How do most of the people you are close to on your floor feel about drinking? 0 They Approve 0 They Don'l Core 0 They Disopprove 6. Do you have any friends on The floor who do not drink 01 all? 0 None 0 One or Two 0 Several 0 Most 0 All 7. Do you have any friends on the floor who drink foirly requlorly, once 0 week or more? 0 None 0 One or Two 0 Several 0 Most 0 All 8. How much do your friends on your floor drink of 0 single sitting? 0 They don't 0 One or Two 0 Three or 0 Five or Drink Drinks Four Drinks More Drinks 9. Hove any of your friends on your floor Tried To get you to drink or to drink more than you really wont to? 0 Never 0 Once or Twice O Severollimes 0 Very Often 113 10. If you were going to do something illegal or that most people think of as wrong, would your friends on your floor try to influence you not to? O Definitely 0 Probably O Probably O Definitely Would Would Would Not Would Not If you actually do something most peOple would think of as wrong, what sort of reaction do you get from your friends on your floor? 0 Show Strong 0 Show Moderate O ShowaLittle 0 Don't React O Show Some Disapproval Disapproval Disapproval at All Approval k. YOUR ROOMMATE This section has to do with your roommate. CIRCLE YOUR ANSWER 12. How does your roommate feel about drinking? S/ he Approves S/ he Doesn’t Care S/he Disapproves Does your roommate drink? Yes No If your roommate does drink, about how much does s/he drink in a single sitting? S/He doesn’t One or Two Three or Four Five or more Drink Drinks Drinks Drinks Has your roommate tried to get you to drink or to drink more than you really want to? Never Once or Twice Several Times Very Often If you were going to do something illegal or that most people think of as wrong, would your roommate try to influence you not to? Definitely Probably Probably Definitely Would Would Would Not ould Not If you actually do something most people would think of as wrong, what sort of reaction do you get from your roommate? Show Strong Show Moderate Show a Little Don't React Disapproval Disapproval Disapproval at All ID. YOUR RESIDENT ASSISTANT This section has to with your Resident Assistant. CIRCLE YOUR ANSWER 18. If you were going to do something illegal or that most people think of as wrong, would your Resident Assistant try to influence you not to? Definitely Probably Probably Definitely Would Would Would Not Would Not 114 19. If you actually do something most people would think of as wrong, what sort of reaction do you get from your Resident ksistant? Show Strong Show Moderate Show a Little Don’t React Show Same Disapproval Disapproval Disapproval at All Approval 20. How does your Resident Assistants feel about drinking? S/ he Approves S/he Doesn't Care S/ he Disapproves l 21. Does your Resident Assistant enforce the alcohol policies on your floor? Never Some of the Time All of the Time IE. DRINKING This section is about alcohol use. CIRCLE YOU ANSWER 22. 23. 24. Have you ever had a drink of beer, wine, or liquor (not just a sip or a taste of someone else’s drink)? Yes No If "No", please go on to Question 50. How long has it been since you last had a drink of beer, wine or liquor? A Week or Less More than 6 Months, but less than 2 years More Than a Week, Less Than a Month 2 Years or More More than a Month, Less than 6 Months Have you had a drink of beer, wine, or liquor more than two or three times in your life? If no, please to to Question 50. Yes No If you have not had a drink since you came to MSU, please go on to Question 50. TH 25. 26. l Tl SPRTI RlNKl FTRRRll T Have you had a drink of beer since you came to MSU this fall? Yes No (If "No“, go to Question 30). On how many days have you had a drink of beer? Days On how many of those days that you drank beer, did you drink beer at two or more different times during the day (for instance, in the afternoon, then again in the evening?) Days 115 28. 29. 30. 31. 33. 34. 35. 36. 37. Think about those times you’ve drunk beer while at MSU. About how many cans or bottles of beer did you usually drink at any one sitting? (A regular can or bottle contains 11 or 12 ounces). II or More Cans or Bottles (about 2 Six Packs) 3 Cans or Bottles 8 to 10 Cans or Battles (about 1 1/2 Six Packs) 2 Cans or Bottles 6 to 7 Cans or Battles (about 1 Six Pack) 1 Can or Bottle 4 Cans or Bottles Half a Can or Bottle What is the greatest amount of beer that you've had at a single sitting since you came to MSU? 11 or More Cans or Bottles (about 2 Six Packs) 3 Cans or Bottles 8 to 10 Cans or Bottles (about 1/2 Six Packs) 2 Cans or Bottles 5 to 7 Cans or Bottles (about 1 Six Pack) 1 Can or Bottle 4 Cans or Bottles Half a Can or Bottle Have you had drink of wine since you came to MSU? Yes No (If "No", go on to Question 35). On how many days did you have a drink of wine? Days On how many of those days that you drank wine, did you drink wine at two or more different times during the day (for instance, in the afternoon, then again in the evening?) Days If you think about those times that you've drunk wine, about how many glasses of wine did you usually drink at any one sitting? (Wine glasses are usually 3 or 4 ounces.) 9 or More Classes 4 Classes 1 Glass 6 to 8 Classes 3 Classes Half a Glass 6 Classes 2 Classes What is the greatest amount of wine that you've had at a single sitting since you came to MSU? 9 or More Classes 4 Classes 1 Glass 6 to 8 Classes 3 Classes Half a Glass 5 Classes 2 Classes Have you had drink of liquor (distilled spirits) since you came to MSU? Yes No (If "No”, go on to Question 40). On how many days have you had a drink of liquor? Days On how many of those days that you drank liquor, did you drink liquor at two or more different times during the day (for instance, in the afternoon, then again in the evening?) Days 116 38. 40. 4f. 42. 43. If you think about those times you have drunk liquor, about how many drinks did you usually have at any one sitting? (A drink usually contains about 1/2 ounces of liquor, or one ”shot"). 8 or More Drinks 6 Drinks 2 Drinks 7 Drinks 4 Drinks 1 Drink 6 Drinks 3 Drinks Half a Drink What is the greatest amount of liquor that you've had at a single sitting since you came to MSU? 8 or More Drinks 5 Drinks 2 Drinks 7 Drinks 4 Drinks 1 Drink 6 Drinks 3 Drinks Half a Drink Since you came to MSU, about how many times have you: Pl 1 n n r. l —2 3-4 5—6 7—9 Never Times Times Times Times 0. Had five or more drinks at a single sitting, either of beer, wine, liquor, or some combination of these? b. Felt a little high or light headed after drinking? c. Gotten Drunk? d. Had a hangover (headache or nausea) morning after you had been drinking? 6. Had a drink first thing in the morning as you woke up? f. Continued your drinking over a period of several days? g. Had some difficulty in remembering what you had done while drinking the night before? How much of your drinking takes place on Thursday evenings? All of It Most of It Half of It Some of It None of It How much of your drinking takes place from Friday evening through Sunday night? All of It Most of It Half of It Some of It None of It How much of your drinking do you do when you are alone? All of It Most of It Half of It Some of It None of It 117 10— f 4 Times 15— 19 Times 20+ Times 44. 46. For each reason listed below, please check the column that shows how important that reason is for your own drinking: Very Pretty Not Too Not Important lmportont Important Important at All 0 . Makes get—togethers more fun. _ ___. __ .— b. Helps me get my mind off my problems. __ __ __ _ . Because you’re expected to drink at most social gatherings. n d. Adds to the pleasure of a good meal. _ _ __ __ e. It makes me feel less tense in social situations. .. _ _ __ __ —.~ . Just adds to the good feeling at a party. _ _ __ __ g. Helps me forget I'm not the kind of person I'd like to be. __ __ _ __ h. Because most people my age drink. __ _ _ _ i. When there are too many pressures on me. Since you came to MSU, has your drinking: Increased Decreased Stayed the Same The next questions focus on things that might have happened in relation to your drinking since you came to MSU. 2—3 4—5 SINCE YOU CAME TO MSU: Never Once Times Times a. How many times have you been criticized by your friends because of your drinking? b. How many times have you missed class because of your drinking? __ __ _ _ c. How often have you had difficulties with your roommate because of your drinking? d. How often have peOple told you that your drinking was creating problems at school? e. How often have you gotten into trouble with the police or been arrested because of something having to do with your drinking? 118 6 or More Times 2-3 4—5 6 or More SINCE YOU CAME TO MSU: Never Once Times Times Times 1. Have you had any accidents at school that may have been due to your drinking? g. How many times have you driven when you’ve had a good bit to drink? h. How many times have your friends, roommate or RA expressed concern about your drinking? i. How many times have you damaged public or private property that did not belong to you after you had been drinking? i. How often have you been loud or disorderly in a public place (including your residence hall) after you had a good bit to drink? k. How many times have you started a fight or a shaving match with a stranger after you had a good bit to drink? I. How often have you gotten into trouble with a member of the Residence Life staff or been documented ("written up”) because of something having to do with your drinking? If you think about your drinking since you came to MSU, has alcohol been a help or a problem for you? It's Been a Great Help It’s Been a Mild Problem It's Been Some Help It’s Been 0 Moderate Problem It's Been No Problem at All It’s Been a Serious Problem Has your doctor ever suggested that you cut down on your drinking or stop drinking altogether because it was having an adverse effect on your health? No Yes If "Yes", what was he/she concerned about? Have you stopped drinking alcohol altogether in the past two years? Yes No If "Yes", what were your main reasons for stopping? 119 APPENDD( B APPENDIX B INTRODUCTORY LETTER acnms APPROVAL FOR THIS project express: AU61619$ and must be renewed wlthin 11 months to continue. September 25, 1995 Student Name Address City State Dear Student: I am a doctoral student working to complete my degree in Higher Adult and Lifelong Education. I am conducting a study to complete my dissertation and I would like to request your assistance. Your name was randomly selected from those Michigan State University 19%-96 freslunen who live in a campus residence hall to participate in my study. In the past few years, there has been a lot of discussion about college student drinking on campus and I have chosen this topic for my study. I would like to ask you to complete the qinestiounaire when it arrives. It will take less than 10 minutes. Students who have completed this questionnaire have found it to be enjoyable as“ it gave them an opportunity to reflect on their alcohol use and relationships with the other students on their residence hall floor. I am hopeful you will also find it an~ opportmity to gain additional insight into your freshmen year experience. Within a week. you will receive a letter and questionnaire from me. Your response is exu'ernely important since only a small but representative percentage of the freshmen class members have been selected to participate in this study. I will greedy appreciate your completing the questionnaire and returning it promptly. ‘Ihank you. Sincerely, J. Ann Hower Graduate Student 120 APPENDIX C APPENDIX C FIRST LETTER OF TRANSMITTAL Omaha 2. 1995 Student Name Student Address . City State Dear Student: During the week of September 25th. you received a letter from me regarding a study I am conducting to complete my doctoral dissertation. You have been randomly selecmd from the MSU freshmen class 1995- 96 to participate in this study I am conducting regarding the relationship between the residence hall living environment and alcohol consumption patterns of freshmen students at MSU. Your participation is very important as I will analyze the results and use the analysis as the basis of my doctoral dissertation. Students who have completed this questionnaire have found it to be enjoyable as it gave them an opportunity to reflect on their alcohol use and relationships with the other students on their residence hall floor. I am hopeful you will also find it an opportunity to gain additional insight into your freshmen year experience. Iknow drisisaverybusytimeinthesemester.ifyouwilltakeone lSminutebreaktocompletethe enclosed survey and post card. it would be greatly appreciated!!! In order that the results will truly represent the views and experiences of the fieslunen student. it is important that you complete and retum the questionnaire. Your input is especially valuable as the accuracy of the results increases with the number of individuals who respond. You may be assured of complete anonymity. There is no way I can identify you from the questionnaire you return. Instead. I ask that you print your name and student number on the enclosed post card and mail it back separately so that I may remove your name from my mailing list. Only the questioruraire should be returned in the postage paid envelope. Mail the sealed envelope and post card by Friday. October 9,1995. If you retum mu survey 13y October 9, I995I mr return m card will a enteg into a Ewing [2r $00.00 gift certificate at the MSU bookstore! The winner will be notified m mail by October 16, 1&5. You indicate your voluntary agreement to participate by completing and returning this questiormaire. Participation in the survey is voluntary and there will be no penalty for not responding. All results will be treated with strict confidence and all respondents will remain anonymous in any report of research findings. You may receive a copy of the results of this research (with these restrictiom) “copy of the results" on the back of the return post card and printing your name and address below it. I would be most happy to answer any questions you might have. The telephone number is 339-4768. Thank you so much for your assistance. J. Ann Hower Graduate Student 121 APPENDIX D APPENDIX D SECOND LETTER OF TRANSMITTAL Ocrober 16. 1995 Student Name Student Address ‘ City State Dear Student: A couple of weeks ago. I wrote to you seeking your assistance with a survey regarding college student drinking at MSU. If you have already returned your complete questionnaire. please disregard this letter. Thank you very much! If you have not completed the questionnaire. may I urge you to do so? Your participation in the study is very important; every student has an unique insight to share. I am writing to you again because each questionnaire is important to the study. Irrorder for the results to be truly representative of freshmen students at MSU. it is essenu'al that each person in the sample return the questionnaire. As I mentioned in my last letter. .you may be assured of complete anonymity. If you return your guestignnaire by Qctober 23I 1995I ygur return 99% will i; cum 9' to a drawing for $50.00 gift certificate at the M89 mkstore! The m 341' be ggfi g by OCtober 30 1995. You indicate your voluntary agreement to participate by corrrpleting and returning this questionnaire. Participation in the survey is voluntary and there will be no penalty for not responding. All results will be treated with strict confidence and all respondentswill remain anonymous in any report of research findings. You may receive a copy of the results of this research (with these restrictions) "copy of the results" on the back of the return post card and printing your name and address below it. In the event that your questionnaire was misplaced. a replacement is enclosed. Once again. please return the questionnaire in the postage paid envelope. Please return the postcard separately. If you have any questions about the purpose of the survey. please contact me 339-4768. Please mail the complete questionnaire and postcard by October 22, 1%; Thank you very much. Sincerely. J. Ann Hower Graduate Student 122 APPENDIX E APPENDIX E THIRD LETTER OF TRANSMITTAL UCRIHS APPROVAL FOR THIS project EXPIRES: AU8161996 and must be . OCtober 30. I995 - 11 months'gngomgthrn Dear name (hand written): In the middle of September. I wrote to you about a Study I am conducting to complete my dissertation on college student drinking at MSU. I need your questionnaire! Enclosed is another questionnaire, postage paid post card and return envelope. Please complete 8 the survey and mail it TODAY! Your assistance would be greedy appreciated. Sincerely, J. A. flower Graduate Student 123 APPENDIX F APPENDIX F RETURN POSTCARD FRONT: LA. Hower First Class 5967 Village Dr. Postage Stamp Haslett. MI 48840 Student Name Address City State Zip Back: October 25. 1995 It’s not too late ........ to retum your questionnaire! If you have already completed and returned you survey, please accept my sincere thanks! If not. please do so today. YOUR response is extremely important to the outcome of this study. COMPLETE THIS SURVEY FOR A CHANCE TO WIN A $25 .00 GIFT CERTIFICATE TO THE MSU BOOKSTORE!!! _ You indicate your voluntary agreement to paru'cipate by completing and returning this questionnaire. Participation in the survey is voluntary and there will be no penalty for not responding. All results will be treated with strict confidence and all respondents will remain anonymous in any report of research findings. You may receive a copy of the results of this research (with these restrictions) "'copy of the results" on the back of the return post card and printing your name and address below it. If by some chance you did not receive a questionnaire, or it has been misplace, please call me at 339-4768 and I will gladly send you a new questionnaire today. Sincerely. J. Ann Hower Graduate Student 124 APPENDIX G APPENDIX G LETTER OF APPROVAL FOR STUDY MICHIGAN STATE U N r v E R s I T Y August 16, 1995 i To: Julia A. flower 5967 Villa e Dr. Haslett, M 48840 RE: IRBI: 95-433 TITLE: A STUDY OF THE RELATIONSHIP BETWEEN COLLEGE STUDENT'S PERCEPTION or SELECTED VARIABLES WITHIN THE RESIDENCE HALL ENVIRONMENT AND COLLEGE STUDENT DRINKING BEHAVIOR REVISION REQUESTED: N/A CATEGORY: l-C APPROVAL DATE: 08/16/95 The University Committee on Research Involving Human Sub octs'(UCRIHS) review of this project is complete. I am pleased to adv so that the rights and welfare of the human subjects appear to be adequately protected and methods to obtain informed consent are appropriate. gerefore, the UCRIHS approved this project and any rCVLsions listed a ove. . RENEWAL: UCRIHS approval is valid for one calendar year, beginning with the approval date shown above. Investigators planning to continue a project beyond one year must use the green renewal form (enclosed with t a original agproval letter or when a project is renewed) to seek u ate certification. There is a maximum of four such expedite renewals ssible. Investigators wishing to continue a project beyond the time need to submit it again or complete reView. REVISIONS: UCRIHS must review any Changes in procedures involving human subjects. rior to initiation of t 9 change. If this is done at the time o renewal, please use the green renewal form. To revise an approved protocol at any other time during the year send your written request to the CRIHS Chair, requesting revised approval and referEnCing the project's IRB I and title. Include in your request a description of the change and any revised instruments, consent forms or advertisements that are applicable. PRoaLENS/ WTWEOF CHANGES: Should either of the followin arise during the course of the RESEARCH work, investigators must noti UCRIHS promptly: (1) problems (unexpected Side effects. comp aints, etc.) involVing uman AND Subjects or (2) changes in the research environment or new GRADUATE information indicating greater risk to the human sub'ects than existed when the protocol was previously reviewed an approved. STUDHES Universityumrnittasoa If we can be of any future helg, lease do not hesitate to contact us llasaarehlavalvlsa at (517)355-2180 or FAX (Sl7)4 2- 171. Mmunfiflnds (WINS) Sincerely, , r [K ‘. Mmmnmemwnw - /<:\S; K \ L... Zhumwmwmammm ‘ K ‘ ' , .4.- “ a V W .\ mm'm'mm'w David E Wr' ht Ph D i " . . i , . . “52““? UCRII-IS Chairg .' \ “”355 2180 DEW: kaa/ le --’ FAX; 517/432-1171 CC: Kathryn M. Moore 1225 APPENDIX H APPENDIX H YOUNG ADULT SURVEY REQUEST J. Ann Hower 6260 Gossard Avenue Eastlsansing Michigan 48823 (517) 3363435 July 20. 1994 Dr. Jessor, Director Institute of Behavioral Science Campus Box 483 University of Colorado, Boulder Boulder, Colorado 80309-0483 Dear Dr. Jessor, I recently wrote to you requesting one of your instruments to be used in my doctoral research. 1 received a response from your office requesting another letter with the signature of the person supervising my research. I have asked my Dissertation Advisor, Dr. Kathryn Moore. to sign this letter. ” As I wrote to you previously, I am currently working on my dissertation in higher education administration at Michigan State University. My research topic is college student alcohol use and I am considering using your Problem Behavior Theory as the theoretical framework. I am interested in researching the impact, if any, of certain aspects of the university environment on student alcohol use and problem behavior, i.e. alcohol use of roommates, alcohol policies and enforcement of policies, educational programming, Resident Assistant alcohol use and attitudes towards use, etc. The role of the perceived university environment upon student alcohol use and behavior is of particular interest to me. I am very intereSted in using one of your instruments in my research although I recognize that I may have to modify it to meet the needs of my particular study. While I have read a great deal about your theory I have not seen an actual instrument and 1 am unsure of how many instruments you may have or their titles. I am hoping that from my brief description of my research topic that you will know which, if any, of your insrruments are appropriate. Any suggestions you can provide would be very much appreciated. I will, of course, pay for any costs associated with the instruments. S incerel y , J. Ann Hower . r. CRamflM. Moore DOCtoral Student irper Michigan State University Department of Educational Administration Michigan State University 126 APPENDIX I APPENDIX I Questionnaire - Variable Codes SURVEY QUESTION VARIABLE NAME ACRONYM SCORING SCALE RANGE SCALE NAME l STUDENTS WHO DRINK STDR 16-0 RESIDENCE HALL MODELS FOR DRINKING SCALE- RHM STUDENT ON FLOOR WHO DRINK STFLDR 5-1 RHM FRIEND HANG DRINK FRHGDR 5-1 CLOSE FRIENDS DRINK CLFRDR 5-1 FRIENDS APPROVAL- DIS SAPROVAL FOR DRINKING FRAPPDR FRIENDS MODELS FOR ABSTENTION FRMODAB FRIENDS MODELS FOR DRINKING FRMODDR 9-2 SOCIAL SUPPORT FOR DRINKING SCALE FRIENDS MODELS FOR QUANTITY DRUNK FRMODQDR 1-4 13-3 FRIENDS DRINKING PRESSURE FRPRDR 1-4 10 FRIENDS CONTROL FRCONT FRIENDS CONTROL ll FRIENDS CONTROL FRCONT 4-0 FRIENDS CONTROL 12 ROOMMATE APPROVAL- DISSAPPROVAL FOR DRINKING RMAPPDR 3-1 l3 ROOMMATE MODEL FOR ABSTENTION RMMODAB 0=1 l=2 127 QUESTION VARIABLE NAME ACRONYM SCORING SCALE RANGE SCALE NAME l4 ROOWATE MODEL QUANTITY DRUNK RMMODQDR 1-4 RMSSDR 15 ROOMMATE DRINKING PRESSURE RMDRPR 4-1 l6 ROOMMATE CONTROL RMCONT 8-2 RMCONT l7 ROOMMATE CONTROL RMCONT RMCONT 18 RA CONTROL RACONT 4-1 RACONT 19 RA CONTROL RACONT 4-1 8-2 RACONT 20 RA APPROVAL- DI SSAPPROVAL FOR DRINKING RAAPP DR 3-1 3-1 21 RA ENFORCEMENT ALCOHOL POLICY 3-1 22 ABSTAINER- DRINKING STATUS ABDRST 0=l l=2 ABDRST 23 CURRENT DRINKER STATUS and RECENCY OF LAST DRINK CURDRST 24 ABSTINER DRINKER STATUS CURDRST ABDRST 25 BEER SINCE MSU BEER 0=l l=2 26 FREQ DRINKING BEER SINCE MSU FRBR VALUE VALUE FRBR 27 FREQ DRINKING BEER SINCE MSU FRBR VALUE VALUE FRBR 28 AVERAGE INTAKE OF BEER PER SITTING QBR ll-O.5 ll-O.5 128 (NJESTTCRJ # \hARJNAde AKZR(HNYTU SCXJRJDKS SCHULE RADUGE SCIJJEPIALJE 29 ‘NLABHLJLDA HTTAJCE<1F IBEIHIFTH! SITTTFK3 NLAXIH! 143 1-8 30 ‘RHFEESHWCE hdSII 0=l l=2 31 FREQ DRINKING WINDE SINCE MSU FFHNI \hALlHE \hAIlflE FRNVI 32 PEEK) IDRJFHCHQCI ‘RflhflZSHRCIE LASLJ FRNVI \hALlHE VTUJJE FRNVI 33 AUJEIUACHE IDWT)JCE(JF ‘WflIflEPTfl! SFTTTFKE (ERR 1-8 1-8 34 NLAXHLJURA HTTAJCE