ABSTRACT MULTIVARIATE PREDICTION OF PSYCHOACTIVE DRUG USE: A SOCIAL-PSYCHOLOGICAL APPROACH By Victor Anthony Battistich The purpose of the present research was to develop and test a comprehensive theoretical system which was capable of accounting for individual differences in the use of psychoac- tive drugs. Two basic assumptions guided this undertaking. First, it was assumed that psychoactive drug use was best con- ceptualized as a deviant social behavior. This assumption fo— cused attention on those factors which may conceivably serve to inhibit or counteract the influence of various forces of social control, as well as those interpersonal influences which may serve to initiate and maintain an individual's use of drugs. Second, it was assumed that neither sociocultural or intraper- sonal variables were capable, individually, of accounting for psychoactive drug use. Rather, what was needed was a multivari- ate framework which utilized both sociocultural and personality variables as part of a social-psychological system. Rotter's (1954) social learning theory of personality, as modified and extended by Richard Jessor and his colleagues (Jessor, et al, 1968, 1973) in their theory of deviant behavior, was selected as an initial framework on which the theoretical system could be constructed. The final theoretical system contained three sociocultural and two personality variables. Specifically, the theoretical framework focused on an individual's access to opportunities for attaining valued goals, the extent of nor- mative constraints against engaging in deviant behavior, oppor— tunities for observing and interacting with drug users, atti— tudinal tolerance of deviant behavior, and expectations about the consequences of drug use. Since it was desirable to obtain subjects who differed as much as possible in terms of drug use and related characteris- tics, two separate samples were obtained: a "College Sample" (a = 51) drawn from individuals utilizing the services of the Counseling Center at a large Midwestern university; and a "Street Sample" (n = 72) composed of individuals utilizing the services of a local Drug Education Center. Although on an a priori basis these two groups were considered to represent different popu- lations, examination of the relevant information revealed that the two samples did not differ significantly in terms of drug use or the majority of sociodemographic characteristics. Con- sequently, the two samples were combined for the major statis- tical analyses. Data was collected by means of a comprehensive, anonymous questionnaire assessing a variety of drug use, sociodemographic, interpersonal, and personality variables. On the basis of this information, measures of drug use and the sociocultural and per- sonality variables included in the theoretical system were con— structed. Three separate measures of drug use were utilized as depen— dent variables. The first consisted of a sample trichotomous classification of user, past user, or nonuser for each indivi— dual on each of the psychoactive substances examined. The sec- ond measure consisted of a frequency-variability index of use for each of the substances. Finally, the third measure of drug use was a measure of the individual's overall pattern of drug use. Frequency—variability scores for each of the psychoactive substances were cluster analyzed, revealing two discrete clus- ters or groups of substances. The first cluster was composed of marijuana, hallucinogens, barbiturates, amphetamines, cocaine, opiates, and nitrous oxide, while the second cluster contained only alcohol and tranquilizers. The distribution of subjects' "cluster scores" on each of these clusters was then dichoto- mized at the median, forming "high" and "low" groups for each drug cluster. These groups were then combined to yield four separate "patterns" of drug use: high use of drugs in both of the clusters, low use of drugs in both of the clusters, or high use of drugs in one cluster but low use of drugs in the other. The independent variables were eight measures of sociocul- tural and personality characteristics. Two measures of access to opportunities for valued goal attainment were constructed: a measure of Socioeconomic Status based on Hollingshead and Redlich's (1958) two-factor index of social position; and a measure of Objective Access based on the respondent's age, race, marital status, and religious background. Normative constraints against deviance was assessed through a measure of Involvement with Prosocial Groups, based on the extent of respondents' par— ticipation in conventional religious groups. The third socio— cultural variable, opportunities for observing and engaging in drug use, was measured through three separate indices: a mea— sure based on the respondent's age, marital status, and the size of the home community (Opportunities for Deviance 1: Sociodemo— graphic Factors): a measure of the percentage of the respondent's friends who used drugs (Opportunities for Deviance g: Perceived Drug y§§_in the Social Environment): and a measure of interac- tion with the respondent's five closest friends (Opportunities for Deviance 3: Extent of Use with Friends). The two person- ality variables were assessed by means of a scale of Positive Expectations for Drug Use based on respondents' subjective pro- bability of future drug use, the percentage of pleasant experi- ences with drugs, and perceived physical and psychological danger in use of the various drugs: and a Tolerance g: Deviance Scale developed by Jessor, Graves, Hanson, and Jessor (1968). The relationship of each of the sociocultural and person— ality variables to both use-nonuse and frequency-variability of use of each of the psychoactive drugs was examined through a series of univariate analyses. In addition, the efficacy of the system of variables in accounting for overall patterns of drug use was examined through a multiple discriminant analysis. In general, the results of these two sets of analyses were com- plementary and supported the hypotheses. For use-nonuse, frequency-variability of use, and overall patterns of drug use, the more extensively an individual was involved with drugs: (1) the less his/her participation in prosocial groups: (2) the more opportunities for observing and engaging in drug use s/he perceived in the immediate social environment: (3) the greater the amount of time spent using drugs with his/her closest friends; (4) the more tolerant of deviant behavior in general s/he was; and (5) the more positive were his/her expectations concerning the consequences of drug use. Of the eight measures of vari- ables in the theoretical system, only the two measures of ac— cess to opportunities for valued goal attainment failed to show any consistent, significant relationship to any of the measures of drug use. The overall pattern of results is consistent with the find— ings of previous studies in demonstrating the importance of peer influences in accounting for psychoactive drug use (Kandel, 1973. 1974; Jessor, g1 g1, 1968, 1973; Sadava, 1973). Of the six measures of sociocultural variables, the best predictors of drug use were measures of interpersonal influences (i.e., perceived drug use among friends and the amount of time spent using drugs with one's closest friends). Thus, these findings support the hypothesis of involvement in a drug-using "subcul- ture" which serves to initiate and maintain the individual's use of drugs (Goode, 1969: Kandel, 1973. 1974). However, the single most predictive variable in the present study was a per— sonality variable: Positive Expectations for Drug Use. In addition, both sociocultural and personality variables were needed to maximally differentiate the four patterns of drug use. These findings thus support the assumption that both socio— cultural and personality variables are needed to adequately ac- count for an individual's use of psychoactive drugs. Finally, a number of modifications and additions to the theoretical system were discussed, including the addition of measures of socialization practices which theoretically mediate between the individual personality and the sociocultural en— vironment. In addition, it was suggested that future research in the form of a longitudinal study would prove invaluable for understanding the complex processes underlying psychoactive drug use. MULTIVARIATE PREDICTION OF PSYCHOACTIVE DRUG USE: A SOCIAL-PSYCHOLOGICAL APPROACH By Victor Anthony Battistich A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Psychology 1976 ACKNOWLEDGMENTS First and foremost, I would like to express my gratitude to Dr. Robert A. Zucker, chairperson of my thesis committee, for his friendship and his continuing guidance and support in this work as well as in my professional development as a psy- chologist. I would also like to thank Dr. Lawrence Messe and Dr. Charles Johnson for their contributions as members of my committee. For their invaluable help in data collection and/or analysis, I would like to thank Dr. Ralph Kron, Dr. Ralph Levine, Dr. Gordon Williams, Bryan Coyle, and Scott, Pete, and Elliot. at the DEC. Finally, I would like to thank my friends, Dr. Michael Dowdle and Martha Montgomery for their love and support. Without their help none of this would have been possible. ii TABLE OF CONTENTS List of Tables List of Figures Introduction . Literature Review Patterns of Drug Use Correlates of Drug Use Demographic Factors Interpersonal Factors Personality Factors Theoretical Perspective and Purpose of the Study Statement of the Problem Rotter's Social Learning Theory The Sociocultural System The Personality System Purpose of the Study and Hypotheses Method Instrument A Note on the Validity of Self- -Report Measures . . . . . . . . . Subjects Procedure The Measurement of Drug Use The Measurement of the Sociocultural System . . . . . . . . . . . . iii Page vii 12 12 14 17 20 20 24 27 34 39 48 48 48 49 57 63 66 The Measurement of the Personality System . The Measurement of Social Sanctions Against Drug Use Analyses Results and Discussion . Drug Use Characteristics of the Two Samples Univariate Analyses The Sociocultural System and Drug Use The Personality System and Drug Use Multivariate Analyses The Patterns of Psychoactive Drug Use The Theoretical Framework and Patterns of Drug Use General Discussion and Appraisal of the Theoretical Framework . . . . . . . . Footnotes References Appendices A. The Questionnaire . . . . . B. Cover Letter . C. Demographic Characteristics of "Incentive" and "No-Incentive" Subjects in the "Street Sample" . . . D. Statement of Procedures For Insuring the Confidentiality of Respondents . . . E. Statistical Procedures for Dealing with Heterogeneity of Variance F. Sex Differences in Drug Use iv Page 75 80 85 87 87 94 94 127 134 135 147 159 173 176 182 182 197 198 201 202 205 Table U1 C‘w N 10. 11. 12. 13. 14. 15. 16. 17- 18. LIST OF TABLES Demographic Characteristics of the Two Samples Sex Differences in Demographic Characteristics Socioeconomic Status Involvement with Prosocial Groups Opportunities for Deviance 1: Sociodemographic Factors . . . . . . . . . . . . . . . . . . Opportunities for Deviance 2: Perceived Drug Use in the Social Environment Opportunities for Deviance 3: Extent of Use with Friends . . . . . . . . . . . . . . . . . . Positive Expectations for Drug Use Intercorrelations of Variables in the Theoretical Framework I I I I I I I I I I I I I I I I I I Social Disaproval and Legal Penalties Associated with Drug Use . . . . . . . . . . . . . . . Drug Use—Nonuse Characteristics of the Two Samples . . . . . Frequency—Variability of Drug Use Characteristics of the Two Samples . . . . . . . . . . Access to Opportunities and Drug Use—Nonuse Access to Opportunities and Frequency-Variability of Drug Use . . . . . . . . . . . . . . . . . . Participation in Prosocial Groups and Drug Use— Nonuse . . . . . Participation in Prosocial Groups and Frequency- Variability of Drug Use . . . . . . . . . . . Opportunity for Deviance and Drug Use-Nonuse Opportunity for Deviance and Frequency—Variability of Drug Use . . . V Page 52 58 68 7O 72 74 75 78 79 82 88 91 99 101 103 106 108 112 Table Page 19. Interaction with Closest Friends and Drug Use- Nonuse . . . . . . . . . . . . . . . . . . . . . . 115 20. Interaction with Closest Friends and Frequency- Variability of Drug Use . . . . . . . . . . . . . 122 21. Tolerance of Deviance and Drug Use-Nonuse . . . . 128 22. Tolerance of Deviance and Frequency—Variability of Drug Use . . . . . . . . . . . . . . . . . . . . . 129 23. Positive Ekpectations for Drug Use and Use- Nonuse . . . . . . . . . . . . . . . . . . . . . . 131 24. Positive Expectations for Drug Use and Frequency- Variability of Use . . . . . . . . . . . . . . . . 133 25. Drug Use Intercorrelations . . . . . . . . . . . . 136 26. Cluster Analysis of Drug Use Frequency—Variability Scores . . . . . . . . . . . . . . . . . . . . . . 136 27. Discriminant Analysis of Group Differences in Frequency—Variability of Drug Use . . . . . . . . 139 28. Differences in Drug Use-Nonuse Between Groups . . 142 29. Accuracy of Prediction of Group Membership From Drug Use Discriminant Functions . . . . . . . . . 146 30. Discriminant Analysis of Group Differences in Sociocultural and Personality Characteristics . . 149 31. Sheffe A Posteriori Comparisons of Group Differ- ences in Sociocultural and Personality Charac— teristics . . . . . . . . . . . . . . . . . . . . 153 32. Accuracy of Prediction of Group Membership From Sociocultural-Personality Discriminant Function . 155 33. Group Differences in Interaction with Closest Friends . . . . . . . . . . . . . . . . . . . . . 156 34. Demographic Characteristics of "Incentive" and "No-Incentive" subjects in the "Street Sample" . . 198 35. Sex Differences in Drug Use . . . . . . . . . . . 205 vi Figure LIST OF FIGURES Sociocultural and Personality Influences on Drug Use . . . Normative Violation and Extent of Drug Use Social Disapproval and Extent of Drug Use Location of Group Centroids in Discriminant Space vii Page 38 96 97 145 INTRODUCTION AND OVERVIEW Although there is ample evidence of the use of psychoac— 1 drugs by a significant percentage of the population (Na- tive tional Commission on Marihuana and Drug Abuse, 1973: O'Donnell, Voss, Clayton, Slatin, & Room, 1976), theoretical conceptuali- zations of the antecedents of drug use remain remarkably incom- plete. This is especially true in regard to accounting for differences in type or pattern of drug use among various groups of users. Thus, while elaborate theories of the dynamics of drug use have been developed for individual drugs such as alco- hol (e.g., Jessor, Graves, Hanson, & Jessor, 1968; McClelland, 1972; Wilsnack, 1973), marijuana (e.g., Jessor, Jessor, & Finney, 1973; Johnson, 1973; Sadava, 1973), opiates and hallucinogens (e.g., Akers, 1973). these theoretical perspectives have yet to be applied to use of the entire range of psychoactive drugs. Consequently, and despite the tremendous amount of research con— ducted in the area, an adequate comprehensive theory of drug use has not yet been fully developed. It is toward the develop— ment of such a theoretical perspective that the present research is directed. Early conceptualizations tended to take an overly simplis- tic view of drug use. The use of any psychoactive drug or drugs was seen as related to particular personality or sociocultural characteristics of the users, and individual differences in type or pattern of drug use were ignored (Johnson, 1973: Russell, 1972). Although studies attempting to contrast this "universal drug user" with individuals not using psychoactive drugs have reported some general differences (Cohen & Klein, 1970, 1972; Kohn & Mercer, 1971), they have commonly failed to show any adequate differentiation. Recent investigators have argued for a more detailed analy— sis of drug use (Goldstein, Gleason, & Korn, 1975). Individuals may differ in such factors as the type of drug used (e.g., alco— hol vs. heroin), number of drugs used (e.g., single vs. multi— ple drug use), as well as in the frequency of drug use (e.g., occasional vs. regular). Given these variations in use, it would seem probable that individuals conforming to different types of use would also differ along other dimensions. Such factors as individual reactions to various drugs and motiva- tions for use, the "popularity" of particular drugs among dif— ferent groups of users, and certain socio-psychological and demOgraphic variables may all relate to individual differences in drug use. Consequently, any conceptualization of drug use which equates all types of use is unlikely to provide an ade- quate explanation of the phenomenon. Before proceeding further it is necessary to define a num- ber of terms. In the discussion which follows, the term "drug use" is used to refer to self—administered, non—prescribed use of psychoactive substances. Likewise, the term "drug user" refers only to individuals engaging in such use. Medically supervised use of any substance and self-administered use of such substances as caffeine and tobacco is therefore not con- sidered "drug use" in the present context. In addition, the terms "type" or "pattern" of drug use are used to refer to dif— ferences in both: (1) the particular psychoactive drugs used, and (2) the frequency and variability with which they are used. Thus, individuals will be said to differ in their type or pat- tern of drug use if they either use different substances, or use the same substance(s) but differ in their frequency of use. Within the context of this research, the use of psychoac— tive drugs is viewed as a behavior resulting from the complex interaction of various personality, interpersonal, and socio— cultural variables. It should also be noted here that the use of psychoactive drugs is seen as predominantly a social behavior. In a review of the majority of relevant studies conducted prior to 1971, the National Commission on Marihuana and Drug Abuse (1973) found that a substantial number of drug users reported that their initiation into drug use resulted from a need for social approval, enhanced sociability, a better understanding of friends, and/or a shared experience with significant others. Across all studies, the reported reasons for drug use relating to social experiences generally outweighed personal reasons. In addition, over 75% of all high school and college users re- ported that they were introduced to drug use by a friend, and that they rarely used drugs alone. The commission concluded that drug use is an experience which is perceived by the user as an exciting and pleasurable social activity.2 Drug use is also a deviant behavior. That is, with the exception of tobacco and caffeine use, moderate use of alcohol, and medically sanctioned use of certain stimulants, depressants, and narcotics, the use of psychoactive drugs violates both so- cial and legal sanctions. Consequently, even where use con- forms to the norms of a particular social group, it is still viewed as a deviant behavior by the larger society, and has at least the potential for incurring negative consequences to the user (e.g., arrest and imprisonment). The problem, then, for any comprehensive theory of psycho- active drug use is the identification of those variables and processes in the person and the sociocultural situation which operate together to make the probability of drug use higher than that of nonuse. These variables and processes should ac- count for the circumstances under which an individual will use drugs, differential drug use by persons in the same situation, and the distribution of drug use in different sociocultural locations. That is, an adequate theory of drug use should not only account for whether or not an individual will use drugs, but his/her particular pattern of drug use as well. One theory which holds promise in this regard is Rotter's (1954) social learning theory of personality. This is the gen- eral theoretical perspective adopted in the present research, and will be described in greater detail in a subsequent chapter. At this point, it suffices to say that from this perspective behavior is viewed as the outcome of a choice, selection, or decision process in which alternative behaviors are "sorted" to determine which has the highest probability of maximizing consequences desired by the actor in a particular situation. To account for any behavior therefore requires knowledge of four interrelated factors: (1) that the behavior has been learned by the actor and is available in his/her repertoire; (2) the expectations held by the actor that the behavior will lead to certain outcomes: (3) the value placed on these out— comes: and (4) the outcomes perceived by the actor as poten- tially available in the particular situation. From this per- spective, then, drug use will occur when it is perceived as having a higher likelihood of maximizing valued goal attain- ment than nonuse. This theoretical perspective is seen as having a number of distinct advantages when applied to psychoactive drug use. First, since all learned behavior is viewed as goal directed, drug use and nonuse are theoretically homogeneous in that the same principles should account for both. Second, the theory is sufficiently general in nature to enable it to account for such a complex behavior as drug use. Previous theories of drug use have generally focused on one set of determinants to the exclusion of others. For example, psycholOgical theories have tended to concentrate on certain "basic personality traits" of the user (e.g., oral dependency needs), while sociological the— ories have focused on sociocultural determinants (e.g., socio- economic status). Neither of these perspectives have been able to provide an adequate explanation of drug use. The psycholOgi- cal approach fails to account for the fact that similar constel- lations of personality traits may result in different behaviors under different circumstances, or for the location and distri- bution in society of these personality attributes. 0n the other hand, the sociolOgical approach provides little understanding of the structures and processes which mediate between a state of society and the occurrence of behavior, and fails to account for individual differences in the relationship between a socio— cultural state and behavior. As Yinger (1963) has emphasized, every behavior is both personal and situational. In Rotter's theory, personality is conceived of as a §y_- Ififla That is, "personality" refers to the organization of rela— tively enduring psychological structures of the person. The interaction of these elements results in tendencies to respond in certain ways. However, personality does not directly deter- mine behavior. Rather, the response is viewed as a joint func- tion of the personality and sociocultural systems. Behavior reflects both personality-determined dispositions and socio— culturally-determined characteristics of the situation in which the response occurs. By considering both of these sets of de- terminants, the theory promises to provide a more comprehensive explanation of the phenomenon. LITERATURE REVIEW The review of the literature which follows is neither ex- haustive nor representative of the existing empirical and theo- retical literature on psychoactive drug use. Instead, only those findings particularly relevant to the present research are reviewed. These findings may be organized into two broad categories: (1) those concerning the patterns or types of drug use: and (2) those studies reporting factors which differentiate users from nonusers. Since the issue of drug use patterns has received relatively little attention in the literature, the studies reviewed in this section represent, to the best of the author's knowledge, all of the reported findings in this area. In regard to the correlates of drug use, an attempt was made to selectively review the literature. The studies reviewed in this section are representative of the major findings in the area. Studies reporting findings which were either redundant to those presented or contributed little additional relevant information are not reviewed. Patterns of Drungse A number of studies report either correlations between the use of several drugs (Blum, 1969; Johnston, 1973: O'Donnel, g: a1, 1976), or the extent of use of several drugs in relation to use of a given drug——predominantly marijuana (Goode, 1969: Johnson, 1973; National Commission on Marihuana and Drug Abuse. 1973)- 7 In general, these studies indicate that use of a given drug or drug class is significantly and positively related to use of all other psychoactive drugs. Thus, for example, regular mari- juana users are highly likely to have used all other psychoac— tive drugs (Blum, 1969: Johnson, 1973). However, these studies do not generally reveal detailed patterns of use for the indi— vidual. That is, they do not reveal either the specific drugs a given individual is likely to use on a regular basis, or the sequence in which he is likely to use them. In an intensive study of student drug-use, Blum (1969) col- lected data from 1,314 drug users attending four colleges and one junior college in the Western United States. A factor anal— ysis of the users' lifetime drug-profile scores revealed four clear factors which accounted for 66.8% of the variance in drug use. Marijuana and hallucinogens loaded highly on Factor I, sedatives and tranquilizers on Factor II, alcohol and tobacco on Factor III, and amphetamines, opiates, and "special substan— ces" (i.e., glue, nitrous oxide, etc.) on Factor IV. Blum con- sidered Factor II to represent a "distress and anxiety dimini- shing" factor, and Factor III a "conventional social-drug use" cluster. Since amphetamines correlated .33 with marijuana, Fac— tors I and IV were considered to be substructures, with Factor I possibly representing "psychedelic enthusiasts" or "drug experi— menters," and Factor IV a "drug immersion factor." While these groupings appear to have a certain logical validity, the extent to which they correspond to actual patterns of use remains unclear. In a survey of over 2,000 Canadian secon- dary school students, Russell (1972) described three patterns of use which accounted for 75% of all drug users: 36% reported marijuana use only; 28% marijuana plus LSD; and 10% marijuana, LSD, and methedrine. 0n the other hand, Freedman and Brotman (1969) reported only two dominant drug—use patterns among a sample of urban, upper-middle class New York high school stu— dents——marijuana only and marijuana plus amphetamines; while Goode (1969) has reported that 88% of all drug use among two samples of New York university students was marijuana only. Thus, with the exception of Russell's marijuana plus LSD usage pattern, none of the reported patterns correspond to Blum's fac— tors. Rather, the two dominant patterns appear to be marijuana use only, and marijuana plus "other substances." The variance in these findings may relate to a number of factors. First, the samples differed in such demographic char- acteristics as age, socioeconomic status, and geographical area of residence. Each of these factors has been shown to be sig- nificantly related to differences in drug use (Blum, 1969: National Commission on Marihuana and Drug Abuse, 1973: Berg, 1970). Second, although there is ample evidence that most users of illicit psychoactive drugs also use alcohol (Blum, 1969: Bogg, Smith, & Russell, 1969; National Commission on Marihuana and Drug Abuse, 1973), only Blum's study included alcohol use in the reported findings. Conversely, many alcohol users un- doubtedly do not use marijuana or other illicit drugs. For example, while only 15% of a national household sample of ado- lescents had used marijuana, 50% had used alcohol outside of a family setting (Josepheson, Haberman, Zanes, & Elinson, 1972). 10 Failure to consider the entire range of psychoactive drugs un- doubtedly alters any reported findings to an appreciable extent. Recent findings concerning the sequence of drug use sug— gest a somewhat different approach to distinguishing various patterns of use. Goldstein, Gleason, and Korn (1975) made both longitudinal and cross—sectional comparisons of drug use among over 3,000 students attending Carnegie-Mellon University. The majority of the sample were males from urban and suburban areas of the middle Atlantic states. Their analysis indicated that users most frequently began use with alcohol and tobacco, and prOgressed to marijuana, depressants, hallucinogens, and nar- cotics, in that order. This hierarchical ordering was suffi- ciently invariable that use of a given drug indicated far be- yond chance that an individual had used all preceding drugs. In addition, as one ascends the continuum from alcohol to nar- cotics, successive drugs are used by fewer individuals (BOgg, Smith, & Russell, 1969; Hager, Vener, & Stewart, 1971: Wolfson, Lavenhar, Blum, Quinones, Einstein, & Louria, 1972). These findings seem to suggest that a primary differenti- ating factor between various patterns of use is the degree to which the user has ascended the "drug-use continuum." Thus, the largest type of use would be expected to be alcohol only, followed by an alcohol and marijuana type, etc (cf. O'Donnell, gt g1, 1976). However, Johnson's (1973) study of marijuana use among a stratified sample of college students indicates that, beyond marijuana, patterns of drug use do not invariably conform to this hierarchical progression. He found evidence for two 11 distinct "subcultures" of drug use separated along racial lines. Black users were about twice as likely as whites to have used cocaine and heroin; while white users were about twice as likely to have used amphetamines, sedatives, and hallucinogens as blacks. Thus, while both groups used marijuana, black users tended to "bypass" drugs in the middle of the continuum and progress directly to narcotics. Although the sequence of drug use within any single group may be linear, it appears that the specific drugs utilized differs among various user groups. Studies of drug use among non—students also report vary- ing patterns of use for different groups. Based on observations made as part of an ethnographic field study of the Haight-Ashbury "hippie" community in San Francisco, Davis and Munoz (1968) des- cribed three types of use. "Heads" were characterized by their regular use of LSD. This group was composed of individuals of both sexes, in their middle-to-late twenties, and predominantly from middle and upper-middle socioeconomic class backgrounds. "Freaks," on the other hand, were characteristically males from working class backgrounds who rarely used LSD, but used methe- drine extensively. The third group was composed of "mixed drug users" who regularly used a wide spectrum of drugs, including both LSD and methedrine. A more recent study (Smith & Gay, 1972) has indicated the existence of a fourth group in the com— munity consisting of regular heroin users. While these studies provide ample evidence for the exis— tence of different types of psychoactive drug use, the exact nature of these patterns remains unclear. Differences in such critical factors as the number or range of drugs considered, as 12 well as the age, sex, and other demographic characteristics of the samples, makes any comparison of these studies difficult. In addition, the frequency of drug use was rarely examined. Individuals not only differ in terms of the particular drugs they use, but in the regularity with which they use them. Thus, two individuals may both use the same drugs, but differ exten- sively in their frequency of use. Disparities such as these may well relate to much of the inconsistency in previous find— ings. Correlates of Drug Use As noted earlier, individual variations in type or pat- tern of drug use are undoubtedly related to differences in other personal characteristics of the users. Unfortunately, with the exception of the Johnson (1973) and Davis and Munoz (1968) stu— dies, the relationship between patterns of drug use and other characteristics of the individual has infrequently been exam- ined. However, a large number of studies have examined the re- lationship between various personality, social, and/or demogra- phic variables and either general drug use or the use of a spe- cific drug—-typically marijuana or LSD. For the purposes of the present discussion, these findings can be conveniently grouped into three categories: (1) demOgraphic, (2) interpersonal or social, and (3) personality factors which have been shown to differentiate drug users from nonusers. Demographic Factors. The use of psychoactive drugs has been found to be positively related to socioeconomic background (Berg, 1970; Haagen, 1970: Leahy, gt gt, 1972), urban as compared 13 to rural environment (Bogg, gt gt, 1969: Blum, 1969; Josephson, gt gt, 1972), a Protestant or Jewish religious background (Blum, 1969; National Commission on Marihuana and Drug Abuse, 1973), a lack of participation in conventional religious groups and activities (Jessor, gt gt, 1973: Lavenhar, gt gt, 1972: Blum, 1969), and parental drug use (Lavenhar, gt gt, 1972: Blum, 1969: Kandel, 1973, 1974). However, few of these factors have been shown to exhibit any differential relation to the various types or patterns of drug use. A large number of studies have found drug use to be higher among males than females (Johnson, 1973: Hager, Vener, & Stewart, 1971: Leahy, Steffenhagen, & Levine, 1971). However, Berg (1970) has noted that this is far from a consistent finding. Josephson, gt gt (1972) and Kohn and Mercer (1971) found no relation be- tween drug use and the sex of the user, while Bogg, Smith, and Russell (1969) found that males were significantly more likely to be users only in urban environments. The inconsistency in these findings may well relate to a failure to consider type of use. The Lavenhar. E; El (1972) study reported that while the sexes contributed equally to the overall use of drugs, males and females differed in their patterns of use. Males were more likely to be regular marijuana users, while females were more likely to use amphetamines and barbiturates frequently. Unfor- tunately, the authors offer no possible basis for accounting for these differences. Drug use would also appear to increase with age (Josephson, gt_gt, 1972: Blum, 1969: Hager, gt gt, 1971) or schoollevel 14 (Leahy, gt gt, 1971). Kohn and Mercer (1971) found no signi- ficant association between drug use and age, but report that use was significantly associated with class level: the major- ity of users were in their later years of study. 0n the other hand, Johnson (1973) found no differences in the extent of drug use with increasing class level. Perhaps age also is differ— entially associated with particular patterns of drug use. Lavenhar, gt_gt (1972) report that the use of marijuana, LSD, and amphetamines increases with age, but not the use of heroin or barbiturates. Interpersonal Factors. While certain demOgraphic factors are undoubtedly related to drug use, the single factor consis- tently found gggt predictive of drug use by any given indivi- dual is the number of his/her friends who use the drug (Lavenhar, gt gt, 1972: Jessor, Jessor, & Finney, 1973: Blum, 1969: Tec, 1972: National Commission on Marijuana and Drug Abuse, 1973; Kandel, 1973. 1974). In fact, the results of Johnson's (1973) study strongly indicate that demographic factors such as sex and religiosity are more highly related to the frequency of drug use than use per se. Thus, a religious, conservative female with many user friends is highly likely to use drugs, but un— likely to be a regular or heavy user. Kandel and her colleagues (Kandel, 1973: 1974: Kandel, Single, Treiman, & Faust, 1974) examined the relative importance of parental and peer influences on marijuana use among a multi— phasic random sample of over 8,000 adolescents representative of New York secondary school students. Twenty-three percent of the adolescents were matched to their parents and best school 15 friend, allowing comparison of the adolescent's perception of parental and peer drug use, and the self—reported use of the parents and best friends. Their results indicated that while adolescent marijuana use was directly related to the perceived frequency of parental use of stimulants and tranquilizers, there was little relation between self-reported parental use of these drugs and the subject's use of marijuana. However, peer influ- ences were found to be much greater. The use of marijuana by the adolescent was directly related to both perceived and self— reported use by the best friend. In addition, frequency of use was also directly related to the frequency of the friend's use. Using a multiple classification analysis (see Cooley & Lohnes, 1971) with 13 predicter variables (e.g., closeness to parents, political attitudes, etc), Kandel, gt gt (1974) accounted for 40% of the variance in subjects' marijuana use. Friend's use alone accounted for 25% of the variance, with the next most predictive factors (i.e., political attitude and involvement with peers) accounting for only 8% of the variance. In addi— tion, demographic variables such as sex, year in school, and urban vs. rural residence were only minimally related when other characteristics were controlled. Kandel, gt gt (1974) conclu- ded that: Marihuana use is clearly the result of attitudinal and interpersonal processes and influences rather than sociodemOgraphic factors...the importance of peers and the role of life-style variables, such as political attitude and degree of involvement in peer activities, support the hypothesis of an adolescent subculture that revolves around the use of marihuana (pp. 16-17)- A number of other longitudinal studies are in substantial agreement with Kandel‘s findings. Sadava (1973) examined 16 marijuana use in a longitudinal study of 151 Canadian high school students. His measure of social support for drug use consisted of a 16-item Likert scale assessing exposure to deviant role models, social reinforcement for drug use, and the absence of negative social sanctions. Social support was found to be sig- nificantly related to frequency of marijuana use in both the spring and fall of the academic year, with a significant in- crease between measurements. In addition, it was the most power— ful predictor of marijuana use. The multiple correlation using all predictors (i.e., internal~external control, tolerance of drug use, perceived positive and negative consequences of use) was .47, while the social support variable alone correlated .42 with marijuana use. Jessor, Jessor, and Finney (1973) examined marijuana use among a random sample of high school and college students. Across sex, school level, and intensity of use, marijuana users were consistently found to perceive less compatibility between peers and parents, greater peer relative to parental influence on their views, and more models, pressures, and peer approval for drug use. Of greater interest, however, are their longi- tudinal findings. Regardless of sex or school level, there was a significant increase in perceived social support for drug use among individual's becoming marijuana users during the year, but not for those remaining nonusers. There is thus a considerable body of evidence suggesting that individuals use those drugs which are used by their friends. The strength and consistency of this finding has led several investigators (e.g., Goode, 1969: Johnson, 1973: Kandel, gt gt, 17 1974) to propose that drug use is primarily a function of the particular subculture in which one is involved. Thus, the pre- dominant differentiating factor among various types of psycho- active drug use may be the "popularity" of the drugs among the users' friends. However, the evidence for such a subcultural theory of drug use has come from studies primarily focused on the use of marijuana (Jessor, gt gt, 1973: Sadava, 1973; Tec, 1972: Kandel, 1973. 1974: Kandel, gt_gt, 1974: Johnson, 1973; Goode, 1969). Consequently, the extent to which these findings are applicable to the use of any or all other psychoactive drugs has yet to be clearly demonstrated. Indeed, there is some evi— dence that the use of at least two drug classes is not directly related to peer influences. Johnson (1973) found that regular use of marijuana increases the probability of having used hal- lucinogenic drugs regardless of whether or not one's friends use them: while Battistich and Huffman (in press) found that opiate use was not related to use by one's friends. Thus, in addition to subcultural influences, such factors as individual reactions to various drugs (Tec, 1972: Haagen, 1972: Faunce & Johnson, 1970) and motivations for use (Wolfson, gt gt, 1972: Ahmed, 1967) may relate to individual differences in type of drug use. Personality Factors. In general, investigations of the relationship between personality characteristics and drug use have been disappointing. The majority of these investigations have been clinical studies, and/or have utilized clinical diag- nostic instruments such as the Minnesota Multiphasic Personality Inventory (Cohen & Klein, 1970: McAree, Steffenhagen, & Zheutlin, 18 1972: Rosenberg, 1969: Pittel, 1972: Lombardi, O'Brian, & Isele, 1968: Hill, Haertzen, & Glaser, 1960). In addition, the major- ity of subject samples in these studies were drawn from psychi- atric and prison populations, or non-institutionalized drug addicts (Cohen & Klein, 1970, 1972: Rosenberg; 1969: Pittel, 1972: Cohen, White, & Schoolar, 1971: Hill, gt gt, 1960: Knight & Prout, 1951: Lombardi, gt gt, 1968: Gerard & Kornetsky, 1954). Consequently, it is not surprising that these studies typically report an abundance of neurotic and psychotic features charac- teristic of drug users (e.g., immaturity, passivity, strong dependency needs, difficulty in coping with anxieties), although rarely conclude that a definite psychopathic syndrome is pre- sent (Pittel, 1972: McKenna—Hartung, Hartung, & Baxter, 1971: Hill, gt gt, 1960; Cohen a. Klein, 1970; Lombardi, gt gt, 1968: Gerard & Kometsky, 1954). While this has been a fairly consistent finding among insti— tutionalized and addict populations, these factors have ggt been found to reliably differentiate users from nonusers among high school and college populations (Blum, 1969: National Commission on Marihuana and Drug Abuse, 1973: McAree, gt gt, 1972). Rather, the factors found more predictive of drug use in these more typ- ical user populations are those personality variables which are more cognitive or social in nature (i.e., beliefs, attitudes, values, expectations, etc.). Thus, drug users have been con— sistently found to be less religious (Jessor, gt gt, 1973: Blum, 1969: Lavenhar, gt gt, 1972: National Commission on Marihuana and Drug Abuse, 1973) and less accepting of conventional atti- tudes and values (Haagen, 1970: Blum, 1969: Cohen & Klein, 19 1972), and to have "left-wing" political attitudes (Kandel, gt gt. 1974: Blum, 1969: National Commission of Marihuana and Drug Abuse, 1973). lower expectations for academic achievement (Kandel, gt gt, 1974: Sadava, 1973: Jessor, gt gt, 1973), and higher expectations for social acceptance (Sadava, 1973: Kandel, gt gt, 1974). However, the personality variables which seem gggt predictive of use are those attitudes and expectations which are more directly related to drug use. That is, drug users have a more positive attitude toward use, expect more positive and less negative consequences of use, and are more tolerant of deviant behavior than nonusers (Sadava, 1973: Jessor, gt gt, 1973: Kohn & Mercer, 1971: Haagen, 1970: Gorsuch & Butler, 1976). In sum, those personality variables which are most pre- dictive of drug use appear to be cognitive and/or social in nature, and generally appear to be indirect measures of the individuals' degree of socialization into conventional society (Gorsuch & Butler, 1976). While interpersonal variables appear to play a much stronger role than personality in drug use, the latter undoubtedly makes a significant contribution to predic- tion. As Jessor, gt gt (1973) concluded from their study of marijuana use: Despite its distal relation to marijuana use, the sys- tem of personality variables alone contributed signi- ficantly to accounting for the variance in marijuana use and...is, indeed, central to variation in drug use, whether there is high social support for it or not (p. 13). THEORETICAL PERSPECTIVE AND PURPOSE OF THE STUDY In this section, attention will be directed toward inte- grating the empirical findings concerning psychoactive drug use into a comprehensive theoretical perspective. First, the vari- ous problems in the extant empirical and theoretical literature will be noted and commented upon. Next, Rotter's (1954, 1955, 1960, 1967) social learning theory of personality will be pre- sented. Finally, an attempt will be made to extend and apply this conceptual framework to the problem of psychoactive drug use: working toward the goal of a comprehensive, lOgically con— sistent theoretical system. Statement of the Problem The preceding review of the literature has indicated a num- ber of problems. First, there are an inadequate number of stu— dies investigating variations in type or pattern of drug use. Where they exist, methodolOgical differences in such factors as the number of drugs examined and whether or not frequency of use was considered, as well as differences in the subcultural characteristics of the samples, have led to substantial incon- sistency. Since the adequacy of any typology is directly re- lated to the comprehensiveness of the original data, there is little useful information regarding differences in type or pat- tern of drug use and associated characteristics among drug users. 20 21 Second, much of the literature relating various personal- ity, interpersonal, and sociodemographic variables to drug use is contradictory or ambiguous. The majority of previous stu- dies have focused on the use of either one or several drugs across a general population of "drug users." Yet, much of the evidence regarding the relationship between interpersonal and sociodemographic variables and drug use suggests that individ- uals may differ extensively in type or pattern of drug use as a function of various cultural and subcultural influences. Not only do individuals generally appear to use those drugs which are used by their friends, but such factors as the age and sex of the user may well have a differential relation to drug use as a function of subcultural norms. In other words, these vari- ables may be highly related to a particular type of drug use, yet minimally related to use in general. Consequently, many of the empirical relationships between drug use and other char- acteristics of users may actually have been derived from spe- cialized subsamples, and hence cannot be generalized to other user populations. If subcultural factors are indeed as impor- tant in psychoactive drug use as the evidence suggests, any satisfactory understanding of the phenomenon necessitates con- sideration of differences in type or pattern of drug use across subcultures. That is, studies of drug use may actually be in- vestigating a particular relationship between culture and per— sonality. If this is indeed the case, the lack of replicabil— ity of the various findings relates to the failure to consider subcultural differences between particular samples. 22 Each of these problems relates to a more basic problem: namely, the lack of a comprehensive theoretical perspective which can serve to integrate the various empirical findings and guide future research. As noted earlier, previous concep- tual schemes have tended to focus on one set of determinants (i.e., personality factors 9; sociocultural factors) to the ex— clusion of others. While this limitation has been in accord with the maintenance of a strict disciplinary focus, the scope and detail of any explanation of social behavior has been lim- ited by the corresponding limitations of the separate paradigms. As Jessor. gt El (1968) have noted, this practice results in numerous "gaps" in the causal chain and reflects the narrow scope of the explanatory network in which social behavior is embedded. What is needed, then, is the development of a more comprehensive explanatory system which can encompass and inte— grate the sociocultural and personality approaches into a single coherent system. That is, an interdisciplinary "social-psycho— logical" framework which "...assumes a knowledge not only of the main facts about the social structure...but also of the main facts about the personalities operating in that structure" (Inkeles, 1959, p. 273). Examples of such interdisciplinary conceptual schemes may be found in the work of Lewin (1951), Parsons and Shils (1951) and Gillin (1954). More recently, this perspective has been applied to the analysis of several areas of social behavior (Yinger, 1968: Inkeles & Levinson, 1963), including deviant behavior (Jessor, gt gt, 1968: Akers, 1973)- 23 In order to realize such a comprehensive explanatory frame— work, the variables selected should meet a number of require- ments (Jessor, gt gt, 1968). First, they should be of suffi- cient abstractness to permit certain essential prOperties of social behavior to be revealed. Thus, the concepts should con— stitute more than mere description of surface characteristics. Second, the concepts selected from each set of determinants should suggest implications for concepts in the other set. That is, personality variables should be selected which allow coordination with sociocultural variables, and vice versa. Fi- nally, the concepts should have clear implications for the be- havior in question, for it is behavior which lies at the inter- section of personal and social processes. To the extent that these conditions are fulfilled, there should result a "struc- tural identity" between the personality and sociocultural sys- tems which provides the basis for their systematic and concep- tual coordination with behavior (Inkeles & Levinson, 1963). One approach which meets the foregoing requirements, and the perspective selected for the present research, is Rotter's (1954, 1955; 1960, 1967) social learning theory of personality. This approach is seen as particularly suited to the present investigation as the component structures are delineated at the level of the "social personality" (Inkeles, 1953), and refer to COgnitive variables which are the outcomes of patterned expo- sure to the sociocultural environment. Consequently, they al- low for the construction of a relatively direct linkage with their correlates in the sociocultural system. In addition, the resulting framework should be sufficiently abstract to be 24 applicable with only minor modifications to analysis of other deviant social behaviors (e.g., juvenile delinquency). In this regard, the author's extreme indebtedness to Richard Jessor and his colleagues (Jessor, gt gt, 1968) for their general the— ory of deviance will be apparent. Rotter's Social Learning Theory Rotter's (1954, 1955, 1960, 1967) social learning theory of personality is essentially concerned with cognitive learning or central processes, rather than with peripheral motor respon- ses. Personality is thus conceived of as the outcome of learn- ing, and is represented in systems of preferences and of expec- tation which mediate goal-directed behavior tendencies. The term "social" conveys the concern of the theory for the inter- personal or societal mediation of the significant learnings, rewards, and punishments experienced by individuals. That is, the individual's beliefs, values, and expectations are consid- ered as outcomes of interpersonal transactions or interactions which take place in a socially defined context. This context, in turn, endows the objects and actions of social interchange with meaning. The fundamental concepts in Rotter's theory are: (1) g;- pectation (E), which refers to the subjective probability held by an individual that a specific behavior will lead to the occur- rence of certain outcomes: (2) reinforcement value (RV), or the individual's preference for these outcomes: (3) behavior poten— tial (BP), or the probability of a behavior occurring: and (4) the psychological situation (S), or the individual's perception or interpretation of the immediate context of action. 25 Prediction or explanation at the personality level is based on the following formula generated from the basic concepts: BP = f(E + RV). That is, the likelihood of any behavior occur- ring in a given situation is some function of the expectation that it will lead to a certain outcome, and the value placed on that outcome. The psychological situation is implicit in that each of the terms in the formula is coordinated to the situa- tion. Consequently, the magnitude of each term varies as the situation varies. Actual behavior always involves a process of selection of that particular behavior in a repertoire which is perceived as having the highest probability of securing grati— fication in a given context. Through the processes of socialization and experience, var- ious specific behaviors become functionally related as a conse- quence of their substitutability in leading to classes of simi- lar goals. Likewise, specific goals become functionally related through their interchangeability in reinforcing certain behav- iors. These ggtg of functionally related behavior potentials are termed need potential (NP). The mean expectancy of obtain- ing gratification characteristic of a set of related behaviors is called freedom gt movement (FM). Finally, the mean prefer— ence value of a set of goals is referred to as need value (NV). Thus, it is possible to state a more generalized formula analo- gous to the basic formula: NP = f(FM + NV). That is, the pro- bability of occurrence of a set of functionally related behav- iors in relation to a set of similar goals is a function of the mean expectancy for these behaviors securing these goals, and the mean value placed on the set of goals. In this manner the 26 theory can be applied to either a general class of behaviors (e.g., deviant behaviors) or a specific member of that class (e.g., drug use) while maintaining the basic theoretical lOgic. The sets of functionally related behavior potentials may be considered the needs or motives of Rotter's theory. How— ever, these needs are not considered dependent on physiologi- cal referents. Instead, they are conceptualized as predisposi- tions to respond in certain ways, whose nature and organization are consequences of the social learning experiences and encul- turation of the individual. In as far as they represent the outcome of social learning, they reflect in large measure the patterns of valuation and success and failure with which the culture has confronted the individual. Finally, the substantial importance attached to the psy- ch010gical situation should be emphasized. In Rotter's theory individuals are not seen as conglomerations of traits or attri— butes which determine behavior irrespective of the context of its occurrence. The situation is seen as not only influential in determining the relative strength of the behavioral tenden- cies, but also in determining the process of selection among them leading to action. Consequently, the situation is descri- bable in terms parallel to those used to describe persons (e.g., in terms of the various goals available, etc.). The psycholo- gical situation is thus the environment of learned meanings con- stituted predominantly of the social definitions of the world in which the individual is embedded. It therefore provides the bridge between personality and the sociocultural environ- ment. 27 This brief description of Rotter's theory will serve as the framework for an analysis of psychoactive drug use. Con- sistent with this perspective, discussion will center around factors seen as influencing the individual's expectation that drug use will lead to certain outcomes, and the corresponding preference or reinforcement value of these outcomes. In over- view, Rotter's theory predicts that the greater the expectation that drug use will lead to preferred or positively-valued out- comes, the greater the probability that it will occur. Since these expectations and valuations are considered outcomes of patterned exposure to the sociocultural system, various socio- cultural factors which are seen as conducive to drug use will be discussed first. The Sociocultural System According to Rotter's theory, drug use, like other behav- iors, is goal directed. That is, it is a learned way of seek— ing and securing gratification or success and of coping with frustration and failure. Any analysis of factors conducive to drug use, then, should begin with consideration of the possible gratifications or goals obtained from engaging in the behavior. Leighton, Clausen, and Wilson (1957) have proposed that the possible goals of drug use fall into two classes: (1) ten— 3 sion release: and (2) enhancement of a sense of group member- ship, gg much gy the shared act gt use as by any intrinsic ef- fects of the drug. While either or both of these may serve as the dominant motivation for drug use in any individual case, the preceding review of the literature strongly suggests that 28 the latter goal is typically of greater import. That is, drug use appears predominantly the outcome of social rather than personalistic motivations. However, the use of drugs is only one of many possible means of attaining these goals. Considering drug use as a deviant behavior, one would ex- pect that individuals typically hold low expectations for drug use leading to desired outcomes. With the exception of certain restricted types noted earlier (e.g., moderate use of alcohol), the use of psychoactive drugs is generally likely to entail more negative than positive outcomes for the user. For example, such costs as the risk of arrest, negative evaluation by non- using peers, the user's family, and other conventional groups (e.g., school and church groups), as well as any negative ef- fects of drugs themselves (e.g., addiction), would be expected to overshadow such positive outcomes as tension release and the satisfaction of affiliation needs. This is particularly true in that these positive outcomes could be obtained through non- deviant behaviors without the large risk of negative consequen- ces. Consequently, the sociocultural factors which induce an individual to use drugs must be those which increase the expec- tation that deviant behavior will lead to positive outcomes, and/or decrease the expectation that such behaviors will be punished (i.e., lead to negative outcomes). Since the outcomes attainable through drug use may also be attained through legitimate, nondeviant behaviors, the socio- cultural factors which serve as an instigation to deviance are likely to be those factors which inhibit the individual's soci— alization into conventional society. One of the primary functions 29 of socialization is to make the individual responsive to the normative structure of the society. This structure may be de- fined as a set of socially defined standards or expectations about appropriate modes of behavior, the violation of which results in the application of some pattern of sanctions--i.e., punishment (Merton, 1959: Jessor, gt gt, 1968). To the extent that this structure is operative, there will be significant pressures against the occurrence of deviant behavior. Thus, the more conventionally socialized the individual, the less likely s/he should be to adopt illegitimate means of goal at- tainment. Conversely, to the extent that normative constraints against deviance are reduced, there should be a greater insti- gation to attain goals through illegitimate means such as drug use. Durkheim (1951) and Merton (1957) have suggested a number of conditions which may result in a breakdown of the normative structure: a condition they refer to as "anomie." One such condition is low consensus on norms. When a norm is not widely shared there is uncertainty about appropriate behavior. Sec— ond, even when norms are widely shared they may lose their moral authority, and hence their effective influence on behavior. Finally, and related to a loss of moral authority, there may be a decline in the application of sanctions for normative vio- lation. As Jessor, gt gt (1968) have noted, the moral author- ity of norms derives in large part from the fact that adherence is rewarding in the long run, even in the negative sense that it avoids the possibilities of punishment. Thus, to the extent 30 that sanctions are not applied, norms may no longer serve to inhibit deviance. One index of the degree to which norms are likely to in- hibit deviance is the extent to which the individual is exposed to the social institutions responsible for instilling a sense of normative obligation. The communication of appropriate modes of behavior is largely the responsibility of social institutions such as the family, school, and church groups. An individual who participates in these groups will be exposed to pressures serving to inhibit deviance, and is therefore unlikely to en- gage in deviant behaviors such as drug use. In other words, participation in conventional, non-deviant groups is incompa- tible with engaging in deviant behavior. The preceding review of the literature suggests that this is indeed the case. Individuals who use psychoactive drugs are unlikely to be active participants in school or church groups (Blum, 1969: National Commission on Marihuana and Drug Abuse, 1973). In addition, drug users have been found to be less ac— cepting of the norms and values of society (Cohen & Klein, 1972), and to have little respect for social institutions (Gorsuch & Butler, 1976). Thus, one sociocultural factor con- ducive to deviance appears to be inadequate interaction with the socializing agents of society, leading to a lack of norma- tive constraints on behavior. A second set of sociocultural factors possibly conducive to the adoption of illegitimate means are those which restrict the individual's access to legitimate, nondeviant means of goal attainment. That is, to the extent the individual feels that 31 valued outcomes are unattainable through nondeviant behaviors, there should be a greater instigation to deviance. The impor- tance of this aspect of the social environment is conveyed in Merton's (1957) concept of "value-access disjunction," Nadel's (1957) emphasis on the degree of command over existing benefits and resources, and may be seen as a large part of the concep— tual meaning of "social class" (Hollingshead & Redlich, 1958). In addition, it is also the focus of Rotter's (1955) concept of "freedom of movement." Merton (1957) documents the overwhelming emphasis placed on success or achievement values in American social life, but points out that the institutionalized or legitimate channels for achieving these goals are not uniformly distributed through— out society. In particular, the lower social strata and cer- tain ethnic and racial minorities occupy disadvantaged posi- tions with respect to access to legitimate means of goal atain- ment (e.g., education, participation in socially influential groups, etc.). According to Merton's theory of anomie, it is this "disjunction" between the pervasive value emphases of the culture and the socially—structured limitations on legitimate access to these goals which generates pressure toward the adOp— tion of deviant, illegitimate means of achieving success and coping with failure. The distribution of access to legitimate Opportunities, and therefore of value-access disjunctions, par— allels the hierarchy of socioeconomic status and membership in ethnic or racial minorities. Consequently, this sociocultural source of pressure for deviance would be expected to be con— centrated in the lower social strata. Thus, location in the 32 "opportunity structure" (Jessor, gt gt, 1968) may serve as an index of the degree of instigation for the use of illegitimate means of goal attainment. In addition, patterned exposure to such differential Opportunity should result in differential expectations of achieving valued goals through legitimate means. This factor may partially explain drug users' disdain for social institutions and conventional norms and values, as well as the findings that use of "hard" drugs such as opiates (as Opposed to "soft" drugs such as alcohol and marijuana) tends to be nega— tively related to socioeconomic status (National Commission on Marihuana and Drug Abuse, 1973) and positively related to mem— bership in ethnic or racial minorities (Johnson, 1973)- Limited access to legitimate opportunities for goal at- tainment and the lessening of normative constraints on behav- ior are thus two sociocultural factors which may serve to in— crease the probability of an individual engaging in deviant behavior. However, these factors are not directly conducive to the use of drugs. That is, they may serve as an instigation to deviance pgt_gg, but not necessarily to drug use in parti- cular. In order to adequately account for drug use, those soci- ocultural factors which bear directly on this particular form of deviance must be examined. It is in this regard that peer influences on drug use would seem of crucial importance. The emphasis here is on socially patterned Opportunities for learning and performing deviant be- haviors, and the nature and operation of the sanction systems for discouraging such behavior. The Opportunity to learn and engage in deviant behaviors is considered an important causal 33 agent in Sutherland's (1955) "differential association" theory of criminal behavior, as well as Cloward's (1959) and Cloward and Ohlin's (1960) analysis of delinquency. Cloward and Ohlin's general thesis is that socially patterned and differentially distributed access to illegitimate means can usefully refer to differences in the exposure of individuals to the everyday man— ifestations of deviant behavior by other persons in the envi- ronment. Consequently, they argue that such opportunities in— fluence not only the likelihood of deviant behavior, but also the specific tgtg of deviance. Thus, this factor may not only serve to account for drug use, but also the particular type or pattern of drug use manifested by the individual. The preceding review of the literature indicated that in— dividuals who use psychoactive drugs are likely to have friends who use drugs. These friends not only supply the Opportunity to observe and engage in deviant behavior, but may also demon- strate the absence of negative consequences, thereby serving to weaken pressures against its occurrence. Consequently, to the extent the individual is exposed to and interacts with drug users, s/he should be more likely to use drugs. In addition, since these individuals serve as the person's models of drug use, the individual's usage pattern should largely conform to that of his/her friends. In summary, three factors in the sociocultural system have been identified which would appear to influence the probability of an individual using drugs: (1) restriction of access to legitimate means of goal attainment: (2) a lessening of normative constraints against deviant behavior: and (3) opportunities to 34 observe and engage in the use of drugs. Each of these factors may be seen as related to the degree of socialization in con- ventional society in that each serves to weaken the effective- ness of the culture's means of social control. In addition, the first two factors may be seen as related to the instigation of deviant behavior in general, while the third is most related to the particular form of deviance--drug use, in this case. It should also be noted that these factors are probably not independent, but interdependent to some greater or lesser degree. For example, restriction of access to legitimate means of goal attainment in itself probably serves to lessen norma- tive constraints against the adoption of illegitimate means. Further, any decrease in the effectiveness of social control may lead to an increase in the frequency of deviant behavior in the social environment. Thus, more Opportunities to learn and perform deviant acts would be provided. Conversely, inter— acting with individuals who engage in deviant behavior is likely to be incompatible with participation or interaction with non- deviant individuals and groups, thereby lessening exposure to normative sanctions, etc. These structures in the sociocul— tural environment may thus be viewed as comprising a system of interrelated structures, each interacting with the others to provide a greater or lesser instigation to engage in deviant acts such as drug use. The Personality System While certain conditions in the sociocultural system may account for the differential distribution of deviant behavior 35 across various social groups, such factors are not capable of accounting for differential behavior within a given segment of society. For example, not every member of the lower social strata adopts illegitimate means of goal attainment: nor do all individuals who interact with drug users become users them— selves. In order to account for these individual differences, characteristics of the individual personality must be examined. Consistent with Rotter's theoretical framework, the per- sonality factors to be discussed are considered a consequence of patterned exposure to the sociocultural system. That is, they refer to attitudes and expectations which are the result of social interaction. In addition, these factors parallel those of the sociocultural environment. They refer in some sense to the extent to which the individual is responsive to agents and institutions of social control, and therefore re- flect the adequacy of socialization into conventional society. However, whereas the factors discussed in the preceding section referred to mechanisms of social control, these factors are con- sidered to represent the degree of personal control over be- havior. One personality factor which seems of crucial importance is the individual's attitude toward deviance. This variable may be seen as reflecting the degree of personal acceptance or tolerance of transgressions from socially approved behavior, and is thus probably an index of the degree to which social norms have been internalized. To the extent that the normative structure has been internalized the person should hold a nega- tive attitude toward deviance, and deviant behavior should be 36 personally aversive. That is, engaging in deviance should re— sult in negative self-attributions, a loss of self-esteem, and other aversive consequences. Therefore, the individual should be less likely to engage in deviant behaviors such as drug use. The relation of the individual's attitude toward deviance and actual behavior may be seen as mediated by the concept of expectancy. That is, holding a neutral or positive attitude toward deviance may indicate a low expectation that such be- havior will lead to punishment, and/or an expectation that de- viant behavior is required to achieve valued goals. Since ex- pectations of punishment operate to inhibit behavior in Rotter's theory, tolerance of deviance should be conducive to its occur- rence. Thus, the less negative the attitude toward deviance, the more likely the use of psychoactive drugs. A second personality factor which is more directly rela— ted to drug use is the individual's expectations concerning use itself. That is, while an individual may be tolerant of deviance in general, s/he may expect the use of drugs to lead to negative outcomes (e.g., addiction, overdose, etc.). There- fore, the use of drugs should be less likely. In addition, these expectations may differ depending on the particular drug, and therefore be related to the type or pattern of drug use. For example, an individual may view use of alcohol or marijuana as relatively harmless, but expect use of opiates or barbitur- ates to lead to extremely negative outcomes. Consequently, the individual would be much more likely to use the former sub— stances than the latter. 37 As with factors in the sociocultural system, these person— ality factors are probably related. For example, holding a positive expectation for drug use may generalize to other forms of deviant behavior and increase the individual's tolerance for deviance, or vice versa. In addition, these factors should theoretically be related to the various sociocultural factors. Restricted access to legitimate means of goal attainment and/or a decrease in normative constraints on behavior may each serve to increase an individual's tolerance for deviance. Likewise, the opportunity to observe and interact with drug users may lead to a greater expectation that drug use will lead to posi— tive outcomes. Groups of friends tend to be highly similar in terms of attitudes and norms (Heider, 1958: Newcomb, 1961), es- pecially in so far as these attitudes are relevant to group functioning (Cartwright & Zander, 1968). Thus, individuals who interact with drug users should be more tolerant of deviant behavior and expect more positive consequences from drug use. Following Rotter's perspective, these interrelations are a log- ical necessity. In order for the individual to mediate the re— lation between the sociocultural environment and deviant beha— vior, his/her personality must reflect to some degree the con- sequences of patterned exposure to that environment. The theoretically expected relations and interrelations between the sociocultural system, the personality system, and drug use are schematically summarized in Figure 1. 38 on: make so moocOSHmcH zpfiamsomhom psm HandeSOOMOOm H mmDUHm uuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu on: ways how wQOfiPMPOOQxO o>flpfiwom J I mofldfl>mfi HO% ®OQNH@HOB Ememwm MBHH¢Zommmm /V AV _ mm: mca>mm “ mo m ....... mo _ wquHmflPMEHOQ Ummmmhowm II on: wspc QH msflwmwco 62m mcfl>homnc how hpflQSPhommo Ememwm Qaa Aasvm Amavm seasons Acmvac Amcvsm mamcam *msempm Hosanna Ammvoo Asmvmm osmsom Amavas Amsvmm mama In: **Kmm ma.m oc.am Hm.s ss.mm stmwm .m.m M Agv m .m.m m Agv m Ams n zv oaasnm pooaem Aam n zv oaaemm owoaaoo oaomasm> moamsmm 039 6:9 Mo mogpnflpopompmno canmmhmosOm H mgm AU.PQOOV H mHm¢e 54 H A H HmQOHmmomohm whowmsma .mhopmemohm .OPO .HHOHQ swampom\UOHHme UOHHmenHEom UOHHmeCD H 66%OHQEoQD o: PQOUSPm COHPmasooo A MAI-\AAA A BMBWWM H VVVVVVOV O\N “if-3' N MA AA 0 wN NN VVO OVVOV HH UN : A Hwavma Assam Hooo.oom A V spas .mq on AmHvoH Hovm Acoo.oom-ooHv apao Amv Asmvoa Ammvma oasosm Asv cs.a om.m Acavaa am.a so.m Hmmvaa Aooo.mm-cav £309 .w>¢ Amv nova Aaavm Aooo.onw v sacs .am Hmv Ammvma AHva HmasM\asmm AHV ONHm thcsesoo ooo.mmw.n Am mmm.:m;mH A: Am Hm AAA Aflmwm \OO\NN m HH vvvvv H H N v v Hm.a mm.m cm.a mc.m mmm.sa-oH v mmm.m-m V 80.me Cw osoosH hHHEmm Hmsss< 30 “NOW AAAA Afimnm :1' (\1 HN m vvvvv N O L\("\\O H .o.m x Asv m .m.m x Asv m Ans u zv oHasmm poohpm Ham n zv oaosom omoHHoo oaomasm> Ac.psoov a mqmse 55 Aova Acmvaa psocspm opmscmau Hov Ammvaa Ammvofi season s Amv +3.H AmHvo 0N.H AmmvHH HOHGSW : Aiv AomvoH Amva whosozmom : Amv Asavma Hsvm cosnmoss oonHoo Amv Asvm o Hoogom swam Hap|1|| **mmmHo Asavs Ammvma oo.s-om.m Hmv Aamvma Ammvms as.m-oc.m Asv ca. Ammvsa as. Homvfla mm.m-om.m Amv Amavo Amavm as.m-oo.m Amv Amva o mm.amw AHV Aaavm Asvm cosmHoocs: Asvm Amvm scape Asavm Amvm osaoacoz o Amvm moopzomop .sz @ .w¢ o Amvm wchoochcm Aaavm Amvm moosamsm Amva Adew QOHPmOscm “nave Amvm moapasmssm Amy: ANvH ooQOHom HMHSsz ABNVNH Amdem ooCOHom HmHoom Hews: OHEopmo< quo mezmopem .m.m m ARV c .m.m m Asv c Ams n zv oHasmm poospm Ham n zv oHasmm omoHHoo oHosHsm> AU.#QOUV H mgm¢e Ho. v m** mo. v Q* 56 H 30:& P.som mm HmSOHmmOHon N mummmsms .mHOPOHHQon N .opo .HHOHQ m :msonoH\BOHHHMm UOHHHanS Ho pmHHmeIHEOm QOHPmQSOOo covoomxm A AA UN :rH m Amva socpo s o cog o: Ammvas pm: Hoonom [\AA .m.m x ARV m .m.m x Agv m ANA n zv oHasmm eoonem Ham n zv oHaamm omoHHoo oHonasm> AU.PQOOV H mqm¢e 57 males, t(88) = 2.77, p < .01, and are less likely to be work- ing toward a professional occupation than males, Kendall's Tau g = -.27, p < .001. Procedure The questionnaires were distributed to subjects by members of the staff of the Counseling Center and DEC. Subjects were asked to read over the cover letter, briefly examine the ques- tionnaire, and decide if they were willing to participate in the study. Individuals who were unwilling to participate were asked to return the blank questionnaire. Otherwise, subjects who received no incentive for completing the questionnaire were allowed to take the questionnaire with them for completion at their convenience. A self—addressed, stamped envelope was en- closed for returning the completed questionnaire. Subjects who received a lottery ticket for their participation in the study were asked to complete the questionnaire immediately. They were given their lottery ticket upon returning the com- pleted questionnaire to the author or a member of the DEC staff who was present at the time. Since the use of most psychoactive drugs is currently il- legal, a number of steps were taken toward insuring the confi- dentiality of responses. A statement of these procedures is included in Appendix D. Since the procedure followed for distribution and collec- tion of the questionnaires allowed individuals to return ques- tionnaires immediately if they decided not to participate in the study, it is unfortunately impossible to determine 58 m ozoz PaosvoHMQH PQOSUOHm mosmccmpp< sopsno AAA NW N Hcmn VVV O H:- Hm: AA A\OU\ O\N\O vvv MON m H ozoz scape ansoh oHHonpmo PQMPnopopm *mocomomopm mSOHmHHOm AA AA N HA0: N N (fie-4:? vvvvv O\CD MO\CD HH AA AAAMM \O O\O\N m vvvvv N MONK) (I) Amavm ocoz o 90:90 Assn amazon AamvoH oaaonemo AmdeH Pcmpmoponm vczowmxomm mSOHmHHom AA N BAA/x :2' major) vvvvv (\Nm MB mm Amva AmHvoH Acmvos coBOUH3\.Honm\.>HQ OOHHHmE m onsHm mzpmpm HmPHHmE AA MHA L\N\O VVV U'NL\N sm.s mm.HN 63.3 as.mm **wmm .m.m x Agv m .m.m x Asv m Asa n zv monsom Ham u zv monz oHansm> mOHPmHHmPOmHmco OHnmmHmosom CH moocmhommHQ xom N mHm AU.PQOOV N mqm¢e 60 m@.H wo.m A AAA [\AN N O NEH HH vvvvv N-fl' L\L\-\O H A mAAAAA A bqavaUX3 VVVVVVO v 52)- H3 4‘) :2' ON H A AAAA 00A moo HO H\O HNHH vv VVVV mm .3 MO\\O HN H NAAWA :fd‘d‘ Hd’ VVVVV H H H-:)' H Am A0 A0 mm.H mm.m Am Am Aw mmmsfiwsm COHPmOzUm mePHQmesm mosmflow Hmpzpwz woCmHom Hmfloom Howfl: OHEmUmo< MHZO mezmmbem HmQOHmwm%opm hmwmcme .Mopmflnmonm .opm .meHv swamnom\vmaaflxm UmaaflxmnHEwm Umaaflxmcb UmzoamemCD PmmUSPm SOHPmmsooo Aooo.oomAAV zpflo .mg A Aooo.oom-ooHV Apflo A magnum A Aooo.mm-oHV C309 .m>< A Aooo.onwV mace .Em A HMASM\SMmm A m #100 v-IN V V V V V V mnwm hPHCSEEOQ .Q.m x ARV m .3 x A§ g Anm u zV mmamsmm Adm n zV mmams manmfipm> AU.PQ00V N mqm AU.PCOUV N mam AU.PQOUV N mHm .05. However, within the Street Sample subjects offered an incentive returned sig— nificantly more questionnaires,)é? (1) = 68.78, p ( .0001. Thus, since subjects who received a lottery ticket did not dif- fer significantly from those who were not offered an incentive except in terms of class level, there is little reason to ex- pect that offering an incentive had any significant effect other than to increase return rate. The Measurement of Drug Use. A central problem in any study of drug use, and a major criticism of the majority of previous studies relates to the method used to measure drug use. Perhaps the most common method has been a simple classi— fication of use—nonuse. However, such a gross measure yields little information about the individual's drug use. For exam- ple, the use of simple discrete categories does not take the frequency with which drugs are used into account. Thus, an 64 individual who has used marijuana on five or six occasions would be classified into the same category as one who has used the drug on over 60 occasions. Yet, few would contend that the drug usage of these two individuals was comparable. What is clearly needed is a more refined measure. In the present research, three separate measures of drug use were utilized. The first was a gross measure of use-nonuse. Respondents were asked the following two questions for each of the 12 substances examined (see Footnote 1): (1) Have you ever used this substance?; and (2) Have you used this substance within the past year or so? On the basis of these two ques- tions, respondents were classified into one of three categories for each of the 12 substances: 1. nonuser - Individuals who had never used the substance. 2. past user - Individuals who had used the sub- stance, but not within the past year. 3. user - Individuals who had used the substance within the past year. As can be seen, this measure does not take either frequency of use into account or the individual's overall pattern of drug use. Consequently, it only allows comparison between users and nonusers of each substance, and not among users as a group. The second measure is more refined, and allows examina- tion of the variation amggg users of each drug. Respondents were asked for each drug which they used: (1) The frequency with which they used the substance within the last year (rang- ing from about once a year to three or more times a day); and (2) Considering all the occasions on which they use drugs, the 65 proportion of the time they use the particular substance (rang- ing from less than 10% to over 90%). These two responses were then combined into an overall frequency—variability index5 of use for each drug (see Cahalan, gt al, 1969). Specifically, the frequency with which an individual used a drug was weigh— ted by his/her variability in use of the drug. Thus, indivi- duals who used a drug with a low frequency, or with a moderate frequency and much variability would receive a low score; while individuals who used a drug with a high frequency, or a moder- ate frequency but little variability would receive a high score. Thus, this index summarizes in a single score frequency and variability of use of psychoactive drugs, permits the calcula— tion of group statistics which are not possible with discrete categories and, since the scores are continuously distributed along a single dimension, allows statistical comparison of group means, correlational analyses, etc. While the frequency-variability index is a considerable refinement over any discrete categorical measure, it does not allow assessment of an individual's type of drug use. That is, it does not differentiate between individuals who only use a single drug and multiple drug users, nor differentiate among multiple drug users who differ in the particular drugs utilized. Thus, a third measure of drug use was constructed which takes into account the number and type of substances used, as well as the frequency-variability with which they are used. The subjects' frequency—variability scores were intercorrelated and subjected to a discrete cluster analysis (see Fruchter, 1954). This procedure classifies the drugs into discrete groups on 66 the basis of their intercorrelations, and allows the computa- tion of "cluster scores" for each subject (analogous to factor scores; see Harman, 1960). These scores can then be used for grouping subjects on the basis of their having similar types or patterns of drug use. The first two measures were utilized for testing the hypo- thesized relationships between drug use and the various socio- cultural and personality characteristics in a series of uni— variate analyses. The third measure was utilized for a multi- variate analysis of drug use, simultaneously examining the re- lationship between the set of personality and sociocultural measures and the various types of drug use characteristic of individuals in the samples. The Measurement of the Sociocultural System. In the pre- ceding theoretical analysis, three sociocultural structures were described which were hypothesized as related to drug use. Specifically, drug use was seen as varying directly with the degree of restriction in access to legitimate opportunities for attaining valued goals, the lessening of normative constraints or social control over deviant behavior, and the degree of op- portunities for observing and engaging in drug use. In the following pages, the methods used to measure each of these socio- cultural structures will be described. Two indices of access to legitimate opportunities were constructed for the present research. The first measure was an index of Socioeconomic Status (SES), based on Jessor, gt al's (1968) modification of Hollingshead‘s (1956) two-factor Index of Social Position. Specifically, Father's Education, Mother's 67 Education, and Annual Family Income were additively combined, yielding SES scores ranging from 3 to 23. These items, their internal splits, and item—total correlations are presented in Table 3. The second measure of access to legitimate opportunities was designed to take into account a broader class of factors, following a similar scale constructed by Jessor gt a; (1968). Briefly, it was reasoned that younger people have more access to opportunities than older individuals, since the young have a greater potential for achieving unattained goals in the fu- ture. In addition, youth itself is seen as an asset in many occupational settings, while age is often considered a social drawback. In a similar vein, older persons who have never mar— ried or have lost their spouses have less access to opportuni- ties than those who are married. Individuals who are unemployed or have only unskilled jobs have less Opportunities for achiev— ing success goals than those who have skilled or higher occu- pations. Whites have greater access to opportunities than mi- norities and, finally, individuals who belong to higher-status Protestant churches have more opportunities than those who are affiliated with less prestigeful congregations or have no reli— gious affiliation at all. Thus, the index of Objective Access was composed of five items, each scored dichotomously: Age (0 = 30 and over; 1 = under 30); Age plus Marital Status ( 0 = 30 and over and single: 1 = under 30 and/or married); Occupa- tion (0 = unskilled or unemployed; 1 = student, semiskilled or higher); Race (0 = other than White; 1 = White): and Reli- gion (0 2 other than Protestant; 1 = Protestant). Obviously, 68 TABLE 3 Socioeconomic Status Variable Q (%) Item—Total1 Correlation Father's Education .64 1 = Grammer school or less 11(9) 2 = Some high school 9(8) 3 = High school graduate 24(21) 4 = Some college 21(18) 5 = College graduate 24(21) 6 2 Post graduate 26(23) Mother's Education .55 1 = Crammer school or less 4(3) 2 = Some high school 13(11) 3 = High school graduate 39(34) 4 = Some college 29(25) 5 = College graduate 25(22) 6 = Post graduate 6(5) Annual Family Income .47 1 = £335,000 6(5) 2 = 5‘99999 16(14) 3 = 10-14.999 22(19) 4 = 15-249999 30(27) 5 = 22$25,000 39(35) Reliability Coefficient (Alpha) = .72 1Corrected to remove effect of item on total score. 69 since these items are not conceptually homogenous (e.g., there is no reason to suspect any relationship between age and reli— gious affiliation), this measure is, a priori, the weakest of the two. Measurement of normative constraints was accomplished through the construction of a scale of Involvement With Pro— social Groups, which focuses on participation in religious groups. Four dichotomous items concerning religous background, religi— ous preference, and church attendance were additively combined, yielding a range of scores from 0 (no involvement) to 4 (high involvement). These items, their internal splits, and biserial correlations with total score are presented in Table 4. Of the three sociocultural structures considered to be related to drug use, opportunities for observing and engaging in drug use is theoretically the most powerful predictor. Con— sequently, three measures of this structure were constructed for the study. The first scale (Opportunities for Deviance L: SociodemOgraphic Factors) is the weakest of the three measures, and is composed entirely of demographic variables. It was rea- soned that individuals between the ages of 18 and 25 have more opportunities for engaging in deviant behavior than individuals 16 or 17 years of age, or individuals between the ages of 26 and 29. Likewise, individuals in these last two groups have more opportunities for deviance than individuals under 15 years of age, or over 30 years of age. Similarly, individuals who are single are seen as having more opportunities for deviance than individuals who are divorced or separated, while the lat- ter have more Opportunities than married people. That is, Involvement 70 TABLE 4 With Prosocial Groups Variable n (%) Item-Total1 Correlation Religious Background .19 O = None 12(10) 1 = Any 109(90) Religious Preference .52 O = None 56(46) 1 = Any 65(54) Church Attendance 1 .58 O = None 66(55) 1 = Infrequent/Frequent 53(45) Church Attendance 2 .40 O = None/Infrequent 106(89) 1 = Frequent 13(11) Reliability Coefficient (Alpha) = .63 1 Corrected to remove effect of item on total score. 71 single individuals and those between the ages of 18 and 25 ty— pically have fewer responsibilities and greater freedom than individuals in the other age groups or those who are married. Likewise, individuals who were raised in a city have more oppor- tunities for observing and engaging in deviance than those who were raised in smaller towns or the suburbs, who, in turn, have more Opportunities than small town or rural residents. Thus, the first measure of opportunities for deviance consisted of three trichotomously-scored items: Age (0 15 or under and 30 or over; 1 = 16-17 and 26—29; 2 = 18—25): Marital Status (0 = married; 1 = divorced or separated; 2 single); and Size of Home Community (0 = farm, rural, or small town; 1 = average town or suburb: 2 = city or large city). Once again the items are not conceptually homogenous (e.g., there is no reason to expect any relationship between age or marital status and com- munity size). In addition, from examination of the demographic characteristics of the two samples (Table 1), one would expect that the range of scores on this measure would be relatively restricted. Thus, there are a number of reasons for expecting that this measure will be somewhat inadequate. Examination of the internal-splits and item-total correlations for these three items presented in Table 5 confirms this expectation. However, although the reliability coefficient for this measure is rela- tively low, it is sufficiently large to allow the measure to be retained. Nonetheless, any results based on this measure should be interpreted with caution. The second measure of opportunities for observing and en- gaging in drug use (Opportunities for Deviance 2: Perceived 72 TABLE 5 Opportunities for Deviance 1: SociodemOgraphic Factors Variable fl (%) Item—Total Correlation1 Ass .36 o = g 15 and 3-30 12(10) 1 = 16-17 and 26-29 17(14) 2 = 18-25 94(76) Marital Status .38 O = Married 17(14) 1 = Divorced/Separated 9(7) 2 2 Single 95(79) Size of Home Community .10 O = Farm, rural, sm. town 29(25) 1 = Avg. town or suburb 53(46) 2 = City or lg. city 33(29) Reliability Coefficient (Alpha) = .40 1Corrected to remove effect of item on total score. 73 qug Egg in the Social Environment) concerns the extent to which the individual interacts with drug users in his general social environment. Specifically, respondents were asked what percen- tage of their friends used each of ten classes of psychoactive substances6 (on a six-point scale ranging from less than 10% to over 90%). The responses to these 10 scales were then addi- tively combined to yield a range of scores from 0 (less than 10% of their friends used any of the drugs) to 60 (over 90% of their friends used each of the drugs). These ten items, their intercorrelations, and item-total correlations are pre— sented in Table 6. The third measure of opportunities for drug use (Opportu- nities for Deviance 3: Extent 9: Use With Friends) concerned interaction with drug-using friends. Respondents were asked for each of their five closest friends: (1) The amount of lei- sure time spent with the friend (scores ranging from O for none to 4 for all); and (2) The amount of time the individual and friend used drugs when they were together (scores ranging from 0 for none to 4 for all). These two scores were then multipli— catively combined for each friend, and then summed across the five friends to yield a range of scores from 0 (none of the five friends used drugs with the subject 9; none of the friends spent any of their time with the individual) to 80 (all five friends used drugs with the subject and they spent all their leisure time tOgether). Basically, it was reasoned that of two friends who both used drugs when with the subject, the friend who spent more time with the subject provided more op- portunities for observing and engaging in drug use. The five .onoom Hapop so amps mo Pommmo o>osoh op pmpoohhooH 74 mm. H Amggaav psmflOHmmmoo spflaflnmflfimm om. - mm mm m: on mm an H: NH mo muflxo msoopflz Aoflv oz. - m: on ma Hm mm on we mo mpcm>aom Amv om. - on m: mm Hm mm mm mo mmpmflgo Amv mm. - mm :m on A: 0: NH msflmooo ARV as. - as no mm 0: mm mmcflsmpmnge< on up. - me am mm ma mmpgpgpflppmm Amv mm. 1 mm mm SH mamaflaflzwsmhe Adv mm. 1 m: mm msowocfiozaamm Amv ms. . mm mamzmfloms Amv mm. - Hoaooaa AHV HMMWWMHMMMMO Aofiv Amv Amv Asv va Amv Asv Amv Amy AHV mu2muWHpMomMmmmmwoomm PC¢EC0MH>QM Hmfloom may Ca om: mssm pw>fioopmm "N mocmfi>om 90% moHPflQSPpommo 0 mqmH H ooo.mmH\mHH mH = m.: o.m mmpmHgo HHH HH = = m.m 0.: mszooo HHH HH = = m.m o.m mmcHs unponme< HH HHH ooo.mHa\mHH m = m.m o.m mmpm _ -Hspanmm H >H ooo.on\mHH m = w.H o.H mHmN :HHHSUQMHB >H H = = m.m 0.: mcmw nocHosaamm :HH H ooo.nma\mos mH ooo.mH\HmmH H m.m o.m mumsHHHmz H mcoz mcoz mcoz m.H o.H HonoOH¢ chHPMH0H> cam: 660: m>HPMEHoz mascmnom Noamm Hom Cowmmwmmom How mo Noch m HoHvsoo Spamsom Hmwoq N zpamcom Hmwmg Ham>onmmem HmHoom wSHQ mm: WSHQ SPHB poPmHoomm¢ mmHPHmcmm Hmwmq paw Hm>ohmmmmflm HmHoom OH mqmHPdEHOCIHwHPQOO =zaoPMHmpos: haco mm COHPmOHmHmmmao mPH wwHMHPmSw Hm>oHQ9mme HmHoom mo msHoP CH wQHxQMH Boa ham>HPmHmH mpH mm Hams mm .mmHSPmHmHMoH opmpm hams wsosm mm: msmstHms mo COHPmNHHms IHEHHoop phagop onHP map .mmHPHmsmm HmHopom oHo>mm op Powmnsm mH om: mcmstHms :w50npa¢ .AHmmH .H zHSh popcosw mmv ommH mo Po< HoHPcoo 69m COHPQm>on mm59¢ mSHQ m>chonoHQEoo : on>mm n >H anm u HHH mpmnmuos u HH HmeHsHs n Hm .mmHPHQOQ ESEHNME N .Am: n mv Hopsoo QOHmesum mSHm awooa mo mmmpm hp wQchmmH >H - - - m.m o.m meon msopsz >H : I I H.: o.m mpsm>aom m20HpmHoH> smog ope: o>HpmsHoz masconom Noamm HOH COHmmommom How Ho NoocH m HoHpsoo hpamcmm Hmwmq m hpamcom ammoq Ham>oHQQMmHQ HmHoom MSHQ AU.PQOOV OH mamaom mmpmHgo ngpo Asv m Asv m Asv Q How: How: 9mmm Homssoz Amsuzv mHmsmm pmmnpm Asv m Asv m Asv m How: How: pmmm Homssoz AHmuzv mamsmm owoaaoo WSHQ AU.PCOUV HH mqmHom 00.0 0H.0 0H.0 HS.HH 00.0 m0.0H mmpmHmo 00090 00.m 00.0 H0.0 0:.0H Hm.: :0.0H cHonm Hs.: 00.0H m0.: 00.0H 00.0 50.0H wchooo 50.: H0.0H 00.0 00.0 :0.0 ~0.0H mmcHempmgms< 0:.: NH.0H H0.: 00.HH 00.0 0:.0H mmpmespHnemm 00.: 0H.0H 0:.m 00.0H 00.: 00.0H mHmNHHHseemsH 00.: :A.0H 0m.0 0N.HH 00.: Hm.0H .HHmm pmspo 00.: HA.HH 00.0 00.0H 00.: 00.0H 0mH 00.: H0.0H 50.0 00.0H 00.: 05.0H mcmsHHems 0H.: 00.0H 0 00.0H 0 00.:H mHogoOH< mSPMPm OflfiocoomOHoom 0.00 0-02 0m-0z .0.m m .0.m m .0.m m mcomHHmmsov pmHHmm How: How: pmmm Hmmssoz MSHQ omscozummb WSHQ pcw mmechpHommo op mmmoo¢ mH mgmHom 0:. 00.: 00. 00.: 00. 00.: mmpmHmo 00000 0 00.: 0:. H0.0 00. 0H.: cHoemm 00. :H.: 0:. 00.: 00. 0H.: mchooo 0m. mo.: Nm. #0.: on. mN.: mwQHewpmsme¢ * 00. 00.: 00. 00.0 00. 00.: *mmpmespHpemm * 00. 00.: 00. 00.0 00. 00.: *mHmNHHHsacmHH 00. 0H.: 00. :H.: 00. 0H.: .HHmm 00000 0:. 0H.: 00. 00.: H0. 00.: 00H * 00. 0H.: 00.H 00.0 00. 0:.: *mcmsHHemz on. 0H.: 0 oo.N o 00.: NHO£00H< mwmoo< 0>Hpomnno 0-00 0-0z 00-00 .0.0 m .0.0 m .0.0 m mcowHMGQEoo pwHHmm Hows How: pmmm HmwSCOZ NSHQ Ap.pcoov mH mqmHom D-Dm Dubz Dm-Dz .Q.m M .Q.m M .Q.m M 0:00H:0Qsoo 00:H0m :00: :00: p000 :00::0z w::m A0.p:oov mH mqmH00 00. 0:.: 00.0 0:.: 00.0 00.: 0000000 00000 H0. 00.0 0:.H 00.: 0:.H 00.: 0Ho000 00.H :0.: 00.H 00.: 00.H 00.: 00H0000 * 0H.H H0.: 00.H H0.0 00.H 00.: 0000H00000000 H0.H 00.: 0H.H 00.: 00.H 0H.: 0000000H0000 00.H 00.: 00.H 00.: 0:.H H0.: 000NHH0000000 00. 00.: :0.H 00.: 00.H :H.: .HH00 00000 00. 00.: ::.H 00.: H0.H 00.: 00H * 00.H 00.: 0:.0 HH.0 H0.H 00.: *00000H000 0:.H 00.: 0 00.0 0 00.: 0H0000H0 000po0m 0H:Q00mosm00Hoom 0-00 0.00 00-00 .0.0 0 .0.0 0 .0.0 0 0:00H00m800 000H0m :00: :00: p00& 0000:0z wane H 000:02-0mb 000Q 0:0 00:0H>0Q :00 th:0p:omgo 0H mqmH00 * * 00.0 :0.00 00.0 00.0H 00.0 H0.0H 000000000 00000 00.: 00.00 00.0 00.0H 00.0 00.0H 000000 0 0 00.0 00.00 HH.0 00.0H 00.0 0H.:H ***00H0000 * * 00.0 :0.00 00.0 00.0H H0.0 00.0H ***000H000000e< * * * 00.0 00.00 00.0 0:.0H 00.0 00.0H ***m000000H0000 * * 00.0 00.0H 00.0 00.00 00.0 00.00 #:000NHHH000000 * * 00.0 00.00 00.0H :0.0H 0H.0 00.0H ***.HH00 00000 0 * H0.0 00.H0 0:.0H 00.00 00.0 00.0H 00:00H 00.0 00.0H 0:.0 00.HH H0.0H 00.0H 000000000 00.0 00.0H 0 00.HH 0 00.00 0H0000H< PQmECOHH>CH Hmfloom 9.3. 03.” mmD WSHQ Um>flmohmm 0-00 0-0z 00-02 .0.0 0 .0.0 0 .0.0 0 0:00H00QEOU 000H00 000: :00: p000 0000:02 000Q H A0.p:oov 0H mHm000 D-D9 51:2 29:32 .Q.m .Q.m M .Q.m M 00000009800 000009 000: 0003 0009 HmwSCOZ mspm AU.PCOUV AH mgm0om * * mn.m m0.m 00.0 mm.m 00.0 00.0 ***mm0m0mo 00000 0B. on.m 0:.0 5:.N No.0 N0.N 000002 * * mm.0 05.m mm.N hm.m in. om.0 ***0000000 * * 20.0 mm.m mm. 00.0 mm. md.0 ***0000E00009E< * * * 20.0 mm.m 00.0 mm.m ow. mm.0 ***00000000D00m * 00.0 00.0 00.0 mm.m om.0 00.0 ***m00N000600000 * * m:.m mm.: mm. 00.0 mm. om.0 ***.0002 00000 * * * mm.m No.0 mm.0 mm.m mm. 00.0 ***200 0000009 000 .0 .0 mm. . 0. . 000500002 0 00cm 00 * * do w: :2 m 02 *** .. 0002 00002 00.0 00.0 o o o 0 00000000 00 000002 M 2:29 2:22 29:22 .2.m M .9.0 M .Q.m M 000000009600 000009 0002 0002 0009 0000002 0002 00000000 0000021002 0002 000 0000009 0000000 0003 00000000000 00 mqm¢e 116 N0. 00. mm. 00. mm. mm. 000000000000 mm. mi. 0m. 02. mm. :m. 00000000000m * mm. mm. mm. 00.0 mm. mm. ***00000H000000B * 00. mm. 00. 00. 00. om. 0.0000 00000 * 0m. 00. 00. 00. mm. mm. 000 00:0 00. 00. 00. 00. 00. 00. 000000002 MWMM0mmsmm mm. 00. 0 00.0 0 00.0 N0000000. 00 000002 m * * mm.N wN.m oo.m mo.N N0.0 00.0 #**000No 0000002 - - 00.0 00.0 00.0 00.0 0000>000 * 0 00.0 00.0 00. 00.0 00.0 00. 0000000000 00200 mm.m 00.0 mm. 00.0 00.0 0N.0 000002 * * mo.m mm.m 0©.N mm.0 mm. no. ***0000000 * * mm.N om.N mm. 20.0 mm. m:. ***0000E00009E¢ * 0 00.0 00.0 00.0 00.0 00. 00. ***00000000n000 * * 00.0 00.0 00. 00. 00.0 00. ***0000000000000 0 * 00.0 00.0 00.0 00.0 00. 00. #000000000000000 00000 0-00 0-02 00-02 .0.0 m .0.0 m .0.0 M 000000009000 000009 0002 0002 0009 0000002 0002 0000000> A0.0000v 00 mqm000 00.0 00.0 00. 00.0 00.0 00.0 0000000 00000 00.0 00.0 no.0 00.0 m0.0 00.0 000000 00.0 00.0 00. 00.0 00.0 00.0 0000000 00.0 00.0 mm. 00.0 00.0 00.0 000000000000 00.0 00.m mm. wo.m mo.0 mm.m 00000000n0mm 00.0 00.0 mm. 0000 00.0 00.0 0000000000000 00.0 00.0 mm. 00.0 00.0 00.0 .0000 00000 0 00.0 00.0 00. 00.0 00.0 00.0 0000 0000wm0 00.0 00.0 00.0 00.0 00.0 00.0 000000000 00MW0W0WM00M 00.0 00.0 0 00.0 0 00.0 N0000000 00 000000 0 00. 00. 00. 00. 00. 00. 00000 0000002 - - 00.0 00.0 00. 00. 0000>000 0 0 mm. 00. 00. 00.0 00. om. 000000000 00000 00.0 00. 00. 00. mm. 00. 000000 00. 00. 00. 00. 00. 00. 0000000 0-00 0-02 00-02 .0.0 0 .0.0 0 .0.0 0 00000000800 0 000000 000: 000: 0000 0005002 ms0m 0000000> AU.PCOUV 00 m0m¢e 118 * * 3N.H mm.H ®N.H m©.H mm. mm. ***.Hawm Hmnpo * * 0:.H mm.H :o.H mm.H mm. mm. ***mmq 902 on mmhflw mH.H :N.H mm.H um.H m:. om. acmsnflpmz ccmwsmm mmm 0H.H mH.H o o o o NH0£00H< m Pcsoe< M mm.fi mm.m :o.H :m.m mm.H mm.m mcflxo msonpflz - - mm.m oo.m mm.H o:.m mpcm>aom * * um.fi Hm.m mu.fi om.m mm.fi mo.m ***mmpmfigo “mayo :H. om.m mm.fi oo.m mm.H mm.m aflopmm * om.fi mm.m om.fi mm.m mo.H mm.H ***mcflmooo * om.H mm.m mm.fi om.m mo.H mm.fi ***mmsflsmpmgge< * * mfi.m mm.m mm.H mo.m mw.fi mm.H ***mmpmp5pfippmm * mu.fi mw.m mm.H H:.m HB.H mo.m *mumafiaflswcmpe * * NB.H mm.m Hm.e mm.m mm.H m©.H ***.HHmm thPo * * :w.H mm.m om.H mm.m m:.H mu.fi ***mmq mwspm * Hm.fi mn.m mm.fi oo.m :0. mm. ***mcmshflpms mucwwmmgwmm Hm.H w:.m o o o o maogooa< m pcsos< m p-3m p-22 pm-pz .m.m m .m.m m .a.m m Hm20wflpwgeoo Umhflmm 9mm: hmm: Pmmm nmmscoz wspm mfipmflpm> AU.PCOUV 0H mgmopmmmmfim m mm. om.H Ho.H 5:.H mfi.a 0H.H mvflxo m:onpflz - - mm.H mo.H 0H.H HH.H mpcm>aom * * om.H mfi.m om. oo.H mo.fi mm. ***mmpmfigo “mayo mm. on.m om. mm.H mH.H mo.H aflopmm * m:.H mm.a mm. m:.H Hm. mm. ***mcflwooo * mm.fi mm.H mo.H NH.H mm. Hm. **mmcH8mpmgm2¢ * * HN.H :o.H mH.H mm.H Ho.fi om. ***mmpmuspflnpmm * * mm.a mm.H HN.H m©.H mm. :5. ***mHmNHHfl:dcth 2-2m p-22 :m-:z .a.m m .a.m w .m.m m Hmcowfimmmsou Umpwmm 9mm: 9mm: Pmmm hmwscoz wSQQ mHQMwhm> AU.PQOUV 0H mflmomm : H mm H mm m d H :H .. z -QMmHo mmm: mm.H mH.H o o o o mHo:00H< m pcsos< m mm. mm.- mm. mm.. mm. mm.- muon msopsz - - on. 03.- mm. mm.- mpcm>Hom m3. mm.. mm. :m.- mm. mm.- mmpmHmo ngpo :H. oH.- ma. mH.- mm. mm.- cHon: 2-:m p-22 pm-:z .m.m m .m.m m .m.m m msomfihmmsoo H Umpwmm pmm: hmm: pmmm pmwscoz mzhm wapmflhm> AU.PEOUV ma mqmHom Dubm Dubz Dmubz .Q.m M .Q.m M .Q.m M wcomHHmmEoo H cmnHmm hmmb Hmmb Pwmm Hmmssoz mama mapmHHm> A@.Pcoov 0H mqmHom * mm.mH 06.6: mm.mH cm.mo Hm.mH 6H.sc *moemHoo Hosea * * mm.m on.mm Hm.oH mm.so oH.mH Ho.mc *sHvom * mm.sH mm.sm sm.w mo.os mm.mH Hm.mc *osHmooo ms.mH mo.mc mo.mH mm.mc om.sH om.mc mosHsmeosesa om.sH oo.am om.oH sm.Ho mm.:H :H.:o mopmsspHpsmm mm.sH me.Ho mm.mH 00.66 om.sH sm.:c onNHHHsssmAe mm.mH mm.mm Hs.mH mm.mm sm.sH mm.mc .HHmm Hoses om.mH Hc.oc sm.sH ss.wm cm.:H mm.mc mmH cm.mH Hc.Ho m:.cH 3:.No sc.mH s:.mc momansmz mm.mH Hm.mc o oo.mm o oo.om NHosooHa p-26 2-:2 sm-sz .m.m m .m.m M .q.m M mcomHHmmsoo voHHmm pom: How: pmmm Homscoz mane H omSCOZIomD mama Ucw HN mocmH>oQ mo oonHoHoe mqmHPHmom mm mgm<9 132 .mmsohm How: pme 020 Monaco: How H n c mm meHmmoQ pom somHHszoo HacHPmHPmpm N .COwHHmQEoo HHOHHoPmoQ m oHHon H Hoo.uv Q *** Ho.nv Q ** mo.uv Q * H * 0H.0m 00.00 00.NH 0s.00 0s.sH 0H.Hm ssocon msoHsz - 0s.HH 0m.0m ss.0H H0.mm meso>Hom 0.00 0-00 00-00 .0.m m .0.m m .0.m m msowHHMQsoo onHmQ How: How: Pwmm Homscoz wSHQ AU.PQODV MN mqmQomHm 00>osom mxHHB P00000Q H00H00000 mpoom H00. v.0.”Hs. u m “00.00 n Ammm.0mvm m.ocm .m0Honesoo 0o spHHmsom 0cm 000. 0H. 00. mm. 0H. 0 0 00. mcon msoHHHz 000. 00. 00.0 00.0 as. 0 0 m:.H mopmHgo H000.”v Hm. sm.0H H0.N mm. 0 mm. 00.H ochooo Hoo. mm. 0H.0 00.m Hm. 5H. 0:. mH.: wocstponQE< Hooo.nv :5. 00.0 00.N m0. 0 mm. 0m.m 009005PHQ00m H000.u. ms. 0H.0 0H.0 00. mm. mm. 0m.m mnosHHHsasoHH H000. v 00. 00.0 00.0 00. 0 00. 00.0 msowosHosHHmm H000. v 00. 00.00 m0.0H 00.: 00.0 00.00 00.00 msmsHH0mz H000.”V 00. 00.00 00.0H 0:.HH 00.Hm mm.HH 00.0w HonooHa cosmoHOHsmHm a AmHH.mvm m..0.m memos memos homo: memos 090H00>HCD mQ5o00 Amdnzv Adanv Amnzv Ammnzv 00H00Q : Q5000 m Q5000 m Q5000 H Q5000 0H00H00> om: 050m 00 hPHHHan0m>uhoco5vo0Q 0H 000000000HQ Q5000 Ho mHmzH00< 9000H8H000HQ mm mqm mPQoHOHmmooo 00Hpoc5m PcmcHsH0owHQ 0000000000Pm AU.PQOUV RN mqmHPHmoQ 000. 00. 00.0 Hm.mH H0.00 H0.H0 HH.00 00.00 oosmH>om Ho 000000H09 H000.v 00. 00.0 00.00 00.0H 00.00 00.00 0H.m0 m oosmH>00 000 00H0H05P00QQ0 N00. 0m. HN.m 00.0 mm.mH oo.mH mm.mH 00.00 N mocmH>0Q 000 moHpHs5p00QQo mm0. mH. mm. 00.H HH.0 ww.m 00.0 00.0 H 0000H>0Q 00% 00HPH05P00QQQ HHH. 00. 00.H 0H.H 00.0 00.0 00.H H0.H mgso00 HmHoomo00 o, ans psoso>H0>cH .m. 000. H0. 00. ms. 0H.0 00.0 00.0 HH.0 nmooo< o>Heoomno 0mm. 0H. 00.H 00.0 00.0H H0.0 00.0H 00.0H 000000 0H8000000Hoom cosmoHHchHw m AmHH.mvm m..m.m 00002 0:002 0:002 0:002 000H00>H00 000000 000nzv H0anv Amnzv Ammuzv 00H00Q 0 Q5000 m Q5000 N Q5000 H Q5000 0H90H00> mOHPmH0090000£0 mpHHmsom00Q 0:0 H005HH500H000 0H 000000000HQ Q5000 00 mHthms¢ pcmcHEH0omHm om mqmHPH00Q mmnm. 0000H>0Q Ho 000000Hoe wMNm. m 0000H>0Q 000 moHpH05P00QQo Hmmo. N 0000H>0Q 000 00HPH05H00QQ0 wNmH. H 0000H>0Q 000 mmeHQ5P00QQo oHoH.I 0Q5000 H0H00000Q :PHB P0080>H0>0H H0mo.u mmmoo< 0>Hpommno mm.HH 05pmpm 0H8000000Hoom H 0H90H00> mponOHHHooo 00H9005m P000H8H000HQ 000H00000090 150 000. 0 H0.0 H00. 0.0 00H. 000. 0 000. 0H 00.0H 000. 0.HH 000. 000. H H000.uv 00 00.00 000. 0.00 000. 000. 0 moCMOHHHQme .Q.Q 00050m mvnsmg 0000B mo 00Hp0H00000 05H0>Qmem 00>osom H30 mxHHz #00000Q H00H00000 . mpoom H000.0v m “00. u 0.100.0H n 0000.0000 0.000 .00Hopeso0 0o spHHmsam 0cm AU.PCOOV on mHmHHH00Q wn.mH HN.m0 Nm.m HN.H0 00.0H HH.00 mm.mH 00.00 0000H>0Q 00 000000Hoe * * 00.0H 00.0H 00.0H 0N.0N 00.0m NN.00 00.0N 0H.m0 m 0000H>0Q 000 00HPH05P00QQ0 * 00.0 00.0H HH.0 00.0H 00.0 00.0H 00.0 00.00 0 00:0H>00 000 00HPH05P00QQ0 0-0 0-0 0-0 0-H 0-H 0-H .0.0 m .0.0 m .0.0 0 .0.0 M H0so0H000000 H00uzv H0Huzv Hmuzv H00nzv 000H0Q 0 Q5000 m Q5000 N Q5000 H Q5000 0H00H00> 00H90H009000000 09HH000000Q 000 H005HH500H000 Hm mgm¢e CH 000000000HQ Q5000 mo mcomH00Q500 H00H00pmoQ < 000000 154 deviance and Group 4 the least tolerant. Thus, overall the pattern of relationships conforms to that expected on the ba- sis of the theoretical framework. The extent to which respondents could be correctly assigned to their respective drug—use groups on the basis of the discri— minant function is summarized in Table 32. Overall, approxi- mately 59% of the subjects were correctly classified. Thus, the single discriminant function based on the single sociocul— tural variable and both of the personality variables was ex- tremely effective in differentiating the respondents in terms of their characteristic patterns of drug use. Finally, the relationship of drug-use group membership to interaction with the subjects' five closest friends is sum- marized in Table 33. In general, these results parallel those of the univariate analyses. The more extensively a respondent is involved with drugs (as indicated by group membership), the more s/he uses drugs with his/her friends, the more s/he uses drugs when his/her friends are not using drugs, and the more s/he uses drugs which his/her friends disapprove of using. Thus, once again the results support the idea of involvement in a drug—using subculture, but do not support the hypothesis that individuals will conform to the pattern of drug use char— acteristic of their friends. In terms of the other specific hypotheses, the results of the multivariate analysis strongly support the hypothesis that individuals who use psychoactive drugs with a relatively high frequency will participate in drug use with other drug users (i.e., Opportunities for Deviance 3) to a greater extent than 155 TABLE 32 Accuracy of Prediction of Group Membership From Sociocultural-Personality Discriminant Function Actual Group Predicted Group Membership I2 1 4 2 3 n (%) n (%) a (%) (%) I3 53 31 (58) 11 (21) 7 (13) 4 (8) 9 1 (11) 6 (67) 2 (22) 0 14 O 1 (7) 10 (71) 3 (22) 47 3 (6) 4 (8) 14 (30) 26 (56) 41‘me Total percentage correctly classified 2 59.3; 32 = 75.29; df = 3; p<.001 individuals with a low frequency of use (Hypothesis 5a). Like— wise, the results support the hypotheses that drug users will be more tolerant of deviance (Hypothesis 7a) and will have more positive expectations and evaluations of drug use than nonusers (Hypothesis 8). Moreover, the obtained results also support the hypothesis that the relationship between drug use and Qpppp- tunities for Deviance 3: Extent pi Use With Friends will be stronger for users of "hard" drugs than users of "soft" drugs (Hypothesis 5b). On the other hand, neither of the hypotheses concerning drug use and access to opportunities were supported. Thus, neither the univariate nor the multivariate analyses con- firm these expectations. Likewise, the hypotheses concerning participation in prosocial groups were not supported to a sta- tistically significant degree, although the pattern of group means was in the expected direction. Finally, although the 156 Hm.H 00.H Nw. H0. 00. 00.H mm.N 00. 00. Nm.H 00.H 0H.H 00. 00. 00. 00.N 00.0 mm. mm. om.H 00.0 00.H 00. N0. 00. Hm.N Nw.N H0. mm. HH.0 00.H 00.H mm. 00.0 00.H 0N.m Hm.N 00. 00.0 00.m ***0m50m 002 0080H0m 080 m PC508< x 0080H0m :PHS P80Qm 08Ha 0050H0H Ho P8508< M 2HQQ 0080H0m 0m 0002 0w50Q 00 000852 m 0H00 0 00 0002 0050a 00 000852 N 00Q0H0Q 000 0 0000 0m 0002 0050a 00 009852 2 0-0 0-0 0-0 0-H 0-H 0-H H 0800H00Q800 000H0Q .Q. 0 Hs0uzv x 0 Q5000 .Q.m Hanzv- m Q5000 .Q.m Ha-ZV - N Q5000 .Q.m M 200nzv H Q5000 0H00H00> mm mamo0gm0 100m 0002 m 005o8< x 0002 m 2080 0050a 00 0000000 2m 00>o000000m M ***902 0D 0000000 0003 0m50m 0002 m 085o8< x :-m :-m m-m 8-0 m-H m-H 0 0000000m8oo 0000mm .Q.m M 208-20 3 Q5000 .Q.m M .Q.m M 2:0nzv Amuzv m @5000 N @5000 .Q.m M Ammuzv 0 Q5000 000000m> A0.PCOUV mm mqm -‘ Present attendance at religious services:, a. Regular b. " Infrequent . c. ' Not at all" ' . Ethnic origin: a. Black b. Spanish/Mexican-American c. Asian d. American Indian e. White f. Other 8. 10. 11. (To 12. 13. 14. 15. 183 Parent's education: Father Mother grammer school or less some high school high school graduate some college college degree post-graduate degree H00.-00"“ . Parent's approximate annual income: a. under $5,000 b. $5,000-10,000 c. $10,000-15,000 d. $1S,OOO-ZS,OOO e. over $25,000 Type of community in which you were raised: (If you are a student, complete questions 12, 13, 14, and 15. If you are not a student, go on to the next farm or rural small town (under 10,000) average town (10,000-99,999) suburb of a city city (100,000-500,000) large city (over 500,000) Occupation page.) be completed by students only) Academic Major Grade Point Average (cumulative): a. below 1.99 b. 2.00-2.49 c. 2.50-2.99 d. 3.00-3.49 .e. 3.50—4.00 Year in school: high school Expected occupation a. b. college freshman —‘_“ M.S.U. c. college sophomore --- -——-— ‘ LOCOC. d. college junior ‘-———— J————— Other e. college senior f. college graduate The next few questions are concerned with howm mong you think different kinds of actions are. MOst peeple think that something like murder isv _gty. wrong, while something like bragging may be considered only slightly_qupg. For the following questions, indicate how wrong you think something is by circling a number from O to 9. Not Slightly Extremely Wrong Wrong Wrong Wrong O l 2 3 4 5 6 _ 7 8 9 If you think an action is Not Wrong, circle 0. If you think an action is Slight1y Wrong, circle 1, 2, or 3. If you think an action is Wrong but not ' extremely wrong, circle 4, 5, or 6. If you think an action is Extremely_Wrong, circle 7, 8, or 9. In each case, circle the one number that indicates exactly how wrong ypg_think an action is. 1. To take little things that do not belong to you. 0123 4.56789 2. To break something that belongs to another person out of spite. O l 2 3 4 5 6 7 8 9 3. To break into a place that is locked just to look around. 0 1 2 3 4 5 6 7 8 9 4. To take something from another person without asking them. 0 1 2 3 4 5 6 7 8 9 5. To do things on the spur of the moment. 0 1 2 3 4 5 6 7 8 9 6. To borrow $5 or so from a friend without really expecting to pay it back. 0 l 2 3 4 5 6 - 7 8 9 7. To damage public or private prOperty that does not belong to you. O l 2 3 4 5 6 7 8 9 8. To lie when applying for a license or job. 0 1- 2 3 4 5 6 7 8 9 10. 11. 12. 1E15 . To go to a movie when you have some work to do. 01234‘562-89 To take something of value from a store without paying for it. 0 1 2 3 4 5 6 7 8 9 To drive over the speed limit on a city street. 012345-61789 To cheat on.an examination. ‘ .0-~1 2 3 t- 5 6 7 8.9 PLEASE READ THE FOLLOWING INSTRUCTIONS CAREFULLY BEFORE COMPLETING THE REMAINDER OF THE QUESTIONNAIRE. On the following pages you will find a series of questions relating to your use of a number of drugs. In all, there are 20 questions about such factors as how often you use certain drugs, how long you have been using them, and your feelings about the use of such substances. On the next page of the questionnaire you will find an answer sheet listing the 13 specific drugs or drug classes we are interested in. Following each drug or drug class there is a row of 20 boxes numbered consecutively from 1 to 20. These box numbers correspond to the question numbers. Each question is followed by a number of possible answers. In answering the questions, place the letter corresponding to your answer in the appropriate box. Thus, if your answer to question 1 is alternative 3) you would place the letter .§_in box number 1. For EACH drug you have used, please answer At; 20 questions. If you have ‘pgt used a particular drug, you should skip questions 4-14 FOR THAT_DRUG. Thus, if you have used every drug, you will answer each question 13 times - once for each drug or drug class. If you have only used some of the drugs or drug classes, you will skip questions 4-14 for the drugs you have Egg used, and answer all 20 questions for those drugs which you have used. .Ho HH Hm pm as Hm Ha Ha Hm Ho No proomou “some. swan. Hunconv z>wHucbz> Amn000..5005v Emu Annuav oammw m>hfidnnzcomzm figmmnmwupm0-02H. mnn.v ambzocHHHwam A5535: 0.55580 on.” .v 0>000000>me Anson. aoauav wZQBdcnmwwimmnosbwp-snowy >20=ma>KHzmm Amemmav oon>Hzm EMwOHz 08mm” 00H>Hmm a00ucst son0ru=0t man.v m>H>Nozw Amucm.mmmonuum. nanny m0finhoc05000m 0 «0002 Asvs 0000 Aomvms Amv0 000V: Amwvmm 00.0 s0. s0. mo. 00000 nso000z ons Ammwm Mmmwmw 0 0V axmwm M0mwmm o o o o Mpso>0om 0 0 0 0 00 s 0m 00 m . . . . 00000 o 00000 ANVN Amvm Avaow o vaN Admva on N mm mm N on .0 000002 Ammvom Anvm Ammvmm 00000 00000 000000 m0.m :0. 00.0 mm. os0noo0 A0mvnN Am0vm0 Avam: AoNvm AONvm Ammvw0 0:.0 0N.N mm.N mm. 000080008080 000v00 000vs0 Aoovsm 00mvs 00000 000v0m mm.m s0.0 ms.0 ms. 000005005000 A0mvsm 000v0 Ammv0m 000V00 00mvs Aomvs0 m0.0 om.m 00.0 m0.m nnon000ssssne 000V00 Ammvmm Awmvms A0mvs Ammv00 Assvo0 00.0 mm.0 00.0 00.0 .0000 nospo AONV00 000V00 ANovdm A0an AmmvN0 0::vm0 000 Awsvmw vas 000vs0 Ammvmm onN Amvm Nm.00 :0.00 00.00 m0.m0 080500002 Ammvmm 00v0 0000 0000080 0 0 00.00 :0.00 0:.00 0m.sm 0osoo0¢ Asvs Asvs 0000 asvs Asvs 00v: .0.0 m .0.0 m 0002 0002 000D 0002 0002 000: 0000 u802 P00m I802 N000080m N00002 00000802 000002 050m 00D 050m 80 0000000000Q 80m mm 00000 "7:11000“