Preventing alcohol, marijuana, and cigarette use among adolescents: Peer pressure resistance training versus establishing conservative norms

PREVENTIVE

MEDICINE

20,

414-430 (1991)

Preventing Alcohol, Marijuana, and Cigarette Use among Adolescents: Peer Pressure Resistance Training versus Establishing Conservative Norms’ WILLIAM

B.

HANSEN,

PH.D.,*,*

AND

JOHN

W.

GRAHAM,

PH.D.t

*Department of Public Health Sciences, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, North Carolina 27103; and tDepartment of Preventive Medicine, School of Medicine, University of Southern California, Los Angeles, California 9ooo7 Background. Two strategies for preventing the onset of alcohol abuse, and marijuana and cigarette use were tested in junior high schools in Los Angeles and Orange Counties, California. The first strategy taught skills to refuse substance use offers. The second strategy corrected erroneous normative perceptions about prevalence and acceptability of use among peers and established conservative groups norms regarding use. Methods. Four experimental conditions were created by randomly assigning schools to receive (a) neither of the experimental curricula (placebo comparison), (b) resistance skill training alone, (c) normative education alone, or (d) both resistance skill training and normative education. Students were pretested prior to the program and post-tested 1 year following delivery of the program. Results. There were main effects of normative education for summary measures of alcohol (P = O.OOll), marijuana (P = 0.0096), and cigarette smoking (P = 0.0311). All individual dichotomous measures of alcohol, marijuana, and tobacco use indicated significant reductions in onset attributable to normative education. There were no significant main effects of resistance shill training. Conclusion. These results suggest that establishing conservative norms is an effective strategy for preventing substance use. 0 1991 Academic Press, Inc.

INTRODUCTION

Alcohol, tobacco, and marijuana have long been known to be drugs of first use (1, 2). Alcohol and tobacco are known to increase long-term risk of disease and premature mortality (3-5). Alcohol use among youth is a risk factor for morbidity and mortality due to accidents (4). In the nation’s current war on drugs, educational programs frequently target these three substances for primary prevention. Research to date has shown great promise for preventing cigarette use among adolescents (620). Fewer studies that have targeted marijuana use have been conducted (9, 11, 12, 14, 16, 21). Nonetheless, there is evidence that programs that prevent the onset of cigarette use also curtail the onset of marijuana use. To date, alcohol use and misuse among adolescents have not been reliably reduced by prevention programming (9, 13, 16, 22-27) although several recent studies suggest small reductions in alcohol use due to preventive efforts (11, 14, 28, 29). Several recent reviews and research reports have expressed serious doubt that ’ Supported by a grant (ROl-AAO6201) from the National Institute on Alcohol Abuse and Alcoholism. 2 To whom reprint requests should be addressed. 414 0091-7435/91$3.00 Copyright ‘0 1991 by Academic Press, Inc. AU rights of reproduction in any form reserved.

SUBSTANCE

USE PREVENTION

415

school-based alcohol prevention is feasible given the widespread acceptance of the behavior (9, 30, 31). Most substance use prevention programs have utilized eclectic strategies, typically with multiple components that defy precise theoretical interpretation. Programs that have been successful to date have tended to include instruction about the nature of peer pressure and skills to resist peer pressure as one component (32). Indeed, even programs that include a number of intervention techniques often ascribe their successprimarily to the impact of teaching students to resist pressure. Nonetheless, the lack of reliable effects on preventing alcohol use and the continual mixing of multiple strategies leads one to question whether peer pressure resistance training, as the single strategy, can account for preventive effects. Particularly, it is possible that it is the establishment of conservative norms regarding use rather than resistance skills that accounts for lower use among particular subgroups (33). Ironically, peer pressure resistance training programs have been consistently confounded with other programmatic strategies that have the potential to establish conservative norms through the correction of erroneous normative beliefs (34). To date, two distinguishable strategies for addressing the issues of social i&hiences have been developed. The first strategy to be tested involves teaching skills to resist peer pressure to use alcohol and other substances. This strategy was developed in responseto the known statistical relationship between substance use among friends and the onset of use among individuals who associate with friendship groups who use (34). Implied by this relationship, and validated to some extent by an examination of the process of onset (35), is the idea that young people perceive peer pressure to participate in substance use in order to be accepted by the peer group. To counter this pressure, programs teach young people to identify situational pressuresto conform to group behavior and teach skills to refuse offers and demands to use while maintaining group membership. The second strategy involves the correction of erroneous perceptions about the prevalence and acceptability of alcohol use. The roots of this strategy are derived from the propensity of young people to overestimate actual prevalence of all forms of substanceuse (34,364O). This overestimation may lead young people to expect use to be normative when, in fact, it is not. This may create an internally generated expectation that the reference group will find use of substances to be desirable and appropriate. To counter this, programs provide students with feedback about actual rates of use and publicize the conservative attitudes toward substance use that exist in the peer group. The Adolescent Alcohol Prevention Trial is an g-year project the purpose of which is to explore the potential of two strategies for deterring the onset of substanceuse in youth. This study involves replicating longitudinal tests of these two strategies in four separate cohorts of students, two cohorts of students that receive programming during the fifth grade, and two that receive initial programming during the seventh grade. This report concerns only the first cohort of seventh-gradestudents. The primary focus of the study was on preventing alcohol use and misuse among adolescents. Cigarettes and marijuana, the use of which correlates with alcohol consumption (l), were also addressed.

416

HANSEN

AND

GRAHAM

METHODS Subjects and Setting

During the 1987-1988school year, 12junior high schools in Los Angeles and Orange Counties, California, were recruited to participate in the study. Prior to delivery of the program, seventh-gradestudents and parents gave informed consent. Experimental Design

Schools were stratified by size, test scores, and ethnic composition and then randomly assigned to receive one of four intervention programs (41). The first program (Information) consisted of four 45-min lessonsabout the social and health consequencesof using alcohol and other drugs. The second program (Resistance Training) consisted of four lessons about the consequences of using substances plus five lessons that taught students to identify and resist peer and advertising pressure to use alcohol and other substances. The third program (Normative Education) included four information lessons plus five lessons that corrected erroneous perceptions of the prevalence and acceptability of alcohol and drug use among peers and established a conservative normative school climate regarding substance use. The fourth program (Combined) included three lessons about information, three and one-half lessons teaching resistance skills, and three and one-half lessons establishing conservative norms. Components of these programs are described in Table 1. These four programs result in two main effects that may be tested: normative education vs no normative education and resistance training vs no resistance training. Programs were delivered entirely by project staff. Program specialists devoted full time effort to teaching. Each had received a minimum of 2 weeks of intensive training. Several had helped with the initial development of both programs and had several years of experience in delivering these types of programs prior to their participation in this study. Previous analyses of the quality of implementation indicate high fidelity of performance (42). Measures

Students were pretested using a questionnaire that assessedtheir use of alcohol, marijuana, and cigarettes. For each of these substances, subjects were asked to report on cumulative lifetime consumption and use during the 30 days prior to the survey. For alcohol, students also reported the number of times they had previously been drunk and their alcohol consumption during the past 7 days. For alcohol measures,responseswere framed in terms of numbers of drinks. Cigarette responseswere framed in terms of numbers of cigarettes. For example, responses to cumulative lifetime alcohol consumption included (a) none or sips (for religious service), (b) sips (not for religious service), (c) part or all of one drink, (d) 2 to 4, (e) 5 to 10, (f) 11 to 20, (g) 21 to 100, and (h) more than 100. Similar scales were constructed for each alcohol and tobacco item. Marijuana items used dichotomous (yes/no) responses.Subjects also reported whether they had discipline problems at school during the previous year and whether they had engagedin violent

Definition of drugs, alcohol is a dw Film/discussion-The Glug (Churchill Films)

Definition of drugs, alcohol is a drug Positive and negative, shortterm, and social consequences of drinking using Socratic Method Parent interview homework, consequences of drinking Prevention Baseballquestion-answer game about consequences of drinking

Question Box-answers to student-generated questions about alcohol and drugs

1

2

4

3

ICU

Session Definition of drugs, alcohol is a dw Positive and negative, shortterm, and social consequences of drinking using Socratic Method Parent interview homework, consequences of drinking Review of consequences, homework Types of social pressure Techniques to refuse Refusal practice Parent interview about peer pressure Assertiveness Role-played refusals

Definition of drugs, alcohol is a dw Positive and negative, shortterm, and social consequences of drinking using Socratic Method Parent interview homework, consequences of drinking Review of consequences, homework Survey about prevalence of alcohol use among students

Survey results Agree/disagree opinion statements Parent interview homework, appropriate/inappropriate alcohol use Class opinions about alcohol use continuum Class discussion of appropriate/ inappropriate alcohol use Nondrinker interview homework-advice

Pressure refusal practice Film/discussion-The Glug

RT

Norm

Curriculum

TABLE 1 PROGRAMCONTENTOFFOVRCVRRKVLA

Types of social pressure Techniques to refuse Refusal practice Parent interview about peer pressure

Survey results Agree/disagree opinion statements

Definition of drugs, alcohol is a drug Positive and negative, shortterm, and social consequences of drinking using Socratic Method Parent interview homework, consequences of drinking Review of consequences, homework Survey about prevalence of alcohol use among students

Combined

Techniques used by advertisers Evaluating alcohol ads Parent/child homework about alcohol advertising Review of homework Develop anti-alcohol advertisements Question box Videotaped commitments to resist pressure to drink alcohol

Developing positive friendships, including nondrinking as a positive quality Prevention Baseball Alcohol rap songs-informal rules about alcohol Question box Videotaped rap songs and personal opinion statements

6

I

8

9

10

Standing up to pressure-report of personal resistance experiences Prevention Baseball

RT

Review of homework Film/discussion-The Glug

Norm

5

Session

Curriculum

TABLE l-Continued

Assertiveness Role-played refusals Parent interview homework, appropriate/inappropriate alcohol use Class opinions about alcohol use continuum Class discussion of appropriate/ inappropriate alcohol use Nondrinker interview homework--advice Techniques used by advertisers Evaluating alcohol ads Parent/child homework about alcohol advertising Alcohol rap songs-informal rules about alcohol Question box Standing up to pressure-report of personal resistance experiences Videotaped rap songs, personal opinion statements and commitments to resist pressure to drink

Combined

SUBSTANCE USE PREVENTION

419

or destructive behavior. Students were then asked whether participating in these problem behaviors occurred when under the influence of alcohol. Subjects were post-tested using the same survey instrument during their eighth-grade year. Students were identified only by coded number. At each administration of the survey, students were reminded that the information they provided would remain confidential and that their nameswould never be associateddirectly with the data. Students were encouraged to report honestly. Because alcohol was the primary focus of the study and because no biological measure of alcohol was appropriate for adolescents, no “bogus pipeline” procedure was employed as has been done previously in tobacco research (43). With strong assurancesof confidentiality and multiple experiences with data collection activities, honest reporting was expected to be improved (37, 44). Validity of self-reports was not expected to vary by condition. Statistical Analysis

Only data for students present at pre- and post-test were included. For cigarettes, alcohol, and marijuana, use indices for individuals were created by averaging statistically standardized3 items. Reliability for each scale was calculated using Cronbach’s (Y.For the alcohol index, Cronbach 0~‘swere 0.84 for seventhgrade and 0.86 for eighth-grade reports. For the marijuana index, 0~‘swere 0.58 and 0.60 for seventh- and eighth-grade measures,respectively. For tobacco use at seventh and eighth grades, a’s were 0.81 and 0.87. Each use item was also dichotomized. For alcohol, the cumulative life use and 30-day and ‘I-day use items were dichotomized with responsesof never used, only sips, and only sips for religious purposes coded as 0, and more than sips coded as 1. For cigarettes, dichotomization was based on no use or only taking puffs vs more than puffs (whole cigarettes or more). Drunkenness and marijuana use (ever and past 30 days) items were dichotomized with a non-vs-any split. Schools were the unit of assignment; however, the unit of delivery was the classroom. Because reception of the program varied at the classroom level, and becausethe power for detecting differences with school as the unit of analysis was prohibitively low, classroom of program delivery was selected as the unit of analysis. For each index and dichotomized item, classroom means were created by averaging individual reports by classroom. The classroom-level standardized indices are relatively uninterpretable in terms of incidence and prevalence but provide an accurate overall measure of consumption. For dichotomous items, classroom means represent the proportion of individuals who were above the cut-point. A general linear model analysis of covariance approach was used with classroom means for each composite index and for each dichotomous item. Analyses tested for main effects of ResistanceTraining (ResistanceTraining and Combined programs vs Information and Normative Education programs) and Normative 3 Standardization involved statistically adjusting each item so that the mean equaled 0 and the standard deviation equaled 1 using PROC STANDARD in SAS.

420

HANSEN AND GRAHAM TABLE 2 SUBJECT CHARACTERISTICS Rates by program Information only ww--

Resistance Normative training education Program plogram

F values Resistance Combined training program (main effect)

Normative education (main effect)

RT x NE (interaction)

Sex

Female Ethnicity Asian Black Hispanic white

SES Family income * P **p ***p t P <

48.8%

49.7%

55.2%

55.4%

0.04

5.9s

9.8% 3.0% 42.6% 33.2%

16.4% 1.7% 21.7% 52.2%

25.9% 0.8% 11.3% 52.2%

17.5% 2.0% 30.5% 38.8%

0.16 0.02 0.11 0.80

12.82*** 2.33 19.52t

0.84

n.45t

$22,140

$24,874

$23,291

$21,846

3.85

8.15**

40.44t

0.02 9.7998 3.91'

62.13

< 0.05. c 0.01. < 0.001. o.ooo1.

Education (Normative Education and Combined programs vs Information and Resistance Training programs) and for interactions between the two programs. Pretest scores associated with post-test variables were used as covariates. Data for sex, ethnicity, and family income4 by condition are presented in Table 2. Additional analysis of covariance were completed with sex, ethnicity, and family income included as covariates. Classroom means were calculated for all demographic variables, including the proportions of each ethnic group in each classroom. Attrition was examined via x2 analyses. An a level (two-tailed) of 0.05 was established for all analyses. RESULTS Sample Characteristics

Of the total sample(N = 3,011)who were pretested, 2,135 (70.%) were located and post-tested in program schools. An additional 77 students who were absent at the time of testing subsequently completed the survey when they returned to school. Of those not located in the program schools, 204 were successfully contacted either at their new schools (N = 188)or at home (N = 16). Thus, 80.2% of eligible students (ZV= 2,416) were successfully tracked and tested. Of the 595 subjects who did not participate in repeated measurement, 330 (11.O%)were located but were excluded becauseof personal or parental decline at the post-test or because they failed to return informed consent at the post-test. The remainder were absent without completed survey (N = 37), had moved from schools involved in the study and could not be located (N = 123),or did not return mailed follow-up surveys (ZV= 105). Thirty-two classrooms received the Information program; 33 received Resis4 Family subjects.

income was estimated from census data corresponding

to the zip code of residence

of

421

SUBSTANCE USE PREVENTION

tance Training; 27 received Normative Education; and 26 classroomsreceived the Combined program. All classrooms were represented at pre- and post-test. Programmatic Outcomes

Analyses of standardized indices revealed that for all three substances, there was a main effect of Normative Education (Table 3). Compared with classrooms that received no Normative Education, those that did receive a version of Normative Education had significantly reduced rates of alcohol consumption, marijuana use, and cigarette consumption. Adding sex, ethnicity, and family income as covariates did not change the pattern of results for alcohol. The only significant covariate was family income (F = 8.48; P = 0.0043).None of the ethnicity proportions was a significant covariate (all F’s < 0.7; P’s > 0.40). There was still a main effect of Normative Education on drinking (F = 9.23; P < 0.0030). Similar results were obtained for the tobacco index; family income was a significant covariate (F = 10.62;P = 0.0015) and the main effect of Normative Education was preserved (F = 6.25; P = 0.0139). Family income was also the only statistically significant additional covariate to predict the marijuana use index (F = 7.59; P = 0.0069). The main effect of Normative Education was significant (F = 4.29; P = 0.0408), and for marijuana use there was a significant interaction between Normative Education and Resistance Training (F = 6.75; P = 0.0107).An examination of the least-squaresmeans suggeststhat the Resistance Training program was least effective in preventing the onset of marijuana use, whereas the Combined program had the greatest effect at reduced onset. The Normative Education and Information programs had intermediate effects. Analyses of averaged dichotomous items provide more interpretable results insofar as the impact of program on incidence and prevalence is concerned. In each of these analyses,pretest use and family income were included as covariates. These analyses yield a similar overall pattern of results (Table 4) with Normative Education most consistently reducing the onset of substance use. TABLE 3 RESULTS OF ANALYSIS OF COVARIANCE (F VALUES) FOR ALCOHOL, MARIJUANA, AND CIGARETTE INDICES AT POST-TEST WITH PRETEST USE AS THE COVARIATE

Pretest (cob&ate)

Index Alcohol Marijuana Cigarettes *P ** P *** P tP<

< 0.05. < 0.01. < 0.001. O.oool.

91.35t 16.92-r 65.74t

Resistance training (main effect)

Normative education (main effect)

Resistance Training x normative education (interaction)

0.60 0.48 0.59

11.24*** 6.93*** 4.76*

0.21 1.21 0.04

422

HANSEN AND GRAHAM TABLE 4

RESULTS OF ANALYSIS OF COVARIANCE (F VALUES) FOR ALCOHOL, MARIJUANA, AND CIGARETTE DICHOTOMOUS ITEMS AT POST-TEST WITH PRETEST USE AND FAMILY INCOME AS COVAIUATES

Pretest (covariate)

Index Alcohol Ever use 30-Day use 7-Day use Drunkenness hoblem use Marijuana Ever use 30-Day use Cigarettes Ever use 30-Day use * ** *** t P

P P P <

Family income (covariate)

113.51t 18.18? 10.11** 49.92t 16.13t

3.72 9.93* 0.68

33.34-t 1.os 33.16t 22.01t

2.62

Resistance training (main effect)

Normative education (main effect) 8.17**

Resistance training X normative education (interaction)

3.22 0.58 0.88 0.03

5.93*

1.90

25.19t

2.24

0.11

4.26*

0.01

2.62 7.48**

3.12 0.71

9.57** 4.59*

1.85 6.20*

6.29* 4.50*

1.43 0.66

5.43* 4.71*

0.63

1.99

5.10*

0.76 5.47*

1.60

< 0.05. < 0.01. < 0.001. 0.0801.

For alcohol, the strongest program effect was observed in delaying the onset of ever being drunk (Fig. 1). From pretest to post-test, the incidence of drunkenness in non-Normative Education classrooms was 1l.l%, whereas a 4.2% increase in the incidence of drunkenness in Normative Education classrooms was observed.

Q 25% z

ac

20%

B ; 15% Q 2 z 10% 5 e n 5%

8TH

+I GRADE

FIG. 1. Changes in incidence of ever being drunk between seventh and eighth grade among students receiving four prevention programs.

423

SUBSTANCE USE PREVENTION

This difference reflects a 62.4% reduction in the rate of onset of drunkenness attributable to Normative Education. Students ever using alcohol was also strongly affected by the Normative Education program. Initially, 28.7% of students reported ever having consumed more than a sip of alcohol. The incidence of ever using alcohol in classes that did not receive Normative Education increased 14.2% between seventh and eighth grades. The averageincrease for students who received Normative Education was 11.O%,a 22.5% relative reduction in the incidence of initial drinking attributable to Normative Education. At pretest, the prevalence of weekly alcohol use was 4.5%. The pre- to post-test increase in prevalence of weekly alcohol use among classrooms that did not receive Normative Education was 5.1%; among those who did receive Normative Education, the increase was 2.8%, a 45.1% relative decrease in incidence. Pretest prevalence of 30-day alcohol use was 8.1%. The seventh to eighth grade increase in prevalence of past month drinking was 9.1% in classrooms that did not receive Normative Education, whereas a 6.2% increase was observed in classes that received Normative Education, a 31.9% relative decrease. The interaction between Normative Education and ResistanceTraining (interpreted by examining least-squaresmeans)was the result of the Combined program being superior to all other programs and the Resistance Training Only program being inferior. Increases in problems attributed to alcohol during the past year were also reduced by Normative Education. Problem use was very low among this age group (overall pretest prevalence of problems during the past year was only 1.4%). From seventh to eighth grade, the non-Normative Education classes increased an average of 2.4%, whereas Normative Education classes increased an average of only 0.3% in the prevalence of problem alcohol use (Fig. 2). This INFORMATION RESISTANCE TRAlNlNG e-sem. NORMATIVE EDUCATION -.w.COMBINED P!?QG!4AM Me-

7TH

8TH GRADE

FIG. 2. Prevalence of problems attributed to alcohol use during the previous 12-month period for seventh- and eighth-grade students receiving four prevention programs.

424

HANSEN AND GRAHAM

represents.a87.5% decreasein the rate of onset of problems resulting from alcohol use. Incidence of marijuana use was also strongly affected by the Normative Education program (Fig. 3). Classesthat did not receive Normative Education had an increased incidence of marijuana use of 6.2% between pre- and post-test. Incidence among Normative Education classesincreased 2.2%, a relative reduction of 64.5% in the rate of incidence. Thirty-day marijuana use also had a significant main effect of Normative Education. At post-test, 1.7% of students who received Normative Education reported using marijuana compared with 2.7% of those who had not received this program. An examination of least-squares means indicates that the significant interaction between Normative Education and Resistance Training was the result of the Combined program being clearly superior to all other programs in preventing the prevalence of monthly marijuana use, whereas the Resistance Training Only program was least effective at preventing onset. In fact, comparing Resistance Training Only and Information Only conditions suggests that Resistance Training, by itself, may have been slightly harmful. Dichotomous measures of cigarette consumption maintain the trend of lower use among Normative Education classes.Normative Education classeshad lower rates of ever smoking cigarettes (8.1% vs 10.3%)and of having smoked during the 30 days prior to the post-test (4.8% vs 6.5%). There was no difference in rates of attrition comparing students who received and did not receive Normative Education (x2 = 0.07; P = 0.79). However, there was differential attrition when comparing ResistanceTraining and non-Resistance Training conditions (x2 = 6.07; P = .Ol). The rate of attrition among Resistance Trained students was 18.0% vs 21.6% among those who did not receive Resistance Training. This difference is not large in absolute terms. However, since 14%

INFDRt.U5lON RESlSTANCETRAlNlNG -e-m*.

Q f2% z

NORhWTlVE EDUCAT!ON -.-.-

k 10% 0 8 a 8% :

5 3 0 5 n

CDMBlNED PRCXSIAM me-

8% 4% 2%

7TH

8TH GRADE

FIG. 3. Changes in incidence of ever using mtijuana students receiving four prevention programs.

between seventh and eighth grade among

SUBSTANCE

USE PREVENTION

425

attrition is related to drug use status (45), a comparison of Resistance Training vs no Resistance Training may slightly underestimate the potential benefit of the program in preventing substance use onset. Since main effects of Resistance Training did not even aproach significance, the interpretation of findings is not threatened. DISCUSSION

A program designed to correct erroneous perceptions among students about the prevalence and acceptability of alcohol, marijuana, and cigarettes significantly deterred the onset of use of these substances.In contrast, a program designed to teach seventh-gradestudents techniques to refuse offers to use substanceshad no discernable positive impact on substance use behavior. As the nation moves toward identifying strategies that are useful for preventing substance use among adolescents, these results provide initial insights about possible effective school-basedstrategiesfor preventing substanceuse. Although this study is exploratory and requires replication, these results may nonetheless be useful in helping the field of substance abuse prevention define its objectives. It has long been suspectedthat peer pressure is a major causeof the onset of use of commonly used substances.Numerous programs that have featured skill building to resist peer pressure have been developed. Popularly known as the “just say no” approach to prevention, this strategy has become a common component of substance use prevention curricula derived to students. The program component that was used for teaching resistance skills training has had a long history of development and refinement. Particularly, the skills training component used in this project evolved directly from Project SMART (12) with an increased emphasis on realistic and structured role play and is the component that is used as the basis of Project STAR and Project I-STAR (16). This component has been shown to be effective in increasing students’ knowledge of peer pressure resistance (39) and skill at resisting offers (46, 47). Our results indicate that this approach, particularly when delivered without also establishing conservative behavioral norms, is not effective in deterring the onset of substanceuse. Our combined experience suggeststhat this program was equal in content and well-delivered. Why it should fail, other than becauseof a lack of relevance of skills to prevention, is unknown. While a major emphasis of prevention has been on teaching skills to resist pressure, even the earliest psychosocial prevention studies (48) included feedback about prevalence rates. Peer pressure resistance training and normative education have been consistently confounded. The normative education program sought to deter experimentation with alcohol, marijuana, and tobacco by establishing conservative group norms regarding use. This program underwent extensive elaboration and revision from what was included in previous studies (39). The intervention sought to correct erroneous group perceptions of prevalence and acceptability. The results of this study indicate that establishing conservative normative expectations and beliefs effectively reduce the demand to use substancesand may be key to preventing substance use in our society. It should be noted that our approach to establishing conservative norms was not prescriptive but instead relied exclusively on Socratic and devil’s advocacy meth-

426

HANSEN

AND

GRAHAM

ods. As a result of responsesto curriculum-generated discussions, both students and their parents almost universally agreed that abstinence was the only acceptable course regarding alcohol use. Students and parents were clearly opposed to drunkenness. Contrary opinions were dealt with as the focus of group debate which, by virtue of the conservative tendency of the participants, strengthened group perceptions of intolerance toward alcohol and drug use. Since the information came directly from the students and parents and not the teachers, the credibility of the information is expected to have been maximized. Problems resulting from alcohol use were also greatly reduced by the normative education program. Because problem use had clearly not yet emerged in the age group being studied, it is premature to conclude the changing norms will have a clinically significant effect on problem alcohol use. Nonetheless, these results provide an initial hope that will be testable as this cohort of subjects matures. Marijuana and cigarette smoking received relatively less attention in the curriculum. Nevertheless, the incidence and prevalence of marijuana use and cigarette smoking were clearly altered by the normative education program. The results regarding use, particularly drunkenness and problem alcohol use, are encouraging. However, it must be noted that these results are based on selfreported behavior. For drunkenness in particular, there may be problems in measurement validity. Seventh- and eighth-grade students may not have an accurate conceptualization of drunkenness. Given the fact that body size and experience with alcohol are important factors that contribute to evaluating one’s mental state and that these are changing for this group, the validity of self-reports is a concern. This study relief on assurancesof confidentiality to obtain honest reports from students. Previous research has indicated that it is primarily initial data collection sessions that suffer from underreports, and subsequently it may be threatening conditions that may foster false self-reports (43,44). Nonetheless, students in this study may have under- or over-reportedtheir alcohol and drug use. In a separate report (37) we examined the relationship between self-reports and independently assessedestimates of subjects from their best friends. In this report, there was evidence that there was corroboration between the two sources of information. These data provided evidence that equals or slightly improves the correspondence between self-report and previously tested biological markers (43). We conclude that these self-report data reflect reasonably valid responses by subjects. We speculate that the reason for the differences attributable to normative education was that the primary causes of alcohol and drug use have to do with the social availability of and internal expectations about the social acceptability of use of these substances.Among young adolescents, alcohol, marijuana, and tobacco are consumed primarily in social settings, particularly at parties and informal get togethers. Among groups of students where underlying conservative beliefs have been revealed, the availability of alcohol at social functions will be reduced since hosts will feel providing alcohol will be unacceptable to guests. Offers to use substancesin all settings will be reduced due to a perceived intolerance toward such offers. To the extent that other types of curriculum components support conservative normative beliefs, they should contribute to the approach fostered by such pro-

SUBSTANCE

USE PREVENTION

427

grams. Therefore, information program components that are delivered in conjunction with programs that reduce students’ perceptions of prevalence and acceptability should be expected to enhance these programs. One potential explanation of the failure of the Resistance Training program for reducing the prevalence of marijuana use is that focusing solely on techniques to resist pressure may actually increase students’ perception of prevalence+if there is pressure to use, it must be common and acceptable to others. We speculate that approaches that unintentionally augment perceived social acceptability will be successfulat prevention. It is possible that the mechanismaccounting for program successin other trials may be related to norm restructuring even when other mediating processesare thought to be operating. Prior research that has focused on alcohol has yielded inconsistent results. Indeed, only a few field studies have demonstrated any reduction in the onset of alcohol use (11, 12, 16, 28, 29,49). The focus of this program was on preventing the onset of alcohol use and misuse. The greatest effects of the program were on changing the rate of incidence of drunkenness. Drunkenness represents the most extreme form of alcohol consumption for adolescents. From the perspective that it was easiest to define drunkenness as an aberrant behavior, the relatively large impact of the normative approach on this behavior is logical. Since drunkenness is particularly bothersome as a risk factor for accidents and later heavy involvement in alcohol, these results are gratifying. Nonetheless, additional research is needed to determine whether or not these findings can be replicated. The samplewas not equivalent across groups, which is a causefor concern. The Normative program schools were overrepresented with Asians and underrepresented with Hispanics. This nonequivalence was modeled by including classroom proportions of each ethnic group as covariates. In no analysis was ethnicity a significant covariate, nor did including ethnicity proportions in the model alter the main drug use findings. This suggests that the effectiveness of the Normative Education program was not the result of ethnic differences. Thus, while the ethnic imbalance of groups could be troublesome, it did not manifest itself in any way that allows one to conclude that the results observed were due to this imbalance. On the other hand, family income was a significant covariate. However, including this covariate in analysesactually strengthenedthe effect of Normative Education and did not affect Resistance Training at all. Controlling for inequalities statistically reduces selection bias as a threat to internal validity, yet some higher-order effect may remain, perhaps a treatment by selection interaction such that normative education programs are most effective among Asians and poor families. While this will remain a possibility, we found no statistical evidence to support this case. The findings of this study suggestthat the key to changing substance use patterns requires changing sociologic factors that account for the onset of use. Where norms exist that proscribe use, social pressure to use is effectively reduced. Particularly, establishing conservative norms withint adolescent friendship groups may effectively reduce demands that otherwise may serve to promote substance use. Even students who are individually well equipped with skills for dealing with social pressure may ultimately fail to use such skills unless significant social support for refusing offers to use substancesexists.

428

HANSEN AND GRAHAM

ACKNOWLEDGMENTS We appreciate the assistance of the following in conducting this project: School Districts-Garden Grove Unified School District and Downey Unified School District; Research Assistants-Nancy Collins, Luanne Rohrbach, and Bonnie Wolkenstein; Health Educators-Kelly Rippentrop, Mamey Barovich, Ginger Hahn, Dana Mann, Anne Raynor, and Laura Ross; Data Collectors-Kathie Heller, Paula Nichols-Rodriquez, Gaylene Gunning, Sallye G’Guynn, Greg Anapol, Betty Beck, Mario Berrerra, Steve Blackman, Eliezer Bonilla, Judith Bowman, Antoinette Carillo, Carl Anthony Craig, Sue Fortune, Sarah Gildea, Jim Green, Dan Guerrerro, William Howells, Jean Liverman, Ray Palmer, Jenny Pierce, Lisa Potts, Manual Rodriguez, Flor Santa Ana, and Marcos Valdez; Support StaffRoberta Searl.

REFERENCES 1. Hansen WB, Graham JS, Sobel JL, Shelton DR, Flay BR, Johnson CA. The consistency of peer and parent infhtences on tobacco, alcohol, and marijuana use among young adolescents. J Behav Med 1987; l&559-579. 2. Kandel DM. Stages in adolescent involvement in drug use. Science 1975; 190:912914. 3. USDHHS. Alcohol consumption and related problems. Alcohol and Health Monograph 1. USDHHS, PHS, ADAMHA, NIAAA. DHHS Pub. No. (ADM) 82-1190, 1982. 4. USHHS. Sixth Special Report to the U.S. Congress on Alcohol and Health from the Secretary of Health and Human Services. USDHHS, PHS, ADAMHA, NIAAA. DHHS Pub. No. (ADM) 87-1519. 5. USDHHS. The Health Consequences of Smoking: Nicotine Addiction. A Report of the Surgeon General, 1988. USDHHS, PHS, CDC, CHPE, OSH. DHHS Pub. No. (CDC) 888406, 1988. 6. Botvin GJ, Eng A. The efficacy of a multicomponent approach to the prevention of cigarette smoking. Prev Med 1982; 11:199-211. 7. Botvin GJ, Eng A, Williams CL. Preventing the onset of cigarette smoking through lie skills training. Prev Med 1980; 9135-143. 8. Del Greco L, Breitbach L, Rumer S, McCarthy RH, Suissa S. Four-year results of a youth smoking prevention program using assertiveness training. Adolescence 1986; 21:631&O. 9. Ellickson PL, Bell RM. Drug prevention in junior high: A multi-site longitudinal test. Science 1990; 247~1299-1305. 10. Flay BR, Ryan KB, Best JA, Brown KS, Kersell MW, d’Avemas JR, Zanna MP. Are social psychological smoking prevention programs effective? The Waterloo study. J Behav Med 1985; 8:37-59.

11. Graham JW, Johnson CA, Hansen WB, Flay BR, Gee M. Drug use prevention programs, gender, and ethnicity: Evaluation of three seventh-grade Project SMART cohorts. Prev Med 1990; 19305-313. 12. Hansen WB, Johnson CA, Flay BR, Phil D, Graham JW, Sobel J. Affective and social influences approaches to the prevention of multiple substance abuse among seventh grade students: Results from project SMART. Prev Med 1988; 17:135-154. 13. Hansen WB, Malotte CK, Fielding JE. Evaluation of a tobacco and alcohol abuse prevention curriculum for adolescents. Health Educ Q 1988; 15:93-114. 14. McAlister AL, Perry C, Killen J, Slinkard LA, Maccoby N. Pilot study of smoking, alcohol, and drug abuse prevention. Am J Public Health 1980; 70~719-721. 15. Murray DM, Davis-Heam M, Goldman AI, Pirie P, Luepker RV. Four and five year follow-up results from four seventh-grade smoking prevention strategies. J Behav Med 1988; 11:395-405. 16. Pentz MA, Dwyer JH, MacKinnon D, Flay BR, Hansen WB, Wang EYI, Johnson, CA. A multicommunity trial for primary prevention of adolescent drug abuse: Effects on drug use prevalence. JAMA 1989; 261332593266. 17. Schinke SP, Gilchrist LD, Snow WH. Skills intervention to prevent cigarette smoking among adolescents. Am J Public Health 1985; 75:665-667. 18. Telch MJ, Killen JD, McAlister AL, Perry CL, Maccoby N. Long-term follow-up of a pilot project on smoking prevention with adolescents. J Behav Med 1982; 5:1-8.

SUBSTANCE USE PREVENTION

429

19. Tell GS, Klepp K, Vellar OD, McAlister A. Preventing the onset of cigarette smoking in Norwegian adolescents: The Oslo Youth Study. Prev Med 1984; 13:2X-275. u). Va&inen E, Pallonen U, McAlister A, Koskela K, Puska P. Effect of two years of educational intervention on adolescent smoking (The North Karelia Youth Project). WHO 1983;61:X%532. 21. Moskowitz JM, Schaps E, Malvin J, Schaeffer G, Condon JW. An Evaluation of an Innovative Drug Education Program: Follow-up Results ERIC Dot 231739, 1981. 22. Duryea EJ, Okwumabua JO. Effects of a preventive alcohol education program after three years. J Drug Educ 1988; 18:23-31. 23. Goodstadt MS, Sheppard MA. Three approaches to alcohol education. J Stud Alcohol 1983; 44~362-380. 24. Goodstadt MS, Sheppard MA, Chan GC. An evaluation of two school-based alcohol education

programs. J Stud Alcohol 1982; 43:352-369. 25. Hopkins RI-I, Mauss AL, Keamey KA, Weisheit RA. Comprehensive evaluation of a model alcohol education curriculum. J Stud Alcohol 1988; 49~38-50. 26. Williams AF, Dicicco LM, Unterberger H. Philosophy and evaluation of an alcohol education program. J Stud Alcohol 1968; 29685-702. 27. Wodarski JS. Teaching adolescents about alcohol and driving: A two year follow-up. J Drug Educ 1987; 11:327-344. 28. Botvin GJ, Baker E, Botvin EM, Filazzola AD, Millman RB. Prevention of alcohol misuse

29.

30.

31. 32.

33.

34.

35. 36. 37. 38. 39.

40.

41.

through the development of personal and social competence: A pilot study. J Stud Alcohol 1984; 45:550-552. Perry CL, Grant M, Emberg G, Florenzano RU, Langdon MC, Myeni AD, Waahlberg R, Berg S, Andersson K, Fisher J, Blaze-Temple D, Cross D, Saunders B, Jacobs DR, Schmid T. W.H.O. collaborative study on alcohol education and young people: Outcomes of a four-country pilot study. Int J Addict 1989; 24:1145-1171. Mauss AL, Hopkins RH, Weisheit RA, and Keamey KA. The problematic prospects for prevention in the classroom: Should alcohol education programs be expected to reduce drinking by youth? J Stud Alcohol 1988; 495 l-61. Moskowitz JM. The primary prevention of alcohol problems: A critical review of the research literature. J Stud Alcohol 1989; 50~54-88. Tobler N. Me&analysis of 143 adolescent drug prevention programs: Quantitative outcome results of program participants compared to a control or comparison group. J Drug Issues 1986; 16537-567. MacKinnon DP, Johnson CA, Penta MA, Dwyer JH, Hansen WB, Flay BR, Wang EYI. How school-based drug prevention works: One year effects of the Midwestern Prevention Project. Health Psychol, in press. Hansen WB. Theory and implementation of the social inthtence model of primary prevention. Prevention research findings: 1988. OSAP Prevention Monograph-3. USDHHS, PHS, ADAMHA, 1989:93-107. Biglan A, Severson HH, Bavry J, McConnell S. Social influence and adolescent smoking: A first look behind the barn. Health Educ 1983; 14-18. Chassin L, Presson CC, Sherman SJ, Corty E, Olshavsky R. Predicting the onset of cigarette smoking in adolescents: A longitudinal study. J Appl Sot Psycho1 1984; 14:224-243. Graham JW, Marks G, Hansen WB. Social intluence processes affecting adolescent substance use. J Appl Psychol, in press. Grube JW, Morgan M, McGree ST. Attitudes and normative beliefs as predictors of smoking intentions and behaviours: A test of three models. Br J Sot Psycho1 1986; X%81-93. Hansen WB, Graham JW, Wolkenstein BH, Lundy BZ, Pearson JL, Flay BR, Johnson CA. Differential impact of three alcohol prevention curricula on hypothesized mediating variables. J Drug Educ 1988; 18143-153. Sussman S, Dent CW, Rauch JM, Johnson CA, Hansen WB, Flay BR. Adolescent nonsmokers, triers, and regular smokers’ estimates of cigarette smoking prevalence: When do overestimations occur and by whom? J Appl Sot Psycho1 1988; l&537-555. Graham JW, Flay BR, Johnson CA, Hansen WB, Collins LM. Group comparability: A multiat-

430

42. 43. 44. 45. 46.

47. 48. 49. 50. 51. 52. 53.

HANSEN

AND

GRAHAM

tribute utility measurement approach to the use of random assignment with small numbers of aggregated units. Eva/ Rev 1984; 8:247-260. Hansen WB, Graham JW, Wolkenstein BH, Rohrbach LA. Program integrity as a moderator of prevention program effectiveness: Results for fifth grade students in the Adolescent Alcohol Prevention Trial. .I Stud Alcohol, in press. Murray DM, Perry CL. The measurement of substance use among adolescents: When is the ‘bogus pipeline’ method needed? Addict Behnv 1987; 12~225-233. Hansen WB, Malotte CK, Fielding JE. The Bogus Pipeline revisited: The use of the threat of detection as a means of increasing self-reports of tobacco use. J Appl Psychol1985; 703789-791. Hansen WB, Collins LM, Malotte CK, Johnson CA, Fielding JE: Attrition in prevention research. J Behav Med 1985; 8:261-275. Graham JW, Rohrbach LA, Hansen WB, Flay BR, Johnson CA. Convergent and discriminant validity for assessment of skill in resisting a role play alcohol offer. Behav Assess 1989; 11:353379. Rohrbach LA, Graham JW, Hansen WB, Flay BR, Johnson CA. Evaluation of resistance skills training using multitrai-multimethod role play skill assessments. Health Educ Res 1987;2:401407. Evans RI, Rozelle RM, Mittelmark MB, Hansen WB, Bane AL, Havis J. Deterring the onset of smoking in children: Knowledge of immediate physiological effects and coping with peer pressure, media pressure, and parent modeling. J Appl Sot Psychol 1978; 8~126-135. Johnson CA, Graham JW, Hansen WB, Flay BR, Rohrbach LA, McGuigan K, Gee M. Project SMART after three years: An assessment of sixth-grade and multiple-grade implementations. USC-IPR Technical Report, September 1987. Duryea EJ. Utilizing tenets of inoculation theory to develop and evaluate a preventive alcohol education intervention. J Sch Health 1983; 53:250-257. Duryea E, Mohr P, Newman IM, Martin GL, Egwaoje E. Six-month follow-up results of a preventive alcohol education intervention. J Drug Educ 1984; 14~97-104. Murray DM, Johnson CA, Luepker RV, Mittelmark MB. The prevention of cigarette smoking in children: A comparison of four strategies. J Appl Sot Psycho1 1984; 14~274-288. Pechacek TF, Murray DM, Luepker RV, Mittelmark MB, Johnson CA. Measurement of adolescent smokoing behavior: Rationale and methods. J Behav Med 1984; 7:123-140.

Received May 22, 1990 Revised December 12, 1990 Accepted December 20, 1990