Preventing Binge Drinking During Early Adolescence: One-and Two ...

Copyright 2001 b the Educadonal Publishing Foundation 10.1037Im93-162X.15.4.360

Psychology of Addictive Behaviors 2001. 15. 4. 360-365

Vol.

0893-164~101155.h DOI:

No.

Preventing Binge Drinking During Early Adolescence: One- and Two-Year Follow-up of a School-Based Preventive Intervention Gilbert J. Botvin, Kenneth W. Griffin, Tracy Diaz, and Michelle Ifill-Williams Weill Medical College of Cornell University The authors examined the effectiveness of a school-based prevention program on reducing binge drinking in a sample of minority, inner-city, middle-school students. Rates of binge drinking were compared who received the pro& beginning in the 7th grade (n = 1.7i3) and a c&rol gou;(n = among 1,328) that did not. The preventionprogram had protective effects in terms of binge drinking at the l-year (8th grade) and 2-year (9th grade) follow-up assessments.The proportion of binge drinkers wa5 over 50% lower in the interventiongroup relative to the control group at the follow-up assessments.There were also several significant program effects on proximal drinking variables, including drinking knowledge. pro-drinking attitudes, and peer drinking norms. These findings indicate that a school-based drug abuse prevention approach previously found to be effective among White youth significantly reduced binge drinking among urban minority youth.

Notwithstanding the fact that it is illegal for adolescents to purchase alcohol$ beverages, the vast majority of youth have some experience with alcohol. Over 50% of 8th graders, 70% of 10th graders, and 80% of 12th graders report having tried alcoholic beverages at least once (Johnston, O'Malley, & Bachman, 2000). More alarming is the fact that 9% of 8th graders, 23% of 10th graders, and 33% of 12th graders reported being drunk in the past month along with the fact that 15% of 8th graders, 26% of 10th graders, and 31% of 12th graders reported binge drinking (i.e., drinking five or more drinks per drinking occasion). As is the case with most forms of substance abuse, alcohol use proceeds along a developmental continuum that begins with initial use and progresses to more serious levels of involvement, including increasing frequency of use, amount consumed per drinking occasion, and episodes of drunkenness (Millman & Botvin, 1992). As the frequency and intensity of alcohol use increase, there is increased risk for alcohol-related health and social problems, such as accidents, assaults, and suicide (Baker, O'Neill, Ginsburg, & Li, 1992; Roizen, 1982). Efforts to reduce alcohol abuse and alcohol-related problems have taken various forms, including interventions targeting the family, schools, workplace, and community. Over the last 20 years considerable attention has been focused on the development of prevention approaches that target adolescents in school settings (G. J. Botvin, 2000). Recognizing the relationship among different forms of substance use, prevention researchers have frequently tested approaches designed to influence the use of alcohol, to-

Gilbert J. Botvin, Kenneth W. Griffin, Tracy Diaz, and Michelle IfillWilliams, Institute for Prevention Research, Weill Medical College of Comell University. This research was supported by funds from the National Institute for Drug Abuse (Grant P50DA7656). Compondence concerning this article should be addressed to Gilbert J. Botvin, Institute for Prevention Research, Weill Medical College of Cornell University, 411 East 69th S m t , New York, New York 10021. Electronic mail may be sent to [email protected].

bacco, and other dmgs (ATOD). Reviews of the extant prevention literature (e.g., G. J. Botvin, 2000; Flay, 1985; Hansen, 1992) and meta-analytic studies (Bangert-Drowns, 1988; Bruvold & Rundall, 1988; Tobler & Stratton, 1997) indicate that psychosocial ATOD prevention approaches that target social influences either alone or in combination with the teaching of general personal and social skills are more effective than other available prevention approaches. Effective psychosocial prevention approaches are based on social learning theory (Bandura, 1977) and problem behavior theory (Jessor & Jessor, 1977) as well as empirical evidence concerning the etiology of adolescent ATOD use (Hawkins, Catalano, & Miller, 1992; Pandina, 1997). The initiation of ATOD use is conceptualized as resulting from the interplay of inter- and intrapersonal etiologic factors (G. J. Botvin, 2000). The use of ATOD is learned through a process of modeling and reinforcement from various social influences, including peers, family members, and the media. vulnerability to these influences is determined by domain-specific cognitions, attitudes, and expectations; degree of bonding to social institutions; and the availability of skills for coping with peer pressure to drink and other life situations confronting adolescents. Preventive interventions based on this formulation emphasize increasing students' awareness of the social influences promoting ATOD use, modifying normative expectations concerning ATOD use, and teaching skills for resisting ATOD pressures (e.g., Pentz et al., 1989). Competence enhancement prevention approaches also emphasize the importance of teaching self-managementskills and social skills (e.g., G. J. Botvin, Baker, Dusenbury, Tortu, & Botvin, 1990) in an effort to decrease motivations to use ATOD and decrease vulnerability to ATOD-use social influences. A growing empirical literature shows that these approaches produce short-term reductions in the frequency and amount of ATOD use relative to untreated controls. However, although there is some evidence that these effects may decay over time (e.g., Ellickson & Bell, 1990). other studies have shown that more comprehensive and intensive prevention approaches, including booster sessions, can produce more durable prevention effects that last until the end of high school

SPECIAL ISSUE: PREVENTING BINGE DRINKING (G. J. Botvin, Baker, Dusenbury, Botvin, & Diaz, 1995; Botvin et al., 2000, Pentz et al., 1989). Despite significant advances over the last 20 years, there are a number of manifest gaps in the prevention literature. One such gap concerns the fact that, because most prevention research has been conducted with predominantly White, middle-class, adolescent populations, there is a dearth of research with inner-city, minority youth. Consequently, little is known about the type of prevention approaches that may work with minority youth or the generalizability to minority populations of approaches shown to be effective with White youth. The few studies that have included minority youth have several notable limitations, such as the inclusion of only a small number of minority youth (Ellickson & Bell, 1990). data derived from small-scale studies (G. J. Botvin et al., 1997; G. J. Botvin, Schinke, Epstein, & Diaz, 1994; G. J. Botvin, Schinke, Epstein, Diaz, & Botvin, 1995), or studies focused only on cigarette smoking (G. J. Botvin, Batson et al., 1989; G. J. Botvin, Dusenbury, Baker, James-Ortiz, & Kerner, 1989; G. J. Botvin et al., 1992). Another important gap is that little is known about the impact of current prevention approaches with respect to more serious levels of ATOD involvement in general and problematic alcohol use in particular. Although the ultimate goal of prevention approaches targeting alcohol is to reduce alcohol abuse, alcohol-related problems, or both, preventive interventions designed for youth typically focus on early-stage use (G. J. Botvin, 2000). Research is also needed to assess the impact of promising prevention approaches on more problematic levels of alcohol use, including binge drinking. Past research has shown that binge drinkers are at increased risk of alcohol-related problems (Wechsler, Dowdall, Maenner, Gledhill-Hoyt, & Lee, 1998). Furthermore, most research on binge drinking has been conducted with older youth, particularly college undergraduates (e.g., Norman, Bennett, & Lewis, 1998; Weingardt et al., 1998). Research is needed to increase our unders&ding of how to effectively prevent binge drinking during early adolescence, before it increases in frequency and becomes associated with various alcohol-related problems. This study was designed to extend the findings o f previous research assessing the impact of a cognitive-behavioral approach called Life Skills Training (LST) on ATOD use among minority youth (G. J. Botvin, Griffin, Diaz, & Ifill-Williams, 2001). The purpose of this study, in which we used data from G. J. Botvin et al.'s (2001) large-scale, randomized trial, was to determine the effectiveness of the LST approach for reducing binge drinking with inner-city minority youth and to determine its impact on several proximal drinking-related variables.

Method Sample Participants were students from 29 schools in New York City. Students

(N = 3,041) who completed surveys in the seventh, eighth, and ninth grades were included in the analyses. The sample was 44% male and 56% female. The racial-ethnic composition of the sample was 57% African American, 24% Hispanic, 8% Asian, 3% White, 2% Native American, and 6% of mixed or other backgrounds. Most students (55%) lived in twoparent households (including households with a stepparent), and more than one third (35%) lived in mother-only households. The sample included a large proportion of economically disadvantaged youth, as indicated by the

361

fact that 62% of students received free lunches at school. Approximately 90% of eligible students participated in this study. The vast majority of

eligible students who did not complete the survey were absent from school, and relatively few students who were present in the classroom refused to participate.

Research Design Before we randomid participating schools to experimental condition, we surveyed them and divided them into high, medium, or low smoking prevalence. Using a block randomized design, we randomly assigned schools from within these groups to either receive the intervention (16 schools) or be in the control group (13 schools). Students in the intemention group (n = 1,713) received 15 sessions in the seventh grade and 10 booster sessions in the eighth grade. Regular classroom teachers administered all intervention ses..ions. Students in the control group (n = 1,328) received the substance use cumculum normally in place in New York City schools.

Procedure Data were collected following a detailed protocol approved by Comell Medical College's Institutional Review Board. Participants completed a pretest survey in the seventh grade prior to the prevention program and at annual follow-up assessments in the eighth and ninth grades. To link pretest and follow-up surveys while maintaining student confidentiality, we used unique identification codes rather than the students' names. Questionnaires were administered during a regular classroom period by a team of several data collectors who were of similar ethnic-mial backgrounds as that of participants. Carbon monoxide breath samples were also collected at all assessments to enhance the validity of self-report data, using a variant of the procedure developed by Evans. Hansen, and Mittlernark (1977). Prior research has shown that this procedure enhances the validity of adolescents' self-reports of alcohol and marijuana use in addition to smoking (E.M. Botvin, Botvin, Renick, f i l m l a , & AUegrante, 1984).

Prevention Program The preventive intervention teaches alcohol and drug resistance skills, norms against drinking and drug use, and material designed to facilitate the development of important personal and social skills. The goal of the prevention program is to provide adolescentswith the knowledge and skills needed to effectively mist social influences to use ATOD. The program also aims to reduce potential motivations to use substances by increasing general personal and social competence. The program teaches students a variety of cognitivebehavioral skills for building self-esteem, resisting advertising pressure, managing anxiety, communicating effectively, developing personal relationships, and asserting one's rights. These are taught with proven skills-training techniques such as group discussion, demonstration, modeling, behavioral reheanal, feedback and reinforcement, and behavioral "homework" assignments for out-of-class practice. The program also teaches problem-specific skills related to alcohol and drug use. Students are taught, for example, ways to use general assertiveness skills in situations wherethey experiencedirect inGkonal pressure from peers to drink alcohol. The program emphasizes information regarding the immediate negative consequences of alcohol and drug use, the decreasing social acceptability of use, and the actual prevalence rates among adults and adolescents.Only minimal information is provided regarding the long-term health consequences of alcohol and drug use, unlike many traditional prevention programs. Material is also provided to reinforce norms against alcohol and drug use. Intervention materials included a teacher's manual with detailed lesson plans, student handouts, and video material demonstrating the skills taught in the prevention program by same-age minority adolescents. Regular

362

BOTVIN, GRIFFIN, DIM, AND IFILL-WILLIAMS

classroom teachers who had attended a 1-day teacher training workshop implemented the prevention program. The purpose of the training workshop was to familiarize the teachers with the content of the prevention program, the rationale for the cognitive-behavioral skills training prevention strategy, and the results of prior studies and to provide an opportunity for teachers to learn and practice the skills needed to successfully implement the prevention prog& (Tortu & Botvin, 1989). This prevention approach has been shown to be effective among suburban White youth (e.g., G. J. Botvin et a]., 1990) with prevention effects lasting until the end of high school (G. J. Botvin, Baker, et a]., 1995; G. J. Botvin et al., 2000). Over the past several years, this approach has been revised for use with minority youth (G. J. Botvin et al., 1992, G. J. Botvin et al., 1997; G. J. Botvin et a]., 1989; G. J. Botvin et al., 1994; G. J. Botvin, Schinke, et al., 1995). and the program has recently been shown to be effective in preventing ATOD use among innercity minority youth in a large randomized trial (G. J. Botvin et a]., 2001). This study is the first, however, to examine prevention effects on binge drinking.

Measures Data were collected from students by means of a questionnaire that included items assessing current alcohol and drug use and a series of scales measuring cognitive, attitudinal, and skills variables believed to be associated with the initiation of alcohol and drug use. The questionnaire also included standard survey items concerning gender, age, family structure, race, and ethnicity. Although all measures were self-reported, data collectors emphasized the confidential nature of the data being collected in order to enhance the truthfulness of student responses. Only those scales retained after preliminary analyses and relevant to the evaluation of the intervention are included. Alcohol use. Quantity of alcohol use per drinking occasion was assessed by asking students "If you drink alcohol, how much do you usually drink each time you drink?" with response options of I ("I don't drink), 2 ("One drink"), 3 ('Two drinks"), 4 ("Three or four drinks"), 5 ("Five or six drinks") and 6 ("More than six drinks"). We created a dichotomous variable to identify students who reported that they typically drink 5 or more drinks per drinking occasion as binge drinkem. Drinking-related variables. Drinking knowledge was assessed with four truefalse items regarding the physiologicaland psychological effects, prevalence, and social acceptability of drinking. Higher scores represented greater drinking knowledge. Five items were used to assess favorable attitudes toward alcohol (a = 36). such as students' perceptions of the perceived social benefits of drinking, with responses made on a 5-point scale anchored by I (srrongly disagree) and 5 (strongly agree). ~ o r m & v e expectations regarding alcohol use were assessed in terms of the perceived prevalence of alcohol use among peers rated on a 5-point scale that ranged from 1 (none) and 5 (all or a l m s r all).

Implementation Fidelity Project staff in randomly selected classrooms monitored program implementation. Using observation forms developed for each prevention session, trained staff members observed teachers and recorded how much of the material allocated for each session was actually covered in the classroom. During the first year of the study, all LST teachers were observed at least once, and more than 95% of teachers were observed in the second year. Most teachers were observed for two or three class sessions. We quantified the completeness of the implementation of the program by calculating the proportion of prevention objectives covered during each session observed (number of objectives covered divided by the total number of actual cuniculum objectives for the particular session observed). The mean number of program points covered was 48% (SD = 20%).

Data Analysis We analyzed the data using chi-square tests, generalized linear models analysis of variance, and logistic regression using the generalized estimating equations (GEE) option in SAS 8.0 (SAS Institute, 2000). Listwise deletion was used for all analyses. First we conducted a series of chi-square tests to determine pretest comparability of the intervention and control conditions. Second, we compared the longitudinal sample used in this study with the full pretest sample to determine the impact of attrition using a series of generalized linear models analysis of variance. Thud, we examined the effectiveness of the pevention program on binge drinking and several drinking-related variables, after controlling for pretest scores and several additional covariates. Because the intervention was randomized and administered at the school level, we conducted analyses of program effectiveness controlling for intracluster correlations (ICCs) among students within schools. The ICC quantifies the degree of similarity of questionnaire responses within schools and in the present study indicates how binge drinking and related variables varied at the school level. Previous school-based drug abuse prevention programs have shown that even small degrees of within-cluster correlation (ICCs in the range of r = .02-.05) can have large effects on variance estimation and hypothesis testing (Murray & Hannan, 1990) and can lead to underestimated standard errors and increased Type I errors. Furthermore, we have found in our previous work that prevalence rates of drug and alcohol use are often lower among African American youth compared to other racial-ethnic groups, and therefore these behaviors may cluster among small intact groups of high-risk youth within some schools, underscoring the need to control for the ICCs (Scheier, Griffin, Doyle, & Botvin, in press). Thus, we conducted analyses using the GEE-independent option (in SAS's PROC GENMOD) to evaluate intervention effects while controlling for the ICCs, as recommended by Norton, Bieler, Ennett, and Zarkin (19%). The GEE method adjusts the estimated standard e m r to account for the within-cluster correlation and generally provides for a more conservative test of the hypothesis when a positive ICC is present.

Results

Pretest Equivalence We used a series of chi-square analyses to examine whether the intervention and control groups differed on any of the alcohol use outcome variables at the pretest assessment. There were no significant pretest differences in rates of binge drinking or in lifetime, monthly, and weekly alcohol use or drunkenness. For example, 1.5% of the intervention students reported binge drinking at baseline compared to 1.8% of the control group students, $(I, N = 3,041) = 0.4, p < S34. We also compared the intervention and control groups in terms of several demographic variables at the pretest. Although there were no differences between conditions in gender composition, there were significant differences in terms of ethnicity. The intervention group had a higher proportion of African American youth (65%) relative to controls (46%), $(l, N = 3,041) = 102.2, p < .0001, and a lower proportion of Hispanic youth (18%)relative to controls (32%), ,$(I, N = 3,041) = 87.3, p < .0001. Furthermore, the intervention group had a lower proportion of students who received free lunch from school (58%) relative to controls (67%),$(I, N = 3,021) = 2 6 . 4 , ~< .0001. To account for these differences, we included race, gender, and free lunch status as control variables in all analyses of program effects.

Attrition Analysis Of the 5,233 seventh-grade students who completed the pretest survey, 4,190 (80%) completed an identical survey at the 1-year

363

SPECIAL ISSUE: PREVENTING BINGE DRINKING

follow-up, and 3,041 (58%) completed the survey at both the 1and 2-year follow-up assessments. Attrition rates were similar across both experimental and control conditions. We conducted analyses to determine whether youth who used alcohol at the pretest had higher attrition rates than those who did not. Findings indicated that those who reported drinking monthly at baseline (5% of the sample) were more likely to drop out (57%) relative to those who did not report monthly drinking (41%). ,&I, N = 5,233) = 25.4, p < .0001. However, the rate of attrition of drinkers did not differ across experimental conditions. Furthermore, males dropped out of the study at a higher rate (47%) than females (37%), g ( 1 , N = 5,221) = 62.3, p < .0001, but this did not differ across experimental condition. Because attrition was higher among participants who may have been initially at higher risk for binge drinking, this may have resulted in a restricted range in alcohol scores in the matched longitudinal sample used to evaluate intervention effects. Thus, this study may be a relatively conservative test of the prevention program.

Intervention Eflects We conducted several analyses to examine the effects of the intervention on binge drinking at the 1- and 2-year follow-up assessments. In evaluating program effects, a variable indicating experimental condition was coded such that participants in the intervention received a 1 and controls received a 0; control variables included race, gender, free lunch, percentage of the prevention program completed, and the matching pretest variable of interest. One-yearfollow-up. As shown in Table 1, the intervention had a significant effect on binge drinking at the 1-year follow-up (eighth grade), with an odds ratio (OR) of 0.41 and 95% confidence interval (CI) of 0.18-0.93. The covariate-adjusted proportion of binge drinkers at the 1-year follow-up was 4.3% of the control group and 1.8% of the intervention group; thus, the prevention program produced a 57% reduction in binge drinking at the 1-year follow-up. In terms of the proximal drinking-related variables (drinking knowledge, pro-drinking attitudes, and peer drinking normative expectations), we conducted separate analyses for each outcome that included experimental condition as the independent variable, along with the same set of covariates used in previous analyses. As shown in Table 2, the preventive intervention had an impact on drinking knowledge at the 1-year follow-up, with intervention students scoring higher than controls. There were also program effects on pro-drinking attitudes and peer drinking normkive expectations, with intervention students scoring lower than controls on these variables. Table 1 Logistic Regression Analyses of Experimental Condition Predicting Binge Drinking at the Follow-Up Assessments Binge drinking

b

SE

OR

One-year follow-up (8th grade) Two-year follow-up (9thgrade)

-0.90* -OM**

0.42 0.31

0.41 0.18-0.93 0.40 0.22-0.74

95% CI

Note. N = 2,982.Covariates for all analyses were gender, race, free

lunch, percentage of program completed, and baseline level of binge drinking. Experimental condition was coded 0 for control group and 1 for experimental group. OR = odds ratio; CI = confidence interval. *p < .05. **p < .01.

Two-year follow-up. As shown in Table 1, the intervention also had a significant protective effect on binge drinking at the 2-year follow-up (ninth grade), with an OR of 0.40 and a 95% CI of 0.22-0.74. The covariate-adjusted proportion of binge drinkers at the 2-year follow-up was 5.2% of the control group and 2.2% of the intervention group; thus. the prevention program maintained a 57% reduction in binge drinking at the 2-year follow-up. As shown in Table 2, the preventive intervention had an impact on normative expectations regarding peer drinking at the 2-year follow-up, with intervention students scoring lower than controls on this variable, although the effects on drinking knowledge and pro-drinking attitudes were no longer significant.

Discussion Advances in preventing adolescent ATOD use indicate that school-based approaches that target psychosocial factors associated with ATOD use can deter the early-stage use of one or more substances. The most effective approaches focus on pro-ATOD use social influences either alone or in combination with methods intended to enhance generic personal and social competence (G. J. Botvin, 2000). However, most prevention research has been conducted with predominantly White suburban youth and has focused on early-stage use. Relatively little is known about the impact of effective school-based prevention approaches on inner-city minority youth and the durability of any observed effects. Even less is known about the effectiveness of these prevention approaches in terms of their potential for ameliorating more serious levels of ATOD use, particularly episodes of binge drinking. To begin addressing these gaps in the prevention literature, we designed this study to provide important new information about the impact of a cognitivebehavioral, school-based prevention approach called Life Skills Training on inner-city, minority youth with respect to binge drinking, defined as typically consuming 5 or more drinks per drinking episode. We examined as well the impact of the prevention program on several drinking-related proximal measures, including drinking knowledge, attitudes, and peer normative expectations. The findings of this study indicate that the prevention program produced measurable changes in binge drinking, as well as several proximal drinking-related variables, at the 1- and Zyear follow-up assessments. The proportion of binge drinkers was more than 50% lower in the intervention group relative to the control group at the two follow-up assessments. One-year follow-up results also indicated that the preventive intervention increased drinking knowledge and decreased pro-drinking attitudes and normative expectations regarding peer drinking. Two-year follow-up results also revealed that the prevention program produced long-term prevention effects with respect to binge drinking and peer drinking norms. The results of this study are important because they show that this prevention approach produces prevention effects on problematic levels of alcohol use with inner-city, minority youth that last for 2 years after the initial year of the prevention program. Prior research with this approach has provided evidence that it is effective with White, middle-class youth and produces long-term effects (e.g., G. J. Botvin, Baker, et al., 1995; G. J. Botvin et al., 2000). More recent data also indicate that this approach can produce short-term effects on alcohol use among minority youth (G. J. Botvin et al., 2001). However, this is the first study to

BOTVIN, GRIFFIN, DIAZ, AND IFILL-WILLIAMS

Table 2 Adjusted Means for Proximal Drinking-Related Variables at the Follow-Up Assessments, by Experimental Condition

Drinking-related variable One-year follow-up (8th gmde) Drinking knowledge Ro-drinking attitudes Peer drinking norms Two-year follow-up (9th grade) Drinking knowledge Pro-drinking attitudes Peer drinking norms

Intervention PUP

Control group

N

M

SE

M

SE

Z

P

2,631 2,718 2.869

1.48 1.52 3.00

0.01 0.02 0.02

1.45 1.59 3.17

0.01 0.02 0.04

1.98 2.39 3.42

.0475 .0170 .OOo6

2,511 2,576 2,873

1.48 1.47 3.30

0.01 0.02 0.03

1.48 1.52 3.41

0.01 0.02 0.03

0.51 1.45 2.43

6091 .I462 .0151

Note. Covariates for all analyses were -gender, race, free lunch, percentage of program completed, and baseline level of outcome variable.

demonstrate that this type of prevention approach can produce significant and durable reductions in binge drinking among minority youth using a school-based intervention. These effects not only were statistically significant, but they also are particularly meaningful, because they indicate that the prevention approach tested reduced the odds of binge drinking by more than 50% for the students who received the LST prevention program. In addition to demonstrating the durability of prevention effects on alcohol use, we found that the prevention program produced significant prevention effects on several proximal drinking-related variables. These included increases among students who received the LST program in drinking knowledge and decreased prodrinking attitudes and normative expectations regarding peer drinking. Similar findings have been observed in other studies with respect to the impact of LST on proximal variables (e.g., G. J. Botvin et al., 1990; G. J. Botvin et al., 1992; G. J. Botvin et al., 1994; G. J. Botvin, Griffin, Diaz, Miller, & Ifill-Williams, 1999). This study was designed to overcome weaknesses of earlier prevention studies. For example, prior prevention research with school-based approaches has been criticized for failing to adequately control for the observed correlations of individuals within assignment clusters (e.g., classrooms or schools), which may inflate the Type I error rate (Murray & Hannan, 1990). A methodological strength of this study, therefore, is that data analysis methods were used that adjusted for school-level ICCs. Other manifest strengths of this study include the use of random assignment of schools to conditions, a predominantly minority sample, long-term follow-up, the use of minority data collectors, previously tested assessment instruments, unique ID codes to ensure confidentiality, and data collection procedures designed to enhance the validity of students' self-reports. This study also has several limitations that should be noted. First, the follow-up was limited to 2 years after the initial posttest. Additional follow-up data would be useful to determine the longer term durability of these prevention effects with this population. The retention rate in the ninth grade was less than ideal and may have been due to the fact that students transferred to senior high between the eighth and ninth grades. Second, because this was a school-based study that relied on students' self-reports, the significant relationships among variables may partly reflect shared method variance, because all data were obtained by self-report

questionnaire. Third, the generalizability of these findings to other minority populations is unclear. It is important to note that this study focused on African American and Hispanic youth, with only a small number of Asians. Future research is needed to determine the extent to which this prevention approach is effective with other minority populations as well as specific Hispanic (Puerto Rican, Dominican, Cuban, etc.) and African American ethnic subgroups. However, because the number of participants from certain ethnic groups was relatively small, we did not have adequate statistical power to test for program effects by subgroup. Fourth, because the prevention approach used in this study is a universal approach intended for all students regardless of risk status, further research is needed to determine if targeted or indicated prevention programs are better suited for students with special needs (e.g., newly immigrated youth or those at highest risk) rather than universal prevention approaches. Finally, the low implementation fidelity we observed (mean number of program points covered was 48%) suggests that there are considerable difficulties with implementing this type of program in inner-city schools and that future research should focus on the barriers to fidelity and how to overcome them. One approach might be to increase the time and effort spent on teacher training. The results of this study provide additional evidence of the effectiveness of a school-based ATOD prevention approach focusing on social influences and generic competence skills. These results extend prior research with this approach by demonstrating its effectiveness in reducing binge drinking among minority youth over time. This study is also important because it used rigorous research methods, including a randomized control group design and data analysis methods that adjusted for school-level ICCs. Future research is warranted to determine the durability of these prevention effects with longer term follow-up assessments, the impact of this prevention approach on specific subgroups of adolescents, and the extent to which these findings generalize to other populations.

References Baker, S. P., O'Neill, B., Ginsburg, M. J., & Li, G . (1992). The injuryfact book (2nd 4 . ) . New Yo& Oxford University Press. Bandura, A. (1 977). Social learning theory. EnglewoodCliffs, NJ: Prentice Hall.

SPECIAL ISSUE: PREVENTING BINGE DRINKING Bangert-Drowns, R. L. (1988). The effects of school-based substance abuse education: A meta-analysis. Journal of Drug Education, 18, 243-264. Botvin. E. M., Botvin, G. J., Renick, N. L., Filazzola, A. D., & Allegrante, J. P. (1984). Adolescents' self-reports of tobacco, alcohol, and marijuana use: h U n i ~ n gthe comparability of videotape, cartoon, and verbal bogus pipeline procedures. Psychological Reports. 55. 379-386. Botvin, G. J. (2000). Preventing drug abuse in schools: Social and competence enhancement approaches targeting individual-level etiological factors. Addictive Behaviors, 25, 887-897. Botvin, G. J., Baker, E., Dusenbury, L. D., Botvin, E. M., & Diaz, T. (1995). Long-term follow-up results of a randomized drug abuse prevention trial in a White middle-class population. Journal of the American Medical Association, 273, 1106-1 112. Botvin, G. J., Baker, E., Dusenbury, L., Tortu, S., & Botvin, E. M. (1990). Preventing adolescent drug abuse through a multimodal cognitivebehavioral approach: Results of a threeyear study. Journal of Consulting and Clinical Psychology, 58, 437-446. Botvin, G. J., Batson, H., Witts-Vitale, S., Bess, V., Baker. E., & Dusenbury, L. (1989). A psychosocial approach to smoking prevention for urban Black youth. Public Health Repons. 104,573-582. Botvin, G. J., Dusenbury, L., Baker, E., James-Odz, S., Botvin, E. M., & Kemer, J. (1992). Smoking prevention among urban minority youth: Assessing effects on outcome and mediating variables. Health Psychology, 11, 290-299. Botvin, G. J., Dusenbury, L., Baker, E., James-Ortiz, S., & Kerner, J. (1989). A skills training approach to smoking prevention among Hispanic youth. Journal of Behavioral Medicine, 12, 279-296. Botvin, G. J., Epstein, J. A., Baker, E., Diaz, T., Ifill-Williams, M., Miller, N., & Cardwell, J. (1997). School-based drug abuse prevention with inner-city minority youth. Journal of Child and Adolescent Substance Abuse, 6, 5-19. Botvin, G. J., Griffin, K. W., Diaz. T., & Ifill-Williams, M. (2001). Drug abuse prevention among minority adolescents: Oneyear follow-up of a school-based preventive intervention. Prevention Science, 2, 1-13. Botvin, G. J., Griffin, K. W., Diaz, T., Miller, N., & Lfill-Williams, M. (1999). Smoking initiation and escalation in early adolescent girls: One-year follow-up of a school-based prevention intervention for minority youth. Journal of the American Medical Women's Association, 54, 139-143. Botvin, G. J., Griffin, K. W., Diaz, T., Scheier, L. M., Williams, C., & Epstein, J. A. (2000). Preventing illicit drug use in adolescents: Longterm follow-up data from a randomized control trial of a school population. Addictive Behaviors, 5, 769-774. Botvin, G. J., Schinke, S. P., Epstein, J. A.. & Dim, T. (1994). Effectiveness of culturally-focused and generic skills training approaches to alcohol and drug abuse prevention among minority youths. Psychology of Addictive Behaviors. 8, 116-127. Botvin, G . J., Schinke, S. P., Epstein, J. A., Diaz, T., & Botvin, E. M. (1995). Effectiveness of culturally-focused and generic skills training approaches to alcohol and drug abuse prevention among minority adolescents: Two-year follow-up results. Psychology of Addictive Behaviors, 9, 183-194. Bruvold, W. H., & Rundall, T. G. (1988). A meta-analysis and theoretical review of school based tobacco and alcohol intervention programs. Psychology and Health, 2, 53-78. Ellickson, P., & Bell, R. M. (1990, March 16). Drug prevention in junior high: A multi-site longitudinal test. Science, 247, 1299-1305. Evans, R. I., Hansen, W. B., & Mittlemark, M. B. (1977). Increasing the validity of self-reports of smoking behavior in children. Journal of Applied Psychology, 62, 521-523. Flay, B. R. (1985). What we know about the social influences approach to smoking prevention: Review and recommen&tions. In C. S. Bell & R. Battjes (Eds.), Prevention research: Deterring drug abuse among chil-

365

dren (NIDA Research Monograph No. 63). Washington, DC Department of Health and Human Services. Hansen, W. B. (1992). School-based substance prevention: A review of the state-of-the-art in cumculum. Health Educafion Research, 7, 403-430. Hawkins. J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin. 112, 64-105. Jessor, R., & Jessor, S. L. (1977). Problem behavior and psychosocial development:A longitudinal study of youth. New York: Academic Press. Johnston, L. D., O'Malley. P. M.. & Bachman, J. G. (2000). Monitoring the Future national survey results on drug use, 1975-1999. Volume I: Secondary school students (NU4 Publication No. 00-4802). Rockville, MD: National Institute on Drug Abuse. Millman, R. B., & Botvin. G. J. (1992). Substance use, abuse, and dependence. In M. Levine, N. B. Carey, A. C. Cmker, & R. T. Gross (Eds.), Developmenml-behavioral pediatrics (2nd ed., pp. 451-467). New York: Saunders. Murray, D. M., & Hannan, P. J. (1990). Planning for the appropriate analysis in school-based drug-use prevention studies. Journal of Consulting and Clinical Psychology. 58. 458-468. Norman, P., Bennett, P., & Lewis, H. (1998). Understanding binge drinking among young people: An application of the theory of planned behaviour. Health Education Research, 13, 163-169. Norton, E. C., Bieler, G. S., Ennett, S. T., & Zarkin. G. A. (1996). Analysis of prevention program effectiveness with clustered data using generalized estimating equations. Journal of Consulting and Clinical Psychology, 64,919-926. Pandina, R. J. (1997). Risk and protective factor models in adolescent drug use: Putting them to work for prevention. In National Conference on Drug Abuse Prevention Research: Presentations, papers, and recommendations. Rockville, MD: U.S. Department of Health and Human Setvices, National Institutes of Health, National Institute on Drug Abuse. Pentz, M. A., Dwyer, J. H., MacKinnon, D. P., Flay, B. R., Hansen, W. B., Wang, E. Y., & Johnson, C. A. (1989). A multicornmunity trial for primary prevention of adolescent drug abuse: Effects on drug prevalence. Journal of the American Medical Association. 261, 3259-3266. Roizen, J. (1982). Estimating alcohol involvement in serious events. In National Institute on Alcohol Abuse and Alcoholism, Alcohol consumption and related problems (Alcohol and Health Monograph No. 1, DHHS Publication No. ADM 82-1 190, pp. 179-219). Washington, DC: U.S. Government Printing Office. SAS Institute. (2000). SAWSTAT user's guide, Version 8. Cary, NC: Author. Scheier, L. M., Griffin, K. W., Doyle, M. M., & Botvin, G. J. (in press). Estimates of intragroup dependence for drug abuse and skills measures encountered in school-based drug abuse prevention trials: An empirical study of three independent samples. Health Education & ~ e h a v b r . Tobler, N. S., & Stranon, H. H. (1997). Effectivenessof school-based drug prevention programs: A meta-analysis of the research. Journal of Primary Prevention. 18, 7 1-1 28. Tortu, S., & Botvin, G. J. (1989). School-based smoking prevention: The teacher training process. Preventive Medicine, 18, 280-289. Wechsler, H., Dowdall, G. W., Maenner, G.. Gledhill-Hoyt, J., & Lee, H. (1998). Changes in binge drinking and related problems among American college students between 1993 and 1997. Journal of American College Health, 47, 57-68. Weingardt, K. R., Baa, J. S., Kivlahan, D. R., Roberts, L. J., Miller. E. T., & Marlatt, G. A. (1998). Episodic heavy drinking among college students: Methodological issues and longitudinal perspectives. Psychology of Addictive Behaviors, 12, 155-167.

Received December 4, 2000 Revision received March 23, 2001 Accepted March 26, 2001