Preventing Tobacco and Alcohol Use Among Elementary ... - CiteSeerX

Preventing Tobacco and Alcohol Use Among Elementary School Students Through Life Skills Training Gilbert J. Botvin Kenneth W. Griffin Elizabeth Paul Araxi P. Macaulay

ABSTRACT. The present study examined the effectiveness of a substance abuse prevention program in preventing tobacco and alcohol use among elementary school students in grades 3 through 6. The prevention program teaches social resistance skills and general personal and social competence skills. Rates of substance use behavior, attitudes, knowledge, normative expectations, and related variables were examined among students (N = 1090) from 20 schools that were randomly assigned to either receive the prevention program (9 schools, n = 426) or serve as a control group (11 schools, n = 664). Data were analyzed at both the individual-level and school-level. Individual-level analyses controlling for gender, race, and family structure showed that intervention students reported less smoking in the past year, higher anti-drinking attitudes, increased substance use knowledge and skills-related knowledge, lower normative expectations for smoking and alcohol use, and higher self-esteem at the posttest assessment, relative to control students. School-level analyses showed that annual prevalence rate was 61% lower for smoking and 25% lower for alcohol use at the posttest assessment in schools that Gilbert J. Botvin and Kenneth W. Griffin are affiliated with the Institute for Prevention Research, Weill Medical College of Cornell University. Elizabeth Paul and Araxi P. Macaulay are affiliated with Teachers College, Columbia University. Journal of Child & Adolescent Substance Abuse, Vol. 12(4) 2003 http://www.haworthpress.com/store/product.asp?sku=J029 2003 by The Haworth Press, Inc. All rights reserved. 10.1300/J029v12n04_01

1

2

JOURNAL OF CHILD & ADOLESCENT SUBSTANCE ABUSE

received the prevention program when compared with control schools. In addition, mean self-esteem scores were higher in intervention schools at the posttest assessment relative to control schools. Findings indicate that a school-based substance abuse prevention approach previously found to be effective among middle school students is also effective for elementary school students. [Article copies available for a fee from The

Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: Website: © 2003 by The Haworth Press, Inc. All rights reserved.]

KEYWORDS. Smoking, alcohol, elementary school, prevention, children

Considerable effort and resources have been expended over the years in an effort to develop effective approaches for reducing the use of tobacco, alcohol, and illicit drugs. These efforts have taken the form of policy initiatives, media campaigns, law enforcement, treatment, and prevention approaches. An area of particular interest and focus has been prevention approaches designed for implementation in schools. These approaches typically target middle or junior high school students utilizing classroom-based interventions. School settings are especially well suited for both the implementation and testing of substance abuse prevention programs because they offer efficient access to large numbers of youth during the years that many begin to experiment with tobacco, alcohol, marijuana, and other drugs. Research over the past two decades has led to an increased understanding of the risk and protective factors associated with substance use initiation (Hawkins, Catalano, & Miller, 1992; Newcomb & Bentler, 1988) and to the development of more complete theoretical formulations of the etiology of substance use (Petraitis, Flay, & Miller, 1995). This research has produced new information regarding the factors promoting and sustaining substance use as well as information concerning its developmental course. The accumulated information from this research is important because it provides much needed guidance concerning the nature and timing of potentially effective preventive interventions. Research delineating the etiologic determinants of substance use has highlighted the importance of an array of interpersonal and intra-personal factors for promoting and sustaining tobacco, alcohol, and illicit drug use. As a consequence, the focus of prevention approaches has shifted from a primary emphasis on knowledge concerning the adverse

Botvin et al.

3

consequences of substance use toward an emphasis on social and psychological factors (Botvin & Griffin, 1999; Hansen, 1992). Similarly, research concerning the developmental progression of substance use indicates that it proceeds in a stage-like sequence from the use of legal and widely available substances to the use of illegal substances (Kandel, 1975). For this reason, preventive interventions have generally targeted the use of tobacco and alcohol, two substances identified as occurring at the beginning of this progression. An implicit assumption of contemporary prevention approaches is that, if they are successful, they will not only reduce the initiation of substance use, but will also disrupt its developmental progression. Until recently, this has been an untested assumption. However, there is growing evidence that prevention programs that effectively target early-stage substance use can disrupt the developmental progression to more serious levels of drug involvement (Botvin, Scheier, & Griffin, in press). These findings contain new support for implementing prevention approaches targeting tobacco and alcohol use with younger populations. Reviews of the extant literature on substance abuse prevention (Botvin & Griffin, 1999; Hansen, 1992) and meta-analytic studies (Bangert-Downs, 1988; Tobler & Stratton, 1997) indicate that considerable progress has been made toward the development of effective strategies for preventing the onset and developmental progression of tobacco, alcohol, and illicit drug use. The most effective prevention approaches target individuals during the beginning of adolescence, a time of increased risk for an array of problem behaviors. These approaches also emphasize the teaching of social resistance skills either alone or in combination with the teaching of personal self-management skills or general interpersonal skills (Botvin, 2000). Despite the success of these prevention approaches, intervening earlier might produce stronger and more durable prevention effects. Indeed, a study by Sarvela et al. (1999) suggests that a critical time period in which to implement tobacco and alcohol prevention programs is between third and fifth grades. Moreover, implementing prevention programs earlier is also warranted by evidence that a substantial number of elementary students experiment with tobacco or alcohol. For example, Elder et al. (1996) found that nearly 5% of fifth graders reported smoking cigarettes. Many of the same etiologic factors appear to play a role in promoting the substance use onset among younger populations (O’Loughlin, Paradis, Renaud, & Sanchez, 1998). Yet, little is known about the type of prevention approach that would be effective with a younger population. For

4


example, it is unclear whether prevention approaches found to be effective with adolescents would also succeed with elementary school students. A limited literature of substance abuse prevention studies with elementary school students shows that several approaches to smoking and/or alcohol prevention have been utilized including smoke-free school policies (Elder et al., 1996), broad-based health promotion efforts (O’Loughlin et al., 1998), and school-based approaches teaching resistance skills and knowledge of immediate negative effects (Shope, Dielman, Butchart, Campanelli, & Kloska, 1992). Few approaches have produced an impact on smoking or alcohol use behavior. However, evidence does exist showing that prevention approaches targeting elementary school students during grades 4, 5, and 6 can have positive effects on drug knowledge and intentions (Price et al., 1998). Moreover, a randomized trial with 5,000 5th and 6th graders testing a prevention program emphasizing resistance training for peer pressure, knowledge regarding immediate effects of alcohol, and risks of alcohol misuse did produce significant program effects on alcohol misuse (Shope et al., 1992). One of the most extensively and rigorously tested approaches to substance abuse prevention emphasizes the teaching of generic personal and social skills, norm setting, and drug resistance skills using a program called Life Skills Training (LST). The LST program was initially developed as a smoking prevention program for junior high school students. Through a series of research studies, the LST approach was demonstrated to be effective in preventing cigarette smoking (Botvin & Eng, 1982; Botvin, Renick, & Baker, 1983; Botvin, Eng, & Williams, 1980), alcohol use (Botvin, Baker, Dusenbury, Tortu, & Botvin, 1990; Botvin, Schinke, Epstein, Diaz, & Botvin, 1995), marijuana use (Botvin, Baker, Botvin, Filazzola, & Millman, 1984; Botvin, Baker, Dusenbury et al., 1990, 1995), and polydrug use (Botvin, Baker, Dusenbury et al., 1995; Botvin, Epstein, Baker et al., 1997). Furthermore, prevention effects have been shown to be both robust and durable. The results of a large-scale randomized trial found that prevention effects were evident more than six years later (Botvin et al., 1995; Botvin et al., 2000). Because of its emphasis on competence enhancement and the development of generic personal and social skills, it would appear to offer potential application to a younger population. The purpose of the present study was to determine the effectiveness of a competence enhancement approach to substance abuse prevention when conducted with elementary school students. More specifically, this study tested the effectiveness of a prevention program previously found effective with adolescents that was revised to make it develop-

Botvin et al.

5

mentally appropriate for elementary school students in grades 3 through 6. The impact of the prevention program was tested with respect to early-stage tobacco and alcohol use as well as several hypothesized mediating variables. METHOD Procedure Twenty suburban elementary schools were randomly assigned to experimental or control conditions. Participating students (N = 1,954) were administered the pretest questionnaire in the classroom. Students were told not to put their name on the questionnaire and were assured about the confidentiality of their responses. Following the pretest, students in the intervention schools participated in the psychosocial prevention program while those in control schools did not. Posttest data were collected three months later, and 1,868 students completed this survey. Unique student identification codes were not collected in order to maintain student anonymity. However, surveys for 1,090 students (58% of posttest respondents) were matched from pretest to posttest based on their school, teacher, gender, and birthday. Reasons for a lack of match include that students were absent from school at the time of data collection or transferred to another school, or that there was missing data for one of the matching variables. Sample The final sample of 1,090 students consisted of primarily of 4th (40%) and 5th graders (32%), with smaller numbers of 3rd (23%) and 6th graders (5%). The sample was 52% male and 48% female. The racial/ethnic backgrounds of students were white (48%), Hispanic (26%), African-American (13%), Asian (3%), and other or mixed ethnicity (10%). Most (70%) came from two-parent families, although several lived with mother only (18%), father only (3%) or with other relatives or guardians (9%). In terms of parental education, most students reported that their father had graduated high school (73%) or attended/graduated from college (19%), and their mother had graduated high school (65%) or attended/graduated from college (24%).

6


Prevention Program The goal of the prevention program was to reduce tobacco and alcohol use by providing youth with the necessary knowledge and skills for resisting social influences to use tobacco and alcohol, as well as to reduce motivation to use these substances. The prevention program consisted of 24 classes (30-45 minutes each) taught over three years with 8 classes per year for elementary school students in grades 3, 4, 5. The main focus of the intervention was on teaching of information for drug resistance skills, personal self-management and general social skills for increasing overall competence and promoting the development of characteristics associated with decreased risk of using drugs in the future. Methods utilized to teach the content of the program included facilitation and group discussion, as well as cognitive-behavioral skills, modeling and practice. To maximize the intervention’s effectiveness, a student-teacher ratio of 25:1 was used. The activities were conducted in both small and large groups. The central role of the intervention provider was to act as a skills trainer or coach because the intervention emphasized the acquisition of personal, interpersonal, and social resistance skills. Cognitive-behavioral skills were taught by combining instruction, demonstration, behavioral rehearsal, feedback, social reinforcement (praise), and extended practice in the form of behavioral homework assignments. We attempted to measure and maximize program fidelity by monitoring teachers randomly as they provided the intervention to students. A monitoring observation form was created that included a checklist of the program points actually covered as well as several qualitative assessments of the classroom teachers that implemented the program. Measures Demographic data on gender, date of birth, family structure, race, and socioeconomic status (receive free school lunch) were collected using standard survey items. Substance use. Students were asked “Have you smoked a cigarette or even part of a cigarette?” and responded yes or no in terms of (a) ever in your life, (b) in the past year, and (c) in the past month. Similar questions were asked for chewing tobacco (“Have you tried chewing tobacco or snuff . . . ?”) and alcohol (“Have you had a drink of beer or wine . . . ?”).

Botvin et al.

7

In addition, a summary score representing the recency of substance use was created for smoking, chewing tobacco, and alcohol use, ranging from 0 (never used) to 3 (used in past month). Peer and parental attitudes. Participants were asked “How would your friends feel if you smoked cigarettes?” and “How would your friends feel if you drank alcohol?” with response options including 1 (they’d be against it), 2 (they wouldn’t care either way), and 3 (they’d be for it). Similar questions were asked in terms of parents’ attitudes regarding smoking and alcohol use. Peer, teen, and adult normative expectations. Participants were asked “How many people your age smoke cigarettes?” and “How many people your age drink beer or wine?” with response options including 1 (none), 2 (only a few), and 3 (most). Similar questions were asked in terms of teenagers’ smoking and alcohol use, and adults’ smoking and alcohol use. Attitudes about smoking and alcohol. Six items adapted from Teenager’s Self-Test: Cigarette Smoking (US Public Health Service, 1974) were used to assess attitudes about smoking (e.g., “Smoking cigarettes make you look cool” and “Kids who smoke have more friends”). Six similar items assessed drinking attitudes. Response options were 1 (agree), 2 (not sure), and 3 (disagree). Knowledge. A variety of true/false items assessing students’ knowledge of the program content were included on the survey. Knowledge was assessed in terms of smoking and drinking (9 items, e.g., “Cigarette smoking can cause your skin to wrinkle”), and skills such as decision-making knowledge (3 items, e.g., “It is best to make decisions quickly”), advertising knowledge (2 items, e.g., “Advertisements always tell the truth”), stress knowledge (2 items, e.g., “Stress can cause you to get sick”), communication skills knowledge (4 items, e.g., “Body language is an important way to communicate your feelings”), social skills knowledge (3 items, e.g., “If you want to have friends, you must be a good friend yourself”), and refusal skills knowledge (3 items, e.g., “You should never say ‘no’ when someone asks you to do something”). Additional psychosocial measures. Two psychosocial factors potentially relevant to the initiation of tobacco and alcohol use were assessed: risk-taking and self-esteem. Four items (␣ = .62) that assess impulsive and daring behavior and were adapted from the Eysenck Personality Inventory (Eysenck & Eysenck, 1975) to measure risk-taking, such as “I would do almost anything on a dare.” Five items (␣ = .68) adapted from the Rosenberg (1965) Self-Esteem scale assessed students’ self-evalua-

8


tions, such as “I am happy with myself.” Response options were on a 3-point scale including 1 (disagree), 2 (not sure), and 3 (agree). Data Analysis Data analysis procedures included chi-square and t-tests to examine pretest equivalence between intervention and control students and schools. Program effects were examined both using the individual and school as the unit of analysis. A series of GLM ANCOVAs were used to examine program effects of substance use, normative expectations, attitudes, and related variables. For the individual-level analyses, control variables included gender, race, family structure, and the pretest score on the outcome variable. For the school-level analyses, student scores were averaged for each school (based on the 1,090 students with matched pre-post data). A series of GLM ANCOVAs were used to examine program effects at the school-level, with the relevant school mean at the pretest assessment used as a covariate in each analysis. RESULTS Attrition Analysis An analysis comparing students that were matched from pretest to posttest to those who were not showed that the matched pre-post sample had a lower lifetime smoking prevalence rate of 8.6% relative to a rate of 11.0% for the unmatched students at the pretest assessment, ␹2 (1) = 3.2, p < .044. Furthermore, the matched sample had a higher proportion of white students (48%) relative to unmatched students (38%), ␹2 (1) = 25.9, p < .001; a higher proportion of two-parent families (70%) relative to unmatched students (32%), ␹2 (1) = 369.8, p < .001; and a higher proportion of households with one or both parents being college graduates (41%) relative to unmatched students (20%), ␹2 (1) = 141.1, p < .001. Thus the matched pre-post sample, which was used for all subsequent analyses in the present study, was generally at less risk than the entire school sample. Pretest Equivalence Additional analyses were conducted to examine pretest equivalence between matched students (N = 1,090) in the experimental versus control conditions. As shown in Table 1, intervention and control students were similar on all substance use measures with the exception of annual prevalence rates for chewing tobacco, for which the intervention students reported a higher rate (1.6%) relative to the control group (0.3%),

Botvin et al.

9

␹2 (1) = 5.7, p < .021. In terms of demographic variables, the intervention group had a lower proportion of White students (44.1%) relative to controls (50.9%), ␹2 (1) = 4.77, p < .017; and the intervention group had a higher proportion of students from two-parent families (73.0%) relative to controls (67.6%), ␹2 (1) = 3.56, p < .034. The intervention and control groups had similar proportions of students in each grade level. As shown in Table 2, when student data were averaged across schools, the intervention and control schools were similar on all substance use and demographic variables, with the exception of the mean annual prevalence rate for chewing tobacco (also higher in intervention schools). Intervention Effects Several GLM ANCOVA analyses were conducted to examine to what extent the program had an impact on the outcome variables of toTABLE 1. Pretest Equivalence for Intervention and Control Students Intervention Students (n = 426)

Control Students (n = 664)

%

%

F2 (df)

p -value

Substance Use Behavior Smoking Prevalence Lifetime

9.9

7.8

1.34 (1)

.148

Annual

4.2

3.2

0.84 (1)

.225

Month

1.6

0.9

1.20 (1)

.208

Chewing Tobacco Prevalence Lifetime

3.5

2.6

0.84 (1)

.231

Annual

1.6

0.3

5.70 (1)

.021

Month

0.7

0.2

2.17 (1)

.169

Alcohol Use Prevalence Lifetime

34.3

34.9

0.04 (1)

.443

Annual

16.9

15.4

0.43 (1)

.283

Month

7.5

6.5

0.42 (1)

.299

Demographic Variables White

44.1

50.9

4.77 (1)

.017

Two-parent families

73.0

67.6

3.56 (1)

.034

Both parents college graduates

20.4

19.9

1.01 (1)

.603

Note. Individual means were based on students in matched pre-post sample (N = 1,090)

10

JOURNAL OF CHILD & ADOLESCENT SUBSTANCE ABUSE TABLE 2. Pretest Equivalence for Intervention and Control Schools Intervention Schools (n = 9)

Control Schools (n = 11)

Mean (SD)

Mean (SD)

t (df)

p -value

Substance Use Behavior Smoking Prevalence Mean lifetime rate

8.5 (7.1)

8.2 (4.2)

0.14 (18)

.894

Mean annual rate

3.2 (3.1)

3.3 (2.7)

0.09 (18)

.932

Mean month rate

1.2 (2.1)

1.1 (1.7)

0.06 (18)

.951

Chewing Tobacco Prevalence Mean lifetime rate

3.2 (2.6)

2.5 (2.3)

0.61 (18)

.552

Mean annual rate

1.6 (1.4)

0.3 (0.8)

2.52 (18)

.027

Mean month rate

0.6 (1.1)

0.2 (0.7)

0.89 (18)

.387

Alcohol Use Prevalence Mean lifetime rate

35.3 (10.7)

34.9 (10.9)

0.09 (18)

.933

Mean annual rate

18.3 (7.4)

16.2 (7.2)

0.65 (18)

.523

Mean month rate

7.9 (7.5)

7.2 (3.4)

0.29 (18)

.778

Mean proportion males

49.0 (7.5)

53.2 (5.4)

1.44 (18)

.167

Mean proportion White

53.4 (38.1)

53.7 (21.8)

0.02 (18)

.983

Mean proportion two-parent families

74.6 (11.5)

68.8 (12.3)

1.10 (18)

.286

Demographic Variables

Note. School means were based on students in matched pre-post sample (N = 1090)

bacco and alcohol use behavior, attitudes, knowledge, and normative expectations, as well as risk-taking and self-esteem. Individual level. Table 3 shows the adjusted posttest means for intervention and control group students. These analyses showed that intervention students at the posttest assessment reported less smoking in the past year, F (1,1059) = 3.15, p < .038, higher anti-drinking attitudes, F (1,1058) = 2.93, p < .044, increased substance use knowledge, F (1,736) = 3.47, p < .031, and increased skills-related knowledge, relative to control students. In addition, intervention students reported lower normative expectations for peer smoking, F (1,1046) = 4.31, p < .019, and alcohol use, F (1,1056) = 12.84, p < .000, as well as higher self-esteem, F (1,1046) = 6.21, p < .006, relative to control students. Intervention students also perceived that their friends had greater anti-drinking attitudes at the posttest relative to control students, F (1,1055) = 3.82, p < .025.

Botvin et al.

11

School level. Table 4 shows the school means for the outcome variables at posttest for intervention and control schools, adjusted for the school mean at the pretest assessment. These analyses showed that intervention schools were significantly lower than control schools at posttest in terms of mean smoking frequency, F (2,17) = 5.92, p < .013, annual smoking prevalence F (2,17) = 9.57, p < .003, and annual drinking prevalence F (2,17) = 2.89, p < .054. The intervention schools also had higher scores for mean anti-drinking attitudes, F (2,17) = 2.99, p < .051, and mean self-esteem F (2,17) = 6.04, p < .013. The mean level of perceived friends anti-drinking attitudes were also higher in intervention schools relative to control schools, F (2,17) = 3.55, p < .038. DISCUSSION The present study tested the effectiveness of cognitive-behavioral substance abuse prevention approach, called Life Skills Training (LST), with elementary school students. Prior research has demonstrated that the LST approach can prevent tobacco, alcohol, and illicit drug use with adolescents attending middle or junior high school (Botvin, Baker et al., 1984; Botvin, Baker, Dusenbury et al., 1990, 1995; Botvin, Eng, & Williams, 1980; Botvin, Epstein et al., 1997; Botvin, Griffin et al., 2000; Botvin, Renick, & Baker, 1983; Botvin, Schinke et al., 1995). In the present study, schools were randomly assigned either to receive a new version of the LST approach designed for elementary school students in grades 3, 4, and 5 or to a comparison (control) group. Results of analyses conducted at both the individual and school levels indicated that the elementary LST program reduced the annual prevalence of cigarette smoking and alcohol use. More specifically, individual-level analyses controlling for gender, race, and family structure showed that intervention students reported less smoking in the past year, higher anti-drinking attitudes, increased substance use knowledge and skills-related knowledge, lower normative expectations for smoking and alcohol use, and higher self-esteem at the posttest assessment, relative to control students. School-level analyses showed that annual prevalence rates for smoking and alcohol use were lower at the posttest assessment in schools that received the prevention program than among control schools. In addition, mean self-esteem scores were higher in intervention schools at the posttest assessment relative to control schools. Taken together, these findings provide preliminary evidence that a school-based substance abuse prevention

12


TABLE 3. Covariate Adjusted Individual-level Means for Intervention and Controls Students at Posttest Assessment Intervention Students (n = 426) Mean SE Substance Use Behavior Smoking frequency Smoke in past year Chewing tobacco frequency Chew tobacco in past year Drinking frequency Drink in past year Substance Use Attitudes Anti-smoking attitudes Anti-drinking attitudes Friends’ anti-smoking attitudes Friends’ anti-drinking attitudes Knowledge Smoking knowledge Substance use knowledge Decision-making knowledge Advertising knowledge Stress knowledge Communication knowledge Social skills knowledge Refusal skills knowledge Normative Expectations Peer smoke norms Teen smoke norms Adult smoke norms Peer drink norms Teen drink norms Adult drink norms Psychosocial Risk-taking Self-esteem

Control Students (n = 664) Mean

SE

F

df

p-value

0.062 0.024 0.019 0.009 0.442 0.149

0.019 0.008 0.010 0.004 0.040 0.016

0.100 0.042 0.015 0.007 0.413 0.168

0.015 0.007 0.008 0.003 0.033 0.013

2.47 3.15 0.08 0.13 0.32 0.90

1,962 1,1059 1,1022 1,1056 1,753 1,1057

.058 .038 .386 .358 .287 .172

2.836 2.790 3.265 3.171

0.011 0.014 0.044 0.044

2.833 2.760 3.197 3.059

0.009 0.011 0.035 0.036

0.05 2.93 1.49 3.82

1,1058 1,1058 1,1055 1,1055

.410 .044 .111 .025

4.389 5.351 1.979 1.563 1.190 2.661 1.747 1.817

0.072 0.091 0.036 0.025 0.034 0.061 0.047 0.052

4.118 5.139 1.873 1.570 1.106 2.516 1.744 1.789

0.053 0.068 0.029 0.020 0.027 0.045 0.035 0.039

9.14 1,736 3.47 1,736 5.33 1,1062 0.05 1,1062 3.70 1,1062 3.59 1,736 0.00 1,736 0.170 1,731

.001 .031 .011 .412 .027 .029 .476 .340

0.574 1.433 1.200 1.241 0.579 1.570

0.030 0.029 0.034 0.035 0.030 0.027

0.653 1.517 1.252 1.402 0.603 1.606

0.024 0.023 0.027 0.028 0.024 0.021

4.31 5.03 1.47 12.84 0.38 1.13

1,1046 1,1050 1,1052 1,1056 1,1052 1,1060

.019 .013 .113 .000 .270 .145

1.753 2.633

0.026 0.022

1.794 2.562

0.021 0.018

1.50 6.21

1,1040 1,1046

.110 .006

Note. Individual means were based on students in matched pre-post sample (N = 1,090); Covariates for all analyses were gender, race, family structure, and pretest level of outcome variable; p-values represent one-tailed significance levels.

approach previously found to be effective among middle school students is also effective for elementary school students. These findings are important because they show that the LST prevention approach produces prevention effects on tobacco and alcohol use with elementary school students. Although previous studies have shown

Botvin et al.

13

TABLE 4. School-level Means for Intervention and Control Schools at Posttest Assessment

Substance Use Behavior Smoking frequency Proportion smoke in past year Chewing tobacco frequency Proportion chew tobacco in

past year

Drinking frequency Proportion drink in past year Substance Use Attitudes Anti-smoking attitudes Anti-drinking attitudes Friends’ anti-smoking attitudes Friends’ anti-drinking attitudes Knowledge Smoking knowledge Substance use knowledge Decision-making knowledge Advertising knowledge Stress knowledge Communication knowledge Social skills knowledge Refusal skills knowledge Normative Expectations Peer smoke norms Peer drink norms Teen smoke norms Teen drink norms Adult smoke norms Adult drink norms Psychosocial Risk-taking Self-esteem

Intervention Schools (n = 9) Mean SE

Control Schools (n = 11) Mean

SE

F

df

p-value

0.049 0.017 0.014 0.006

0.017 0.007 0.009 0.004

0.105 0.043 0.014 0.006

0.016 0.006 0.008 0.004

5.92 9.57 0.00 0.00

2,17 2,17 2,17 2,17

.013 .003 .497 .482

0.436 0.134

0.057 0.019

0.520 0.178

0.051 0.017

1.18 2.89

2,17 2,17

.146 .054

2.841 2.792 3.286 3.181

0.014 0.013 0.044 0.053

2.831 2.762 3.214 3.047

0.012 0.011 0.040 0.048

0.28 2.99 1.43 3.55

2,17 2,17 2,17 2,17

.300 .051 .125 .038

4.321 5.327 1.977 1.565 1.211 2.590 1.724 1.796

0.080 0.106 0.043 0.054 0.069 0.100 0.057 0.050

4.279 5.280 1.907 1.637 1.210 2.587 1.796 1.877

0.072 0.095 0.038 0.048 0.061 0.090 0.051 0.046

0.14 0.10 1.44 0.93 0.00 0.00 0.88 1.39

2,17 2,17 2,17 2,17 2,17 2,17 2,17 2,17

.358 .376 .123 .174 .495 .492 .181 .127

0.541 1.292 1.447 0.550 1.201 1.595

0.044 0.064 0.043 0.056 0.050 0.033

0.612 1.418 1.522 0.586 1.262 1.625

0.040 0.058 0.039 0.051 0.045 0.030

1.41 2.11 1.65 0.22 0.82 0.47

2,17 2,17 2,17 2,17 2,17 2,17

.126 .082 .108 .322 .189 .251

1.716 2.678

0.035 0.026

1.764 2.593

0.032 0.023

1.01 6.04

2,17 2,17

.164 .013

Note. School means were based on students in matched pre-post sample (N = 1,090); Covariates for all analyses were school-wide level of outcome variable at pretest; p-values represent one-tailed significance levels.

that this approach is effective with middle or junior high school students, this is the first study to show that it is also effective with a younger population. Moreover, this is one of the first randomized trials to show that a school-based prevention approach targeting elementary school students can produce significant and meaningful reductions in

14


both tobacco and alcohol use. For example, there was a 61% reduction in the annual rate of smoking (Eta-squared = .36), and a 25% reduction in the annual rate of alcohol use (Eta-squared = .17).1 In addition, this study has several strengths that provide further support for the validity of these findings. First, the study used a true experimental design in which schools were randomly assigned to either prevention or control conditions. This eliminated the potential of contamination across conditions that would otherwise be present in a design in which classes within the same school were assigned to experimental conditions. Second, many prevention studies testing school-based approaches have been criticized for failing to adequately control for the observed correlations of individuals within assignment clusters (e.g., schools), which may inflate the Type I error rate (Murray et al., 1990). A variety of methods exist for dealing with ICCs including either statistically adjusting for school-level ICCs or conducting analysis on the cluster (school) level using approaches similar to that used in the current study. Other strengths of the current study include the use of developmentally appropriate assessment instruments and data collection procedures designed to enhance the validity of students’ self-reports. At the same time, this study also has several limitations that should be noted. First, the findings presented are from the immediate posttest after a single year of intervention. Additional follow-up data are needed to determine the longer-term durability of these prevention effects with this population. Second, since this was a school-based study that relied on students’ self-reports, the significant relationships among variables may partly reflect shared method variance (all data was obtained by self-report questionnaire). Third, in an effort to use measures that were age-appropriate, some measures used fewer response options than in research using similar measures with older populations. Moreover, only some of the outcome measures used in this study were affected by the intervention. Fourth, it is unclear whether these findings generalize to populations different from that included in this study. For example, because students who are absent during data collection are likely to be at somewhat higher risk than those who provided pretest and posttest data, it is unclear to what extent the findings of this study are applicable to high-risk students. It is also unclear to what extent these findings are applicable to different ethnic subgroups. Because the population included in this study was racially/ethnically diverse, it is likely that this prevention approach is effective with multiple populations. However, neither the design of the current study nor the sample size was adequate to empirically test the relative efficacy of this approach with the different populations comprising this sample. Similarly, it was not possible to determine the relative efficacy of the prevention program by age or grade

Botvin et al.

15

level. Additional research is needed to determine the extent to which this prevention approach is effective with different populations as well as to determine whether there were developmental differences in the efficacy of the prevention program for the different age groups/grade levels involved in the study. The results of this study provide the first evidence for the effectiveness of a comprehensive school-based prevention approach designed for elementary school students that teaches substance use resistance skills, normative education, personal self-management skills, and general social skills. These results extend prior research with the LST approach by demonstrating its effectiveness with a younger population. This study is also important because it uses rigorous research methods including a randomized control group design and data analysis methods that adjusted for school-level intra-cluster correlations. Finally, this study is important because prevention programs administered during elementary school either alone or in combination with similar programs during middle or junior high school offer the potential of producing a greater impact on adolescent substance use. Future research is needed to replicate the findings of this study and determine the impact of this prevention approach on a wider array of outcome measures. Research is also needed to determine the durability of these prevention effects with longer-term follow-up assessments, the impact of this prevention approach on high risk youth and other subgroups of youth, and the extent to which these findings generalize to other populations. NOTE 1. The percent reductions in smoking and drinking were calculated by taking the difference score in annual prevalence rates at the posttest assessment between the experimental and control conditions and dividing this score by the annual prevalence rate in the control condition at posttest. For example, the 61% reduction in the annual rate of smoking was calculated as follows: 0.043-0.017 = 0.026/0.043 = 0.605, and the 25% reduction in the annual rate of alcohol use was calculated as follows: 0.178-0.134 = 0.044/0.178 = 0.247.

REFERENCES Bangert-Downs, R. L. (1988). The effects of school-based substance abuse education: A meta-analysis. Journal of Drug Education, 18, 243-265. 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, 1106-1112.

16


Botvin, G. J., Baker, E., Dusenbury, L., Tortu, S., & Botvin, E. M. (1990). Preventing adolescent drug abuse through a multimodal cognitive-behavioral approach: Results of a three-year study. Journal of Consulting & Clinical Psychology, 58, 437-446. Botvin, G. J., Baker, E., Renick, N., Filazzola, A. D., & Botvin, E. M. (1984). A cognitive-behavioral approach to substance abuse prevention. Addictive Behaviors, 9, 137-147. 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: Examining the comparability of videotape, cartoon, and verbal bogus pipeline procedures. Psychological Reports, 55, 379-386. Botvin, G. J., & Eng, A. (1982). The efficacy of a multicomponent approach to the prevention of cigarette smoking. Preventive Medicine, 11, 199-211. Botvin, G. J., Eng, A., & Williams, C. L. (1980). Preventing the onset of cigarette smoking through life skills training. Journal of Preventive Medicine, 9, 135-143. 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 & Adolescent Substance Abuse, 6, 5-19. Botvin, G. J., & Griffin, K. W. (1999). Preventing drug abuse. In A. J. Reynolds, R. P. Weissberg, & H. J. Walberg (Eds.), Positive outcomes in children and youth: Promotion and evaluation (pp. 197-228). Thousand Oaks, CA: Sage Publications. Botvin, G. J., Griffin, K. W., Diaz, T., Scheier, L. M., Williams, C., & Epstein, J. A. (2000). Preventing illicit drug use in adolescents: Long-term follow-up data from a randomized control trial of a school population. Addictive Behaviors, 25, 769-774. Botvin, G. J., Renick, N., & Baker, E. (1983). The effects of scheduling format and booster sessions on a broad spectrum psychosocial approach to smoking prevention. Journal of Behavioral Medicine, 6, 359-379. Botvin, G. J., Scheier, L. M., & Griffin, K. W. (in press). Preventing the onset and developmental progression of adolescent drug use: Implications for the gateway hypothesis. In D. B. Kandel (Ed.), Stages and pathways of involvement in drug use: Examining the gateway hypothesis. Oxford: Cambridge Press. 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. Elder, J. P., Perry, C. L., Stone, E. J., Johnson, C. C., Yang, M., Edmundson, E. W., Smyth, M. H., Galati, T., Feldman, H., Cribb, P., & Parcel, G. S. (1996). Tobacco use measurement, prediction, and intervention in elementary schools in four states: The CATCH study. Preventive Medicine, 25, 486-494. Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual of the Eysenck personality questionnaire. London: Hodder & Stoughton. Hansen, W.B. (1992). School-based substance abuse prevention: A review of the state of the art in curriculum, 1980-1990. Health Education Research: Theory & Practice, 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.

Botvin et al.

17

Kandel, D. (1975). Stages in adolescent involvement in drug use. Science, 190, 912-914. Murray, D. M., & Hannan, P. J. (1990). Planning for the appropriate analysis in school-based drug-use prevention studies. Journal of Consulting & Clinical Psychology, 58, 45-468. Newcomb, M.D. & Bentler, P.M. (1988). The consequences of adolescent drug use: Impact on the lives of young adults. Newbury Park, CA: Sage Publications, Inc. O’Loughlin, J., Paradis, G., Renaud, L., & Sanchez, G. L. (1998). One-year predictors of smoking initiation and of continued smoking among elementary schoolchildren in multi-ethnic, low-income, inner-city neighborhoods. Tobacco Control, 3, 268-75. O’Loughlin, J., Renaud, L., Paradis, G., Meshefedjian, G., & Zhou, X. (1996). Prevalence and correlates of early smoking among elementary schoolchildren in multiethnic, lower-income inner city neighborhoods. Annals of Epidemiology, 8, 308-18. Petraitis, J., Flay, B. R., & Miller, T. Q. (1995). Reviewing theories of adolescent substance use: Organizing pieces in the puzzle. Psychological Bulletin, 117, 67-86. Price, J. H., Beach, P., Everett, S., Tellijohann, S. K., & Lewis, L. (1998). Evaluation of a three year urban elementary school prevention program. Journal of School Health, 68, 26-31. Rosenberg, M. (1965). Society and the adolescent self-image. Princeton: Princeton University Press. Sarvela, P. D, Monge, E. A., Shannon, D. V., & Nawrot, R. (1999). Age at first use of cigarettes among rural and small town elementary school children in Illinois. Journal of School Health, 69, 398-402. SAS Institute (2000). SAS/STAT User’s Guide, Version 8. Cary, NC: SAS Publications. Shope, J. T., Dielman, T. E., Butchart, A., Campanelli, P. C., & Kloska, D. D. (1992). An elementary school-based alcohol misuse prevention program follow-up evaluation. Journal of Studies on Alcohol, 53, 106-121. Tobler, N. S., & Stratton, H. H. (1997). Effectiveness of school-based drug prevention programs: A meta-analysis of the research. Journal of Primary Prevention, 18, 71-128.