Executive Summary
"Can Abstinence Work? An Analysis of the Best Friends Program"
By Robert Lerner, Ph.D.

INTRODUCTION

The study summarized here and published in the Adolescent & Family Health journal analyzed data from the Best Friends program in order to ascertain the extent of its success in preventing smoking, alcohol use, drug use, and early onset sex among middle school aged girls. The Best Friends program is one of the oldest abstinence programs in the United States. Founded in 1987, the program operates today in more than 100 schools in the United States.

The Best Friends program implements an extensive yearlong curriculum. It stresses the development of decision-making skills among junior high and middle school aged girls in order to assist them in avoiding alcohol use and drug use as well as sex. The program is continued at the high school level with the Diamond Girls Leadership program. The male counterpart program is Best Men, which was established in 2000 and not studied here.

METHODOLOGY

While an important qualitative study of the Best Friends program was carried out some time ago, Best Friends has not been studied using quantitative techniques until now.

The data analyzed in this study consist of responses to questionnaires filled out by program participants at the beginning and at the end of the school year. These are compared to the responses of girls that were part of the Youth Risk Behavior Surveys (YRBS) of Washington, D.C. These surveys are modeled after the national Youth Risk Behavior Survey conducted by the Center for Disease Control every other year. This study compares Best Friends participants in grades 6 through 8 and Diamond Girl Leadership participants in grades 9 through 12 to the same school grade YRBS data.

Best Friends girls' data were obtained for the 1997-98, 1998-99, 2000-01, and 2001-02 school years. Comparison data were obtained from the Washington, D.C. junior high and middle school YRBS surveys for the years 1997, 1999, and 2001.

RESULTS

Washington, D.C. Best Friends girls attend schools in challenging circumstances. These schools have average reading scores similar to and math scores lower than those from the District as a whole (not shown here). Best Friends schools are also located in wards with higher rates of out-of-wedlock births and higher rates of teen pregnancy than in the city as a whole (data not shown here).

The results reported below were obtained by computing multiple logistic regression equations for each of the risk factors comparing Best Friends and YRBS girls. These predictive equations controlled for respondents' grade, age, race/ethnicity, and the year in which the survey was completed.

Girls who have attended the Best Friends program are substantially less likely to smoke, drink, take drugs, and have sex than are YRBS respondents. The relative odds of Best Friends girls not smoking are 2.40 to 1, the relative odds of Best Friends girls not taking drugs are 8.10 to 1, the relative odds of Best Friends girls not drinking are 1.91 to 1, and the relative odds of Best Friends girls not initiating sex are 6.48 to 1. All these findings are strongly statistically significant (See Table "Risk Indicators...") at the p<.001 level of statistical significance or better.

Even stronger results were found when comparing girls in the high school program, Diamond Girls, with girls surveyed in the D.C. high school YRBS. (This analysis was restricted to the Diamond Girls samples for 2001 and 2002 and the D.C. YRBS high school sample for 2001, due to the unavailability of other data). The results reported below were obtained by computing multiple logistic regression equations for each of the risk factors comparing Diamond Girls and YRBS girls. The predictive equations controlled for respondents' grade, age, and race/ethnicity.

Girls who have attended the Diamond Girls program are substantially less likely to smoke, drink, take drugs, and have sex than are YRBS respondents. The relative odds of Diamond Girls not smoking are 8.91 to 1, the relative odds of Diamond Girls not taking drugs are 26.13 to 1, the relative odds of Diamond Girls not drinking are 3.05 to 1, and the relative odds of Diamond Girls not initiating sex are 119.58 to 1. All these findings are strongly statistically significant (See Table "Diamond Girls vs. High School YRBS...") at the p<.001 level or better.

An analysis of girls who started the Best Friends program but failed to complete a post-test survey shows that dropouts do not differ from those remaining in the program with respect to their risk behavior (not shown here).

DISCUSSION

The results presented provide empirical evidence supporting the efficacy of the Best Friends program and of its high school counterpart, Diamond Girls, in restraining drug use, alcohol use, smoking and the initiation of sex. The program's success occurs despite the fact that the schools in which the program operates are located in neighborhoods with higher rates of out-of-wedlock births and higher rates of teen pregnancy than for the city as a whole.

While the results of the study support the view that the Best Friends program is effective, they have certain limitations that should be noted. The most important limitation was that the YRBS only surveyed girls at the end of the school year. Having YRBS surveys at the beginning as well as at the end of the year would have allowed for a direct comparison of change over time between Best Friends and the YRBS respondents. Other limitations include that this is not a multi-year study and did not follow the girls for more than one school year, which did not allow the tracking of long-term changes in girls' behavior and the lack of questions on both surveys that would have allowed comparison of how decisions are made by Best Friends girls and by those not enrolled in the Best Friends program. Despite these limitations, the study has provided support for the view that the Best Friends program is successful in teaching junior high and middle school aged girls to resist trying drugs, alcohol, smoking, and sex as compared to their peers not enrolled in the program.

TABLES

Risk Indicators by Best Friends Program Membership With Controls (N=2,730)
Not Smoked
Not Used Drugs
Not Drank
Not Had Sex
B
Odds Ratio
B
Odds Ratio
B
Odds Ratio
B
Odds Ratio
YRBS-BF
0.88
2.40***
2.09
8.10***
0.65
1.91***
1.87
6.48***
AGE
-0.27
0.76***
-0.43
0.65***
-0.26
0.77***
-0.67
0.51***
GRADE
-0.31
0.74***
-0.27
0.76**
-0.40
0.67***
-0.11
0.90
YEAR
1997
0.36
1.44
1.03
2.81*
-0.15
0.86
0.78
2.18*
1999
0.16
1.17
1.13
3.11**
0.19
1.21
0.83
2.29*
2001
0.3
1.35
1.09
2.98**
0.41
1.50*
0.79
2.20*
RACE
Black
-0.05
0.95
-1.01
0.36*
-0.21
0.81
-0.78
0.46
Indian
0.34
1.40
-0.90
0.41
0.15
1.17
-0.72
0.49
Asian
1.08
2.94**
-0.44
0.65
1.16
3.19**
0.19
1.20
Other
0.42
1.53
-0.05
0.95
-0.5
0.61
0.43
1.54
Hisp.
0.48
1.61
-0.75
0.47
-0.07
0.93
-0.26
0.77
Mixed
-0.11
0.90
-0.61
0.54
-0.57
0.57
-0.68
0.51
Constant
2.16
8.68***
4.24
69.33***
2.81
16.56***
4.32
75.27***
*p<0.05; **p<0.01; *** p<0.001

 

Diamond Girls Versus High School YRBS with Controls (N=847)
Not Smoked
Not Used Drugs
Not Drank
Not Had Sex
B
Odds Ratio
B
Odds Ratio
B
Odds Ratio
B
Odds Ratio
YRBS-BF
2.19
8.91***
3.26
26.13***
1.11
3.05**
4.78
119.58***
AGE
-0.25
0.78*
-0.41
0.66***
-0.07
0.93
-0.83
0.44***
GRADE
0.23
1.26
0.23
1.26
-0.08
0.92
0.42
1.52**
RACE
Black
-1.24
0.29
0.09
1.10
-1.19
0.30
0.47
1.60
Indian
0.31
1.36
0.05
1.06
0.11
1.11
-0.44
0.64
Asian
1.21
3.36
0.65
1.91
0.10
1.10
0.26
1.30
Other
0.57
1.76
0.85
2.34
0.68
1.97
0.51
1.66
Hisp.
3.33
27.95
3.00
20.02
3.91
49.98
3.44
31.04
Mixed
-0.31
0.73
0.62
1.86
-0.41
0.67
0.53
1.71
Constant
-0.93
0.39
0.75
2.11
0.49
1.64
1.11
3.04
*p<0.05; **p<0.01; *** p<0.001

© 2005 Best Friends Foundation. All Rights Reserved.
Send questions or comments to webmaster@bestfriendsfoundation.org

Best Friends Foundation: News   Donor Info   Parent Tips   Start a New Program   Results   Staff   National Programs   Foundation Home   Girls Home   Guys Home
Best Friends: What is Best Friends?   What We Believe   News   Events   Start a Program
Diamond Girls Leadership   National Sites   Safety Rules
Best Men: What is Best Men?   What We Believe   News   Events   Start a Program
Best Men Leadership   National Sites   Safety Rules