Contact Name
Barbara Dietsch
562-985-9488

Contact Email
bdietsc@WestEd.org

Douglas Kirby, Ph.D.
ETR Associates

As this country has become increasingly conservative, there has been an increased emphasis upon abstinence-only programs in schools. This emphasis is manifested by the federal allocation of $50 million per year for abstinence-only programs and by the increasing number of schools that implement abstinence-only programs. If this much money and an increasing amount of school time and effort are directed toward abstinence-only programs, it is important to assess what research tells us about the actual impact of these programs.

Before examining the studies measuring the impact of abstinence-only programs, it is important to realize that abstinence-only programs are a very diverse group of programs that are defined by one common quality-their emphasis upon abstinence as the only appropriate choice for young people. Thus, some abstinence-only programs are curriculum-based, while others include a wide variety of youth development activities. Some last for 15 to 20 sessions, while others last only one or two sessions. Some emphasize that it is immoral to have sex prior to marriage, while others encourage youth to postpone sex until a later age. Some are very religious and begin with a prayer for God's guidance, while others are far more secular. Some are very didactic, while others engage the participants in group activities and use role playing and other active-learning strategies to change group norms and teach assertiveness skills. Truly, abstinence-only programs are a heterogeneous group of programs.

Measuring the short-term impact of some abstinence-only programs upon knowledge, attitudes and values is relatively easy. If the duration of programs is short, the pretest-posttest evaluations can measure any changes that occurred during the program. If longer term effects are measured, then a good comparison or control group is needed to adjust for maturation and other factors that may change attitudes.

Several studies have documented that abstinence-only programs can increase knowledge and change attitudes and values in the direction favorable to abstinence. Some of these studies measured very short term effects, while others measured longer term effects up to 18 months or longer.

However, these results should be viewed somewhat cautiously for several reasons. First, response biases may have substantially inflated the findings and reduced their validity. For example, when a teacher emphasizes that it is wrong to have sex before marriage and then promptly asks students to complete a questionnaire which asks them whether premarital sex is wrong, some students may select the "correct" answer even though their own attitudes may not have changed. Second, although attitudes and values are modestly related to actual initiation of sex, changes in attitudes and values may not translate into significant delay in the initiation of sex. Finally, some studies indicate that the impact of abstinence programs upon attitudes and values does diminish with time.

Measuring the impact of abstinence-only programs upon delay in initiation of sex is much more challenging. There are at least two reasons for this. First, the percentage of a cohort of youth who have initiated sex can only increase with time. Thus, simple pretest-posttest comparisons cannot show a positive impact by decreasing the percentage who have ever had sex. Rather, a large and very well matched comparison group is needed to demonstrate that a larger percentage of youth in the comparison group initiated sex in a given period of time than youth in the abstinence program. Second, a relatively small percentage of youth initiate sex in a short period of time. Thus, researchers must measure long term effects in order for a significantly larger percentage of youth in the control group than in the abstinence intervention group to initiate sex.

To date, only five studies have measured the impact of abstinence-only programs upon initiation of sex, and none of them found both a consistent and significant impact upon delaying the onset of intercourse; at least one study provided strong evidence that the program did not delay the onset of intercourse. Thus, the weight of the evidence indicates that abstinence-only programs do not delay the onset of intercourse.

On the other hand, this evidence is not strong, because all but one of these evaluations had significant methodological limitations that could have obscured program impact. For example, two of the studies measured the impact of the program for only six weeks after the end of the program, and during that brief period of time, too few youth in the comparison group initiated sex in order for the program group to have had significantly fewer youth initiate sex. Another study included only 91 study participants.

It is also true that just as some people are concerned that sex education programs that cover both abstinence and contraception may hasten the onset of sex or increase sexual behavior, other people are concerned that abstinence-only programs may decrease the use of contraception once participants initiate sex. Thus far, only one study has measured that impact well, and it failed to find any significant impact one way or the other upon contraceptive use.

This review has summarized the results found in published studies. Another review by Brian Wilcox conducted a meta-analysis of many AFLA-supported studies, both published and unpublished. That review reached similar conclusions-the studies were very weak and there was no evidence indicating that abstinence-only programs had a positive or negative impact upon sexual and contraceptive behavior.

It seems likely to this author that some abstinence-only programs are likely to delay the onset of sex, especially those which incorporate the characteristics of effective sex and HIV education programs, some of which delayed the onset of intercourse. On the other hand, that cannot be known for certain until rigorous studies with random assignment, sufficiently large sample sizes and long term measurement of behavior are completed.

The results of these studies of abstinence-only programs are in contrast to the results of some sex and HIV/STD programs which have been demonstrated to reduce unprotected sex either by delaying sex, reducing the frequency of sex, reducing the number of sexual partners, or increasing the use of protection against STD and pregnancy. There are now several programs with strong evidence that they are effective at changing behavior, and there are other programs with weaker evidence that they also reduced unprotected sex.

In conclusion, given the great diversity of abstinence-only programs with possible differential effects, and given the many limitations of most previous studies that measured changes in attitudes and values or that measured change in sexual behavior, there is simply too little evidence to determine whether different types of abstinence programs actually delay the onset of intercourse or have other positive effects upon sexual and contraceptive behavior. If educators, policy makers and others want to implement programs with strong evidence for delaying sex or reducing unprotected sex, then they need to implement effective programs that emphasize both abstinence and the use of contraception for sexually experienced youth.