Abstract

This paper synthesizes 35 evaluations of specific school-based programs designed to reduce sexual risk-taking behavior, including sex and HIV education programs, school-based health centers, and school condom-availability programs.  This paper summarizes the effects of those programs on several sexual and contraceptive behaviors.  Although the studies have several limitations that restrict any conclusions, the evidence suggests that a few programs have produced credible evidence that they have reduced sexual risk-taking behavior by delaying the onset of intercourse, by reducing the frequency of intercourse, by reducing the number of sexual partners, or by increasing the use of condoms or other forms of contraception.  The majority of studies, however, found no statistically significant impact on risk-taking behaviors.  Of considerable importance in this controversial field, none of the studies found statistically significant results indicating that any of the programs increased any measure of sexual activity, as is often feared.
 

School-Based Programs

For a variety of reasons, schools have the potential for playing an important role in reducing sexual risk-taking behaviors among adolescents.  Schools are the one institution in our society regularly attended by most young people—nearly 95% of all youth aged 5–17 years are enrolled in elementary or secondary schools (National Center for Education Statistics, 1993).  Moreover, virtually all youth are attending schools before they initiate sexual risk-taking behaviors, and a majority are enrolled at the time they initiate intercourse.  Additionally, schools are especially well suited to educate youth, particularly about topic areas such as sexuality in which different concepts should be taught during different developmental stages.  They are also well designed to involve youth in programs that may motivate them to delay early childbearing.  Finally, schools are capable of identifying at-risk youth and then either providing health and social services directly or referring those youth for such services.

For decades there has been and continues to be widespread support for sexuality and HIV education in schools.  For example, in 1988, a Harris poll indicated that 85% of adults approved of sexuality education in schools (Louis Harris and Associates, 1988).  Further, a 1987 Gallup poll showed that 94% of parents approved of HIV education (Gallup, 1987).

The approval for sexuality and HIV education also is manifested in state policy: 23 states require that sexuality education be taught in schools, while an additional 15 states require that schools offer STD/HIV education (NARAL Foundation, 1995, 1997).  In addition, 47 states either recommend or require the teaching of sexuality education, while all 50 states recommend or require AIDS education programs (DeMauro, 1990; Gambrell & Haffner, 1993).  Thus, the controversies surrounding sexuality and HIV education programs do not focus on whether those programs should be offered in school, but rather on what topics should be taught.  Some groups believe that only abstinence until marriage should be emphasized, whereas other groups believe that contraception and other topics related to sexuality should be covered.

There has also been considerable support for the provision of condoms or contraceptives through school condom-availability programs and school-based health centers.  For example, a 1991 Roper poll indicated that 64% of American adults favor making condoms available in high schools (Roper Organization, Inc., 1991), and a 1988 Harris poll revealed that 73% of American adults favor making contraceptive information and contraceptives available through school-based clinics (Louis Harris and Associates, 1988).

Given the need for effective educational programs and public support for such programs, schools have responded.  A vast majority of junior and senior high schools have developed and/or implemented programs aimed to reduce unprotected sexual intercourse, either by delaying the onset of intercourse or by increasing the use of protection against HIV, other STDs, and pregnancy.  The most common programs are sexuality education and HIV prevention programs.  Indeed, according to a 1994 national study by the Centers for Disease Control and Prevention, approximately 86% of all schools surveyed required instruction on HIV prevention and 80% required instruction on human sexuality (Collins et al., 1995).  Additionally, some schools have school-based or school-linked health centers, some of which provide contraceptives.  Other schools have implemented school condom-availability programs to increase youth access to condoms.  Still others have implemented multicomponent programs, which typically include an educational component and a component aimed to increase access to condoms or other contraceptives.

Studies under Review

This paper reviews the published research on those school-based programs, focusing primarily on their impact on sexual and contraceptive behaviors.  A total of 35 studies are included, which represent the studies known by the authors that meet the following four conditions: (1) they focused on school-based programs (or curriculum-based programs that could be implemented in schools); (2) they employed experimental or quasi-experimental designs; (3) they measured program impact on sexual or contraceptive behavior or pregnancy rates; and (4) they have been published as major reports, have been published in peer-reviewed journals or volumes, or are about to be published in peer-reviewed publications.

Not surprisingly, the quality of the research methods employed and the strength of the resulting evidence for the impact of the programs vary greatly from study to study.  For example, to minimize selection biases, some studies used random assignment of youth to the intervention or comparison conditions, whereas others did not.  Some studies involved large sample sizes to ensure adequate statistical power in detecting statistically significant results, while others used relatively small samples, reducing the chance for finding programmatically significant differences.  Some measured program effects for only three months, while others measured longer-term impacts up to 24 months.  And finally, some used more rigorous statistical analyses, while others failed to control statistically for design limitations.  Because of that variation in quality, the strength of the evidence from each study should be considered when reviewing the results.  Additionally, the strengths of any conclusions about the impact of programs are confined by the limitations both of individual studies and of all the studies as a group.

The Effects of Curriculum-Based Sex and HIV Education Programs

Most adolescents in this country know a considerable amount about the risks of unprotected sexual intercourse and the methods of preventing those risks.  For example, nearly all youth know that unprotected sexual intercourse can lead to pregnancy or STD, and most know that condoms can be obtained at stores and provide protection against pregnancy and STD.  They learn that and other information through a variety of sources, such as their school sex and HIV education programs, the media, their parents and other adults, their peers, and others.  Indeed, innumerable studies have demonstrated that sex and HIV/AIDS education programs do increase knowledge.  Presumably, this information does reduce the amount of unprotected sex among teenagers.  There remain, however, the following important questions:  Given this body of knowledge among adolescents, does additional instruction about different aspects of sexuality affect their sexual behaviors in positive or negative ways?  What are the characteristics of programs that have positive effects?

As a result of the increasing body of research on the effectiveness of sex and HIV education programs, it is now possible to formulate preliminary conclusions regarding their impact on sexual risk behaviors and to analyze characteristics that distinguish effective from ineffective programs.

Abstinence-Only Programs

Abstinence-only programs focus on the importance of abstinence from sexual intercourse, typically abstinence until marriage.  Abstinence programs either do not discuss contraception or briefly discuss contraceptive failure to provide complete protection against pregnancy and STD.  To date, six studies of abstinence programs have been published (Christopher & Roosa, 1990;  Jorgensen, Potts & Camp, 1993; Kirby, Korpi, Barth & Cagampang, 1995; Roosa & Christopher, 1990; St. Pierre, Mark, Kaltreider & Aikin, 1995; Young, Core-Gebhart & Marx, 1992).

None of the six studies found both a consistent and a significant impact on delaying the onset of intercourse, and 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 programs do not delay the onset of intercourse.

On the other hand, that evidence is not strong, because all but one of the evaluations had significant methodological limitations that could have obscured the program's 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 for the program group to have had significantly fewer youth initiate sex (Christopher & Roosa, 1990; Roosa & Christopher, 1990).  Other studies included as few as 91 study participants (Jorgensen, Potts & Camp, 1993).  Given those limitations, there is too little evidence to determine whether or not different types of abstinence programs can delay the onset of intercourse.

Sex and HIV Education Programs That Include Both Abstinence and Contraception

These programs differ from the abstinence-only programs in that they discuss both condoms and other methods of contraception as methods of providing protection against STDs or pregnancy.  This group includes a wide variety of programs, ranging from sex or AIDS education programs taught during school classes, to those taught on school campuses but after school, to programs implemented in community settings.  Because some of the programs originally evaluated in community settings are now taught in classroom settings, we include them in this review, even though their original evaluations were not school-based.

Although this review focuses primarily on the impact of programs on behavior, it should be noted that nearly all evaluations of programs have demonstrated that programs do increase students' knowledge about different aspects of sexuality and contraception.  Although that knowledge may or may not lead to behavior change, it does help build a foundation for better decision making.

Impact on Sexual Activity.  The studies of sexuality and HIV education programs strongly support the conclusion that the programs do not increase sexual intercourse, either by hastening the onset of intercourse or by increasing the frequency of intercourse.  Of the 16 evaluations of middle school, high school, or community sexuality or HIV education programs that measured the impact of the programs on the initiation of intercourse, none found that their respective programs significantly hastened the onset of intercourse (Eisen, Zellman & McAlister, 1990; Ekstrand et al., 1996; Howard & McCabe, 1990; Jemmott, Jemmott & Fong, 1992; Kipke, Boyer & Hein, 1993; Kirby, 1984; Kirby, Barth, Leland & Fetro, 1991; Kirby, Korpi, Adivi & Weissman, 1997; Levy et al., 1995; Main et al., 1994; Nicholson & Postrado, 1991; St. Lawrence, Jefferson, Alleyne & Brasfield, 1995; Thomas et al., 1992; Walter & Vaughn, 1993; Warren & King, 1994).

Similarly, none of the 11 studies that examined the impact of programs on the frequency of intercourse found a significant increase (Howard & McCabe, 1990; Jemmott, Jemmott & Fong, 1992; Kirby, 1985; Kirby, Barth, Leland & Fetro, 1991; Kirby et al., 1997; Levy et al., 1995; Main et al., 1994; Moberg & Piper, 1990; Smith, 1994; St. Lawrence et al., 1995; Walker & Vilella-Velez, 1992).  Thus, the data strongly indicate that sex and HIV education programs do not significantly increase sexual activity as some people have feared.

Furthermore, those studies indicate that some, but not all, of the programs reduced sexual behavior, either by delaying the onset of intercourse or by reducing the frequency of intercourse.  Four of the 13 studies that examined the impact of programs on the initiation of intercourse found evidence that their respective programs significantly delayed the onset of intercourse (Ekstrand et al., 1996; Howard & McCabe, 1990; Kirby, Barth, Leland & Fetro, 1991; St. Lawrence et al., 1995).  Five of the 11 studies that measured program impact on frequency of intercourse found evidence that their programs reduced the frequency of intercourse (Howard & McCabe, 1990; Jemmott, Jemmott & Fong, 1992; Levy et al., 1995; Smith 1994; St. Lawrence et al., 1995).

Impact on Number of Partners.  Consistent with the results regarding impact on the initiation and frequency of sexual activity, none of the seven studies that examined program impact on the number of sexual partners found a significant increase (Jemmott, Jemmott & Fong, 1992; Kipke, Boyer & Hein, 1993; Kirby et al., 1997; Levy et al., 1995; Main et al., 1994; St. Lawrence et al., 1995; Walter & Vaughn, 1993).  To the contrary, about half the studies (four of seven) that measured program impact on the number of sexual partners found that programs decreased the number of partners.

Impact on Use of Condoms and Other Contraception.  Studies examining program impact on condom and other contraceptive use suggest that some, but not all, of the programs increased condom use or contraceptive use more generally.  Four of the eight studies that examined program impact on condom use found that the programs did increase some measure of condom use (Jemmott, Jemmott & Fong, 1992; Kipke, Boyer & Hein, 1993; Kirby et al., 1997; Levy et al., 1995; Main et al., 1994; St. Lawrence et al., 1995; Walter & Vaughn, 1993; Warren & King, 1994).  Similarly, 4 of the 10 studies that examined program impact on the use of contraception more generally found significant positive results (Eisen, Zellman & McAlister, 1990; Howard & McCabe, 1990; Kirby, 1984; Kirby, Barth, Leland & Fetro, 1991; Kirby et al., 1997; Nicholson & Postrado, 1991; Schinke et al., 1981; Smith, 1994; Thomas et al., 1992; Walker & Vilella-Velez, 1992).  In combination, those results are quite positive, indicating that some sex and HIV education programs can significantly increase condom or contraceptive use, while other programs do not.

A disproportionate number of the programs that significantly increased condom or contraceptive use were AIDS education programs.  Four of the six AIDS education programs found significant effects on condom use, while 4 of 10 sex education programs found significant effects.  It cannot yet be determined whether the AIDS education programs under review are inherently more effective than sex education programs that cover pregnancy, STD, HIV, and other topics, or whether AIDS education programs had simply been better funded, had better training, had studies with larger sample sizes, targeted youth who voluntarily participated, or had some other advantage that might improve measured results.

The data also suggest that sex and AIDS education programs may be more effective with African Americans than with other ethnic groups.  The only two middle school curricula that have presented positive results are Postponing Sexual Involvement and Healthy Oakland Teens.  Notably, both achieved those positive results when implemented among African American youth.  When Postponing Sexual Involvement was implemented among many different ethnic groups in California, no impact was found.  In addition, both Be Proud! Be Responsible! and Be a Responsible Teen, two of the four high school curricula with the strongest evidence for effectiveness, were implemented among African Americans.  Because HIV is more prevalent among heterosexual African Americans than among heterosexual whites or Hispanics, African Americans may be more receptive to the messages of those AIDS curricula.

In general, the evaluation results of curriculum-based sex and HIV education programs are encouraging, but methodological caveats are in order.  Only three of the studies of abstinence education, sex education, or HIV education programs included random assignment, large sample sizes, long-term follow-up, measurement of behavior, and proper statistical analyses (Kirby et al., 1995; Kirby et al., 1997; Thomas et al., 1992), and none of them found significant effects on behavior.  Thus, the strength of the evidence from studies with positive results was reduced by one or more methodological limitations discussed above.  Furthermore, there have been very few replications of programs and their evaluations.  Postponing Sexual Involvement and a second curriculum were implemented together in Atlanta and found to be effective, but when Postponing Sexual Involvement was implemented alone in California, the curriculum did not have any positive effects on behavior (Kirby et al., 1995).  Thus, there do not exist any curricula that have been independently implemented and evaluated in two or more settings and found to be effective at changing behavior.

Effective Curricula

The studies of educational programs raise important questions.  What curricula are most effective at changing sexual risk-taking behaviors, either by delaying or reducing sexual activity or by increasing the use of protection? What are their characteristics?

The Division of Adolescent and School Health within the Centers for Disease Control and Prevention (CDC) has identified four curricula as having particularly strong evidence for success in changing sexual risk-taking behaviors (Education Development Center, 1996): Be a Responsible Teen (St. Lawrence, 1994); Be Proud! Be Responsible! (Jemmott, Jemmott & McCaffree, 1994); Get Real about AIDS (Comprehensive Health Education Foundation, 1994); and Reducing the Risk (Barth, 1996).

When those four curricula and others with less strong evidence of success are compared with curricula without positive behavioral results, the effective curricula share several characteristics, which may be linked to their success, while the ineffective curricula lack one or more of those characteristics.  The characteristics of effective curricula were first published by a panel of experts selected by CDC (Kirby et al., 1994) and subsequently updated by Kirby.  Some of them have also been identified in other reviews of impact studies (Frost and Forrest, 1995; Miller and Paikoff, 1992; Moore, Sugland, Blumenthal, Glei & Snyder, 1995).  Those characteristics reflect different aspects of effective pedagogy.  In addition, they are similar to the characteristics of educational programs found to be effective at reducing substance abuse (Dusenbury & Falco, 1995).

The nine characteristics are:

1.  Effective programs focused clearly on reducing one or more sexual behaviors that lead to unintended pregnancy or HIV/STD infection.  There are two aspects to this characteristic.  First, the effective programs focused narrowly on a small number of specific behavioral goals, such as delaying the initiation of intercourse or using contraception; relatively little time was spent addressing other sexuality issues, such as gender roles, dating, and parenthood.  (It is unclear whether programs covering a more comprehensive array of topics are also effective, because few of them have been well evaluated.) Second, the effective programs gave an unambiguous message by continually reinforcing a clear stance on those behaviors.  They did not simply lay out the pros and cons of different sexual choices and implicitly let the students decide which was right for them; rather, most knowledge facts, most activities, most values, and most skills were directed toward convincing the students that avoiding sex, using condoms, or using other forms of contraception was the right choice.

2.  The behavioral goals, teaching methods, and materials were appropriate to the age, sexual experience, and culture of the students.  For example, programs for younger youth, few of whom had engaged in intercourse, focused on delaying the onset of intercourse.  Programs designed for high school students, some of whom had engaged in intercourse, emphasized that students should avoid unprotected intercourse, either by not having sex or by using contraception if they did have sex.  And programs for higher-risk youth, many of whom were already sexually active, emphasized the importance of using condoms and avoiding high-risk situations.
Some of the curricula, for example, Be a Responsible Teen and Be Proud! Be Responsible!, were designed for specific ethnic groups and emphasized statistics, values, and approaches that were tailored to those groups.

3.  Effective programs were based on theoretical approaches that have been demonstrated to be effective in influencing other health-risk behaviors, for example, social cognitive theory (Bandura, 1986), social influence theory (McGuire, 1972), social inoculation theory (Homans, 1965), cognitive behavioral theory (Bandura, 1986; Schinke, Blythe, Gilchrist & Burt, 1981), and the theory of reasoned action (Fishbein & Ajzen, 1975).  Those theories address many of the individual sexuality-related antecedents of sexual behavior.  Thus, those programs strive to go far beyond the cognitive level; they focus on recognizing social influences, changing individual values, changing group norms, and building social skills.

4.  Effective programs lasted a sufficient length of time to adequately complete a variety of activities.  Shorter programs appeared to have less effect, while longer programs provided the opportunity to complete many of the activities discussed below.  Effective programs tended to fall into two categories: those that lasted 14 or more hours and those that lasted a smaller number of hours but implemented the curriculum in small group settings with a leader for each group.  Those features may have enabled the programs to involve the youth more completely, to tailor the material to each group, and to cover more material and more concerns more quickly in each group.

5.  Effective programs employed a variety of teaching methods designed to involve the participants and have them personalize the information.  Instructors reached students through active learning methods of instruction, not through didactic instruction.  Students were involved in numerous experiential classroom and homework activities: small group discussions, games or simulations, brainstorming, role-playing, written rehearsal, verbal feedback and coaching, locating contraception in local drugstores, visiting or telephoning family planning clinics, and interviewing parents.

In addition to using experiential activities, a few effective curricula used peer educators or videos with characters (either real or acted) who resembled the students and with whom the students could identify.  All those activities helped the students personalize the information.

6.  Effective programs provided basic, accurate information about the risks of unprotected intercourse and methods of avoiding unprotected intercourse.  Although increasing knowledge was not the primary goal of those programs, effective programs provided basic information that students needed to assess risks and avoid unprotected sex.  Typically, that information was not unnecessarily detailed or comprehensive.  For example, the curricula did not provide detailed information about all methods of contraception or different types of STD.  Instead, they emphasized the basic facts needed to make behaviorally relevant decisions.

7.  Effective programs included activities that address social pressures on sexual behaviors.  The activities took several forms.  For example, several curricula discussed situations that might lead to sex.  Most of the curricula discussed "lines" that are typically used to get someone to have sex, and some discussed social barriers to using contraception (e.g., embarrassment about buying condoms) and how to overcome those barriers.  Some curricula also addressed peer norms.

8.  Effective programs provided modeling and practice of communication, negotiation, and refusal skills.  Typically, the programs provided information about the skills, modeled effective use of the skills, and then provided some type of skill rehearsal and practice (e.g., verbal role-playing or written practice).  There were, however, significant variations in the quality and amount of time devoted to skill practice.

9.  Effective programs selected teachers or peers who believed in the program they were implementing and then provided training for those individuals.  The training ranged from approximately six hours to three days.  In general, the training was designed to give teachers and peers information on the programs as well as practice using the teaching strategies included in the curricula (e.g., conducting role plays or leading group discussions).

Despite these commonalities, there is very little evidence regarding which of the factors or combinations of factors contributes most to the overall success of the programs.  For example, simply increasing knowledge is not likely to change behavior.  To assume, however, that the positive behavioral effects resulted from just the skill practice or the instruction on social influences would be premature.

In summary, the data from abstinence, sex education, and HIV prevention programs indicate that education programs do not increase sexual intercourse, either by hastening the onset of intercourse, increasing the frequency of intercourse, or increasing the number of sexual partners.  The data also indicate that some, but not all, programs may either reduce one or more measures of sexual activity or increase the use of condoms or other forms of contraception.  Those studies do, however, have severe limitations.

The Effects of School-Based and School-Linked Health Centers

School-based health centers are clinics located in schools that offer services to the students; school-linked clinics are adolescent clinics located near schools that provide many of the same services and can be integrated into the schools.  The clinics typically provide basic primary health care services; some of them also dispense contraceptives.  When the clinics are well staffed and well run and dispense contraceptives, they have many of the characteristics of ideal reproductive health programs, for example, their location is convenient to the students, they can reach both females and males, they provide comprehensive health services, they are confidential, their staff is selected and trained to work with adolescents, they can easily conduct follow-up, their services are cost-free, and they can integrate education, counseling, and medical services.  On the other hand, they may not be open during summer months and they may not easily reach older males, who are most likely to father children born to adolescent females.

School-based and school-linked clinics do provide contraceptives to substantial percentages of sexually experienced youth.  For example, in a study of four clinics that provided prescriptions or actually dispensed contraceptives, the proportion of sexually experienced females who obtained contraceptives through the clinic varied from 23% to 40% (Kirby, Waszak & Ziegler, 1991).

Six studies have examined the impact of those health centers (Edwards, Steinman, Arnold & Hakanson, 1980; Kirby, Waszak & Ziegler, 1991; Kirby et al., 1993; Kisker, Brown & Hill, 1994; Newcomer & Duggan, 1996; Zabin, Hirsh, Smith, Streett & Hardy, 1986).  Five of those studies examined programs in three or more schools.  The outcomes they measured and their quasi-experimental designs varied considerably.  In general, the quasi-experimental designs were not strong.  For example, some did not collect baseline data, and some did not have equivalent comparison groups or comparison groups at all.  In addition, the studies measured population effects.  That is, they measured the effects on the entire school population and not just on those students who actually used the clinics for family planning services.  Consequently, inferences should be drawn cautiously from these studies.

Three of the six studies measured the impact of school-based clinics and school-linked clinics on sexual and contraceptive behaviors (Kirby, Waszak & Ziegler, 1991; Kisker, Brown & Hill, 1994; Zabin et al., 1986).  They evaluated the impact of 6 clinics, 19 clinics, and 1 clinic, respectively.  The Kirby, Waszak, and Ziegler study found that the presence of the clinic did not affect the onset of sexual intercourse either positively or negatively.  In contrast, both the Kisker, Brown, and Hill study and the Zabin study found some data indicating that the clinic and its educational programs may have delayed the onset of intercourse.  The Kirby, Waszak, and Ziegler study and the Kisker, Brown, and Hill study also found that clinic presence was not associated with greater frequency of intercourse.  In combination, those studies indicate that providing contraceptives on campus does not hasten the onset of intercourse nor increase its frequency, as some people have feared.

Regarding contraceptive use, the Kisker, Brown, and Hill study (1994) produced the most negative results—clinic presence was associated with lower rates of contraceptive use.  It is not clear what produced that anomalous result, but it should be noted that their study did have important limitations, for example, baseline data could not be collected.

The results of the Kirby, Waszak and Ziegler study (1991) were more mixed.  At one site where the clinic focused on high-risk youths, emphasized pregnancy prevention, and dispensed oral contraceptives, there was a significantly greater use of oral contraceptives among females than among females in the comparison school; there was no significant difference in condom use.  At two other sites that dispensed both condoms and oral contraceptives but did not have strong educational components, no significant differences were found between the clinic and comparison schools in use of condoms by male students or use of oral contraceptives by female students.  At those schools there clearly were substitution effects—although many sexually experienced students obtained contraceptives from the clinics, most of those students would have obtained contraceptives elsewhere if the clinics had not been there.  The Zabin et al. study (1986) found that clinic presence was associated with greater use of contraception.

All six studies examined the impact of their programs on either pregnancy or birth rates.  Edwards et al. (1980) gave a great impetus to school-based clinics with a report that birth rates declined in three different schools in St. Paul after clinics providing reproductive health care (including prescriptions for contraception) were opened.  But those conclusions were based on only one baseline year for each school and on the staff's knowledge of births among students.  Subsequently, Kirby et al. (1993) overcame those limitations by generating birth rates from school and public records in St. Paul for five baseline years and multiple postclinic years.  That study found large year-to-year variations in schoolwide birth rates, but no evidence indicating that the clinics significantly reduced birth rates.

Kirby, Waszak, and Ziegler (1991) found that clinic presence was not significantly related to pregnancy rate in any of the six school-based clinic sites, after background characteristics of the students were statistically controlled.  Similarly, Kisker, Brown, and Hill (1994) found that clinic presence was not related to pregnancy rate, and Newcomer and Duggan (1996) found that clinic presence was not related to birth rate.  The Zabin study (1986) found that the pregnancy rate increased slightly and then decreased.

In sum, the data consistently demonstrate that providing contraceptives in school-based or school-linked clinics does not hasten or increase student sexual activity.  Data also indicate that those clinics do provide contraceptives to substantial numbers of sexually experienced students in their schools.  Other results are more mixed.  Although one study did report increased contraceptive use after a clinic was opened, the weight of the evidence suggests that those clinics do not increase schoolwide contraceptive use significantly.  To the contrary, the data indicate that there is a large substitution effect, although not all students who obtained contraceptives from the clinics would have obtained them elsewhere.  Similarly, despite the results of one study, the weight of the evidence suggests that school-based or school-linked clinics also do not reduce schoolwide pregnancy rates or birth rates significantly.

The Effects of School Condom-Availability Programs

Given the threat of AIDS, as well as the threat of other STDs and pregnancy, more than 300 schools without school-based clinics have begun making condoms available through school counselors, nurses, teachers, vending machines, or baskets (Kirby & Brown, 1996).  Those schools are in addition to the 92 schools that make condoms available to students through school-based clinics.

The number of condoms obtained by students from schools varies greatly from program to program; in some schools students obtain very few condoms from the school, while in other schools they obtain large numbers (Kirby & Brown, 1996).  In general, students in smaller alternative schools obtain many more condoms per student than students in larger schools or students in mainstream schools.  In addition, when schools make condoms available in baskets (a barrier-free method), students obtain many more condoms than when they must obtain condoms from school personnel.  Finally, if schools have clinics, students obtain many more condoms than they do if schools do not have clinics.

Thus far, three studies have presented results on how the availability of condoms at school affects students' sexual behavior.  The study with the strongest evaluation design measured the effects of making condoms available through vending machines in five Seattle schools without school-based clinics and through vending machines and baskets in five additional Seattle schools with preexisting school-based clinics.  Schoolwide data were collected both before condoms were made available in the schools and two years later.  In neither group of schools was there an increase in either sexual activity or use of condoms during last intercourse (Kirby et al., forthcoming).  Notably, all 10 schools had a strong educational intervention, but because it existed before the baseline data collection, the effect of the educational component was not measured.

A second study measured the impact of making condoms available in baskets in a single high school in Los Angeles (Schuster et al. 1996).  Schoolwide data were collected both before and after the condoms were made available.  Because a change in school policy required a change from passive parental consent to active parental consent, a much smaller percentage of the students (and potentially a less representative sample of students) completed the posttest survey than the pretest survey.  Comparison of the pretest and posttest surveys indicated that there was not a significant change in the percent of students who had ever had sex nor in the frequency of sexual activity.  Among males who were sexually active during the year, however, there was an increase in the percentage who used a condom every time they had sex.  Similarly, among males who had initiated intercourse recently, there was an increase in the percentage who used a condom the first time they had sex.  There were no significant changes among females.

The third study evaluated a New York program.  The school district implemented a comprehensive AIDS prevention program in the city high schools.  It included additional instruction about AIDS, schoolwide activities, and condom availability.  Because baseline data could not be collected in New York, analyses compared students in New York schools with a matched sample from Chicago schools.  Results indicated that students in the New York schools were not more likely to have initiated intercourse, but were more likely to have used a condom the last time they had sex than were Chicago students (Guttmacher et al., 1997).

In summary, the results from these three studies are similar to those of school-based clinics—they confirm that making condoms available on school campuses does not increase sexual activity, but the impact of the availability of condoms at school on their use is mixed.  It is unclear why results suggest that school condom availability may have increased condom use in New York and Los Angeles but not in Seattle.

The Effects of Multicomponent Programs

One program included multiple components, some of which were implemented in the schools.  It was designed to reduce pregnancy in a small rural South Carolina community (Koo, Dunteman, George, Green & Vincent, 1994; Vincent, Clearie & Schluchter, 1987).  Teachers, administrators, and community leaders were given training in sexuality education.  Sex education was integrated into all grades in the schools.  Peer counselors were trained.  The school nurse counseled students, provided male students with condoms, and took female students to a nearby family planning clinic.  The local media, churches, and other community organizations highlighted special events and reinforced the messages of avoiding unintended pregnancy.  After the program was implemented, the pregnancy rate for 14- to 17-year-olds declined significantly for several years.  After parts of the program ended (for example, the school nurse resigned; linkages to contraceptives were terminated; and some teachers left the school), the pregnancy rates returned to preprogram levels.

Summary and Conclusions

Although this paper has reviewed many studies, at least four factors limit the conclusions that can be drawn from those studies.  First, given the complexity of adolescent sexual behavior and the diversity of programs addressing that behavior, there are simply too few studies evaluating the effectiveness of each of the different approaches.  That research challenge is increased by the fact that programs are implemented in different settings with different target groups, and their success may vary with setting and target population.

Second, many of these studies were limited by methodological problems or constraints.  Far too often, studies did not use experimental designs and thus suffered from self-selection effects; had sample sizes that were too small and thus failed to detect programmatically important outcomes or may have produced anomalous results; or did not use proper analytic techniques.  Thus, the results that are published are undoubtedly biased in unknown ways.

Third, the studies often produced inconsistent results.  Some programs appeared to affect behavior, while other seemingly similar programs did not.  Fourth, there are very few replications of evaluations, and when they have occurred, the subsequent studies failed to replicate the previous positive results.

For those reasons, it is difficult to reach many conclusions that are well supported by evidence.  As a result, funds and efforts are undoubtedly not directed toward the most effective approaches.

Given those important caveats about the evaluations of programs, conclusions about the impact of programs must be expressed cautiously.  Nevertheless, some tentative conclusions exist:

• The overwhelming weight of the evidence demonstrates that programs that focus on sexuality, including sex and AIDS education programs, family planning programs, and school-based clinics and condom-availability programs, do not cause harm, as some people fear.  None of the evaluations of any program found significant results indicating that any of them increased any measure of sexual activity.
• Nearly all sex and AIDS education programs that have been evaluated have produced some outcome deemed socially desirable by our society, for example, an increase in knowledge.  Studies of a few programs have produced credible evidence that their respective programs reduced sexual risk-taking behavior either by delaying sex, reducing the frequency of sex, reducing the number of sexual partners, or increasing the use of either condoms or other forms of contraception.  Typically, but not always, those behavioral effects were only measured for the short term, not the long term.  Studies of other sex and AIDS education programs have failed to find positive effects on behavior.  The nine characteristics discussed above may distinguish effective from ineffective programs.
• Although abstinence-only programs may be appropriate for many youth, especially middle-school youth, there does not currently exist any evidence that they have actually delayed the onset of sexual intercourse or reduced any other measure of sexual intercourse.  At this point in time, their impact upon sexual behavior is simply unknown.
• Substantial percentages of sexually experienced female students in schools with school-based or school-linked clinics obtain contraceptives from those clinics; nonetheless, the weight of the evidence indicates that those clinics typically do not significantly increase the schoolwide proportion of students using contraception, nor do they significantly decrease the schoolwide pregnancy and birth rates.
• The few studies of school condom availability programs provide inconsistent results.  If schools provide condoms in a manner that reduces barriers to obtaining condoms (e.g., through baskets in clinic bathrooms), then students obtain large numbers of condoms from that source.  Studies have produced mixed results, however, regarding the impact of school condom availability on actual condom use.

On the other hand, studies also indicate that some programs can have some success and can modestly reduce one or more sexual behaviors for at least a brief period of time.  Thus, sex and HIV education programs may be an important component in a larger, more comprehensive initiative to reduce adolescent sexual risk-taking behavior.

References
 

Bandura, A. (1986). Social foundations of thought and action. Englewood Cliffs, NJ: Prentice Hall.

Barth, R. P. (1996). Reducing the risk: Building skills to prevent pregnancy, STD & HIV (2nd ed.). Santa Cruz, CA: ETR Associates.

Christopher, F. S., & Roosa, M. W. (1990). An evaluation of an adolescent pregnancy prevention program: Is "just say no" enough? Family Relations, 39(1), 68–72.

Collins, J. L., Leavy Small, M., Kann, L., Collins Pateman, B., Gold, R.S., & Kolbe, L.J. (1995). School health education. Journal of School Health, 65(8), 302–311.

Comprehensive Health Education Foundation. (1994). Get real about AIDS. Seattle: Comprehensive Health Education Foundation.

DeMauro, D. (1990). Sexuality education 1990: A review of state sexuality and AIDS curricula. SIECUS Report, 18(2), 1–9.

Dusenbury, L., &  Falco, M. (1995). Eleven components of effective drug abuse prevention curricula. Journal of School Health, 65(10), 420–425.

Education Development Center, Inc. (1996, April 12). National Training Partnership Directors of Training Meeting, Atlanta, GA.

Edwards, L., Steinman, M., Arnold, K., & Hakanson, E. (1980). Adolescent pregnancy prevention services in high school clinics. Family Planning Perspectives, 12(1), 6–14.

Eisen, M., Zellman, G. L., & McAlister, A. L. (1990). Evaluating the impact of a theory-based sexuality and contraceptive education program. Family Planning Perspectives, 22(6), 262.

Ekstrand M. L., Siegel, D. S., Nido, V., Faigeles, B., Cummings, G. A., Battle, R., Krasnovsky, F., Chiment, E., & Coates, T. J. (1996, July). Peer-led AIDS prevention delays onset of sexual activity and changes peer norms among urban junior high school students. Paper presented at the XI International Conference on AIDS, Vancouver.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior. Reading, MA: Addison-Wesley.

Frost, J. J., & Forrest J. D. (1995). Understanding the impact of effective teenage pregnancy prevention programs. Family Planning Perspectives, 27(5), 188–195.

Gallup, A. (1987). The 19th annual Gallup polls of the public's attitudes toward the public school. Gallup Poll, 69(1).

Gambrell, A. E., & Haffner, D. (1993). Unfinished business: A SIECUS assessment of state sexuality education programs. New York: SIECUS.

Guttmacher, S., Lieberman, L., Ward, D., Freudenberg, N., Radosh, A., & DesJarlais, D. (1997, in press, September 1997). Condom availability in New York City public high schools:  Relationship to condom use and sexual behaviors. American Journal of Public Health.

Homans, G. C. (1965). Group factors in worker productivity. In H. Proshansky and L. Seidenbert (Eds.), Basic studies in social psychology. New York: Holt, Reinhart & Winston.

Howard, M., & McCabe, J. (1990). Helping teenagers postpone sexual involvement. Family Planning Perspectives, 22(1), 21–26. [AU: PLEASE PROVIDE ISSUE NUMBER.]

Jemmott, J. B., III, Jemmott, L. S., & Fong, G. T. (1992). Reductions in HIV risk–associated sexual behaviors among black male adolescents: Effects of an AIDS prevention intervention. American Journal of Public Health, 82(3), 372–377.

Jemmott, L. S., Jemmott, J. B., III, & McCaffree, K. A. (1994). Be proud! Be responsible! New York: Select Media.

Jorgensen, S. R., Potts, V., & Camp, B. (1993). Project Taking Charge: Six-month follow-up of a pregnancy prevention program for early adolescents. Family Relations, 42(4), 401–406.

Kipke, M. D., Boyer, C., & Hein, K. (1993). An evaluation of an AIDS risk reduction education and skills training (ARREST) program. Journal of Adolescent Health, 14(7), 533–539.

Kirby, D. (1985). The effects of selected sexuality education programs: Toward a more realistic view. Journal of Sex Education & Therapy, 11(10), 28–37.

Kirby, D. (1984). Sexuality education: An evaluation of programs and their effects. Santa Cruz, CA: Network Publications.

Kirby, D., Barth, R., Leland, N., & Fetro, J. (1991). Reducing the risk: A new curriculum to prevent sexual risk-taking. Family Planning Perspectives, 23(6), 253–263.

Kirby, D., Brener, N., Brown, N., Peterfreund, N., Hillard, P., & Harrist, R. (forthcoming). School condom-availability in Seattle schools.

Kirby, D., & Brown, N. (1996). School condom availability programs in the United States. Family Planning Perspectives, 28(5), 196–202.

Kirby, D., Korpi, M., Adivi, C., & Weissman, J. (1997). An impact evaluation of Project SNAPP: An AIDS and pregnancy prevention middle school program. AIDS Education and Prevention, 9(Suppl. A), 44-61.

Kirby, D., Korpi, M., Barth, R. P., & Cagampang, H. H. (1995). Evaluation of Education Now and Babies Later (ENABL): Final Report. Berkeley, CA: University of California, School of Social Welfare, Family Welfare Research Group.

Kirby, D., Resnik, M. D., Downes, B., Kocher, T., Gunderson, P., & Blum, R. W. (1993). The effects of school-based health clinics in St. Paul upon school-wide birth rates. Family Planning Perspectives, 25(1), 12-16.

Kirby, D., Short, L., Collins, J., Rugg, D., Kolbe, L., Howard, M., Miller, B., Sonenstein, F., & Zabin, L. (1994). School-based programs to reduce sexual risk behaviors: A review of effectiveness. Public Health Reports, 109(3), 339–360.

Kirby, D., Waszak, C., & Ziegler, J. (1991). Six school-based clinics: Their reproductive health services and impact on sexual behavior. Family Planning Perspectives, 23(6), 6–16.

Kisker, E. E., Brown, R. S., & Hill, J. (1994). Health caring: Outcomes of the Robert Wood Johnson Foundation's school-based adolescent health care program. Princeton: Robert Wood Johnson Foundation.

Koo, H. P., Dunteman, G. H., George, C., Green, Y., & Vincent, M. (1994). Reducing adolescent pregnancy through a school- and community-based intervention:  Denmark, SC revisited. Family Planning Perspectives, 26(5), 206–211, 217.

Levy, S. R., Perhats, C., Weeks, K., Handler, A., Zhu, C., & Flay, B. R. (1995). Impact of a school-based AIDS prevention program on risk and protective behavior for newly sexually active students. Journal of School Health, 65(4), 145–151.

Louis Harris and Associates. (1988, May). Public attitudes toward teenage pregnancy, sex education and birth control. [Poll conducted for the Planned Parenthood Federation of America, New York].

Main, D. S., Iverson, D. C., McGloin, J., Banspach, S. W., Collins, K., Rugg, D., & Kolbe, L. J. (1994). Preventing HIV infection among adolescents: Evaluation of a school-based education program. Preventive Medicine, 23, 409–417.

McGuire, W. (1972). Social psychology. In P. C. Dodwell (Ed.), New horizons in psychology. Middlesex, England: Penguin Books.

Miller, B. C., & Paikoff, R. L. (1992). Comparing adolescent pregnancy prevention programs: Methods and results. In B. C. Miller, J. J. Card, R. L. Paikoff, & J. L. Peterson (Eds.), Preventing adolescent pregnancy (pp. 265-284). Newbury Park, CA: Sage Publications. [AU: PLEASE PROVIDE PAGE NUMBERS.]

Moberg, D. P., & Piper, D. L. (1990). An outcome evaluation of Project Model Health: A middle school health promotion program. Health Education Quarterly, 17(1), 37–51.

Moore, K. A., Sugland, B. W., Blumenthal, C., Glei, D., & Snyder, N. (1995). Adolescent pregnancy prevention programs: Interventions and evaluations. Washington, DC: Child Trends, Inc.

NARAL Foundation. (1995). Sexuality education in American: A state–by–state review. Washington, DC: NARAL Foundation.

NARAL Foundation. (1997). Who decides? A state-by-state review of abortion and reproductive rights. Washington, DC: NARAL Foundation.

National Center for Education Statistics. (1993). Digest of education statistics, 1993. Washington, DC: U.S. Department of Education, Office of Educational Research and Improvement.

Newcomer, S., & Duggan, A. (1996, May 9). Do school-based clinics influence adolescent birth rates? Paper presented at the annual meetings of the Population Association of America, New Orleans.

Nicholson, H. J., & Postrado, L. T. (1991). Girls Incorporated preventing adolescent pregnancy: A program development and research project. New York: Girls Incorporated.

Roosa, M, & Christopher, S. (1990). Evaluation of an abstinence-only adolescent pregnancy prevention program: A replication. Family Relations, 39(4), 363–367.

Roper Organization, Inc. (1991). AIDS: Public attitudes and education needs. New York: Gay Men's Health Crisis.

Schinke, S. P., Blythe, B. J., Gilchrist, L. D., & Burt, G. A. (1981). Primary prevention of adolescent pregnancy. Social Work with Groups, 4(1–2), 121–135.

Schuster, M. A., Bell, R. M., Berry, S. H., & Kanouse, D. E. (1996). Evaluation of a high school condom availability program. I. Impact on sexual behavior, use of condoms, and attitudes. Santa Monica: RAND Corporation.

Smith, M. A. B. (1994). Teen incentives program: Evaluation of a health promotion model for adolescent pregnancy prevention. Journal of Health Education, 25(1), 24–29.

St. Lawrence, J. S. (1994). Becoming a Responsible Teen: An HIV risk reduction intervention for African-American adolescents. Jackson, MI: Jackson State University.

St. Lawrence, J. S., Jefferson, K. W., Alleyne, E., & Brasfield, T. L. (1995). Comparison of education versus behavioral skills training interventions in lowering sexual HIV risk behavior of substance dependent adolescents. Journal of Consulting and Clinical Psychology, 63(2), 221–237.

St. Pierre, T. L, Mark, M. M, Kaltreider, D. L., & Aikin, K. J. (1995). A 27-month evaluation of a sexual activity prevention program in Boys & Girls Clubs across the nation. Family Relations, 44(1), 69–77.

Thomas, B., Mitchell, A., Devlin, M., Goldsmith, C., Singer, J., & Watters, D. (1992). Small group sex education at school: The McMaster teen program. In B. Miller, J. Card, R. Paikoff, & J. Peterson (Eds.), Preventing adolescent pregnancy (pp. 28-52).
Newbury Park, CA: Sage Publications.

Vincent, M., Clearie, A., & Schluchter, M. (1987). Reducing adolescent pregnancy through school and community-based education. Journal of the American Medical Association, 257(24), 3382–3386.

Walker, G., & Vilella-Velez, F. (1992). Anatomy of a demonstration. Philadelphia, PA: Public/Private Ventures.

Walter, H. J., & Vaughn, R. D. (1993). AIDS risk reduction among a multi-ethnic sample of urban high school students. Journal of the American Medical Association, 270(6), 725–730.

Warren, W. K., & King, A. J. C. (1994). Development and evaluation of an AIDS/STD/sexuality program for grade 9 students. Kingston, Ontario: Social Program Evaluation Group.

Young, M., Core-Gebhart, P., & Marx, D. (1992). Abstinence-oriented sexuality education: Initial field tests of the Living Smart curriculum. Family Life Educator, 10(4), 4–8.
Zabin, L. S., Hirsh, M. B., Smith, E. A., Streett, R., & Hardy, J. B. (1986). Evaluation of a pregnancy prevention program for urban teenagers. Family Planning Perspectives, 18(3), 119–126.