more effective contraceptive practice. It is difficult for research to resolve this debate definitively because of the wide variation in the content and depth of the sex education offered, differences in the research definitions of "sex education," the reluctance of schools and agencies to allow evaluation of the effects of such courses, and deficiencies in some study designs (Furstenberg et al., 1985; Kirby, 1984; Scales, 1981; Spanier, 1976).
There are, however, several bodies of information that shed light on this topic. National survey data present mixed results, but indicate that adolescents who receive sex education are more likely to use contraception than those who do not receive such instruction (Ku et al., 1993, 1992; Dawson, 1986; Marsiglio and Mott, 1986; Furstenberg et al., 1985; Zelnik and Kim, 1982). Retrospective surveys, however, cannot provide causal associations; such associations can only be made through evaluations with experimental or quasi-experimental designs. Unfortunately, few evaluations are so methodologically rigorous, and most fail to measure behavior change and long-term program effects (Chapter 8).
In an attempt to address the lack of rigorous assessment, Kirby (1984) used quasi-experimental designs to evaluate 15 well-regarded sex education curricula from the 1970s and early 1980s. He concluded that the programs did increase knowledge about various topics in reproductive health, but did not change sexual behavior or contraceptive use.
This discouraging picture appears to be changing. In a 1994 review, Kirby and colleagues suggest that both programs and evaluation methods have improved. Evidence from more than 20 surveys and studies of school-based sex and HIV and AIDS education programs indicates that specific programs delayed the initiation of intercourse, reduced the frequency of intercourse, reduced the number of sexual partners, or increased the use of contraceptives. In addition, available data indicate clearly that participation in these sexuality education programs has not been found to encourage adolescents to initiate sexual intercourse, or to increase the frequency of intercourse among adolescents who were sexually active before the program.
The sex education programs reviewed by Kirby and colleagues (1994) clustered into three types: (1) abstinence-only programs that do not discuss contraception, (2) sexuality or AIDS education programs that discuss both abstinence and contraception, and (3) programs that provide comprehensive reproductive health education covering many topics including contraception and abstinence, as well as clinical services. Abstinence-only programs appear to affect attitudes regarding premarital intercourse, but the few evaluations that measure behavior change are limited by methodological problems, and there is insufficient evidence to determine whether abstinence programs delay the age of first intercourse or affect other sexual and contraceptive behaviors. Effects of programs of the second type are mixed, but those that most successfully delay sexual intercourse or increase contraceptive use appear to focus on the "particular facts, values, norms, and skills necessary to avoid sex or unprotected