Introduction

Adolescent pregnancy has long been a concern and its prevention a point of debate in the United States. Researchers, practitioners, and policymakers, regardless of ideological differences, welcomed the decline in teen pregnancy rates, which from 1991 to 2005 dropped 34% (Centers for Disease Control 2009). However, even during this time of decreasing pregnancy rates, the U.S. still had the highest rate among developed countries (Singh and Darroch 2000). Now, continued or renewed attention to this issue is needed as after the 14 year decline there was a 5% rise from 2005 to 2007 (Hamilton et al. 2009).

Recent reviews and discussions of policy and research on prevention of adolescent pregnancy have raised important questions about relative effectiveness of different approaches for working directly with youth to postpone sexual involvement and reduce the numbers of adolescents having children. Over the past decade, there has been increased emphasis at federal and state levels on encouraging abstinence as the primary mechanism for avoiding out of wedlock pregnancy (U.S. Department of Health & Human Services 2002). Some in the research community note that abstinence approaches may be appropriate for many youth (Rector 2002), especially for younger adolescents. The experts caution, however, that rigorous studies of abstinence curricula have not yet shown any overall effect on sexual behavior. In a longitudinal examination of four abstinence education programs, Trenholm and colleagues (2007) found participation in these programs did not increase the number of youth who abstained from sex compared to control group participants. In a review of 9 abstinence education programs, only 2 were found to delay initiation of sex (Kirby 2008). Kirby (2008) also stresses the evidence in favor of abstinence programs provided by these two evaluations is weak due to quasi-experimental design and problematic methods.

There is accumulating evidence in favor of comprehensive approaches that include broader coverage of responsible sexual behavior in the prevention of adolescent pregnancy and sexually transmitted diseases (Kirby 2008). Fifteen of 48 comprehensive sex and HIV/STD education programs delayed initiation of sex among participants (Kirby 2008). Programs that include job-related or community service components have also shown promising results (Philliber et al. 2002). In her review, Scher (2008) concludes that the most promising results have occurred for intensive multi-component, youth development programs for at-risk youth, although additional rigorous evaluations are needed to substantiate these results. Another assessment of prevention programs effectiveness indicates that effective programming must include both contraceptive knowledge building and distribution (Franklin and Corcoran 2000). Similar to evaluations of abstinence education, cautions also apply to more comprehensive approaches. While there are a few studies indicating that comprehensive or multi-component programs decrease adolescent pregnancy rates (The National Campaign to Prevent Teen & Unplanned Pregnancy 2009), research is still limited in providing compelling evidence of the reliability of comprehensive programs’ effects on sexual behavior and likelihood of adolescent pregnancy.

Research examining the decline in teen pregnancy rates found that for youth ages 15–17 years old, 77% of the decline could be attributed to increases in improved contraceptive use and 23% to decreased sexual activity (Santelli et al. 2007). While the majority of the decline was attributable to contraceptive use, it is still important that almost a quarter of the decline was due to abstinence. This supports the argument that there may be a role for encouraging abstinence as one of many ways to prevent teen pregnancy. Indeed, in a report on “what works,” one of five categories of effective programming is education that promotes both abstinence and contraceptive use (The National Campaign to Prevent Teen & Unplanned Pregnancy 2009). Additionally, Kirby (2008) indicates that we do not yet have the final answer for the abstinence-comprehensive sex education debate. The current study contributes to the field by comparing the outcomes for two types of prevention programs. Evaluation data were collected from two community-based service demonstration projects to compare immediate effects on youths’ life plans of abstinence-based versus a more comprehensive service learning approach.

Program Description

The stated purpose of each of the two grant-funded programs was to prevent adolescent pregnancy. Both programs were framed on the theory of reasoned action/planned behavior (Albarracin et al. 2001; Ajzen 2002; Richard et al. 2003), and a primary measure of outcomes was youths’ forecasts of their life plans before and after participating in the program. The small-group curriculum-guided programs were provided through a community-based, non-profit social services agency in central Texas, and were targeted to youth identified by their parents, teachers, or counselors as being at high risk for adolescent pregnancy.

The main differences between the programs were: the central tenet of the respective curricula, the age ranges targeted, breadth of coverage of responsible sexual behavior, and length of the program. The abstinence-based program was funded through the federal Office of Adolescent Pregnancy Programs (OAPP). Targeted to youth ages 9–14, it emphasized the benefits of postponing sexual involvement and skills for recognizing and resisting peer and media pressures to have sex. It was guided by a locally developed curriculum with experiential activities drawn from Managing Pressures Before Marriage developed at Emory University in Atlanta, Georgia (Howard 1993). The series of 9–16 meetings in a support group cycle for the abstinence-based program usually was completed within one school semester. The service learning program, funded through the Texas Department of Health, was targeted to youth ages 13–17 and emphasized school success and career exploration through volunteer work and summer employment as means of protecting youth from teen pregnancy and other negative behaviors. This program, which used the curriculum developed in St. Louis, Missouri for the Teen Outreach Program (Allen et al. 1997), included material on contraception and family planning. The 11–24 meetings in a support group cycle for the service learning program usually extended across 2 school semesters.

Direction and structure of the two programs were very similar. Both were conducted at schools and summer camps in Austin/Travis County, Texas, and facilitated by professionally trained social workers and health educators. Each support group accommodated approximately 5–12 youth of similar ages (e.g., 6th and 7th graders or 7th and 8th graders but not 6th and 8th graders) and same gender. Groups for boys had male facilitators and those for girls had female facilitators. Since both programs were prevention focused, the participating youth were not parents or pregnant at time of program. Both programs were curriculum guided and experientially-based with substantial overlap in topics covered—values clarification, affirmation of positive personal qualities, decision making, communication skills, assertiveness skills, and relationships.

The similarities in purpose, structure, and theoretical foundation enabled comparison of immediate outcomes recorded for participating youth in the overlap of ages targeted—i.e., ages 13 and 14. For this secondary analysis of the program evaluation data, the sample of youth were matched on gender and risk behaviors at the time they entered their respective programs. The research question was whether these programs would have an effect on the participant’s life plans. If so, it was predicted that observed changes in youths’ life plans from before to after their group experience would be closely related to the emphasis of the respective curriculum approaches. The hypothesis was that youth who participated in the abstinence-based education groups would become more likely to forecast futures in which they postponed sexual involvement until after age 20 and after marriage, whereas those participating in the service learning program would become more likely to forecast futures in which they postponed childbearing but not necessarily sexual involvement until after age 20 and after marriage.

Method

Study Design

Life plans of young adolescents were compared before and after they participated in abstinence-based (ab) or service learning (sl) curriculum-guided small groups for which the primary goal was preventing adolescent pregnancy. Life plans were measured by asking youth to forecast the age at which they thought they would pass certain developmental milestones (e.g., graduate from high school). This set of questions was based on items in the Prevention Minimum Evaluation Data Set (Sociometrics 1999). All youth who participated in the programs were invited to participate in the evaluation. Active consent was obtained from youths’ parents together with assent from the youth.

Participants

A matched sample was drawn from the evaluation data sets supplied by a total of 518 ab youth and 117 sl youth who completed both a pre- and a post-test for their respective program. Matching ab and sl youth on age, school semester in which the youth began participating in his or her respective program, gender, and risk behaviors reported by the youth on the pre-test produced groups constituted of 44 ab and 44 sl youth. Average age for the ab sample was 13.83 (SD = 0.38) and for sl 13.99 (SD = 0.47).

Characteristics of the study participants are displayed in Table 1. Each group included 27 females and 17 males. More than two-thirds (77%) were age 14 while 23% were age 13 when they entered their respective prevention education support group. The majority (64%) reported living with two parents, but nearly one-fourth (23%) reported living with their single mother, and 9% reported living with grandparents or aunts or other arrangements. Nearly all (89%) were youth of color with half (56%) describing themselves as Mexican or other Hispanic heritage, 19% describing themselves as Black or African American, and 14% describing themselves as Mexican/Black or Hawaiian or Black/Indian or Asian. Significantly fewer of the ab youth described their heritage as mixed or other than Black or Hispanic or White than did sl youth (χ2 = 6.18, df = 1, p < .05). There were no other significant differences in demographic characteristics or risk behaviors of the ab and sl youth.

Table 1 Demographic characteristics of study participants
Full size table

Indicators documented on the pre-tests of risk for adolescent pregnancy among these young people were being “in trouble at school” (41%), using alcohol (28%) or tobacco (24%) during the 4 weeks before completing the pretest, giving an answer to a pretest question that indicated they believe they “probably” or “definitely” will have sexual intercourse as an unmarried teenager (27%), and indicating on their pretest life plans that first sexual intercourse would be at an age younger than their current age, indicating perhaps that they already had engaged in sexual intercourse (13%).

Data Collection

Youth completed the pretest during the first meeting with their group and the posttest during the last or next to last meeting. They were asked to “write the age” they thought they would be when they graduate from high school, get their first full-time job, get married, move into their own house or apartment, have sexual intercourse for the first time, and have their first baby. For any behavior they believed they would never do, they were instructed to write an “X” or the words “not me.”

Data Analysis

Life plans forecast by the youth before and after participating in their respective prevention education programs were summarized by tallying the percent of program participants reporting they would pass a particular developmental milestone (e.g., get first full-time job) before age 20. Average age for passing each milestone was calculated. Counts were made case by case of number of youth forecasting they would have sexual intercourse and/or have first baby at an age earlier than the age at which they forecast they would marry, get their first full-time job, or move into their own house or apartment. To enable quantitative analyses, the youths’ forecasts were recoded such that before age 20 received a score of 1 whereas equal to or after age 20 was scored 0. Similarly, before marriage was scored 1 and equal to or after marriage was scored 0. To compare results of ab and sl programs, scores were submitted to SPSS and a two groups by two time points repeated measures analysis using the general linear model command was conducted. Post-hoc analyses were conducted to evaluate possible bias contributed by the small but significant difference in proportion of youth describing their heritage as mixed or other than Black, Hispanic, or White.

Results

Virtually all of the participating youth indicated they will graduate from high school (98%), get a full-time job (97%), and move into their own house or apartment (94%). The vast majority also indicated they will marry (85%), have sexual intercourse (82%), and have a baby (83%). A few youth indicated they “never” will have sexual intercourse (10%), marry (9%), or have a baby (5%). For those with plans to pass the milestones, the average ages forecast in order of the anticipated debut were have first sexual intercourse at 17.37 ± 4.89, graduate from high school at 17.76 ± 0.73, get first full-time job at 17.99 ± 2.96, move into own house or apartment at 19.53 ± 2.75, have first baby at 24.34 ± 4.87, and marry at 25.19 ± 8.24. There were no significant differences by self-described ethnic heritage. The majority forecast a future in which they will graduate from high school, get their first full-time job, have sexual intercourse, and move to their own house or apartment before they are 20. These data are displayed in Table 2.

Table 2 Changes in percent of youth whose life plans indicated they would pass certain developmental milestones before age 20 and/or before marriage
Full size table

Nearly all of the youth planned to postpone marriage and children until they are older than 20. However, even after participating in the prevention program, 10 of the 88 youth (12%) forecast futures in which they will have their first baby before they are 20. Three of these 10 had life plans indicating they will have their first baby before they are 18.

Before and after both the ab and the sl program, the majority of participants forecast futures in which they will have sexual intercourse at least a year before they marry. There were no statistically significant differences between the programs or from before to after the program in youths’ forecasts regarding the age at which they will make their sexual debut. It is noted that although not a significant change, the percentage of participants who reported that they would have sexual intercourse before marriage and before age 20 unexpectedly increased from pre-test to post-test in the ab group.

The majority of the participating youth indicated they will get their first full-time job before they marry (91–95%). Virtually all of the youth (94–95%) indicated they will get their first full-time job before they have their first baby. There were no statistically significant differences between the programs or from before to after the program in youths’ forecasts regarding the age at which they will graduate from high school, get their first full-time job, or move into their own house or apartment.

At pre-test a substantial number of youth indicated they expected to have their first baby before marriage. At post-test a significant main effect for change occurred in the desired direction for participants regarding timing of childbirth. Analysis of the 0–1 scores showed the change was statistically significant (F = 5.04, p < .05, partial η2 = .07). Post-hoc analyses looking at these groups independently revealed that a significant change occurred for participants in the ab program (F = 6.98, p < .01, partial η2 = 0.16), whereas the change in the sl group was not significant. Specifically, the mean scores in the ab program were 0.32 before and 0.16 after compared with 0.39 before and 0.32 after in the sl program, indicating that at post-test more youth indicated their plans to have a baby after getting married There was not a significant interaction effect of program by time for childbirth planning.

Discussion

Overall, these programs did not appear to effect youths’ life plans. Results of this study demonstrated a modest positive change from before to after for the abstinence-based education support group program in youths’ plans to postpone having their first baby until after marriage. Yet there were no significant changes in plans for sexual debut in either group. In both the abstinence-based program and the service learning program, the majority of youth indicated plans or expectations to have sexual intercourse for the first time before they are 20 and before they marry. The participants’ anticipated average age of sexual debut (M = 17.37, SD = 4.89) was higher than actual mean age of sexual debut (M = 15.2, SD not available) reported in a study by Shafii et al. (2007) who obtained their estimate from data in the National Longitudinal Study of Adolescent Health from 1994 to 2002. Thus, there is evidence to suggest that actual age of sexual debut may be sooner than anticipated for these youth.

Lack of reported change in youths’ plans regarding sexual debut is a serious concern because of the implications of the theory of reasoned action/planned behavior (Ajzen 2002) on which both programs are based. According to the theory and the research on which it is founded, a key determinant of human behavior is intention to perform or not perform target behaviors. If, as is indicated in this study, youth who are identified as being at risk for adolescent pregnancy in fact intend or plan to become sexually involved before they are 20 and before marriage, it is highly likely they will do so. In so doing, they put themselves at risk for sexually transmitted diseases and pregnancy, which could lead to constrained opportunities for success in school and at work.

Interestingly, examination of the youths’ forecasts regarding when they expect to marry and to have their first baby showed them to be remarkably similar to actual ages of first marriage and first birth for women in the United States. On average, the youth in this study expected to have their first baby at age 24.3 and to marry at 25.2. In 2000, when this study began, national survey data showed average age at first birth for women in the US was 24.9 (Mathews and Hamilton 2002) and average age at first marriage was 25.1 (Johnson and Dye 2005).

Limitations in this study included a lack of opportunity to follow-up with the participating youth to obtain information about the actual age at which they make their sexual debut, get married, graduate high school, and experience their first pregnancy. Current reviews of studies with longitudinal data indicate mixed outcomes for pregnancy prevention programs (Kirby 2008; Trenholm et al. 2007). However, length of time for follow up varies greatly across studies, so it remains difficult to compare effectiveness. This current study was not a longitudinal study and thus there was no opportunity to find out if participants’ intentions were reliable predictors of their future behaviors. Additionally, youth attendance records were not available for this analysis so it is not known the actual number of sessions attended by these participants. Regardless, youth included in the study attended at least one of the first two sessions and then participated over the course of the curriculum such that they were available to take the post-test at the last session. In addition, the number of sessions differed between the two programs. Despite these overall limitations, we believe this study is important because it provides results from two different prevention education approaches as actually practiced in the community.

Implications

Results of this study offer both research and practice implications for social workers interested in addressing adolescent pregnancy. This study provides an example of asking youth to forecast their personal life plans—i.e., to report the ages at which they plan to graduate from high school, get their first full-time job, move into their own house or apartment, marry, have sexual intercourse for the first time, and have their first child—as a method for evaluating short term outcomes of prevention education programs and having the youth participate in this future orientation exercise. However, longitudinal research is further needed in order to evaluate whether forecasts are a valid proxy for behavioral outcomes.

Practice implications include the need for programs to (1) distinguish sexual activity and childbearing, (2) tailor information to fit participants’ age and activity level, and (3) focus specifically on life plans for childbearing/becoming parents. The current study, like other studies that have probed adolescents’ perspectives on childbearing, revealed that some youth intend to become teen parents (Adler and Tschann 1993; Zabin et al. 1993). Despite a statistically significant reduction in the abstinence-based program in plans for having first baby before marriage, the post-tests showed approximately 3% of prevention program participants forecasting futures in which they will have their first baby at ages 15, 16 or 17. Outcomes likely could be improved if, as is recommended elsewhere (Kalmuss et al. 2003), prevention programs acknowledge and address sexual activity and childbearing as distinct constructs. Exercises and discussions specifically designed to help young people explore their beliefs about childbearing, particularly in regards to the likely role of the baby’s other parent and the baby’s grandparents in the lives of the young person and his or her baby should be included.

Results of the current study provide support for tailoring prevention programs to the level of sexual experience that is likely to be encountered within different age groups (Franklin and Corcoran 2000; Johnson et al. 2003). Several youth (13%) in the current study indicated on their pre-test life plans that first sexual intercourse would be at an age younger than their current age. Thus, promotion of abstinence is not likely to be an effective prevention approach for these youth. Given that youth participating in this study were identified by their parents, counselors or teachers as being at high risk for adolescent pregnancy, this finding is not surprising. These results also suggest that perspectives on childbearing should be one of the emphases of prevention programs and their evaluations, especially when youth are already sexually active.

A final recommendation that reaches beyond the results of this study is for social work researchers, administrators, and practitioners to leverage and strengthen the potential power of existing research-based prevention programs. For example, better results may be obtained if youth-focused programs are embedded in larger and more sustained efforts to involve the whole village in raising the children (Manlove et al. 2003). Almost 90% of a nationally representative sample of youth ages 12–19 surveyed in the US in 2003 said it would be much easier to postpone sexual activity and avoid teen pregnancy if they were able to have more open, honest conversations with their parents about sex, contraception, and pregnancy (The National Campaign to Prevent Teen Pregnancy 2003). Other research demonstrates that teenagers who perceive that their mothers disapprove of their having sex are less likely to become sexually active (Jaccard and Dittus 2000). Finding innovative and effective ways to involve parents is, therefore, an especially attractive option for enhancing the effects of prevention education programs for youth (Lederman et al. 2004). Building a web-based “second line of defense” to help young people obtain needed information (Wilson 2000), positioning interactive computer kiosks and health education videos in the waiting rooms at health clinics (Paperny and Hedberg 1999), supporting the work of physicians as adolescent sexuality educators (Clark et al. 2003), and continued effort to reduce poverty and disadvantage are other recommended strategies for involving the broader community in helping young people to make healthy choices.

Conclusion

By comparing the outcomes for two types of prevention programs, this study contributes to the knowledge base for adolescent pregnancy prevention efforts. Findings from both programs suggest that sexual activity and childbearing need to be addressed as distinct constructs in prevention programs. While most youth plan to have sex before marriage, there was a significant decrease in plans to have a baby before getting married for youth in the abstinence-based program. The study also demonstrated that asking youth to forecast their own futures can be a useful way to obtain feedback about their intentions and plans regarding sexual debut, marriage, and childbearing. It is striking that their age forecasts for child-bearing and marriage were close to actual statistics for these phenomena. In addition, results support the need for social workers to tailor prevention education efforts to the level of sexual experience of targeted youth.