Educational outcomes in teenage pregnancy and parenting programs: results from a demonstration
Warrick, L.; Christianson, J.B.; Walruff, J.; Cook, P.C.
Family Planning Perspectives 25(4): 148-155
1993
ISSN/ISBN: 0014-7354 PMID: 8405340 DOI: 10.2307/2135922Document Number: 327716
Hispanic adolescents, and particularly Mexican Americans adolescents, have a very high rate of pregnancy and dropping out of school. The aim of this article was to provide the results of an evaluation of a school- based demonstration project (Teenage Pregnant and Parenting Demonstration: TAPP). TAPP was implemented at 3 sites in Phoenix or Tucson, Arizona, during 1985-89 and compared with 4 other sites. TAPP enrolled students during pregnancy and aimed to keep pregnant and parenting students in school. Each program provided comprehensive services. 5 program models were represented in the 7 participating program sites. Model 1 included the 2 sites for the TAPP program, which mainstreamed pregnant students with regular students and recruited pregnant school dropouts over the summer months. The unique feature was case management for each student, which was coordinated with student's classroom teachers. Models 2-5 are described. 789 students were enrolled in the evaluation. Of the 39% of nonparticipants, 44% had dropped out of school in Models 1 and 2, 42% gave birth before they could participate; in Models 4 and 5, 27% had dropped out, 24% gave birth before enrollment, and 34% refused IN Model 3, there were not program dropouts. Ethnic distribution was 40% Hispanic, 34% whites, 18% blacks, and 8% Native Americans. Survival analysis was used to examine dropouts and logistic regression was used to analyze program model and educational persistence. The results showed that Model 1 was the most successful at keeping study participants in high school at 12 and at 24 months. Model 1 was also the most effective at school retention for those in grades 9, 10, 11 and when controlling for the effects of academic grades. Model 1 was also the most successful at retaining students who had given birth. The regression analysis indicated that living with one's boyfriend or receiving public assistance reduced the likelihood of education persistence. Those enrolled in the study in their last trimester were less likely to continue in school. The strongest and most consistent predictor of education persistence was previous success in school. A stepwise regression, which was performed separately by ethnic group, showed that Model 1 had a stronger impact for Hispanic students; the log likelihood test did not isolate the distinctive differences. The evaluation reinforced that comprehensive community-linked and school-based programs are particularly successful for Hispanics and younger students, those living with their partner, and those enrolled early in pregnancy. Critical program features were strong campus and community-wide outreach and case management.