Study Details

Citation

Weed, S. E., Birch, P. J., Ericksen, I. H., Olsen, J. A. (2011). Testing a predictive model of youth sexual intercourse initiation. Unpublished manuscript.

Program or Component Study?
Program
Program or Component Name

Heritage Keepers Abstinence Education I

Show Evidence of Effectiveness
Yes
Study Rating and Explanation
Moderate —

Quasi-experimental study that met all criteria for a moderate rating; findings show a positive, statistically significant impact for at least one behavioral outcome

Program Information

Program Type
Sexual risk avoidance
Program Length
Fewer than 10 sessions

Evaluation Setting

Evaluation Setting
In school: Middle school

Study Sample

Average Age Group
14 to 17
Majority Racial/Ethnic Group
African American or Black
Gender
Youth of any gender

Research Design

Assignment Method
Quasi-Experimental Design
Sample Size

2215

Number of Follow-Ups

2

Length of Last Follow-Up

12

Year of Last Data Collection
2004

Study Findings

Result Sexual Activity
Potentially favorable evidence
Reviewed Studies
Moderate-Quality Quasi-Experiment
Protocol Version
Version 2.0
Details

This study used a quasi-experimental design involving South Carolina middle school and high school students. The study compared 1,828 students from 34 schools that implemented the Heritage Keepers Abstinence Education I program with 387 students from seven schools that did not implement the program. Surveys were administered immediately before and after the program and again a year after the program had ended.

The study found that a year after the program had ended, students in the schools that implemented the program were less likely to report having ever had sex than students in the schools that did not offer the program. The study also examined program impacts on measures of recent sexual activity and number of sexual partners. Findings for these outcomes were not considered for the review because they did not meet the review evidence standards. Specifically, findings for these outcomes did not statistically adjust for behaviors measured at baseline.