Research

The WWC reviewed two studies of the effectiveness of financial incentives for teen parents. The Ohio LEAP program study met WWC evidence standards, and the California Cal-Learn program study met evidence standards with reservations.

The evaluation of the Ohio LEAP program (Long, Gueron, Wood, Fisher, & Fellerath, 1996) that met WWC evidence standards was a randomized controlled trial. A total of 7,017 teens in seven Ohio counties were randomly assigned to the intervention and control groups. Some sample members were then excluded from the analysis, because implementation problems in the startup year resulted in an inconsistent treatment. The study followed the same rules in excluding intervention and control members, however, thus maintaining the integrity of random assignment. Of the 2,967 teens who remained after exclusions, the study collected outcomes using a survey fielded to a random sample of 1,178 teens three years after random assignment. The analysis was conducted using data for the 913 respondents to the survey.

The evaluation of the Cal-Learn program (Mauldon, Malvin, Stiles, Nicosia, & Seto, 2000) that met WWC evidence standards with reservations was a randomized controlled trial with attrition problems. A total of 4,859 teens in four California counties were randomly assigned to research groups. Of those teens, 2,682 responded to the first survey about 13 months after they entered the program. After the survey, the study excluded sample members who lost custody of their children, moved to a nonresearch county or out of state, left welfare, or did not receive welfare for at least six months, resulting in a sample of 2,156. The study administered a second survey about 26 months after program entry, with 1,562 respondents. In addition to the low response rates, the WWC had reservations about the study because sample members were excluded from the second survey based on conditions that could have been affected by the financial incentives, such as high school completion within six months of random assignment. As a result, the remaining teen parents in intervention and comparison groups may no longer have been equivalent.