Study 1
Schinke and colleagues (2000) evaluated the effects of Project Learn on youth in public housing in New York City, N.Y., Cleveland, Ohio, Oakland, Calif., Tampa, Fla., and Edinburgh, Texas. This program was implemented by Boys & Girls Clubs of America (BGC) in communities they already served. For the purpose of the evaluation, a site in each city that had BGC, but not the Project Learn enhancement, was matched on demographics, youth proportion, size, and public housing service population of the Treatment sites to serve as the Comparison site. Additionally, a Control site was matched to each Treatment site using a youth facility that was not BGC and did not provide educational enhancements to their adolescent students. In total, 15 sites were used (five Treatment, five Comparison, and five Control), matched on geographic and demographic variables. Youth and parents at each site completed informed consent statements. The total number of participants at baseline was 283. At the 18-month follow-up, the number of participants had decreased to 249, and to 191 by the final 30-month follow-up. An attrition analysis showed participant drop-off did not differ significantly between intervention groups and sites.
Of the 283 youth across the 15 sites at baseline, the average age was 12.3 years. Forty percent were female. The participant group was 63 percent Black, 19 percent Hispanic, 13 percent white, and 5 percent Asian or of another ethnicity. There were no significant differences in age, gender, or ethnicity between groups. Although treatment and comparison youth came from BGCs, control youth were regular users of their facilities, which included after-school, recreational, and human services programs. Treatment sites were encouraged to participate in the evaluation when chosen; comparison sites were motivated by the receipt of the program after the evaluation; and control sites received a financial incentive at each measurement.
Data from schools were collected to assess students’ performance in mathematics, English grammar, composition, reading, spelling, history, science, social studies, and geography. Schools also provided attendance records and behavioral incidents. Researchers used analysis of variance to test for differences between the treatment, comparison, and control groups. The CrimeSolutions review looked at the differences between the treatment sites that implemented the Project Learn enhancements to the BGCs and control sites that had no BGCs, at the 30-month follow up. Subgroup analyses were not conducted.