School-to-Work and Academy Demonstration Programs
The School-to-Work and Academy Demonstration programs, funded under the Edna McConnell Clark Foundation's "Programs for Disadvantaged Youth," attempt to improve school retention and transitions to work for high school students in seven cities.
This evaluation of the School-to-Work and Academy Demonstration programs entails both process and outcome components. The process component relies upon a case study description of each program, including a summary of its setting, design, management structure, and target group. Information for this segment of the evaluation derives from site visits, staff questionnaires, interviews, observation, and discussions with program managers.
The outcome component employs a comparison group design in which both program students and a matched group of nonprogram students were tracked from the year prior to a program's start through each year of its operation. Data were collected on student retention, attendance, credits earned, courses failed, grade point averages, and standardized test scores. In addition, questionnaires were administered to program students before and after receiving program services.
Seven sites were included in the evaluation, which covers the 1986–87 school year: Chicago, Cleveland, Denver, Oakland, Pittsburgh, Portland, and Washington, D.C. Programs varied from site to site in both structure and age. Cleveland and Oakland, for example, were operating in their second year; Chicago, Denver, Pittsburgh, and Portland in their third; and Washington, D.C., in its sixth.
Four sites—Chicago, Pittsburgh, Portland, and Washington—offered multi-year, combined academic and job preparation programs; Denver offered a one-year academic and job preparation program; Oakland offered two or three weeks of job search training; and Cleveland offered curriculum development training to teachers, with the intent of implementing revised curriculum during 1987-88.
Evaluation of student outcomes revealed varying effects from site to site and variable to variable. Most differences between program and comparison groups were not statistically significant