Career Academies: Impacts on Students' Engagement and Performance in High School

More than 1,700 applicants across 9 high schools

The same nine-site US urban randomised trial as Kemple & Willner (2008), with outcomes measured through scheduled high-school graduation.

Same random assignment; recorded separately because it carries the test-score null, which is the half of the trial that gets left out of advocacy summaries. The dropout finding is on a pre-specified high-risk third of the sample, so it is a subgroup result and weaker than the intention-to-treat estimate. Read at secondary depth via the eight-year report and the review literature, not from the 2000 report itself.

Among students at high risk of dropping out, dropout before senior year fell from 32% in the control group to 21% in the academy group — an 11-point reduction — with better attendance, more academic course-taking, and more credits earned toward on-time graduation. Career Academies did NOT improve standardized mathematics or reading scores: academy students performed no better and no worse than controls. This is the archive's contrast inside a single randomised trial — engagement and persistence move, tested achievement does not.

None material. Individual random assignment.

The in-high-school half of the same trial reported in Kemple & Willner (2008), so replication is not applicable. The test-score null is consistent with every later CTE design except Connecticut's +0.18 SD, and with the Baltimore RD's negative GPA result.