The Knowledge Is Power Program: Preparing Youth for College Success

A college education opens the door to better pay, steadier employment, more career opportunities, and greater social mobility. Yet many youth from underserved communities find the door to college admission and success hard to open.

To address this disparity, the Knowledge Is Power Program (KIPP) established a national network of free, open-enrollment college preparatory public schools in underserved communities throughout the United States. The KIPP approach to education includes training outstanding leaders, giving students more time in the classroom, setting high expectations, providing a rigorous college-preparatory curriculum, and developing a strong culture of support. Founded in 1994, KIPP has grown to a national network of 57 schools in 17 states and Washington, D.C. 

Mathematica's five-year, $3.8 million evaluation of KIPP focuses on the following questions:

• What are the impacts of KIPP middle schools on student achievement and other outcomes, and how do these outcomes compare with those of students at other schools?
• Does the performance of KIPP students—both within and beyond core academic areas—suggest that they are on a path toward college attainment and persistence?
  

Mathematica is using both an experimental design and a quasi-experimental design to evaluate student outcomes over a broad range of KIPP middle schools. These methodologies will provide sound data on the impact of KIPP on student achievement and college readiness.

The experimental component consists of a randomized control trial conducted in KIPP schools that are “oversubscribed”—with more applicants than spaces available—and that use lotteries to determine which students are offered admission. The lotteries randomly assign sample members into a treatment group (comprised of those students with access to a KIPP education) or a control group (comprised of students without such access). Student outcomes over the follow-up period will be measured for both groups and may include the use of school records, principal surveys, student and parent surveys, and results from a test of higher-order thinking skills.

For the nonexperimental component, Mathematica will collect multiple years of data from school records on KIPP middle school students and students at nearby traditional public schools. The information will be used to rigorously estimate the KIPP effect by comparing outcomes for KIPP students with outcomes for a comparison group of students identified as most similar to KIPP students based on pre-middle-school trends in test scores and other characteristics.


Back to Top