Making Evidence Locally

February 14, 2017

CEPR Faculty Director Thomas J. Kane rethinks education research under the Every Student Succeeds Act in the following EdNext article....

 

I’ve seen the potential for districts to gather local evidence on the efficacy of their programs in this way through the Proving Ground project at the Center for Education Policy Research (CEPR) at Harvard University, where I am director. In 2015, CEPR began working with 13 school agencies to develop a model to easily conduct low-cost, local pilots. By pooling data across a network of agencies, we help smaller districts and charter organizations meet the 200-classroom target sample size, with 100 classrooms each in the treatment and control groups. That enables them to evaluate interventions aimed at increasing student achievement by 2 percentile points, or .08 standard deviations. We also provide a network for sharing lessons and facilitating discussions about what districts might try next, based on the evidence they have collected.

From an initial assessment of their data, district leaders develop hypotheses about interventions they wish to try. CEPR helps them decide how many classrooms to include in a pilot. If the district is willing to use random assignment, we provide them with the software tools to select treatment and control groups. If not, we use algorithms to identify comparison students, employing a standard approach to matching on prior test scores and achievement.

 

...

Continue reading at educationnext.org