*Names in bold indicate Presenter
For the first time, through its Fast Response Survey System (FRSS), the U.S. Department of Education has collected nationally representative data on the prevalence of a wide range of district-level dropout prevention practices, including tutoring, self-paced coursework, career and technical education, transition supports, mentoring, programs to reduce behavioral problems, early warning data systems, community partnerships, informational strategies, and district-level data analysis and strategic planning. However, no nationally representative evidence is currently available regarding the extent to which these district-level practices are actually related to graduation rates.
To address this gap in understanding, we will link FRSS data on school district dropout prevention practices with high school graduation rates estimated using grade-level enrollment data from the U.S. Department of Education’s Common Core of Data and aggregated to the school district level. This will allow us to present findings on the relationship between district dropout prevention practices and high school graduation rates both overall and after controlling for key district- and state-level variables related to high school graduation rates, using Hierarchical Linear Modeling.
We will also use HLM to generate district graduation rate predictions using key variables (e.g., district poverty, state compulsory schooling age), and then compare the dropout prevention practices of school districts that substantially outperform predicted graduation rates with those that substantially underperform predictions.
In sum, this study will identify promising district-level practices for dropout prevention, thereby informing critical decisions at the local and state level regarding how best to allocate scarce resources in pursuit of improved graduation rates.