Overview of Methods and Sample

Overview of Methods

Reinvigorating the Pipeline: Insights into Proposed and Approved Charter Schools


NACSA collected and analyzed charter school applications—both approved and not approved—from 19 states and the District of Columbia[1] (hereafter referred to as 20 states) over a five-year period (Fall 2013 to Spring 2018) in order to describe trends and types of applications being proposed, approved, withdrawn, and denied.

Data Acquisition

Research teams from NACSA and Public Impact[2] collected charter school applications from two primary sources: direct submissions from charter school authorizers and downloads from authorizer or state department of education websites. All types of applications were collected (e.g., initial applications, appeals, replications). Authorizers also provided or confirmed the status of each application (e.g., approved, not-approved, pending). The project has received 2,943 applications to date.[3]

Application Coding

A team of trained researchers and analysts coded each application across more than 50 domains comprising over 180 variables. Variables include the application’s proposed school models/features, information about the applicant, and many other application characteristics.[4] Project leadership agreed to definitions for each variable.[5] Industry-standard methods and targets were used to establish a high rate of coder agreement throughout the coding process. 


Authorizers participating in the study oversaw 81 percent of charter schools in the 20 states included in this study. To maximize the resources available for the study and the comprehensiveness of the findings, those 20 states were chosen because of (a) their relative charter sector size, (b) authorizer willingness to provide data to researchers, and (c) the availability of data. The research team attempted to acquire applications from every current authorizer in those 20 states (i.e., those that oversaw at least one charter school across the five-year study period).[6] Researchers followed up with non-responding authorizers, prioritizing authorizers with many charter schools in their portfolio. In all states except one,[7] the research team received applications from authorizers overseeing two-thirds or more of charter schools in that state.[8]


[1] The 20 jurisdictions are Arkansas, Arizona, California, Colorado, Connecticut, Washington, D.C., Florida, Georgia, Illinois, Indiana, Louisiana, Massachusetts, Missouri, Minnesota, New Jersey, New York, North Carolina, Ohio, Tennessee, and Texas.
[2]Public Impact’s mission is to improve education dramatically for all students, especially low-income students, students of color, and other students whose needs historically have not been well met. They are a team of professionals from many backgrounds, including former teachers. They are researchers, thought leaders, tool-builders, and on-the-ground consultants who work with leading education reformers. For more on Public Impact, please visit www.publicimpact.com. Public Impact provided critical thought leadership to the project and led the application coding process. The authors are extremely grateful for their competence, diligence, partnership, and thoughtfulness in this project.
[3] This count includes applications appealed to an appellate body with the authority to authorize directly. Specifically, in California, county and state agencies are empowered to directly authorize appealed applications. The count also includes multi-campus applications (i.e., a single application for five schools was counted as five applications). For most analyses, applications to an appellate authorizer are removed, but multi-campus applications are included.
[4] The full list of variables is available upon request.
[5] Please see the Glossary of Terms for variable descriptions.
[6] The research team did not request applications from all “potential authorizers,” entities that state law empowers to be authorizers but have not yet approved a charter school. In a handful of instances, the research team did request applications from “potential authorizers” known to have received an application during the study period.
[7] Participating Ohio authorizers only oversaw approximately 35 percent of existing charter schools in the state. Consequently, we have less confidence the state-level data in Ohio is an accurate depiction of application activity and it should be interpreted with caution.
[8] The research team used the percent of charter schools overseen by participating authorizers in each state as a proxy for where applications are likely submitted. This information is available upon request.


Additional Report Content: Reinvigorating the Pipeline