• Introduction
• Summary of Major Findings and Recommendations
• References

This report is the second in a series that will provide a comprehensive assessment of data quality in the Common Core of Data (CCD) as it relates to coverage, classification, reliability, validity, survey design, and estimation.¹The purpose of this evaluation is to analyze and document classification issues. The results can be used for ongoing process improvement of the CCD statistical program. This evaluation also explores such issues as school district geography, enrollment, governance, history, and service delivery and relates each to agency and school classification.

During the reference frame of this evaluation, the CCD statistical program consisted of four separate surveys. These were the Public Elementary/Secondary Education Agency Universe Survey (hereafter referred to as the agency survey), the Public Elementary/Secondary School Universe Survey (hereafter referred to as the school survey), the State Non-Fiscal Survey (hereafter referred to as the state aggregate survey), and the National Public Education Fiscal Survey (NPEFS).

This evaluation was undertaken primarily to examine the agency portion of the survey, but it does integrate the school portion for certain purposes. Agencies are authorized under state law to perform certain services and to operate certain types of schools. A comparison of school types reported by various agencies can be found in the complete report.

A detailed analysis of CCD definitions contained in both the glossary and field description portions of the instructions for completing the survey is also included. The purpose of this analysis is to describe potential definitional inconsistencies, omissions, and redundancies. The analysis is not intended to serve as a framework for a new set of definitions, but it describes possible validity and reliability problems as they relate to survey definitions.

The evaluation process was initiated by researching state statutes and administrative codes to determine the powers, governance, service area, and services provided by agencies existing in each state. This research was based on existing research done by the Bureau of the Census in connection with the 1992 Census of Governments and updated based on state legislation passed after the reference date of that census. These findings are included in the complete report.

• The CCD survey is used as a basis for many other surveys within the Department of Education. It would be an ideal platform on which to unify public elementary/secondary education definitions for all related surveys.
• The full potential of the CCD survey has yet to be realized. An improved CCD survey could take on several important roles.
• CCD definitions are generally inadequate. Many are neither exhaustive nor exclusive and present obvious problems for respondents. The inadequacy of and misinterpretation of survey definitions lead to unreliable and invalid data.
• Schools and agencies are canvassed in separate surveys, but CCD definitions do not provide a clear distinction between the two components.
• To reduce reporting errors, refinements in survey methodology should be explored. Possibilities include assigning a central reference person to interpret survey definitions, establishing state-to-federal definitional crosswalks, directly canvassing local education agencies, and restructuring the survey to meet state reporting capabilities.
• CCD definitions should be revised so that they are both flexible and exhaustive. If individual arrangements in states differ from conventional approaches, and cannot be made to fit into the definitional framework, they must be clearly documented and explained.
• In order to prevent duplication of effort and unnecessary respondent burden, all education surveys that use the CCD as their sampling frame should be coordinated. Definitions should be unified where possible, and information should be shared among surveys where possible.
• Enrollments from the school, agency, and state aggregate surveys are not comparable. Enrollment figures represent different things in different states on different surveys. The basis for enrollment must be adequately defined.
• Certain geographic coding schemes in the CCD survey may produce misleading results. With the completion of the school district mapping project,²the utility of these codes may be in question.
• The advent of a new period of education reform makes this an opportune time to revisit the structure and content of the CCD survey.