Appendix A: Methodology and Technical Notes

Common Core of Data survey system. The CCD is a program of the U.S. Department of Education’s National Center for Education Statistics (NCES) that annually collects fiscal and nonfiscal data about all public schools, public local education agencies, and state education agencies in the United States. The system consists of both nonfiscal and fiscal survey components. The State Nonfiscal Survey of Public Elementary/Secondary Education, the Local Education Agency (LEA) Universe Survey, and the Public Elementary/Secondary School Universe Survey are the nonfiscal components. These three surveys, which are the subject of this report, include information that describes schools and LEAs, including name, address, and phone number; as well as descriptive information about students, teachers, and staff. The School District Finance Survey and the National Public Education Financial Survey are the fiscal components. The fiscal surveys include school finance information such as revenues and current expenditures.

Data collection for nonfiscal CCD. SEAs report data for the three nonfiscal CCD universe surveys through the U.S. Department of Education’s EDFacts collection system. EDFacts is an initiative to centralize and coordinate administrative data reporting to the Department of Education for elementary and secondary public education. SEAs deliver data to the EDFacts collection system through a series of file groups that fall into different reporting schedules throughout the year. The school year (SY) 2011–12 EDFacts collection of CCD data opened in January 2012, with groups of data collected over a requested schedule throughout the year. Depending on the specific variable or state, the data in this report represent data reported from January 2012 through April 2013. The SY 2011–12 EDFacts collection will remain open to SEAs for revisions and corrections for approximately 3 years after the opening of the initial collection, continuing through the end of 2015. For more information about the EDFacts data collection, please visit the following website: http://www2.ed.gov/about/inits/ed/edfacts/index.html.

Participation in SY 2011–12. SEAs from the 50 states, the District of Columbia, Puerto Rico, and the U.S. Virgin Islands reported in EDFacts for SY 2011–12; the Commonwealth of the Northern Mariana Islands and Guam reported directly to the CCD. The Department of Defense Education Activity (DoDEA), the Bureau of Indian Education (BIE), and American Samoa did not report data for SY 2011–12.

Data quality. NCES performed an extensive quality review of all CCD data submissions. Data analysts asked state EDFacts and CCD coordinators to correct or confirm any numbers that appear out of range when compared to other states' data or with the state's reports in previous years. If no correction or explanation for anomalous data was provided by the SEA, NCES either edited the value (for example, replace a reported total value with the sum of detail) or suppressed the data. For example, if a state reported enrollment of 12^th grade students that was substantially larger than the enrollment of 11^th grade students in the previous year, and the state could not explain the discrepancy, NCES would suppress the reported number of 12^th grade students.

As part of the quality review, NCES also checked for consistency among the school, LEA, and state-level files. In some states where inconsistencies were identified and SEAs were not able to provide a correction or explanation, NCES also applied additional data edits across all three data files to reduce data anomalies. For example, in SY 2011–12, some states reported more full-time equivalent (FTE) counts at the school level than the LEA level, and the states confirmed that the LEA figures were accurate. NCES edited the FTE data at the school level to match the data at the LEA level by proportionally adjusting the counts.

Totals. "United States" and "reporting states" totals in the tables include only the 50 states and the District of Columbia. These totals exclude data from the BIE, DoDEA, Puerto Rico, or the other jurisdictions of American Samoa, Guam, the Commonwealth of the Northern Mariana Islands, or the U.S. Virgin Islands. The term "reporting states" is used when not all states have been able to report a data item.

Missing data. Not all states or jurisdictions collect and report all of the data items requested in CCD surveys. NCES attempts to complete missing data first by drawing on other sources. For example, if an SEA does not report the number of teachers in the State Nonfiscal Survey of Public Elementary/Secondary Education, but reports the number of teachers for each LEA in the LEA Universe Survey, NCES would complete the state-level number by summing the number of teachers from the state’s LEAs. In some cases, if a state is unable to report data during the collection period but publishes them later through a written report or website, NCES may import data from these other published sources to complete missing items. When this is done, table footnotes identify all instances in which NCES summed or imported data from other CCD surveys or outside sources. NCES uses editing procedures to complete missing data for any CCD data item in any of the three CCD nonfiscal surveys, as necessary. NCES may carry some directory information, such as address or telephone number, forward from a prior year’s report if it is missing in the current year. In general, NCES does not carry forward prior year statistical information, such as the number of students in membership, to replace missing data items. An exception was made in 2011–12 for the DoDEA, the BIE and American Samoa, which did not report data for the 2011–12 school year; the numbers and statuses of the schools and agencies in these jurisdictions were carried over from prior year files and all the other data items were left as missing. NCES also carries over the state directory information from prior year files with all the other data items left as missing.

When reporting results, NCES treats missing data within individual states differently than it treats missing data across all states and the District of Columbia. If information is missing for some items, but at least 85 percent of items have been reported for totals across the 50 states and District of Columbia, NCES calculates totals and identifies them as "reporting states" totals (rather than totals for the United States). In order for NCES to report a state total of an item, the SEA must have reported at least 80 percent of possible cases within that state.

State nonfiscal imputed data. NCES imputes missing values in the State Nonfiscal Survey of Public Elementary/Secondary Education to create data files that more accurately reflect student and staff counts and to improve comparability among states. Imputations are performed on data from the 50 states and District of Columbia only. Imputations provide plausible values in cases where the data are missing. An imputation assigns a value to the missing item, which increases subtotals containing this item by the amount of the imputation. An example of one of the imputation methods used is the following: If a state is unable to report the number of prekindergarten teachers employed by the state for a particular school year, the number of prekindergarten teachers for the state would be imputed by calculating a median growth rate for all states that reported prekindergarten teachers in the previous and current year, and applying this median growth rate to the state’s prior year reported number of prekindergarten teachers. For more information about the imputation procedures used to assign values to missing items, please refer to the documentation for the State Nonfiscal Survey of Public Elementary/Secondary Education, which can be accessed at http://nces.ed.gov/ccd/stnfis.asp.

NCES does not currently impute missing items in the Public Elementary/Secondary School Universe Survey or the LEA Universe Survey. Data users should use caution in making inferences when comparing totals tabulated from the school and LEA universe files compared to prior year files, as the files may have different occurrences of missing school or LEA data from year to year.

Edits. NCES uses statistical editing procedures to identify potential errors and apply corrections for reported values. An example of one of the editing procedures is the following: if a state reports a total count of teachers that exceeds or is less than the sum of associated detail counts, but none of the associated components have been imputed, NCES edits the reported total to equal the sum of the associated detail counts. NCES notifies the SEA of the potential errors and asks the SEA to verify the information or provide corrected information. However, when the data cannot be corrected or validated, a trained CCD analyst uses information provided by the SEA in the current and prior years, together with information from outside sources such as SEA or LEA published information to determine if and when corrections or suppressions should be made to the data.

Beginning in SY 2009–10, NCES implemented a revised editing methodology for agency and school data to improve the precision of our edits in identifying errors. The revised methodology used data from multiple years and was applied to both the reported data items and related ratios. The methodology placed an additional requirement for an item to be identified as a potential error that both the count and a related ratio must be identified as potential errors. For example, for total student membership to be flagged as a potential error, both total membership and a related ratio, such as pupil/teacher ratio, must be flagged. This methodology was applied to school-level membership data used in table 4 of this report. For more information about the procedures used to edit reported data, please refer to the documentation associated with each of the CCD Surveys at http://nces.ed.gov/ccd/data_resources.asp.

Blank Responses. The EDFacts collection system accepted blank responses in SY 2011–12 and did not require that states distinguish among missing, not applicable, and "zero" values. NCES used statistical editing procedures to change blank responses to missing, not applicable, or zero using information available from SEAs or from prior year reporting. However, it is possible that some blank responses may have been categorized incorrectly. For example, blank staff counts may have been categorized as missing when the actual count was zero.

School and agency operational and membership status. In order to ensure continuity over time, the CCD includes schools and agencies that may not be operating during the school year reported. "Inactive" schools and agencies are those that are closed temporarily with the intention that they will be reopened; these schools and agencies retain their original NCES identification code. When an LEA or school ceases to operate permanently, the CCD includes these as "Closed" schools or agencies for one school year after closing. "Future" schools are those that are scheduled to open, but have not yet begun to operate.

Some operational schools or agencies may legitimately not report students. A vocational school or a LEA operating only vocational schools may provide classes for students from a number of regular schools or school districts. In this case, the students are usually reported in the membership of their school of record, and the vocational school (identified as a "shared time" school) shows no student membership. It is also possible that an operational school that is not "shared time" can be reported with no membership. For example, the number of students may be missing, or the school could have not yet enrolled students when it was reported.

School type edits. Beginning with SY 2007–08, the CCD edits the reported school type if it does not agree with the CCD definition. For example, if a school name indicated that it focused primarily on the needs of students with disabilities or if the state reported in EDFacts that at least 80 percent of the students enrolled in the school had special education individual education programs, CCD analysts reclassified the school as a special education school. Similarly, if a school name suggested that it was a vocational or alternative school, NCES researched the school through different resources (e.g., school website or education association) or by calling the school to determine if the school should be reclassified as a vocational or alternative school. NCES asked CCD Coordinators to review these reclassifications, and if they provided evidence that a school met CCD criteria for the originally reported type, the school type was not changed.

Reportable programs. NCES makes every effort to ensure that the CCD and EDFacts files agree in the numbers of schools they contain. Because the EDFacts system collects data for a number of programs in the U.S. Department of Education, it includes some entities that do not meet the CCD’s definition of a school (e.g., a self-contained magnet or alternative program within a school is counted as a separate school in EDFacts, but not in the CCD). These entities, referred to as "reportable programs," were excluded from this report.

BIE Schools. Beginning with the 2010–11 school year, BIE reported each school on both the school file and the agency file. In prior years, the schools were listed only on the school file.

Teacher counts. SEAs report all teacher counts in FTE units. This is the amount of time required to perform an assignment stated as a proportion of a full-time position. It is computed by dividing the amount of time an individual is employed by the time normally required for a full-time position.

State agencies vary in their staff and teacher data collection and reporting systems, which may cause inconsistency among states in how they report teacher and staff data. Several states collapse two or more categories of staff (including teachers) into one. In these cases, for the state nonfiscal data only, NCES imputes this number by distributing the reported number of staff across the several categories based on the average distribution of these staff in the reporting states. NCES does not impute teacher or staff counts in the LEA and school surveys.

Agreement across survey levels. For LEA membership counts, SEAs report students in the LEA that initially receives funds for their education and has responsibility for their education. For school membership counts, SEAs report students in the school they attend. Each student may only be reported for a single school and for a single agency. Some students receive a public education outside a regular school district (for example, they may attend a state-operated residential school). Some students in a regular school district may not be served by a school. Hospitalized and homebound students, for example, may be reported in the membership for a regular school district but not for any of the district’s schools. NCES derives the numbers of students and staff shown in the tables. NCES considers the numbers reported in the State Nonfiscal Survey of Public Elementary/Secondary Education to be the official statistics for a state.

For more detailed explanations on CCD methodology and technical information, see the data file documentation for the Common Core of Data that can be accessed at http://nces.ed.gov/ccd/ccddata.asp.

Top