Skip navigation
Skip Navigation
Search
Click for menu... About NAEP... Click for menu... Subject Areas... Help Site Map Contact Us Glossary NewsFlash
Sample Questions Analyze Data State Profiles Publications Search the Site
NAEP Technical Documentation
The Nation's Report Card (home page)

Table of Contents  |  Search Technical Documentation  |  References

The NAEP Database → Creation of the Database → Merging Files

Merging Files

Prior to merging, a reorganization of the files is performed and includes the following tasks:

  1. Files are separated by subject area to improve maintenance and efficiency.

  2. The files are restructured, eliminating unused (blank) areas to reduce the size of the files.

  3. In cases where students chose not to respond to an item, the missing responses are recoded as either an "omitted response" or a "not reached item."

  4. A school file is created by merging the school characteristics and policies questionnaire file and can be associated with a student record in order to report school information for students.

Following the reorganization of data files, the following merging steps take place:

  1. Final student reporting weights data are merged with the student-weight files.

  2. The resulting file is then merged with the students with disabilities/limited-English-proficient (SD/LEP) student questionnaire and teacher questionnaire data.

The matching criteria used in these steps are:

  • In all steps, the 10-digit booklet identification number is used as the matching criterion. The first three numbers correspond to the 3-digit booklet number common to every booklet with the same blocks of items (see examples of common booklet numbers). The next six digits correspond to the 6-digit serial number unique to the booklet a student is given, and the last number is a single-digit check.

  • The teacher data can be linked to the student data through four data variables: primary sampling unit (PSU), school code, teacher ID, and classroom period.

  • The PSU and school codes are used as the matching criterion for the national school data; the Federal Information Processing Standards (FIPS) code is used for state school data. Since some schools did not return a questionnaire, some of the records in the school file contained only school-identifying information. The school data can be linked to the student data through the PSU or FIPS codes.

  • Whenever new data values (such as composite background variables or plausible values) are derived, they are added to the appropriate database files using the same matching procedures described above.

Student names are not part of the files that can be matched. School and student data are matched using unique identifiers and combined only to produce results focused on students (e.g., What percentage of students attends school in which mathematics is identified as a special priority?).

Last updated 19 June 2008 (MH)

Printer-friendly Version