Skip to main content
Skip Navigation

Table of Contents  |  Search Technical Documentation  |  References

NAEP Technical DocumentationDevelopment of Initial School Nonresponse Cells for the 2004 Assessment

The cells for nonresponse adjustments are generally functions of the school sampling strata. (Sampling strata definitions differ depending on the school type, public or private.)     

Separate non-response adjustments were made for operational assessment, mathematics bridge assessment, and reading bridge assessment, each age group (9, 13, and 17), and upper and lower grades corresponding to each age group. This way, 18 public and 18 private subsamples were created. The rationale for creating upper and lower grades was that each age group corresponds to more than one grade in a school. Some schools would only allow access to students in an age group from one grade, usually the modal grade, with the largest number of age eligible students. Upper and lower grade categories were created to adjust for under-representation of students from the age category of interest that was not available for assessment. A school was placed in the upper grades category as a respondent if it allowed testing in the upper grades, and in the lower grades category as a nonrespondent if it didn't allow testing in the lower grades. Similarly, a school was placed in the lower grades category as a respondent if it allowed testing in the lower grades, and in the upper grades category as a nonrespondent if it didn't allow testing in the upper grades.   

Student grade enrollment distribution displays the proportion of students age 9, 13, or 17 distributed across relevant grades, by region. It shows the rationale for creating the upper and lower grades. The grade levels were defined as follows:

  • Age 9: lower grades - 1 through 3; upper grades - 4 through 12;
  • Age 13: lower grades - 1 through 7; upper grades - 8 through 12; and
  • Age 17: lower grades - 1 through 10; upper grades 11 through 12.

Within each of the 18 public school subsamples, nonresponse cells were developed using the following nesting structure:

  • Certainty or non-certainty primary sampling unit (PSU) status,
  • school location stratum, and
  • race/ethnicity status stratum.

Within each of the 18 private school subsamples, nonresponse cells were developed using the following nesting structure:

  • Certainty or non-certainty PSU status,
  • NAEP Region,
  • private school type, and
  • race/ethnicity status stratum.

Last updated 17 April 2009 (JL)

Printer-friendly Version