NAEP Technical DocumentationDevelopment of Initial School Nonresponse Cells

The cells for nonresponse adjustments are generally functions of the school sampling strata. (Sampling strata definitions differ for public schools and private schools.) The public school samples formed initial weighting cells within jurisdiction using the following nesting cell structure:

Trial Urban District Assessment (TUDA) district vs. the balance of the state for states with TUDA districts;
charter school status (grade 4 only);
urbanicity classification; and
minority classification, achievement level or median income, or grade enrollment.

The cell definitions for the fourth-grade samples attempted to incorporate charter school status to accommodate the national charter school sample integrated into the public school sample. However, only Arizona, the District of Columbia, Michigan, California (excluding the Los Angeles and San Diego school districts), and Texas (excluding the Houston school district) had enough charter schools to warrant such a breakdown. Charter schools within each remaining jurisdiction and each TUDA district were grouped into the same cells as noncharter schools.

In general, the nonresponse cell structure used minority classification as the lowest level variable. However, where there was only one minority category within a particular level of urbanicity, categorized achievement or median income data were used instead. For Virgin Islands in both fourth and eighth grades, and American Samoa and Puerto Rico in eighth grade, schools were grouped according to student enrollment size due to the lack of data on urbanicity, achievement, income, and race/ethnicity.

Private school samples formed nonresponse adjustment cells hierarchically as follows:

private school type (e.g., Catholic, Lutheran, Conservative Christian, Other Private);
census division stratum;
school location stratum; and
minority status stratum.

Last updated 06 November 2008 (RF)

Printer-friendly Version