NAEP Technical DocumentationStratification of Public Schools in the 2012 Long-Term Trend (LTT) Assessment

Prior to stratification, the public school sampling frame was divided into age-specific files, one each for ages 9, 13, and 17. For each age-specific frame file, separate implicit stratification schemes were used to sort schools into certainty primary sampling units (PSUs) and noncertainty PSUs. In all cases, the implicit stratification was achieved via a "serpentine sort."

For certainty PSUs, the schools were hierarchically sorted by

census region;
urbanization classification (urban-centric locale);
race/ethnicity strata; and
estimated age enrollment.

If there were fewer than two expected sampled schools for a particular urbanization classification cell (nested within the Census region), the cell was collapsed with a neighboring urbanization classification cell. If the expected sampled schools exceeded four, then the race/ethnicity strata were defined based on the percentage of Blacks, Hispanics, and American Indians/Alaska Natives. The strata were defined so that there were at least two expected sampled schools for each racial/ethnic stratum. If the urbanization classification stratum had an expected sample size less than four, no race/ethnicity strata were generated, and the final sort variable was percent Blacks, Hispanics, and American Indians/Alaska Natives rather than estimated age enrollment.

Schools in noncertainty PSUs were hierarchically sorted by

PSU stratum;
urbanization classification (urban-centric locale); and
percent race/ethnicity (Blacks, Hispanics, American Indians/Alaska Natives).

The collapsing of cells within the noncertainty PSUs was implemented in a fashion similar to that described for certainty PSUs.

Last updated 09 August 2012 (DB)

Printer-friendly Version