Table of Contents  |  Search Technical Documentation  |  References

NAEP Sample Design → NAEP 2003 Sample Design → 2003 State Assessment Sample Design → School Sample Selection → Sample Selection

NAEP Technical DocumentationSample Selection

			Public School Sampling Frame and New-School Sample Sizes Comparisons of the State School Samples Against the CCD Frame Evaluation of State Achievement Data in the Sampling Frame

The measure of size assigned to each school s in jurisdiction j was

where

• E_js is the expected number of hits for school s in jurisdiction j,
  

• b_j is the proportionality constant for jurisdiction j for the
  measured grade (fourth or eighth),
  

• MOS_js is the initial measure of size for school s in jurisdiction j, and
  

• u_j is the upper bound constraint for the expected number of hits.

The schools were ordered within each jurisdiction using the serpentine sort with the following stratification variables: charter school status, Trial Urban District Assessment (TUDA)/urbanicity status, minority status, and achievement score or ZIP Code area median income. A systematic sample was then drawn using this serpentine sort and the measures of size.

In addition, new and newly eligible schools were sampled from the new-school sampling frame. The assigned measures of size for these schools is similar to the measure of size for originally selected schools:

The formula used the b_j and u_j values from the main school sample for the jurisdiction (i.e., the same sampling rates as for Common Core of Data (CCD) schools within each jurisdiction). The new schools were ordered by district (as the serpentine sort was not used for the small numbers of these schools), and a systematic sample drawn.

In addition to the samples described earlier, sample schools for the TUDA and fourth-grade charter school studies were a part of the 2003 State NAEP sample design. The TUDA schools are located in 10 districts:

• Los Angeles, California;
• San Diego, California;
• Atlanta, Georgia;
• Chicago, Illinois;
• Boston, Massachusetts;
• Albuquerque, New Mexico;
• New York City, New York;
• Charlotte, North Carolina;
• Cleveland, Ohio; and
• Houston, Texas.

The supplemental charter school samples are located in California, Michigan, and Texas.

The goal of deeply stratifying the school sample in each jurisdiction is to reflect the population distribution as closely as possible, with as little sampling error as possible. The success of this approach can be seen by comparing the proportion of minorities enrolled in schools (CCD values for each school), median income, and type of location (viewed as an interval variable) reported in the original frame against the school sample. In addition, state assessment achievement scores for the original frame can be compared with the scores for the school sample for those jurisdictions for which state assessment achievement data are available.

Printer-friendly Version