NAEP Technical DocumentationNew-School Sampling Frame for the 2005 State Assessment

The Common Core of Data (CCD) file used for the frame corresponds to the 2002–03 school year, whereas the assessment year is the 2004–05 school year. During this two-year period, some schools closed, some changed structure (one school becoming two schools, for example), and others came into existence.

To achieve as close to full coverage as possible, the school frame was supplemented by a sample of new schools obtained from a sample of districts which was sent a list of the CCD schools and asked to add in any new schools or old schools that had become newly eligible for fourth or eighth grades.

Since asking every school district to list new and newly eligible schools would have generated too much of a burden, a sample of districts was contacted to obtain a list of new schools. To represent the unsampled districts in the full sample of schools, weights for schools included in the new-school sample were adjusted to reflect the district selection process probability.

The goal was to allow every new school a chance of selection, thereby fully covering the target population of schools in operation during the 2004–05 school year. The first step in this process was the development of a new-school frame through the construction of a district-level file from the CCD school-level file. To develop the frame, the district-level file was divided into two files: one for “small” districts and a second for “medium and large” districts.

Small districts contained no more than three schools on the aggregate frame including no more than one school at each fourth, eighth, and twelfth grade. New schools in small districts were identified during school recruitment and added to the sample if the old school was sampled. From a sampling perspective, the new school was viewed as an “annex” to the sampled school that had a well-defined probability of selection equal to that of the old school. The “frame” in this case was, in fact, the original frame; when the old school was sampled in a small district, the new school was automatically sampled as well.

The remaining districts were defined as “medium and large” districts. In these districts, a frame of new schools was developed based on information provided by the district. To limit the required effort, the new-school frame was created through developing information on a sample of "medium and large" public school districts in each jurisdiction. All districts were selected in the following classes of districts:

jurisdictions where all schools were sampled with certainty at either grade 4 or grade 8 (so that all new schools would be selected with certainty as well),
very large districts with more than 6,000 fourth-grade students and 105 fourth-grade schools,
state-operated districts,
districts in states with fewer than 10 districts, and
TUDA districts.

The remaining districts in each jurisdiction (excepting the certainty jurisdictions) were separated into two strata of large and medium districts. These strata were defined by computing an aggregate percentage of enrollment for each district within the state (removing districts in the certainty strata defined above) and sorting in descending order by percentage of jurisdiction enrollment represented by the district. All districts up to and including the first district at or above the 80th cumulative percentage were defined as large districts. The remaining districts were defined as medium districts.

An example is given below. A state's districts are ordered by percentage enrollment. The first six become large districts and the last six become medium districts.

Exemplar distribution of large and medium districts, state assessment, by percent enrollment, stratum, and district: 2005
District	Percent enrollment	Cumulative percent enrollment	Stratum
District 1	20	20	L
District 2	20	40	L
District 3	15	55	L
District 4	10	65	L
District 5	10	75	L
District 6	10	85	L
District 7	5	90	M
District 8	2	92	M
District 9	2	94	M
District 10	2	96	M
District 11	2	98	M
District 12	2	100	M
SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2005.

The target sample sizes for each jurisdiction were 10 districts total: 8 large and 2 medium. More than 10 districts were sampled in the jurisdictions that were oversampled. In the example above, all six large districts and four of the medium districts were selected for the new-school inquiry.

If sampling was needed in the medium stratum (i.e., it was not a certainty jurisdiction), the medium districts were selected with equal probability. If sampling was needed in the large stratum, the large districts were sampled with probability proportional to enrollment. These probabilities were retained and used in all later sampling and weighting, as the district probability then represents the number of other districts which were not sampled to be surveyed for new schools.

The selected districts in each jurisdiction were then sent a listing of all their schools that appeared on the 2002–03 CCD file and were asked to provide information about the new schools not included in the file and grade span changes. These listings provided by the selected districts were used as sampling frames for selection of new public schools. This process was conducted through the NAEP State Coordinator in each jurisdiction. The coordinators were sent the information for all sampled districts in their respective states and were responsible for returning the completed updates.

The eligibility of a school was determined based on the grade span. A school also was classified as “new” if a change of grade span had occurred such that the school status changed from ineligible to eligible in a particular grade.

Additional detail can be found in the report Supplemental Tables from NAEP 2005 Sample Design (Westat 2005).

Last updated 16 March 2009 (RF)

Printer-friendly Version