NAEP Technical DocumentationWithin-School Student Sampling Weight for the 2000 National Main Assessment

The within-school student sampling weight component reflects the sampling of students within a school. The within-school sampling weight STUSA_WT is calculated as follows:

STUSA_WT = SAMP_INT x SPL x SUBADJ

where

SAMP_INT is the adjustment to account for the sampling of students within the school, including the initial student sample, the minority (Black and Hispanic) oversample, and/or the students with disabilities and limited-English-proficient (SD/LEP) students oversample;
SPL is the adjustment to account for the allocation of the sampled students to a subject; and
SUBADJ is the adjustment to account for the difference in estimated grade enrollment between a participating substitute school and its corresponding original school.

The within-school sampling interval SAMP_INT depends on the overall sampling rate within the school, as well as whether minority or students with disabilities or limited-English-proficient (SD/LEP) students are oversampled. Minority oversampling occurs in public schools with low minority enrollment. Oversampling of students with disabilities or limited-English-proficient (SD/LEP) students occurs in public schools regardless of the level of minority enrollment (high or low) and private schools.

In schools in which minority oversampling occurs, minority students are sampled at twice the rate of non-minority students. In schools in which SD/LEP oversampling occurs, SD/LEP students are sampled at twice the rate of non-SD/LEP students. In schools in which both types of oversampling occur, non-SD/LEP minority students are sampled at twice the rate of non-SD/LEP non-minority students, non-minority SD/LEP students are sampled at twice the rate of non-SD/LEP non-minority students, and minority SD/LEP students are sampled at four times the rate of non-SD/LEP non-minority students. Note that oversampling a group of students by a factor of x decreases their weights by a factor of 1/x.

Example: A public school with high minority enrollment has 400 fourth-graders, 80 of whom are SD/LEP and 320 of whom are non-SD/LEP. 100 fourth-graders were selected in the initial student sample (20 SD/LEP and 80 non-SD/LEP). This leaves 60 SD/LEP students eligible for the oversample, and 20 are selected. There is no Black or Hispanic oversample. Thus the student sample in this school contains 80 non-SD/LEP students and 40 SD/LEP students. Each non SD/LEP student has SAMP_INT = 4 and each SD/LEP student has SAMP_INT = 2.

The subject allocation weighting factor (SPL) adjusts for allocating sampled students across the subjects assigned to a school. Its value is a function of grade, subject, the number of sessions assigned to a school, the number of sessions of the given subject assigned to the school, and school type. For example, in a school in which one session each of math and science were assigned, each student has SPL = 2.

The substitution adjustment factor adjusts for the difference in grade enrollment prior to sampling between the cooperating substitute school and its corresponding original school. It applies only to cooperating substitute schools and is calculated as follows:

SUBADJ = GRD_ENRL ÷ SUB_ENRL

where

GRD_ENRL is the grade enrollment of the original school; and
SUB_ENRL is the grade enrollment of its corresponding substitute school.

This adjustment is necessary in order for the weight of the substitute school to reflect the weight of the original school. For original schools, SUBADJ is set to 1.0. For substitute schools, SUBADJ can be less than, equal to, or greater than 1 depending on the size of the substitute school versus the size of the original school.

Last updated 17 April 2008 (TS)

Printer-friendly Version