Table of Contents  |  Search Technical Documentation  |  References

NAEP Sample Design → Sample Design for the 2000 Assessment → National Main Assessment Sample Design in 2000 → Sampling of Schools in the 2000 National Main Assessment → Assigning Measure of Size and Selecting School Samples for the 2000 National Main Assessment → Grade 4, 8, and 12 Private School Samples for the 2000 National Main Assessment

NAEP Technical DocumentationGrade 4, 8, and 12 Private School Samples for the 2000 National Main Assessment

The 2000 national main assessment expanded the oversampling of private schools to allow the reporting of assessment results for up to six groups defined by religious orientation and/or private school association membership. Schools on the private school frame were grouped into seven strata for each grade separately. The seven strata corresponded to the six reporting groups plus an additional stratum for those private schools whose religious orientation and private school association membership were unknown. The first five reporting groups listed are mutually exclusive and exhaustive.

• "Independent" schools were those with a membership in the National Association of Independent Schools (NAIS).

• Catholic schools were those oriented or affiliated with the Roman Catholic Church.

• Lutheran schools were those oriented or affiliated with one of the following churches:

  • Lutheran Church - Missouri Synod, or

  • Evangelical Lutheran Church in America.

• Conservative Christian schools were those with a membership in one of the following organizations:

  • Accelerated Christian Education,

  • American Association of Christian Schools,

  • Association of Christian Schools International, or

  • Oral Roberts University Educational Fellowship.

• Other religious schools were those that did have a religious orientation or affiliation but were not Catholic, Lutheran, or conservative Christian

• Nonsectarian schools were those with no religious orientation or affiliation.

Student assessment data from an independent school potentially contributed both to an independent school estimate and to the estimate for one of the other five groups (Over 70 percent of independent schools are nonsectarian. An additional 22 percent would be classified as "Other Religious.") In creating the sampling strata, NAIS membership took precedence over any other affiliation. Also because only a very small percentage of Catholic schools were NAIS members, Catholic schools for which NAIS membership was unknown were designated to the Catholic stratum, not to the NAIS stratum.

Schools within each explicit sampling stratum for three grade-specific private school frames were sorted in a serpentine fashion by the following:

1. school size (school size for private schools is a dichotomous variable defined as having at most 19 grade-eligible students or at least 20 grade-eligible students);

2. Primary sampling unit (PSU) stratum;

3. school type (school type for private schools includes Catholic, other religious, and nonsectarian); and

4. estimated grade enrollment.

For grades 4, 8, and 12 private school samples, schools were sampled at a rate designed to meet specific student-level sample sizes for each sampling stratum except for the unknown affiliation stratum. An equal probability sample of 25 schools was selected from the unknown affiliation stratum for each grade. In the remaining six sampling strata, school sample sizes were determined so that the national target of students to be assessed would be selected (adjusted for nonparticipation of schools and students and exclusion of students) and that the probability of selection of a student would be uniform over the United States (except students from very small schools), given that all students in schools with fewer than a specified number of eligible students would be selected and that only this specified number of students would be selected in larger schools. These school-level sample sizes also took into account that very large schools could be sampled more than once in grade 8 (2 times) and grade 12 (up to 3 times), as discussed in on the web page titled Initial Student Sample.

The target number of completed assessment booklets for Catholic schools was 3,000 for mathematics assessment and 3,000 for science at each of grades 4, 8, and 12. The target was 750 per subject and grade for each of the other five reporting groups. At grade 4 alone, a reading assessment and a special study for the mathematics assessment were conducted, with overall targets of 2,000 and 1,000 completed booklets, respectively. The samples for these last two subjects were allocated in rough proportion to the national distribution of fourth-graders across the six private school reporting groups. Because not all sampled schools and students were expected to participate, some allowance was made for attrition: the target student sample sizes were inflated to offset loss due to school ineligibility, student exclusion and student nonresponse. In years past, private school sample attrition has been consistently greater at the higher grades and for non-Catholic schools. Consequently, the inflation factors used to set the before-attrition student sample size targets for the 2000 assessment, and based on NAEP 1998 survey experience, varied by grade and Catholic affiliation.

Printer-friendly Version