NAEP Technical DocumentationReplicate Group Assignments for SD/LEP Students

The R2 reporting population includes all of the non SD/LEP students, but only SD/LEP students in sessions in which accommodations were not offered. The R3 reporting population includes all of the non SD/LEP students, but only SD/LEP students in sessions in which accommodations were offered. Thus some SD/LEP students are left out of each reporting population. This means that certainty schools were treated as certainties when forming replicate groups for the non SD/LEP students, but these same certainty schools were treated as noncertainties when forming replicate groups for the SD/LEP students. The replicate group assignments for SD/LEP students in noncertainty schools remained the same as that used for non SD/LEP students in those schools. The re-assignment of replicate groups for SD/LEP students in certainty schools was implemented as follows:

All schools except those in American Samoa, Guam, Virgin Islands and the Department of Defense Domestic Dependent Elementary and Secondary Schools DDESS Grade 8. The assignment of schools to replicate groups was done separately for various subgroups of the SD/LEP sample. The certainty schools (being treated as noncertainties) were first sorted by "sample type" within jurisdiction. Within each sample type, the schools were sorted by their new school status and sample selection order. In those jurisdictions where the number of replicate groups for schools did not exceed 62, the schools in the sorted list were assigned group numbers, two to a group, beginning where the previous assignments for the noncertainty schools with non SD/LEP students stopped. If the number of schools was odd, then the last three schools were assigned to the same replicate group. If the number of replicate groups for schools exceeded 62, then the group numbering started at 62 and proceeded backwards, assigning pairs of schools to the same replicate group. If the number of schools to be assigned was odd, the last three schools were assigned to the same replication group. In situations where there was only one certainty school with SD/LEP students assessed, the school was assigned to the last replicate group used for the noncertainty schools with non SD/LEP students. If there was an odd number of such schools, then the triple was broken up into two doubles and the school in question was assigned to the last double.

American Samoa, Guam, Virgin Islands, DDESS Grade 8. Schools were assigned to replicate groups in the following way:

All schools were grouped into a single replicate group.
Schools were randomly sorted.
Starting with the second school, replicates were formed by consecutively leaving out one of the remaining n - 1 schools; each replicate included the first school.

Last updated 20 March 2008 (MH)

Printer-friendly Version