Skip to main content
Skip Navigation

Table of Contents  |  Search Technical Documentation  |  References

NAEP Technical DocumentationComputing Student-Level Replicate Weights

This section describes the computation of replicate student base weights. The corresponding calculation of student base weights is given in Computation of Student Base Weights. The replicate student base weights differ considerably between certainty and noncertainty schools. For noncertainty schools, school selection is the first stage of selection and replication is based on the noncertainty school strata. The students “inherit” the replicate weights of their parent noncertainty schools. For sampled students in noncertainty schools the replicate base weights are defined as follows (for school s, student k, replicate weight r):

STUWGTsk(r) =

ws(r) * SCHSESs(A) * SNRADJc(r) * SUBADJs * TRIMs * WINSCHLs * SESsk(A) * YRRNDs

where

The school weighting factors are replaced with replicate weighting factors, but none of the student weighting factors. For students in certainty schools, students are the first stage of selection. In this case, the student weighting factors are replicated to reflect student selection. These replicate factors are given as REPFACsk(r). The overall student replicate weight for students in certainty schools is as follows:

STUWGTsk(r) =

ws * REPFACsk(r) * SCHSESs(A) * SNRADJc(r) * SUBADJs * TRIMs * WINSCHLs * SESsk(A) * YRRNDs

Note that ws is not replicated, but the school nonresponse factor is. Certainty schools will have replicate weighting factors that differ from the full-sample weighting factor. Note that almost all absolute certainty schools were in the fourth and eighth grades' public school jurisdiction alpha samples.1

1 With the exception of one absolute certainty school in the twelfth-grade private school sample.


Last updated 19 February 2009 (RF)

Printer-friendly Version