Table of Contents  |  Search Technical Documentation  |  References

NAEP Weighting Procedures → Weighting Procedures and Variance Estimation for the 2000 National Main Assessment → Modular Weights for the 2000 National Main Assessment

NAEP Technical DocumentationModular Weights for the 2000 National Main Assessment

After student trimming, weighting continues in two parallel processes. Reporting weights are the result of one of these processes, and modular weights are the result of the other. Modular weights are used for analyses involving one sample type or for comparing one sample type against the other. Each sample type separately weights up to the target population. The weights are calculated for each sample type by grade and subject for the national main assessment. For example, modular weights could be used to compare the NAEP performance of limited-English-proficient (LEP) students in the S2 sample with the performance of LEP students in the S3 sample.

Modular weights differ from reporting weights in two ways:

• Modular weights do not contain reporting factors.
• Modular weights are poststratified like reporting weights, except that each sample type within each grade and subject are poststratified separately. Some adjustment factors are quite variable for the same adjustment cell across different sample types for the same grade and session. This indicates that the individual samples by sample type may not be particularly stable and that comparing the sample types to each other should be done with caution.

Printer-friendly Version