NAEP Technical DocumentationWeighted Enrollment Checks from the 2002 Main Study Responding School Sample

The responding school sample consists of the eligible schools that cooperated with the NAEP 2002 national main assessment. This sample is a subset of the full school sample. The weights for responding schools received school nonresponse adjustments to account for the nonresponding schools. If the nonresponse adjustments adjusted successfully for nonresponse biases, the weighted estimated grade enrollment aggregations for the responding schools should have estimated the same population totals as the full school sample using the school base weights. Thus, significant differences between these totals may have pointed to nonresponse biases which are not accounted for in the nonresponse adjustment process.

Measuring the standard error of the difference between these weighted aggregations (to test the null hypothesis of no difference) was a nontrivial undertaking. Both have sampling error. The approach derived an approximate sampling error for the difference by concatenating the original school sample and responding school sample files together (so that each responding school was represented twice on the file), and treating it for this purpose as a single “sample.” The weights for the original school sample were the school base weights and the weights for the responding school sample were the nonresponse-adjusted school weights. The sample error was computed for the difference using these special file weights.

For the fourth-grade public school sample, the responding school sample percentage of Black students was 17.5 percent, significantly different than the 17.0 percent for the full school sample. The Native American percentage was also significantly different (1.1 percent for the responding school sample and 1.2 percent for the full school sample). Median income was 400 dollars lower for the responding school sample as opposed to the full school sample. No significant differences were found for fourth-grade private schools.

For the eighth-grade public school sample, the statistically significant differences were as follows. The responding school sample percentage of Black students was 15.8 percent, as opposed to 15.2 percent for the full school sample. Hispanic student percentages tipped in the same direction: 15.0 percent for the responding school sample and 14.4 percent for the full school sample. It can be noted that most of this difference was within the Southeast Region. The Native American percentage was 1.0 percent for the responding school sample and 1.2 percent for the full school sample. Median income was 370 dollars lower in the responding school sample as compared to the full school sample. Private schools showed no significant differences for the eighth grade.

For the twelfth-grade public school sample, only median income and type of location showed statistically significant differences. Median income was 840 dollars lower for the responding school sample as compared to the full school sample. The type of location mean was 0.06 units higher in the responding school sample, on a numeric scale in which a larger number indicates lesser urbanicity (a greater degree of 'rurality'): the extreme values were 1 (center city) and 8 (rural). No significant differences were found in the private school sample.

Last updated 24 September 2008 (LR)

Printer-friendly Version