The target sample sizes for fourth- and eighth-grade public school samples for the 2003 assessment were specified at the jurisdiction level, rather than at the national level, so that the samples would satisfy the needs of both the national estimates and the participating jurisdictions estimates. For each participating jurisdiction (except the Trial Urban District Assessment districts), the sample size requirement was approximately 2,500 assessed students per grade and subject. This sample size was sufficient to meet national sample requirements for these two subjects at grades 4 and 8 for public schools. To compensate for the possibility of states and jurisdictions dropping out of the state assessment, within each jurisdiction probability subsamples of schools were identified for the national main assessment. Private school samples were selected for the national main assessment at grades 4 and 8. School samples for grade 12 were only selected for pilot testing items and not for the operational, reportable sample.