Table of Contents | Search Technical Documentation | References
For new schools, measures of size were computed as follows:
where is the probability of selection of the school's district into the new-school district sample. If zjs was less than 1, then it served as the new school's probability of selection. If zjs was greater than 1, in some cases it was trimmed. (For fourth grade, all jurisdiction zjs values were trimmed to 3 if they were greater than 3, except for Alaska which had a trimming point of 8. In eighth grade, the trimming level was generally 3, except for Alaska (8), Delaware (6), Hawaii (5), and Rhode Island (4). Note that Delaware for the fourth grade and Delaware, Hawaii, and Rhode Island for the eighth grade all had manual modifications made to their Ejs values.)
The new-school samples were implicitly stratified using a hierarchical ordering by jurisdiction, district, and estimated grade enrollment. The sample sizes are given in the linked table below.
In a small number of jurisdictions—American Samoa, District of Columbia, DDESS, DoDDS, and Virgin Islands—all of the newly identified schools and their students were included in the 2003 state assessment. Schools and students in these jurisdictions were also treated as "take-all" in the state assessment samples based on the CCD frame.
In addition to the regular new-school procedure described above, new schools that were discovered by field staff in small districts during the field period were added to the sample. Since all schools in small districts were included in the sample, the new schools in these districts were also included with no sampling of the schools. The distribution of these schools by state is shown in the table below.
State | Fourth grade | Eighth grade |
---|---|---|
Georgia | 3 | 0 |
Iowa | 1 | 0 |
Kansas | 0 | 1 |
South Carolina | 1 | 0 |
SOURCE: U.S. Department of Education, Institute of Education Sciences, National Center for Education Statistics, National Assessment of Educational Progress (NAEP), 2003. |