How to Read SEELS Tables
  You will go through a three-step process to select the data you would like to see: (1) Data Source, (2) Conceptual Area, and (3) Topic. You will then be able to view your selection by selecting the "View Tables" button. Below we provide information regarding how to use and interpret the data tables.  
  • Your selected data are presented in one single page containing 10 tables (5 data tables and 5 standard error tables). A table of contents frame at the left allows you to quickly move from table to table.
  • The five (5) data tables present your selected topic cross-tabulated by student disability category, gender, age group, family income, and ethnicity/racial group, each in a separate table.
  • The percentages in the data tables are organized by column, so that the number indicates the weighted estimate of respondents nationally who responded in that way. In the example below, 13.8 percent of students with speech impairments were reported to be in 'good health.'
  • Following each data table is a table indicating standard errors associated with the data in the preceding table. Standard errors are an indication of the precision of the reported percentage. Small standard errors allow for greater confidence to be placed in the value. Large standard errors, on the other hand, require a greater degree of caution in interpretation.
  • The size of individual standard errors is linked to the unweighted number of respondents in a specific category of students. In SEELS, there are several groups of students whose numbers in the unweighted sample are comparatively small and are often associated with large standard errors. Among these groups are the following: students classified by the school as deaf-blind, Native American, or ages 13 and over. Users should use caution in interpreting results involving these groups.
  • Standard errors can also be higher because certain questions are only asked of a subset of respondents based on their responses to previous questions.
  • Some questions were asked to comparatively small numbers of respondents (as low as n=100). Tables displaying these data will show values only if the unweighted subgroup was more than 35. So, it is possible to have valid values for the population, but values for the subgroups may be blank.
  • Use the following format to cite the source of data coming from this site:
    Special Education Elementary Longitudinal Study. (2002). SEELS data tables. Available at:
  • Finally, we hope that you enjoy the tables and find them useful. We are always trying to improve the site's usefulness, so if you have feedback, please contact us at: SEELS Project Team.



Example Data Table

Explains elements of two tables:

Last Updated 05/03