Information needed to administer the federal SSI program resides on the Supplemental Security Record (SSR) file, which in December 2001 contained records for approximately 6.4 million individuals who received a federal SSI benefit for that month. The SSI Public-Use Microdata File drawn from the SSR represents only persons who received a federal SSI payment in December 2001; it excludes persons who received a federally administered state supplementary payment but no federal SSI payment in December 2001.

The SSI Public-Use Microdata File contains an extract of data fields from the SSR and consists of a 5 percent random, representative sample of persons who received a federal SSI payment in December 2001. Records in the underlying SSR for persons who received a federal SSI payment for December 2001 were ordered by state, by disability diagnosis within state, by age within disability diagnosis, and then by sex within age; a 5 percent systematic random sample was then selected. The file contains approximately 320,000 records, with 13 data fields on each record, and is available in a standard SAS file structure as well as the CSV file format.

Since the sample for the SSI Public-Use Microdata File is a representative 5 percent systematic random sample of persons receiving an SSI payment in December 2001, all records have weight equal to 20. Variance and standard errors can be approximated with the standard formulas used for simple random sampling.

Descriptions of data fields are provided in the Data Dictionary for this file. The descriptions cover the content of each field, the method of presentation, and the disclosure avoidance steps taken to provide confidentiality. Because information on the internal SSR record file is used to administer the SSI program, data quality is extremely high because data are validated and kept current. Note that the information on each file refers to a single time period—December 2001.

Individual identifiers have been removed from all records on this file and other distinguishing characteristics have been modified to prevent identification of persons to whom a record pertains. Records are sequenced in random order.

Topcodes and bottomcodes are employed for all numeric fields to avoid showing extreme field values on a data record. Values beyond the topcode or bottomcode are replaced by the average of the values in excess of the respective topcode or bottomcode. Topcode and bottomcode values were derived at the national level, and the replacement values are derived and applied at the state level for this file (see table). Values shown for categorical fields are groupings of detailed information from the internal file, and numeric variables are rounded to the nearest $5. All topcode and bottomcode values, replacement values, and information on detail groupings are provided to users as part of file documentation.