Sampling Errors for SESTAT

SESTAT combines information from three surveys sponsored by the National Science Foundation:

• The National Survey of College Graduates (NSCG),
• The Survey of Doctorate Recipients (SDR), and
• The National Survey of Recent College Graduates (NSRCG)

The combined SESTAT database represents U.S. residents with a baccalaureate degree or higher who were trained as scientists or engineers as of the reference period of April 15 for the four survey years of 1993, 1995, 1997 and 1999 or who were working as scientists or engineers as of the week of April 15, 1993.

Evaluating the Accuracy of SESTAT Estimates

The discussion presented below provides information about sampling errors for the combined SESTAT surveys and includes:

• An explanation of how sampling errors were developed
• An overview of standard errors and the resulting confidence intervals
• A general discussion of methods used to derive generalized variance functions for estimating standard errors
• A description of how users can calculate standard errors for estimates from SESTAT
• A description of the generalized variance functions used for SESTAT.

Click on the underlined topic below for more information:

Understanding Sampling Errors: An overview of standard errors and the resulting confidence intervals

Generalized Variance Functions: A Methodology for Estimating Standard Errors: A general discussion of methods used to derive generalized variance functions for estimating standard errors.

Calculating Standard Errors: A description of how users can calculate standard errors for estimates from SESTAT. Also includes a description of the GVFs used for SESTAT.

Download This Methodology Report: You may find it more convenient to refer to a hard-copy version of this discussion. Click here to download a complete copy of Sampling Errors of SESTAT: 1993/1995, 1997, and 1999 -- approx. 22 pages (PDF Format)

'
Updated: September 26, 2001