Award: Evaluation of Noise Infusion for Large-Scale Demographic Sample Survey (Survey of Doctorate Recipients) (SDRN)

Award Date: 09/27/2023
Project Number: SDRN-23-N02
Project Title: Evaluation of Noise Infusion for Large-Scale Demographic Sample Survey (Survey of Doctorate Recipients)
Project Performer: Knexus Research Corp.
Project Partners: Tumult Labs, Strategix
Award Amount: $798,962.00
Objective: The objective of this research and development project is to evaluate the use of noise infusion for a demographic survey as a possible privacy-preserving method in the use of federally confidential data. Investigations will focus on use cases where noise infusion may be appropriate and use cases where noise infusion may introduce quality issues that reduce confidence in the use of estimates for decision making. The Survey of Doctorate Recipients (SDR) is a sample survey that provides data on the characteristics of science, engineering, and health doctorate degree holders. The SDR provides data useful in assessing the supply and characteristics of U.S.-trained science, engineering, and health doctorates employed in educational institutions, private industry, professional organizations, and government in the U.S., as well as in other countries worldwide. To address disclosure concerns with the SDR while maximizing data utility, noise infusion is being explored as an alternative to augment other disclosure limitation methodologies currently in use with the SDR, both for the restricted-use data and the public-use microdata file. This project will inform the National Secure Data Service Demonstration Project (NSDS-D) by exploring noise infusion for a sample survey and assessing the disclosure protections and quality considerations for the resulting estimates.
Period of Performance: September 2023 – September 2025
Point of contact: Christine Task| Knexus Research Corp. | christine.task@knexusresearch.com

Request for Solutions (Closed)

Overview Release Date Submission Date Documents
The Government would like to evaluate the use of noise infusion for a demographic survey as a possible privacy-preserving method in the use of federally confidential data. Investigations will focus on use cases where noise infusion may be appropriate and use cases where noise infusion may introduce quality issues that reduce confidence in the use of estimates for decision-making. 06/26/2023 07/17/2023