A tutorial in assessing disclosure risk in microdata.

Area of Science:

• Statistics
• Data Privacy
• Information Security

Background:

• Statistical agencies release microdata for research and policy, necessitating robust confidentiality protection.
• Current disclosure risk assessment often relies on subjective judgment, lacking objective criteria.
• Altered data releases may still pose disclosure risks, highlighting the need for improved methods.

Purpose of the Study:

• To provide an overview of the statistical disclosure control framework for microdata.
• To focus on the risk analysis stage within this framework.
• To define existing quantitative disclosure risk measures and their estimation.

Main Methods:

• Review of established statistical disclosure control methodologies.
• Definition and explanation of various quantitative disclosure risk measures.
• Discussion on estimating these risk measures using available software.

Main Results:

• Identified limitations of subjective methods in assessing disclosure risk.
• Presented a range of quantitative risk measures for objective evaluation.
• Demonstrated how to estimate these measures, facilitating practical application.

Conclusions:

• Quantitative risk measures offer a more objective approach to assessing disclosure risk in microdata.
• Adoption of these measures enhances the statistical disclosure control framework.
• This tutorial equips agencies with tools for more informed and secure data release decisions.