Australian Bureau of Statistics
1504.0 - Methodological News, Mar 2013
Previous ISSUE Released at 11:30 AM (CANBERRA TIME) 20/03/2013
|Page tools: Print Page Print All RSS Search this Product|
Disclosure-Protected Inference with Linked Micro-data using a Remote Analysis Server
Large amounts of micro-data are collected by data custodians in the form of Censuses and administrative sources. Often, data custodians will collect different information on the same individual. Many important questions can be answered by linking micro-data collected by different data custodians. For this reason, there is a very strong demand from analysts, within government, business and universities, for linked micro-data. However, many data custodians are legally obliged to ensure the risk of disclosing information about a person or organisation is acceptably low. Different authors have considered the problem of how to facilitate reliable statistical inference from analysis of linked micro-data while ensuring that the risk of disclosure is acceptably low. The methodology area of ABS wrote a paper for its Methodology Advisory Committee (MAC) meeting in November 2012 that considered this problem from the perspective of an Integrating Authority.
An Integrating Authority is trusted to link the micro-data and to facilitate analysts' access to the linked micro-data via a remote server. A remote server allows analysts to fit models and view the statistical outputs without being able to observe the underlying linked micro-data itself. One disclosure risk that must be managed by an Integrating Authority is that one data custodian may use the micro-data it supplied to the Integrating Authority and statistical outputs released from the remote server to disclose information about a person or an organisation. The MAC paper measures the utility and disclosure risk of a proposed method in simulation and with a real example. The evaluations show that the protections prevent disclosure in a high risk scenario and have only a small impact on inferences for analysis involving moderate sample sizes.
The paper is available from James Chipperfield and may be made available on the Australian Bureau of Statistics website.
These documents will be presented in a new window.
This page last updated 27 June 2013