Once users find material, they should assess its quality and identify sources of uncertainty [error.] Statistical principles are based on the idea that all observation contains uncertainty. Identifying it increases rather than reduces its credibility.
In order to do this, users will need:
Material about the procedures used to collect the information
Material that documents its origins and authenticity
A description of the data collection, compilation and processing methods used by the data creators.
Measures of uncertainty—i.e. the results of statistical estimates that quantify the likely degree of uncertainty
Many users need to conduct their own assessment of the limits of information they use.
These kinds of users need
Copies of the original Micro Data Sets. These are frequently released as Public Use Data sets which have been examined to be sure that the identity of the subjects—persons, providers, hospitals and payers cannot be identified.
Each agency provides detailed information about the availability of its records for analysis. Many allow researchers who promise to protect privacy access under controlled circumstances designed to allow research but at the same time protect subject’s privacy.
Page 39 of 40
Last reviewed: 03 January 2008 Last updated: 03 January 2008 First published: 03 January 2008 Metadata| Permanence level: Permanent: Dynamic Content