Data Presentation Standards Within the Federal Statistical System

• Activity Number: 682
• Type: Topic Contributed
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Government Statistics Section
• Author(s): Jennifer Madans* and Ron S. Jarmin* and Art Hughes* and Virginia Lesser*
• Companies: National Center for Health Statistics and U.S. Census Bureau and Substance Abuse and Mental Health Services Administration and Oregon State Univeristy
• Keywords: precision ; suppression ; standards ; SAMHSA ; Census ; NCHS

Abstract

Federal statistical reports regularly present information tabulated for population, geographic and other subdomains. These reports can contain a wide variety of variables with a range of properties, including rare dichotomous outcomes and highly skewed distributions. The combination of rare outcomes with small subdomains can lead to highly variable estimates. To balance producing estimates for a wide range of users and producing stable estimates, agencies utilize programmable presentation guidelines to ensure adequate precision. Presentation guidelines based on sample sizes, standard errors, confidence intervals, or other factors have been used. These guidelines can differ across data systems, agencies and data systems within agencies. In this session, the panelists will first describe data presentation guidelines for tabular data at their federal agency. Second, the panelists and the discussant will discuss commonalities and differences in these standards among agencies, as well as considerations and assumptions made among agencies when developing or revising the standards.