Applications of Statistical Models That Detect Daily Changes Using Key Estimates from the American Community Survey Due to the U.S. Census Bureau Regional Office Restructure

• Activity Number: 341
• Type: Topic Contributed
• Date/Time: Tuesday, August 6, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Survey Research Methods Section
• Author(s): Lindsay McMillan*+ and Robyn Sirkis
• Companies: US Census Bureau and U.S. Census Bureau
• Keywords: Data Manipulation ; Responsive Design ; American Community Survey

Abstract

In 2012, the U.S. Census Bureau shifted field operations from 12 to six regional offices (ROs). To monitor this shift in management we built models for key variables for many different surveys to determine if there were significant changes in our estimates. Models were run on unedited Computer Assisted Personal Interview (CAPI) data on a daily basis. This paper discusses the models used for the American Community Survey (ACS). First, we will give a brief overview of the scope of this project and discuss the key variables, predictor variables, and how models were built using 2011 ACS unedited data. Second, we discuss the automated system used to generate graphs for each key variable. These graphs track the daily change on the coefficient for the management structure indicator variable from the models. Finally, we present the results of some of our models and discuss possible ways to improve/use these models in the future.