|
Abstract:
|
To determine where to count people who have multiple places to stay (e.g., second home, relatives' homes) on Census Day, Census coverage operations use a set of "cycle" questions that ask how often a person goes back and forth between places. These questions offer predefined patterns (e.g., going between places every week) and ask where the person spent most of the time during a specified time period (e.g., March and April). Because the cycle questions presume set patterns of living situations that may not reflect reality, we investigate an alternative method of assigning residency. The new "dates" method involves collecting dates of stays for each address and calculates where a person should be counted on census day given the dates. This paper presents findings from an experiment that evaluated the effectiveness of this alternative method of assigning residency.
|
