Project Scope
Researchers asking participants to verify the month and year of each move gathered 2,782 addresses from 1,771 respondents who had relocated at least once.
This report examines the effectiveness of two techniques for cleaning and correcting incomplete address information. It utilizes data collected from residents of White Center, Wash., as part of Annie E. Casey Foundation’s Making Connections study. Making Connections was an ambitious 10-site, longterm initiative of Casey's devoted to advancing the premise that supportive communities can help empower families and enhance outcomes for children.
Researchers utilized the retrieval approach to complete 82% of the addresses in their dataset while the administrative cleaning crew gathered full details for just 33% of their assigned addresses.
Researchers asking participants to verify the month and year of each move gathered 2,782 addresses from 1,771 respondents who had relocated at least once.
In this study, retrieval took nearly twice as long as the administrative cleaning option and could prove costly for researchers crunching larger datasets.
Questionnaires administered in English were 1.8 times as likely to yield addresses relative to questionnaires delivered in Spanish or Vietnamese.
Respondents who self-identified as black were 1.4 times more likely to report data-friendly addresses compared to their white counterparts.
This report sizes up two different data cleaning approaches. Option one: Administrative cleaning, where research assistants polish data via free resources and on-hand information. Option two: Retrieval, where a field-savvy team fills in knowledge gaps by calling residents and utilizing investigative, fee-based web tools. The cut-to-the-chase conclusion? Retrieval yields better results, hands down.
