A Comparative Methodological Study of Underenumeration and Related Problems
Charles A. Valentine and Betty Lou Valentine
KEY WORDS: coverage error, participant observation, match study, undercount, erroneous enumeration, match study, disclosure, privacy, trust, household structure
This classic study, conducted in the late 1960s, greatly influenced the Census Bureau’s understanding of the sources of the differential undercount affecting its census and surveys. In a double-blind study, resident ethnographers conducted a participant observation study of 25 households in one block of a poor inner city community. Census Bureau interviewers conducted interviews in the same households, and the ethnographers’ observations of who resided in the households were matched against household rosters collected by the interviewers. The ethnographers found that many more adult males lived in the households than were reported to survey interviewers. The omissions resulted in misleading survey statistics that greatly overstated the fraction of female-headed households in the community.
Based on the characteristics of households that were misreported, the authors argued that adult men were not reported when disclosing their presence to governmental authorities could risk vulnerable sources of financial support to households. Confusion about the concept of household or misunderstandings between respondents and interviewers also contributed to some errors, but they were erroneous inclusions rather than omissions.
The authors argued that various procedural improvements proposed to address or measure census undercounts were unlikely to succeed because they did not address the basic problem identified in this study: the relationship between those seeking the information and those asked to provide it. The authors proposed that the research designs of survey programs should regularly include independent ethnographic checks on the quality of information collected in surveys.
CITATION: Charles A. Valentine and Betty Lou Valentine. 1971. Missing Men: A Comparative Methodological Study of Underenumeration and Related Problems. Report prepared under contract for the U. S. Census Bureau. Washington, D. C.