Socio-Economic Status and Names: Relationships in 1880 Male Census Data
Rebecca J. Vick, University of Minnesota
Record linkage is the process of connecting records from two or more data sources for the same individual. Linked files are uniquely rich in information about individual life changes such as migration, occupational mobility and household composition. The historical linking process would not be possible without name data. A powerful way to minimize false links is to NOT link when a record has high similarity with multiple records. This often occurs with common names which leads to a disproportionate number of those with common names being excluded from linked sets. This paper aims to check for relationships between name commonness and socio-economic status, which would be of interest to record linkers whose top concern is avoiding bias in linked sets.