Journal of Official Statistics, Vol.20, No.4, 2004. pp. 687–703
Maintaining Race and Ethnicity Trend Lines in U.S. Government Surveys
Elizabeth Greenberg, Jon Cohen and Dan Skidmore
Abstract:In 1997 the U.S. Office of Management and Budget issued guidelines that require federal government surveys to allow respondents to report multi-racial identifications. Maintaining trend lines when reporting racial breakdowns requires a linkage between the old and new measures of race and ethnicity. For reasons of both cost and statistical efficiency, linking studies are best based on data in which the same sample is asked both the old and new race/ethnicity questions. However, if respondents are influenced by the order in which the questions are asked, complicated statistical modeling will be required to estimate and adjust for the influence of the joint administration of the items to a common sample. This article uses data from the National Assessment of Educational Progress 2000 field test to show that the sequence in which race/ethnicity questions are administered has a small but important effect on the likelihood that respondents will identify more than one race/ethnicity.
Keywords:Race and ethnicity linking, racial identification, statistical trend, priming effect
Copyright © Statistics Sweden 1996-2018. Open AccessISSN 0282-423XCreated and Maintained by OKS Group