Arias E, Lucas JW [PDF - 176 KB]. Exploring Alternative Methods for Bridging Multiple Race Data. National Center for Health Statistics Data Analysis Group Presentation, January 2002.
Abstract – One of the provisions of the 1997 revised OMB standards for the collection of racial and ethnic data. This analysis presents a model-based approach to bridging data, which uses characteristics of respondents to predict how they might select a single race group. Data from the 1986-1994 NHIS were used to link to the 1990 Census public use STF3A data (to include county-level population characteristics). We estimate multinomial logistic regression models to develop predicted probabilities of selecting a primary race (based on the NHIS follow-up race question) as a function of selected individual and contextual-level characteristics. The results indicate that the combination of factors associated with selecting a primary race group differs greatly for the various multiple race groups, as well as important differences within each bi-racial group in the selection of "primary" vs. "multiple" or refusal to select a single race. The findings suggest that developing algorithms for assigning multiple race persons to a single race group may not be straightforward; group-specific algorithms should be developed that take into account the relevant factors for each group.
Lucas JW [PDF - 125 KB]. Measuring Racial and Ethnic Disparities in Health Using Data from the National Health Interview Survey. Presented at the National Committee on Vital and Health Statistics, Subcommittee on Subpopulations Hearings, February 11-12, 2002.
Abstract - The elimination of health disparities, including those that affect racial and ethnic groups in the United States, is a major objective of the U.S. Department of Health and Human Services. Central to the issue of tracking changes in health outcomes for racial and ethnic populations is the measurement of race and ethnicity (Hispanic origin). This presentation provides examples of how data from the National Health Interview Survey are used to examine racial and ethnic disparities in health. It also illustrates some of the challenges that exist in measuring racial and ethnic disparities in health in light of the revised OMB standards for collecting data on race and ethnicity.
Lucas JW, Madans JH [PDF - 94 KB]. Consistency of race reporting in two national health surveys. Presented at the Population Association of America Annual Meeting, April 2002.
Abstract - The 1997 revised race and ethnicity standards represented a significant change in the way federal data on race and ethnicity are collected, and raise questions about how this change in collection methodology will affect analyses involving race and a variety of outcomes. This study furthers the understanding of the complexity of race reporting by using data from two large national health surveys to assess the consistency of Hispanic origin, single race, and multiple race reporting. We used data from the linked 1993 National Health Interview Survey (NHIS) – 1995 National Survey of Family Growth (NSFG) file to examine the consistency in race/Hispanic origin reporting among respondents to these two surveys. A summary measure of agreement on the race responses was created for these analyses, and SUDAAN was used to compute point estimates, standard errors, and conduct significance testing and logistic regression modeling. The results indicated that the agreement between survey responses was high for Hispanic origin, but significantly lower for race. Adjusting for all factors, Hispanic origin, nativity, marital status, education, and region of residence all play a role in the consistency of race responses. These findings suggest that the understanding of and ability to report race remains a problem for persons of Hispanic origin. The high level of agreement on Hispanic origin reporting and significantly lower level of agreement on race reporting - particularly multiple race reporting – also raises some important questions about measuring multiple race groups.
Lucas JW [PDF - 63 KB]. Assessing the health status of multiracial persons: data from the National Health Interview Survey. Presented at the DHHS National Leadership Conference on Eliminating Racial and Ethnic Disparities, July 2002.
One major result of the changes to the federal data standards for collecting racial and ethnic data is the need to study and understand patterns of health and health behavior among single and multiple race groups in the United States. In this presentation, descriptive analyses of demographic and health data from the 1997-1999 National Health Interview Surveys are presented to provide information on health outcomes for single and multiple race population groups. The findings indicate that single and multiple race groups differ significantly from one another on both demographic and health measures. Additionally, multiple race groups differ both from one another and from their single race counterparts. Patterns of association for health outcomes among single and multiple race groups are not necessarily intuitively obvious, and require further analysis. Additionally, because the multiple race population remains relatively small, the ability to study health outcomes in this population is limited.