NHIS - Race and Hispanic Origin Information
Frequently Asked Questions
Race and Hispanic origin are variables that are very important in analyzing the data we collect. For example, such information helps us learn whether people in one group are more or less likely to have health insurance coverage or access to health care services than people in another group. In turn, this information helps policymakers better understand how to improve the health care system.
Since race and Hispanic origin information is very important to our understanding of the health data we collect, we encourage respondents to report as much information as they are able and willing to give us. However, we understand that race and Hispanic origin are complex concepts and that some respondents may have difficulty answering these questions, including those who are reporting race and Hispanic origin for other members of the household (i.e., proxy responses). In the event that respondents feel that they cannot answer questions on race or are unwilling to do so, they can indicate "Don’t Know" or "Refused" as their response, as they can on any other question on the survey.
In the 1995-2005 NHIS sample design, the final stage of the development of person-level weights for the NHIS includes procedures that ensure that NHIS estimates for approximately 88 age-sex-race-Hispanic origin categories are consistent with independent population estimates that are developed by the U.S. Bureau of the Census. These procedures require that each person in the NHIS sample have a valid response for age, sex, race, and Hispanic origin. Those respondents who do not have a response for race and/or Hispanic origin are assigned one through imputation.
The imputation procedures used on the NHIS race and Hispanic origin data are based on those developed by the U.S. Bureau of the Census for the creation of the Modified Race Data Summary file. Imputation flags are also available on the NHIS public use data file for users who wish to analyze unimputed race data. More information on the NHIS imputation procedures can be found in Appendix II of the 2004 NHIS Survey Description Document.
The names of the race and Hispanic origin variables have changed over the years whenever some change was made to either the data collection procedures or the editing and/or imputation procedures. In 1999, the names of the race and Hispanic origin variables were changed to indicate the full implementation of the revised OMB race standards in the NHIS. In 2000, they were changed again to indicate the implementation of hot-deck imputation procedures. In 2003, the names of the race variables were changed to indicate the editing out of "Other race" responses.
We are unable to release detailed data on some of the smaller racial and ethnic population groups on our public use data files because of confidentiality and disclosure regulations.
Researchers who wish to access confidential NHIS data on race and Hispanic origin may submit a research proposal to use the NCHS Research Data Center. Procedures for submitting the proposal and other important information can be found here: NCHS Research Data Center.
All information on the race and Hispanic origin data for a given year of the redesigned NHIS can be found in Appendix II of the Survey Description Document for that data year. The information for the 2005 NHIS file release can be found in the 2005 Survey Description Document.
Yes. The NHIS sample is redesigned about every 10 years, following the Decennial Census. The next redesign will cover the 2006-2014 survey years, and will include, for the first time, an oversample of Asian households. Similar to past sample redesigns that have included oversampling of black and Hispanic households, this sample redesign will help to improve the precision of estimates for Asian respondents to the NHIS.
You may contact:
National Center for Health Statistics
Division of Health Interview Statistics
3311 Toledo Road, Room 2333
Hyattsville, MD 20782