SAMPWEIGHT is a 12-digit numeric variable with three implied decimals. That is, values of 012345678912 should be interpreted as 12345678.912. The command files delivered with IPUMS extracts automatically divide SAMPWEIGHT by 1,000, so no further adjustment is needed.
SAMPWEIGHT is an IPUMS NHIS-constructed variable that represents, with a few exceptions, the random selection of a sample person in the household to complete a supplement survey. (The exceptions to this generalization are the 1983 health insurance supplement and 1993 health insurance, access to care, and family resources supplements, which were asked of all persons in the second half of the year.)
For survey years 1997 forward, SAMPWEIGHT is based on the Final Annual Sample Adult and Sample Child Weights in the original NHIS public use files. This weight should be used with variables taken from the sample adult or sample child supplements for 1997 on (as indicated by the supplement title in the "Source" section toward the top of each variable description).
SAMPWEIGHT also contains the sampling weights for a subset of the pre-1997 supplements, for which the definition of a sample person, and the rules governing the selection of sample persons, are less consistent. The universe statements for SAMPWEIGHT document who was included in the supplements requiring use of this weight on a year-by-year basis. SAMPWEIGHT consistently represents the inverse probability of selection into a supplement, adjusted for non-response with additional post-stratification adjustments for age, race/ethnicity, and sex using the Census Bureau's population control totals.
Users should review the documentation for extracted variables--most notably the "Weights" section toward the top of each variable description--to ascertain which weight is the appropriate choice for a given survey year. See the User Notes on the use of sampling weights for additional information.
The comparability of SAMPWEIGHT is affected by changes including the 2019 redesign. SAMPWEIGHT is comparable for all years from 1997 forward, with the stipulation that each year's values are adjusted using contemporaneous Census data. The definition of a sample person is less consistent for the pre-1997 supplements that require use of SAMPWEIGHT. Users should note the universe statements for SAMPWEIGHT to discern whom the related supplements represent in each year.
The NHIS questionnaire was substantially redesigned in 2019 to introduce a different data collection structure and new content. For more information on changes in terminology, universes, and data collection methods beginning in 2019, please see the user note. Additionally, NCHS implemented design changes in 2020 to accommodate challenges presented by the COVID-19 pandemic. The original 2020 sample was reduced in quarters 3 and 4 to free up resources to administer the survey to a group of respondents who already completed the 2019 survey. Please see the user note on COVID-related changes to the NHIS for more information.
- 1974: Sample persons age 17+.
- 1975-1977: Sample persons age 20+.
- 1979-1980: Sample persons age 17+ and all persons age 17+ needing help with basic daily functions.
- 1981: Sample persons under age 18.
- 1983; 1993: All persons in Qt 3 and Qt 4.
- 1984: All persons age 65+ and sample person age 55-64.
- 1985; 1987; 1989; 1990: Sample persons age 18+.
- 1988; 1991: Sample persons age 18+ and sample persons under age 18.
- 1992: Sample persons 18+ in Qt 1, Qt 2, and 2 weeks of Qt 3.
- 1997-2018: Sample adults age 18+ and sample children under age 18.
- 2019-2022: Sample adults age 18+ and sample children age 0-17.
- 1974-1977, 1979-1981, 1983-1985, 1987-1993, 1997-2022
- 2019-2022 : SAMPWEIGHT