MORTWT is a variable with eight columns and no implied decimal places.
MORTWT reports the sampling weight that adjusts for ineligible respondents in mortality analysis based on linking NHIS sample members during survey years 1986 to 2014 to the National Death Index (NDI).
Because adult survey participants who provided insufficient data for linking may differ from those who provided enough data for linking, ignoring the ineligible adult respondents could lead to biased mortality analysis. To correct this bias, NCHS developed the eligibility-adjusted weight in MORTWT (available for 1986-2014) for use with variables from the NHIS person files and MORTWTSA (available for 1997-2014) for use with variables from the sample adult files. NCHS cautions that the MORTWT, rather than the standard person weight (PERWEIGHT), should be used when analyzing mortality variables in conjunction with variables from the person files. Similarly, MORTWTSA should be used when analyzing mortality variables in conjunction with variables from the sample adult files.
Only persons who were age 18+ at the time of interview and who provided sufficient data were eligible for linking to the NDI and eligible for inclusion in the linked public use mortality files (indicated by IPUMS NHIS code 1 in MORTELIG). Linked mortality data for children are not available for public use; researchers interested in access to such data, or in linked mortality data with additional detail, should review the technical documentation for linked restricted access mortality files at NHIS Linked Mortality Files.
For additional information, see the variable description for MORTELIG, the IPUMS NHIS User Note on the use of sampling weights, and the related technical documents from NCHS, including the material on the creation and analysis of the NHIS Linked Mortality Files and the NHIS Analytic Guidelines (2010) at NHIS Linked Mortality Files.
For other mortality variables included in the IPUMS NHIS, refer to the MORTSTAT variable description.
With the exceptions of 1986 and 1992, MORTWT is completely comparable over time. In 1986, the NCHS did not adjust the original NHIS person weight to control for selection bias, so in analyses using the 1986 mortality data, analysts should use person weight (PERWEIGHT) instead of MORTWT. MORTWT values in 1986 contain the weights from the original file, which, per correspondence with NCHS data linkage staff, are incorrect. In 1992, MORTWT values include the Hispanic oversample. For more information about how the Hispanic oversample may affect analyses that pool 1991 and 1992, please refer to this IPUMS NHIS user note.
- 1986: All persons age 18+.
- 1987-2014: All persons age 18+ eligible for mortality follow-up.
- 1986-2014 : MORTWT