IPUMS Online Data Analysis System
This system uses high-speed tabulation software developed at UC-Berkeley's Computer-assisted Survey Methods Program.
Users should reference our resources in the MEPS user guide, the Considerations for SDA Use, and the SDA Data Training Exercise for technical guidance on how to best utilize MEPS and SDA.
MEPS Combined SDA Dataset | Combined multi-year MEPS SDA dataset allows for longitudinal analysis |
---|
1996-2022: All MEPS Combined SDA Dataset |
MEPS Single Year SDA Datasets | Single year MEPS SDA datasets allow for faster data analysis |
---|
-
2020s 2022 2021 2020 -
2010s 2019 2018 2017 2016 2015 2014 2013 2012 2011 2010 -
2000s 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 -
1990s 1999 1998 1997 1996
Considerations when using these SDA Datasets:
- Variable Availability: These SDA datasets contain only person-level variables derived from the Full Year Consolidated files (i.e., those listed under the "Annual" variable drop down menu).
- Survey Design Variables: STRATA and CLUSTER (PSU) variables account for the complex sample design of MEPS and are used to obtain appropriate variance estimates (standard errors) when computing annual or pooled estimates. See our user note on variance estimation for more information.
- The Combined MEPS SDA Dataset uses the survey design variables STRATAPLD and PSUPLD for pooled estimates adjusted for the complex sample design of the MEPS (default setting).
- The Single-Year MEPS SDA Datasets uses the survey design variables STRATANN and PSUANN for annual estimates adjusted for the complex sample design of the MEPS (default setting).
- Weights: These SDA datasets contain multiple weights. The sampling weight users should use for most analyses in all SDA datasets is PERWEIGHT.
- Other weights available for these datasets are SAQWEIGHT and DIABWEIGHT, designated for use in analyses of variables from the Self-Administered Questionnaire and the Diabetes Care Supplement, respectively. Users should reference our guidance on the use of sampling weights with IPUMS MEPS for more information.
- NOTE: When generating weighted estimates using SDA, the option to display unweighted case counts will result in unweighted totals that do not match the total sample size reported on our sample size page because cases with a sampling weight value of zero are excluded.
- Notes on using Combined MEPS SDA Dataset:
- Limit analyses using the YEAR variable if you do not intend to pool all years of MEPS data for your analysis.
- When using pooled data, certain SDA estimates (i.e., population counts) will be incorrect as it does not automatically divide the weight by the number of samples. For estimates of the average population count for a time period range greater than a year, we advise downloading an extract and adjusting sampling weights manually.
- SDA User Guide: This SDA User Guide was developed for the previous version of SDA but is still accurate for use with the current SDA version and Health Surveys data. A new version is currently under development.