Veterans Self-Administered Questionnaire (VSAQ) Overview

Overview of the VSAQ

The Veteran Self-Administered Questionnaire (VSAQ), a self-administered paper-and-pencil questionnaire, was fielded in 2018 (for Panel 22, Round 5 and Panel 23, Round 3) and in 2019 (for Panel 23 Round 3 and Panel 24 Round 1). The VSAQ collects data on veteran health and health care services at VA facilities and outside VA facilities. See our Surveys page for the VSAQ questionnaires.

Households received a VSAQ based on their response in the Renumeration (RE) section of the CAPI instrument whether there was someone on active duty in the past but not now. The VSAQ is only available in the English language. The variable VETELIG indicates the person's eligibility status for the VSAQ.

The completion of the VSAQ is determined by the answer provided by ARMFEV. For confidentiality reasons, AHRQ restricted the variables that indicate when the veteran was in active duty, whether or not they received a disability rating, was a POW, and their discharge information from the public data available through IPUMS. Users interested in restricted MEPS data should consult with AHRQ to apply for use of restricted MEPS data.

Key Variables Description
VETELIG VETELIG reports the person's eligibility status for the VSAQ.
ARMFEV ARMFEV reports whether the person has served on active duty in U.S. Armed Forces, Reserves, or National Guard in the past, but does not currently. This variable establishes if the person is a veteran and sets the universe for the remainder of the VSAQ variables. ARMFEV's variable description contains the full list of all VSAQ variables available through IPUMS.
VETWEIGHT VETWEIGHT is the person-level weight for VSAQ.

Information on VETWEIGHT

A person-level weight, VETWEIGHT, was developed by AHRQ for use with the VSAQ data. This weight was assigned to each person with a final person-level weight (PERWEIGHT>0) who also completed the VSAQ and self-reported as being a veteran. The weighting process paralleled the interviewing process of veterans. To determine veterans eligible for the study, the interviewing process called for first having the RU respondent identify those members of an RU who he or she thought were veterans. The criteria for identifying a veteran were the same used for the American Community Survey (ACS). Note that veterans are identified in the ACS by a household respondent and are not asked to confirm whether this identification is correct. All RU members so identified as a veteran were asked to complete the VSAQ. The first question of the VSAQ asked the respondent to self-identify as a veteran (ARMFEV). A small percentage of the VSAQ respondents indicated that they were not veterans and thus were ineligible to complete the remainder of the survey.

The weighting process was established as follows. Those identified as veterans by the RU respondent were assigned as an initial weight the final person-level weight value PERWT18F. These weights were then raked to control totals based on 2018 or 2019 ACS data, respectively, on veterans to achieve consistency with these estimates. A nonresponse adjustment was then undertaken to account for nonresponse to the VSAQ as well as to account for the fact that some of the individuals identified as veterans indicated that they were not.

In 2018, veterans were oversampled for Panel 23 but not Panel 22. 1,157 people were assigned a VSAQ weight (VETWEIGHT>0) in 2018. The sum of the VSAQ weights is 16,563,024, an estimate of the adult population self-reporting as being veterans. The sum of these weights is, by design, smaller than the estimated number of veterans based on ACS figures. This is consistent with the fact that some VSAQ respondents self-reported as 'not a veteran' while this opportunity is not provided in the ACS. According to AHRQ, the 2018 ACS estimate of veterans was about 17.96 million and the corresponding MEPS estimate of veterans based on the reports of the MEPS RU respondent was about 17.72 million.

In 2019, veterans were oversampled for Panel 23 but not Panel 24. 1,131 people were assigned a VSAQ weight (VETWEIGHT>0) in 2019. The sum of the VSAQ weights is 16,158,443, an estimate of the adult population self-reporting as being veterans. The sum of these weights is, by design, smaller than the estimated number of veterans based on ACS figures. This is consistent with the fact that some VSAQ respondents self-reported as 'not a veteran' while this opportunity is not provided in the ACS. According to AHRQ, the 2019 ACS estimate of veterans was about 17.69 million and the corresponding MEPS estimate of veterans based on the reports of the MEPS RU respondent was about 18.18 million.

Source for all information on this page were the 2018 FYC and 2019 FYC documentation.