How do I know if information is self or proxy reported in IHIS?

I use person-level IHIS data since 1997. I use only persons who are at least 18 years old. I have some questions about the IHIS survey.

  1. So, one row of the (rectangular) data will have information about a certain person, call him John. Is there any variable that can tell me if it is John who responded questions about himself or if it was somebody else answering for him?

  2. In my analysis I also keep only observations with non-missing values on some variables which are in the sample adult files. So, if John’s observation remains in my dataset, it is because he was also sampled to be part of the adult file. The NHIS documentation (e.g. p.43 here: http://www.cdc.gov/nchs/data/nhis/srv…) claims that: “sample adults generally respond for themselves, although in a small number of cases, proxy responses are allowed if the selected adult had a physical or mental condition prohibiting him/her from responding.” Ignoring this small number of cases, then this means that if John is in the sample adult file, then he answered all the questions about himself: am I missing something here?

ps: thanks to the IHIS team at the MPC. The website is an excellent resource for researchers and teachers alike.

Since 2001, the variable PROXYSA is available to indicate whether a sample adult (see ASTATFLG) required a proxy. Consistent with the NHIS documentation you cite, at least 95% of sample adults in a given year are identified as not using a proxy (PROXYSA=1). The remaining cases are either proxy responses (PROXYSA=2) or have an unknown proxy status (PROXYSA=9). We do know that most of the unknown cases are partially completed sample adult interviews that did not use a proxy.

Hope this helps.