I'm not aware of any such list, though it sounds like something AAPOR should
develop. At the beginning of a project, we generally create a set of coding
rules to make sure variables across datasets are consistent with regard to
values, size, labels, etc. For missing, we generally use system missing to
represent NIU, as that is the only true legitimate missing, and then .R for
refused, .D for Don't Know and whatever other neumonics we can think of. Of
course, that's in English and speakers of other languages might think
otherwise.
Bryan Sayer
Statistician, SSS Inc.
-----Original Message-----
From: Joseph Coveney
To: Statalist
Sent: 7/8/03 10:12 PM
Subject: Re: st: STATA equivalent to SAS subsetting "else if" statements?
Bryan Sayer, responding to a question from Nick Cox, posted:
.N is a special missing, to distinguish NIU, from don't know, refused,
invalid, etc.
------------------------------------------------------------------------
-------
Is there a convention for special missing values to distinguish the
reason for
missing values, such as .N reserved for not-in-universe (NIU)?
If so, is there a reference for an agreed-upon list of special missing
values
of this nature? In a quick Google search, I could only find a limited
list in
a couple of documents by Stephen Blumberg from the U.S. Centers for
Disease
Control. If there's a formal convention among survey researchers, or
even a
widespread informal standard, I'd rather use it than something ad hoc.
Joseph Coveney
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
