Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Re: Problem with combining sibling data


From   daniel klein <klein.daniel.81@gmail.com>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   Re: st: Re: Problem with combining sibling data
Date   Mon, 13 May 2013 15:01:58 +0200

Sorry for the follow up post, but to elaborate on my statement

I do not underrstand why you want to only exclude -2 from the
calculations, as you already correctly noted, that we cannot "be sure
of what [any of the negative values] mean", in terms of the "true"
(underlying) value.

What Annabel might be tempted to do here (and this can be done with
-egen-'s -rowtotal()- after defineing missing vales) is to add up all
valid values in the four variables. Thus, all observations that have
at least one valid value on any of the fur original variables will
have have a "valid" value on the newly created variable holding the
number of siblings.

Now my problem with this approach is, that someone who ticks e.g. "one
older brother in the same house" and has missing values (i.e. -1 or
-2) on all other variables, will be treated as if he or she has a
total of one sibling. This migth be true -- but we cannot know. If any
of the missing values in the other three variables "maskes" one or
more other sibling, our results will be wrong. Therefore I strongly
recomment to exclude all _cases_ from the calculation if there is one
or more missing value in the original variable. Ther are (more or less
complex) ways of dealing with missing values later on.

Best
Daniel

--

I will try and answer as good as I can. If -1 and -2 represent missing
values of some kind, it seems natural to exclude those from the
calculation (that is before or while adding the variables up). You can
do so using -mvdecode- (see my first reply), -recode-, -replace-, or
some technique as demonstaretd in Nick's reply.

I do not underrstand why you want to only exclude -2 from the
calculations, as you already correctly noted, that we cannot "be sure
of what [any of the negative values] mean", in terms of the "true"
(underlying) value.

Best
Daniel
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index