[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
RE: st: RE: the impute command
At 05:40 PM 5/19/2004 +0100, Nick Cox wrote:
It works in the case of the example I gave, where -logit- is used. The
-round- function could potentially be problematic if you were using the
-impute- command, because you could conceivably have an imputed value of,
say, 1.51, which would get rounded to two. A more all-purpose recode would be
. replace y = round(y)
is an alternative to -recode- here.
recode y (min/.50 = 0)(.50/max = 1).
The manual makes the claim that impute "is often much better than deleting
cases with any missing data." I agree that that statement is very
debatable. Paul Allison in his book "Missing Data" makes a pretty good
argument that listwise deletion is often better than some of the
traditional alternatives that have been used, including the sort of thing
done by impute. He also discusses more advanced techniques that can be better.
2. It is not at all clear that -impute-
is overall a very good way of doing it.
I agree. I'd like to see Stata implement some of the more powerful methods
that are out there. I'd also like to see a portion of the Users Guide
devoted to discussion of ways of handling missing data.
3. Various users have implemented some
better ways to do it.
4. Official Stata's support for good
ways of doing it is not that great, right
now. Information on what StataCorp will
add in the near future follows this colon:
(no, sorry, it's missing; impute your own
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
WWW (personal): http://www.nd.edu/~rwilliam
WWW (department): http://www.nd.edu/~soc
* For searches and help try: