Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Combine Listwise Deletion and Multiple Imputation?
From
Stas Kolenikov <[email protected]>
To
[email protected]
Subject
Re: st: Combine Listwise Deletion and Multiple Imputation?
Date
Thu, 11 Aug 2011 10:23:35 -0400
On Thu, Aug 11, 2011 at 9:43 AM, Andrea Bennett <[email protected]> wrote:
> I am dealing with a data set (field experiment) containing missing values. I wonder, is it ok to combine both methods, that is "Listwise Deletion" and "Multiple Imputation" in order to deal with missing values?
>
> I've been reading Paul D. Allison's "Missing Data" and cannot find a clear answer there. I want combine both methods because the loss in observations is small except for one particular variable where I have lots of "don't know" answers (question: highest completed education of parents). I would then only apply MI for one variable where Listwise Deletion would result in quite a severe loss.
>
> Does anyone of you, who is more familiar with MI, disagree in general? Do you recommend reading a specific book, potentially tailored to Stata users?
Hm. "I would like to be on a diet to control my calory intake, and
will be eating turkey with veggies instead of burgers and fries, but I
am going to substitute milk with whipped cream".
If you have access to MI, why don't MI everything? I could understand
combining MI with listwise deletion for a methodological study of how
things perform (and then I would still MI everything and then randomly
delete observations, real or imputed), but frankly I don't see any
point for this awkward procedure in any real data analysis setting.
--
Stas Kolenikov, also found at http://stas.kolenikov.name
Small print: I use this email account for mailing lists only.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/