Dear Keith,
This is indeed very helpful, many thanks.
Best,
Shige
On Mon, Jan 26, 2009 at 7:52 AM, Keith Schnakenberg <
keith.schnakenberg(a)gmail.com> wrote:
This should help.
http://gking.harvard.edu/files/cem.pdf (Section 6.2)
On Jan 26, 2009, at 7:42 AM, Shige Song wrote:
Dear All,
I am trying to estimate a model using Zelig on matched data (using
MatchIt). Some of the variables have large amount of missing values. In this
situation, what will be the ideal way to proceed? Imputing first then
matching and then modeling?
I would appreciate any suggestions and examples. Many thanks.
Best,
Shige