I'm still not sure if I'm understanding your question, but what about just recoding the weights after you match? You can set it to 0 if the industry variable equals zero for example.
Kosuke Imai
Department of Politics, Princeton University
http://imai.princeton.edu
Sent from my iPad
On Nov 11, 2011, at 3:50 PM, Daniel Armanios <daniel.armanios(a)gmail.com> wrote:
> Dear Kosuke:
>
> My question was with regards to the weight scores given by the MatchIt function. I would like the weights to only be 1 for those groups when variable is 1 together and NOT also when variable is 0. Is that possible? I wanted to avoid subsetting as this data is part of a larger set and wanted to subset based on the weight scores so I can then only used those matched samples into regression analyses such as lm, glm, etc directly.
>
> If I have to go to subsetting because this is not possible by MatchIt, I will but wanted to avoid it if at all possible.
>
> Thanks for your patience and help with my inquiries!
>
> Daniel
>
> On Thu, Nov 10, 2011 at 2:27 PM, Kosuke Imai <kimai(a)princeton.edu> wrote:
> Hi Daniel,
>
> I am not sure if I understand your question. If the industry variable is binary, then exact matching on that variable should group the observations whose value of that variable is 1 together and should group the rest of the observations whose value is 0 together. Maybe, there is something strange that is happening in relationship with optmatch package. One thing you can do easily is to subset the data based on the industry variable and then perform the matching within each subset.
>
> Kosuke
>
> Department of Politics
> Princeton University
> http://imai.princeton.edu
>
>
> On Nov 9, 2011, at 1:54 PM, Daniel Armanios wrote:
>
> > Dear Dr. Imai:
> >
> > My name is Daniel Armanios, and I am 3rd year PhD student using your R MatchIt Package to do exact matching for firms I am studying in China. The treatment I have is whether firm is in a science park or not and the matching I am doing is exact matching by industry (0/1) and founding year (year). The issue I am having is that when I am using MatchIt, the industry is being matched when the variable is 0, i.e. when the firm is NOT in that industry. I would like to only match when my industry variable =1 or when the firm is in that industry. Is there a way to restrict the condition on the indicator variable in such a manner in MatchIt so as to only match when the indicator variable is of a certain value, something like IndustryA==1?
> >
> > Thanks so much for your patience and help with this inquiry!
> >
> > Daniel
> >
> > --
> > Daniel Armanios
> > PhD Student, Management Science and Engineering
> > Stanford Technology Ventures Program
> > Huang Engineering Center
> > 475 Via Ortega
> > Stanford University
> > Stanford, California 94305-4121
> > Email: armanios(a)stanford.edu
> > Phone: 404-316-0041 (mobile)
> >
> > (Mailing Address: 51 Dudley Lane, Apt. #330, Stanford, CA 94305)
>
>
>
>
>
> --
> Daniel Armanios
> PhD Student, Management Science and Engineering
> Stanford Technology Ventures Program
> Huang Engineering Center
> 475 Via Ortega
> Stanford University
> Stanford, California 94305-4121
> Email: armanios(a)stanford.edu
> Phone: 404-316-0041 (mobile)
>
> (Mailing Address: 51 Dudley Lane, Apt. #330, Stanford, CA 94305)
Hi Daniel,
I am not sure if I understand your question. If the industry variable is binary, then exact matching on that variable should group the observations whose value of that variable is 1 together and should group the rest of the observations whose value is 0 together. Maybe, there is something strange that is happening in relationship with optmatch package. One thing you can do easily is to subset the data based on the industry variable and then perform the matching within each subset.
Kosuke
Department of Politics
Princeton University
http://imai.princeton.edu
On Nov 9, 2011, at 1:54 PM, Daniel Armanios wrote:
> Dear Dr. Imai:
>
> My name is Daniel Armanios, and I am 3rd year PhD student using your R MatchIt Package to do exact matching for firms I am studying in China. The treatment I have is whether firm is in a science park or not and the matching I am doing is exact matching by industry (0/1) and founding year (year). The issue I am having is that when I am using MatchIt, the industry is being matched when the variable is 0, i.e. when the firm is NOT in that industry. I would like to only match when my industry variable =1 or when the firm is in that industry. Is there a way to restrict the condition on the indicator variable in such a manner in MatchIt so as to only match when the indicator variable is of a certain value, something like IndustryA==1?
>
> Thanks so much for your patience and help with this inquiry!
>
> Daniel
>
> --
> Daniel Armanios
> PhD Student, Management Science and Engineering
> Stanford Technology Ventures Program
> Huang Engineering Center
> 475 Via Ortega
> Stanford University
> Stanford, California 94305-4121
> Email: armanios(a)stanford.edu
> Phone: 404-316-0041 (mobile)
>
> (Mailing Address: 51 Dudley Lane, Apt. #330, Stanford, CA 94305)
--
MatchIt mailing list served by Harvard-MIT Data Center
List Address: matchit(a)lists.gking.harvard.edu
Subscribe/Unsubscribe: http://lists.gking.harvard.edu/?info=matchit
MatchIt Software and Documentation: http://gking.harvard.edu/matchit/
Browse/Search List Archive: http://lists.hmdc.harvard.edu/lists/matchit/
