Hi,
I am using Coursened Exact Matching to improve balance in my sample. The summary matching results after applying CEM are that 11,128 pixels have been matched:
Sample sizes:
Control Treated
All 19741 8283
Matched 6728 4400
Unmatched 13013 3883
Discarded 0 0
Firstly, is it possible to specify 1:1 matching with CEM? I want to only retain matched pairs, which requires the same amount of pixels in my treatment & control groups.
I am also having trouble getting a unique identifier for each matched pair so that I can calculate the difference between the treatment and control. I am using the code:
pair: Produces a paired sample out of a CEM match solution
When I execute the below command, it says there are 0 matches.
> # we want a set of paired units
> pairsample <- pair(match.out_cem, data)
Total number of units paired in CEM strata: 0
Total number of units matched: 0
Can anyone help or point me in the right direction?
Thanks, Tori
Victoria Graham |
MPhil., PhD candidate
Macquarie University NSW 2109 Australia
T:
+61 2 9850 6298
| E: victoria.graham@mq.edu.au
Twitter:@ToriFGraham