Nearest Neighbour Matching

[QiYee Diu]

Dear all, I am not familiar with Stata, so I would like to ask you guys about the nearest-neighbour matching. These are the codes I found in the Internet, is there anything I need to correct? *Outcome variable: GreenIssuer *Treatment: GreenBond *Independent variable: Total Assets PBratio RNDTA import excel "G:\CSR\DATA\Issuer 2.0.xlsx", sheet("exclude underwriter") firstrow estimate pscore// pscore GreenBond Total Assets PBratio RNDTA, pscore(mypscore) blockid(myblock) comsup logit psgraph, treated(GreenBond) pscore(mypscore) using nnm(ATET)// attnd GreenIssuer GreenBond, pscore(mypscore) comsup dots detail looking for the matched pairs after using nnm// gen id = _n sort id teffects psmatch (GreenIssuer) (GreenBond Total Assets PBratio RNDTA), gen(match20) atet nn(1) generate GreenIssuer_1 = GreenIssuer if GreenBond==1 generate GreenIssuer_0 = GreenIssuer if GreenBond==0 replace GreenIssuer_1 = GreenIssuer[match201[_n]] if GreenBond==0 replace GreenIssuer_0 = GreenIssuer[match201[_n]] if GreenBond==1 gen id_of_first_match = id[match201[_n]] list id* GreenBond match201 GreenIssuer* if inlist(id,1,481), noobs ab(20) list id* GreenBond match201 GreenIssuer* if inlist(id,1-481), noobs ab(20) list id* GreenBond match201 GreenIssuer* if inlist(id,1-481) I have some questions as well. 1. After I got the pscore, it showed "the inferior bound, the number of treated and the number of controls for each block". The total number is 371 (343 controls + 28 treated). Are 371 the matched pairs? 2. And then I tried to use Nearest Neighbour Matching, and I got these results: n. treatment (28) n. contr. (46) ATT (0.179) Std. Err, (0.062) t (2.867). Are the results mean 28 treated match with 46 controls (total: 74)? 3. I would like to try to use nnm with replacement by using the code "attnd", but I failed. 4. As I wanted to know the exact matched pairs, so I used the commands I showed above. However, the total matched pairs are 436, it's different from either 371 or 74.