https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908744#M45107 All resolved! Tue, 19 Dec 2023 00:43:00 GMT Callam1 2023-12-19T00:43:00Z https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908532#M45095 <P>Hi</P><P>I am struggling to understand the result of my matching with replacement with k=5. The output is showing that 99% of the 30,000 treated observations are successfully matched and 88% of the 28,000 control observations are matched. But when I open the resulting dataset (matched_data in the code below), although all the treatment observations have a matching ID,&nbsp; 50% of control observations have a missing matching ID. It appears that&nbsp; the information provided in the output is not accurate as it is counting as matched the control observations with a missing matching ID. The matching appears very good based on the Standardised difference, but can I trust it? In the matched sample there is no matching set where a&nbsp; treatment observation was matched to more than 1 control. I do have several treated observations matched to one control observation (as expected given the matching with replacement). Is it that SAS cannot process accurately a matching with replacement with k&gt;1 when the control sample is smaller than the treatment sample ? If I select k=1 the output is accurately reflecting the matching and the matched dataset has no missing matching ID. However the matching performance&nbsp; is not as good, despite the fact that the number of matched treatment and control observations with a matching ID is the same as in k=5. Any views?</P><P>Many thanks in advance!</P><P>&nbsp;</P><P>ods graphics on;<BR />proc psmatch data= dataset1 region=cs;<BR />class is_treated binary_covariates categ_covariates;</P><P>psmodel is_treated(Treated='1') =categ_covariates continuous_covariates interaction_effect;<BR />match method=replace(k=5) stat=lps caliper=0.2;<BR />assess lps var=(binary_covariate continuous_covariates)<BR />) / plots=(BoxPlot StdDiff barchart);<BR />output out(obs=match)=matched_data matchid=MID ;<BR />run;</P> Mon, 18 Dec 2023 00:17:35 GMT https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908532#M45095 Callam1 2023-12-18T00:17:35Z https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908744#M45107 All resolved! Tue, 19 Dec 2023 00:43:00 GMT https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908744#M45107 Callam1 2023-12-19T00:43:00Z https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908851#M45118 <P>How was it resolved? Kind of went from an interesting problem to "nothing to see here."</P> <P>&nbsp;</P> <P>SteveDenham</P> Tue, 19 Dec 2023 17:13:39 GMT https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908851#M45118 SteveDenham 2023-12-19T17:13:39Z https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908959#M45119 Hi, basically the problem was that in the last line of the code I only had the name for one machid. When k=n, you need n names. So it should be machid=(mid1 mid2 mid3 …midn).<BR />In this way there are no more missing machid in the matched dataset.<BR /><BR /> Wed, 20 Dec 2023 04:26:16 GMT https://communities.sas.com/t5/Statistical-Procedures/proc-psmatch-with-replacement-and-k-gt-1-when-control-sample/m-p/908959#M45119 Callam1 2023-12-20T04:26:16Z