Hi Statalist,
I am matching cities from cross-sectional dataset based on a few matching variables. I am trying to include a caliper option in the teffects psmatch command:
The caliper of 0.04 is equal to 0.2 times the standard deviation of the propensity scores, as per usual guidelines.
Unfortunately, an error message comes up:
"5 observations have no propensity-score matches within caliper .04; they are identified in the osample() variable"
Just excluding the osample()==1 variables and re-estimating the matching does not work, as the sample is different and I get a similar new error message.
I would like to find a workaround that is similar to the caliper option for psmatch2, where observations are automatically dropped for which no match within the caliper distance can be found. I have seen several threads (such as Link) with similar problems, but I could not find a sufficient answer as I get the same error again and again. Does anyone have a solution for this problem?
Thank you!
Tillmann
I am matching cities from cross-sectional dataset based on a few matching variables. I am trying to include a caliper option in the teffects psmatch command:
Code:
teffects psmatch (sb_date) (lending_1500 ln_pop_1816 c_latitude c_longitude $controls1800city), atet vce(robust) caliper(0.04) osample(outofrange)
Unfortunately, an error message comes up:
"5 observations have no propensity-score matches within caliper .04; they are identified in the osample() variable"
Just excluding the osample()==1 variables and re-estimating the matching does not work, as the sample is different and I get a similar new error message.
I would like to find a workaround that is similar to the caliper option for psmatch2, where observations are automatically dropped for which no match within the caliper distance can be found. I have seen several threads (such as Link) with similar problems, but I could not find a sufficient answer as I get the same error again and again. Does anyone have a solution for this problem?
Thank you!
Tillmann
