Dear All,

I am using propensity score matching to match observations in current year with observations in prior year based on some characteristic variables. I searched online and found that there is a [in range] option for both pstest and psmatch2.

http://repec.org/bocode/p/psmatch2.html
http://fmwww.bc.edu/repec/bocode/p/pstest.html

But I could not find any examples using the [in range] option.

If I simply put a number to indicate the starting observation, such as

psmatch2 X1 X2 X3 in 1000, out(Y) common

X1 is the main explanatory variable, X2 and X3 are control variables, Y is the outcome variable

Stata shows "insufficient observations"

If I put two numbers or "-" between the numbers

psmatch2 X1 X2 X3 in 1000 2000, out(Y) common

Stata shows "invalid '2000' "

psmatch2 X1 X2 X3 in 1000-2000, out(Y) common

Stata shows "'1000-2000' invalid observation number"

Could anyone teach me how to correctly specify the range in propensity score matching? Many thanks!