For PSM, I use "pscore" to gain the propensity score and to learn about the balance. the help document " pscore treatment varlist [weight] [if exp] [in range] , pscore(newvar)
[blockid(newvar) detail logit comsup level(#) numblo(#) ]" the level(#) allows to set the significance level of the tests of the balancing property. The default is 0.01
my question is what is better to set the level, the default 0.01 or 0.001 as many examples in PSM handbook? because when I run the pscore with the same covariants, the balance can be achieved in level(0.001) but did not in level(0.01), which means i have to change the set of covariants and lead to a different estimated result.
Thanks!
[blockid(newvar) detail logit comsup level(#) numblo(#) ]" the level(#) allows to set the significance level of the tests of the balancing property. The default is 0.01
my question is what is better to set the level, the default 0.01 or 0.001 as many examples in PSM handbook? because when I run the pscore with the same covariants, the balance can be achieved in level(0.001) but did not in level(0.01), which means i have to change the set of covariants and lead to a different estimated result.
Thanks!
Comment