{smcl}
{* 03dec2008}{...}
help for {cmd:optselect}
{hline}

{title:Command optselect}

{pstd}
Calculates the optimal cutoff point for the propensity score (to calculate Optimal Subpopulation Average Treatment Effect, OSATE) 
based on the propensity score distribution, as proposed by Crump, R.K., V.J. Hotz, G.W. Imbens and O.A. Mitnik (2008) 
"Dealing with limited overlap in estimation of average treatment effects", Biometrika. The cutoff is calculated assuming homoskedasticity
and a weight function that gives equal weight to all observations.


{title:Syntax}

{p 8 17 2}
{cmd:optselect}
{varname}
{ifin} 


{title:Description}

{pstd}
Provide a variable for the propensity score in {varname}. The results are saved in r() (see below). The cutoff is saved in r(bound). 
If the optimal cutoff is 0, then r(maxsup) is set to 1, otherwise is set to 0. If the optimal cutoff cannot be calculated 
r(maxiter) is set to 1, otherwise is set to 0. In that case, the cutoff is set to 0.10. 


{title:Example}

{phang}{cmd:. optselect pscore}


{title:Saved results}

{pstd}
{cmd:optselect} saves the following in {cmd:r()}:

{synoptset 20 tabbed}{...}
{p2col 5 20 24 2: Scalars}{p_end}
{synopt:{cmd:r(bound)}}Optimal cutoff for the propensity score{p_end}
{synopt:{cmd:r(maxsup)}}Indicator for optimal cutoff = 0{p_end}
{synopt:{cmd:r(maxiter)}}Indicator for maximum number of iterations attained (cutoff not calculated, set to 0.10){p_end}
{p2colreset}{...}