{smcl} {* 11oct2018}{...} {cmd:help sensimatch} {hline} {title:Title} {p2colset 5 18 20 2}{...} {p2col:{hi:sensimatch}{hline 1}}Data-driven sensitivity analysis for Matching estimator{p2colreset}{...} {title:Syntax} {p 8 17 2} {hi:sensimatch} {it:outcome} {it:treatment} [{it:varlist}] {ifin} {weight}{cmd:,} {cmd:mod}{cmd:(}{it:{help sensimatch##modeltype:modeltype}}{cmd:)} {cmd:sims}{cmd:(}{it:number}{cmd:)} [{cmd:fac}{cmd:(}{it:varlist_f}{cmd:)} {cmd:seed}{cmd:(}{it:number}{cmd:)} {cmd:save_sens}{cmd:(}{it:filename}{cmd:)} {cmd:vce}{cmd:(}{it:vcetype}{cmd:)} {cmd:gr_dep_var}{cmd:(}{it:text}{cmd:)} {cmd:gr1_save}{cmd:(}{it:filename}{cmd:)} {cmd:gr1_title}{cmd:(}{it:text}{cmd:)} {cmd:gr1_xtitle}{cmd:(}{it:text}{cmd:)} {cmd:gr1_ysize}{cmd:(}{it:size}{cmd:)} {cmd:gr1_xsize}{cmd:(}{it:size}{cmd:)} {cmd:gr2_save}{cmd:(}{it:filename}{cmd:)} {cmd:gr2_title}{cmd:(}{it:text}{cmd:)} {cmd:gr2_xtitle}{cmd:(}{it:text}{cmd:)} {cmd:gr2_ysize}{cmd:(}{it:size}{cmd:)} {cmd:gr2_xsize}{cmd:(}{it:size}{cmd:)}] {pstd}{cmd:fweight}s, {cmd:pweight}s, {cmd:iweight}s are allowed only for model "reg"; see {help weight}. {title:Description} {pstd} {cmd:sensimatch} provides a sensitivity test for checking the robustness of the selection-on-observables assumption in treatment effect observational studies, both within a regression adjustment and a propensity-score matching approach. Rooted in the machine learning literature, this sensitivity analysis is based on a "leave-one-covariate-out" (LOCO) approach. This method recalls a bootstrap over different subsets of covariates, and simulates various estimation scenarios to be compared with the baseline results obtained by the analyst. The main output of {cmd:sensimatch} is graphical, thus providing the user with an easy-to-interpret robustness check of his/her study results. {title:Options} {bf:General options} {phang} {cmd:mod}{cmd:(}{it:{help sensimatch##modeltype:modeltype}}{cmd:)} specifies the model to be estimated, where {it:modeltype} must be one of the following models: "reg" or "match". It is always required to specify one model. {phang} {cmd:sims}{cmd:(}{it:number}{cmd:)} specifes the number of model simulations for each set of included covariates. {phang} {cmd:fac}{cmd:(}{it:varlist_f}{cmd:)} specifies that factor variables have to be included among the regressors. It is optional in each model. {phang} {cmd:seed}{cmd:(}{it:number}{cmd:)} specifies the random generation seed for each simulation. {phang} {cmd:save_sens}{cmd:(}{it:filename}{cmd:)} allows to save the variables used to generate the sensitivity graph. {phang} {cmd:vce}{cmd:(}{it:vcetype}{cmd:)}: allows to choose {it:vcetype} as either {it:robust}, or {it:cluster clustvar}. This option works only for model "reg". {phang} {cmd:gr_dep_var}{cmd:(}{it:text}{cmd:)} allows to customize the name of the dependent variable appearing in the sensitivity graph. {bf:Graph options} {phang} {it:Graph 1 - ATET} {phang} {cmd:gr1_save}{cmd:(}{it:filename}{cmd:)} allows to save the sensitivity graph for ATET by a user specifed filename. {phang} {cmd:gr1_title}{cmd:(}{it:text}{cmd:)} allows to customize the overall title of the sensitivity graph for ATET. {phang} {cmd:gr1_xtitle}{cmd:(}{it:text}{cmd:)} allows to customize the x-axis title of the sensitivity graph for ATET. {phang} {cmd:gr1_ysize}{cmd:(}{it:size}{cmd:)} allows to customize the size of the y-axis title of the sensitivity graph for ATET. {phang} {cmd:gr1_xsize}{cmd:(}{it:size}{cmd:)} allows to customize the size of the x-axis title of the sensitivity graph for ATET. {phang} {it:Graph 2 - T-Student} {phang} {cmd:gr2_save}{cmd:(}{it:filename}{cmd:)} allows to save the sensitivity graph for the T-Student by a user specifed filename. {phang} {cmd:gr2_title}{cmd:(}{it:text}{cmd:)} allows to customize the overall title of the sensitivity graph for T-Student. {phang} {cmd:gr2_xtitle}{cmd:(}{it:text}{cmd:)} allows to customize the x-axis title of the sensitivity graph for T-Student. {phang} {cmd:gr2_ysize}{cmd:(}{it:size}{cmd:)} allows to customize the size of the y-axis title of the sensitivity graph for T-StudentT. {phang} {cmd:gr2_xsize}{cmd:(}{it:size}{cmd:)} allows to customize the size of the x-axis title of the sensitivity graph for T-Student. {marker modeltype}{...} {synopthdr:modeltype_options} {synoptline} {syntab:Model} {p2coldent : {opt reg}}Regression estimated by ordinary least squares (OLS){p_end} {p2coldent : {opt match}}Propensity-score matching{p_end} {synoptline} {title:Remarks} {pstd} Please, before running this program, remember to have the most recent up-to-date version installed. {title:Examples} *---------------------------------------------------------------------------------------------------------- {inp:. webuse nlsw88 , clear} {inp:. global y "wage"} {inp:. global w "union"} {inp:. global xvars "age race married never_married grade south smsa c_city collgrad hours ttl_exp tenure"} {inp:. global factors "industry occupation"} *---------------------------------------------------------------------------------------------------------- {inp:. sensimatch $y $w $xvars if c_city==1 , mod(match) sims(5) vce(robust) ///} {inp:fac($factors) save_sens(data_sens) seed(1010) ///} {inp:gr_dep_var("Wage") ///} {inp:gr1_title("") ///} {inp:gr1_xtitle(Number of included covariates) ///} {inp:gr1_ytitle(ATET) ///} {inp:gr1_xsize(small) ///} {inp:gr1_ysize(small) ///} {inp:gr1_save(mygraph1) ///} {inp:gr2_title("") ///} {inp:gr2_xtitle(Number of included covariates) ///} {inp:gr2_ytitle(T-Student) ///} {inp:gr2_xsize(small) ///} {inp:gr2_ysize(vsmall) ///} {inp:gr2_save(mygraph2)} *---------------------------------------------------------------------------------------------------------- {title:Reference} {phang} Cerulli, G. 2015. {it: Econometric Evaluation of Socio-Economic Programs: Theory and Applications}. Springer, Berlin. {p_end} {phang} Nannicini, T. 2007. {it: Simulation-based sensitivity analysis for matching estimators}. The Stata Journal, 7, Number 3, pp. 334-350. {p_end} {phang} Rosenbaum, P. R. 1987. {it:Sensitivity analysis to certain permutation inferences in matched observational studies}. Journal of the Royal Statistical Society, Series B, 45, pp. 212-218. {p_end} {phang} Rosenbaum, P. R., and Rubin, D. B. 1983. {it:Assessing sensitivity to an unobserved binary covariate in an observational study with binary outcome}. Journal of the Royal Statistical Society, Series B, 45, pp. 212-218. {p_end} {title:Author} {phang}Giovanni Cerulli{p_end} {phang}IRCrES-CNR{p_end} {phang}Research Institute for Sustainable Economic Growth, National Research Council of Italy{p_end} {phang}E-mail: {browse "mailto:giovanni.cerulli@ircres.cnr.it":giovanni.cerulli@ircres.cnr.it}{p_end} {title:Also see} {psee} Online: {helpb sensatt}, {helpb rbounds} {p_end}