{smcl}
{* 28may2006}{...}
{hline}
help for {hi:sensatt}
{hline}
{title:A simulation-based sensitivity analysis for matching estimators}
{p 8 24}
{cmd:sensatt}
{it:outcome treatment} [{it:varlist}]
[{it:weight}]
[{cmd:if} {it:exp}]
[{cmd:in} {it:range}]
[ ,
{cmd:alg}{cmd:(}{it:att*}{cmd:)}
{cmd:p}{cmd:(}{it:varname}{cmd:)}
{cmd:p11}{cmd:(}{it:#}{cmd:)}
{cmd:p10}{cmd:(}{it:#}{cmd:)}
{cmd:p01}{cmd:(}{it:#}{cmd:)}
{cmd:p00}{cmd:(}{it:#}{cmd:)}
{cmdab:r:eps}{cmd:(}{it:#}{cmd:)}
{cmd:ycent}{cmd:(}{it:#}{cmd:)}
{cmd:se}{cmd:(}{it:se_type}{cmd:)}
{cmd:pscore}{cmd:(}{it:scorevar}{cmd:)}
{cmdab:logit }
{cmdab:index }
{cmdab:comsup}
{cmdab:boot:strap }
]
{p 8 8}{cmd:fweight}s, {cmd:iweight}s, and {cmd:pweight}s are allowed;
see help {help weights}.
{p 8 8}{cmd:sensatt} makes use of the commands for the propensity-score
matching estimation of average treatment effects written by
Becker and Ichino (2002): {cmd:attnd}, {cmd:attnw},
{cmd:attk}, {cmd:attr}. Before using {cmd:sensatt}, you should install them
and be familiar with their utilization. Download: {net sj 5-3 st0026_2}
{title:Description}
{p} {cmd:sensatt} implements the sensitivity analysis for matching estimators
proposed by Ichino, Mealli and Nannicini (2006).
The analysis builds on Rosenbaum and Rubin (1983a) and Rosenbaum (1987),
and simulates a potential binary confounder in order to assess the robustness of
the estimated treatment effects with respect to specific deviations from the
Conditional Independence Assumption (CIA).
{p} As a first step, the Average Treatment effect on the Treated (ATT) is estimated
by using one of the following propensity-score matching estimators: Nearest Neighbor ({cmd:attnd},
{cmd:attnw}); Radius ({cmd:attr}); Kernel ({cmd:attk}). The options that are common to these commands specify how
the baseline ATT is estimated (see Becker and Ichino, 2002).
{p} As a second step, a potential binary confounder (U) is simulated in the data, on the
basis of four parameters: pij (with i,j=0,1). Defining Y as the outcome (or as a binary transformation of the outcome
in the case of continuous outcomes) and T as the binary treatment, each simulation parameter pij represents
the probability that U=1 if T=i and Y=j.
The user can either specify the four parameters (with the options: {cmd:p11}{cmd:(}{it:#}{cmd:)}
{cmd:p10}{cmd:(}{it:#}{cmd:)}, {cmd:p01}{cmd:(}{it:#}{cmd:)}, {cmd:p00}{cmd:(}{it:#}{cmd:)})
or let the distribution of U mimic the distribution of some relevant covariate
(with the option: {cmd:p}{cmd:(}{it:varname}{cmd:)}).
{p} Finally, U is considered as any other covariate and is included in the set of matching variables
used to estimate the propensity score and the ATT. The imputation of U and the ATT estimation are
replicated many times (according to the option: {cmdab:r:eps}{cmd:(}{it:#}{cmd:)}), and a simulated
ATT is retrieved as an average of the ATTs over the distribution of U. This estimate is robust
to the specific failure of the CIA implied by the parameters pij. A comparison of the simulated ATT and the baseline ATT
tells us to what extent the latter is robust, with respect to the specific deviation from the CIA that we are assuming.
{p} In order to further emphasize the characteristics of the failure of the CIA implied by the simulated confounder
(i.e., by the chosen pij), the estimated effect of U on the selection into treatment - {it:selection effect} - and the
estimated effect of U on the outcome of untreated subjects - {it:outcome effect} - are also reported as odds ratios.
{title:Remarks}
{p} The treatment must be binary.
{p} When the outcome is not binary, the potential confounder U is simulated on the basis of a binary
transformation of the outcome: Y=1 if the outcome is above the mean, the 25, the 50 or the 75 centile, according
to the choice of the user (see option {cmdab:ycent}). The ATT, of course, is still estimated for the continuous outcome.
{title:Options that are specific to {cmd:sensatt}}
{p 0 4}
{cmdab:alg(}{it:att*}{cmd:)} specifies the name of the command (i.e., of the matching algorithm) that is
used in the ATT estimation. One of the following commands can be specified: {cmd:attnd}, {cmd:attnw},
{cmd:attk}, {cmd:attr}. The default is {cmd:attnd}.
{p 0 4}
{cmdab:p(}{it:varname}{cmd:)} indicates the binary variable which is used to simulate the confounder.
The parameters pij used to simulate U are set equal to the ones observed for {it:varname}.
Instead of selecting this option, the user can directly specify the parameters pij.
{p 0 4}
{cmd:p11}{cmd:(}{it:#}{cmd:)}, {cmd:p10}{cmd:(}{it:#}{cmd:)}, {cmd:p01}{cmd:(}{it:#}{cmd:)} and
{cmd:p00}{cmd:(}{it:#}{cmd:)} jointly specify the parameters pij used to simulate U in the data.
Since they are probabilities, they must be between zero and one. For each parameter, the default is zero.
{p 0 4}
{cmdab:r:eps}{cmd:(}{it:#}{cmd:)} specifies the number of iterations, i.e., how many times
the simulation of U and the ATT estimation are replicated. The default is 1,000.
{p 0 4}
{cmd:ycent}{cmd:(}{it:#}{cmd:)} is relevant only with continuous outcomes. It means that U
is simulated on the basis of the binary transformation of the outcome: Y=1 if the continuous outcome
is above the {it:#} centile. Three centiles are allowed: 25, 50, 75.
If {cmd:ycent}{cmd:(}{it:#}{cmd:)} is not selected but the outcome is continuous, U is simulated
on the basis of the transformation: Y=1 if the outcome is above the mean.
{p 0 4}
{cmd:se}{cmd:(}{it:se_type}{cmd:)} allows the user to decide which standard error should be displayed
with the simulated ATT. Three {it:se_type}s are possible: {cmd:set} uses the total variance
in a multiple-imputation setting; {cmd:sew} uses the within-imputation variance;
{cmd:seb} uses the between-imputation variance. The default is {cmd:set}.
{title:Options that are common to {cmd:attnd}, {cmd:attnw}, {cmd:attk} and {cmd:attr}}
{p 0 4}
{cmdab:pscore(}{it:varname}{cmd:)} specifies the name of the user-provided
variable for the estimated propensity score. If no name is provided,
the propensity score is estimated on the basis of the specification in {it:varlist}.
{p 0 4}
{cmd:logit} uses a logit model to estimate the propensity score
instead of the default {cmd:probit} model when the option
{cmdab:pscore(}{it:varname}{cmd:)} is not specified by the user.
{p 0 4}
{cmd:index} requires the use of the linear index as the propensity score
when the option {cmdab:pscore(}{it:varname}{cmd:)} is not specified by the user.
Otherwise no effect is produced.
{p 0 4}
{cmdab:boot:strap} bootstraps the standard error of the estimated ATTs.
{p 0 4}
{cmd:comsup} restricts the computation of the ATT to the region of
common support.
{title:Saved results}
{p} The program stores the simulated treatment effect in {inp:r(att)}.
{p} Standard errors are stored in: {inp:r(se)} (multiple-imputation);
{inp:r(sew)} (within-imputation); {inp:r(seb)} (between-imputation).
{p} The estimated outcome effect and selection effect of the simulated confounder are respectively
stored in {inp:r(yodds)} and {inp:r(todds)}.
{title:Examples}
{inp:. sensatt emp training male age yschool, p(male) or}
{inp:. sensatt emp training male age yschool, p(male) alg(attk) se(seb) boot}
{inp:. sensatt emp training male age yschool, p11(0.6) p10(0.5) p01(0.5) p00(0.2)}
{inp:. sensatt emp training male age yschool, p11(0.6) p10(0.5) p01(0.5) p00(0.2) logit alg(attr) r(100) comsup boot}
{title:Author}
{browse "http://www.tommasonannicini.eu":Tommaso Nannicini}
(Universidad Carlos III de Madrid)
Email {browse "mailto:tnannici@eco.uc3m.es":tnannici@eco.uc3m.es} if you observe any problem.
{title:Also see}
{p}Help for {help pscore}, {help attnd}, {help attnw},
{help attk}, {help attr}.
{p}Analytical formulas and further details on the sensitivity analysis implemented
by this program can be found in: Ichino, A., Mealli, F. and Nannicini, T. (2006),
"From Temporary Help Jobs to Permanent Employment: What Can We Learn from
Matching Estimators and their Sensitivity?", {browse "ftp://ftp.iza.org/dps/dp2149.pdf":IZA DP No. 2149}.
{title:Useful references}
{p 0 2}
Becker, S.O. and Ichino, A. (2002), "Estimation of average treatment effects
based on Propensity Scores", The Stata Journal, 2, 4, 358-377.
{p_end}
{p 0 2}
Dehejia, R.H and Wahba, S. (1999), "Causal Effects in Non-Experimental Studies:
Re-Evaluating the Evaluation of Training Programmes", Journal of the American
Statistical Association, 94, 1053-1062.
{p_end}
{p 0 2}
Dehejia, R.H. and Wahba, S.(2002), "Propensity Score Matching Methods for Non-Experimental
Causal Studies", Review of Economics and Statistics, 84(1), 151-161.
{p_end}
{p 0 2}
Heckman, J.J., Ichimura, H. and Todd, P.E. (1997), "Matching As An Econometric Evaluation
Estimator: Evidence from Evaluating a Job Training Programme", Review of Economic Studies,
64, 605-654.
{p_end}
{p 0 2}
Heckman, J.J., Ichimura, H. and Todd, P.E. (1998), "Matching as an Econometric Evaluation
Estimator", Review of Economic Studies, 65, 261-294.
{p_end}
{p 0 2}
Ichino, A., Mealli, F. and Nannicini, T. (2006), "From Temporary Help Jobs to Permanent Employment:
What Can We Learn from Matching Estimators and their Sensitivity?", IZA DP No. 2149.
{p_end}
{p 0 2}
Imbens, G.W. (2003), "Sensitivity to Exogeneity Assumptions in Program Evaluation",
AEA Papers and Proceedings, 93, 2, 126-132.
{p_end}
{p 0 2}
Imbens, G.W. (2004), "Nonparametric Estimation of Average Treatment Effects Under Exogeneity:
A Review", The Review of Economics and Statistics, 86, 1, 4-29.
{p_end}
{p 0 2}
Rosenbaum, P.R. (1987), "Sensitivity analysis to certain permutation inferences
in matched observational studies", Biometrika, 74, 1, 13-26.
{p_end}
{p 0 2}
Rosenbaum, P.R. and Rubin, D.B. (1983a), "Assessing Sensitivity to an Unobserved
Binary Covariate in an Observational Study with Binary Outcome",
Journal of the Royal Statistical Society, B, 45, 212-218.
{p_end}
{p 0 2}
Rosenbaum, P.R. and Rubin, D.B. (1983b), "The Central Role of the Propensity Score in
Observational Studies for Causal Effects", Biometrika, 70(1), 41-55.
{p_end}
{p 0 2}
Smith, J., and Todd, P.E. (2005), "Does Matching Overcome Lalonde's Critique of
Nonexperimental Estimators?", Journal of Econometrics, 125, 305-353.
{p_end}