___ ____ ____ ____ ____ ® /__ / ____/ / ____/ Stata 18.0 ___/ / /___/ / /___/ MP—Parallel Edition Statistics and Data Science Copyright 1985-2023 StataCorp LLC StataCorp 4905 Lakeway Drive College Station, Texas 77845 USA 800-782-8272 https://www.stata.com 979-696-4600 service@stata.com Stata license: Single-user 2-core perpetual Serial number: 501806221597 Licensed to: Carlos Mendez Nagoya University Notes: 1. Stata is running in batch mode. 2. Unicode is supported; see help unicode_advice. 3. More than 2 billion observations are allowed; see help obs_advice. 4. Maximum number of variables is set to 5,000 but can be increased; see help set_maxvar. . do analysis.do . **************************************************** . * Evaluating a Cash Transfer Program (RCT) . * with Panel Data in Stata . * . * Companion do-file for the tutorial at: . * carlos-mendez.org/post/stata_rct/ . * . * Dataset: dataSIM4RCT.dta . * 2,000 households, balanced panel (2021--2024) . * Stratified randomization, imperfect compliance . * True treatment effect: 0.12 log points (~12%) . * . * Usage: . * 1. Open Stata . * 2. Run: do analysis.do . * 3. All graphs are saved as PNG files . **************************************************** . . clear all . set more off . . *--------------------------------------------------- . * Section 4: Data loading and exploration . *--------------------------------------------------- . . use "https://github.com/quarcs-lab/data-open/raw/master/ametrics/dataSIM4RCT. > dta", clear . . * Describe key variables . des y age edu female poverty treat D Variable Storage Display Value name type format label Variable label ------------------------------------------------------------------------------- y float %9.0g Log monthly consumption age float %9.0g edu float %9.0g female float %9.0g poverty float %9.0g treat float %9.0g Assignment to offer (Z) D float %9.0g Receipt of cash transfer . . * Summary statistics at baseline . sum y age edu female poverty treat D if post==0 Variable | Obs Mean Std. dev. Min Max -------------+--------------------------------------------------------- y | 2,000 10.0154 .4348886 8.454445 11.48253 age | 2,000 35.126 9.650839 18 68 edu | 2,000 12.0275 1.9889 6 18 female | 2,000 .5085 .5000528 0 1 poverty | 2,000 .3125 .4636283 0 1 -------------+--------------------------------------------------------- treat | 2,000 .518 .4998009 0 1 D | 2,000 0 0 0 0 . . * Summary statistics at endline . sum y age edu female poverty treat D if post==1 Variable | Obs Mean Std. dev. Min Max -------------+--------------------------------------------------------- y | 2,000 10.1137 .4382183 8.638689 11.55002 age | 2,000 35.126 9.650839 18 68 edu | 2,000 12.0275 1.9889 6 18 female | 2,000 .5085 .5000528 0 1 poverty | 2,000 .3125 .4636283 0 1 -------------+--------------------------------------------------------- treat | 2,000 .518 .4998009 0 1 D | 2,000 .4615 .4986402 0 1 . . * Declare panel structure . xtset id year Panel variable: id (strongly balanced) Time variable: year, 2021 to 2024, but with gaps Delta: 1 unit . . *--------------------------------------------------- . * Section 5: Baseline balance checks . *--------------------------------------------------- . . * 5.1 T-tests and proportion tests . preserve . keep if post==0 (2,000 observations deleted) . . ttest y, by(treat) Two-sample t test with equal variances ------------------------------------------------------------------------------ Group | Obs Mean Std. err. Std. dev. [95% conf. interval] ---------+-------------------------------------------------------------------- 0 | 964 10.02523 .0140108 .4350124 9.997735 10.05273 1 | 1,036 10.00626 .0135081 .4347839 9.979754 10.03277 ---------+-------------------------------------------------------------------- Combined | 2,000 10.0154 .0097244 .4348886 9.996333 10.03447 ---------+-------------------------------------------------------------------- diff | .0189696 .0194617 -.0191977 .0571369 ------------------------------------------------------------------------------ diff = mean(0) - mean(1) t = 0.9747 H0: diff = 0 Degrees of freedom = 1998 Ha: diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(T < t) = 0.8351 Pr(|T| > |t|) = 0.3298 Pr(T > t) = 0.1649 . ttest age, by(treat) Two-sample t test with equal variances ------------------------------------------------------------------------------ Group | Obs Mean Std. err. Std. dev. [95% conf. interval] ---------+-------------------------------------------------------------------- 0 | 964 35.33506 .3164756 9.826045 34.714 35.95612 1 | 1,036 34.93147 .2947006 9.485516 34.35319 35.50975 ---------+-------------------------------------------------------------------- Combined | 2,000 35.126 .2157993 9.650839 34.70278 35.54922 ---------+-------------------------------------------------------------------- diff | .4035951 .4318923 -.4434114 1.250601 ------------------------------------------------------------------------------ diff = mean(0) - mean(1) t = 0.9345 H0: diff = 0 Degrees of freedom = 1998 Ha: diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(T < t) = 0.8249 Pr(|T| > |t|) = 0.3502 Pr(T > t) = 0.1751 . ttest edu, by(treat) Two-sample t test with equal variances ------------------------------------------------------------------------------ Group | Obs Mean Std. err. Std. dev. [95% conf. interval] ---------+-------------------------------------------------------------------- 0 | 964 11.97407 .0628956 1.952805 11.85064 12.09749 1 | 1,036 12.07722 .0628074 2.02158 11.95398 12.20046 ---------+-------------------------------------------------------------------- Combined | 2,000 12.0275 .0444731 1.9889 11.94028 12.11472 ---------+-------------------------------------------------------------------- diff | -.1031537 .0889963 -.277689 .0713817 ------------------------------------------------------------------------------ diff = mean(0) - mean(1) t = -1.1591 H0: diff = 0 Degrees of freedom = 1998 Ha: diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(T < t) = 0.1233 Pr(|T| > |t|) = 0.2466 Pr(T > t) = 0.8767 . prtest female, by(treat) Two-sample test of proportions 0: Number of obs = 964 1: Number of obs = 1036 ------------------------------------------------------------------------------ Group | Mean Std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- 0 | .4844398 .0160961 .452892 .5159876 1 | .530888 .0155046 .5004996 .5612764 -------------+---------------------------------------------------------------- diff | -.0464482 .022349 -.0902514 -.002645 | under H0: .022372 -2.08 0.038 ------------------------------------------------------------------------------ diff = prop(0) - prop(1) z = -2.0762 H0: diff = 0 Ha: diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(Z < z) = 0.0189 Pr(|Z| > |z|) = 0.0379 Pr(Z > z) = 0.9811 . prtest poverty, by(treat) Two-sample test of proportions 0: Number of obs = 964 1: Number of obs = 1036 ------------------------------------------------------------------------------ Group | Mean Std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- 0 | .3070539 .0148566 .2779356 .3361723 1 | .3175676 .0144633 .28922 .3459152 -------------+---------------------------------------------------------------- diff | -.0105136 .0207342 -.0511518 .0301246 | under H0: .0207424 -0.51 0.612 ------------------------------------------------------------------------------ diff = prop(0) - prop(1) z = -0.5069 H0: diff = 0 Ha: diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(Z < z) = 0.3061 Pr(|Z| > |z|) = 0.6122 Pr(Z > z) = 0.6939 . . * 5.2 Balance table (iebaltab) . capture ssc install ietoolkit, replace . iebaltab y age edu female poverty, grpvar(treat) . . * 5.3 Visual balance plot . capture net install balanceplot, from("https://tdmize.github.io/data") replac > e . balanceplot y age edu i.female i.poverty, group(treat) table nodropdv Base category = 0_0 Reference category = 1_1 N Used In Balance Calculations - N for treat = 0_0 : 964 - N for treat = 1_1 : 1036 Results stored in matrix: bias_0_1 Difference in Means Across Groups of treat: base(0_0) vs ref(1_1) | mean_base mean_ref ttest_pval std_diff -------------+----------------------------------------------- y | 10.025 10.006 0.330 -4.362 age | 35.335 34.931 0.350 -4.179 edu | 11.974 12.077 0.247 5.190 1.female | 0.484 0.531 0.038 9.296 1.poverty | 0.307 0.318 0.612 2.268 . graph export "stata_rct_balance_plot.png", replace width(1200) file stata_rct_balance_plot.png written in PNG format . . * 5.4 AIPW as a formal balance test . teffects aipw (y age edu i.female i.poverty) (treat age edu i.female i.povert > y) Iteration 0: EE criterion = 2.872e-17 Iteration 1: EE criterion = 3.407e-30 Treatment-effects estimation Number of obs = 2,000 Estimator : augmented IPW Outcome model : linear by ML Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATE | treat | (1 vs 0) | -.0244086 .018861 -1.29 0.196 -.0613754 .0125582 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.02792 .0138363 724.75 0.000 10.0008 10.05504 ------------------------------------------------------------------------------ . . * Diagnostic checks . tebalance overid Iteration 0: Criterion = .20558532 Iteration 1: Criterion = .20774777 (backed up) Iteration 2: Criterion = .21224786 Iteration 3: Criterion = .22059216 (backed up) Iteration 4: Criterion = .22082465 (backed up) Iteration 5: Criterion = .22290893 Iteration 6: Criterion = .22491111 Iteration 7: Criterion = .23183975 Iteration 8: Criterion = .2329628 Iteration 9: Criterion = .23527828 Iteration 10: Criterion = .23608747 Iteration 11: Criterion = .23672004 Iteration 12: Criterion = .23713449 Iteration 13: Criterion = .23716776 Iteration 14: Criterion = .23718314 Iteration 15: Criterion = .23718369 Iteration 16: Criterion = .23718379 Overidentification test for covariate balance H0: Covariates are balanced chi2(5) = 3.21614 Prob > chi2 = 0.6667 . tebalance summarize Covariate balance summary Raw Weighted ----------------------------------------- Number of obs = 2,000 2,000.0 Treated obs = 1,036 1,000.1 Control obs = 964 999.9 ----------------------------------------- ----------------------------------------------------------------- |Standardized differences Variance ratio | Raw Weighted Raw Weighted ----------------+------------------------------------------------ age | -.0417918 .0002505 .9318894 .9446877 edu | .0519015 -6.96e-06 1.071677 1.078214 | female | 1 | .0929611 6.51e-06 .9970775 .9999996 | poverty | 1 | .0226764 .0002864 1.018475 1.000233 ----------------------------------------------------------------- . . tebalance density y . graph export "stata_rct_density_y.png", replace width(1200) file stata_rct_density_y.png written in PNG format . . teffects overlap . graph export "stata_rct_overlap_baseline.png", replace width(1200) file stata_rct_overlap_baseline.png written in PNG format . . restore . . *--------------------------------------------------- . * Section 8: Cross-sectional estimation at endline . *--------------------------------------------------- . . preserve . keep if post==1 (2,000 observations deleted) . . * 8.1 Simple difference in means . reg y treat, robust Linear regression Number of obs = 2,000 F(1, 1998) = 35.43 Prob > F = 0.0000 R-squared = 0.0174 Root MSE = .43449 ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. t P>|t| [95% conf. interval] -------------+---------------------------------------------------------------- treat | .1157465 .0194443 5.95 0.000 .0776132 .1538798 _cons | 10.05374 .014001 718.07 0.000 10.02628 10.0812 ------------------------------------------------------------------------------ . . * 8.2 Regression Adjustment -- ATE and ATT . teffects ra (y c.age c.edu i.female i.poverty) (treat), ate Iteration 0: EE criterion = 7.437e-29 Iteration 1: EE criterion = 6.125e-31 Treatment-effects estimation Number of obs = 2,000 Estimator : regression adjustment Outcome model : linear Treatment model: none ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATE | treat | (1 vs 0) | .1125431 .0190927 5.89 0.000 .0751221 .1499641 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.05503 .0138703 724.93 0.000 10.02785 10.08222 ------------------------------------------------------------------------------ . teffects ra (y c.age c.edu i.female i.poverty) (treat), atet Iteration 0: EE criterion = 7.445e-29 Iteration 1: EE criterion = 2.549e-31 Treatment-effects estimation Number of obs = 2,000 Estimator : regression adjustment Outcome model : linear Treatment model: none ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATET | treat | (1 vs 0) | .1132537 .0191498 5.91 0.000 .0757208 .1507865 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.05623 .0140082 717.88 0.000 10.02878 10.08369 ------------------------------------------------------------------------------ . . * 8.3 Inverse Probability Weighting -- ATE and ATT . teffects ipw (y) (treat c.age c.edu i.female i.poverty), ate Iteration 0: EE criterion = 2.872e-17 Iteration 1: EE criterion = 2.842e-31 Treatment-effects estimation Number of obs = 2,000 Estimator : inverse-probability weights Outcome model : weighted mean Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATE | treat | (1 vs 0) | .1126713 .0190886 5.90 0.000 .0752583 .1500844 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.05495 .0138651 725.20 0.000 10.02778 10.08213 ------------------------------------------------------------------------------ . teffects ipw (y) (treat c.age c.edu i.female i.poverty), atet Iteration 0: EE criterion = 2.872e-17 Iteration 1: EE criterion = 1.559e-31 Treatment-effects estimation Number of obs = 2,000 Estimator : inverse-probability weights Outcome model : weighted mean Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATET | treat | (1 vs 0) | .1134031 .0191397 5.93 0.000 .0758899 .1509162 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.05608 .0140004 718.27 0.000 10.02864 10.08352 ------------------------------------------------------------------------------ . . * 8.4 Doubly Robust (IPWRA) -- ATE and ATT . teffects ipwra (y c.age c.edu i.female i.poverty) /// > (treat c.age c.edu i.female i.poverty), vce(robust) Iteration 0: EE criterion = 2.872e-17 Iteration 1: EE criterion = 1.480e-30 Treatment-effects estimation Number of obs = 2,000 Estimator : IPW regression adjustment Outcome model : linear Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATE | treat | (1 vs 0) | .112639 .0190901 5.90 0.000 .0752231 .1500549 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.055 .0138677 725.07 0.000 10.02782 10.08218 ------------------------------------------------------------------------------ . . teffects ipwra (y c.age c.edu i.female i.poverty) /// > (treat c.age c.edu i.female i.poverty), atet vce(robust) Iteration 0: EE criterion = 2.872e-17 Iteration 1: EE criterion = 2.098e-31 Treatment-effects estimation Number of obs = 2,000 Estimator : IPW regression adjustment Outcome model : linear Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATET | treat | (1 vs 0) | .1133162 .0191469 5.92 0.000 .0757889 .1508435 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.05617 .0140019 718.20 0.000 10.02873 10.08361 ------------------------------------------------------------------------------ . . * 8.5 Doubly Robust (AIPW) -- ATE . teffects aipw (y c.age c.edu i.female i.poverty) /// > (treat c.age c.edu i.female i.poverty) Iteration 0: EE criterion = 2.872e-17 Iteration 1: EE criterion = 1.848e-31 Treatment-effects estimation Number of obs = 2,000 Estimator : augmented IPW Outcome model : linear by ML Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATE | treat | (1 vs 0) | .1126412 .0190903 5.90 0.000 .075225 .1500574 -------------+---------------------------------------------------------------- POmean | treat | 0 | 10.055 .013868 725.05 0.000 10.02782 10.08218 ------------------------------------------------------------------------------ . . restore . . *--------------------------------------------------- . * Section 9: Difference-in-Differences . *--------------------------------------------------- . . * 9.3 Basic DiD with panel fixed effects . gen treat_post = treat * post . label var treat_post "Treated x Post (1 only for treated in 2024)" . . xtset id year Panel variable: id (strongly balanced) Time variable: year, 2021 to 2024, but with gaps Delta: 1 unit . xtdidregress (y) (treat_post), group(id) time(year) vce(cluster id) Treatment and time information Time variable: year Control: treat_post = 0 Treatment: treat_post = 1 ----------------------------------- | Control Treatment -------------+--------------------- Group | id | 964 1036 -------------+--------------------- Time | Minimum | 2021 2024 Maximum | 2021 2024 ----------------------------------- Difference-in-differences regression Number of obs = 4,000 Data type: Longitudinal (Std. err. adjusted for 2,000 clusters in id) ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. t P>|t| [95% conf. interval] -------------+---------------------------------------------------------------- ATET | treat_post | (1 vs 0) | .1347161 .0272737 4.94 0.000 .0812282 .188204 ------------------------------------------------------------------------------ Note: ATET estimate adjusted for panel effects and time effects. . . * 9.4 Doubly Robust DiD (DRDID) . capture ssc install drdid, replace . drdid y c.age c.edu i.female i.poverty, ivar(id) time(year) treatment(treat) > dripw 0 Doubly robust difference-in-differences Number of obs = 4,000 Outcome model : least squares Treatment model: inverse probability ------------------------------------------------------------------------------ | Coefficient Std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATET | treat | (1 vs 0) | .1374784 .027387 5.02 0.000 .0838008 .191156 ------------------------------------------------------------------------------ . . * Alternative: Stata 17+ built-in command . xthdidregress aipw (y c.age c.edu i.female i.poverty) /// > (treat_post c.age c.edu i.female i.poverty), group(id) note: variable _did_cohort, containing cohort indicators formed by treatment variable treat_post and group variable id, was added to the dataset. Computing ATET for each cohort and time: Cohort 2024 (1): . done Treatment and time information Time variable: year Time interval: 2021 to 2024 Control: _did_cohort = 0 Treatment: _did_cohort > 0 ------------------------------- | _did_cohort ------------------+------------ Number of cohorts | 2 ------------------+------------ Number of obs | Never treated | 1928 2024 | 2072 ------------------------------- Heterogeneous-treatment-effects regression Number of obs = 4,000 Number of panels = 2,000 Estimator: Augmented IPW Panel variable: id Treatment level: id Control group: Never treated (Std. err. adjusted for 2,000 clusters in id) ------------------------------------------------------------------------------ | Robust Cohort | ATET std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- year | 2024 | .1374784 .027387 5.02 0.000 .0838008 .191156 ------------------------------------------------------------------------------ Note: ATET computed using covariates. . . *--------------------------------------------------- . * Section 10: Endogenous treatment (Advanced) . *--------------------------------------------------- . . * 10.2 Endogenous treatment regression . preserve . keep if post==1 (2,000 observations deleted) . . etregress y c.age i.female i.poverty c.edu, /// > treat(D = treat c.age i.female i.poverty c.edu) vce(robust) Iteration 0: Log pseudolikelihood = -1797.6297 Iteration 1: Log pseudolikelihood = -1797.6297 Linear regression with endogenous treatment Number of obs = 2,000 Estimator: Maximum likelihood Wald chi2(5) = 92.23 Log pseudolikelihood = -1797.6297 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ | Robust | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- y | age | .003187 .0010016 3.18 0.001 .001224 .0051501 1.female | .0801465 .0189552 4.23 0.000 .042995 .117298 1.poverty | -.1030302 .0205984 -5.00 0.000 -.1434023 -.062658 edu | .0182634 .0045243 4.04 0.000 .0093959 .0271308 1.D | .1471 .0246775 5.96 0.000 .0987329 .1954671 _cons | 9.705642 .0694641 139.72 0.000 9.569495 9.841789 -------------+---------------------------------------------------------------- D | treat | 2.55806 .0802103 31.89 0.000 2.40085 2.715269 age | .0014954 .0039348 0.38 0.704 -.0062166 .0092074 1.female | -.0400483 .0758177 -0.53 0.597 -.1886483 .1085516 1.poverty | -.0655305 .0821766 -0.80 0.425 -.2265936 .0955327 edu | .0214726 .0191507 1.12 0.262 -.016062 .0590072 _cons | -1.844408 .2847883 -6.48 0.000 -2.402582 -1.286233 -------------+---------------------------------------------------------------- /athrho | -.0060068 .0481062 -0.12 0.901 -.1002933 .0882796 /lnsigma | -.8567973 .0156472 -54.76 0.000 -.8874653 -.8261293 -------------+---------------------------------------------------------------- rho | -.0060068 .0481045 -.0999584 .088051 sigma | .4245195 .0066426 .411698 .4377404 lambda | -.00255 .0204219 -.0425761 .0374761 ------------------------------------------------------------------------------ Wald test of indep. eqns. (rho = 0): chi2(1) = 0.02 Prob > chi2 = 0.9006 . . gen byte esample = e(sample) . . * ATE of receipt . margins r.D if esample==1 Contrasts of predictive margins Number of obs = 2,000 Model VCE: Robust Expression: Linear prediction, predict() ------------------------------------------------ | df chi2 P>chi2 -------------+---------------------------------- D | 1 35.53 0.0000 ------------------------------------------------ -------------------------------------------------------------- | Delta-method | Contrast std. err. [95% conf. interval] -------------+------------------------------------------------ D | (1 vs 0) | .1471 .0246775 .0987329 .1954671 -------------------------------------------------------------- . . * ATT of receipt . margins, predict(cte) subpop(if D==1 & esample==1) Predictive margins Number of obs = 2,000 Model VCE: Robust Subpop. no. obs = 923 Expression: Conditional treatment effect, predict(cte) ------------------------------------------------------------------------------ | Delta-method | Margin std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- _cons | .1471 .0246775 5.96 0.000 .0987329 .1954671 ------------------------------------------------------------------------------ . . restore . . * 10.3 Doubly robust estimation of receipt effect . preserve . keep if post==1 (2,000 observations deleted) . . teffects ipwra (y y0 c.age i.female i.poverty c.edu) /// > (D c.age i.female i.poverty c.edu treat), vce(robust) Iteration 0: EE criterion = 4.191e-16 Iteration 1: EE criterion = 2.472e-29 Treatment-effects estimation Number of obs = 2,000 Estimator : IPW regression adjustment Outcome model : linear Treatment model: logit ------------------------------------------------------------------------------ | Robust y | Coefficient std. err. z P>|z| [95% conf. interval] -------------+---------------------------------------------------------------- ATE | D | (1 vs 0) | .1172686 .0322495 3.64 0.000 .0540608 .1804764 -------------+---------------------------------------------------------------- POmean | D | 0 | 10.03361 .0171459 585.19 0.000 10 10.06722 ------------------------------------------------------------------------------ . . * Diagnostic checks . tebalance summarize age edu i.female i.poverty Covariate balance summary Raw Weighted ----------------------------------------- Number of obs = 2,000 2,000.0 Treated obs = 923 999.0 Control obs = 1,077 1,001.0 ----------------------------------------- ----------------------------------------------------------------- |Standardized differences Variance ratio | Raw Weighted Raw Weighted ----------------+------------------------------------------------ age | -.0229917 -.0091187 .9957372 1.132753 edu | .0714012 .0130735 1.045596 .9719336 | female | 1 | .0589745 -.0079862 .9978819 1.000002 | poverty | 1 | -.0062371 .0210919 .9951149 1.016377 ----------------------------------------------------------------- . tebalance summarize, baseline Covariate balance summary Raw Weighted ----------------------------------------- Number of obs = 2,000 2,000.0 Treated obs = 923 999.0 Control obs = 1,077 1,001.0 ----------------------------------------- ----------------------------------------------------------------- | Means Variances | Control Treated Control Treated ----------------+------------------------------------------------ age | 35.22841 35.0065 93.3567 92.95874 | female | 1 | .4948932 .524377 .2502062 .2496763 | poverty | 1 | .3138347 .3109426 .2155426 .2144897 | edu | 11.96193 12.10401 3.871226 4.047739 treat | .1569174 .9393283 .1324173 .0570525 ----------------------------------------------------------------- . . tebalance density y0 . graph export "stata_rct_density_y0_receipt.png", replace width(1200) file stata_rct_density_y0_receipt.png written in PNG format . . tebalance density age . . teffects overlap . graph export "stata_rct_overlap_receipt.png", replace width(1200) file stata_rct_overlap_receipt.png written in PNG format . . restore . . *--------------------------------------------------- . * End of analysis . *--------------------------------------------------- . . di _newline(2) . di "============================================" ============================================ . di " Analysis complete." Analysis complete. . di " True treatment effect: 0.12 log points" True treatment effect: 0.12 log points . di " See comparison table in the tutorial." See comparison table in the tutorial. . di "============================================" ============================================ . end of do-file