----------------------------------------------------------------------------------------------------- log: F:\Stata 9.0\oct25.log log type: text opened on: 25 Oct 2006, 13:51:47 . desc Contains data from E:\Metodos\Leblang\class_sweden.dta obs: 10,732 Extract from Swedish Exit Poll Data vars: 14 8 Sep 2004 11:39 size: 203,908 (80.6% of memory free) ------------------------------------------------------------------------------- storage display value variable name type format label variable label ------------------------------------------------------------------------------- eu byte %40.0g eu Do you think Sweden should resign from the EU or stay in the Union party byte %39.0g party What political party would you vote for in a parliamentary election today gender byte %14.0g gender Gender birthyear int %14.0g birth_year What year were you born citizen byte %14.0g citizen Are you a Swedish citizen union byte %37.0g union Are you a member of a labor union leftright byte %22.0g pol_scale On the left-right political scale, where would you place yourself trust byte %14.0g trust Generally speeking, how much trust do you have for politicians employed byte %67.0g employment What is your employment situation immigration byte %33.0g imm_vote How important was the issue of immigration for how you decided to vote democracy byte %33.0g dem_vote How important was democracy for how you decided to vote interestrate byte %33.0g intrate_vote How important was the possibility for Sweden to decided its interest rate for ho ownecon byte %33.0g ownecon_vote How important was the question of your own economy for how you decided to vote yesno byte %9.0g yes=voted for referenda ------------------------------------------------------------------------------- Sorted by: . summ Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- eu | 8874 1.718391 .4498091 1 2 party | 9736 9.432621 19.8957 1 82 gender | 10446 1.490714 .4999377 1 2 birthyear | 10248 1959.318 16.35441 1911 1985 citizen | 10572 .9643398 .1854504 0 1 -------------+-------------------------------------------------------- union | 10248 2.730386 1.252851 1 4 leftright | 10333 2.891222 1.14788 1 5 trust | 10543 2.481647 .7524196 1 4 employed | 10284 12.25428 26.55986 1 81 immigration | 9354 2.822857 1.209105 1 5 -------------+-------------------------------------------------------- democracy | 9586 1.674317 .9428764 1 5 interestrate | 9440 2.204025 1.241436 1 5 ownecon | 9674 2.10306 1.142772 1 5 yesno | 10383 .4711548 .4991913 0 1 . tab union Are you a member of a labor union | Freq. Percent Cum. --------------------------------------+----------------------------------- Yes, LO-union (Blue-collar workers) | 2,641 25.77 25.77 Yes, TCO-union (WHite COllar WOrkers) | 1,873 18.28 44.05 Yes, SACO-union (Academics) | 1,342 13.10 57.14 No | 4,392 42.86 100.00 --------------------------------------+----------------------------------- Total | 10,248 100.00 . tab union, nolab Are you a | member of a | labor union | Freq. Percent Cum. ------------+----------------------------------- 1 | 2,641 25.77 25.77 2 | 1,873 18.28 44.05 3 | 1,342 13.10 57.14 4 | 4,392 42.86 100.00 ------------+----------------------------------- Total | 10,248 100.00 . recode leftright (1=1) (2=1) (3=2) (4=3) (5=3), gen(lr) (8960 differences between leftright and lr) . tab lr leftright RECODE of | leftright | (On the | left-right | political | scale, | where | On the left-right political scale, where would you would you | place yourself place yo | Clearly t Somewhat Neither l Somewhat Clearly t | Total -----------+-------------------------------------------------------+---------- 1 | 1,373 2,438 0 0 0 | 3,811 2 | 0 0 3,335 0 0 | 3,335 3 | 0 0 0 2,314 873 | 3,187 -----------+-------------------------------------------------------+---------- Total | 1,373 2,438 3,335 2,314 873 | 10,333 . tab lr RECODE of | leftright | (On the | left-right | political | scale, | where would | you place | yo | Freq. Percent Cum. ------------+----------------------------------- 1 | 3,811 36.88 36.88 2 | 3,335 32.28 69.16 3 | 3,187 30.84 100.00 ------------+----------------------------------- Total | 10,333 100.00 . tab union Are you a member of a labor union | Freq. Percent Cum. --------------------------------------+----------------------------------- Yes, LO-union (Blue-collar workers) | 2,641 25.77 25.77 Yes, TCO-union (WHite COllar WOrkers) | 1,873 18.28 44.05 Yes, SACO-union (Academics) | 1,342 13.10 57.14 No | 4,392 42.86 100.00 --------------------------------------+----------------------------------- Total | 10,248 100.00 . gen sindicato = 1- (union ==4) . summ union sindic Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- union | 10248 2.730386 1.252851 1 4 sindicato | 10732 .5907566 .4917172 0 1 . tab union, nolab Are you a | member of a | labor union | Freq. Percent Cum. ------------+----------------------------------- 1 | 2,641 25.77 25.77 2 | 1,873 18.28 44.05 3 | 1,342 13.10 57.14 4 | 4,392 42.86 100.00 ------------+----------------------------------- Total | 10,248 100.00 . tab sindic sindicato | Freq. Percent Cum. ------------+----------------------------------- 0 | 4,392 40.92 40.92 1 | 6,340 59.08 100.00 ------------+----------------------------------- Total | 10,732 100.00 ** Ok, todos los tipo union=4 ahora son sindicato=1 . gen female = gender ==1 (10446 real changes made) . tab female female | Freq. Percent Cum. ------------+----------------------------------- 0 | 5,412 50.43 50.43 1 | 5,320 49.57 100.00 ------------+----------------------------------- Total | 10,732 100.00 . . gen edad = 2002 - birthyear (484 missing values generated) . summ edad birth Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- edad | 10248 42.68199 16.35441 17 91 birthyear | 10248 1959.318 16.35441 1911 1985 . tab lr // nueva variable con 3 categorias RECODE of | leftright | (On the | left-right | political | scale, | where would | you place | yo | Freq. Percent Cum. ------------+----------------------------------- 1 | 3,811 36.88 36.88 2 | 3,335 32.28 69.16 3 | 3,187 30.84 100.00 ------------+----------------------------------- Total | 10,333 100.00 . save , replace // salvando la base modificada file E:\Metodos\Leblang\class_sweden.dta saved ** Usando OLS para una variable categorica (1 left, 2 center, 3 right) . reg lr edad female sindicato Source | SS df MS Number of obs = 9944 -------------+------------------------------ F( 3, 9940) = 115.25 Model | 226.089446 3 75.3631485 Prob > F = 0.0000 Residual | 6499.71104 9940 .653894471 R-squared = 0.0336 -------------+------------------------------ Adj R-squared = 0.0333 Total | 6725.80048 9943 .676435732 Root MSE = .80864 ------------------------------------------------------------------------------ lr | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- edad | .0008964 .0005011 1.79 0.074 -.000086 .0018787 female | -.1149188 .0162422 -7.08 0.000 -.1467568 -.0830809 sindicato | -.2786293 .0165404 -16.85 0.000 -.3110517 -.2462068 _cons | 2.126801 .0254498 83.57 0.000 2.076915 2.176688 ------------------------------------------------------------------------------ ** MODELO ORDERED LOGIT // Variable dependiente categorica ordenada . ologit lr edad female sindicato Iteration 0: log likelihood = -10898.697 Iteration 1: log likelihood = -10730.384 Iteration 2: log likelihood = -10730.204 Ordered logistic regression Number of obs = 9944 LR chi2(3) = 336.99 Prob > chi2 = 0.0000 Log likelihood = -10730.204 Pseudo R2 = 0.0155 ------------------------------------------------------------------------------ lr | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- edad | .0020757 .0011406 1.82 0.069 -.0001599 .0043112 female | -.2608645 .0372023 -7.01 0.000 -.3337797 -.1879493 sindicato | -.6319981 .038045 -16.61 0.000 -.7065649 -.5574314 -------------+---------------------------------------------------------------- /cut1 | -.9723004 .0596821 -1.089275 -.8553257 /cut2 | .3993825 .0589551 .2838326 .5149324 ------------------------------------------------------------------------------ ** Efectos marginales para el grupo-lefty (outcome 1) . mfx, predict(p outcome(1)) Marginal effects after ologit y = Pr(lr==1) (predict, p outcome(1)) = .36322476 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- edad | -.0004801 .00026 -1.82 0.069 -.000997 .000037 42.6368 female*| .0602796 .00857 7.03 0.000 .043474 .077085 .497687 sindic~o*| .1430808 .00838 17.08 0.000 .126658 .159504 .584775 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 ** Efectos marginales para el segundo grupo (outcome 2) . mfx, predict(p outcome(2)) Marginal effects after ologit y = Pr(lr==2) (predict, p outcome(2)) = .3289461 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- edad | .0000378 .00002 1.75 0.080 -4.6e-06 .00008 42.6368 female*| -.0047539 .00097 -4.89 0.000 -.00666 -.002848 .497687 sindic~o*| -.0063975 .00168 -3.80 0.000 -.009693 -.003102 .584775 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 ** Efectos marginales para el tercer grupo--right (outcome 2) . mfx, predict(p outcome(3)) Marginal effects after ologit y = Pr(lr==3) (predict, p outcome(3)) = .30782914 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- edad | .0004423 .00024 1.82 0.069 -.000034 .000919 42.6368 female*| -.0555258 .00791 -7.02 0.000 -.071032 -.04002 .497687 sindic~o*| -.1366833 .00829 -16.48 0.000 -.152939 -.120427 .584775 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 ** Modelo ORDERED PROBIT // Variable dependiente categorica ordenada . oprobit lr edad female sindicato Iteration 0: log likelihood = -10898.697 Iteration 1: log likelihood = -10730.124 Iteration 2: log likelihood = -10730.117 Ordered probit regression Number of obs = 9944 LR chi2(3) = 337.16 Prob > chi2 = 0.0000 Log likelihood = -10730.117 Pseudo R2 = 0.0155 ------------------------------------------------------------------------------ lr | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- edad | .0012506 .0006985 1.79 0.073 -.0001183 .0026196 female | -.1598753 .0227072 -7.04 0.000 -.2043806 -.1153701 sindicato | -.3854371 .0231534 -16.65 0.000 -.4308169 -.3400573 -------------+---------------------------------------------------------------- /cut1 | -.5990014 .0362764 -.6701018 -.5279011 /cut2 | .2480839 .0360169 .177492 .3186757 ------------------------------------------------------------------------------ ** Efectos marginales... . mfx, predict(p outcome(1)) Marginal effects after oprobit y = Pr(lr==1) (predict, p outcome(1)) = .36415961 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- edad | -.0004697 .00026 -1.79 0.073 -.000984 .000044 42.6368 female*| .0599984 .00851 7.05 0.000 .043327 .07667 .497687 sindic~o*| .1423003 .00836 17.02 0.000 .125912 .158689 .584775 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . mfx, predict(p outcome(2)) Marginal effects after oprobit y = Pr(lr==2) (predict, p outcome(2)) = .32720536 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- edad | .0000294 .00002 1.73 0.084 -3.9e-06 .000063 42.6368 female*| -.0037594 .00077 -4.88 0.000 -.00527 -.002249 .497687 sindic~o*| -.0050864 .00133 -3.83 0.000 -.007692 -.002481 .584775 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 . mfx, predict(p outcome(3)) Marginal effects after oprobit y = Pr(lr==3) (predict, p outcome(3)) = .30863503 ------------------------------------------------------------------------------ variable | dy/dx Std. Err. z P>|z| [ 95% C.I. ] X ---------+-------------------------------------------------------------------- edad | .0004404 .00025 1.79 0.073 -.000042 .000922 42.6368 female*| -.056239 .00798 -7.05 0.000 -.071876 -.040602 .497687 sindic~o*| -.1372139 .0083 -16.54 0.000 -.153476 -.120952 .584775 ------------------------------------------------------------------------------ (*) dy/dx is for discrete change of dummy variable from 0 to 1 ** MODELO MULTINOMIAL LOGIT // Variable dependiente categorica no ordenada . mlogit lr edad female sindicato Iteration 0: log likelihood = -10898.697 Iteration 1: log likelihood = -10711.777 Iteration 2: log likelihood = -10711.076 Iteration 3: log likelihood = -10711.076 Multinomial logistic regression Number of obs = 9944 LR chi2(6) = 375.24 Prob > chi2 = 0.0000 Log likelihood = -10711.076 Pseudo R2 = 0.0172 ------------------------------------------------------------------------------ lr | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- 2 | edad | .0059508 .0015291 3.89 0.000 .0029538 .0089479 female | .0412526 .0488406 0.84 0.398 -.0544733 .1369785 sindicato | -.4667908 .0507699 -9.19 0.000 -.566298 -.3672835 _cons | -.118158 .0788364 -1.50 0.134 -.2726745 .0363585 -------------+---------------------------------------------------------------- 3 | edad | .0027155 .0015413 1.76 0.078 -.0003053 .0057364 female | -.3624815 .0498497 -7.27 0.000 -.4601851 -.2647778 sindicato | -.8384621 .0509448 -16.46 0.000 -.938312 -.7386122 _cons | .3841968 .0772173 4.98 0.000 .2328538 .5355399 ------------------------------------------------------------------------------ (lr==1 is the base outcome) ** Interpretacion: ** 1. Las mujeres tienen una menor probabilidad de declarar ** ser de derecha que ser de izquierda (grupo de comparacion). ** 2. Las mujeres tienen la misma probabilidad de declararse de centro que de ser ** de izquierda (grupo de comparacion) ** 3. Los sindicalizados tienen mas probabilidad de ser de izquierda que de centro ** y mucho menos que de derecha. ** Ojo: los efectos marginales requieren usar MFX o CLARIFY... pero viendo los ** coefs podemos decir todo lo anterior. ** MLOGIT con otro grupo de comparacion: . mlogit lr edad female sindicato, baseoutcome(3) Iteration 0: log likelihood = -10898.697 Iteration 1: log likelihood = -10711.777 Iteration 2: log likelihood = -10711.076 Iteration 3: log likelihood = -10711.076 Multinomial logistic regression Number of obs = 9944 LR chi2(6) = 375.24 Prob > chi2 = 0.0000 Log likelihood = -10711.076 Pseudo R2 = 0.0172 ------------------------------------------------------------------------------ lr | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- 1 | edad | -.0027155 .0015413 -1.76 0.078 -.0057364 .0003053 female | .3624815 .0498497 7.27 0.000 .2647778 .4601851 sindicato | .8384621 .0509448 16.46 0.000 .7386122 .938312 _cons | -.3841968 .0772173 -4.98 0.000 -.5355399 -.2328538 -------------+---------------------------------------------------------------- 2 | edad | .0032353 .001532 2.11 0.035 .0002327 .0062379 female | .403734 .0509521 7.92 0.000 .3038698 .5035983 sindicato | .3716714 .0511322 7.27 0.000 .2714542 .4718885 _cons | -.5023548 .0778995 -6.45 0.000 -.6550351 -.3496746 ------------------------------------------------------------------------------ (lr==3 is the base outcome) ** Noten que los coeficientes del grupo 1 son el inverso de los del grupo 3 ** del modelo anterior donde basecategory era el grupo 1... ** Pruebas de hipotesis . test female // Ho: female es insignificante para todos los grupos ( 1) [1]female = 0 ( 2) [2]female = 0 chi2( 2) = 75.88 Prob > chi2 = 0.0000 . mlogit lr edad female sindicato, baseoutcome(2) Iteration 0: log likelihood = -10898.697 Iteration 1: log likelihood = -10711.777 Iteration 2: log likelihood = -10711.076 Iteration 3: log likelihood = -10711.076 Multinomial logistic regression Number of obs = 9944 LR chi2(6) = 375.24 Prob > chi2 = 0.0000 Log likelihood = -10711.076 Pseudo R2 = 0.0172 ------------------------------------------------------------------------------ lr | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- 1 | edad | -.0059508 .0015291 -3.89 0.000 -.0089479 -.0029538 female | -.0412526 .0488406 -0.84 0.398 -.1369785 .0544733 sindicato | .4667908 .0507699 9.19 0.000 .3672835 .566298 _cons | .118158 .0788364 1.50 0.134 -.0363585 .2726745 -------------+---------------------------------------------------------------- 3 | edad | -.0032353 .001532 -2.11 0.035 -.0062379 -.0002327 female | -.403734 .0509521 -7.92 0.000 -.5035983 -.3038698 sindicato | -.3716714 .0511322 -7.27 0.000 -.4718885 -.2714542 _cons | .5023548 .0778995 6.45 0.000 .3496746 .6550351 ------------------------------------------------------------------------------ (lr==2 is the base outcome) . test [1]sindicato = -[3]sindicato // Ho: Sindicato tiene efecto inverso en grupo 1 y 3. ( 1) [1]sindicato + [3]sindicato = 0 chi2( 1) = 1.16 Prob > chi2 = 0.2811 ** No podemos rechazar que el coef de sindicato es de la misma magnitud y signo inverso ** en los grupos 1 y 3 respecto del grupo 2: comparado con los centristas, los sindicalizados ** tienen la misma probabilidad de ser izquierda que de no ser derecha... ** Ok, la interpretacion suena rara pero es solo un ejemplo... ** USANDO CLARIFY para hacer predicciones y comparaciones . net from http://gking.harvard.edu/clarify ----------------------------------------------------------------------------------------- http://gking.harvard.edu/clarify/ Clarify: Software for Interpreting and Presenting Statistical Results ----------------------------------------------------------------------------------------- Michael Tomz Jason Wittenberg Gary King Stanford Univ. Univ. of Wisconsin, Madison Harvard Univ. For a brief description of Clarify, type net describe clarify To download documentation to your working directory, type net get clarify To install Clarify program files, type type net install clarify To update from Clarify 2.0 or later, type net install clarify, replace To update from versions earlier than 2.0, please see the documentation PACKAGES you could -net describe-: clarify Software for Interpreting and Presenting Statistical Results ----------------------------------------------------------------------------------------- . net get clarify checking clarify consistency and verifying not already installed... copying into current directory... copying clarify.pdf ancillary files successfully copied. . net install clarify checking clarify consistency and verifying not already installed... installing into c:\ado\plus\... installation complete. ** CLARIFY TIENE TRES COMANDOS BASICOS: ** estsimp-- estimates simulated parameters (via bootstrap), it is a prefix ** setx -- sets indep variables X at values of interest, it has many options ** simqi -- simulates quantities of interest after previous two commands, it has many options . estsimp mlogit lr edad female sindicato // it is just a prefix to almost ANY stata estimation Iteration 0: log likelihood = -10898.697 Iteration 1: log likelihood = -10711.777 Iteration 2: log likelihood = -10711.076 Iteration 3: log likelihood = -10711.076 Multinomial logistic regression Number of obs = 9944 LR chi2(6) = 375.24 Prob > chi2 = 0.0000 Log likelihood = -10711.076 Pseudo R2 = 0.0172 ------------------------------------------------------------------------------ lr | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- 2 | edad | .0059508 .0015291 3.89 0.000 .0029538 .0089479 female | .0412526 .0488406 0.84 0.398 -.0544733 .1369785 sindicato | -.4667908 .0507699 -9.19 0.000 -.566298 -.3672835 _cons | -.118158 .0788364 -1.50 0.134 -.2726745 .0363585 -------------+---------------------------------------------------------------- 3 | edad | .0027155 .0015413 1.76 0.078 -.0003053 .0057364 female | -.3624815 .0498497 -7.27 0.000 -.4601851 -.2647778 sindicato | -.8384621 .0509448 -16.46 0.000 -.938312 -.7386122 _cons | .3841968 .0772173 4.98 0.000 .2328538 .5355399 ------------------------------------------------------------------------------ (lr==1 is the base outcome) Simulating main parameters. Please wait.... % of simulations completed: 12% 25% 37% 50% 62% 75% 87% 100% Number of simulations : 1000 Names of new variables : b1 b2 b3 b4 b5 b6 b7 b8 . summ b* Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- b1 | 1000 .0059455 .001511 .0003019 .0105552 b2 | 1000 .0413949 .0495448 -.1162226 .1913942 b3 | 1000 -.4673169 .0518243 -.6552606 -.3019966 b4 | 1000 -.1177766 .07832 -.395796 .1620496 b5 | 1000 .0026936 .0015144 -.0023063 .0089249 -------------+-------------------------------------------------------- b6 | 1000 -.3626331 .0510846 -.4913138 -.2186288 b7 | 1000 -.8351774 .0509477 -.9981426 -.6801283 b8 | 1000 .3828034 .0760216 .1352265 .5931341 ** CLARIFY guarda las betas estimadas en la simulacion, con todo y sus std dev. . setx mean // fija todas las X en sus medias . setx // describe los valores fijados en memoria You have set the following values for the explanatory variables: ------------------------------------ Variable | Value Description ----------+------------------------- edad | 42.63676 mean female | .497687 mean sindic~o | .5847747 mean ------------------------------------ . simqi // simla cantidades o probabilidades relevantes, dado setx Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- Pr(lr=1) | .3659675 .0050343 .3561332 .375391 Pr(lr=2) | .3254018 .0048218 .3156473 .3352482 Pr(lr=3) | .3086306 .0049694 .2990501 .3181146 ** Ahora, simulando female y sindicato = 1 . setx edad mean female 1 sindicato 1 . setx You have set the following values for the explanatory variables: ------------------------------------ Variable | Value Description ----------+------------------------- edad | 42.63676 mean female | 1 1 sindic~o | 1 1 ------------------------------------ . simqi Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- Pr(lr=1) | .4456142 .0085627 .4290182 .4618562 Pr(lr=2) | .3334613 .0078376 .3187407 .3485599 Pr(lr=3) | .2209244 .0066801 .2073952 .2333689 ** Una mujer sindicalizada con edad promedio tiene 44% de prob de ser lefty y solo 22% ** de ser righty. . setx edad mean female 1 sindicato 0 . simqi Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- Pr(lr=1) | .2994505 .0086166 .2825707 .3152965 Pr(lr=2) | .3568051 .0089964 .3390937 .3735984 Pr(lr=3) | .3437445 .0091492 .3256003 .3619084 ** Una mujer promedio no sindicalizada tiene 29% de prob de ser de izquierda, etc. . setx edad 22 female 1 sindicato 1 . simqi Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- Pr(lr=1) | .4692208 .0112419 .4466649 .4911531 Pr(lr=2) | .3106945 .0096957 .2912363 .3297635 Pr(lr=3) | .2200847 .0081234 .2047687 .2366863 ** Una mujer de 22 sindicalizada tiene 46.9% de prob de ser de izquierda, etc. . setx edad 22 female 1 sindicato 0 . simqi Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- Pr(lr=1) | .3184575 .0103992 .2987475 .3386323 Pr(lr=2) | .3357331 .0102933 .3152839 .3556977 Pr(lr=3) | .3458094 .010604 .3252761 .3660856 ** Pero si no es sindicalizada, su prob(left) es solo de 31.8% . setx You have set the following values for the explanatory variables: ------------------------------------ Variable | Value Description ----------+------------------------- edad | 22 22 female | 1 1 sindic~o | 0 0 ------------------------------------ ** Calculando el CAMBIO EN LA PROBABILIDAD dados ciertos CAMBIOS EN X: ** usamos opcion: fd = first difference y changex = variables que cambian y como . simqi, fd(pr) changex(female 0 1 sindicato 0 1) First Difference: female 0 1 sindicato 0 1 Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- dPr(lr = 1) | .189398 .0137592 .16401 .2166948 dPr(lr = 2) | .0275936 .0123046 .0042688 .0520834 dPr(lr = 3) | -.2169916 .0127662 -.2432448 -.1927709 ** La probabilidad de ser lefty aumenta en 18.9% si comparamos un hombre no sindicalizado ** con una mujer sindizalizada. La prob de ser righty baja 21% en el mismo escenario. . setx You have set the following values for the explanatory variables: ------------------------------------ Variable | Value Description ----------+------------------------- edad | 22 22 female | 1 1 sindic~o | 0 0 ------------------------------------ . simqi, fd(pr) changex(edad 22 40) First Difference: edad 22 40 Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- dPr(lr = 1) | -.0166022 .0052956 -.0269718 -.0068437 dPr(lr = 2) | .0183564 .0052655 .0081228 .0284809 dPr(lr = 3) | -.0017542 .0053577 -.0123605 .0089621 ** La prob de ser lefty de una mujer no sindicalizada cambia muy poco al pasar de 22 a 40 anios. . simqi, fd(pr) changex(edad 22 40 sindicato 0 1) First Difference: edad 22 40 sindicato 0 1 Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- dPr(lr = 1) | .1326893 .0106772 .1120438 .1530419 dPr(lr = 2) | .0025051 .0099526 -.0169623 .0210168 dPr(lr = 3) | -.1351944 .0107058 -.1564473 -.1149134 . simqi, fd(pr) changex(edad 22 40 sindicato 0 1 female 0 0 ) First Difference: edad 22 40 sindicato 0 1 female 0 0 Quantity of Interest | Mean Std. Err. [95% Conf. Interval] ---------------------------+-------------------------------------------------- dPr(lr = 1) | .1342345 .0104835 .1139951 .1544289 dPr(lr = 2) | .0096288 .0095135 -.0093897 .0274637 dPr(lr = 3) | -.1438632 .0112638 -.1663143 -.1225234 ** Noten como dejamos female sin cambio alguno... puede ayudar al hacer multiples ** combinaciones . log close log: F:\Stata 9.0\oct25.log log type: text closed on: 25 Oct 2006, 14:40:45 -----------------------------------------------------------------------------------------