Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "Jason P. Kelly" <jpk2@princeton.edu> |
| To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
| Subject | st: predicted probabilities and marginal effects after xtmelogit and xtlogit |
| Date | Tue, 8 Feb 2011 18:00:22 +0000 |
Dear Statalist, We are trying to generate predicted probabilities and marginal effects after mixed-effects and fixed effects logistical regression models (xtmelogit and xtlogit), but getting some odd results. Our dependant variable is whether or not an appeals judge votes to uphold a trial court decision, and the controls are types of judicial selection systems and public opinion in the states with those systems. After xtmelogit, for the predicted probabilities, I follow the direction provided here: http://www.ats.ucla.edu/stat/stata/faq/xtmelogit_prob.htm and run: . margins party, at(retention=1 retpo3=.75 polappoint=0 nonparelec2=0 papo3=0 partpo3=0) predict(mu fixedonly) Predictive margins Number of obs = 7014 Expression : Predicted mean, fixed portion only, predict(mu fixedonly) at : polappoint = 0 retention = 1 nonparelec2 = 0 papo3 = 0 retpo3 = .75 partpo3 = 0 ------------------------------------------------------------------------------ | Delta-method | Margin Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- party | 0 | .598739 .0488954 12.25 0.000 .5029059 .6945721 1 | .488338 .0455832 10.71 0.000 .3989966 .5776793 ------------------------------------------------------------------------------ These results seam fine to me. But the only way I can get stata to produce marginal effects after running xtmelogit is: . margins, predict (mu fixedonly) dydx(*) at(retention=1 retpo3=.5 polappoint=0 nonparelec2=0 papo3=0 partpo3=0) Average marginal effects Number of obs = 7014 Expression : Predicted mean, fixed portion only, predict(mu fixedonly) dy/dx w.r.t. : 1.party copkill rape rob multivic vic_fem grounds polappoint retention nonparelec2 papo3 retpo3 npartpo3 partpo3 at : polappoint = 0 retention = 1 nonparelec2 = 0 papo3 = 0 retpo3 = .5 partpo3 = 0 ------------------------------------------------------------------------------ | Delta-method | dy/dx Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- 1.party | -.1173236 .0304326 -3.86 0.000 -.1769705 -.0576768 copkill | .0953764 .0300031 3.18 0.001 .0365714 .1541815 rape | .0240124 .0160042 1.50 0.134 -.0073552 .0553801 rob | .0284853 .0124724 2.28 0.022 .0040399 .0529307 multivic | .0697821 .0135588 5.15 0.000 .0432074 .0963568 vic_fem | .0184944 .0141448 1.31 0.191 -.0092289 .0462176 grounds | .030395 .0030072 10.11 0.000 .024501 .0362891 polappoint | -.3219874 .3974086 -0.81 0.418 -1.100894 .4569191 retention | .5976574 .2906474 2.06 0.040 .027999 1.167316 nonparelec2 | 1.575365 .2475787 6.36 0.000 1.09012 2.060611 papo3 | 1.758983 .4573814 3.85 0.000 .8625317 2.655434 retpo3 | .3767294 .2466326 1.53 0.127 -.1066615 .8601203 npartpo3 | -.8024966 .2529398 -3.17 0.002 -1.29825 -.3067436 partpo3 | 1.260203 .3036206 4.15 0.000 .665117 1.855288 ------------------------------------------------------------------------------ Note: dy/dx for factor levels is the discrete change from the base level. Most of these results also seem fine. For instance, the marginal effect for party seems to be the difference between party = 0 and party = 1 from the predicted probabilities. But I'm confused by the values over 1. Similarly, after xtlogit, I ran the following to generate the predicted probabilities: . margins party, at(retention=1 retpo3=.75 nonparelec2=0 papo3=0 partpo3=0) predict(pu0) Predictive margins Number of obs = 6609 Model VCE : OIM Expression : Pr(uphold|fixed effect is 0), predict(pu0) at : retention = 1 nonparelec2 = 0 papo3 = 0 retpo3 = .75 partpo3 = 0 ------------------------------------------------------------------------------ | Delta-method | Margin Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- party | 0 | .9834969 .0268031 36.69 0.000 .9309638 1.03603 1 | .0003144 .2352597 0.00 0.999 -.4607861 .4614149 ------------------------------------------------------------------------------ These results seam way off, so I also ran: . predict puphold (option pc1 assumed; probability of success given one success within group) (255 missing values generated) . gen puphold2 = exp(puphold)/(1+exp(puphold)) (255 missing values generated) This produced a column of puphold2 values, which I am led to believe are predicted probabilities, but I'm unsure how to interpret them (I assume they're the probability that a high court will uphold, given values of all the other independent variables for the observation in question)? Is it possible to produce predicted probabilities for party = 0 and 1, when the other independent variables are set at their mean, or some other stated value, as in the case above? I also tried to produce marginal effects at xtlogit as follows: . margins, predict (pu0) dydx(*) at (retention=0 nonparelec2=0 npartpo3=0 retpo3=0 party=1 pa > rtpo3=0.5) Average marginal effects Number of obs = 6609 Model VCE : OIM Expression : Pr(uphold|fixed effect is 0), predict(pu0) dy/dx w.r.t. : party copkill rape rob multivic vic_fem grounds retention nonparelec2 papo3 retpo3 npartpo3 partpo3 at : party = 1 retention = 0 nonparelec2 = 0 retpo3 = 0 npartpo3 = 0 partpo3 = .5 ------------------------------------------------------------------------------ | Delta-method | dy/dx Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- party | -.2119108 88.65475 -0.00 0.998 -173.972 173.5482 copkill | .0076088 3.608257 0.00 0.998 -7.064444 7.079662 rape | .0018991 .9006103 0.00 0.998 -1.763265 1.767063 rob | .0021728 1.030404 0.00 0.998 -2.017381 2.021727 multivic | .0053517 2.537873 0.00 0.998 -4.968788 4.979491 vic_fem | .0017628 .8359773 0.00 0.998 -1.636723 1.640248 grounds | .0025043 1.187582 0.00 0.998 -2.325114 2.330123 retention | .0292498 13.8709 0.00 0.998 -27.15722 27.21572 nonparelec2 | .115432 54.74035 0.00 0.998 -107.1737 107.4045 papo3 | .1622443 76.93982 0.00 0.998 -150.637 150.9615 retpo3 | .0536925 25.46213 0.00 0.998 -49.85116 49.95854 npartpo3 | -.0536713 25.45208 -0.00 0.998 -49.93883 49.83149 partpo3 | .0982661 46.59992 0.00 0.998 -91.2359 91.43244 ------------------------------------------------------------------------------ My questions are many. First, am I using the correct commands to produce the predicted probabilities and marginal effects for these two models? Second, when both models are run correctly, are the predicted probabilities and marginal effects from the two models comparable (i.e. can they be interpreted in the same way, after taking into account the assumptions and restrictions of the two models)? Third (from above), for the marginal effects in the xtmelogit model, if these marginal effects indicate the change in the predicted probability, how can the value be over 1? And fourth (also from above), after the xtlogit, is it possible to produce predicted probabilities for party = 0 and 1, when the other independent variables are set at their mean, or some other stated value, as in the case above? I realize this is a lot to ask (and to sift through), but any assistance is greatly appreciated. Thanks, Jason * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/