I am trying to estimate marginal effects after having run a triple interaction (in a linear probability model) with one of the main effects being caught by the fixed effect. Therefore the coefficient of one of the main effects is omitted. I assume that that is why I cannot estimate marginal effects after running this model (from my understanding this should lead to the constant being "unestimable", right?). Is there a way for me to work around this (other than just running the model without fixed effects)? It is really the slope I am interested in, more so than any intercepts - so perhaps there is a way of just inserting a/any value for the intercept?! Also, in case it helps, the main effect which drops out is for a binary variable (which makes split sample analyses an option, but I wanted to see whether there are other ways of dealing with this first and leave that as a last resort).
In Stata terms, I am trying to estimate the following two lines below. I have also added a data example at the bottom of this, in case it helps illustrate my point.
Code:
areg Y c.X1##c.X2##i.X3, a(FE) margins, dydx(X1) at(X2 = (0 (0.5) 4.5) X3=(0 1)) vsquish
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input long FE float(X2 Y X1 X3) 10143018 1.0986123 1 1.231193 1 10143018 0 0 .7493059 1 10143009 0 1 1.3820202 1 10143009 1.0986123 0 1.275593 1 10142872 2.70805 1 1.307223 0 10142872 3.135494 0 .9065389 0 10142848 2.772589 0 1.2386667 1 10142848 0 1 1.2820165 1 10142827 .6931472 0 1.2553422 1 10142827 0 1 1.379391 1 10142595 2.944439 1 1.0276814 1 10142595 4.3438053 0 1.0960358 1 10142594 0 0 .3706292 0 10142594 1.7917595 1 1.3194864 0 10142451 2.70805 1 1.0763656 1 10142451 0 0 .9885153 1 10142448 1.94591 1 1.2935138 1 10142448 0 0 1.277683 1 10142194 .6931472 1 1.271437 0 10142194 .6931472 0 1.0061655 0 10142182 .6931472 1 1.1844742 0 10142182 1.94591 0 1.2094254 0 10142124 0 1 .9228793 0 10142124 .6931472 0 .9958407 0 10142121 2.890372 0 1.0443183 1 10142121 3.4011974 1 1.1112077 1 10142115 2.397895 1 1.2370248 0 10142115 1.94591 0 1.1493943 0 10142076 1.0986123 0 1.2654196 1 10142076 0 1 1.3615023 1 10142013 0 0 .6853231 1 10142013 0 1 .27892447 1 10141965 2.0794415 0 1.1266971 1 10141965 .6931472 1 1.3776045 1 10141736 1.3862944 0 1.206344 1 10141736 0 1 .7457477 1 10141708 0 1 .6492612 0 10141708 .6931472 0 .57538915 0 10141638 4.1271343 1 1.3236197 1 10141638 2.3025851 0 1.3121105 1 10141615 0 0 .6377352 1 10141615 .6931472 1 1.289306 1 10141577 1.7917595 1 1.255946 1 10141577 0 0 .6630335 1 10141562 2.995732 1 .9476664 0 10141562 1.94591 0 .9631982 0 10141494 0 1 .7453708 0 10141494 0 0 .4729676 0 10141457 2.1972246 1 .7859399 1 10141457 2.397895 0 .8258712 1 10141314 1.3862944 1 1.349841 0 10141314 1.3862944 0 .55477506 0 10141191 1.7917595 1 1.0532235 0 10141191 1.0986123 0 .951867 0 10141120 1.7917595 1 .7505929 1 10141120 2.995732 0 1.2242912 1 10141086 .6931472 0 1.1190642 0 10141086 0 1 1.0493766 0 10141084 2.0794415 1 .9811458 1 10141084 1.94591 0 .9422299 1 10140945 0 1 .8014972 0 10140945 0 0 .6099473 0 10140814 0 0 .26569033 0 10140814 1.7917595 1 1.3924794 0 10140812 .6931472 0 .34921825 0 10140812 0 1 1.4015034 0 10140800 2.70805 1 .5356797 0 10140800 3.9512436 0 .9135412 0 10140781 0 0 1.0477619 1 10140781 1.0986123 1 1.0129087 1 10140746 4.6051702 1 1.1419833 1 10140746 4.844187 0 1.1674879 1 10140699 0 1 .9487832 1 10140699 .6931472 0 1.068078 1 10140628 2.1972246 0 1.0852195 0 10140628 2.564949 1 1.0986866 0 10140598 3.5263605 1 1.1449307 1 10140598 0 0 .6565773 1 10140532 0 0 1.1055694 1 10140532 3.0445225 1 1.1195079 1 10140387 3.295837 1 1.158425 0 10140387 3.871201 0 1.1757511 0 10140376 3.135494 1 1.209297 1 10140376 1.3862944 0 1.0646921 1 10140367 0 0 .9131414 0 10140367 1.3862944 1 1.1242832 0 10140360 0 1 1.3866943 0 10140360 0 0 .4267809 0 10140339 0 0 .59409094 0 10140339 0 1 .6886784 0 10140333 4.0775375 0 .7903495 0 10140333 5.32301 1 1.2403438 0 10140308 1.0986123 0 .6452845 0 10140308 0 1 1.3927143 0 10140302 0 1 1.2086025 1 10140302 .6931472 0 1.273538 1 10140301 .6931472 1 .9430985 0 10140301 2.1972246 0 1.0170584 0 10140294 1.0986123 1 1.2636527 1 10140294 1.94591 0 1.2083483 1 10140283 3.218876 0 1.1499057 1 10140283 3.89182 1 1.1415253 1 10140278 3.89182 0 1.1802843 0 10140278 3.988984 1 1.1738459 0 10140264 1.0986123 0 1.0200553 0 10140264 3.3322046 1 1.1923115 0 10140208 0 1 .4746172 0 10140208 0 0 .5337738 0 10140177 2.484907 1 .7916193 0 10140177 1.0986123 0 .7291901 0 10140166 . 1 . 1 10140166 0 0 .9897379 1 10140106 1.609438 0 1.2267642 1 10140106 2.772589 1 1.1130809 1 10140011 1.609438 0 1.1362092 0 10140011 1.3862944 1 .9261208 0 10140009 .6931472 0 .8951876 1 10140009 0 1 1.2029034 1 10140000 .6931472 0 1.2000805 0 10140000 0 1 1.1441292 0 10139988 0 1 .9697916 0 10139988 0 0 .8390914 0 10139858 0 0 .6079006 0 10139858 2.0794415 1 1.2491387 0 10139843 0 0 .54556745 0 10139843 0 1 .6603725 0 10139812 1.94591 0 .9233347 1 10139812 1.609438 1 1.0238228 1 10139787 1.0986123 0 .9764532 0 10139787 0 1 .9589562 0 10139361 2.1972246 1 .90837 1 10139361 2.6390574 0 .9134418 1 10139270 2.6390574 1 .820991 0 10139270 .6931472 0 .4257423 0 10139169 0 0 .53702605 1 10139169 .6931472 1 .57743293 1 10139115 1.609438 1 .8493823 0 10139115 1.94591 0 .8016933 0 10138951 0 0 .5689031 1 10138951 3.433987 1 .7854894 1 10138812 4.3307333 1 1.0107682 1 10138812 1.609438 0 .8471205 1 10138759 1.3862944 1 .52097213 1 10138759 2.484907 0 1.2818906 1 10138697 0 0 .9308263 0 10138697 2.397895 1 .9227511 0 10138356 0 0 .7657002 1 10138356 2.6390574 1 1.1339812 1 10138339 4.553877 0 .9919426 1 10138339 4.394449 1 .9746974 1 10138323 5.214936 1 1.0302064 1 10138323 3.4011974 0 .9377924 1 10138293 3.2580965 1 1.1161833 0 10138293 3.0445225 0 1.0981154 0 10138169 2.0794415 1 1.0170199 1 10138169 0 0 .8199392 1 10138120 2.70805 1 1.1370072 1 10138120 1.0986123 0 1.1125196 1 10138106 1.7917595 0 .6866983 0 10138106 1.94591 1 .7396418 0 10137864 1.94591 1 .7463221 1 10137864 1.609438 0 .7103068 1 10137812 2.0794415 1 .5715956 0 10137812 0 0 .5133577 0 10137739 5.147494 0 .3986844 1 10137739 3.6888795 1 .3789936 1 10137651 0 1 .8744928 1 10137651 2.397895 0 .8746155 1 10137625 1.3862944 0 .7511013 1 10137625 3.89182 1 1.0652461 1 10137546 0 1 1.080466 1 10137546 .6931472 0 .7146025 1 10137523 1.609438 0 .8756774 0 10137523 .6931472 1 .8779976 0 10137472 2.0794415 0 1.1346102 0 10137472 0 1 .9702966 0 10137304 1.0986123 1 1.3779285 0 10137304 2.70805 0 .3735727 0 10137271 3.9318256 1 1.1716378 1 10137271 1.3862944 0 .38255095 1 10137269 3.4011974 1 1.0100572 1 10137269 3.433987 0 .9836367 1 10137265 3.9318256 0 1.3762392 0 10137265 3.135494 1 1.2902777 0 10137260 2.995732 1 .7448317 1 10137260 2.833213 0 .7563806 1 10137229 3.218876 0 1.1252269 0 10137229 3.8501475 1 .5704803 0 10137208 0 1 1.1243953 1 10137208 1.0986123 0 1.187624 1 10137180 2.484907 0 .3911843 1 10137180 2.70805 1 .3868511 1 10137135 0 0 .3389531 1 10137135 . 1 . 1 10136909 2.1972246 1 1.3949153 1 10136909 0 0 .4510143 1 10136862 0 1 1.0193574 1 10136862 1.609438 0 .9723536 1 10136850 4.976734 1 .6832581 1 10136850 3.9318256 0 .6890621 1 end label values Y realized label def realized 0 "Not Observed", modify label def realized 1 "Observed Patent-Lawyer Pair", modify
0 Response to Estimating Marginal Effects after Regression in which one Main Effect is Caught by Fixed Effect
Post a Comment