Coefficient comparison across two sub-samples regression with reg3

[Hakan Gunduz]

Hi all,

I was wondering if there is a way to make meaningful inferences on the coefficients that are estimated by reg3 based on two sub-samples? Assume the following:

reg3 (winv2 wndf wdiv wsize wcf wmtb cage wtang l.stdev l.winv2 i.sectorno) (wndf winv2 wdiv wmtb wcf wtang wsize wol l.stdev i.sectorno) if ablev>0
reg3 (winv2 wndf wdiv wsize wcf wmtb cage wtang l.stdev l.winv2 i.sectorno) (wndf winv2 wdiv wmtb wcf wtang wsize wol l.stdev i.sectorno) if ablev<0

I used Hausman test and rejected the null hypothesis that difference in coefficients are not systematic for each of the equations. But can we compare the actual impacts of coefficients on the basis of one standard deviation? I thought it didn't make sense because one standard deviation in one sub-sample isn't equal to one standard deviation in the other sample.

Is there a way to compare these coefficients by reporting actual number?

Thanks

[Andrew Musau]

Hakan Gunduz, if you provide a data example using dataex, I can try to work something out for you.

[Hakan Gunduz]

Hi Andrew,

thank you for the reply. Please see the data set below:

* Example generated by -dataex-. To install: ssc install dataex
clear
winv2 wndf wdiv wsize wcf wmtb cage wtang stdev sectorno wol ablev
          .           .           .  19.50374           .  .4618229         0   .7269824  .03368416231878015 6  .41277805           .
 .024691245  -.03991951           0 19.387636 -.006440642  .6042107  .6931472   .7204822  .04713174251722159 6  .55332214  -.10400563
 .028318446  -.05022202           0 18.899504   .06990485  .6226496 1.0986123   .7114984  .04443969249369145 6  .50469434  -.20094493
  .08371407   .07575002           0 18.930893  .018279947  .7402942 1.3862944   .6251968  .04684967609206084 6   .6728042  -.11267912
  .10236384 -.008288767           0 18.962965   .17908525  .6928178  1.609438   .5795303 .049014399684528004 6   .6017379   -.0979009
  .14686301  -.04049312           0 18.915792   .13066146  .7741757 1.7917595   .6815129 .044052101331119725 6   .6565078  -.13665149
   .2191839    .2145593           0 19.004965   .13193488  .9771302   1.94591   .7171664  .03517801753907024 6    .649655   .19856507
  .16321194  -.08137736           0   19.0076    .1505007  .9372945 2.0794415   .7164302  .06346885766196175 6   .6877512   .14417413
  .12834354  .027703894           0 19.112946   .13078226  .9024966 2.1972246   .6085404 .024077012903894725 6   .8970788   .16462883
-.034131523  -.14559884           0 18.992754    .1214931  .8497508 2.3025851   .4697834 .023698773087239448 6   .9208653 -.020822406
 -.08357146   -.0781861  .068876386   18.7812   .14973433  .9172774  2.397895    .483317  .03752829107796552 6  1.1195805  -.17695177
  .14300196 -.005790711           0 18.866583   .13023695  .9647316  2.484907   .4457953  .03710473721832368 6   .9416743  -.27385563
   .1916079 .0014446864           0  18.96345   .16384515  .8181329  2.564949  .45354545 .027387082575012413 6   .9169894 -.020567954
   .3246178 .0012546746           0  19.09279   .01981947 1.1015437 2.6390574   .4949109  .03569683730385639 6   .8425992   .10268438
 .021570254    .2522725           0 19.045732   .03535558 1.0347221   2.70805   .3833725 .050586985751680795 6   .9783086   .14955625
  .04428934   -.0880947           0 18.954088   .04382805 1.1724584  2.772589    .407982  .06583092446917034 6  1.0783521   .14131278
   .1411534   .12554465           0  19.05104   .11036421  1.129929  2.833213   .4036872  .03305227828019108 6   .9236502    .2733822
  .11203937 -.008995033           0 19.143196    .1285517 1.3794178  2.890372   .3882919  .06144819581078098 6   .8720471    .0369972
  .08250243 -.017084198           0 19.266495   .14833838  1.824916  2.944439   .4036074  .04360832658142907 6   .8521075  .005359933
  .14364812 -.032362588           0  19.36747   .14674781 1.7001905  2.995732   .4494838 .019374613211923866 6   .8502552  .029852495
   .0890753  -.15297946           0 19.272543   .08633637 3.5271196 3.0445225    .429927 .044973816996687795 6  1.1696298  -.14119488
  .08748706    .1558068           0   19.1524    .0430828 2.9826565 3.0910425    .436003  .04244806305032617 6   1.385409   .03886082
          .           .           .  20.08273           . 4.6283073         0  .17406407 .011501785050569731 6  .01433468           .
          0           0   .06067634  20.19148           .  5.534523  .6931472   .2240611 .010144432109301675 6 .031508595  -.05333525
  .25228322   .14034115   .05348955  20.29712           .  5.537065 1.0986123   .3641776 .018875923888989583 6 .034978915 -.036099136
 -.01895924 -.002906487   .05817629   20.4015           .  5.537065 1.3862944  .29597566 .027869267881923728 6  .04375292  -.04441379
 -.04876679   -.0834287   .04544178  20.23779           .  5.537065  1.609438   .2164926 .023480240700322192 6  .03764143  -.03750107
  .14566031 -.032637075    .0653229 20.378294           .  5.537065 1.7917595   .2351745 .025745615236995933 6  .03846154  -.04369204
  .04765503           0   .05562521  20.57177           . 4.0087576   1.94591  .19266558   .0251858297866997 6  .03776683  -.04766003
  -.0220502           0   .06990382  22.42165    .4576883  .6838123 2.0794415  .02737132 .028274933474665188 6  .01433468  -.08217592
   .4761157     .809868   .03856762 23.304066    .4576883 1.0064584 2.1972246  .53511447  .01918574392236586 6   .6801264   .10641608
-.012108396  -.03399211   .01171742 23.304066           .   .930879 2.3025851   .4929564 .017959545117698168 6   .6427481  -.05896762
  .26635092   .20115957   .01071733 23.304066    .1526843   1.06403  2.397895  .52469754 .014840366366058469 6   .5564235   .03892303
  .14382939  .074386045  .009332854 23.304066    .1341646  .9549593  2.484907  .53334606 .018513724445093326 6   .6803806   .12362427
  .16976377  .012413558   .00976652 23.304066   .14895223  .9359657  2.564949   .5616834  .01790387932535679 6   .6565481   .16270062
   .1437912  .007013781 .0082870005 23.304066   .11729426  .7410959 2.6390574   .6610067  .03691615310408653 6   .6641485   .08365405
  .12104151  .004534249 .0040644873 23.304066   .05804054  .8054844   2.70805   .6882233 .033429295355453706 6   .6065313   .13811028
    .077224  -.04767992  .011766548 23.304066   .15041697  .8117781  2.772589   .6502563  .02360537616002247 6   .6023756  .016979486
  .12988639 .0020390023  .010799096 23.304066    .1039843  .7318522  2.833213   .6296045 .023661493924114048 6   .6063972   .07866913
  .11155552  .013959862   .01362039 23.304066     .129883  .7221554  2.890372    .621285  .02006399626231021 6   .6451005  .068672836
  .06899363  -.07193795  .014435836 23.304066   .09838372  .7945376  2.944439   .5542009 .015025745511735458 6   .5021027   -.0741708
   .0970176  -.02215145  .014638268 23.304066   .08893054  .7174414  2.995732   .6488728 .015253171266204683 6  .51660275  -.10143024
   .0964209  .007190172  .013741212 23.304066   .08095114   .660577 3.0445225   .7050218 .018648308444115014 6  .50097746 -.063893855
   .0715137  .024868606  .006906599 23.304066   .03997409  .7517173 3.0910425   .6789489  .02513981753452406 6   .4695507  -.09401637
  .11804034   .02530533  .008115676 23.304066  .034511793  .7940683  3.135494   .4914198  .01660884372709664 6   .4379458 -.020429224
 .037242737  -.08636665  .007362473 23.304066   .04537373  .6467206  3.178054    .534554 .021619679999356396 6   .6202945  -.06014177
          .           .           .  19.80093           .  .9499344         0   .8555799  .02063368658727732 6  .01433468           .
          0  -.10391106   .01864433 19.698017   .07866998  .9978263  .6931472   .9233871 .016878507060774848 6 .014616936   .04701465
 -.03974272  .013827434  .009813017    19.696    .1628961  .9481625 1.0986123   .8254671 .011904073946015533 6  .01433468  -.01046008
  -.0388576  -.23523733   .01689461  19.70875   .13557924  .9497796 1.3862944   .6359466 .018403688841080963 6  .06005004  -.12824821
   .5960457     .601864   .01999682  20.02122    .0996333  .9506331  1.609438   .7913244 .015241859060166686 6   .2102156   .08821619
 -.10699653  -.10537416  .016211595 19.925644    .1167235  .8913053 1.7917595   .6978595 .023702321593397975 6          .  .003471255
   .1089258  -.06407394  .016851747  19.59644  .016119063 1.0992718   1.94591   .8587916 .028233198710199876 6 .035302106           .
          .           .           . 21.110344           .  5.537065         0    .598644                   . 4  .04506956           .
  .17063554  -.03289135  .013551805  20.88893    .0653006  2.469688  .6931472    .711356  .04413443594414609 4   .0588968           .
   .2581953   .13811292   .02200247  21.06784   .11843421 3.3387394 1.0986123   .7768636 .034923636451326456 4  .05656077 -.007603541
  .20438647  -.02323816    .0201862   21.1317    .1295116  3.949275 1.3862944   .8893459  .02337118456178556 4 .067903556 -.017446838
    .465461     .809868    .0205072  22.10822   .16020714 1.6865592  1.609438   .7971312 .025955989276564565 4  .04393427   .27299842
  .14578836  -.20516984  .007623124  21.88675   .07457565 1.8683873 1.7917595   .6728119   .0252648376230002 4  .05517056  -.08764333
  .05164712   .04189909  .010012466 22.032024   .11195294  1.698537   1.94591   .6089992 .029062703029912634 4 .073584236   .04451913
   .1609963     .096312 .0091045275 22.281023   .12898552 2.2951076 2.0794415   .5948225 .026946112305060173 4  .06631873  -.04312167
     .08772   .07654653  .007100513 22.485586   .12302153 1.7015243 2.1972246   .5265121  .02254938769408602 4 .063277364   .04996207
  .06479966   -.1917641   .00332716  21.86689  .062678814  2.481087 2.3025851  .48372585 .027121642610166172 4  .06845967  -.17109942
  .13904311   .52077895   .01874207 22.629435   .22000475 1.9478763  2.397895   .4852657 .017649629887843573 4  .08746553   .05788054
  .06267651   -.1493234           0 22.420816    .0770199  2.022054  2.484907   .4823215 .020527960449550638 4   .0675638  -.09850451
          .           .           . 17.905233           .  3.468647         0    .320008                   . 5   .0855588           .
  .13849369  -.02897801    .1128888 18.080645   .21280737  2.906028  .6931472   .2520009  .02143966877217036 5   .2992596           .
  .03004423    .1902609   .10687619 18.369621   .19652915 2.1813853 1.0986123  .18733883 .016232860040543546 5  .22132038   .05468811
  .08959439  -.40579975    .0864225  18.72471   .13167405 1.5459754 1.3862944  .14792225 .022746534398275325 5   .3175229  -.21018767
    .849883     .809868   .08007877 19.919827   .08107814  1.294274  1.609438   .4109506 .019682997911068336 5   .0780619  .013189822
  .05085528   .14416128   .01917827  20.02551   .08328384 1.0438381 1.7917595   .3985683  .01822667698506742 5  .26411834 -.013533056
 .020802874  .067735516           0  20.13952  -.03397217  1.252552   1.94591   .4311205 .032570400223860006 5   .3952188   .11186975
 .032024607   -.0870007  .014408337    20.145   .07760371 1.1344051 2.0794415   .4512839  .02271254680669363 5  .27438143   .04684603
 .032327857 -.064043194  .013160215  19.96674   .07154278 1.2587402 2.1972246   .4747382 .020185142215692578 5   .3323719  .009275317
  .10585071    .0835989   .01524293  20.12812   .09886117 1.3997736 2.3025851   .5002612 .024833442866533287 5  .29461834 -.012232274
          .           .           . 17.935513           .   .602916         0   .5981374 .019854317633116667 5   .7304401           .
  .06165434   .03896915           0 18.218035           .  .5436325  .6931472    .449808  .05166529602142017 5   .7085074    .2163376
  .05927089   .03789732           0 18.471264           .   .549671 1.0986123   .3463033 .042594877713488695 5   .6697303    .3060099
  .04767768   .02173832           0 18.602423           .  .5270401 1.3862944     .32691 .004686953915248824 5   .6862354    .2874061
  .07322656 -.027966734           0  18.67219           .  .5272969  1.609438   .4000801 .019854317633116667 5   .8437181   .23906404
  .06024484   .00250668           0 18.690495           . .55158514 1.7917595   .4130885  .04160251471689219 5   .9223216   .26194495
  .05145223   .15729146           0 18.830471           . 1.2117745   1.94591   .3824959 .030100879383093263 5   1.051324  -.10744092
 .021755483  -.09432362           0 18.833801           . 1.4999743 2.0794415   .3831382 .053193495267941426 5  1.0191038  -.08311658
  .05952628  .000531118           0 18.925098           . 1.9906553 2.1972246   .3962263 .020257279600383732 5  1.1024358  -.07709116
   .0545768    -.012639           0  19.10485   .12431119 1.6147667 2.3025851  .36946535  .02272487237600944 5  1.1651723  -.08856204
 .028551577   .14192589           0 19.221315   .18655677 1.7855568  2.397895   .3798598 .014513742405125199 5  1.0990434  -.01604268
  .07611669   .14756052           0  19.51047   .12532893  1.258158  2.484907  .38733006 .016674921831987127 5  1.1477838   -.1053649
 .018116869  -.01921054           0  19.39928   .07107555  1.306708  2.564949   .4196508    .027328531168725 5   .8662283  -.07516566
  .03038292 -.032127142           0 19.416594   .08512443  1.409431 2.6390574   .4151096 .014157853203202226 5  1.2282313  -.09107217
  .02368709 -.019250236           0 19.508427   .10021506  1.213931   2.70805   .3742242 .015176061457407544 5  1.3923794  -.14728978
 .006714966 -.005015858           0  19.58177   .11875656  1.257734  2.772589  .33181575 .013482959926412217 5  1.2189606   -.1650915
 .029930424  .016341198           0  19.68343    .1219693  1.226237  2.833213  .28912425 .013899234487445278 5  1.0839162  -.07738438
  .05612695   .12412202           0   19.8799    .2182824 1.1535556  2.890372  .27543595 .013525098975194622 5   .9065547  .016893685
 .005746168 -.067032896           0  19.69574    .1866189 1.0126789  2.944439   .3191726 .016537325059804462 5   .7069895  -.04839963
  .02065695  .021337155           0 19.645157  .072999425   .938682  2.995732   .3438683  .01645669031552646 5   .5918934  -.04044253
  .03136212  .018070191           0 19.700144   .09260893 1.1590134 3.0445225   .3409738  .02107465882086997 5   .8018831  -.14924273
 .013261742  .006363776           0 19.696724   .08418106 1.0518227 3.0910425   .3523365 .015288838188419229 5   .8316814  -.05611551
          .           .           . 16.721132           . 2.2811081         0 .014532244                   . 5    2.58426           .
 .005449591  -.04541326   .03814714  16.76337    .1709555 1.1806704  .6931472 .010883762  .04340900487263372 5    2.58426           .
.0025491605           0   .07477538 16.956097   .20843934  1.658024 1.0986123 .007358094 .033925921450175964 5    2.58426  .004865156
end

[Andrew Musau]

Hi Hakan Gunduz, when I impose the variance is the same across the two samples (using an interaction term), I get different coefficient estimates. I am therefore not sure that you can do this in reg3 in the way that I imagined. In case I can make any further progress, I will post.

[Hakan Gunduz]

Hi Andrew, once I followed a similar approach and failed to get a meaningful result. Thank you.