Tobit regression in stata - F statistic reported instead of Chi2 statistic

[Mohd Fahad]

Dear Members,

Greetings!

I am executing Tobit regression but I do not see Chi2 statistic in the output with robust standard errors. When I use vce(robust) then I get F statistic. However if I do not use robust option I get chisquare statistic. I am studying the effect of distance on the equity in acquisitions. I have controlled for industry fixed effects. The value of equity ranges from 0.8% to 100%.

Since most of the papers I have referred on tobit regression report chi square statistic, I was curious to know if only reporting F statistic is OK? Is my model OK?

Thank you very much.
Regards
Fahad

Output with vce(robust)

tobit sought ind1 ind2 ind3 ind4 ind5 ind6 ind7 ind8 ind9 ind10 lndistance , ll(0.8) ul(100) vce(robust )

Tobit regression Number of obs = 11,205
F( 11, 11194) = 25.41
Prob > F = 0.0000
Log pseudolikelihood = -10600.256 Pseudo R2 = 0.0133

------------------------------------------------------------------------------
| Robust
sought | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ind1 | 46.14627 6.109372 7.55 0.000 34.17083 58.12172
ind2 | 13.52896 9.534614 1.42 0.156 -5.160563 32.21848
ind3 | 16.23009 2.815702 5.76 0.000 10.71082 21.74936
ind4 | -3.733479 4.873112 -0.77 0.444 -13.28564 5.818678
ind5 | 65.68343 14.4861 4.53 0.000 37.28812 94.07874
ind6 | 32.57879 2.611634 12.47 0.000 27.45953 37.69806
ind7 | 24.50673 11.10461 2.21 0.027 2.739732 46.27372
ind8 | 27.49329 19.30412 1.42 0.154 -10.34618 65.33276
ind9 | 28.20282 5.932656 4.75 0.000 16.57377 39.83187
ind10 | 35.62353 6.967574 5.11 0.000 21.96586 49.2812
lndistance | -5.228204 .886592 -5.90 0.000 -6.96608 -3.490328
_cons | 201.3265 7.322474 27.49 0.000 186.9732 215.6799
-------------+----------------------------------------------------------------
/sigma | 66.16526 .8951188 64.41067 67.91985
------------------------------------------------------------------------------
0 left-censored observations
1,424 uncensored observations
9,781 right-censored observations at sought >= 100



Output without robust standard errors

. tobit sought ind1 ind2 ind3 ind4 ind5 ind6 ind7 ind8 ind9 ind10 lndistance , ll(0.8) ul(100)

Tobit regression Number of obs = 11,205
LR chi2(11) = 286.23
Prob > chi2 = 0.0000
Log likelihood = -10600.256 Pseudo R2 = 0.0133

------------------------------------------------------------------------------
sought | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ind1 | 46.14627 5.992422 7.70 0.000 34.40007 57.89247
ind2 | 13.52896 9.43937 1.43 0.152 -4.973868 32.03178
ind3 | 16.23009 2.813524 5.77 0.000 10.71509 21.74509
ind4 | -3.733479 4.951673 -0.75 0.451 -13.43963 5.972671
ind5 | 65.68343 14.43998 4.55 0.000 37.37853 93.98832
ind6 | 32.57879 2.70716 12.03 0.000 27.27228 37.88531
ind7 | 24.50673 10.75022 2.28 0.023 3.434401 45.57905
ind8 | 27.49329 18.51904 1.48 0.138 -8.807295 63.79388
ind9 | 28.20282 6.029533 4.68 0.000 16.38388 40.02177
ind10 | 35.62353 6.827263 5.22 0.000 22.2409 49.00617
lndistance | -5.228204 .8692839 -6.01 0.000 -6.932153 -3.524255
_cons | 201.3265 7.403358 27.19 0.000 186.8146 215.8384
-------------+----------------------------------------------------------------
/sigma | 66.16526 1.543082 63.14055 69.18998
------------------------------------------------------------------------------
0 left-censored observations
1,424 uncensored observations
9,781 right-censored observations at sought >= 100

[John Mullahy]

Mohd: These FAQ replies may be helpful.

https://www.stata.com/support/faqs/s...ons/index.html

https://www.stata.com/support/faqs/s...est/index.html

[Carlo Lazzaro]

Mohd:
welcome to the list.
Two comments about your query:
- why imposing -ll()- when your sample has no left-censored observation;
- the F-test you got is due to -robust- option. The main point is whether -robust- is necessary or not or, again, whether you typed -robust- but meant -cluster-.(on industry).

PS crossed in the cyberspace with John's helpful reply.

[Mohd Fahad]

Dear John and Carlo, Many thanks for the useful insights. I now understand the difference between chi-squared and F statistic and the reasons for showing up in the results.

One last question: I have to now additionally incorporate year and acquirer country dummies and use cluster option. I have included them and clustered my observations by both acquirer and target countries. I am now facing issues with the F statistic missing. I have read earlier posts and potential problem might be from singleton indicator for all values in the cluster. I then by trial and error removed those acquirer countries which did not result in the output and finally got the below output.

Can I report the model with few country fixed effects as shown in point 3 (after removing those which made the F-statistic missing)? When I use only vce(robust) option I can keep all the country dummies and the results are almost identical (point 4). I need to keep country fixed effects and also cluster using acquire-target country.

Request your kind attention and thoughts on the same

Thank you once again.
Regards
Fahad

1. I generated acqtargetcluster by egen acqtargetcluster = group (acqcountry target country)

2. Results with vce( cluster acqtargetcluster) and including year and acquirer country dummies

Tobit regression Number of obs = 11,205
F( 39, 11161) = .
Prob > F = .
Log pseudolikelihood = -10403.331 Pseudo R2 = 0.0314

(Std. Err. adjusted for 56 clusters in acqtargetcluster)
----------------------------------------------------------------------------------
| Robust
sought | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-----------------+----------------------------------------------------------------
ind1 | 28.95434 8.679032 3.34 0.001 11.9419 45.96677
ind2 | -.1804859 5.856522 -0.03 0.975 -11.6603 11.29933
ind3 | 11.7315 7.261456 1.62 0.106 -2.50223 25.96524
ind4 | -6.567614 7.925296 -0.83 0.407 -22.10259 8.967366
ind5 | 56.50434 12.34017 4.58 0.000 32.31542 80.69327
ind6 | 21.73909 4.478606 4.85 0.000 12.96023 30.51795
ind7 | -3.949759 11.94903 -0.33 0.741 -27.37197 19.47245
ind8 | 33.07513 5.150037 6.42 0.000 22.98015 43.17011
ind9 | 21.30043 6.045543 3.52 0.000 9.450094 33.15076
ind10 | 25.7181 7.796796 3.30 0.001 10.435 41.0012
acqaustria | -61.76405 13.00665 -4.75 0.000 -87.25939 -36.26871
acqbelgium | 9.776575 9.636192 1.01 0.310 -9.112063 28.66521
acqcanada | -23.19578 7.517801 -3.09 0.002 -37.93199 -8.459558
acqfrance | -55.41884 20.01478 -2.77 0.006 -94.65133 -16.18634
acqgermany | -44.4326 11.68696 -3.80 0.000 -67.34111 -21.52409
acqhongkong | -54.21855 4.542629 -11.94 0.000 -63.1229 -45.31419
acqindia | -7.022447 4.34194 -1.62 0.106 -15.53342 1.488523
acqireland | -9.40499 12.66857 -0.74 0.458 -34.23763 15.42765
acqnetherlands | -8.415209 5.096455 -1.65 0.099 -18.40516 1.574742
acqnorway | -35.05557 10.44927 -3.35 0.001 -55.53799 -14.57315
acqsweden | -12.45356 18.5191 -0.67 0.501 -48.75426 23.84713
acqswitzerland | -39.74898 12.3654 -3.21 0.001 -63.98735 -15.51061
acqunitedkingdom | -8.715045 11.40349 -0.76 0.445 -31.06791 13.63782
acqunitedstates | -11.26975 10.15913 -1.11 0.267 -31.18343 8.643935
d2012 | -13.21224 20.15347 -0.66 0.512 -52.7166 26.29212
d2011 | -16.3711 19.66075 -0.83 0.405 -54.90964 22.16744
d2010 | -16.70763 18.8091 -0.89 0.374 -53.57679 20.16153
d2009 | -17.2257 17.10712 -1.01 0.314 -50.75867 16.30726
d2008 | -10.96625 19.0312 -0.58 0.564 -48.27076 26.33827
d2007 | -11.50866 18.03035 -0.64 0.523 -46.85134 23.83401
d2006 | -7.578071 18.34721 -0.41 0.680 -43.54184 28.3857
d2005 | -5.936666 19.31347 -0.31 0.759 -43.79448 31.92114
d2004 | -15.87665 22.76666 -0.70 0.486 -60.50333 28.75003
d2003 | -17.58267 18.77463 -0.94 0.349 -54.38426 19.21892
d2002 | 26.55836 13.93764 1.91 0.057 -.7618748 53.87859
d2001 | 14.40257 13.66338 1.05 0.292 -12.38006 41.18521
d2000 | 9.92744 11.53522 0.86 0.389 -12.68363 32.53851
d1999 | 9.676738 13.07275 0.74 0.459 -15.94816 35.30163
d1998 | 11.92408 14.60971 0.82 0.414 -16.71354 40.5617
d1997 | 1.831815 13.61493 0.13 0.893 -24.85585 28.51947
d1996 | -16.58809 15.03139 -1.10 0.270 -46.05226 12.87609
d1995 | -21.41993 16.09653 -1.33 0.183 -52.97198 10.13211
d1994 | -3.471853 14.19769 -0.24 0.807 -31.30183 24.35813
lndistance | -7.050493 2.830292 -2.49 0.013 -12.59837 -1.50262
_cons | 240.3073 25.21831 9.53 0.000 190.875 289.7396
-----------------+----------------------------------------------------------------
/sigma | 63.17455 2.406201 58.45797 67.89113
----------------------------------------------------------------------------------
1 left-censored observation at sought <= .80000001
1,423 uncensored observations
9,781 right-censored observations at sought >= 100



3. Results after removing country dummies which had issues.


Tobit regression Number of obs = 11,205
F( 40, 11165) = 847.18
Prob > F = 0.0000
Log pseudolikelihood = -10419.294 Pseudo R2 = 0.0299

(Std. Err. adjusted for 56 clusters in acqtargetcluster)
----------------------------------------------------------------------------------
| Robust
sought | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-----------------+----------------------------------------------------------------
ind1 | 29.06006 8.643049 3.36 0.001 12.11816 46.00197
ind2 | .3142019 5.762334 0.05 0.957 -10.98099 11.60939
ind3 | 11.57932 6.865136 1.69 0.092 -1.877563 25.03619
ind4 | -6.184742 7.917967 -0.78 0.435 -21.70535 9.335871
ind5 | 57.28056 12.21621 4.69 0.000 33.33463 81.22649
ind6 | 24.39482 4.947533 4.93 0.000 14.69679 34.09286
ind7 | -3.139844 12.1148 -0.26 0.796 -26.88699 20.6073
ind8 | 33.91545 5.152372 6.58 0.000 23.81589 44.01501
ind9 | 19.26805 6.353419 3.03 0.002 6.814232 31.72188
ind10 | 31.75275 7.669721 4.14 0.000 16.71874 46.78675
d2012 | -15.88156 20.36737 -0.78 0.436 -55.8052 24.04207
d2011 | -19.32927 19.74672 -0.98 0.328 -58.03632 19.37778
d2010 | -19.70737 18.99779 -1.04 0.300 -56.94639 17.53166
d2009 | -20.07021 17.35691 -1.16 0.248 -54.09282 13.9524
d2008 | -12.27413 19.21017 -0.64 0.523 -49.92945 25.38119
d2007 | -13.40811 18.10778 -0.74 0.459 -48.90256 22.08633
d2006 | -9.133593 18.50975 -0.49 0.622 -45.41597 27.14878
d2005 | -6.883578 19.34061 -0.36 0.722 -44.79459 31.02743
d2004 | -15.51461 22.65038 -0.68 0.493 -59.91335 28.88413
d2003 | -17.85977 19.03003 -0.94 0.348 -55.16199 19.44245
d2002 | 24.84571 13.95725 1.78 0.075 -2.512958 52.20438
d2001 | 12.66007 13.80328 0.92 0.359 -14.3968 39.71694
d2000 | 8.367565 11.68156 0.72 0.474 -14.53035 31.26548
d1999 | 8.440533 13.23063 0.64 0.524 -17.49383 34.37489
d1998 | 11.11132 14.75312 0.75 0.451 -17.8074 40.03004
d1997 | 1.05147 13.7735 0.08 0.939 -25.94703 28.04997
d1996 | -17.61112 15.28407 -1.15 0.249 -47.57058 12.34835
d1995 | -22.52817 16.33913 -1.38 0.168 -54.55575 9.499409
d1994 | -4.153599 14.46469 -0.29 0.774 -32.50694 24.19974
acqaustria | -20.15628 15.35028 -1.31 0.189 -50.24554 9.932984
acqfrance | -14.69007 22.30578 -0.66 0.510 -58.41334 29.0332
acqgermany | -4.239732 15.50314 -0.27 0.784 -34.62863 26.14916
acqireland | 32.54998 15.99777 2.03 0.042 1.191531 63.90843
acqnetherlands | 32.92341 11.96311 2.75 0.006 9.473606 56.37321
acqsweden | 28.97731 20.4596 1.42 0.157 -11.12712 69.08173
acqcanada | 18.48951 11.13249 1.66 0.097 -3.332146 40.31116
acqaustralia | 41.33313 11.36618 3.64 0.000 19.05343 63.61284
acqunitedstates | 29.21843 14.17257 2.06 0.039 1.437693 56.99918
acqunitedkingdom | 32.21066 15.71346 2.05 0.040 1.409506 63.01181
lndistance | -7.031063 2.691279 -2.61 0.009 -12.30644 -1.755682
_cons | 200.507 26.4883 7.57 0.000 148.5852 252.4287
-----------------+----------------------------------------------------------------
/sigma | 63.46207 2.378156 58.80046 68.12367
----------------------------------------------------------------------------------
1 left-censored observation at sought <= .80000001
1,423 uncensored observations
9,781 right-censored observations at sought >= 100


4. Results by using vce(robust) option and keeping all country fixed effects. I additionally included target country effects also.


Tobit regression Number of obs = 11,205
F( 68, 11137) = 16.40
Prob > F = 0.0000
Log pseudolikelihood = -10247.004 Pseudo R2 = 0.0459

----------------------------------------------------------------------------------
| Robust
sought | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-----------------+----------------------------------------------------------------
ind1 | 17.60675 6.220079 2.83 0.005 5.414293 29.79921
ind2 | -1.780251 9.7778 -0.18 0.856 -20.94647 17.38597
ind3 | 13.69387 3.402735 4.02 0.000 7.023907 20.36383
ind4 | -17.68031 4.964616 -3.56 0.000 -27.41183 -7.948781
ind5 | 36.25254 14.0854 2.57 0.010 8.642655 63.86242
ind6 | 23.53088 3.304852 7.12 0.000 17.05278 30.00897
ind7 | -13.37999 10.6438 -1.26 0.209 -34.24372 7.483731
ind8 | 23.25823 18.13544 1.28 0.200 -12.29044 58.8069
ind9 | 18.71059 6.395468 2.93 0.003 6.174345 31.24684
ind10 | 18.4403 6.924812 2.66 0.008 4.866437 32.01415
d2012 | 12.17784 19.32223 0.63 0.529 -25.69716 50.05284
d2011 | 9.724545 19.09185 0.51 0.611 -27.69886 47.14795
d2010 | 10.37588 18.98669 0.55 0.585 -26.84139 47.59316
d2009 | 10.03909 18.91802 0.53 0.596 -27.04358 47.12177
d2008 | 16.48531 18.84039 0.87 0.382 -20.44519 53.41582
d2007 | 16.53518 18.73683 0.88 0.378 -20.19232 53.26268
d2006 | 21.56334 18.7994 1.15 0.251 -15.28682 58.4135
d2005 | 22.51404 18.97325 1.19 0.235 -14.6769 59.70497
d2004 | 18.50039 19.30023 0.96 0.338 -19.33148 56.33226
d2003 | 19.23361 20.50604 0.94 0.348 -20.96185 59.42908
d2002 | 44.62304 20.25301 2.20 0.028 4.923553 84.32254
d2001 | 26.71661 18.83186 1.42 0.156 -10.19716 63.63038
d2000 | 22.54283 18.57981 1.21 0.225 -13.87688 58.96253
d1999 | 24.15364 18.50886 1.30 0.192 -12.12701 60.43428
d1998 | 27.07952 18.55551 1.46 0.144 -9.292565 63.4516
d1997 | 19.14146 18.59698 1.03 0.303 -17.3119 55.59483
d1996 | 6.207391 18.56834 0.33 0.738 -30.18984 42.60462
d1995 | 1.687011 18.82078 0.09 0.929 -35.20504 38.57907
d1994 | 17.61164 20.41234 0.86 0.388 -22.40016 57.62345
acqaustria | 11.15834 15.12144 0.74 0.461 -18.48236 40.79903
acqbelgium | 23.89432 28.16497 0.85 0.396 -31.31399 79.10264
acqcanada | 12.02093 9.878076 1.22 0.224 -7.341847 31.38371
acqfrance | -1.872696 10.54916 -0.18 0.859 -22.55092 18.80553
acqgermany | -11.66487 10.38405 -1.12 0.261 -32.01944 8.689705
acqhongkong | 15.79963 10.49529 1.51 0.132 -4.772985 36.37225
acqindia | -9.25814 11.23303 -0.82 0.410 -31.27686 12.76058
acqireland | 25.5863 13.50301 1.89 0.058 -.8819906 52.05458
acqnetherlands | 29.08688 11.30369 2.57 0.010 6.92966 51.24411
acqnorway | 6.548332 22.22767 0.29 0.768 -37.02183 50.1185
acqsweden | 42.79438 27.64213 1.55 0.122 -11.38909 96.97786
acqswitzerland | 28.65409 16.24103 1.76 0.078 -3.181206 60.48939
acqunitedkingdom | 30.6734 9.331024 3.29 0.001 12.38294 48.96386
acqunitedstates | 29.91959 8.52109 3.51 0.000 13.21674 46.62243
tarargentina | -32.73895 10.72166 -3.05 0.002 -53.7553 -11.72261
taraustralia | -33.91809 7.697592 -4.41 0.000 -49.00674 -18.82945
tarbelgium | -7.350593 23.94713 -0.31 0.759 -54.2912 39.59001
tarbrazil | -42.34371 10.6525 -3.98 0.000 -63.2245 -21.46291
tarcanada | -2.821097 7.445981 -0.38 0.705 -17.41654 11.77434
tarchile | -25.12768 10.16996 -2.47 0.013 -45.06259 -5.192771
tarchina | -66.4153 6.41594 -10.35 0.000 -78.99168 -53.83892
tarcolombia | -31.80082 9.779485 -3.25 0.001 -50.97034 -12.6313
tardenmark | -67.38638 31.89078 -2.11 0.035 -129.898 -4.874797
tarfrance | -27.53874 6.494763 -4.24 0.000 -40.26963 -14.80786
targermany | -46.57653 4.819992 -9.66 0.000 -56.02457 -37.12849
tarindia | -57.31505 11.0548 -5.18 0.000 -78.98442 -35.64569
tarireland | -25.57103 17.23415 -1.48 0.138 -59.35301 8.210952
tarisrael | -40.98642 14.83866 -2.76 0.006 -70.07281 -11.90002
taritaly | -43.63606 11.75445 -3.71 0.000 -66.67686 -20.59526
tarmexico | -38.64645 4.995666 -7.74 0.000 -48.43884 -28.85406
tarnetherlands | -21.42557 9.927322 -2.16 0.031 -40.88488 -1.966265
tarnewzealand | 13.33233 13.76915 0.97 0.333 -13.65764 40.32231
tarnorway | -29.89871 30.36601 -0.98 0.325 -89.42146 29.62404
tarperu | -47.49353 10.67294 -4.45 0.000 -68.41437 -26.57268
tarsouthafrica | -78.08022 11.1277 -7.02 0.000 -99.89247 -56.26797
tarsweden | -15.40051 13.66073 -1.13 0.260 -42.17796 11.37693
tarswitzerland | -10.08412 11.444 -0.88 0.378 -32.51639 12.34814
tarunitedkingdom | -3.098985 4.903423 -0.63 0.527 -12.71056 6.512593
lndistance | 2.129481 2.115762 1.01 0.314 -2.017787 6.276749
_cons | 124.2466 30.26137 4.11 0.000 64.92892 183.5642
-----------------+----------------------------------------------------------------
/sigma | 61.01702 .9039511 59.24511 62.78892
----------------------------------------------------------------------------------
1 left-censored observation at sought <= .80000001
1,423 uncensored observations
9,781 right-censored observations at sought >= 100

[Mohd Fahad]

Dear Members,

Any thoughts on how to proceed. I need it a bit urgently.

Thanks
Fahad

[Nick Cox]

For general advice on bumping and specific advice on claims of urgency please see https://www.statalist.org/forums/help#adviceextras #1.