Dear _all,
I am fitting the same model to cases and controls and
would like to test if the estimated coefficients in
the two models are the same. I think I can use suest
for this but the test results appear to be to
conservative. This is what I am doing:
xi:regress homocum i.alle vitb1p1 drkb1p1_1
smkb1p1_1 i.m_race mage if case==0
noi est store Controls
xi:regress homocum i.alle vitb1p1 drkb1p1_1
smkb1p1_1 i.m_race mage if case==1
noi est store Cases
matrix A=e(b)
local x: colnames A
suest Controls Cases
local y=substr("`x'",1,length("`x'")-5)
foreach var in `x' {
noi test [Controls_mean]`var'=[Cases_mean]`var'
}
Results are found after my signature. I particularly
question the results obtained for the -alle- variable.
Is this correct or am I using suest incorrectly?
Regards,
Ricardo.
i.alle _Ialle_0-1 (naturally
coded; _Ialle_0 omitted)
i.m_race _Im_race_0-3 (naturally
coded; _Im_race_0 omitted)
Source | SS df MS
Number of obs = 153
-------------+------------------------------
F( 8, 144) = 0.96
Model | 23.0809151 8 2.88511438
Prob > F = 0.4699
Residual | 432.835163 144 3.00579974
R-squared = 0.0506
-------------+------------------------------
Adj R-squared = -0.0021
Total | 455.916078 152 2.99944788
Root MSE = 1.7337
------------------------------------------------------------------------------
homocum | Coef. Std. Err. t P>|t|
[95% Conf. Interval]
-------------+----------------------------------------------------------------
_Ialle_1 | .1580478 .4602291 0.34 0.732
-.7516295 1.067725
vitb1p1 | -.0142011 .3059768 -0.05 0.963
-.6189871 .5905849
drkb1p1_1 | -.61993 .3360724 -1.84 0.067
-1.284202 .0443424
smkb1p1_1 | .3558431 .3637287 0.98 0.330
-.363094 1.07478
_Im_race_1 | .1907797 .390875 0.49 0.626
-.5818141 .9633736
_Im_race_2 | -.5546023 .6097384 -0.91 0.365
-1.759796 .6505914
_Im_race_3 | .9383422 1.780466 0.53 0.599
-2.580882 4.457566
mage | .0561531 .029505 1.90 0.059
-.0021656 .1144719
_cons | 6.118908 .8357678 7.32 0.000
4.466951 7.770866
------------------------------------------------------------------------------
i.alle _Ialle_0-1 (naturally
coded; _Ialle_0 omitted)
i.m_race _Im_race_0-3 (naturally
coded; _Im_race_0 omitted)
Source | SS df MS
Number of obs = 337
-------------+------------------------------
F( 8, 328) = 1.62
Model | 72.7850755 8 9.09813444
Prob > F = 0.1171
Residual | 1838.43973 328 5.60499916
R-squared = 0.0381
-------------+------------------------------
Adj R-squared = 0.0146
Total | 1911.2248 336 5.68816905
Root MSE = 2.3675
------------------------------------------------------------------------------
homocum | Coef. Std. Err. t P>|t|
[95% Conf. Interval]
-------------+----------------------------------------------------------------
_Ialle_1 | .9186363 .3813954 2.41 0.017
.1683466 1.668926
vitb1p1 | -.2829661 .265222 -1.07 0.287
-.8047168 .2387847
drkb1p1_1 | .1332106 .2952119 0.45 0.652
-.447537 .7139583
smkb1p1_1 | .4791066 .3093139 1.55 0.122
-.1293827 1.087596
_Im_race_1 | .0411555 .384826 0.11 0.915
-.715883 .7981939
_Im_race_2 | -.1755515 .4548408 -0.39 0.700
-1.070325 .7192217
_Im_race_3 | .9152593 2.38543 0.38 0.701
-3.777413 5.607932
mage | .0470688 .0236712 1.99 0.048
.0005024 .0936353
_cons | 7.77337 .7035925 11.05 0.000
6.389247 9.157494
------------------------------------------------------------------------------
Simultaneous results for Controls, Cases
Number of obs = 490
------------------------------------------------------------------------------
| Robust
| Coef. Std. Err. z P>|z|
[95% Conf. Interval]
-------------+----------------------------------------------------------------
Controls_m~n |
_Ialle_1 | .1580478 .3431422 0.46 0.645
-.5144986 .8305942
vitb1p1 | -.0142011 .3015033 -0.05 0.962
-.6051367 .5767345
drkb1p1_1 | -.61993 .3593731 -1.73 0.085
-1.324288 .0844284
smkb1p1_1 | .3558431 .4023459 0.88 0.376
-.4327404 1.144427
_Im_race_1 | .1907797 .374775 0.51 0.611
-.5437659 .9253253
_Im_race_2 | -.5546023 .4616476 -1.20 0.230
-1.459415 .3502104
_Im_race_3 | .9383422 .4126832 2.27 0.023
.1294979 1.747186
mage | .0561531 .031217 1.80 0.072
-.0050311 .1173374
_cons | 6.118908 .7973963 7.67 0.000
4.55604 7.681776
-------------+----------------------------------------------------------------
Controls_l~r |
_cons | 1.100544 .1145661 9.61 0.000
.8759982 1.325089
-------------+----------------------------------------------------------------
Cases_mean |
_Ialle_1 | .9186363 .4902185 1.87 0.061
-.0421743 1.879447
vitb1p1 | -.2829661 .2579176 -1.10 0.273
-.7884753 .2225432
drkb1p1_1 | .1332106 .3097102 0.43 0.667
-.4738102 .7402315
smkb1p1_1 | .4791066 .2928241 1.64 0.102
-.0948182 1.053031
_Im_race_1 | .0411555 .3700614 0.11 0.911
-.6841516 .7664625
_Im_race_2 | -.1755515 .444215 -0.40 0.693
-1.046197 .6950939
_Im_race_3 | .9152593 .2957542 3.09 0.002
.3355917 1.494927
mage | .0470688 .0227471 2.07 0.039
.0024853 .0916523
_cons | 7.77337 .6455117 12.04 0.000
6.508191 9.03855
-------------+----------------------------------------------------------------
Cases_lnvar |
_cons | 1.723659 .079775 21.61 0.000
1.567303 1.880015
------------------------------------------------------------------------------
( 1) [Controls_mean]_Ialle_1 - [Cases_mean]_Ialle_1
= 0
chi2( 1) = 1.62
Prob > chi2 = 0.2037
( 1) [Controls_mean]vitb1p1 - [Cases_mean]vitb1p1 =
0
chi2( 1) = 0.46
Prob > chi2 = 0.4982
( 1) [Controls_mean]drkb1p1_1 -
[Cases_mean]drkb1p1_1 = 0
chi2( 1) = 2.52
Prob > chi2 = 0.1124
( 1) [Controls_mean]smkb1p1_1 -
[Cases_mean]smkb1p1_1 = 0
chi2( 1) = 0.06
Prob > chi2 = 0.8044
( 1) [Controls_mean]_Im_race_1 -
[Cases_mean]_Im_race_1 = 0
chi2( 1) = 0.08
Prob > chi2 = 0.7763
( 1) [Controls_mean]_Im_race_2 -
[Cases_mean]_Im_race_2 = 0
chi2( 1) = 0.35
Prob > chi2 = 0.5541
( 1) [Controls_mean]_Im_race_3 -
[Cases_mean]_Im_race_3 = 0
chi2( 1) = 0.00
Prob > chi2 = 0.9637
( 1) [Controls_mean]mage - [Cases_mean]mage = 0
chi2( 1) = 0.06
Prob > chi2 = 0.8141
( 1) [Controls_mean]_cons - [Cases_mean]_cons = 0
chi2( 1) = 2.60
Prob > chi2 = 0.1068
Ricardo Ovaldia, MS
Statistician
Oklahoma City, OK
______________________________________________________
Yahoo! for Good
Donate to the Hurricane Katrina relief effort.
http://store.yahoo.com/redcross-donate3/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
