 |
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: RE: nestreg discrepancy
<>
I cannot see the "discrepancy". -nestreg- adds covariates one after another,
and compares the models via LR, while -regress- adds them all at once and
gives you a p-value for the test whether the population parameter is zero.
Also note that for -nestreg-, the order of the -varlist- supplied to it
matters, while for -regress- it does not...
*************
webuse census4, clear
nestreg, lr qui: regress brate (medage) (medagesq) (reg2) (reg4)
regress brate medage medagesq reg2 reg4
nestreg, lr qui: regress brate (medage) (medagesq) (reg4) (reg2)
regress brate medage medagesq reg4 reg2
*************
HTH
Martin
-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Eduardo Nunez
Sent: Donnerstag, 3. Dezember 2009 19:54
To: [email protected]
Subject: st: nestreg discrepancy
Hi everyone,
I apologize if this is a naive question, but what I understand is that
the LR test p-value should be close to the p-value obtained from t (or
wald statistic).
That's why I wonder the huge discrepancy I got (the likelihood ratio
test p-value estimated with nestreg differs so dramatically from the t
statistic p-value from regress).
Here are the results (see block 5=fe_i and block 6=lnai....both
variables are continuous):
. xi: regress fc_i edad tas_i tad_i lnhb fe_i lnai lnlinfos lnpcr
ingprevio i.sexo*i.acxfa hta bcrdhh diuretico_previo iecaprevia
bbloq_previo tnihigh
i.sexo _Isexo_0-1 (naturally coded; _Isexo_0 omitted)
i.acxfa _Iacxfa_0-1 (naturally coded; _Iacxfa_0 omitted)
i.sexo*i.acxfa _IsexXacx_#_# (coded as above)
Source | SS df MS Number of obs =
1041
-------------+------------------------------ F( 18, 1022) =
31.23
Model | 290798.843 18 16155.4913 Prob > F =
0.0000
Residual | 528729.619 1022 517.347964 R-squared =
0.3548
-------------+------------------------------ Adj R-squared =
0.3435
Total | 819528.462 1040 788.008136 Root MSE =
22.745
----------------------------------------------------------------------------
--
fc_i | Coef. Std. Err. t P>|t| [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
edad | -.2057782 .0710256 -2.90 0.004 -.345151
-.0664054
tas_i | -.0866522 .0296956 -2.92 0.004 -.1449235
-.0283809
tad_i | .3368084 .0533858 6.31 0.000 .23205
.4415668
lnhb | 232.2925 62.88337 3.69 0.000 108.8973
355.6878
fe_i | -.1953163 .0531564 -3.67 0.000 -.2996245
-.0910081
lnai | -15.30499 4.041561 -3.79 0.000 -23.2357
-7.374289
lnlinfos | 4.467231 1.00385 4.45 0.000 2.497387
6.437075
lnpcr | 2.542625 .8110956 3.13 0.002 .9510214
4.134228
ingprevio | -5.532232 1.636256 -3.38 0.001 -8.743038
-2.321426
_Isexo_1 | -.7361195 2.043631 -0.36 0.719 -4.746311
3.274072
_Iacxfa_1 | 26.54113 2.159557 12.29 0.000 22.30346
30.7788
_IsexXacx_~1 | -7.450133 2.879739 -2.59 0.010 -13.10101
-1.799257
hta | -3.516089 1.90403 -1.85 0.065 -7.252343
.220166
bcrdhh | -5.498154 2.602638 -2.11 0.035 -10.60528
-.3910296
diuretico_~o | -5.002201 1.751223 -2.86 0.004 -8.438604
-1.565798
iecaprevia | -3.147983 1.555494 -2.02 0.043 -6.20031
-.0956568
bbloq_previo | -9.443874 1.706497 -5.53 0.000 -12.79251
-6.095236
tnihigh | 6.651076 1.666292 3.99 0.000 3.381331
9.920821
_cons | -1037.17 314.5672 -3.30 0.001 -1654.442
-419.8988
----------------------------------------------------------------------------
--
. xi: nestreg, lr quietly: regress fc_i edad tas_i tad_i lnhb fe_i
lnai lnlinfos lnpcr ingprevio i.sexo*i.acxfa hta bcrdhh
diuretico_previo iecaprevia bbloq_previo tnihigh
i.sexo _Isexo_0-1 (naturally coded; _Isexo_0 omitted)
i.acxfa _Iacxfa_0-1 (naturally coded; _Iacxfa_0 omitted)
i.sexo*i.acxfa _IsexXacx_#_# (coded as above)
Block 1: edad
Block 2: tas_i
Block 3: tad_i
Block 4: lnhb
Block 5: fe_i
Block 6: lnai
Block 7: lnlinfos
Block 8: lnpcr
Block 9: ingprevio
Block 10: _Isexo_1
Block 11: _Iacxfa_1
Block 12: _IsexXacx_1_1
Block 13: hta
Block 14: bcrdhh
Block 15: diuretico_previo
Block 16: iecaprevia
Block 17: bbloq_previo
Block 18: tnihigh
+----------------------------------------------------------------+
| Block | LL LR df Pr > LR AIC BIC |
|-------+--------------------------------------------------------|
| 1 | -4940.801 14.58 1 0.0001 9885.603 9895.499 |
| 2 | -4935.84 9.92 1 0.0016 9877.68 9892.524 |
| 3 | -4889.166 93.35 1 0.0000 9786.332 9806.123 |
| 4 | -4879.5 19.33 1 0.0000 9769.001 9793.74 |
| 5 | -4879.492 0.02 1 0.8948 9770.983 9800.671 |
| 6 | -4879.02 0.94 1 0.3313 9772.039 9806.675 |
| 7 | -4875.408 7.22 1 0.0072 9766.815 9806.399 |
| 8 | -4868.665 13.48 1 0.0002 9755.331 9799.862 |
| 9 | -4857.788 21.75 1 0.0000 9735.576 9785.055 |
| 10 | -4854.467 6.64 1 0.0100 9730.934 9785.361 |
| 11 | -4765.272 178.39 1 0.0000 9554.545 9613.92 |
| 12 | -4761.665 7.22 1 0.0072 9549.329 9613.653 |
| 13 | -4757.121 9.09 1 0.0026 9542.243 9611.514 |
| 14 | -4754.762 4.72 1 0.0298 9539.524 9613.743 |
| 15 | -4746.742 16.04 1 0.0001 9525.484 9604.651 |
| 16 | -4743.267 6.95 1 0.0084 9520.533 9604.648 |
| 17 | -4728.035 30.46 1 0.0000 9492.071 9581.134 |
| 18 | -4719.984 16.10 1 0.0001 9477.968 9571.978 |
+----------------------------------------------------------------+
Regards,
Eduardo
*
* 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/
*
* 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/