Interpreting results of MMSEL

[stanzin.lama]

Hello everyone.

I am working with the command MMSEL in Stata 13.0 found here: http://fmwww.bc.edu/repec/bocode/m/mmsel.html. The command was developed as per the Machada and Moto (2005) methodology to decompose discrimination. The command I used was as follows:

mmsel lnfee totincome poor floor hhsize fedufinal medufinal ethni Currentclass Age Schooltype distschool rural, group(Sexnew) filename(ooo)

I am having problem interpreting the results. The results look like this:

-----------------------------------------------------------------------------------------------------------------------------
name: <unnamed>
log: C:\Users\hp\Documents\sawtee\2016\Decomposition\da ta\results/ooo.log
log type: text
opened on: 5 Apr 2016, 11:49:34
olsdiff = -.17324434
se_olsdiff = .04835614

+-------------------------------------------------------------------------------------------------------------------+
rawgap pred1 se_pred1 chars1 se_cha~1 coef1 se_coef1 |
|---------------------------------------------------------------------------------------------------------------------|
1. | -.1022787 -.1776473 .0426857 -.0036508 .0480825 -.1739965 .0489399 |
2. | -.1388364 -.1491552 .0342681 -.034267 .0370442 -.1148882 .0346013 |
3. | -.1329927 -.1396689 .0294674 -.0352226 .0325142 -.1044464 .0283426 |
4. | -.1335316 -.1298484 .0244935 -.040792 .0276894 -.0890564 .0260383 |
5. | -.1612682 -.1179296 .0230237 -.0464047 .023858 -.0715248 .0244542 |
|---------------------------------------------------------------------------------------------------------------------|
6. | -.1549225 -.1114721 .0215768 -.0494378 .023289 -.0620343 .0230374 |
7. | -.1508231 -.1019103 .0204266 -.0476519 .0222666 -.0542584 .0224477 |
8. | -.1541505 -.1016649 .020854 -.0498677 .0216903 -.0517972 .0218829 |
9. | -.119369 -.1025686 .0194939 -.0553659 .0206565 -.0472027 .0230392 |
10. | -.1343126 -.1048271 .0200643 -.0600594 .0210562 -.0447678 .0228864 |
|----------------------------------------------------------------------------------------------------------------------|
11. | -.1476917 -.1174552 .0188654 -.0661965 .0218921 -.0512588 .0232788 |
12. | -.1662807 -.127436 .0193686 -.0739177 .0224647 -.0535182 .0224644 |
13. | -.1972799 -.1408008 .0207176 -.0734816 .0227485 -.0673192 .0221677 |
14. | -.2406569 -.1663471 .0237409 -.0879007 .0260754 -.0784464 .0253634 |
15. | -.2891645 -.1942958 .0264329 -.1094988 .028164 -.084797 .0245965 |
|---------------------------------------------------------------------------------------------------------------------|
16. | -.3235483 -.225695 .0333931 -.1433594 .0321975 -.0823356 .0325428 |
17. | -.1720552 -.2667316 .0410306 -.1873217 .0421963 -.0794098 .0382037 |
18. | -.1785908 -.2791542 .0552776 -.2696151 .0536711 -.0095391 .0535282 |
19. | -.0677624 -.1475457 .0603991 -.1426838 .0652772 -.0048619 .0548883 |
20. | . . . . . . . |
+---------------------------------------------------------------------------------------------------------------------+
(81 real changes made, 81 to missing)
name: <unnamed>
log: C:\Users\hp\Documents\sawtee\2016\Decomposition\da ta\results/ooo.log
log type: text
closed on: 5 Apr 2016, 11:49:34
-----------------------------------------------------------------------------------------------------------------------------

Please pardon my poor table editing skills but I can't upload the log file here for some reason. If I am not mistaken the pred1 represents the total gap, the chars1 represents the gaps explained by differences explanators, and the ceof1 represents the gaps explained by coefficients. What I am not sure is what do the serial numbers at the first column (1, 2, 3, .... 20). My interest is primarily in quantiles 20, 40, 60 and 80 and I can't figure out which rows represent coefficients for the said quantiles.

With best regards
Lama

[ayeb moez]

I am working with the command MMSEL in Stata 13.0 found here: http://fmwww.bc.edu/repec/bocode/m/mmsel.html. The command was developed as per the Machada and Moto (2005) methodology to decompose discrimination. I am having problem when I execute the command, every time the error message " xmbm already defined " appears. Please help me

[Nick Cox]

The program is billed as written by

Sami Souabni, Swansea University, [email protected], who bears no responsibility for any errors.

That really is a very bad sign.

The package is a set of several programs together with plugins, precisely the situation in which suggesting what is going on is most difficult.

[ayeb moez]

thank you a lot