[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

# st: [Fwd: re: simultaneous equations]

 From mmolina@uniroma3.it To statalist@hsphsun2.harvard.edu Subject st: [Fwd: re: simultaneous equations] Date Sun, 8 Mar 2009 16:42:13 +0100 (CET)

```Dear Nick,
Thanks again. The problem were the outliers. I decided to work with
classes and it worked.
Now I have another problem: as I add more variables, I received the
message "insufficient observations". Could you please tell me how to deal
with this?
Regards,
Alejandra Molina

A Stata answer to this is to look at the help for -if-, which gives
examples of compound conditions.

Otherwise, omitting outliers is in my view only the best way to proceed
if one or more of the following apply:

1. You conclude that they are unbelievable (e.g. some gross error in
measurement or recording) and cannot replace them with better values.

2. On sincere reflection you decide they are outside your project.

3. You are comparing a model with and without outliers and wish to
assess sensitivity.

Nick
n.j.cox@durham.ac.uk

mmolina@uniroma3.it

Thank you very much Nick.
In fact there were some substantial outlier. The model seems to work in
analyzed by sectors:
reg3 (normscore p14 p4_1b skilled trained p71a expshare) (expshare p4_1b
p14 age normscore skilled)if ind_code=="Autoparts"
Now I would like to add another condition. That is separate within the
sector the big outliers. How can I add a second condition?
Regards,
Alejandra Molina

With coefficients of the order of 1e8 there seems to be a massive
problem with the data. Look (again?) at the data, e.g. using a scatter
plot matrix. There could be a substantial outlier, or some built-in
dependency, or something else. Whatever it is, it is not obvious that
any model can make sense of the predictor set you are using.

Nick
n.j.cox@durham.ac.uk

Dear Statlist,
Here I am. As you adviced me I tried ivregress 2sls ... and obtained:

ivregress 2sls expshare b2b l1 age l4a (normscore = l11a l10),
note: l10 dropped because of collinearity

Instrumental variables (2SLS) regression               Number of obs =
198
Wald chi2(5)  =
40.30
Prob > chi2   =
0.0000
R-squared     =
.
Root MSE      =
7.9e+07

------------------------------------------------------------------------------
expshare |      Coef.   Std. Err.      z    P>|z|     [95% Conf.
Interval]
-------------+----------------------------------------------------------------
normscore |   4.04e+08   3.93e+08     1.03   0.304    -3.66e+08
1.18e+09
b2b |  -238087.7     387449    -0.61   0.539    -997473.9
521298.5
l1 |   66040.49   14947.01     4.42   0.000     36744.89
95336.09
age |   134492.7   220864.6     0.61   0.543    -298393.9
567379.3
l4a |  -95654.77   20067.76    -4.77   0.000    -134986.9
-56322.68
_cons |  -1.33e+08   1.18e+08    -1.13   0.260    -3.65e+08
9.85e+07
------------------------------------------------------------------------------
Instrumented:  normscore
Instruments:   b2b l1 age l4a l11a

I agree with and don't know why the constant term is dropped.
I was looking for some causality link between normscore and expshare, and
supossed Simultaneous Equations may help me.

Thanks and regards,

Alejandra Molina

st: re: simultaneous equations

MMolina said

I run this command:

reg3 (normscore l1 b2b age l4a l11a l10) (expshare b2b l1 age normscore l4a)

And obtained this results. Could you please tell me if the Std. Error

found for the expshare variable may invalidate the analysis or create any
robustness problem?

I find it very strange, if expshare is really a fraction, that you get
coefficients on the order of 10^8 on any regressor. Would you try running

ivreg expshare b2b l1 age l4a (normscore = l11a l10), first

and show us what you get from that regression? (you could also use
ivregress 2sls ... , first if you are using Stata 10)

Also, is there a good reason (given definitions of the variables) that the
constant term is dropped in your first equation?

This system is triangular, in that normscore can be estimated with OLS as
it does not contain expshare. You may still want to estimate it with 3SLS,
but something fishy is going on here. See what the single- equation
estimates of your second equation reveal.

Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html

*
*   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/
```

 © Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index