--- Hanno Scholtz <[email protected]> wrote:
> year was only an example - if I linearly transform any other variable
> (e.g. logged GDP), the complete regression results change as well.
Logging a variable is not a linear transformation, so this
transformation might lead to a better (or worse) way of contolling the
effectss of the other variables for differences in GDP, and can thus
lead to differences in the other estimates.
If you have a model that is somewhat unstable I would suggest making
sure that for every variable the zero is within or close to the
observed range, and scale the variables in such a way that the sizes of
the effects are not too different, i.e. don't add GDP in euros, but
ln(GDP/10^12), consider measuring change over time in decades since
1950 instead of years since 0, etc.
Sometimes models with discrete dependent variables get unstable if one
of the categories becomes very small. This can become a problem if you
have many categories.
There are other operationalizations of democracy than those provided by
Freedom House. As a robustness check I would also look at:
http://www.cidcm.umd.edu/polity/data/
http://www.prio.no/page/Project_detail//9649/42472.html
Hope this helps,
Maarten
-----------------------------------------
Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands
visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434
+31 20 5986715
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
__________________________________________________________
Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.com
*
* 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/
