Re: st: biasness in panel FE

[Maarten Buis <maartenlbuis@gmail.com>]

On Tue, Dec 18, 2012 at 9:04 AM, Isabella King wrote:
> I am currently working with panel fixed effects model. My dataset has 176
> obs. My question is : whether this small sample size will result in biasness
> in estimation results, especially when I try to use IV to instrument one
> endogenous variable?

The first step is to specify what it is exactly that you want to
estimate. Remember that bias is just a shorthand for "the coefficient
does not estimate what I want to know". So whether or not a
coefficient is biased depends on what it is  it is supposed to
estimate. The problem is that a statement like "I just want to know
the effect of x on y" is often not precise enough as it can be
interpreted in different ways and these different ways often matter as
they lead to different conclusions. What I find often helpful is to
think in terms of two (hypothetical) groups and the effect is some
comparison (difference or ratio) of means or medians or ... of the
dependent variable. So the trick is to think carefully about what
defines these groups and what is it that you want to compare.

After that you can look at your data and possible methods and see how
each approximate the thing you want to know. Typically, you'll never
get there entirely. The trick is not to try to give the "final
answer", but instead a contribution to the scientific debate where you
add a piece of imperfect evidence to the body of evidence available in
this debate.

Hope this helps,
Maarten

---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/