Hi Carlos,

In order to be a valid an instrument, your input should meet several
assumptions - not just be statistically significant correlated with your
endogenous regressor. However, most of these assumptions are not testable
but sometimes defensible on theoretical grounds. The most accessible and
clear introduction to the subject of regression and causal inference,
including IV,  that I know of is Gelman, chapters 9 and 10 (with  lots of R
code and further references).

http://www.stat.columbia.edu/~gelman/arm/

All the best,



Antonio.

**********************************************************
             Political Methodology E-Mail List
   Editors: Melanie Goodrich, <[log in to unmask]>
            Delia Bailey, <[log in to unmask]>
**********************************************************
        Send messages to [log in to unmask]
  To join the list, cancel your subscription, or modify
           your subscription settings visit:

          http://polmeth.wustl.edu/polmeth.php

**********************************************************