st: [Non Stata] Estimation strategy for a belief learning model.

[Antoine Terracol <Antoine.Terracol@univ-paris1.fr>]

Dear _all,

In the context of an economic experiment, I have data on the elicited 
beliefs of individuals on the probability that their opponent plays 
strategy A, B or C in the next round.

I want to fit a learning model (Cheung and Friedman (1997) "Individual 
Learning in Normal Form Games: Some Laboratory Results," Games and 
Economic Behavior, 19, 46-76.) on the data to see wether it describes 
reasonably well the learning process of individuals.

For a given strategy, say strategy A, I can estimate the parameters of 
the theoritical model by running the following maximuml likelihood 
regression:

Actual Belief = Theoretical Belief + epsilon

where epsilon is a disturbance random variable

However, because I only use the belief for a given strategy, I throw 
away valuable information (beliefs on strategy B)

I obvioulsy (??) can't simply multiply the likelihood contributions for 
both beliefs since they are almost mechanicaly negatively correlated; 
and I'm not sure wether using a bivariate distribution for the error 
terms would be appropriate.

I'd appreciate any hint or advice on the subject...

Best,
Antoine.



--
Ce message a �t� v�rifi� par MailScanner
pour des virus ou des polluriels et rien de
suspect n'a �t� trouv�.

*
*   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/