How should I describe my regression results?

[odovakar]

I have a slight dilemma. I have estimated a total of 6 panel data equations. According to the Hausman test that I conducted in Stata (I am still waiting to do the same in Eviews), 5 of them were correctly specified with a fixed effects model. However, according to Murphy's Law, of course the Hausman test had to indicate that the last one would be better estimated with a dynamic effects model. The problem is that I would still prefer to use the results from the fixed effects model in my table of regression results, constructed in Word. The simple reason is that some of the estimated parameters in the dynamic model equivalent of the last estimation were not significant.

Therefore: What would you have done? In the table, would you have accounted for the fixed effects model for the 6th case as well (i.e. by showing standard deviation, t statistics and so on in proper order) or would you have introduced a new column or something like that, and then presented the results of the dynamic model?

Or perhaps I should present the results from the fixed effects regression in the table and then put an asterisk (alternatively a "D") to indicate that a more proper specification in that very case would be a dynamic effects model?

Does anyone know what the standard procedure would be in a case such as this?