Re: [Gretl-devel] for translators

On Wed, 5 Apr 2017, Sven Schreiber wrote: > I think it's time to ditch the entry "Hausman test matrix is not > positive definite (this result may be treated as\n > \"fail to reject\" the random effects specification).\n" > > Although I have a 'classic' paper named "The Hausman test statistic > can be negative even asymptotically" which shows that that would > happen only if H0 is wrong, I have tolerated the existence of that > entry until now. > But for quite some time gretl also has had an alternative > regression-based Hausman test for Random Effects where the problem > cannot happen. So why do we still need this misleading gretl message? The matrix-diff method for the Hausman test is still an option (for comparability with other software). But OK, we can trash the parenthetical clause in the above entry.