Re: [Gretl-users] Goldfeld-Quandt heteroscedasticity test