Sigert Ariens

[Sigert Ariens @AriensSigert]

@kenmcalinn @janne_adolf @EikoFried BIC only imposes assumptions to the extent that the model it is based on imposes assumptions. For the regression models we consider in the paper (iid normal errors, v_t), BIC can be used. It is furthermore accessible for our target audience, as are the LR tests we also discuss.