Am 06.12.18 um 09:40 schrieb Olasehinde Timmy: Hi, I suspect the reason you haven't gotten an answer is that these questions seem unrelated to gretl. What commands or functions are you using for PMG and MG? With respect to unit root tests, more data is of course better. But you "can" conduct the test. Off the top of my head I cannot tell you whether there are better small-sample approximations for the test distributions (critical values) in the literature. It's best to ask questions on this list in relation to some gretl command or to clarify some gretl output. cheers, sven