On Sat, 16 Jun 2018, John C Frain wrote: > Perhaps there is some justification in model reduction as implemented in > David Hendry's Pc-gets or in the Grocer package in Scilab. This does use a > form of model reduction but it is very restricted. Otherwise what are the > benefits of systems of stepwise regression. If you do arrive a a somewhat > sensible looking answer you can make no structural econometric inference. > How do you overcome the problems mentioned by Clive Nicolas. You then > need a completely new and independent data set to make statistical > inferences on the new theory. Perhaps you are simply interested in data > reduction. > > Can you give me one good example in which a stepwise regression program > produces results that are not subject to the objections mentioned by > Clive. OLS, VARS etc may be abused by some but such routines are the basis > of much good work. Stepwise regression is not.
On a personal level, I agree 100%. As Clive remarked, the statistical issues implied by automatic model selection are indeed thorny. The only convincing framework in which you can (to some extent) substitute human judgement with CPU cycles is BMA, but apart from that I myself would never base any critical statistical procedure on tools like stepwise regression. Having said that, however, I believe that the purpose of the gretl project is not to tell the world how econometrics should be done. I've never liked those schools of econometric thought that lend themselves to almost religious allegiance: I'm old enought to remember (and shudder at) the zeal of some converts to the LSE approach back in the 80s and I'm frankly puzzled by the gospel-like status a certain not-so-harmless book has gained in recent years. In my opinion, our goal is to provide a tool for doing econometrics, whose two main features are 1) political: gretl is free (in the "libre" sense: see https://en.wikipedia.org/wiki/Gratis_versus_libre) 2) technical: gretl should offer the best possible compromise between ease of use and technical efficiency (where the word "best" is of course open to interpretation). One of the consequences of the points above is that allowing people to make mistakes is part of our job. Sometimes you will want to do the wrong thing: when teaching, for example, to expose fallacies. Or, in a research context, you may want to replicate other people results, and do so by using free tools. Gretl must be up to the job, and therefore I'm all in favour of implementing statistical techniques I would never trust myself. ------------------------------------------------------- Riccardo (Jack) Lucchetti Dipartimento di Scienze Economiche e Sociali (DiSES) Università Politecnica delle Marche (formerly known as Università di Ancona) r.lucchetti(a)univpm.it http://www2.econ.univpm.it/servizi/hpp/lucchetti -------------------------------------------------------
