期刊名称：South-Eastern Europe Journal of Economics (SEEJE)
印刷版ISSN：1109-8597
电子版ISSN：1792-3115
出版年度：2007
卷号：5
期号：1
出版社：Association of Economic Universities of South and Eastern Europe and the Black Sea Region
摘要：Although there is a lot of econometrics around - perhaps more than the average economist can digest! - a book on model selection and model choice was really badly needed. I am not sure whether this book fills the void but, nevertheless, this is a book the applied economist can benefit from. The central question is old and simple. Many models have been proposed, and we must choose the “best” one. How do we proceed? This is a central question in most empirical analyses. The applied economist can benefit from the good review of model selection criteria in chapter 4. This is a superb exposition of all information criteria currently used in practice (Akaike, Schwarz etc.). I found this the best part of the book. The rest of it is more or less standard econometric theory and the authors’ own empirical work. There is some new material in chapter 5, regarding information criteria in Box-Cox models and Monte Carlo simulations, which is particularly interesting and deserves some attention from the theoretical econometrician. The many tables might, however, discourage the applied economist, and the main points are not easy to digest. Chapter 6 continues in the same spirit and could have been omitted altogether.
