Limited-Dependent and Qualitative Variables in Econometrics (Econometric Society Monographs) (Paperback) by G. S. Maddala, Andrew Chesher (Series Editor), Matthew Jackson (Series Editor)

Paperback: 401 pages

Publisher: Cambridge University Press; Reprint edition (June 27, 1986)

ISBN: 0521338255

Customers who bought this book also bought

This book surely is a classic treatment of limited dependent variable models. However, I suspect most researchers, and definitely almost all graduate stduents will NOT need this book. 1) First, it is VERY outdated. Note that this book has been written in 1983. That is more than *20* years ago. There is nothing here on nonpametric and semiparametric models, which are becoming the standard in this field. For this, a good reference would be Pagan and Ullah, which also contain results for nonparametric and semiparametric general density and regression estimation. The fact that this book is outdated not only means that lots and lots of important tools are not covered here, but it also means that it covers many results and estimators that are no longer considered important. 2) Second, and especially if you are a graduate student, all or almost all the results you will need for your courses is covered very well in textbooks that also cover a wealth of other materials (such as Wooldridge, Amemiya, Ruud, Davidson and McKinnon etc.).

[此贴子已经被作者于2005-4-20 13:58:38编辑过]

不过现在许多计量的书对于受限因变量的介绍相对比较少，特别是在实证研究中遇到的问题这些教科书上都不讲。

不如一个联立方程，其中一个内生变量是0，1变量，在这种情况下如何利用软件去估计?

等等问题，理论上容易，实践上比较困难。

在顶一下，看看谁有啊

I do not agree with Hanszhu's points. The basic principle in econometrics is to find the best estimators for DGP (data generating process). New techniques may be developed over the years, however, Maddals's book paves the way for learning new methods. The main advantage of this book is its well-organized contents. Readers can learn these methods systematically.