I am looking for a few good books pitched at different levels to help analyse discrete data.
Specifically: Specification, Estimation and the application of econometric methods to model discrete choices by individuals, households or firms.
$\begingroup$
$\endgroup$
1
-
2$\begingroup$ Maybe you could change the title to better reflect the specificity of the question? Maybe something like this: "Book recommendation for the econometrics of discrete choice models." $\endgroup$– jmbejaraFeb 25, 2015 at 9:13
Add a comment
|
3 Answers
$\begingroup$
$\endgroup$
1
Personally, for choice analysis, I like
- Discrete Choice Methods with Simulation by Ken Train (pdf)
- Applied Choice Analysis: A Primer by Hensher and Greene (2nd edition, book)
- Modeling Ordered Choices by Hensher and Greene (pdf)
- Regression Models for Categorical Dependent Variables Using Stata by Long and Freese (3rd edition, book)
(1) is a fairly short, readable classic. (2) is much more introductory, with lots of intuition, though it devotes a lot of space to the nlogit software. (3) has the same features as (2), but for a more focused set of topics. (4) is also pretty introductory and focuses on using Stata. It is perhaps the least "economisty" of the four, but still quite good.
For count data, I like Cameron and Trivedi's count data book, followed by Winkelmann's.
The books you have listed are good, particularly Wooldridge, but they cover much more ground than just discrete data at the expense of depth.
-
$\begingroup$ Great resources. Worth pointing at that 1) and 3) are available for free at the links provided. Thanks $\endgroup$– JamzyFeb 25, 2015 at 21:55
$\begingroup$
$\endgroup$
2
I would suggest to look also into Alan Agresti's "Categorical Data Analysis".
Here, you can see detailed Contents, and get a sense of the level of the book. Here, a user review discussing the differences between 2nd and 3d edition. Here, you can check out the 2003 2nd edition.
The book is not focused on econometric applications, and this is exactly why I am suggesting it. It will give you the more general point of view on the matter, which sometimes brings fresh ideas in.
My experience is with the 2nd edition. The book does not target other scholars but students and practitioners. From the extras, I particularly like ch. 16 "Historical Tour of Categorical Data Analysis", since it always benefits my understanding of the material to have a sense of how a scientific field evolved.
answered Feb 27, 2015 at 0:55
-
-
$\begingroup$ Indeed. And as you say, the fact that it is not focused on "econometrics" is worth some extra points $\endgroup$– user4239May 5, 2015 at 19:10
$\begingroup$
$\endgroup$
- Greene, W.H. (2008), Econometric Analysis, 6th edition, Prentice Hall.
Very extensive. Worth having as a reference at least because it covers an enourmous set of models and estimation techniques. It is a little bit light on the intuition though. Also, a kindle edition is available which is a lot cheaper than the very expensive hard cover
- Winkelmann, R. and Boes, S. (2009), Analysis of Microdata, 2nd edition, Springer
I haven't looked at this one so can't speak from experience. It seems popular.
- Wooldridge, J.M. (2010), Econometric Analysis of Cross Section and Panel Data, 2nd edition, MIT Press.
In my experience, this one is a bit easier to read and does a better job of attempting to explain itself. It isn't as extensive as Greene. I would go to this book first because I find it more accessible. If I don't get the answer I need, it will probably be in Greene.