Customer Reviews

Regression Models for Categorical Dependent Variables Using Stata, Second Edition

5 star:		(3)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

This is a very good book on how to do econometrics on limited dependent variables using Stata. The author also has another book with more theories if you need. A lot of different models (e.g., logit, probit, ordinal, multinomial, and count data) are illustrated and Stata codes are provided. It's an excellent book for applied econometrians. The only thing that is very important but not covered in this book is panel data models on limited dependent variables, e.g., panel logit or panel probit models.

This book is undoubtedly the best resource on discrete choice models for applied econometricians. The insights, applications, and tools provided by Long and Freese help not only make sense out of discrete choice models, but are extremely useful for clearly explaining the results to decision-makers untrained in econometrics. Importantly, the book explicitly details the assumptions embedded in each model and the appropriate tests/corrections.

The book won't teach someone all the nuances of discrete choice models on its own, and the authors point this out from the outset. However, when combined with "Regression Models for Categorical and Limited Dependent Variables" the two books provide an perfectly balanced dose of underlying intuition, math, and applied work insights (plus excellent tools, and the best explanation of maximum likelihood estimator that I have ever come across).

The prior reviewer is right that this version of the book doesn't cover discrete choice panel models. And, as the authors point out, it also doesn't cover censored/truncated data models. I see this less a shortfall and more as testament of the authors goal to go deeper rather than broader. As it is, the book covers a wide range of topics. That said, I truly look forward to day when they incorporate those topics into the next version(s) of this very pratical book. (Two thumbs...way up)

This book is a must for anybody interested in using Stata to examine categorical and limited dependent variable regression models. I've been keeping this book within an arm's length while writing programs and estimating models for my dissertation and it's been extremely helpful.