Theory of Point Estimation

Erich Lehmann wrote this classic text on point estimation as the companion to his earlier masterpiece on hypothesis testing. The book lived up to the standard set by the other text. This edition, now published by Springer-Verlag was revised with the help of George Casella who is also an excellent writer. Another great reference book that statisticians should have on their bookshelf.

I am a bit surprised by the rating of this book. First of all, I will like to say, this book is intended for doctoral study or someone who are intersted in the theory of statistics. If you are looking for something basic to implement your projects, this is not the book for you. The proofs in the books may be "sketchy", however, as I've mentioned earlier, for those who are taking class, should have taken some of the basic theories course. And, one thing I will like to mention, measure theory is not required, in the introduction, it's briefly touched on. However, some knowledges in different mode of convergence will be helpful. The proofs may make use of some theorems that are proved using measure theory, but deep understanding of the proofs of those theorems are not required. I find this book is quite well-written, as compare to a lot of contemporary stat or math books.

Good book, but a little abstract for me. It requires some knowledge of measure theory. Some definitions and theories are stated in a more "mathematical" language. "Statistical Inference" by George Casella would be a good reference for it.

I found this book (and the companion volume Testing Statistical Hypotheses) remarkable for its breadth of coverage and yet consistency of treatment. It could be used as a reference book, but probably only after reading it fully. I found the treatment and terminology to be different from what I was used to on many topics and it needs to be read cover-to-cover as each topic builds on the last. The problem is, this takes so long! The text is long and densely packed with useful observations, examples and notes and the breadth of coverage means it is very difficult to skim through without missing a lot. Worth the effort though!