Regression and Other Stories (Analytical Methods for Social Research)

I have a BS in math and a MS in statistics. I am fortunate that I was able to work in full-time positions while doing graduate school part-time, because there are some aspects to the practical side of statistics that just aren't sufficiently covered when you're spending your time learning statistical theory from texts like Casella and Berger's Statistical Inference. The first two chapters of this text contain the clearest motivations for some of the practical aspects of statistics that are too often omitted in statistics texts. These chapters should be required reading for any person who is doing any sort of research or data analysis, period. Beyond these first two chapters, we start learning more about linear regression and transformations, statistical inference, simulation, and then take a deep dive into regression and eventually GLMs and causal inference. R code is brought in throughout the text without any prior programming background assumed. I really appreciate the emphasis on the practical aspects of these topics brought throughout the text, especially the section in chapter 4 titled "Problems with the concept of statistical significance." As a faculty member, my opinion is that this would be an excellent text for a second course in statistics for students who have a strong pre-calculus background, or a graduate-level social sciences course. However, there is a ton of detail packed into this text - I suspect if I ever use this text for a class, I'll have to spend quite a bit of time figuring out what aspects of each topic I want to cover, which is completely fine. This text should be on every social-science and health-science researcher's bookshelf. Not only is it a well-written self-study text, it's an excellent reference: the index is organized extremely well. My only criticism is that I wish this text had been written years ago! I expect that this text will become a classic in statistics eventually and highly recommend this text.

One of the previous reviews mentioned this book would make a great text for a second course/read on stat. methods -- I just wanted to echo that sentiment. It is 100% correct. Just for reference, I have a bachelors in mathematics and a masters in stats, and I work as an analyst in biomedical devices. When I was doing my stat. methods and theory sequence, the texts were Kutner et al., Casella & Berger, Hogg -- the typical treatment. If you've been exposed to those texts you'll definitely be prepared/over-prepared for this text. This book is a bit more conversational, and really teases out the rationale behind building statistical models. It's got a decidedly Bayesian feel but does a fair job of addressing the traditional approaches to modeling. It's also a great reference manual for the rstanarm package (which is GREAT for out-of-the-box Bayesian modeling). If you're looking to further your understanding and intuition of statistical modeling and best practice -- this is the book for you. (I also highly suggest visiting Andrew Gelman's statblog, as it also has some additional bits of wisdom posted pretty frequently) Postscript: my sole criticism is that authors use unconventional terminology to refer to type I/II errors (they use type M and S). It's not a huge deal, but when it pops up you've got to reconcile the difference -- interrupts the flow a bit. Don't let that affect your decision to pick up this book, though!

Surprisingly, the chapter about power analysis and minimum sample size was the clearest exposition of power analysis that I have ever read. The last few chapters on causal inference was a little over my head.

If you are a researcher or a scientist, get this book. Gelman and Co. explain how to perform Bayesian analysis in real life scenarios. This book describes every model, linear and non-linear models, in simple terms. And they complete all their statistical analyses in Rstanarm package, an adaptive RStan package in R software. The book is written like a novel. When I started reading the first few chapters, I couldn't keep down the book. However, this book is practically oriented, so for the mathematical background, you should read their earlier book, Bayesian Data Analysis. I highly recommend to buy this book along with Statistical Rethinking for a deeper understanding in using Bayesian analysis.

Great

Incredibly well written and provides a great breadth of social science examples. if I ever find myself teaching introductory econometrics again this will be the course text. A hemisemidemi-quibble about the typesetting in my paperback version: Annotations appear in the the left hand margins on both left and right pages. For the right pages this means they are partially obscured by the binding.

Fantastic introductory text. It is my first recommendation for anyone starting in statistics or data science.

Best applied regression analysis book out there! A wonder!!!!!

Muy bien libro, y llegó en buen estado.

Bargain compared to bookstores in town. Good for master's level stats applications

The Kindle edition comes as statically sized pages - as you would expect regular pdf files to look like only now you're stick with this file in your Kindle app. If you're going for the electronic version of this book then it's probably a good idea to purchase a proper pdf directly from the publisher so you get to choose where and how you read this.

Ottimo, spiega nel dettaglio diversi tipi di regressione con approccio bayesiano

Prof Gelman and friends explain well known statistical methodolog in new ways