Applied Linear Statistical Models

by Michael H. Kutner, John Neter, Christopher J. Nachtsheim

This new edition of Applied Linear Statistical Models retains the book's uniquely straightforward writing style and format while providing you with the latest information and knowledge. Updates include developments and methods in partial regression and residual plots, an entirely new introduction to the "Design of Experiments" section that frames and outlines the organization and concepts of design and ANOVA, and more.

Genres: Mathematics, Science, Reference

1424 pages, Hardcover

First published January 1, 1974

Reviews

[Sequoyah · Rating 4 out of 5]

The best mathematics (well, statistics) text book I have used. Both the subject and this book (and some R knowledge) helped me understand statistics like my statistical theory course was never able to do. I suppose that is to be expected once you apply mathematics to real examples (for example, I really didn’t truly understand what an “estimator” even was until I had to find the MLE’s for the standard error and the least squares estimators from this text and APPLY THEM, all I knew was how to prove it existed prior). It’s a shame that this will be the last math book I ever read, as I have truly and finally found my interest in mathematics (shame that grad school stats requires so much analysis that I would fail out, in addition to whatever Bayesian statistics is (and the always mentioned but never explained Monte Carlo Simulation)).

Unlike in most other text books I have used, this one actually had questions that you could solve solely by reading the text. I don’t understand why this is so difficult for most mathematics authors to do (particularly for questions requiring proofs), but in almost every textbook I have used, many questions cannot be answered through I reading of the chapters. However, here it is the opposite. This subject has some of the most interesting and widely applicable usage in math (at least in my experience). I genuinely had fun applying this knowledge to the data sets provided in this book. Never expected predicting things to be so easy, even given relatively small samples. I even used the knowledge I got from this text book for my final project in my Data Science with R class (actually got to use the Mallows Cp value (and adjusted R-squared!!) to find the best subset for a linear regression model given a diamond’s nine characteristics to predict price (just an FYI, only depth percentage was taken out of the best subset model due to multicollinearity)).

If you are even remotely interested in why math matters in EVERY subject, get your hands on this book or take a Udemy class on linear modeling. And stop complaining about how it’s been “another year without having to find the value of x.” Because I can, using linear modeling, predict how small your IQ is based on whether you complain about math being taught in school or not, thanks to this book.

[Noshgul · Rating 4 out of 5]

This is the book that actually made me consider switching from my mathematics major to statistics. It's very clear, contains a lot of useful references to more advanced texts, contains good examples, ... The only drawback is that it falls short on the mathematical derivation of the formulas, but then again it isn't supposed to be a mathematical statistics book (see Linear Regression Analysis by Seber and Lee for the mathematical rigor, these two books go perfectly together).

[William Schram · Rating 5 out of 5]

Applied Linear Statistical Models is a textbook by John Neter, William Wasserman, and Michael H. Kutner. The authors initially released the book in 1974, but my edition is from 1987. The authors realized that the computer is a phenomenal aid and included more content based on computing.

The first chapter is a crash course in Probability and Statistics. It covers tables, decision rules, and more. The examples used are clearly defined and explained. It may go without saying, but you should be familiar with Calculus for this book. The book discusses partial derivatives in some sections.

After the foundational chapters, the book covers linear regression modeling from single or multiple variables. As before, it clearly defines and covers the topics. I have no issues with the book besides its age, but it isn't as though mathematicians developed a new form of statistics. The book also covers nonlinear regression, which is heavily computer-oriented. Should you find this book somewhere, I think it's still serviceable. I managed to find it at my local library.

I enjoyed the book. Thanks for reading my review, and see you next time.

[Jacob Bornheimer · Rating 3 out of 5]

Readable but strangely laid out. Read for class.

[Crystal Swafford · Rating 4 out of 5]

Pretty good explanations. 5th edition is much improved over the 2nd edition

[Hai Le · Rating 5 out of 5]

I’m a huge fan of this book. It discusses statistical regression at a good amount of depth, from the assumptions behind the model, to model fitting, to evaluation, inference, prediction, diagnostic checks and remedial measures, i.e. everything you want to know about regression. The book has a very good structure and many realistic examples. I wish there were more math derivation in certain chapters. But still, it’s such a delight to read. I learned a lot out of it. Highly recommend.

[Caleb Cheam · Rating 3 out of 5]

Typical regression analysis book with more focus on economic intepretation of models rather than in depth focus on the statistical aspect and regularization concepts. Would not recommend this book for anyone who is serious about working with data and building useful models. There are other regression analysis books (eg generalized linear models by simon wood) that are comparatively better. Basic knowledge in statistics is required.

[Joe Cole · Rating 5 out of 5]

After the introductory chapter, the authors gave just the right amount of theory to explain the topic at hand and give extensive footnotes for further information. Lots of graphs and example software output are included, all very helpful. I found the text to be well-organized, with coverage given to explanation and examples of each topic.

[Joe Cole · Rating 5 out of 5]

I've found this an excellent textbook in both undergrad and graduate classes. It covers a wide variety of issues experienced in regression with (I have only used the first half of the text) how to diagnose them and how to fix them. Great book to have.
Every good textbook for statistics.

[Bob · Rating 4 out of 5]

Used by Prof. John Lamperti, Mathematics Dept., Dartmouth College as textbook for Topics in Statistics course, Spring 1979.

[Steve · Rating 3 out of 5]

I like this book for its rigor and plentitude of examples across disciplines.