Statistical Methods for the Social Sciences

I am a chemist. Chemists,unfortunately know very little about statistics. I was placed on a project which required me to use multiple linear regression, ANOVA, and logistic regression. I purchased this book, hoping that it would help me to learn these methods. It did. This book does an excellent job of explaining a wide variety of statistical methods in a clear manner. It is easy to read and you will be amazed at how quickly you are able to read this entire book and develop a strong statistical knowledge. I suggest you do the examples yourself as you move through the text using Excel. I think that this adds a great deal to the learning experience, and gives you confidence.

I bought this book in graduate school for my statistics class, and it has continued to help me ever since in my professional career. Unlike many of the statistics books that use complex formulas to explain statistical methods, this book breaks each formula down in an easy to follow format. After explaining the concept, the authors use an example to illustrate the point, which makes things much easier to understand. Also, the Appendix which has SPSS and SAS coding for each of the statistical methods described in the book is a lifesaver!

I was subjected to an earlier edition when I took statistics as an undergraduate, and I've used the 2nd and 3rd editions as a lecturer and professor, and I believe there is no preferable alternative.Agresti and Finlay are, above all, clear and accurate. Over the last decade, I've looked at several dozen alternatives, hoping to find one that's strong in the areas where this text is weak. I've been enticed by different layouts, writing styles, even overall motifs, but am always reminded of why I (and others) have relied on this text for so long.Some alternatives are just sloppy - poor editing, excessive typographic errors, incorrect answers in the answer keys. Some others border on incompetent, confusing basic issues and not clarifying the disputes on border issues. And some, while achieving rapport through comics, comedy, or simply light humor, lose some of the subtle finesse that statistics entails.Now, this one ain't perfect. The subtleties and disputes are side-stepped rather than highlighted. The text and layout are a bit wordy and eye-hard. And the examples are more practical than pedagogical. The data examples could be a bit sexier.But the meat is all there, and correct, and clear. And that's what you want in a statistics textbook. You don't need something that pretends stats is inherently fun or exciting. The lecturer should convey the power of p, the coolness of coefficients, and the holy grail of "r-squared". The text book should cover the material accurately and in detail, and this one does.

Agresti and Finlay did an excellent job with this textbook. It includes most of the important topics generally covered in a two-semester introductory statistics sequence. These authors skillfully use graphs, practical examples, and a diversity of exercises relevant to the social sciences students' experience to illustrate the concepts studied. Summaries at the end of each chapter provide a clear overview of the important points to remember. Readers need no more than a firm high-school math background to understand its contents. This book, in my opinion, is an excellent choice for both undergrad and beginning grad students interested in acquiring a solid foundation on statistical methods.

Your typical undergraduate student who is not a fan of mathematics education will find this book intimidating. But that's not really saying much.A student who's not math-phobic will enjoy it. It's not one of those statistics texts that tries to give just the concepts and not the underlying math. This one goes for the math -- which is the foundation of the concepts. (Conceptual explanations in plain English are here too! The book is not pure math. Anyway, statistics is equal parts numbers and reasoning.)The illustrations and diagrams are generally excellent. Each chapter ends with a large selection of questions and exercises (answers to some of these are provided at the back of the book) and a bibliography for further reading. Yes, really useful further reading -- not just academic texts, but popular science magazine articles, biographies of mathematicians, etc. Notation and terms are boxed off within the text, to be clearly noticeable upon review of the chapter. Helpful for studying.Sample computer output is given frequently, which is a nice bonus. Sometimes the output of popular statistics software can seem cryptic to the uninitiated. This initiates people. An appendix covers SAS and SPSS usage for each topic in the textbook.Of course it's up to the reader (or instructor) to choose how much material to cover; you could easily just ignore the last few chapters if you don't need the advanced material. But it's here, which makes this a nice book. (You might want the advanced material SOMEday...)There are 17 chapters running from "Sampling and Measurement", "Descriptive Statistics" and "Probability Distributions" through regression and correlation, *multiple* regression and correlation, ANOVA, and on to ANCOVA, "Model Building with Multpile Regression", "Logistic Regression", and then a single chapter at the end which briefly talks about the existence of factor analysis, structural equations, and other "Advanced Topics".It's a well written and quite in-depth textbook. A good choice for learning about statistics; a good choice for keeping on your bookshelf.