Principles and Procedures of Statistics [Hardcover]

I found that Steel et al. covered a broad range of methodology needed for biological research, and that the scope of this text is comparable to any statistical manual available. It is very functional as a reference material for troubleshooting any design or analysis problems. As a professor of research methodology, I highly recommend consulting this text for improving the statistical analysis of biological research.

This book is descent but a poor substitute in light of better alternatives, namely Zar's Biostatistical Analysis or Sokal and Rohlf's Biometry. Zar's book is my favorite. I am not as familar with Sokal and Rohlf's but know enough that I prefer the organization and context of Zar. The limitations of Steel et al. is the needless use of matrix algebra, the lack of calculations in many cases (instead, the linear model is presented without decomposition into formulas), the brief discription for many of the analyses, lack of good examples, and difficult to follow writing. On the positive side, Steel et al. have a chapter on experimental design, which Zar and Sokal and Rohlf lack, although it is not an easy chapter to follow. My recommendation is to use Zar as your primary "go to" reference on biostatistics.

I found this book somewhat disappointing. I found the book to be a run-of-the-mill statistics book that just happens to draw examples from Biology as opposed to other areas. The book is perfectly usable, but in my opinion, falls somewhat short of its intended purpose.

As statistics books go, this book is above average, but it is still not the clearest book out there. I do like the inclusion of example data within the book, but often the mathematics is not presented in the clearest way. I think there are a number of better statistics books out there, books that take the reader to a more advanced level while being easier to read. My favorite book that in my opinion achieves these goals is "Mathematical Statistics and Data Analysis" by John Rice.

I also would have appreciated much more discussion of biology. The authors do not do a good job of communicating what makes the application of statistics to biology special. There are important distinctions between the application of statistics in biology and its application in other fields, in terms of the usefulness of various techniques, and the sorts of problems and complications that arise. It is clear that the "biometrical approach" influenced the authors' choice of topics, but I wish they had more explicitly discussed how the mathematics relates to concepts and phenomena that are specific to biology. This potential connection to biology was the main reason I picked up this book, and I was quite disappointed to find that the connection wasn't developed to the degree I had expected.