Customer Reviews

Statistics and Experimental Design for Toxicologists, Third Edition

5 star:		(0)
4 star:		(1)
3 star:		(0)
2 star:		(1)
1 star:		(1)

 

 

I thought I had never met a book on statistics I didn't like (i.e., find something useful therein), until I met this book. I bought it in spite of other reviews that mentioned glaring errors that had not been fixed. I felt I knew enough about statistics to fix the errors, and I was interested primarily in the application of statistics to toxicology, for which topic few good texts exist. However, I could never have imagined the monumental scale of the corrections needed.

In this book (supposedly in its 3rd edition), in my opinion there is essentially NOTHING in the text without errors and omissions, except possibly the SAS programs given, which are possibly useful to the neophyte in SAS. Formulae are unintelligible, missing parentheses and using improper symbols and typesetting. Symbols are referenced that are never defined, etc. Apparently the author just created a montage of clips from other works without regard to antecedent explanation and had zero concern for correctness (he should be ashamed).

Finally, in my opinion, the author does not appear from his writings to be the expert in the field that one would expect from the book title. He does appear to be a toxicologist, but his knowledge of statistics appears rudimentary and confused.

I was truly disappointed by this book, because my needs remain unsatisfied. Its true value lies solely in the table of contents (so you can look up elsewhere the techniques referenced) and the SAS program snippets. Everything else is untrustworthy, incomplete and generally contaminated by errors. If the author cared the least bit about accuracy and content, he could have made his book enormously more useful in the 1st, 2nd or 3rd edition. The eternal optimist may hope for this in the 4th edition.

The third edition has a larger scope and is a useful upgrade from the previous edition. Beginning with descriptive statistics and using examples drawn from toxicology studies, the book covers issues and methods for inferential statistics in toxicology at an accessable level. Meta-analysis, assessment of carcinogenicity, risk assessment, and quantitative structure-activity relationships are all addressed.

A strength of the book is that the interpretation of the statistics in biological context is emphasized. The biological interpretation is an example of the author leading us to follow the advice of R.W. Hamming who suggested that the purpose of computation is insight, not numbers.

The third edition achieves its goal of being useful as a textbook for students and a source for practicing toxicologists.

While this book is a good attempt at placing statistical topics necessary to toxicology in one spot, the mistakes are inexcusable. Many formula are incorrect as well as text referring to the wrong tables or data. Therefore, one begins to lose trust in the information presented. Be careful if you use this book!