Monday, March 31, 2014

Review: Biomeasurement: A Student's Guide to Biological Statistics

Biomeasurement One thing I'm not short of is statistics textbooks, so it takes quite a lot to get me to review a new one. But this one I like.

Biomeasurement: A Student's Guide to Biological Statistics. Dawn Hawkins, Oxford University Press, 2014 ISBN: 9780199650446

In her preface, Dawn Hawkins makes clear that she takes a "by biologists for biologists" approach, and explains "how to carry out techniques and why they are needed, it does not explain the underlying math". This will have the purists sucking their teeth, but many years of experience have shown me that teaching biologists maths doesn't work. What they need is applied maths so that they can carry out sound analysis, and that's what we're talking about here.

  1. Why am I reading this book?
  2. Getting to grips with the basics
  3. Describing a single sample
  4. Inferring and estimating
  5. Overview of hypothesis testing
  6. Tests on frequencies
  7. Tests of difference: two unrelated samples
  8. Tests of difference: two related samples
  9. Tests of difference: more than two samples
  10. Tests of relationship: regression
  11. Tests of relationship: correlation
  12. Generalized Linear Model I: General Linear Model
  13. Generalized Linear Model II: Logistic Model
  14. Choosing the right test and graph - as always, the most important chapter in any statistics text, so why is it at the end?

So that's the good news, but there's bad news too. The writing and explanation is not as clear as Andy Field's, but on the other had it's not as over the top and intrusive as Andy Field's writing has become in recent years. The big flaw is, in spite of the advertised use of R, the book is actually all about SPawn of Satan Statistics (SPSS), with the R content allegedly on a website ( - it wouldn't let me log in, and students certainly won't try as hard as I did), which is reason I won't be recommending it to students.

Other than that, quite a good effort.

No comments:

Post a Comment