Essential Biostatistics: A nonmathematical approach.  Harvey Motulsky

Essential Biostatistics:
  • Explains the ideas of statistics without describing the mathematical underpinnings 
  • Appeals to  the many students and scientists who prefer verbal explanations over mathematical proofs 
  • Focuses on how to avoid falling into common conceptual traps 
  • Points out ambiguities in potentially confusing terms and phrases 
  • Covers a wide breadth of topics in a quick and concise manner
  • Can be used as a stand alone text or as a supplement to longer texts.
  • Explains essential concepts missing from many books, including multiple comparisons, the False Discovery Rate (FDR), outliers and lognormal distributions.

Some ways in which this book is unique

  • It doesn't explain how to calculate any statistical tests. In fact, it only includes two equations.

  • Chapter 1 is a fun chapter that explains how common sense can lead you astray and why we therefore need to understand statistical principles.

  • Chapter 2 is a unique approach to appreciating the complexities of probability.
  • I introduce statistical thinking with Chapter 4, which explains the confidence interval of a proportion. This lets me explain the logic of generalizing from sample to population using a confidence interval before having to deal with concepts about how to quantify the scatter. 

  • I explain comparing groups with confidence intervals (Chapter 12) before explaining P values (Chapter 13) and statistical significance (Chapters 14 and 15). This way I could delay as long as possible dealing with the confusing concept of a P value and the overused word “significant”.

  • Chapter 16 explains how common Type I errors are, and the difference between a significance level and the false discovery rate.

  • Chapter 17 explains how multiple comparisons are an issue in almost all analyses. The issue of multiple comparisons goes way beyond followup tests after ANOVA.

  • Chapter 19 explains all common statistical tests as a series of tables.

  • I include topics often omitted from introductory texts, but that I consider to be essential, including: the false discovery rate, p-hacking, lognormal distributions, geometric mean, normality tests, outliers and nonlinear regression.

  • Nearly every chapter has a Lingo section that explains how statistical terminology can be misunderstood.

  • Nearly every chapter includes a Common Mistakes section, and Chapter 25 explains more general mistakes to avoid.