The Ten Commandments of Statistical Inference

Written by: Richard Lenski

Primary Source: Telliamed Revisited

As Handed Down to Lenski by Sir Ronald Fisher

1.  Remember the type II error, for therein is reflected the power if not the glory.

2.  Thou shalt not pseudoreplicate or otherwise worship false degrees of freedom.

3.  Respect the one-tailed test, for it can make thine inferences strong.

4.  Forget not the difference between fixed treatments and random effects.

5.  Thou shalt not commit unplanned comparisons without adjusting the rate of type I error for thy transgressions.

6.  Honor both thy parametric and thy nonparametric methods.

7.  Consider not the probability of a particular set of data, but rather the probability of all those sets as or more extreme than thine own.

8.  Thou shalt confuse neither manipulation and observation, nor causation and correlation.

9.  Thou shalt not presume statistical significance to be of scientific importance.

10.  Thou shalt not be fearful of paying homage to a Statistician or His Holy Book, especially before planning an experiment; neither shalt thou be fearful of ignoring the Word of a Statistician when it is damnable; for thou art alone responsible for thine acceptance or rejection of the hypothesis, be it ever so false or true.

The Golden Rule:  Review unto others as you would have them review unto you.

***

Notes:  I wrote this for a graduate course on quantitative methods in ecology and evolutionary biology that I taught in Spring, 1989, at UC-Irvine.  The course focused on experimental design and frequentist methods for drawing inferences.

The following two tabs change content below.
Richard Lenski
Dr. Lenski is an evolutionary biologist. He writes about evolutionary science – past, present, and future – and some of the goings-on in the life of a scientist.
Richard Lenski

Latest posts by Richard Lenski (see all)