Variation is the stuff of evolution. Darwin made the breakthrough discovery
that natural selection cannot work without variation among individuals. Under-
standing evolution therefore requires ways to measure and analyze variation.
That need stimulated R. A. Fisher and other evolutionary biologists to lay the
foundations of modern statistics. Since then, statistics has grown immensely,
both in what it is able to do and the number of areas where it is used.
We use statistics in two basic ways. The first is to describe things. When we say
that the mean weight of male African elephants (Loxodonta africana) is about
7000 kg, we are conveying information about what a typical elephant is like. A
second use of statistics is to test hypotheses. Are female elephants smaller than
males? They are: females weigh on average 3600 kg. Statistics tells us that we
can be very confident that the difference between male and female weights is
real, and not simply because we happened to measure some unusually large
males and some unusually small females.
This appendix starts by introducing the key concept of a probability distribu-
tion. We then briefly review how statistics are used to describe populations,
estimate quantities, and test hypotheses. We end with a brief overview of two
major frameworks of statistical analysis, likelihood and Bayesian inference. While
statistics is a branch of mathematics, this appendix keeps the math to a minimum.
For more details, with many examples drawn from evolutionary biology, we rec-
ommend the excellent text by Whitlock and Schluter [2].^1(^1) Statistics can be intimidating because it is such a large and technical field. Luckily for evolutionary
biologists, the text by Whitlock and Schluter [2] is clear, friendly, and (best of all) filled with examples
from evolutionary biology.
Appendix:
A Statistics Primer
A
23_EVOL4E_APP.indd 1 3/22/17 1:52 PM