Experimental Design and Statistics

Statistics is applied philosophy of science.

We have learned huge amounts in biology by doing experiments which generate clear and simple results. But many parts of biology are governed by smaller effects, which are sometimes hard to measure exactly. For example: Does an amino acid substitution change the efficiency of an enzyme? Do female frogs prefer mates with lower-pitched croaks? Does the chance of developing diabetes depend on your genotype at a specific locus? To answer these kinds of questions, we have to use statistics, not only to decide on our answer, but also to make some determination of how much we believe it. To actually use those statistics, you need two things: understanding of the theory behind the statistics and the ability to apply them. To that end, the primary goals of the course are:

- to introduce to the field of statistics and its uses for describing data and testing hypothesis, with a special emphasis on biological examples.
- to develop an understanding of the considerations of designing experiments to answer biological questions clearly and efficiently.
- to learn how to implement statistical tests and calculations using classical formulas, as well as through computational tools, specifically the R language.
- to develop the ability to communicate biological, statistical, and computational concepts, results, and techniques through prose and statistical graphics.

**Lecture** MWF 10:10—11AM, 349 Park Science Center
**Lab** Thurs 1:10-4PM, 100 Park Science Center

*The Analysis of Biological Data, Second Edition* by M Whitlock and D Schluter

Also recommended:

*R Cookbook* by P Teetor

*R Graphics Cookbook* by W Chang

**DataCamp**
Online tutorials in data science, with a number of different courses for different levels of learning. Some of theses will be assigned as part of the class. They will be free while you are in the class, but if you want to continue afterward they ask you to pay for some of the more advanced courses.

**Try R**

Another online introduction to R that runs in your browser. Nothing to install, and a goofy pirate theme.

**Swirl**

An R package for learning about R from within R. You can install and run this on your own computer, which is good for getting used to the RStudio interface. A number of modules are available, including both basic and more advanced tutorials.

**Online learning resources from RStudio**

**R Cheatsheets** Particularly handy are the Data Wrangling (`dplyr`

and `tidyr`

) and Data Visualization (`ggplot2`

) sheets. You may also find the R Markdown Cheatsheet and Reference Guide helpful to have. (note: PDF links)