Biologist's Analytic Toolkit

Invitation:

To all:

This website is intended as a place to post descriptions of statistical procedures that others might find useful. It's open to everyone. The idea is to provide background information plus explanations using examples so that data can be analyzed and interpreted correctly with minimal fuss.

In my opinion, a useful post consists of at least some of the following elements:

  1. Simple descriptions of techniques in an introductory manner that helps beginners understand motivation for analysis.
  2. Detailed explanations of how to input data and produce suitable output, along with a description of assumptions and meaning of results.
  3. A worked example using an available dataset performed by a functional R script.
  4. More detailed lecture-style notes showing interim calculations and statistical logic for individuals interested in somewhat greater depth.

If you have suggestions about what also might be included, or would like to include a pet lab analysis or prototype of your own, please let me know! The whole point is open communication and sharing so we don't always have to start from scratch when something new comes along. For now, and this is just a start, Most of what's posted are notes I've made for one reason or another.

The website address is:

biotoolbox.binghamton.edu

On the main page, one will find two versions of "Getting started with R". One is an adobe postscript (*.pdf file) and the other a generic text file (*.txt). The first is more suitable for printing, but the second is easier to cut and paste into the R interpreter for direct results.

Other sections of the website include different areas of statistical analysis in very broad categories. A place to start might be the subsection labeled "Biostatistics". Here you will find my prototypes for the Biostatistics course in the Biology Department using Biostatistical Analysis"5th edition by Jarrold Zar.

Clicking inside the current course notes (first line) you will find, for instance, an entry labeled "Biostatistics 150 Paired t-test". Clicking on it, you will find a prototype that attempts to cover points 1-4 above. A general description may be found in a choice of formats. Probably the most useful will be the static description as in Adobe postscript (*.pdf). However, I also include the dynamic file in MathCad (*.mcd) that actually performs the calculations. In addition to this, you will typically find a worked example using a dataset from Zar (to allow comparison with the text's presentation), and an R script also showing how to do the test.

To run the example yourself, it is necessary to download the dataset to an appropriate local folder on your computer that your local version of R knows about (see about the working directory in "Getting Started with R"). The R script can then be pasted into the R interpreter. It should work. However, if you find errors, please let me know so I can fix them.

Hope you find this useful.

Wm Stein

stein@binghamton.edu