A bit on our philosophy
As ecologists we're used to the messy realities of nature. We think carefully about how to approach a problem, spend inordinate amounts of time researching methods and equipment, and then, when everything breaks, find creative work-arounds to make our experiments work somehow. Yet when it comes to analyzing our messy, complex data we seem to look for the quick way out—canned statistical packages that we don't understand, tortured transformations to shoe-horn our data into a form that we can plug into a statistic we already know, or worst of all, ignoring lots of data because we don't know what to do with it. Well not any more.
We think that as ecologists we should spend as much effort trying to understand our data and answer the questions we first posed as we do collecting the data themselves. We'll need to be creative, we'll need to spend a good deal of mental effort and maybe even some time in the statistical literature to find our way through, but in the end we'll have a much stronger sense of what we found and what we can infer.
Keep in mind
- If you can't write down your problem, you don't understand it
- There are almost always other, sometimes more elegant ways to answer a question or get (R) to do something, but who cares about elegance?
- Most of what you learn in this class, and your graduate education will take place outside of the class room
- Sharing code makes you a good person
- Commenting on and improving other people's code makes you an even better person
- Our goal is to make you "dangerous," not perfect, so don't worry too much about mistakes (but do fix them!)
Other places to find help (especially with R)
- Ben Bolker's Ecological Models and Date in R wiki page with an errata and other useful items
- StatsRus (aka Rtips), which is a useful collection of tips and examples for working with (R)
- Jack Weiss' "Statistical Analysis in Ecology and Evolution" course from 2006, which has some pretty terrific lecture notes. His "Statistics for Ecology and Evolution" course covers more advanced statistical topics.
- The (R) Journal
- (R) site search, which allows you to search the R-help mail archives, etc.
- A list of useful R functions
- One tip a day is occasionally useful. It's worth searching for that quick solution, although sometimes his solutions are more complex than they need to be.
- crantastic.org is a community site for R packages where you can search for, review and tag CRAN packages.
- It is not really helpful, but there is some interesting history in this New York Times article about R.
