LZ Garamszegi, S Calhim, N Dochtermann, G Hegyi, PL Hurd, C Jorgensen, N Kutsukake, MJ Lajeunesse, KA Pollard, H Schielzeth, MRE Symonds, S Nakagawa (2009) Changing philosophies and tools for statistical inference in behavioral ecology.Behavioral Ecology 20: 1363-1375.
What’s it about?
A review of new statistical approaches (Akaike Informatio Criterion, Bayesian analysis) in behavioural ecology. Kind of an advocacy piece, with some illustrations of the types of analysis that can be carried out using these approaches.
What’s the story behind it?
At the 2008 International Society for Behavioural Ecology conference at Cornell University in New York, I was invited to take part (very hungover) in a post-conference symposium on statistical methods in the behavioural ecology. I was surprised to be invited, but prepared a talk on my experience of using Akaike’s Information Criterion in my analyses of bird abundance and species richness (only later realising that the reason I had probably been invited was because of my work on the effects of phylogenetic inaccuracy on comparative analyses). Anyhow, after the symposium, the participants were invited by Laszlo Garamszegi and Shinichi Nakagawa to take part in writing a paper on the changing application of statistical methods to behavioural ecology. It was a very odd paper to write. I felt somewhat out of my depth, not least because it was written through a wiki, so that anybody could come in to add stuff, or change text they didn’t agree with. The consequence was that every time I contributed something, it felt like it was rapidly overwritten. Consequently in the main body of the text probably only a handful of sentences scattered throughout the paper are actually mine. However, at the review stage we were asked to include more examples and I was able to re-do an analysis from my Symonds and Johnson (2006b) paper and present that as a worked example. So that box in the paper is entirely my writing – and so I did at least eventually feel that an aspect of the paper had my muddy paws all over it.