Cressie et al paper

A few quick thoughts about the Cressie paper:

• the editors invited 7 different commentaries on the article; I've posted links to those commentaries. After you've read the paper it might be interesting to read some of the commentaries (I don't know if we should try to do this for next week or for the week after?)
• I've written to Jay Ver Hoef to see if we can get the data from the original paper. (Update: he's agreed to send them, as soon as he gets back from the Vancouver Olympics and gets caught up ...)
  
• I have a hard time evaluating how useful the discussion of these techniques is because I'm already pretty familiar with them (so I tend to fill in any missing pieces for myself); I will be interested to hear others' reactions. As the authors say, they're not really presenting new methodology here, just trying to emphasize some points for a new(ish) audience.

What do you think?

Wittemyer et al

Just wanted to throw out a few thoughts to begin the discussion on Wittemyer et al 2008 ...

• I thought the techniques they used were reasonable, but the use of the term "autocorrelation" was potentially confusing or misleading. There's a standard procedure in animal movement modeling of characterizing the autocorrelation in movement direction; the authors say that they didn't find much (perhaps this isn't surprising given that their time step is 3-hourly). Instead, they focused on autocorrelation in movement length. The reason I think this is confusing is that what they are really after is temporal patterns in general -- cyclic or otherwise -- and not just the relationship between the movement length at times t and t+1 (which is lag-1 autocorrelation). Spectral or Fourier analysis essentially transforms the (temporal, in this case) autocorrelation curve or autocorrelogram into a power spectrum which contains the same information but decomposes the variation in the response variable (move length) according to the (temporal) frequency, or wavelength -- thus one can easily detect e.g. daily cycles in movement. Wavelets do a (vaguely) similar decomposition, but have the advantage that one can more easily detect changes in temporal pattern over time.
• It felt like the techniques used were standard time-series techniques applied in a somewhat vague way to an interesting data set. I would have preferred an analysis that addressed behavior more directly: for example, can one characterize 'bouts' of behavior from the data and characterize the length of time, or the distribution of lengths of time, for which individuals display particular behaviors in different environments etc.?