Asher et al., 2009, Recent advances in the analysis of behavioural organization

Ref: Recent advances in the analysis of behavioural organization and interpretation as indicators of animal welfare. J. R. Soc. Interface, 6, 1103-1119.

They talk about recent advances in analyzing behavioral organization. They compare this to standard statistical techniques revolving around discrete data collection, e.g., via ethograms.

They talk about 4 methodologies, and discuss pros and cons. They are:
(1) fractal analysis, (2) temporal analyses (Markov chain methods and time-pattern approaches), (3) social network analysis, and (4) agent-based modeling and simulation.

They are pretty into all four approaches. In talking about ABM, they claim that "a model can only reveal whether the heuristics and parameters entered are sufficient to generate observed behavior." There is certainly value in this kind of test, seeing whether given rules can yield observed behavior. However, ABM can also be used in more creative ways. For example, White and Smith (2007) used it to test the validity of various sampling techniques to capture real patterns in behavior.

One nice thing here is that it points out the need for new analytical methods. "Clearly, current behavioural measures of welfare are not ideal and new methods of analysis are needed. There are reasons to believe that analysis of behaviour that incorporates measurement or understanding of behavioural organization might be of particular relevance when it is to be used as a measure of welfare." They then give some nice elaboration on this (p. 1104).