Posté par David Sumpter ven, février 26, 2010 05:43:02 I am a bit late reading this, but I have just read Edwards and Pratts paper on rational decision-making by Temnothorax ants. It seems to be a bit of a fashion just now to test whether groups are rational or not. The standard technique is to first give a studied organism a choice between two options A and B and establish their preference for these two options. After this a choice between A, B and a decoy C is given, the decoy being of lower quality than A or B and thus irrelevant to the rational decision-maker. It seems the individual animals and humans do change their frequency of choosing A and B dependent on the attributes of C, exhibiting irrationality. Stephen Pratt's house-hunting ants were however rational, exhibiting the same preference pattern when choosing between two nest sites in the presence and absence of decoys.

There is a suggestion that groups can be more rational than individuals because they maintain some level of independence. When moving to a new home, most of the ants do not see all of the options and are thus less likely to be influenced by decoy alternatives. The very reason that individuals are making irrational decisions is that they have access to too much information.