Table 1.

Comparison of experimental results with predictions of the Tug-of-War (ToW) model and the Sequential Choice model (SCM).

	ToW	SCM	Results for individuals	Results for colonies
Pooling Poor and Good sites	Longer in binary, due to comparative evaluation	Shorter in binary, due to cross censorship	Longer in binary (ToW)	Shorter in binary (SCM)
Only Poor site		Much shorter in binary, due to strong cross censorship	No difference	Shorter in binary (SCM)
Only Good site		Weakly shorter in binary, due to weak cross censorship	Longer in binary (ToW)	No difference

The second and third columns show each model’s expected differences between the latency to choose an option from a binary set and the latency to take the same option when it is the only one available. The fourth and fifth columns show the latency differences actually observed for individuals and colonies, respectively. Note that latencies of decision making are compared only within each subject group (individuals or colonies). Significant latency differences (p < 0.05) are shown in boldface, with the supported model indicated in parentheses.