Abstract Many social processes are described only in ordinary language, though often there are advantages to stating the ideas more precisely using mathematics. Intuitive ad hoc models are less valuable than are models representing explicit theories because theoretical models are useful beyond particular situations for which they were created. We describe 13 ways in which individuals might use information about status structures and communication from others to guide their behavior, and construct simple mathematical models of the processes. Three models derive from explicit theoretical assumptions and 10 are based in intuitive ideas about how individuals process information. Next we assess empirical success of the 13 models by comparing their predictions to behavior in an experimental setting. All the models fare well by comparison with a null hypothesis of no effects of structure or interaction, but the theoretically-based models make the best predictions. We discuss some reasons for the outcomes and for the intuitive appeal of intuitive models. Both the poorer predictions and the intuitive appeal of intuitive models, oddly enough, may trace to their lack of conditionalizing their predictions.