How do rats cope with the two-way escape problem in a homogenous shuttle box?

The behavior of 25 rats trained in a homogeneous shuttle box to escape unsi gnalled grid-shock was analyzed. Three categories of escape were distinguis hed: (1) species-specific fly away from the charged grid, (2) long-latency crossing preceded and accompanied by other behaviors that compete with the escape response, and (3) short-latency escape which followed an anticipator y postural pose. The animals displayed species-specific fly away only durin g the initial trials of a session. Subsequently long-latency crossings deve lops, reflecting a resistance to enter the opposite compartment. A measure based on a comparison of escape latency distributions in the two halves of the Ist session discriminates between good and poor learners. Subgroups of good and poor learners differed in performance efficiency in all five train ing sessions. Good learners were able to overcome the resistance to enter t he opposite compartment and recall the learned short-latency escape.