The effects of response-dependent and response-independent methods on delayed alternation

Two frequently used methods of producing delayed alternation were compared at two ITI levels, 20 see and 6 min. One method alternated reward according to a fixed schedule (response independent), while the other demanded placement of reward opposite to the S’s previous response (response dependent). The response-dependent method produced a higher number of correct responses and alternations at both ITIs. Increasing the ITI duration from 20 sec to 6 min produced reliable decreases in the number of correct choices and alternations produced by Ss trained according to either procedure.