Color alternation learning in the pigeon under fixed-ratio schedules of reinforcement.

Pigeons were trained on a non-spatial delayed alternation task in which the correct stimulus was that color not responded to on the preceding trial. Subjects required to emit either 15 or 30 pecks to the correct stimulus within a trial learned the task, those required to emit only one or five pecks did not. Also, alternation was learned more easily after an incorrect than after a correct trial. Later experiments showed that a minimum fixed-ratio value was required for successful color alternation to occur, even though no fixed-ratio requirement was necessary when a position cue was available. The mechanism of the fixed-ratio effects derived from the pigeons' tendency to repeat their response in the presence of the color reinforced on the last trial. Whereas subjects trained on larger fixed-ratios corrected this error tendency within a trial, subjects trained on smaller fixed ratios did not.