Absence of anticipatory contrast in rats trained on multiple schedules.