1) A change in the rate of one response that occurs when either the rate of a second response or the reinforcement rate produced by that response is changed in the Opposite direction, where the reinforcement rate maintaining the first response remains constant. For example, the rate of reinforced responding in one multiple-schedule component typically increases if reinforcement is reduced or discontinued in the other. The effect is most appropriately measured relative to a baseline performance in which responses in both components are maintained by the same reinforcement rate but has also been measured relative to rates in prior nonbaseline conditions. The term is usually restricted to responses in the presence of successive stimuli, as in multiple schedules, though similar phenomena occur within concurrent schedules.

2) A feature sometimes incorporated into concurrent schedules to prevent sequences in which one response is closely followed by a reinforcer produced by a concurrent response. As usually arranged, the COD provides that no response can be reinforced within some time after a changeover has been completed (but CODs are occasionally timed from other events, such as the last response before the initiation of a changeover).