Final answer:
Token systems start with continuous reinforcement to establish the behavior and then transition to partial reinforcement to maintain the behavior effectively over time.
Step-by-step explanation:
Usually, token systems begin with a continuous schedule of reinforcement, before fading to a partial schedule of reinforcement. Initially, to establish a new behavior, a continuous reinforcement schedule is used where every instance of the desired behavior is reinforced. This is the quickest way to teach someone a behavior and is especially effective in training new behaviors.
After the behavior is established, trainers may shift to a partial reinforcement schedule. In a fixed interval reinforcement schedule, the behavior is rewarded after a set amount of time, while in a fixed ratio reinforcement schedule, a set number of responses must occur before the behavior is rewarded. By fading to a partial reinforcement schedule, such as a variable ratio or variable interval schedule, reinforcement becomes less predictable and is based on either varying numbers of responses or varying amounts of time, which maintains the behavior more effectively and makes it resistant to extinction.