Differential Reinforcement of Low Rates of Behavior

Description of the Strategy

Differential reinforcement of low rates of behavior (DRL) is one of a family of reinforcement schedules that offer an alternative to using punishment to decrease a target behavior. When a DRL schedule is in effect, a response is reinforced after a specific time period elapses since a response last occurred. An increase in interresponse time (IRT > t) reduces response rate. An increase in interresponse time reduces but may not eliminate the rate of the incompatible response. Thus, the DRL contingency is useful when the overall frequency of behavior needs to be reduced, not eliminated. Children are not required to be “perfect”; they can improve and thereby earn reinforcement.

Research Basis and Target Population

Laboratory studies with animals have demonstrated that the time between responses (the interresponse time or IRT) is a property of behavior that can be reinforced. The DRL schedule most frequently used in applied settings is full session DRL. Under the full session DRL contingency, reinforcement is delivered at the end of a session if less than a prescribed number of responses occur (the DRL limit). The session might be at home, work, school, or wherever the problem behavior occurs. For example, two students with autism were exposed to a DRL schedule to reduce gazing at the ceiling and inappropriate vocalizing. For one student, the opportunity to look through a gyroscope (the reinforcer) was contingent on ceiling gazing below the criterion set for the session. The DRL procedure for one student was programmed to progressively lower the rate of gazing (longer intervals between instances of gazing), until the behavior no longer interfered with classroom instruction. For the other student, the opportunity to listen to music was contingent on a gradually decreased rate of inappropriate vocalizing. Full session DRL has also been used to reduce the rate of speech (in a student who spoke too rapidly), to decrease the rate of off-task talking, to reduce self-stimulatory behaviors, and to reduce noise levels among groups of school children.

A second DRL variation that is frequently used in laboratory research is spaced responding DRL. In spaced responding DRL, a behavior is reinforced if it has been separated from a previous instance of that behavior by a minimum amount of time (the IRT). It is most effectively used with behaviors that are primarily a problem because of their high rate. A DRL procedure was used to decrease rapid eating in four individuals with severe disabilities. Interresponse times between food bites were reinforced and systematically increased. The eating rate was reduced from 10 to 12 bites per 30 seconds to 3 to 4 bites per 30 seconds. This schedule has been reported to reduce inappropriate classroom questions and stereotypical behaviors of normal and intellectually challenged children. The primary problem is its limited usefulness in settings that don't permit constant monitoring of the behavior.

A final variation of interval DRL specifies that reinforcement will only occur if fewer than a specified number of responses are emitted during the interval. If more than the criterion number of responses occur prior to the end of the interval, the DRL interval resets and reinforcement is delayed. For example, an interval DRL procedure was used to reduce the number of times a child interrupted other children. A DRL interval of 2 minutes was initially implemented.[Page 804]Reinforcement at the end of the interval was contingent on the emission of fewer than two interruptions. When a second interruption occurred prior to the end of the interval, the 2-minute interval was reset and the reinforcer was delayed. Similarly, a teacher gradually reduced a student's teeth grinding by initially reinforcing the student when no more than four grindings occurred during a specific time period. When the student met this criterion, reinforcement was next contingent upon grinding teeth no more than three times. This criterion was systematically lowered until the behavior was at an acceptable rate. Other classroom behaviors such as excessive question asking, talking out, and leaving one's seat have also been effectively reduced with interval DRL procedures.

...