Behavior Modification Program
Chapter 22:
Differential Reinforcement
Definition of Differential Reinforcement
- Reinforcing one response class
- When dealing with reducing problem behavior, this involves
- Reinforcing a behavior other than problem behavior
- Reinforcing a reduced rate of problem behavior
- Withholding reinforcement for another
Diagram of Differential Reinforcement
SD
“Play by your self while I wash the dishes”
Response
Child plays with toys
SR+
Praise delivered
Playing with toys occurs more often in the future when the individual has been deprived of attention for periods of time
EO
Deprived of one-on-one attention for a period of time
Response
Child tantrums
SO
Praise withheld
Tantrums occur less often in the future when the individual has been deprived of attention for periods of time
DRI
- DRI: Differential Reinforcement of Incompatible Behavior
- Reinforce a behavior that cannot occur with problem behavior
- Withhold reinforcement for instances of problem behavior
DRA
- DRA: Differential Reinforcement of Alternative Behavior
- Reinforce occurrences of desirable alternative to problem behavior but that is not necessarily incompatible
- Reinforcement is withheld for problem behavior
A note on terminology…
- Sometimes when reinforcer is a negative reinforcer:
- DNRI
- Differential negative reinforcement of incompatible behavior
- DNRA
- Differential negative reinforcement of alternative behavior
Guidelines for Implementing DRI/DRA
- Select incompatible/alternative behavior
- Already exists in repertoire
- Requires equal or less effort than problem behavior
- Emitted at a rate that provides sufficient opportunities for reinforcement
- Likely to be reinforced in natural environment
Guidelines for Implementing DRI/DRA
- Select potent reinforcers that can be controlled
- Identify via stimulus preference assessment
- Identify via functional behavior assessment
- Use same consequence as is maintaining problem behavior for appropriate/incompatible behavior
Guidelines for Implementing DRI/DRA
- Reinforce incompatible/alternative behavior immediately and consistently
- Withhold reinforcement for problem behavior
- Some “mistakes” may be tolerable
- Combine with other procedures
DRO
- Differential Reinforcement of Other Behavior
- Deliver reinforcer whenever the problem behavior has not occurred for a specific time
- “Reinforcement for not responding”
Forms of DRO
- Fixed-interval DRO (FI-DRO)
- Omission requirement is applied at the end of successive time intervals of equal duration
- To apply:
- Establish interval
- Deliver reinforcement at end of interval if problem behavior didn’t occur during the interval
- If problem behavior occurs, reset interval
Forms of DRO
- Variable-interval DRO (VI-DRO)
- Omission requirement is applied at the end of successive time intervals of variable and unpredictable durations
- To apply:
- Establish variable interval schedule
- Deliver reinforcement at end of interval if problem behavior didn’t occur during the interval
- If problem behavior occurs, reset interval
Forms of DRO
- Fixed-momentary DRO (FM-DRO) and Variable-momentary DRO (VM-DRO)
- Omission requirement is applied only at the end of successive time intervals of fixed or variable durations (contingency not in place during interval)
- To apply:
- Establish interval
- Deliver reinforcement at end of interval if problem behavior didn’t occur at the end of the interval
What type of DRO to use?
- Interval more widely used than momentary
- Interval more effective for more suppressing problem behavior
- Momentary may be most useful for maintaining reduced levels of problem behavior
Guidelines for Using DRO
- Recognize limitations
- Reinforcement provided if absence of target problem behavior.
- If another, nontargeted problem behavior occurs, it is reinforced.
- May need to shorten interval
- May need to include other problem behaviors in definition
Guidelines for Using DRO
- Recognize limitations
- With Momentary DROs, reinforcement is delivered if problem behavior is not occurring at end of interval, even if it occurred throughout the majority of the interval
- Change to interval DRO
- Shorten interval
Guidelines for Using DRO
- Set initial DRO intervals that assure frequent reinforcement
- Calculate mean baseline interresponse time (IRT)
- Set interval that is equal to or slightly less than mean IRT
- Do not inadvertently reinforce other undesirable behaviors
- Make rule: must have absence of target problem behavior and other inappropriate behaviors
Guidelines for Using DRO
- Gradually increase the DRO interval
- Three options:
- Increase by constant duration of time
- Increase intervals proportionately
- Increase based on learner’s performance
- Extend to other settings and times of day
- Combine with other procedures
DRL
- Differential Reinforcement of Low Rates of Responding
- Use to decrease the frequency of the occurrence of a behavior, but not eliminate it all together
DRL
- Full-session DRL
- Reinforcement is delivered at the end of a session if during the entire session, the target behavior occurred equal to or fewer times than a predetermined criterion
DRL
- Interval DRL
- Divide the total session into a series of equal intervals of time
- Provide reinforcement at the end of each interval in which the number of occurrences of target behavior is equal to or below predetermined criterion
DRL
- Spaced-responding DRL
- Deliver reinforcer following an occurrence of a behavior that is separated by at least a minimum amount of time from a previous behavior
- In other words, reinforcement is contingent on increasingly longer IRTs
Guidelines for Using DRL
- Recognize limitations
- DRL is slow and does not reduce a behavior quickly
Guidelines for Using DRL
- Choose most appropriate DRL procedure
- Spaced responding is the only DRL procedure that delivers reinforcement immediately following response and maintains lower rates
- Use full-session and interval DRO when it is okay to have either no or low rates of target behavior
- Spaced responding provides higher rates of reinforcement
Guidelines for Using DRL
- Use baseline data to guide selection of initial response or IRT limits
- Set at mean baseline or slightly lower
- Gradually thin the DRL schedule
- Full-session DRL: set new criterion based on learner’s current performance
- Interval DRL: gradually decrease number of responses per interval
- Spaced-responding: adjust IRT criterion based on performance
Guidelines for Using DRL
- Provide informational feedback to the learner
- Enhance effectiveness by helping learner monitor performance