Differential Reinforcement Justin Daigle MA BCBA LBA Program

Differential Reinforcement Justin Daigle, MA, BCBA, LBA Program Director

Introduction �A lot of information will be presented. You do not have to memorize or fully understand all of the information. Some information is lagniappe for those that are interested. Understand the concepts and use this as an introduction.

A Note Before Beginning �All interventions based on Differential Reinforcement are known to be lengthy processes. Do not expect them to be a ‘quick fix’. �Client does not have to know the contingencies.

Response Class �A group of responses (behaviors) of varying topographies, that serve the same function. �Ex: Hitting and Kicking for escape

Differential Reinforcement �Using different “values” of reinforcer for different topographies of behaviors. �Ex: I get 5 skittles for high five and 2 skittles for “knuckles”. You must want me to give you high fives more!

Matching Law (Simplified) �Rates of responding are proportionate to the reinforcement value received from each choice alternative.

1 -Second of Food 5 -Seconds of Food 5: 1 xƒ = 5 xƒ = 1

“+5 Value” “+1 Value” 5: 1 xƒ = 5 xƒ = 1

DRI (Incompatible) �Reinforce a behavior that cannot be done while the problematic behavior is occurring.

Example �Eoin is a 25 -year-old man who makes “psychotic statements” such as “There is a purple moose in my head named Chuckles. ” A functional analysis reveals that the behavior is attention maintained.

Example (cont) �Anytime Eoin made a ‘proper statement’, his aides would attend for 10 -seconds. �Anytime Eoin made a ‘psychotic statement’, his aides would ignore. �Seems basic?

Example (cont) � 1) You cannot make a “psychotic statement” and a “non-psychotic statement” at the same time. They are incompatible. � 2) You are using differential reinforcement �Therefore, you are using a DRI

DRA (Alternative R) �Reinforce a behavior that is an alternative, but is not necessarily incompatible with the problem behavior

Example �Eoin is a 12 -year-old with limited vocal behaviors. The function of problem behavior (kicking) has been determined to be access (lack of manding).

Example �Eoin taught to use a PECS card to request what he wants. �Eoin can kick and grab the PECS card at the same time (not incompatible) �However, both behaviors serve the same function (a way to request)

DNRA or DNRI �Differential Negative Reinforcement of Alternative Behaviors �Differential Negative Reinforcement of Incompatible Behavior

DNRA – Example �Giving a PECS card that says “Break” will remove an adverse stimuli (the instructor). �Kicking will no longer remove the adverse stimuli

Considerations for DRI/A �Select Alt or I R prior to beginning + r �Powerful S delivered consistently + r �Immediate and consistent S + r �Withhold S for problem R �Teach appropriate behaviors �Can be combined with other procedures

Conceptual Question �We all know that we can reinforce a behavior. �Can an organism be reinforced for a lack of behavior?

Conceptual Question �Yes! �You can be reinforced for not doing a thing! �This is the foundation for:

DRO (Other Behaviors) �Delivers a reinforcer whenever the problem behavior has not occurred during a specified set of time. �Interval DRO �Fixed Interval (FI-DRO) �Variable Interval (VI-DRO)

DRO (Other Behaviors) �Can also be at a specified time �Momentary Time Sampling �Momentary DRO �Fixed-Momentary DRO (FM-DRO) �Variable-Momentary DRO (VM-DRO)

DRO (Other Behaviors) �By any other name: �DR of Non-occuring Behavior �DR of Zero Responding �Omission Training

Goals of a DRO �To sustain a behavior at a level of zero occurrence (+/-2)

Limitations of a DRO �Difficult to come into contact with + r S �Behavior may occur to frequently �Only focuses on 1 topography �While focusing on hitting, there is no intervention for biting

Considerations of a DRO �Recognize the limitations + r �Set initial interval to assure S + r �Do not S other undesirable behaviors �Gradually increase interval �Monitor generalization (time of day/setting) �Use with other procedures

DRL (Low - Occurring R) �Delivers a reinforcer whenever the problem behavior has occurred less frequently than a set criteria. �Full Session DRL (interval is entire session) �Interval DRL (interval is specified) �Spaced-Responding DRL (responses occur with a specified time between them)

Goals DRL �Decrease problem behavior but not to eliminate it.

Considerations of a DRL �Recognize the limitation (very slow) �Choose the best procedure (of the 3) �Use baseline data to set criteria �Gradually thin schedule

DRH (High – Occurring R) �Used to increase a target behavior �Delivers a reinforcer whenever the target behavior has occurred more frequently than a set criteria.

Considerations of a DRH �Exact same as the DRL

Review �DRA (Alternative) �DRI (Incompatible) �DRO (Non – Occurrence) �DRL (Low – Occurring) �DRH (High – Occurring)