Page Status: Reviewed

Definition: Removing access to reinforcement for a behavior/response that previously accessed reinforcement.

Example when a behavior, for example crying,  has a existing/previous history of accessing verbal attention from others and now all persons stop providing all forms of reinforcing attention.

Typical Effects: Experimental research shows that the typical effects of implementing extinction,  is an immediate escalation of the frequency (and potentially magnitude, duration or other dimension of the behavior),  if the behavior continues to be ineffective in accessing the previous forms of reinforcement,  often a shift in the topography of behavior is seen (the learner alters the form of behavior in an attempt to access the previously accessed forms of reinforcement),  if these do not access reinforcement, eventually the learner will cease to use them.  This phenomenon is often referred to as an Extinction Burst 

Warning:  the learner is seeking reinforcement,  if it is unavailable for behavior that has worked well in the past, they will attempt to access it by engaging in other behavior.  Failure to make it available at the rates, forms and quality that the learner previously accessed will invariably result in topography shifts.   It is essential that clinicians establish adaptive/appropriate responses that are effective in accessing the specific forms and quality of reinforcement (functional communication responses). Careful consideration should be given to endure that the response effort of the adaptive response is less or equal to currently effective problem behavior. If the new response is too much work and is less efficient in accessing reinforcement, it may not be readily acquired.

Key considerations in teaching alternative responses:A) response effort when compared to the behavior being replaced (should be easier); B)  Speed of access to Reinforcement (should access it as quickly or faster); C) Quality of reinforcement accessed compared to the behavior to be decreased (should be as good or ideally better than the forms accessed by the problem behavior)

Common problems/concerns when using Extinction in applied settings and Clinical recommendations :

1. Extinction Burst: In many settings the level of and topographies of escalation are predictable and may be too dangerous or socially problematic for the individuals in the learners environment to implement extinction.

Clinical recommendations:  In conditions where a learner has a demonstrated history of significant and dangerous escalation when extinction is implemented, and these behavior will result in reinforcement delivery,  Extinction should not be used as an initial treatment procedure.  Alternatives: Implement a non-contingent schedule of access to the identified reinforcer at a rate and quality so that behavior to access reinforcement becomes less probable (for example if the average Inter response time between instances of access to reinforcement for the individual  is 10  minutes,  implement NCR at 5 minutes or less). This should then be followed by teaching one or more functional communication response that access the reinforcer maintaining problem behavior. Once these responses are acquired and demonstrated in a wide range of settings and wit ha variety of people,  so that they reliably use them independently to access  reinforcement,  then an extinction procedure can be implemented.  Under these conditions, a shift in topography occasioned by extinction is more likely to be to a topography of behavior that is in the learners repertoire and with an recent history of accessing the forms of reinforcement the learner is seeking (the forms most recently effectively established by you).

2. Spontaneous Recovery: This term refers to the phenomenon of the return of behavior previously placed on extinction and effectively decreased to zero or near zero rates.   While it is a phenomenon in need of further study, the return of extinguished behavior,  may be indicative of the fact that replacement behavior is no longer effectively accessing the forms of reinforcement (type,  quality, duration etc.) used t0 establish the replacement behavior.

Clinical Recommendations: Deprivation from reinforcement is the most likely (and parsimonious) explanation for this phenomena.  This hypothesis is also easily testable.   If instructional sessions can be implemented where the problem behavior is put back on extinction and the replacement behavior(s) are re-paired with access to reinforcement.  If the hypothesis is correct rates of the problem behavior should decrease to zero or near zero rates, while the alternative response should increase in frequency. If this does not occur,  then the consequence being delivered as a potential reinforcer, may not function as a reinforcer for these behaviors (since the data do not indicate a reinforcement effect or an extinction effect, this must be true).




Mark O’Reilly, Chaturi Edrisinha, Jeff Sigafoos, Giulio Lancioni, Wendy Machalicek, Massimo Antonucci
J Appl Behav Anal. 2007 Winter; 40(4): 731–735. doi: 10.1901/jaba.2007.731-735
VN:F [1.9.22_1171]
Rating: 0.0/5 (0 votes cast)