Effects of extinction alone and extinction plus functional communication training on covariation of problem behaviors.
Never use extinction by itself—pair it with FCT or another reinforcement tool to block severe behavior from surging.
01Research in Context
What this study did
The team compared two ways to stop problem behavior. One group got plain extinction. The other group got extinction plus functional communication training.
They watched to see if mild behaviors dropped and if severe behaviors popped up.
What they found
Extinction plus FCT stopped every form of problem behavior.
Extinction alone only stopped the mild forms. The tough, dangerous forms actually got worse.
How this fits with other research
McLaughlin et al. (1972) saw the same warning sign first. When they ignored talking out, kids also played less and disrupted more.
Hatton et al. (1999) counted the risk in 41 cases. Almost half showed bursts or aggression when extinction ran solo.
Migan-Gandonou Horr et al. (2021) flipped the script. They paired extinction with noncontingent reinforcement and got a 98 % drop in perseverative speech for two years.
Cengher et al. (2020) even used the side effect on purpose. Extinction made new response forms pop out, and they shaped those into complex mands.
Why it matters
If you run extinction alone, be ready for bigger, uglier behavior to take its place. Add FCT, NCR, or another reinforcement piece up front. You will protect the client, spare yourself a crisis, and still kill the problem behavior.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Teach one simple mand first, then start extinction—no ignoring until the client can request.
02At a glance
03Original abstract
In this investigation, extinction (EXT) was applied alone or in combination with functional communication training (FCT) to less severe topographies of problem behavior while more severe topographies continued to be reinforced. EXT alone decreased less severe and increased more severe topographies of problem behavior (i.e., response covariation), whereas EXT with FCT reduced all topographies of problem behavior to near-zero levels.
Journal of applied behavior analysis, 1996 · doi:10.1901/jaba.1996.29-565