The effects of two response-elimination procedures on reinforced and induced aggression.
Extinction and DRO both cut reinforced aggression, but schedule-induced post-reinforcement aggression can linger—so keep watching after you deliver a reinforcer.
01Research in Context
What this study did
Corfield-Sumner et al. (1977) compared two ways to stop aggression.
They used extinction and DRO with lab animals.
The team watched for two kinds of aggression: aggression that got food and aggression that popped up after food delivery.
What they found
Both extinction and DRO cut the aggression that earned food.
Neither method fully stopped the extra aggression that showed up right after food was given.
The surprise: schedule-induced aggression can live on even when reinforced aggression stops.
How this fits with other research
Harrison et al. (1975) showed extinction can CREATE aggression. Corfield-Sumner et al. (1977) now shows extinction can also END aggression. The difference is timing: M saw aggression during extinction, K saw it after extinction was complete.
Neisworth et al. (1985) took the extinction idea from the lab to real kids. They found the same quick drop in behavior, but gains faded for one of two children. This extends K et al. by warning us that extinction effects may not last without booster sessions.
McClannahan et al. (1990) tested token-based DRO on self-injury. Like K et al., DRO worked fast, but E saw crying when tokens were withheld. This conceptual replication tells us DRO side effects can show up across different problem behaviors.
Silva et al. (2025) built on K et al.'s extinction-plus-DRO combo. They added context fading to stop renewal. This successor study shows the 1977 procedures still matter, but we now know how to make them stick longer.
Why it matters
When you stop reinforcing aggression, watch for two waves. First wave: the old reinforced bites or hits should drop. Second wave: new bursts right after a reinforcer may still appear. Track both. If the second wave shows up, add context-fading steps or booster sessions as Silva et al. (2025) and Neisworth et al. (1985) suggest.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Count aggressive responses for 5 minutes after each reinforcer delivery to spot lingering schedule-induced aggression.
02At a glance
03Original abstract
Pecks against a stuffed pigeon were reinforced according to a fixed-interval schedule for one group of pigeons and a variable-interval schedule for a second group. Red and green stimulus lights were alternately illuminated. Subsequently, food deliveries no longer occurred during one color (extinction). In the presence of the other color, food was presented only when no attack occurred for 30 sec. When attack produced food, all pigeons generally exhibited characteristic fixed-interval or variable-interval response patterns. Two birds in each group frequently exhibited postreinforcement schedule-induced aggression. Attack was reduced to low levels at approximately the same rate by extinction and differential reinforcement of other behavior. For birds that had previously exhibited schedule-induced aggression the initial reduction of attack during the second experimental phase was followed by induced attack immediately after food delivery in the differential-reinforcement-of-other-behavior component and upon onset of the extinction component, Either extinction or differential reinforcement of other behavior may eliminate reinforced aggression but may be relatively ineffective for reducing induced attack.
Journal of the experimental analysis of behavior, 1977 · doi:10.1901/jeab.1977.27-5