ABA Fundamentals

An extinction-induced increase in an aggressive response with humans.

Kelly et al. (1970) · Journal of the experimental analysis of behavior 1970
★ The Verdict

Stopping reinforcement can spark sudden aggression in teens, so pair extinction with an alternative reward.

✓ Read this if BCBAs running extinction with teens or any client who can hit.
✗ Skip if Clinicians only teaching new skills without withholding reinforcement.

01Research in Context

01

What this study did

Nine teens earned coins for pulling a lever. When the coins stopped, researchers watched for punching a nearby doll.

The team turned the money on and off four times. They wanted to see if stopping reward made the kids hit more.

02

What they found

Seven of the nine teens punched the doll harder when the money ended. The hitting spiked right after the payoff vanished.

Each time the coins came back, punching dropped. When coins stopped again, punching rose. Extinction caused the aggression.

03

How this fits with other research

The same lab first saw this in monkeys. Thomas et al. (1968) showed extinction made primates bite. Two years later the human test matched the animal data.

Hatton et al. (1999) later checked 41 kids with self-injury. About half showed the same aggression burst when staff withheld reinforcement. The clinic study extends the 1970 lab demo to real clients.

Shahan et al. (2025) gives a fix. Bigger alternative rewards during extinction cut burst size in rats. Their work supersedes the old warning by showing you can prevent the spike.

04

Why it matters

When you stop reinforcement, plan for possible aggression. Watch for hitting, kicking, or yelling the first few minutes. Offer another reward at the same time to blunt the burst.

Free CEUs

Want CEUs on This Topic?

The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.

Join Free →
→ Action — try this Monday

Before you start extinction, set a rich alternative reward and watch for hitting in the first three minutes.

02At a glance

Intervention
extinction
Design
reversal abab
Sample size
9
Population
neurotypical
Finding
negative

03Original abstract

Nine subjects, 14 to 18 yr old, pulled a knob on a schedule of monetary reinforcement. Concurrently, they escaped or avoided periodic presentations of a tone by pressing a button that required 1.5 lb (6.67 N) of force or by punching a padded cushion that required 20 lb (88.96 N) of force. The punching response was designated as an aggressive response because the force of this response together with its topography was comparable to responses of humans that deface objects and produce escape or counter aggression from other humans. It was found that button pressing was the preferred concurrent avoidance response and there were few punches. However, when the monetary reinforcer was discontinued (extinction) punching increased for seven of the nine subjects, but there was no consistent change in the rate of button pressing. When the punching response was replaced by another non-preferred but non-aggressive response, neither this response nor button pressing increased during extinction. Hence, the increase in punching during extinction cannot be attributed solely to the fact that it was a concurrent response or a non-preferred response.

Journal of the experimental analysis of behavior, 1970 · doi:10.1901/jeab.1970.14-153