ABA Fundamentals

Behavioral contrast as a function of the duration of an immediately preceding period of extinction.

Wilton et al. (1971) · Journal of the experimental analysis of behavior 1971
★ The Verdict

A 40-minute extinction period before reinforcement gives the largest behavioral contrast; longer waits add no extra boost.

✓ Read this if BCBAs running multiple-schedule or DRA programs where reinforcement returns after breaks.
✗ Skip if Clinicians who use only continuous reinforcement with no extinction components.

01Research in Context

01

What this study did

N and coworkers worked with four pigeons in a small lab chamber. The birds pecked a colored key for food on a variable-interval schedule.

Before each food session, the team introduced a second key where no food was available (extinction). The duration of this extinction period varied randomly across sessions: 0 seconds, 10 seconds, 30 seconds, 2 minutes, 10 minutes, 40 minutes, or 120 minutes.

The question was simple: does a longer extinction period produce a bigger jump in pecking when food becomes available again?

02

What they found

Longer extinction periods made the pigeons peck faster when reinforcement restarted.

Behavioral contrast increased with increasing extinction durations and leveled off at 40 minutes. At 120 minutes the effect was no greater than at 40, showing a clear ceiling.

In plain words, a 40-minute break from reinforcement gave the biggest behavioral contrast; additional time did not add to the effect.

03

How this fits with other research

Craig et al. (literature) later showed the opposite trend: repeating short extinction sessions makes future resistance weaker, not stronger. The two papers seem to clash, but they test different things. N et al. looked at one long wait before the next reinforcement, while Craig et al. cycled many short waits. One long wait boosts contrast; many short ones wear the behavior down.

R et al. (literature) used the same basic setup and added a twist: they measured how long each peck lasted. Contrast raised rate but did not shorten peck duration, showing the effect is about how often, not how the movement looks.

P et al. (literature) appeared to contradict the need for long extinction at all. They got peak-shift after only three minutes of change. Their brief stimulus switch produced stimulus-control errors without sustained contrast, suggesting different mechanisms drive short-term shifts versus the robust rate boost seen here.

04

Why it matters

When you reintroduce reinforcement after a long break, expect a surge in responding. Plan for this burst if you run DRO, DRA, or return from extinction.

An extinction period of roughly 40 minutes may give you the strongest rebound, so adjust criteria and safety plans accordingly.

Use the spike as a teaching moment rather than a surprise.

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

After a long extinction stretch, watch for a sudden jump in target behavior and set response limits before the session starts.

02At a glance

Intervention
extinction
Design
single case other
Sample size
4
Population
not specified
Finding
positive

03Original abstract

For four pigeons key-peck responding was reinforced on a variable-interval reinforcement schedule in the presence of a vertical white line. When response rates had stabilized a horizontal white line was introduced, in the presence of which reinforcement was not available (extinction). The horizontal line was presented once per session, immediately before the vertical line was presented. The duration of the horizontal line varied randomly from session to session, being either 0 sec (i.e., no presentation), 10 sec, 30 sec, 2 min, 10 min, 40 min, or 120 min. When the horizontal line was present for more than 0 sec, behavioral contrast was obtained in the presence of the following vertical line. Contrast increased with increasing durations of the horizontal line, asymptoting when the horizontal line was present for 40 min.

Journal of the experimental analysis of behavior, 1971 · doi:10.1901/jeab.1971.16-425