Maintenance of behavior controlling the duration of discriminative stimuli.

The team asked a simple question. What keeps a response alive longer: making good things appear, or making bad things go away?

They set up a task where one response turned on a positive stimulus and started food delivery. Another response simply shut off an annoying stimulus. They then watched which response stayed strong over time.

The response that produced the good stimulus and started reinforcement kept firing at high rates. The escape-only response stayed alive, but barely.

In plain words, you need the carrot, not just the stick, to keep behavior robust.

Dardano (1971) showed you can first teach duration control with plain differential reinforcement. Harrison et al. (1975) now add the next layer: once the control is there, tying the response to positive-reinforcer onset is what keeps it going.

Bennett et al. (1998) later found interval schedules make behavior last longer than ratio schedules. Both papers agree: richer, more predictable good stuff equals longer-lasting behavior.

Reid et al. (2003) looked at kids with autism and saw that free, non-contingent preferred items made stereotypy harder to stop. That lines up with M et al.—extra positive reinforcement, even if accidental, pumps persistence.

When you write a program, make the target response turn on the good stuff—tokens, praise, iPad—rather than only cancel something bad. If your escape-to-break plan is weak, add a brief preferred activity right after the break ends so the learner’s response still produces positives. You’ll see the skill stick around longer and need less re-teaching later.