Reinforcement magnitude and responding during treatment with differential reinforcement.

Oliver et al. (2002) asked a simple question: does giving a child more time with a reinforcer make the new skill stick better?

They taught three children to ask for a break instead of engaging in problem behavior. Each child got the same teaching, but the break lasted 20, 60, or 300 seconds.

After the skill looked solid, the team stopped giving the break. They counted how many times the child kept asking anyway.

Longer breaks did not make the new request stronger. Twenty-second and sixty-second breaks produced the same staying power.

The five-minute break only slowed the child down right after the break ended. Resistance to extinction stayed flat across all sizes.

Lowe et al. (1974) saw the same pause pattern in rats decades earlier. Bigger reward, longer pause. The 2002 kids echoed that tiny pause effect, showing the lab result travels to humans.

Staddon (1970) also found longer reinforcers stretched the pause inside fixed-interval schedules. Oliver et al. (2002) now show the pause stays small and barely matters when you move to extinction.

Shahan et al. (2025) looked at the flip side: large reinforcers before extinction create bigger extinction bursts. Oliver et al. (2002) found no burst benefit, so together the papers draw a line: magnitude changes the burst, not the long-term staying power.

You can stop hoarding session minutes for giant reinforcers. A short, timely break works just as well at keeping a new communicative response alive. Spend the saved time on extra practice trials or teaching the next skill.