Effects of two variations of differential reinforcement on prompt dependency.

Cividini-Motta et al. (2013) tested three ways to give reinforcers while teaching kids with autism. One group got the best toy only when they answered alone and a so-so toy after a prompt. A second group got the best toy for alone answers and nothing after a prompt. A third group got the same toy no matter what.

They used an alternating-treatments design. Each child saw all three styles in mixed order during the same lesson.

Three of the four children learned fastest with the first style: big reinforcer for independent responses, small reinforcer for prompted ones. The fourth child did best with the second style: big reinforcer for independent, nothing for prompted.

Both styles beat the equal-reinforcer control. Kids made fewer prompt-dependent errors and reached the learning goal sooner.

Campanaro et al. (2020) asked when to start differential reinforcement, not which size to give. They found starting right away beats waiting. You can blend the two papers: begin DR on trial one and use a bigger prize for independent answers.

Stuesser et al. (2020) and Briere et al. (2025) later used DR without extinction plus stimulus fading for medical tasks. Their data line up with Catia: you can skip extinction if you add fading or adjust reinforcer size.

Wilder et al. (2020) paired DR with three-step guided compliance. Again, DR plus another support beat either one alone, echoing Catia’s call to mix tactics.

Next time you see a child waiting for prompts, try giving the top reinforcer only after unprompted answers. Hand a lesser item or no item after a prompt. Track independence for a few sessions. Most kids will start answering alone faster, and you avoid the stress of full extinction.