Effects of the Onset of Differential Reinforcer Quality on Skill Acquisition.

Cariveau et al. (2022) asked a simple timing question. When you start differential reinforcement for unprompted corrects, should you reinforce right away or wait?

They taught new skills to three children with developmental delay. Each child got two lesson sets. One set gave a better reinforcer the moment an unprompted correct happened. The other set waited a few trials before the better reinforcer showed up.

The team counted how many teaching rounds each child needed to reach mastery.

Waiting won. Every child reached mastery faster when the good reinforcer was delayed.

Fewer exposures were needed, so total teaching time dropped. No extra prompts or fancy steps were added—just a short pause before upgrading the reward.

Three earlier studies say the opposite. van Timmeren et al. (2016), Majdalany et al. (2016), and Xue et al. (2024) all found that even tiny delays—seconds, not trials—slowed learning for children with autism.

The clash is real, but it makes sense. Those studies delayed every reinforcer by a set number of seconds. Tom et al. delayed only the special, differential reinforcer and did it by trials, not seconds. Kids still got regular praise right away.

So the papers do not cancel each other. Seconds hurt; trials can help when you reserve the best prize for later corrects.

If you run discrete trials, try holding back the top reinforcer for a handful of correct unprompted responses. You might reach mastery in fewer lessons without adding extra staff or materials. Start with a short trial delay—say three corrects—then track exposures to see if efficiency improves for that learner.