An evaluation of the number of presentations of target sounds during stimulus-stimulus pairing trials.

Two boys with autism who rarely spoke got 10-minute pairing sessions. The adult said one speech sound and gave a toy or snack right after. In another phase the adult said three sounds in a row before the treat.

Each child had both conditions. The team counted how many speech-like sounds the boys made during each type of pairing.

One sound per treat made both boys babble more. Three sounds per treat cut their vocal play in half.

The single-sound rule kept the reward close to the target noise. The triple-sound rule delayed the treat and the boys lost interest.

Bishop et al. (2020) and Muharib et al. (2021) later added echoic prompts and speech tablets. They still kept the reward tight after one clear model, echoing the single-sound lesson.

Rosales et al. (2012) ran the same pairing game with typical preschoolers learning Spanish. One sound worked there too, showing the rule holds across languages and diagnoses.

Kaneda et al. (2025) seem to disagree. They turned off tablet voice output and still got more real words. The trick is they also delayed the reinforcer, just like three sounds did. Their kids were older and already had some speech, so a brief wait did not shut them down. For brand-new vocalizers, instant payoff still wins.

If you are building first sounds with a minimally-verbal child with autism, pair one adult sound with one quick reinforcer. Keep the gap under a second. Skip long word chains or extra instructions. This tiny timing tweak can double early vocal play and set the stage for echoic training and speech devices later.