Development and evaluation of a speech-generating AAC mobile app for minimally verbal children with autism spectrum disorder in Mainland China.

Only ten children were studied, and there was no control group, so the study cannot show that Yuudee caused the request-training gains. All ten children also attended ABA training during the five-week Yuudee training period, so the contribution of Yuudee cannot be separated from ABA. There were no follow-up sessions, so it is unknown whether the children maintained their requesting skills after training ended. Training and evaluation occurred only in a classroom, so it is unknown whether requests generalized to home or other settings. The study did not record whether parents trained children to use Yuudee at home between sessions, and such home training may have influenced the results. Some children were not trained in every phase, and two children skipped phase IVA because of a trainer oversight, so the phase sequence was not the same for all participants. The accuracy threshold was set at 50% because training was short and included five phases, so a child labeled accurate may still have needed prompts in many trials. Vineland scores did not predict training effectiveness, and the authors note the small sample may explain this. The study did not test whether turning off speech output promotes spontaneous speech, although a few informal observations were reported.

The supplied text lists Beijing Municipal Science & Technology Commission as the funder. It does not provide a further conflict-of-interest declaration.

Do not conclude that Yuudee alone improves requesting skills. The study had no control group, all ten children also attended ABA during the five-week training period, and there were no follow-up sessions or outside-classroom tests.