Teaching Children With Hearing Loss to Recognize Speech: Gains Made With Computer-Based Auditory and/or Speechreading Training.

Group assignment was not truly random. Children were placed into a training group and given a talker by the order they were recruited, so the three groups may differ in ways the study did not measure. No child went without training. Because there was no untrained comparison group over the four weeks, the improvements cannot be cleanly separated from ordinary development during that period. Every child could be fitted with hearing aids or cochlear implants and had good aided hearing, so the results say nothing about children with poorer aided access to sound. The children were tested only straight after training finished, so the study says nothing about whether the gains lasted. The same test order was used for every child rather than varied, and the three kinds of speech test (words, phrases, sentences) were not compared with each other. All training stimuli were spoken by white Midwestern-English actors in their twenties, so it is unclear how well the findings carry over to other accents and talkers. The gains are modest, roughly 9% and 7% words correct on average, and come from a single program developed by the authors.

The text supplied does not include a funding or competing-interest declaration. It states only that the children received a tablet, a speaker, and $10 an hour for taking part, and that the protocol was approved by the Washington University institutional review board.

Do not read this as proof that the training caused the improvement. Every child received training and there was no untrained group to compare against, so the gains could partly reflect ordinary development over the four weeks. Do not assume the improvements lasted either, since the children were tested only immediately after training.