Learning speech-internal cues to pronoun interpretation from co-speech gesture: a training study.

The final sample was small: 21 female and 21 male children, with 14 per condition, and eight children did not complete. The training and testing schedules varied widely, including average 11.2 days to complete the study and gaps between training and post-test. The significant increase appeared only in the Gesture condition; the other conditions did not change significantly, and the difference between conditions was not significant. The study did not test children with language impairment, so it does not show that this training works as a clinical intervention. There was no independent measure showing that children understood the gestures, and the authors note only some children may have learned to read them. The post-hoc analyses of starting point and change were small and inconsistent, and the authors advise caution. The same test video was used before and after training, so familiarity could be a concern, although the authors argue it would not explain increased first-mentioned responses. The training narratives were all produced by the same speaker, so a speaker-specific pattern cannot be fully ruled out.

No conflicts-of-interest or funding declaration is included in the supplied text. The publication types list NIH extramural and non-U.S. government research support.

Do not conclude that gesture training is a proven clinical intervention or clearly better than other training. The Gesture condition was the only condition with a significant within-condition increase, but the difference between conditions was not significant, the sample was small, children with language impairment were not tested, and gesture and pronoun ambiguity were not separated.