Perception-production relationship in children with phonological disorder during speech therapy.

There was no control group. All 16 children received the intervention, so the rise in production scores cannot be attributed to the therapy rather than to normal development, other input, or repeated testing. The sample is small and chosen by convenience, and the authors themselves say the severity findings should be interpreted with caution because of the limited sample size. Seven of the 16 children had not acquired the target sounds by the end and some individual scores were unchanged or lower, so the group averages make the therapy look more consistently effective than it was. The correlations between perception and production, which carry the paper's main argument, are small (0.08 to 0.12 on the scale the authors describe, where 0 means no relationship). The intervention model is described as unpublished and still being developed and validated, so the exact treatment cannot be reproduced from this paper. All the children were monolingual speakers of Brazilian Portuguese aged 4 to 8 with liquid substitution, so the findings may not carry to other languages, ages or sound patterns.

Public research funding only: CAPES (finance code 001), FAPESP (grants 2019/12749-2 and 2020/03990-5) and CNPq (grant 301735/2019-0), all Brazilian government agencies. No competing interests are declared and the section gives no other declarations.

Do not read the rise in production scores as proof that this therapy worked, or that adding self-perception work is what helped. All 16 children received the intervention, there was no comparison group, seven of the 16 had still not acquired the target sounds at the end, and the links the authors report between perception and production are weak (0.08 to 0.12 on their own scale, where 0 means no relationship).