Validade e fidedignidade do instrumento virtual de avaliação da fala

Abstract

The Speech Sound Disorder (SSD) is a broad term that encompasses deficits in motor and phonological aspects involved in speech production. The Speech-Language Pathologist is the professional qualified to work in the assessment, diagnosis, and treatment of SSD and, for that purpose, must have assessment tools capable of identifying difficulties and classifying the individual's speech in relation to the types of SSD present. The present study aimed to verify the reliability and validity for the word naming and word imitation tasks of the Virtual Instrument for Speech Assessment (UBERTI, 2023). This research was properly registered and approved by the Research Ethics Committee (REC) of the Federal University of Santa Maria, under the number 3.972.480. The reliability sample consisted of 20 children with phonological disorder (PD), aged between four years and eight years and eleven months (M=6.2; SD=1.4). The criterion validity sample consisted of 45 children aged between four and nine years (M=6.11; SD=1.54). Among them, 26 had PD, and 19 had Typical Speech Development (TSD). Inclusion criteria required ages between 2 and 17 years, Brazilian Portuguese as the first language, normal or corrected visual and auditory acuity, adequate performance in comprehension and expression of oral and/or written language. Exclusion criteria considered the presence of relevant socioemotional alterations, dental and/or facial alterations or disproportions, syndromes and/or genetic mutations, or complex neurobehavioral disorders. PD diagnosis was based on the INFONO. The children were divided into two groups. Subsequently, the word naming and word imitation tasks of the Virtual Instrument for Speech Assessment (UBERTI, 2023) were administered. For inter-rater reliability, two examiners retranscribed and contrastively analyzed the words from each task for 20 children. The data were evaluated using the Kendall's concordance test. For criterion validity analysis, the T-test was used to compare two independent samples. In the inter-rater analysis, the agreement in the word naming task ranged from 63% to 100%. In the word imitation task, the agreement ranged from 47% to 100%, being significant for all phonemes in both tasks. In the criterion validity analysis both tasks were able to differentiate the groups. There was no significant difference in the word naming and word imitation tasks for nasals and plosives, except for /g/ in the word naming task. The instrument showed satisfactory criterion validity evidence, as it is capable of differentiating the groups. It is noteworthy that the lack of significant differences in phonemes is consistent with the substitutions presented by children with PD, as the majority had already acquired them.