Classification of speech nasality of individuals with cleft lip and palate with distinct ordinal scales

To investigate whether there is a difference in the classification of speech hypernasality by inexperienced listeners using different ordinal scales; to verify the agreement of the listeners in the analyses when using these scales; and to verify whether the order in which the scales are presented influences the results.

Twenty Speech-Language Pathology students classified the degrees of hypernasality of 40 (oral) samples from patients with cleft lip and palate. Ten performed the classifications using a 4-point scale (absent, mild, moderate, and severe) and, after two weeks, using a 3-point scale (absent, slightly hypernasal, and very hypernasal). Other ten students performed the same classifications, but in reverse order. The classifications were made remotely and documented on a form.

The average percentage of correct responses by the students, in relation to the gold standard, was significantly higher for the 3-point scale. There was no significant interaction between the order of presentation and the scale for the percentage of correct classifications. The students' agreement with the gold standard assessment was fair (3-point scale) and moderate (4-point scale). The mean percentage of agreement of the intra-rater analyses was significantly higher for the 3-point scale. There was no significant interaction between presentation order and scale for the percentage of intra-rater classifications. The Kappa coefficient index showed more favorable intra-rater agreement for the reduced scale.

The reduced scale favored the classification of speech hypernasality by listeners and can be considered an important strategy to favor the initial evaluations of students in Speech Therapy during their training.