Revista CEFAC (Jun 2012)
Validação de um modelo linguístico fuzzy para classificar a gravidade do desvio fonológico Validation of a fuzzy linguistic model to classify the severity of phonological disorder
Abstract
OBJETIVO: verificar se os critérios adotados no Modelo Linguístico Fuzzy, para classificar a gravidade do Desvio Fonológico (DF), são empregados na prática fonoaudiológica e se estão de acordo com a finalidade do Modelo. Além disso, analisar a concordância entre a gravidade do DF classificada pelo Modelo e a julgada por fonoaudiólogas, bem como, verificar as principais dificuldades. MÉTODO: a pesquisa foi constituída por duas amostras. A primeira composta por um número representativo de sistemas fonológicos desviantes (n=52) classificados pelo Modelo quando o mesmo foi proposto. A segunda por dois grupos de fonoaudiólogas GF-I e GF-II. Foi solicitado que as fonoaudiólogas julgassem a gravidade dos sistemas fonológicos desviantes, mapeados no Modelo Implicacional de Complexidade de Traços - MICT, em Grave, Moderado-Grave, Moderado-Leve e Leve. Em seguida os critérios e as principais dificuldades foram descritas em um questionário. Empregou-se a Estatística Kappa, com nível de significância de pPURPOSE: to check if the criteria adopted from the Fuzzy Linguistic Model to classify the severity of Phonological Disorder (PD) are employed in the practice of speech and language therapists and whether they are in conformity with the purpose of such model. It is furthermore intended to analyze the acquiescence between PD severity classified by the Model and the one deemed by speech and language therapists and check the main difficulties. METHOD: the study included two samples. The first one was comprised of a representative number of deviating phonological systems (n=52) and classified by the Model when it was proposed. The second one comprised two groups of speech and language therapists GF-I and GF-II. The speech and language therapists were asked to deem the severity of deviating in phonological systems, mapped in the Implicational Model of Feature Complexity - MICT as Severe, Moderate-Severe, Moderate-Mild, and Mild. Next, criteria and main difficulties were described in a questionnaire. Kappa statistics was used, with a significance level of p<0.05. RESULTS: the criteria used in the proposal were generally used by GF-I and considered as adequate by GF-II. There was an acquiescence between classification of severity obtained by the Model and the one deemed by the speech and language therapists. The most frequently reported difficulty was distinguishing between intermediate degrees. CONCLUSIONS: the Fuzzy Linguistic Model criteria are used to classify PD severity when based on MICT. Furthermore, the criteria are in acquiescence with the purpose of the Model. Due to the difficulties, other forms of classifying severity can be added in order to characterize PD over other major aspects.
