Questionário vivendo com disartria (VcD): validação de conteúdo e consistência interna com base na Teoria de Resposta ao Item (TRI)

Abstract

This dissertation researched the quality of life in people with dysarthria based on the validation of an instrument called Living with Dysarthria (VcD). This dissertation is divided into two articles, the objectives of which were: 1) to present the content validation of the VcD Questionnaire for Brazilian Portuguese; and 2) present evidence of validity based on the internal structure of the VcD, with the application of Item Response Theory (IRT). The content validation article included an expert panel of five judges and involved the translated and adapted version of the VcD into Brazilian Portuguese, with 50 items. The judges judged the relevance of the item, theoretical and textual relevance. When calculating the content validity index (CVI) for item relevance, items 6, 9, 10, 15, 37 and 38 had low scores and were considered bad, and for this reason were excluded from the questionnaire. The theoretical and textual relevance of the 50 items was assessed. Items 3, 4, 6, 7, 8, 9, 10, 12, 15, 18, 24, 27, 30, 32, 37, 38 and 45 had a final Content Validity Coefficient (CVC) below that recommended, and Therefore, the 17 items were excluded from the questionnaire, which resulted in the final version of the VcD with 33 items. The second article was an internal consistency validation study and used data from 269 individuals, 127 people with neurological diagnoses and 142 people without a neurological diagnosis. The following analyzes were performed: Cronbach's Alpha, Exploratory Factor Analysis (AFE), Confirmatory Factor Analysis (CFA) and Item Response Theory (IRT). The study proceeded in two stages. The first stage demonstrated good internal consistency of the instrument, based on Cronbach's Alpha coefficient. The EFA was performed and the tests indicated an adequate sample with a significant Bartlett's test of sphericity. In the Measure of Sample Adequacy (MSA) test, all items scored above the recommended level. The AFE did not eliminate any item and indicated the factorial structure of a single factor. A single-factor CFA was carried out, where six items out of the thirty-three were excluded, namely items 5, 12, 18, 19, 22 and 25, confirming a factorial structure with 27 items. The second stage of the study was carrying out the Item Response Theory (IRT). The items with the highest difficulty (b) and discrimination (a) values were observed, in addition to the item characteristic curves (ICCs), it was seen that the items are polytomous. Thus, it is concluded that the VcD presents adequate validity evidence based on the internal structure. The current version of the instrument is composed of 27 polytomous items, which reflect a response key on a six-point Likert scale, and one factor.