Emissões vocais âncoras em um sítio de treinamento e naavaliação perceptivo-auditiva de vozes: análise da efetividade

• Ano de defesa: 2018
• Autor(a) principal: Priscila Campos Martins
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: voz; qualidade da voz; treinamento da voz; percepção auditiva; disfonia; Voz; Disfonia; Percepção Auditiva; Treinamento da Voz; Qualidade da Voz
• Link de acesso: http://hdl.handle.net/1843/FRSS-BB9GR4

Introduction The auditory-perceptual analysis of the human voice is the main way of evaluation in vocal clinical and speech-language-pathology research. There are several advantages to this: it allows perceptual descriptions that cover several vocal parameters, it is a quick, painless and comfortable method to the patient, and besides, it does not depend on equipment, generating a low cost. However, the auditory-perceptual assessment of voice depends on several internal standards, which probably justifies the high variability in intra and inter-rater agreement and makes this evaluation subjective. Some tasks are pointed out as factors that can optimize the result of auditoryperceptual evaluation: the use of scales, anchor stimuli, synthesized stimuli and auditory training. Purpose To analyze the effect of auditory-perceptual training with anchor stimuli of natural voices in the inter-rater agreement of vocal evaluation. To analyze the effect of auditory-perceptual evaluation with anchor stimuli of synthesized voices in the intra and inter-rater agreement of vocal evaluation. Methods This is a quantitative study. This project was structured in two blocks. In the first block, it was developed atraining site composed of anchors of natural voices, in which the GRBASI (R - roughness, B - breathiness, A - asthenia, S - tension and I - instability) parameters scale were trained and evaluated. Fifteen inexperienced evaluators were selected to perform training and vocal assessment, subjects of both genders aged between 18 and 28 years old. This block was composed of two Activities: Control Activity and Auditory Training Activity. The same evaluators performed the two activities, with na interval of 15 days between them, and the order of accomplishment of activities was drawn for each participant.In the Control Activity, the evaluators should, in the first stage, evaluate 57voices classifying them according to the degree of deviation (0 no deviation, 1 light deviation, 2 moderate deviation or 3 intense deviation) of each GRBASI parameter scale. In the second stage, they should read a text. Finally, in the third stage, they should carry out the same evaluation of the first stage. Voices were randomized at each stage.In the Auditory Training Activity, the first and third stages were the same as the Control Activity. In the second stage, the RBASI parameters were trained separately. In the training, the evaluators should listen to the voice to be evaluated,listen to the anchor stimulus and again listen to the stimulus to be evaluated and classify it as to the degree of deviation of the parameter under training. For each parameter, four voices were evaluated, except for the parameter A, in which two voices were evaluated, totaling 56 voices. The anchors stimuli were composed by natural voices.In the second block, an Auditory Calibrator was developed, consisting of two activities in which the roughness and breathiness parameters of the GRBASI scale were classified as 0 no deviation, 1 light deviation, 2 moderate deviation or 3 intense deviation. Thirty two inexperienced evaluators, subjects of both genders aged from 18 years old, were selected for this block and performed the two activities, with an interval of 15 days between them. The order of accomplishment of the activities was drawn for each participant. In Evaluation Activity, 25 voices were evaluated with the support of anchored stimuli of synthesized voices. The evaluators should listen to the natural voice to be evaluated, listen to the anchors of synthesized voices for each degree of the R parameter, and, again, listen to the voice being evaluated, classifying it as 0 no deviation, 1 light deviation, 2 moderate deviation or 3 intense deviation for this parameter. They repeated the process to classify the same voice for the breathiness parameter (B) In the Control Activity the same voices were evaluated, however, without the support of the anchor stimuli.For the analysis of the intra and inter-rater agreement, the AC1 coefficient of Gwet and Kappa was used, and to compare the concordances, the confidence interval (CI) and the percentage variation of agreement were used. Statistical analysis was performed using R version 3.3.1 software. A significance level of 5% was considered for all analyzes. Results As a result of the first block of this research, it was found that there was a decrease of inter-rater agreement in the Control Activity for the RBS and I parameters. In the Auditory Training Activity with natural anchor stimuli, the interrater agreement increased for all parameters except for asthenia, but this improvement did not reach statistical significance. As a result of the second block of this research, it was verified that the inter-rater agreement was significantly higher for the intense degree of the breathiness parameter in Evaluation Activity when compared to Control Activity, as well as the intra-rater agreement of the roughness parameter.Conclusion Training associated with natural voice anchor stimuli suggests an increase in inter-rater agreement in the auditory-perceptual evaluation of voice, indicating that new internal references have been established. The use of vocal anchors stimuli synthesized in the vocal evaluation seems to improve the intra and inter-rater agreement in auditory-perceptual analysis of the voice.Keywords: voice, voice quality, dysphonia, voice training, auditory perception.