Independência funcional de operantes verbais vocais em crianças implantadas

Abstract

One of the possibilities of promoting more effective interventions in speaking skills in cochlear implant rehabilitation is to distinguish the control of stimuli exercised over the errors emitted when vocalizing. Previous studies showed that errors occurred more frequently in naming tasks (pictures as stimuli control) than reading task (textual as stimuli control). There was no vocalization analysis in echoic tasks (spoken words as stimuli control) reported. The objective of the current study was to characterize the vocalization errors produced by implanted children in echoic, naming, and textual behavior tasks. Six children aged from 8 to 11 years old took part in the study; they were cochlear implant users diagnosed with sensorineural, bilateral hearing loss. The participants were evaluated according to their auditory age and their levels of reading and writing. During the initial assessment, a pre-training was conducted aiming at teaching “matching to sample task”, that was used in some stages of this study. Thirty six stimuli that included all phonemes and allophones from Brazilian Portuguese were adopted. After the word recognition stage, the participants performed three tasks: echoic behavior, naming and textual tasks. The speech responses emitted were recorded and transcribed phonetically. Vocalizations classified as totally or partially correct were analyzed according to different controlling antecedent stimuli (dictated words for echoic, figures for naming, and written words for textual) and audiological criteria (correspondence, number of syllables, pattern of tonic syllable, initial and final phonemes, and vowel production) to identify the types of errors. Results showed that most vocal distortions were emitted during the naming task according to previous studies. In this task, distortions in the number of the syllables and in the stressed syllable indicated that the participants had difficulties in emitting grapheme-phoneme conversion. In the textual task the participants had few errors because the printed syllable exercised more precise control over the vocalized syllable. The same occurred in the echoic task, which the participants, in general, made few mistakes once the antecedent stimulus had a formal correspondence with the response emitted, controlling the speech emission more effectively. This result validates previous studies that demonstrated that, after training, the accurate emission of echoic behavior improves the speech quality of implanted children. Data indicate that the most significant distortions and the greatest amount of errors occurred in the vocalizations issued in the naming task in which the stimulus control was by non-verbal stimuli, which required the vocalization to be emitted under control of stimuli that were not present at the time of the task. In case of echoic and textual tasks, different stimuli (spoken word and printed word) controlled the accuracy speech. We argue that improving instructional procedures to teach speaking skills should go through specific understanding of these stimuli control, promoting integration between verbal operant. The instructional procedures should consider that if the control of stimuli of different natures over the speech sound units is not yet properly established, it is an indicative of a failure in this process of planning of teaching condition.