O efeito da segmentação e da velocidade de legendas no processamento linguístico de surdos e ouvintes em vídeos de campanhas políticas: um estudo de rastreamento ocular

Abstract

The present study is part of the EXLEG (Experimental Research in Subtitling) project, coordinated by Dr. Vera Lúcia Santiago Araújo form The State University of Ceará (UECE). With the support of the eye tracking methodology, our main goal was to investigate, in a descriptive and quantitative way, the influence of the speed of subtitles, as well as of the good and ill linguistic segmentations, in the reception of SDH (Subtitles for the Deaf and Hard of Hearing) by deaf and hearing people when watching videos of political campaigns with oneline subtitles. Our corpus was composed of previously collected data, provided by Monteiro (2016), which consisted of the eye path of 64 participants. Our goal was to clean and organize this data to analyze them statistically. The dependent variables that we examined with the assistance of the eye tracking technique were the number and duration of fixations, the regressions, and the deflections. For the statistical analysis, we used the Linear Mixed Model, through the RStudio software. Our results demonstrated that the RBS (fast and well segmented) condition seems to be the most favorable for both groups of participants. We also found evidence that indicates that the speed influences the linguistic processing of subtitles by deaf and hearing people, and that adequate linguistic segmentation facilitates this processing.