Estudo de construções negativas do português do Brasil em enunciados de questões objetivas de concursos

Abstract

This study focuses on the analysis of negative constructions in Portuguese present in the statements of exam questions. It aimed to investigate the gap in the analysis of negation in objective exam questions. An experimental methodology was used, employing eye tracking and analysis of a corpus of questions presented by exam boards. The study was developed in two stages. The first stage involved an experiment with eye tracking to verify how negative sentences were read compared to affirmative sentences. Online studies were conducted, with the reading of negative and affirmative sentences using the eye-tracking device, in order to verify the readers' gaze behavior, including eye movements such as regressions, fixations, and saccades during the reading of negative sentences, and subsequently compared with the reading of the same sentences in their affirmative versions. Based on the analysis of the differences observed in the readings, it was noticed that reading negative structures seemed more costly than reading affirmative structures. Bearing this in mind, the second stage involved analyzing exam questions with statements containing negative words to explore a possible relationship between the use of negative commands and the number of successful candidates. The intention was to investigate whether the higher processing cost of reading negative sentences could somehow interfere with the formulation of exam questions and the number of successful candidates. An analysis of questions from three exam boards was conducted through searches on the boards' websites (Vunesp, AOCP, and Publiconsult) and preparatory courses for competitive exams, as well as online interviews with members of the commissions. The results of the experiment in the first stage and the question analyses suggested that negative structures appear to have a higher linguistic processing cost, and therefore, the results obtained suggest that statements containing these structures may demand greater cognitive effort during reading processing. In this sense, the hypothesis arises that the use of such structures, in evaluative contexts, may influence the performance of candidates. However, this possibility should be interpreted with caution, since it is not a causal relationship directly tested in this study, but an exploratory inference that requires further investigation in future research.