Dissertação de Mestrado
Lei de cotas: um estudo da reserva de vagas em uma instituição federal de ensino através da descoberta de conhecimento em bases de dados
Fecha
2014-09-26Autor
Márcio Teodoro Dias
Institución
Resumen
The work reports a case study to analyze the effectiveness of the places reserved in federal educational institutions by the Brazilian Quota Law. The study contextualizes the law and its implementation at the Federal Institute of Education, Science and Technology of Minas Gerais (IFMG), and sought to evaluate the actual occupancy of the reserved places; to verify the extent to which people who have used their right to the reserve; find out if there are people registered to places without the right to do so; and find out if there are significant differences in the profiles of candidates as to be enrolled in quotas in relation to candidates enrolled in wide competition. Based on concepts of Knowledge Discovery in Databases, the study describes data extraction and preparation and applied a data-mining tool to the extracted data. Data refer to entries in the selection process and enrollment in IFMG for a period of just over a year. The results are analyzed, showing that a significant proportion of reserved places were not occupied; that many candidates who could enroll in reserved spaces did not; that many candidates competed for the reserved places without owning the requirements prescribed by law; and there were no significant differences in the admission of students to the requirements by the quota system or the wide competition, which suggests that the quota system would be unnecessary for the institution studied. Finally, the results suggest the need for greater advertisement of the quota policy, call attention to the importance of verifying the data provided by the candidates who ran for the places and pointed to the need for future studies in other federal educational institutions on the effectiveness of the quota policy, to contribute to the debate on the relevance of this policy for the country.