Banca de DEFESA: SANDRA MARIA CONCEICAO PINHEIRO

Uma banca de DEFESA de DOUTORADO foi cadastrada pelo programa.
DISCENTE : SANDRA MARIA CONCEICAO PINHEIRO
DATA : 17/06/2021
HORA: 08:00
LOCAL: videoconferência na plataforma RNP (sala PEI-UFBA)
TÍTULO:

An approach to long-term models for analyzing student dropout survival in engineering courses: EPUFBA as a case study.

PALAVRAS-CHAVES:

Dropout. Initial semesters. Engineering. Performance measure. Survival analysis. Long-term model.

PÁGINAS: 130
GRANDE ÁREA: Engenharias
ÁREA: Engenharia de Produção
RESUMO:

Dropout and graduation rates in higher education institutions have been the subject of constant discussions, not only in Brazil, but in other countries around the world. In Brazil, the theme was part of one of the guidelines of the Plan for Restructuring and Expansion of Federal Universities (REUNI), a program that allowed federal universities conditions to expand access and permanence in higher education. There are several factors that can lead students to drop out of programs, but studies have pointed out the association of academic factors with program dropout rates. This work seeks to contribute to the construction of a cure fraction survival model, to study the effect that performance levels have on the dropout process of students from engineering programs. Student performance was measured using a proposed measure that was applied to math and physics courses of the basic cycle. Nine engineering programs from the Polytechnic School of the Federal University of Bahia (EPUFBA) were analyzed. Engineering programs were considered in this work because they present high percentages of students dropping out, and because of the current relevance of training qualified and skilled engineering professionals to work in an increasingly demanding world, due to the advancement of technological innovations. The results showed that only four of the nine engineering programs had graduate percentages between 50% and 70%, the remaining programs had graduate percentages below 50%. The programs that presented the lowest graduation percentages are those that presented the highest percentages of voluntary dropouts and the worst performance in the math and physics courses of the first semesters. The yield coefficient of students who have dropped out of programs is generally twice as low as the yield coefficient of students who have managed to complete their programs. The first four semesters of the programs are crucial for students, since more than 50% of them drop out during this period. The measure of performance evaluation in the courses, proposed in this study, proved to be efficient in identifying the performance ranges of higher risk for dropping out. The results found in this study confirm statements made in other studies, when they relate the risk of dropout to poor performance in the initial semesters. The long-term model of promotion time presented consistent estimates, and adjusted satisfactorily the academic covariates to the students' time to dropout. Through the estimates obtained by fitting the model, it was found that the probabilities of students not being susceptible to dropout are generally higher when low performance occurred in courses taught in the 2nd or 3rd semester.

MEMBROS DA BANCA:
Externo(a) ao Programa - 1429447 - ANGELO MARCIO OLIVEIRA SANT ANNA
Externo(a) ao Programa - 1502395 - EDUARDO DO NASCIMENTO
Interno(a) - 1677204 - ELAINE CHRISTINE DE MAGALHAES CABRAL ALBUQUERQUE
Presidente - 2574950 - KARLA PATRICIA SANTOS OLIVEIRA RODRIGUEZ ESQUERRE
Externo(a) ao Programa - 285400 - SONIA MARIA ROCHA SAMPAIO