Tamanho ótimo amostral e análise fatorial e correlacional do desempenho de indivíduos sob a influência de plataformas computacionais de apoio ao ensino

• Ano de defesa: 2018
• Autor(a) principal: Ribeiro, Taffarel Brant
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Ciência da Computação
• Palavras-chave em Português: Tamanho ótimo de amostra; Optimal sample size; Análise de variância fatorial; Factorial analysis of variance; Correlação de Spearman; Spearman's Rank Correlation; Learning analytics; Computação; Ensino auxiliado por computador; Tecnologia educacional; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufu.br/handle/123456789/24190; http://dx.doi.org/10.14393/ufu.te.2019.310

Usage of new technologies in educational scope raises several questions about the efficiency of these approaches and which benefits they provide to the academic field. Investigations in this area cover a line of research called Learning Analytics and, in the literature, many papers that analyze new technological proposals are only aimed at observing improvements that the use of tools can cause. Such researches do not analyze whether the sample size is robust to ensure reliability of results or whether the tool enhancement tends to maintain some influence over students' performance. Based on this, this thesis determined an optimal sample size of 25 students for the performance analysis of students who do not use teaching support technologies and of 20 students for classes in contact with educational platforms. An Experiments Manager was also developed to organize the visibility of Classroom eXperience (CX) platform functionalities and, using this Experiments Manager, a Factorial Analysis of Variance and a Correlation Analysis were performed. It was observed that students' performance was influenced by the interaction between CX functionalities and the courses taken by students. In all undergraduate classes, there were significant increases in student performance in a comparison between the absence of CX and its use with the platform functionalities. Theoretical and mathematical undergraduate courses also presented moderate correlations between the platform usage level and students' performance. Thus, the platform usage positively influenced the grades of undergraduate students and it was inferred that students who interacted more with CX also obtained the best grades in their classes. In graduate classes, there was no significant difference in students performance between CX levels of usage, nor the occurrence of correlations that indicated something similar to what happened with undergraduates, although there have also been increases in students' performance at this academic level.