Níveis de maturidade e capacidade de pensamento computacional dos alunos ao término do ensino fundamental explicados por modelos de equações estruturais: uma investigação dos fatores de chances por meio de ajustes de regressão logística
Carregando...
Data
2021-02-19
Autores
Souza, Eduardo Cardoso de
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O conceito de Pensamento Computacional tem sido polêmico e estudos recentes mostram que esse tipo de pensamento não está bem estabelecido, sobretudo no currículo da Educação Básica. Contudo, esse trabalho teve como objetivo construir uma métrica que representa a capacidade de Pensamento Computacional dos alunos ao término do Ensino Fundamental. Para isso, foi utilizado um instrumento espanhol de diagnóstico de Pensamento Computacional, que por meio de uma abordagem quali-quantitativa de característica exploratória-descritiva, atendendo ao rigor das técnicas estatísticas de análise multivariada de dados, possibilitou construir e validar um modelo integrado por três dimensões estruturantes: algoritmo, reconhecimento de padrão e decomposição. Buscou-se responder à pergunta chave de quais são as habilidades que os alunos ao término do Ensino Fundamental de uma cidade do oeste paulista mobilizam e articulam diante da resolução de problemas, por meio de tarefas desplugadas de fundamentos da Ciência da Computação. Um total de 462 alunos responderam voluntariamente ao teste de Pensamento Computacional no período de outubro a dezembro de 2019, o que corresponde a 74,6% dos concluintes do Ensino Fundamental no município de Santa Cruz do Rio Pardo. Os resultados mostraram uma concentração elevada de alunos nos níveis de maturidade mais baixo das escalas (Níveis 1 e 2) e o ajuste de regressão logística para um Pensamento Computacional adequado revelou que o sexo dos alunos e o tempo de game digital são variáveis que não interferem nessa capacidade. A dependência administrativa da unidade escolar do aluno influencia significativamente o nível de maturidade e capacidade do Pensamento Computacional dos alunos (p<0,01), o que tornou viável apenas a comparação de alunos de mesma dependência. A chance de alunos de escola particular terem um nível adequado de Pensamento Computacional é 7 vezes a chance de alunos de escola pública. O percentual de alunos no mais alto nível de maturidade da escala (Nível 4) são mais similares quando oriundos de escolas particulares, enquanto os alunos das escolas públicas têm menor variabilidade nos menores níveis. A chance de um aluno com desempenho excelente em Matemática ter um Pensamento Computacional adequado é 2,3 vezes a de um aluno com desempenho insuficiente na disciplina. Não foram encontradas evidências estatisticamente significativas de que a autoeficácia diante de tarefas desplugadas de fundamentos da Ciência da Computação interferem na capacidade de Pensamento Computacional. Assim, esta pesquisa desconstruiu as crenças sobre meninas terem menos aptidão no campo da Ciência da Computação e esse mito deve ser rejeitado por professores e alunos. Os resultados apontam uma preocupação com o fato de poucos alunos oriundos de escolas públicas se encontrarem nos níveis mais altos de proficiência (Níveis 3 e 4), o que pode indicar que o país não está se preparando para formar adequadamente alunos para funções mais complexas, que demandam conhecimento de fundamentos da Ciência da Computação, como as da engenharia e áreas afins.
The concept of Computational Thinking has been controversial and recent studies show that this type of thinking is not well established, especially in the Basic Education curriculum. However, this work aimed to build a metric that represents the computational thinking capacity of students at the end of elementary school. For this, a Spanish Computational Thinking diagnostic instrument was used, which, through an exploratory-descriptive quali-quantitative approach, meeting the rigor of the statistical techniques of multivariate data analysis, made it possible to build and validate a model integrated by three structuring dimensions: algorithm, pattern recognition and decomposition. We sought to answer the key question of what are the skills that students at the end of elementary school in a city in the west of São Paulo mobilize and articulate in the face of problem solving, through tasks unplugged from the fundamentals of information science. A total of 462 students who responded to the Computational Thinking test in the period from October to December 2019, which corresponds to 74.6% of the graduates of elementary school in the municipality of Santa Cruz do Rio Pardo. The results showed a high concentration of students at the lowest maturity levels of the scales (Levels 1 and 2) and the logistic regression adjustment for adequate computational thinking revealed that the students' sex and the digital game time are variables that do not interfere in that capacity. The administrative dependence of the student's school unit significantly influences the level of maturity and capacity of students' Computational Thinking (p <0.01), which made only the comparison of students of the same dependency viable. The chance of private school students having an adequate level of computational thinking is 7 times the chance of public school students. The percentage of students at the highest level of maturity on the scale (Level 4) are more similar when they come from private schools, while students from public schools have less variability at the lower levels. The chance of a student with excellent mathematics performance having adequate computational thinking is 2.3 times that of a student with insufficient performance in the discipline. No statistically significant evidence was found that self-efficacy in the face of unplugged tasks from computer science fundamentals interferes with computational thinking capacity. Thus, this research has defused beliefs about girls having less aptitude in the field of computer science and this myth must be rejected by teachers and students. The results point to a concern that few students from public schools are at the highest levels of proficiency (Levels 3 and 4), which may indicate that the country is not preparing to adequately train students for more complex functions, that require knowledge of computer science fundamentals, such as engineering and related areas
The concept of Computational Thinking has been controversial and recent studies show that this type of thinking is not well established, especially in the Basic Education curriculum. However, this work aimed to build a metric that represents the computational thinking capacity of students at the end of elementary school. For this, a Spanish Computational Thinking diagnostic instrument was used, which, through an exploratory-descriptive quali-quantitative approach, meeting the rigor of the statistical techniques of multivariate data analysis, made it possible to build and validate a model integrated by three structuring dimensions: algorithm, pattern recognition and decomposition. We sought to answer the key question of what are the skills that students at the end of elementary school in a city in the west of São Paulo mobilize and articulate in the face of problem solving, through tasks unplugged from the fundamentals of information science. A total of 462 students who responded to the Computational Thinking test in the period from October to December 2019, which corresponds to 74.6% of the graduates of elementary school in the municipality of Santa Cruz do Rio Pardo. The results showed a high concentration of students at the lowest maturity levels of the scales (Levels 1 and 2) and the logistic regression adjustment for adequate computational thinking revealed that the students' sex and the digital game time are variables that do not interfere in that capacity. The administrative dependence of the student's school unit significantly influences the level of maturity and capacity of students' Computational Thinking (p <0.01), which made only the comparison of students of the same dependency viable. The chance of private school students having an adequate level of computational thinking is 7 times the chance of public school students. The percentage of students at the highest level of maturity on the scale (Level 4) are more similar when they come from private schools, while students from public schools have less variability at the lower levels. The chance of a student with excellent mathematics performance having adequate computational thinking is 2.3 times that of a student with insufficient performance in the discipline. No statistically significant evidence was found that self-efficacy in the face of unplugged tasks from computer science fundamentals interferes with computational thinking capacity. Thus, this research has defused beliefs about girls having less aptitude in the field of computer science and this myth must be rejected by teachers and students. The results point to a concern that few students from public schools are at the highest levels of proficiency (Levels 3 and 4), which may indicate that the country is not preparing to adequately train students for more complex functions, that require knowledge of computer science fundamentals, such as engineering and related areas
Descrição
Palavras-chave
Pensamento computacional, Ensino fundamental, Níveis de maturidade e capacidade, Análise multivariada de dados, Modelagem de equações estruturais, Modelo de regressão logística, Computational thinking, Elementary school, Maturity and capacity levels, Multivariate data analysis, Structural equation modeling, Logistic regression model