Ansiedade matemática e estatística: evidências psicométricas e influências na escolha profissional de universitários brasileiros
Carregando...
Data
Orientador
Santos, Flávia Heloísa dos 

Coorientador
Gomides, Mariuche Rodrigues de Almeida
Pós-graduação
Psicologia do Desenvolvimento e Aprendizagem - FC
Curso de graduação
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Tipo
Tese de doutorado
Direito de acesso
Acesso restrito
Resumo
Resumo (português)
O enfrentamento de conteúdos quantitativos no ensino superior pode gerar vivências emocionais negativas, como a ansiedade matemática (AM) e a ansiedade estatística (AE), comprometendo o desempenho acadêmico e influenciando decisões de carreira. Essas formas específicas de ansiedade têm despertado crescente interesse na literatura internacional, mas ainda carecem de investigações aprofundadas e instrumentos validados no contexto brasileiro. O presente estudo teve como objetivo investigar a AM e a AE, suas evidências psicométricas e as influências dessas variáveis na escolha profissional de universitários brasileiros por meio de 3 estudos os dois primeiros examinaram as propriedades psicométricas da AMAS (Abbreviated Mathematics Anxiety Scale) e da SAS (Statistical Anxiety Scale) respectivamente, o terceiro investigou as influências da AM e AE na escolha de carreira em diferentes cursos de graduação, divididos por carga matemática de disciplinas obrigatórias. O estudo foi encaminhado e aceito pelo comitê de ética em pesquisa. A amostra contou com 951 universitários das 5 regiões brasileiras com recortes (Estudo 1: n= 646; estudos 2 e 3, n= 451), obtidos após aplicação de critérios de exclusão. Os participantes responderam a um formulário on-line, que incluiu as duas escalas principais (AMAS e SAS) e variáveis gerais e específicas, relacionadas à matemática, além do questionário sociodemográfico e de questões sobre as áreas acadêmicas. As análises incluíram estatísticas descritivas, comparativas, correlações e modelos de regressão logística e linear hierárquica. Os resultados revelaram que maiores níveis de ansiedade, aliados a baixa autoeficácia e pouco gosto pela matemática, aumentam a probabilidade da não escolha e evasão de áreas STEM. Os resultados confirmaram a validade e confiabilidade da AMAS e da SAS no Brasil, ambas com estrutura bifatorial e diferenças estatísticas significativas por gênero. Além disso, mostrou que estudantes de cursos com baixa carga matemática apresentaram maiores níveis de AM e AE, enquanto os de alta carga tiveram maior autoeficácia, gosto e desempenho. A AM e a ansiedade de teste foram identificadas como preditoras da AE, enquanto a autoeficácia matemática se destacou como fator protetivo. Os achados reforçam a natureza multidimensional de AM e AE, sua associação com crenças de autoeficácia e com variáveis afetivo-cognitivas. A pesquisa contribui ao oferecer instrumentos validados para o contexto brasileiro, aprofundar a compreensão das interfaces emocionais com o aprendizado quantitativo e propor subsídios para práticas educacionais e de orientação profissional mais eficazes. Limitações incluem o delineamento transversal, a amostragem online não probabilística e o uso de medidas autorreferenciadas. Sugere-se que futuros estudos longitudinais e intervenções aplicadas possam aprofundar a compreensão e ampliar estratégias de apoio a estudantes vulneráveis. Conclui-se que AM e AE são construtos centrais para explicar trajetórias acadêmicas e escolhas profissionais, influenciados por gênero, autoeficácia e motivação.
Resumo (inglês)
Confronting quantitative content in higher education can generate negative emotional experiences, such as math anxiety (MA) and statistical anxiety (SA), compromising academic performance and influencing career decisions. These specific forms of anxiety have generated growing interest in the international literature, but they still lack in-depth research and validated instruments in the Brazilian context. This study aimed to investigate MA and SA, their psychometric evidence, and the influence of these variables on the career choices of Brazilian university students through three studies. The first two examined the psychometric properties of the AMAS (Abbreviated Mathematics Anxiety Scale) and the SAS (Statistical Anxiety Scale), respectively. The third investigated the influence of MA and SA on career choices in different undergraduate programs, divided by the mathematics load of required courses. The study was submitted to and accepted by the research ethics committee. The sample comprised 951 university students from the five Brazilian regions (Study 1: n=646; Studies 2 and 3, n=451), obtained after applying exclusion criteria. Participants completed an online form, which included the two main scales (AMAS and SAS), general and specific variables related to mathematics, in addition to a sociodemographic questionnaire and questions about academic fields. The analyses included descriptive and comparative statistics, correlations, and logistic and hierarchical linear regression models. The results revealed that higher levels of anxiety, combined with low self-efficacy and little interest in mathematics, increase the likelihood of not choosing and dropping out of STEM fields. The results confirmed the validity and reliability of the AMAS and SAS in Brazil, both with a bifactor structure and significant statistical differences by gender. Furthermore, it showed that students in courses with a low mathematics load had higher levels of MA and SE, while those with a high load had higher self-efficacy, interest, and performance. MA and test anxiety were identified as predictors of SE, while mathematical self-efficacy emerged as a protective factor. The findings reinforce the multidimensional nature of MA and SE, their association with self-efficacy beliefs and affective-cognitive variables. This research contributes by offering validated instruments for the Brazilian context, deepening the understanding of the emotional interfaces with quantitative learning, and proposing support for more effective educational and career guidance practices. Limitations include the cross-sectional design, non-probability online sampling, and the use of self-reported measures. We suggest that future longitudinal studies and applied interventions could deepen understanding and expand support strategies for vulnerable students. We conclude that MA and SE are central constructs in explaining academic trajectories and career choices, influenced by gender, self-efficacy, and motivation.
Descrição
Palavras-chave
Ansiedade matemática, Ansiedade estatística, Estudantes universitários, Escolha de carreira, STEM e não-STEM, Math anxiety, Career choices, Statistics anxiety
Idioma
Português
Citação
BONOME-VANZELLI, Silvia Regina Cassan. Ansiedade matemática e estatística: evidências psicométricas e influências na escolha profissional de universitários brasileiros. 168 f. Tese (Doutorado em Psicologia do Desenvolvimento e Aprendizagem) – Faculdade de Ciências, Universidade Estadual Paulista (UNESP), Bauru, 2025.

