ABSTRACT
Introduction: Science motivation is important for students' scientific literacy. Yet, there is a lack of valid and reliable measurement tools for the Brazilian context. This study presents the Brazilian Portuguese version of the Science Motivation Questionnaire (SMQ-II) and motivational baseline data. Method: The instrument was translated into Brazilian Portuguese using cross-cultural validation procedures. For structural validity evidence, the responses of 646 secondary school students were subjected to exploratory and confirmatory factor analysis, as well as measurement invariance. For reliability evidence, Cronbach's alpha (a) and McDonald's omega (w) were calculated. Students' motivation was analysed using 2 (gender) x 4 (grade levels) x 3 (study modality) MANOVA. Results: 24 items measuring intrinsic motivation, career motivation, grade motivation, and self-efficacy supported a four-factor structure with adequate reliability against the original five-factor structure (self-determination was not salient). Measurement invariance was established across the gender and study modalities, but not for grade levels. Higher-grade level Brazilian students were less motivated, and girls reported higher intrinsic and career motivation, but lower self-efficacy than boys. Conclusion: These findings lay the foundation for the assessment of Brazilian students' science motivation, although they also reveal problems in the latent structure of the SMQ-II and call for the development of instruments rooted in contemporary motivational theories.
Introdução: A motivação científica é importante para a alfabetização científica dos estudantes. No entanto, há uma falta de ferramentas de medição válidas e confiáveis para o contexto brasileiro. Este estudo apresenta a versão em português brasileiro do Questionário de Motivação Científica (SMQ-II) e dados de base motivacionais. Método: O instrumento foi traduzido para o português brasileiro utilizando procedimentos de validação transcultural. Para construir provas de validade, as respostas de 646 alunos do ensino médio foram submetidas à análise exploratória e confirmatória de fatores, bem como invariância de medidas. Para a evidência de confiabilidade, foram calculados o alfa de Cronbach (a) e o ômega de McDonald's (w). A motivação dos estudantes foi analisada usando 2 (gênero) x 4 (notas) x 3 (modalidade de estudo) MANOVA. Resultados: 24 itens medindo a motivação intrínseca, motivação de carreira, motivação de grau e auto-eficácia suportaram uma estrutura de quatro fatores com confiabilidade adequada contra a estrutura original de cinco fatores (a autodeterminação não foi saliente). A invariância da medição foi estabelecida através de gênero e modalidade de estudo, mas não para o nível de nota. Os estudantes brasileiros de grau superior estavam menos motivados, e as meninas relataram maior motivação intrínseca e de carreira, mas menor auto-eficácia do que os meninos. Conclusão: Estas descobertas abrem o caminho para a avaliação da motivação científica dos estudantes brasileiros, mas também revelam problemas na estrutura latente do SMQ-II e exigem o desenvolvimento de instrumentos enraizados em teorias motivacionais contemporâneas.
ABSTRACT
The development of positive attitudes toward science is one of the main priorities in science education. However, there is a lack of reliable and valid instruments to measure Spanish-speaking elementary students' attitudes towards school science. In this study, the translation and validation of the Spanish School-Science Attitude Survey (S-SSAS) is reported. The instrument was administered to 643 students enrolled in 3rd to 6th elementary grades. Psychometric evaluation of the S-SSAS provided sound evidence for validity (face, content, construct and criterion) and reliability (internal consistency and temporal stability). Content validity was confirmed through a panel of experts who reached great consensus in linking items to attitudinal constructs, with an ICC = .956. Think-aloud interviews confirmed that students have easily understood and correctly interpreted all items included, thus providing face validity for the S-SSAS. Consistent with theoretical expectations, predictive validity ranged between -.334 to 543 and concurrent validity was examined through S-SSAS correlation with two external measures of conceptual convergence that ranged from.301 to .560, thus confirming criterion validity. Construct validity was assessed by obtaining consistent results with the original scale in terms of reporting no statistically significant differences in attitudinal profiles towards school science between girls and boys and between students from urban and rural schools. Cronbach αfor the entire scale was .704, with item-total correlation ranging from .243 to .560, which reports acceptable internal consistency. Temporal stability with a 10-days span was good, with ICC = .873 and r = .464-790. Taken together, these results indicate that the Spanish single-items School-Science Attitude Survey is easy to administer and equally interpreted by both girls and boys enrolled in rural and urban elementary schools, thus being a valid and reliable instrument for measuring attitudes towards school science.
