Mediating effect of learning approaches on the relationship between student engagement in experiments and scientific inquiry competencies in technical institutions in Tanzania

Abstract

Previous studies revealed profound benefits of fostering students’ Scientific Inquiry Competencies (SICs) in science learning, yet students face challenges in developing these competencies. Limited studies exist on the learning factors for promoting students’ SICs. This study examined the mediating effect of learning approaches on the relationship between students' engagement in experiments and SICs in technical institutions in Tanzania. The study objectives were to: compare students’ level of SICs based on gender, grade level, institution nature, and science course preferences; compare level of engagement based on similar factors and SICs performance groups; assess the effect of student engagement on SICs and learning approaches; assess the effect of learning approaches on SICs and examine the mediating effect of learning approaches on student engagement and SICs. The study was guided by Astin’s and Kahn’s engagement theories. A positivist paradigm and a cross-sectional survey design were adopted. A proportionate sampling was used to draw 337 from 477 students. Data was collected using SICs tests, learning approaches, and student engagement questionnaires and analysed by t-tests, ANOVA and mediation analysis. Results revealed significant differences between male and female students in total SICs (p = .002), hypothesis formulation (p = .001), data analysis and interpretation (p = .032), and drawing scientific conclusions (p = .002) in favour of males. Also, significant differences were found between students from public and private technical institutions in total SICs (p = .002), planning and designing experiments (p = .038), and data analysis and interpretation (p = .002) in favour of public technical institutions. Further, significant differences were found between second- and third-year students in cognitive (p = .011) and social (p = .026) engagements in favour of second- and third-year students, respectively and between lower, moderate, and higher SICs performance groups in agentic (p = .009), cognitive (p = .000), emotional (p = .003), and social (p = .001) engagements, in favour of higher SICs performing students. Besides, students’ agentic, cognitive, emotional, and social engagement positively affect SICs (p = .000, .000, .000, .000) and deep learning approach (p = .000, .000, .000, .001) while not affecting surface learning approach (p = .553, 434, 061, 466) in each of the four mediation models. Also, students’ deep (p = .000, 000, 000, 000) and surface (p = .000, 000, 000, 000) learning approaches were positively and negatively affecting SICs, respectively in each of the four-mediation model. Only students’ deep learning approach found to positively and partially mediated the relationship between agentic (CI [.095, .423]), cognitive (CI [.166, .731], emotional (CI [.166, .718], and social (CI [.105, .565]) engagement and SICs. Conclusively, male students and those from public technical institutions had higher SICs than females and those from private technical institutions. Second- and third-year students had higher cognitive and social engagement, respectively and high levels of agentic, cognitive, emotional, and social engagement were linked to high SICs. Deep learning approach partially mediated the relationship between students’ engagement and SICs. The study recommends that instructors create laboratory settings conducive to all students and foster four forms of engagement and deep learning approach to enhance SICs.