Comparative Analysis of STEM Teaching Efficacy Across School Types

Thesis Abstract

The study addresses the persistent inequities in STEM teaching efficacy across different school types, examining how instructional practices, teacher beliefs, and contextual resources influence student engagement and achievement in Science, Technology, Engineering, and Mathematics. Despite widespread policy emphasis on improving STEM outcomes, there is limited empirical understanding of how school type (public, private-religious, and public charter) differentially shapes teaching efficacy and its translation into student learning. The aim is to compare STEM teaching efficacy across school types and identify the mechanisms through which school context affects instructional quality. Specific objectives are (1) to assess differences in teacher self-efficacy in delivering STEM instruction across school types; (2) to evaluate variations in observed instructional practices using a standardized STEM Pedagogical Observation Protocol; (3) to examine the relationship between teacher efficacy, classroom practices, and student STEM achievement; (4) to investigate how school-level factors such as teacher workload, professional development opportunities, and resource availability moderate this relationship; and (5) to generate evidence-based recommendations for policy and practice to promote equitable STEM teaching across school contexts. The methodology employs a comparative cross-sectional design. The population comprises secondary school STEM teachers and students from three school types within a metropolitan region public urban, private-religious, and public charter schools. A multi-stage sampling strategy yields 180 teachers (60 per school type) and approximately 4,500 students (1,500 per school type) who participate in paired teacher-student assessments. Data collection instruments include the STEM Teaching Efficacy Beliefs Instrument (a validated scale adapted to regional context), a Classroom Observation Checklist aligned with the Framework for Effective STEM Instruction, standardized STEM achievement tests administered at the end of the academic year, and a School Context Survey capturing resources, professional development, and workload indicators. Instrument validity and reliability are established through confirmatory factor analysis (CFA) and internal consistency metrics (Cronbach’s alpha ? .80). Data analysis integrates quantitative and qualitative elements. Descriptive statistics summarize central tendencies and dispersion for efficacy beliefs, observed practices, and achievement. Multivariate analysis of variance (MANOVA) tests for differences in teacher efficacy and observed practices across school types, followed by hierarchical linear modeling (HLM) to account for student-level outcomes nested within classrooms and schools. Structural equation modeling (SEM) examines the mediating role of classroom practices between teacher efficacy and student achievement, while multi-group SEM assesses potential moderation by school type. Additionally, regression analyses identify predictors of high-quality STEM instruction, including resource adequacy and professional development intensity. Qualitative components comprise thematic analyses of teacher interviews to contextualize quantitative findings and to identify mechanisms by which school environments influence instructional efficacy, interpreted through the Self-Efficacy Theory (Bandura) and the Sociocultural Theory of Learning. Key expected findings include statistically significant differences in teacher STEM self-efficacy and observed instructional quality across school types, with private-religious schools showing higher efficacy and more frequent use of inquiry-based practices, potentially linked to distinctive professional development ecosystems and resource access. It is anticipated that higher teacher efficacy and richer classroom practices will be positively associated with student STEM achievement, with stronger effects in school types offering robust STEM resources and targeted professional development. The SEM is expected to reveal that classroom practices partially mediate the effect of teacher efficacy on student outcomes, and that school context moderates this mediation. The study contributes to knowledge by offering robust cross-context evidence on how school type shapes STEM teaching efficacy and its relation to student learning, informing policy debates on resource allocation, professional development design, and equitable STEM education. Recommendations will target policymakers to tailor intervention packages by school type, school leaders to strengthen STEM-identity and instructional capacity, and teacher education programs to emphasize context-responsive efficacy development. The conclusion emphasizes the critical role of aligned school resources and professional development in translating teacher efficacy into effective STEM instruction across diverse school environments, with actionable guidelines for scaling equitable practices.