Enhancing senior high school student engagement and academic performance using an inclusive and scalable inquiry-based program

The study addresses how an inclusive, inquiry-based, term-long program (Discovery) can improve engagement and academic outcomes for senior high school STEM students, particularly within non-specialized public schools. The context highlights a gap between high school’s knowledge-based instruction and post-secondary experiential, mastery-based learning, which can discourage students from persisting in STEM. Ontario’s curriculum structure makes advanced STEM courses in Grades 11–12 optional, so maintaining interest and ensuring success are critical. Prior models such as specialized STEM schools and university-led outreach show benefits but are not universally accessible or often lack immersive, iterative experiences. The purpose is to implement and evaluate Discovery—anchored in biomedical engineering (BME)—to immerse entire classes in authentic, problem-based learning that fosters critical thinking, collaboration, and application of curriculum to real-world challenges. The study hypothesizes that such an environment will enhance student engagement, attitudes toward STEM, and performance, especially among under-performing and multi-term participants.

Existing literature indicates that early and varied exposure to STEM is vital for persistence, but many studies are US-centric and context differs in Ontario. Specialized STEM schools and longer-term immersive experiences (after-school, camps) support self-determination, inquiry, and research engagement, leading to stronger persistence in STEM. Effective university-led programs emphasize hands-on learning, real-life applications, immediate feedback, and problem-based projects. Student attitudes strongly influence intention to persist in STEM, potentially more than self-efficacy, although self-efficacy remains important for motivation and persistence. Learning communities and cooperative learning improve attitudes, motivation, and persistence. However, access to specialized programs is limited by geography and resources, supporting the need for scalable, inclusive models embedded in public schools that deliver authentic, inquiry-driven experiences aligned with curriculum.

Design: Quasi-experimental observational study of a mandatory, course-integrated, inquiry-based program (Discovery) implemented over five consecutive academic terms in one public high school (Grades 11–12 Biology, Chemistry, Physics), with replication over two terms in a second school. Ethics approvals were obtained; parental consent required for inclusion in research datasets. Participants: 268 instances of student participation representing 170 unique students (1–4 terms per student). Distribution by term and subject is detailed (Table 1). The primary school partner is high LOI (high external challenges/low SES). A second school with lower LOI (nominally higher SES) participated for replication. Program structure: Each 12-week term followed a capstone-style workflow around a BME global research topic subdivided into subject-specific research questions aligned with Ontario curriculum. Teams of 3–4 students iterated through: background research essay; skill-building lab visit; proposal development; client meeting (feedback from graduate student ‘clients’); two experimental lab visits in university facilities; progress report; data analysis; final research poster and oral presentation at a symposium. Entire class cohorts participated; Discovery contributed 10–15% to the course grade, assessed solely by high school teachers via standardized rubrics. Instruction/mentorship: University of Toronto graduate and undergraduate trainees (primarily BME) served as mentors (~1:8 ratio), trained in pedagogy and safety, and worked consistently with assigned teams; teachers led assessment. Facilities: U of T undergraduate teaching labs and design studios. Data collection: Teachers anonymized and provided per-student component grades (essay, client meeting, proposal, progress report, poster, presentation), cumulative Discovery grade, final course grade (excluding Discovery portion to avoid confounding), and attendance (Discovery sessions and regular classes). Pre/post surveys assessed perceptions of difficulty, comfort with lab environments, and likelihood of pursuing STEM. Surveys underwent PCA validation. Subgroup definitions: Exceeds Expectations (EE): students with Discovery grade ≥1 SD (18.0%) above their final course grade. Multi-Term (MT): students participating in more than one term (76 individuals, 174 instances). Longitudinal analyses for students with 3–4 terms. Statistical analysis: Paired t-tests compared Discovery vs. course grades. Linear regression assessed Discovery vs. course grade correlation and longitudinal trends for MT students. Normality checked via D’Agostino-Pearson; subgroup comparisons via Mann–Whitney U-tests. Significance threshold α ≤ 0.05. Attendance histograms binned (Discovery: 1 day; class: 4-day bins).

- Overall performance: Students scored significantly higher in Discovery than in their course grade excluding Discovery (p < 0.0001). Discovery and course grades were positively correlated (linear regression p < 0.0001), indicating alignment while showing elevated performance in Discovery. - Subgroups identified: EE (N = 99 instances; ≥18.0% higher Discovery vs. course) and MT (174 instances from 76 students). 60.5% of MT students had at least one EE term. - Group formation: Students self-organized into teams with similar overall course performance, reducing concern about high/low performers adversely affecting group outcomes. - Attendance: ≥91% of students attended all Discovery sessions each term. In contrast, 60.8% missed >4 regular classes and 14.6% missed ≥16 classes in a term. - EE outcomes: Compared to non-EE, EE students had lower course grades (p < 0.0001) but higher overall Discovery grades (p = 0.0004). EE students outperformed peers on Discovery deliverables: client meeting (p = 0.0078), proposal (p = 0.0018), progress report (p = 0.036), poster (p = 0.0017); essay difference was not significant (p = 0.097) and final presentation was not significant (p = 0.14). Discovery attendance trended higher (p = 0.074), classroom attendance not different (p = 0.85). - MT outcomes (cross-sectional): MT students had similar course grades to single-term students (p = 0.29) but higher Discovery grades (p = 0.0067). MT performed significantly better on the progress report (p = 0.0021), with trends toward higher proposal (p = 0.081) and presentation (p = 0.056). Discovery attendance similar (p = 0.22); MT missed fewer regular classes (p = 0.010). - MT longitudinal (3–4 terms): No significant course-grade trend over time (p = 0.15) but significant improvement in Discovery grade (p = 0.0011). Deliverables improved across terms: essay (p = 0.0295), client meeting (p = 0.0003), proposal (p = 0.0004), poster (p = 0.0005), presentation (p = 0.0295); progress report showed strong performance with limited room to improve (p = 0.16). - Replication: In a second, lower-LOI school, Discovery vs. course grades diverged significantly from a 1:1 line (p = 0.040), consistent with main findings. - Surveys: 91–100% rated the program as at least moderately challenging each term; 58–86% rated it highly challenging (≥4/5). 94.6% reported increased comfort working in university labs; 72.4% more likely to pursue STEM (10.1% less likely, 17.5% no influence); 84.9% would participate again. - Teacher observations: Students prioritized Discovery deliverables, showed better behavior and focus during Discovery, and struggling students often excelled in the inquiry-focused setting.

Findings indicate that an inclusive, inquiry-based, university-partnered program embedded in regular courses enhances engagement and performance, particularly among students who underperform in traditional, knowledge-focused settings. Higher Discovery grades despite lower course grades in the EE group suggest that hands-on, iterative, problem-based tasks allow these students to demonstrate mastery of skills such as critical thinking, collaboration, and application of concepts. Strong attendance, despite logistical challenges (commuting to campus, early start times), and teacher-reported shifts in behavior underscore engagement. MT students’ repeated participation led to sustained improvements in Discovery-specific competencies rather than course grades, aligning with the program’s emphasis on process skills over rote knowledge. The positive replication in a different SES context suggests scalability. Overall, the model bridges the gap between secondary and post-secondary STEM practices, building relevant skills and positive attitudes that are likely to support persistence in STEM pathways.

The Discovery framework provides a scalable, inclusive model for integrating inquiry-based, BME-contextualized STEM learning into non-specialized high schools. Students achieved significantly higher performance on Discovery deliverables than in their regular coursework, with particularly strong gains among underperforming (EE) and repeat (MT) participants. Engagement indicators—attendance, survey responses, and teacher observations—were robustly positive. The program cultivates critical thinking, problem solving, and collaboration while familiarizing students with post-secondary lab environments. Future research should include longitudinal tracking of academic and career trajectories, expansion across diverse schools and subjects, refinement of assessments to capture process skills, and evaluation of implementation supports for teachers to maximize scalability and equity.

- Generalizability and longitudinal impact not yet established; immediate outcomes may be influenced by cohort effects and classroom factors. Longer-term tracking of participants is needed to assess persistence in STEM. - Replication to date includes only two schools; larger, diverse samples across SES contexts are required to isolate socioeconomic influences. - Implementation may be constrained by resources, institutional support, access to post-secondary partners, and subject-specific logistics; scaling may require adaptations (e.g., focus on subjects with simpler lab logistics). - Attendance comparisons between Discovery and regular classes are approximate due to differing session frequencies and durations. - No a priori power analysis was conducted; participation in program was mandatory (though inclusion in research required consent), which may introduce selection or motivational biases. - Data privacy restrictions limit public availability of individual-level data.