Adaptive Virtual Labs for Technical Education in Resource-Limited Settings

Thesis Abstract

In resource-limited technical education contexts, access to fully equipped laboratories remains a critical bottleneck, constraining practical skill development and adoption of industry-relevant competencies. This study investigates the design, deployment, and evaluation of adaptive virtual labs (AVLs) as a technology-driven intervention to bridge gaps in hands-on experience, laboratory throughput, and learning equity. The aim is to determine how AVL systems, informed by real-time student data and adaptive learning algorithms, influence practical proficiency, self-regulated learning, and motivation among technical education students. Specific objectives include (1) developing an AVL framework that adjusts difficulty, feedback, and available instrumentation to individual learner profiles; (2) assessing the impact of AVLs on practical assessment scores across core disciplines (electrical, mechanical, and computer engineering technology) in resource-limited institutions; (3) examining changes in time-on-task, error rates, and learner engagement using biometric and interaction-log data; (4) validating the theoretical grounding of AVLs in the Activity Theory and Self-Determination Theory (SDT) to explain enhanced autonomy and competence; and (5) proposing an implementation blueprint for scalable adoption in similar settings. A mixed-methods research design was employed. The quantitative strand used a quasi-experimental, pretest–posttest control-group design involving 240 diploma and bachelor's-level students across four technical colleges, with 120 in the AVL intervention and 120 in traditional lab settings. Practical assessments, timed lab tasks, and standardized rubrics were administered at baseline, mid-term, and end of semester. Data sources included system-generated interaction logs (time-on-task, sequence of operations, error frequency), assessment scores, and retention metrics. Advanced analytics comprised multilevel linear modeling to account for nested data (students within courses), ANCOVA to control pre-existing differences, and regression analyses to identify predictors of performance gains. The qualitative strand consisted of semi-structured interviews with 28 instructors and 36 students, focus group discussions with 6 laboratory supervisors, and thematic analysis to extract perceived barriers, facilitators, and experiential nuances of AVL use. The theoretical framing integrated Activity Theory to interpret tool-mediated learning processes and SDT to elucidate motivational dynamics. The AVLs were developed with modular virtual instrumentation, scenario-based experiments, and real-time adaptive feedback that modulates instrument fidelity, task complexity, and scaffolding according to learner mastery indicators. The adaptive engine leveraged reinforcement learning and Bayesian updating to personalize task sequences, while ensuring alignment with national technical education competencies. Validity and reliability of instruments were established through pilot testing, expert reviews, and Cronbach’s alpha checks (alpha > 0.78 for all scales). Data analysis revealed statistically significant improvements in practical performance for the AVL group (mean post-test score increase of 18.6 percentage points, p < 0.001) compared with the control group (11.2 points, p < 0.01), with a medium effect size (d = 0.55). Multilevel models indicated stronger gains among first-year students and those with weaker prior laboratory exposure, suggesting AVL effectiveness in leveling prior disparities. Interaction logs showed higher engagement rates and reduced task completion times in the AVL cohort (p < 0.05). Qualitative findings highlighted enhanced perceived autonomy, immediate feedback, and perceived realism of simulations as key drivers of motivation, while concerns included initial setup demands and digital literacy requirements. The study contributes to knowledge by providing empirical evidence of adaptive virtual laboratories as a scalable, equity-oriented solution to laboratory access constraints in resource-limited settings, extending the theoretical integration of Activity Theory and SDT within practical engineering education contexts. It offers a rigorously tested implementation model, including technical architecture, curriculum alignment, instructional strategies, and professional development needs for instructors. Practical implications include guidelines for cost-effective deployment, data governance, and sustainable maintenance in constrained environments. Recommendations emphasize phased scaling to diverse institutions, continuous refinement of adaptive algorithms to local contexts, and policy support for investing in high-speed connectivity and device provisioning to maximize AVL impact. The study concludes that adaptive virtual labs can significantly enhance practical competencies, learner motivation, and instructional efficiency, recommending broader adoption and longitudinal follow-up studies to assess long-term retention and industrial workflow transfer.