Adaptive Mobile Microlearning for Adult STEM Learners in Low-Resource Settings

Thesis Abstract

Adaptive mobile microlearning offers a scalable ICT-driven approach to enhance STEM learning outcomes for adults in low-resource settings. The study addresses the persistent gap in access to high-quality, just-in-time STEM education for adult learners who face constraints such as limited internet bandwidth, low digital literacy, and irregular study schedules. The aim is to design, implement, and evaluate an adaptive mobile microlearning intervention that personalizes content delivery based on learner profiles, prior knowledge, and context, thereby improving knowledge acquisition, retention, and transferring of STEM skills to practice. Specific objectives are (1) to develop an adaptive microlearning framework integrating spaced repetition, formative assessments, and context-aware content delivery; (2) to implement a mobile learning prototype using offline-capable modules with lightweight multimedia suitable for low-resource environments; (3) to evaluate the impact of adaptation algorithms on learning gains in quantitative terms and on learner engagement qualitatively; (4) to examine user perceptions of usefulness, ease of use, and perceived autonomy guided by the Technology Acceptance Model and Self-Determination Theory; (5) to identify facilitators and barriers to adoption in local communities and inform scalable deployment strategies. A pragmatic mixed-methods research design is employed, combining quasi-experimental comparison with a longitudinal case study. The population consists of 420 adult STEM learners enrolled in vocational and higher-education programs across four district centers in a low-resource region. A stratified random sample of 210 participants is assigned to the adaptive microlearning intervention, while 210 form a control group receiving standard non-adaptive mobile content over a 12-week period. Data collection instruments include a validated STEM knowledge test administered at pretest, immediate posttest, and a delayed posttest at eight weeks, a 5-item System Usability Scale (SUS), and a 12-item Likert-scale survey based on the Unified Theory of Acceptance and Use of Technology (UTAUT) to gauge adoption factors. Qualitative data comprise semi-structured interviews with 40 participants (10 from each site) and 12 focus groups with instructors and program coordinators to capture contextual influences and implementation experiences. Data analysis employs multiple statistical techniques descriptive statistics and normality checks, ANCOVA to compare posttest gains controlling for pretest scores, and hierarchical linear modeling to account for nested data (students within centers). For qualitative data, thematic analysis is conducted using coding frames informed by Activity Theory and Self-Determination Theory, with triangulation across interviews and focus groups to corroborate findings. The analytical framework also integrates a regression discontinuity design component to explore threshold effects of prior knowledge on the efficacy of adaptive content. Expected findings include statistically significant improvements in learning gains for the adaptive group relative to controls, with effect sizes in the small-to-moderate range (Cohen’s d ? 0.35–0.60). Enhanced retention and transfer of skills are anticipated, supported by higher posttest and delayed posttest scores. Qualitative insights are expected to reveal that offline-capable, bite-sized modules with adaptive pacing increase perceived usefulness and autonomy, while challenges such as intermittent power supply and device heterogeneity influence usability. The study contributes to knowledge by empirically validating an adaptive microlearning architecture grounded in Self-Directed Learning and Cognitive Load Theory, operationalized through a mobile platform suitable for low-resource contexts. It advances understanding of how personalization, pacing, and offline functionality interact to improve STEM learning outcomes for adult learners who balance work, family, and education commitments. The findings will inform scalable deployment strategies for regional education authorities and vocational training providers, including guidelines for content modularization, data-sparing analytics, and alignment with national curricula. The conclusion highlights that adaptive mobile microlearning can bridge access gaps and elevate STEM proficiency among adults in resource-constrained settings, provided there is investment in device equity, capacity-building for instructors, and ongoing evaluation of augmentation strategies. Recommendations include expanding offline content libraries, refining personalization algorithms with local contextual variables, integrating micro-credentials for workforce relevance, and conducting longitudinal studies to assess long-term impact on employability and lifelong learning trajectories.