Adaptive micro-credentialing via AI-driven learning analytics for technical education

Thesis Abstract

Adaptive micro-credentialing via AI-driven learning analytics for technical education addresses the growing demand for personalized, portable, and verifiable skills in rapidly evolving technical sectors. The study responds to the persistent gap between formal credentialing and actual practitioner competencies, where traditional degrees fail to capture discrete micro-skills aligned with industry needs. The aim is to design, implement, and evaluate an AI-driven adaptive micro-credentialing system that continuously assesses learner performance, personalizes learning pathways, and issues verifiable badges contingent on mastery demonstrated across technical domains. Specific objectives include (1) identifying key skill tokens and competency frameworks relevant to welding, robotics maintenance, and IoT systems installed in manufacturing environments; (2) developing a learning analytics model that integrates formative assessments, project artifacts, and in-plant performance data to generate personalized micro-credentials; (3) evaluating the validity, reliability, and transferability of issued credentials across industry partners; (4) examining the impact of adaptive micro-credentialing on learner engagement, completion rates, and time-to-competency; and (5) proposing a governance and ethical framework for credential issuance, data privacy, and accountability. The methodology adopts a mixed-methods design anchored in the constructivist epistemology and informed by the social cognitive theory of self-regulated learning and the capability approach. The population comprises technical education learners enrolled in advanced manufacturing and electrical/electronics programs across three technical colleges in a metropolitan region. A stratified random sample of 320 students will be drawn, with 240 participants assigned to the intervention group using the AI-driven adaptive micro-credentialing platform and 80 to a control group receiving standard competency tracking for comparison. Data collection instruments include a) platform-generated analytics dashboards capturing engagement metrics, mastery checkpoints, and credential issuance records; b) standardized practical assessments aligned with ISO/IEC 17024-inspired micro-credentials; c) validated self-report instruments measuring motivation and perceived usefulness of micro-credentials; and d) semi-structured interviews with learners, instructors, and industry partners to triangulate quantitative findings. Validity and reliability of instruments will be established through pilot testing, content validity indices, Cronbach’s alpha checks, and inter-rater reliability for performance assessments. Analytical procedures comprise quantitative and qualitative techniques. Quantitative analyses will use descriptive statistics to profile baseline characteristics, multilevel regression to assess factors predicting mastery and time-to-credential, and repeated-measures ANOVA to evaluate changes in engagement over time. Structural equation modeling (SEM) will test the hypothesized pathways from platform engagement to mastery, to credential issuance, and to job-fit indicators collected through employer surveys. For the qualitative strand, thematic analysis will be conducted on interview transcripts, using a constant comparative method to identify emergent themes related to trust in AI-driven credentials, perceived portability, and alignment with industry standards. A convergent parallel design will integrate findings to produce a holistic interpretation of how AI-driven learning analytics influence adaptive credentialing efficacy. Expected findings include (i) demonstrable improvements in time-to-competency and in-platform mastery rates for the intervention group compared with controls; (ii) strong predictive validity of composite mastery scores for job-readiness as evidenced by subsequent internship performance and employer feedback; (iii) higher learner engagement and satisfaction with adaptive pathways; and (iv) robust evidence for the portability and acceptability of AI-generated micro-credentials across partner organizations. The study contributes to knowledge by operationalizing a scalable, data-driven model for adaptive credentialing that integrates learning analytics with formal and informal skill verification, expanding the theoretical discourse on credential economics, self-regulated learning in technical education, and the legitimacy of AI-augmented assessments. Practical implications include guiding policy on credential standards, informing platform design for interoperability with existing LMS and HR systems, and providing a governance blueprint addressing data privacy, consent, and accountability. The main conclusion is that AI-driven adaptive micro-credentialing can significantly enhance targeted skill development and credential credibility in technical education, provided that rigorous validation, transparent governance, and close alignment with industry competency frameworks are maintained. Recommendations include expanding industry partnerships to diversify credentialing domains, exploring long-term outcomes such as career progression and wage impacts, and developing open interoperability standards to facilitate cross-institutional credential recognition.