Abstract
In high-stakes professional fields like maritime education, institutional summative assessments serve as critical gateways for validating student competency prior to graduation. This study executed a convergent parallel mixed-methods evaluation of the 100-item Comprehensive Examinations administered to the 2024–2025 graduating batch for the Bachelor of Science in Marine Transportation (BSMT) and Bachelor of Science in Marine Engineering (BSMarE) programs at the University of Cagayan Valley. By integrating quantitative psychometric metrics from Classical Test Theory (CTT) with qualitative document analysis, the research analyzed an archived psychometric dataset representing 712 examinees (425 BSMT and 287 BSMarE students) alongside a content audit of item structures. The integrated analysis revealed four primary findings. First, a severe psychometric range compression exists across both programs, with 97% of BSMT items and 96% of BSMarE items clustered exclusively within the "Easy" and "Average" difficulty bands, leaving the "Very Difficult" thresholds completely unrepresented (0%). Second, critical item-level structural design flaws were isolated, including instances of negligible discrimination (e.g., BSMT Item 9, D = 0.10) and severe inverted discrimination (e.g., BSMT Item 30, D = -0.08) where lower-performing student cohorts paradoxically outscored higher-performing peers due to ambiguous stems or keying errors. Third, a distinct programmatic asymmetry emerged, revealing that the marine engineering pathway skews significantly easier (43% Easy or Very Easy) than the marine transportation track, which clusters predominantly in the middle tier (65% Average). The study concludes that a pronounced curriculum-assessment alignment gap exists, as the current examinations over-index on lower-order cognitive recall rather than validating the complex, high-order diagnostic troubleshooting demanded by international Standards of Training, Certification, and Watchkeeping (STCW) mandates. To satisfy the quality requirements of United Nations Sustainable Development Goal 4 (Quality Education) and the academic innovation benchmarks established by the World University Rankings for Innovation (WURI), the College of Maritime Education must transition from passive data archiving to active quality engineering by rewriting defective items, balancing difficulty distributions with scenario-based case studies, and creating a unified institutional item-banking policy.
