<![CDATA[This study aims to evaluate the consistency and fairness of undergraduate thesis assessments at IAI Nurul Hakim Kediri, West Lombok, using Generalizability Theory (G-Theory). Data were collected from the assessments of 13 students by three examiners using four standardized evaluation criteria. Analysis using the EduG software revealed that the largest source of error stemmed from the interaction among persons, raters, and criteria (PRQ). The generalizability coefficient of 0.93 indicates excellent reliability of the assessment instrument. To enhance fairness, efforts should be made to reduce variability in facet interactions. Simulations showed that increasing the number of raters and criteria could improve reliability. While Option 5 yielded the most optimal results, Options 3 and 4 are also worth considering as they offer a balance between reliability and resource efficiency. This study underscores the importance of well-designed assessment systems to ensure fair and consistent undergraduate thesis evaluations.]]>
Copyrights © 2025
