Minimum Passing Standard Setting based on Data-Driven Decision Making for English Online Exam using Fuzzy Algorithm

Abstract

This study aims to establish a minimum standard for student performance in Basic English Language Learning (BELL) through a pre-assessment, the Integrated Standard English Exam (i-SEE). The i-SEE evaluates five key areas: listening, reading, structure and grammar, social skills, and picture/audio comprehension, each consisting of 10 questions, for a total of 50 questions. The test results are presented in tabular form, detailing the performance in each assessed aspect. The recapitulated results serve as the basis for developing a personalized Learning Plan (LP), guiding both tutors and students to focus on areas where students show weaknesses. This ensures a more targeted and comprehensive learning experience. The research employs the Fuzzy Tsukamoto algorithm to determine the minimum standard student scores. The study follows a Data Driven Decision Making approach, leveraging initial data to inform the analysis. The results indicate that the minimum standard for each material is 12.46 out of 20, while the overall minimum standard is 57.36 out of 100. The accuracy of the fuzzy calculation was evaluated using the Mean Absolute Percentage Error (MAPE), yielding 92.08% for category-based calculations and 99.49% for overall calculations. These findings highlight the efficacy of fuzzy logic in setting minimum performance standards and enhancing the precision of educational assessments.