On using fuzzy logic for grading highly subjective assessment material — A case study

It is difficult to remove subjectivity from assessment, especially when the assessed material is highly subjective by nature. While one assessor deems work to be of high quality, a second may view it as poor. Moreover, although two assessors may agree that the work is of satisfactory quality, they may nevertheless proceed to grade it significantly different, based on their interpretation of what is meant by satisfactory. While this is part of human nature, it is important that we strive for an assessment mechanism that is fair to all students, particularly when multiple personnel are involved in the grading process. This paper investigates the suitability of using fuzzy logic, and its ability to model linguistic terms, for the purposes of achieving fair assessment. A case study is carried out in which both the fuzzy logic and the standard numerical grading approaches are applied to a postgraduate poster competition. The results from this case study are presented and discussed within.