Fuzzy learning performance assessment based on decision making under internal uncertainty

The paper delves into decision making with partial knowledge focused on students´ behaviour in multiple-choice testing. To address to this problem, we first provided a binary knowledge model and, then, relaxing some assumptions, allowed for the more realistic framework of partial knowledge which, in turn, adds uncertainty to the assessment of students´ knowledge due to their incentive to make a guess in multiple-choice testing. The aim of this paper is to propose a fuzzy assessment model formalised according to Reiter´s Theory of Diagnosis to reduce this uncertainty and to draw a distinction between the level of students´ ability and the degree of guessing. The provided assessment model was tested on modelled answers with respect to the knowledge frameworks and validated in a real-world context. The findings of this research present the fuzzy learning performance assessment model which may enable a teacher to estimate the level of partial knowledge and, thus, to specify a student´s score as well as the results of computational experiments that confirm the validity of the theoretical outcomes.