ESPRIT adventure: Assessing hybrid fuzzy-crisp rule-based AI method effectiveness in teaching key performance indicators

Abstract

Key performance indicators (KPIs) are a fundamental tool for understanding the outcomes of business policies. Teaching these indicators is conventionally accomplished through verbal lecturing using mathematical equations, accompanied by presentation graphics and tabular information representing the company's operational, financial, and strategic achievements. The conventional teaching method enables the student to understand what a KPI means and how it is calculated, however, it does not enable the student to understand easily how to interpret it. Here, a hybrid fuzzy-crisp rule-based AI system is presented and its effectiveness when applied with a group of management and organization students at the University of Belgrade, Faculty of Organizational Sciences, during the 2022/2023 academic year is evaluated. The potential of this tool is reflected in its ability to simulate adaptive and realistic business economic situations, its high interactivity and user-friendly interface, and intuitive graphical and tabular results that are easy to interpret enhanced with natural-language-inference-explanations. Univariate analysis along with some complementary statistical tests was used to examine students' satisfaction with the system and to distinguish between students who used the system, and students who attended conventional classes, both in terms of acquiring knowledge and retaining knowledge. In general, the results showed high students' satisfaction with the system. The scores obtained by the former group on the objective knowledge assessment test were on average significantly higher compared to those of the latter, proving the positive effect of the hybrid fuzzy-crisp rule-based AI teaching approach.