FEATURES OF SOLVING PROBLEMS FROM THE COURSES OF BUSINESS MATHEMATICS AND STATISTICS IN ENGLISH USING GENERATIVE ARTIFICIAL INTELLIGENCE

Abstract

This publication explores the features of using generative artificial intelligence (GAI) to solve problems from business mathematics and statistics courses taught in English. Given the widespread integration of GAI into the educational process, this topic is highly relevant for both students and educators.

The study focused on some of the most widely used GAI tools today: ChatGPT, Gemini, Copilot, and Grok. The problems, provided in English, had known methods and solutions in advance, ensuring objective analysis. To enhance the validity of the research, the tasks were categorized by complexity: simple and complex.

The study found that:

Business mathematics and statistics problems are correctly interpreted by GAI tools when presented in English with a partially formalized description. Simple tasks (such as calculations, percentage problems, and data series processing) are successfully solved by all tested GAI tools. However, the level of explanation detail varies; for example, Copilot provides more concise answers without additional clarification. Complex problems—such as transportation tasks or those requiring the use of combined statistical methods—pose significant challenges for GAI. The results may be partially correct or entirely incorrect. Especially problematic are tasks with many variables or modified conditions (e.g., mismatches between supply and demand in transportation problems).

The likelihood of using GAI as a basis for plagiarism in solving complex business mathematics and statistics problems in English remains low.

Based on the findings, the following recommendations are made: Avoid including problems in individual student assignments that are easily solvable by GAI tools. Solve such problems in classrooms, using individualized data sets for each student. Actively use GAI tools to create educational materials, particularly variable problem sets. Implement assignments focused on explaining the logic of the solution, rather than merely obtaining the final answer.