Cognitively-oriented prompting in higher education: designing prompts for large language models based on Bloom’s Taxonomy

The aim of the study is to develop and substantiate a typology of prompts oriented towards Bloom’s Taxonomy levels to enhance the cognitive effectiveness of interaction between participants in higher education and large language models within the educational environment. The article analyzes the structure of learning prompts formulated by faculty and students of the Pedagogical Institute of the Mari State University, identifying linguistic markers and cognitive patterns corresponding to the six levels of the taxonomy – from remembering to creating. Special attention is given to the connection between prompt formulation and the quality of the AI-generated response, as well as to pedagogical strategies that facilitate the transition from reproductive information consumption to the development of higher-order thinking. The scientific novelty of the research lies in the fact that it is the first one to propose a systematized typology of prompts correlated with Bloom’s Taxonomy levels and adapted to the context of AI integration in learning. As a result of an empirical analysis of 327 prompts, it was found that the overwhelming majority of requests (68.2%) focused on lower levels of cognitive complexity, while prompts aimed at analysis, evaluation, and creation accounted for 19.5%. It is demonstrated that the conscious design of prompts, considering Bloom’s Taxonomy, significantly enhances the depth, structure, and critical content of AI responses, transforming AI from an information source into an intellectual partner in the educational process.