Mapping human–AI communication in a Russian higher education context: a psychometric network analysis of AI literacy and attitudinal dispositions
TL;DR
The findings support the reconceptualization of AI literacy as a communicative-cultural construct in the Russian cultural-communicative context and extend the human-machine communication framework to Russian higher education, a setting underrepresented in AI literacy research.
Abstract
As generative and chat-based artificial intelligence (AI) systems evolve into communication partners, students' attitudes toward these machine interlocutors have become a central question in communication research. These attitudes are not culturally neutral; they are shaped by the media ecology, linguistic repertoire, and social norms of the communication environment. Drawing on a humanmachine communication framework, this study reconceptualizes AI literacy as communicative competence with machine interlocutors and attitudes toward AI as culturally mediated evaluative orientations. The aim was to map their conditional architecture in the Russian cultural-communicative context. In a cross-sectional design, 668 undergraduate students from three Russian universities completed the Russian-adapted Meta Artificial Intelligence Literacy Scale and the General Attitudes towards Artificial Intelligence Scale. A standardized partial correlation network was estimated. Centrality, bridge centrality, network invariance, and community structure were examined. The strongest bridge was found between practical interaction with machines (Apply AI) and positive attitude. Exploratory network analysis placed Apply AI within the attitude cluster. Critical evaluation was located at the structural center of the network. Negative attitude remained peripheral. Persuasion and emotion regulation competencies merged into a single ESEM dimension. The network remained invariant across gender and frequency of use. This suggests that machine interaction and its evaluation form an integrated communicative-relational domain. A linguistic-cultural reading of this pattern is offered as a post hoc interpretation. The findings extend the human-machine communication framework to Russian higher education, a setting underrepresented in AI literacy research. They also support the reconceptualization of AI literacy as a communicative-cultural construct.