Digital Competence in Generative AI Contexts Through a Second-Level Digital Divide Lens
International Conference on AI Research · 18.11.2026 · 7 (1) · 454-461
DOI: 10.34190/icair.7.1.5650 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Conference on AI Research, online erschienen am 18.11.2026. Verfasst von Yinlei Wang. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).
Abstract (Originalsprache)
As generative artificial intelligence (GenAI) becomes part of university study, students may differ in how they evaluate its outputs, incorporate generated content, manage personal data and direct their interactions with AI tools. Drawing on the second-level digital divide, this study considers differences in the judgements students bring to GenAI use. It examines whether students’ general perceptions of higher-order digital competence are associated with their responses to related GenAI-specific indicators, and explores how students describe these practices. Using an explanatory sequential mixed-methods design, questionnaire responses from 95 university students in China were analysed alongside follow-up interviews with three respondents. Four competences from DigComp 2.2 guided the study: information evaluation, content integration, protection of personal data and privacy, and creative use of digital technologies. The questionnaire included four general self-reported proficiency indicators and eight GenAI-contextualised self-report items. Eight theoretically matched associations were examined using Spearman correlations, with confidence intervals estimated by bootstrapping and p-values adjusted for multiple testing. Interview accounts were analysed thematically and considered alongside the quantitative findings. Most matched associations were small and uncertain; seven of the eight correlation confidence intervals included zero. The only association that remained significant after adjustment linked general privacy competence with awareness of personal-data risks in AI systems (ρ = .284, adjusted p = .043). The same general indicator showed little association with knowledge of data-protection regulation. Interviews suggested that verification depended on prior knowledge, perceived uncertainty and task demands; participants also differed in how actively they guided AI interactions. Awareness of privacy risks did not consistently translate into understanding or protective action. These findings suggest that general competence self-reports and GenAI-contextualised indicators should be examined separately, even when they address the same DigComp competence. Differences in item specificity and response format may also explain the limited associations. The study therefore does not establish whether competence transfers across contexts, but identifies decisions and conditions that merit closer attention when assessing students’ GenAI use.
Zitieren
APA
Wang, Y. (2026). Digital Competence in Generative AI Contexts Through a Second-Level Digital Divide Lens. International Conference on AI Research, 7(1), 454-461. https://doi.org/10.34190/icair.7.1.5650
Vancouver
Wang Y. Digital Competence in Generative AI Contexts Through a Second-Level Digital Divide Lens. International Conference on AI Research. 2026;7(1):454-461. doi:10.34190/icair.7.1.5650
BibTeX
@article{Wang2026,
author = {Wang, Yinlei},
title = {Digital Competence in Generative AI Contexts Through a Second-Level Digital Divide Lens},
journal = {International Conference on AI Research},
year = {2026},
volume = {7},
number = {1},
pages = {454--461},
doi = {10.34190/icair.7.1.5650}
}
EUVIMED