VERIFICATION OF GENERATIVE AI OUTPUTS BY PRIMARY SCHOOL TEACHERS: RELATIONSHIPS BETWEEN ERROR EXPERIENCE, TRUST, AND REPORTED PRACTICES

HONCHAROVA, Olena (2026) VERIFICATION OF GENERATIVE AI OUTPUTS BY PRIMARY SCHOOL TEACHERS: RELATIONSHIPS BETWEEN ERROR EXPERIENCE, TRUST, AND REPORTED PRACTICES. In: INTERNATIONAL CONFERENCE “FROM RESEARCH TO APPLICATION”, 18-19 November, 2026.

[img] Text
Proceedings Book 2026-195-203.pdf
Download (570kB)

Abstract

This study examined how primary school teachers’ experience of working with generative artificial intelligence (GenAI) outputs, their reported verification practices, trust in AI-generated information, and their final prioritisation of professional risks and opportunities were related within the same survey context. A cross- sectional survey was conducted among primary school teachers, with analyses focused on respondents who reported using AI tools in their professional practice. Quantitative data were analysed using descriptive statistics and correlation analysis, while responses to two final open-ended questions were examined through inductive content coding. The findings showed that more frequent experience of detecting errors in AI-generated outputs was related to more frequent verification and to lower reported use of AI-generated materials without in-depth checking. Trust in AI-generated information was more strongly related to the reported use of such materials without thorough verification. Perceived efficiency was not related to verification frequency, although it showed a weaker relationship with the use of AI-generated materials without in-depth verification. In the final open-ended responses, teachers most frequently prioritised time saving and efficiency as professional opportunities, while reduced critical or independent thinking and unreliable information were the most frequently identified risks. The results showed that error experience, trust, verification practices, and final open-ended prioritisation represented related but non-interchangeable aspects of teachers’ self-reported professional experience with GenAI.

Item Type: Conference or Workshop Item (Paper)
Subjects: L Education > L Education (General)
Divisions: Faculty of Law, Arts and Social Sciences > School of Education
Depositing User: Unnamed user with email zshi@unite.edu.mk
Date Deposited: 10 Nov 2026 13:31
Last Modified: 10 Nov 2026 13:31
URI: http://eprints.unite.edu.mk/id/eprint/2712

Actions (login required)

View Item View Item
Universiteti i Tetoves is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.