Date
Publisher
arXiv
The purpose of this study was to examine, from the perspective of secondary education students, the educational affordances and constraints of using Artificial Intelligence (AI) in teaching and learning. The sample consisted of 45 students from the 2nd year of General Lyceum (11th grade, ages 16-17) in Greece, who, after becoming familiarized with ChatGPT-4o and completing six activities, filled in an open-ended questionnaire related to the research purpose. Open, axial, and selective coding of the data revealed that students recognize five educational affordances: the creation of new knowledge building on prior knowledge, immediate feedback, friendly interaction through messaging, ease and speed of access to information, and skills development. Concurrently, three main constraints were identified: content reliability, anxiety about AI use, and privacy concerns. The study concludes that students are positive toward AI use in education.
Study at a glance
- Question: What are the educational affordances and constraints of using ChatGPT-4o in secondary school teaching and learning?
- Setting and data: Attica, Greece; three 11th-grade General Lyceum classes; May 2024; 45 students aged 16–17.
- Method: Thematic analysis (open, axial, and selective coding) of open-ended questionnaires following classroom activities.
Why it matters
Most research on ChatGPT focuses on version 3.5 and higher education, leaving the educational affordances and constraints of GPT-4o in secondary education largely unexamined.
Key findings
- Most Greek 11th-graders recognized learning benefits, highlighting the creation of new knowledge building on prior knowledge (33 students), immediate feedback (28), and a familiar messaging interface (25).
- Students identified key constraints to using ChatGPT-4o, led by uncertainty regarding content truthfulness and quality (16 students), followed by anxiety regarding use (8) and privacy concerns (5).
- While 42 students initially believed ChatGPT-4o never made errors, their post-intervention experience made them considerably more concerned about whether to trust its answers.
Summary drafted with AI from the full paper and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
