Development and Validation of the Atma AI Ethics in Academic Writing Scale (AAEWS)
Keywords:
AI ethics, perception of AI ethics, scientific writing, student assignments, written workAbstract
The integration of artificial intelligence (AI) in education presents new opportunities and ethical challenges, especially concerning academic integrity in student papers. Misuse of AI can result in issues such as plagiarism, data privacy violations, and theoretical reference hallucinations. To address this, a standardized measurement tool is needed to evaluate students' ethical perceptions on using AI for academic writing. This study aims to develop and test the psychometric properties of a measurement tool to assess students' ethical perceptions of AI in writing scientific papers. Ethical Perception in this test tool is defined as awareness and commitment to the responsible use of AI in research and scientific writing, which includes student’s responsibility for the accuracy and integrity of results, substantial author contributions, and transparency and recognition of AI use with five dimensions: reporting standards, data access and retention, originality and acknowledgment of sources, citations, then the use of generative AI and AI-assisted technologies in scientific writing and in figures, images, and artwork. With convenience sampling, the sample size in this study for the item analysis was 34 participants and for validity and reliability testing was 120 participants. The initial number of items was developed from 60 to 42. This measurement tool has an omega reliability of 0.918, and its five dimensions have alpha reliability of 0.335-0.742. For validation, a comparative measurement tool from Meiriza et al.'s (2024) research served as external validity, yielding r(118) = 0.288, p < 0.001. In conclusion, the results of this psychometric testing indicate that the AAEWS is valid according to external validity, correlation with the other test that had similar construct and reliable. Therefore, based on external validity and reliability, the AAEWS can confidently measure students' ethical perceptions of the use of AI in scientific writing.
