Research Area:  Machine Learning

Abstract:

The advancement of artificial intelligence (AI) and the expanding digitalisation of societies more broadly are impacting different sociocultural, geopolitical, and economic domains (Mohamed et al., 2020; Wang, 2019). A transformative development in this field is the shift from narrow-task AI, which focuses on language pattern detection and predictive analytics, to Generative AI (GenAI). GenAI employs advanced neural networks and massive datasets to produce multimodal outputs, encompassing text, images, videos, and complex simulations (Cooper, 2023). Large language models (LLMs), a subset of GenAI, have seen substantial global adoption. People worldwide use GenAI for creating artefacts, learning, solving problems, and conducting research. Educational institutions reflect societal trends, including the adoption and use of digital technologies. One estimate suggests that 1% of all journal articles published in 2023 contained AI-generated text (Gray, 2024). In light of these developments, there is a need to critically examine the role of GenAI in academic research and publication, ensuring that its integration can enhance, rather than undermine, the integrity and quality of science education research. As editors for Research in Science Education (RISE), we feel a responsibility to actively contribute to the ongoing discussions about using transformative technologies in science education research, such as GenAI. Our aim here is to contribute to discussions that others have initiated, including for instance, the Journal of Research in Science Teaching (JRST, Sadler et al., 2024). The Sadler et al. editorial broadly highlights guidelines about the use of GenAI when submitting to JRST. It outlines considerations such as transparency, trustworthiness, and biases inherent in GenAI platforms. While these guidelines are insightful and a valuable catalyst for further discussion, we extend the conversation in several additional areas, including ethics, philosophy, copyright, authorship attribution and the inclusion of images. We also anticipate that more manuscripts will report the use of GenAI in science education, including those that will be curated in our “Artificial Intelligence in Science Education” Topical Collection. As such, this editorial serves to provide informed dialogue among researchers who intend to submit manuscripts involving the use of GenAI to RISE. In this editorial, we outline our positions on two interconnected issues. First, we examine the key philosophical principles that inform our views of using GenAI in educational research. We also discuss several ethical and legal issues to set the context for potential applications, benefits and risks of using GenAI. Guided by these principles and issues, we then outline the guidelines for submitting to RISE, highlighting key expectations, changes to the submission portal, and examples of good practice. By contributing to these emerging discussions, we aim to enrich the dialogue, share our perspectives on using GenAI in research, and more effectively guide authors using GenAI when submitting manuscripts. In this discussion, we seek to balance the opportunities presented by the emergence of the technology while also acknowledging the potential risks and dangers associated with this technology.

Keywords:  

Author(s) Name:  Kok-Sing Tang, Grant Cooper & Wendy Nielsen

Publisher name:  Springer

DOI:  10.1007/s11165-024-10192-3

Volume Information:  Volume 54, Pages 797-807, (2024)

Paper Link:   https://link.springer.com/article/10.1007/s11165-024-10192-3