Search

The adoption of large language models (LLMs) in education holds much promise. However, like many technological innovations before them, adoption and access can often be inequitable from the outset, creating more divides than they bridge. In this paper, we explore the magnitude of the country and language divide in the leading open-source and proprietary LLMs with respect to knowledge of K-12 taxonomies in a variety of countries and their performance on tagging problem content with the appropriate skill from a taxonomy, an important task for aligning open educational resources and tutoring content with state curricula. We also experiment with approaches to narrowing the performance divide by enhancing LLM skill tagging performance across four countries (the USA, Ireland, South Korea and India-Maharashtra) for more equitable outcomes. We observe considerable performance disparities not only with non-English languages but with English and non-US taxonomies. Our findings demonstrate that fine-tuning GPT-3.5 with a few labelled examples can improve its proficiency in tagging problems with relevant skills or standards, even for countries and languages that are underrepresented during training. Furthermore, the fine-tuning results show the potential viability of GPT as a multilingual skill classifier. Using both an open-source model, Llama2-13B, and a closed-source model, GPT-3.5, we also observe large disparities in tagging performance between the two and find that fine-tuning and skill information in the prompt improve both, but the closed-source model improves to a much greater extent. Our study contributes to the first empirical results on mitigating disparities across countries and languages with LLMs in an educational context.

Cheating behaviors have been construed as a continuing and somewhat vexing issue for academic institutions as they increasingly conduct educational processes online and impose metrics on instructional evaluation. Research, development, and implementation initiatives on cheating detection have gained new dimensions in the advent of artificial intelligence (AI) applications; they have also engendered special challenges in terms of their social, ethical, and cultural implications. An assortment of commercial cheating-detection systems have been injected into educational contexts with little input on the part of relevant stakeholders. This paper expands several specific cases of how systems for the detection of cheating have recently been implemented in higher education institutions in the US and UK. It investigates how such vehicles as wearable technologies, eye scanning, and keystroke capturing are being used to collect the data used for anti-cheating initiatives, often involving systems that have not gone through rigorous testing and evaluation for their validity and potential educational impacts. The paper discusses accountability- and policy-related issues concerning the outsourcing of cheating detection in institutional settings in the light of these emerging technological practices as well as student resistance against the systems involved. The cheating-detection practices can place students in a disempowered, asymmetrical position that is often at substantial variance with their cultural backgrounds.

Purpose: This purpose of this paper is to report on the development of an evidence-informed framework created to facilitate the formulation of generative artificial intelligence (GenAI) academic integrity policy responses for English medium instruction (EMI) higher education, responding to both the bespoke challenges for the sector and longstanding calls to define and disseminate quality implementation good practice. Design/methodology/approach: A virtual nominal group technique engaged experts (n = 14) in idea generation, refinement and consensus building across asynchronous and synchronous stages. The resulting qualitative and quantitative data were analysed using thematic analysis and descriptive statistics, respectively. Findings: The GenAI Academic Integrity Policy Development Blueprint for EMI Tertiary Education is not a definitive mandate but represents a roadmap of inquiry for reflective deliberation as institutions chart their own courses in this complex terrain. Research limitations/implications: If repeated with varying expert panellists, findings may vary to a certain extent; thus, further research with a wider range of stakeholders may be necessary for additional validation. Practical implications: While grounded within the theoretical underpinnings of the field, the tool holds practical utility for stakeholders to develop bespoke policies and critically re-examine existing frameworks. Social implications: As texts produced by students using English as an additional language are at risk of being wrongly accused of GenAI-assisted plagiarism, owing to the limited efficacy of text classifiers such as Turnitin, the policy recommendations encapsulated in the blueprint aim to reduce potential bias and unfair treatment of students. Originality/value: The novel blueprint represents a step towards bridging concerning gaps in policy responses worldwide and aims to spark discussion and further much-needed scholarly exploration to this end.

While ChatGPT has recently become very popular, AI has a long history and philosophy. This paper intends to explore the promises and pitfalls of the Generative Pre-trained Transformer (GPT) AI and potentially future technologies by adopting a speculative methodology. Speculative future narratives with a specific focus on educational contexts are provided in an attempt to identify emerging themes and discuss their implications for education in the 21st century. Affordances of (using) AI in Education (AIEd) and possible adverse effects are identified and discussed which emerge from the narratives. It is argued that now is the best of times to define human vs AI contribution to education because AI can accomplish more and more educational activities that used to be the prerogative of human educators. Therefore, it is imperative to rethink the respective roles of technology and human educators in education with a future-oriented mindset.

A descriptive bibliometric analysis of works on artificial intelligence (AI) in science education is provided in this article to help readers understand the state of the field's research at the time. This study's main objective is to give bibliometric data on publications regarding AI in science education printed in periodicals listed in the Scopus database between 2002 and 2023 end of May. The data gathered from publications scanned and published within the study's parameters was subjected to descriptive bibliometric analysis based on seven categories: number of articles and citations per year, countries with the most publications, most productive author, most significant affiliation, funding institutions, publication source and subject areas. Most of the papers were published between 2016 and 2022. The United States of America, United Kingdom, and China were the top-3 most productive nations, with the United States of America producing the most publications. The number of citations to the publications indexed in Scopus database increased in a progressive way and reached to maximum number in 2022 with 178 citations. Most productive author on this topic was Salles, P. with four publications. Moreover, Carnegie Mellon University, University of Memphis, and University of Southern California have the maximum number of publications as affiliations. The National Science Foundation was the leader funding institution in terms of number of publications produced. In addition, "Proceedings Frontiers in Education Conference Fie" have the highest number of publications by year as a publication source. Distribution of the publications by subject area was analyzed. The subject areas of the publications were computer sciences, social sciences, science education, technology and engineering education respectively. This study presents a vision for future research and provides a global perspective on AI in science education.

The emergence of ChatGPT, a Generative AI program, has sparked discussions about its teaching and learning value, and concerns about academic integrity in higher education (HE). An extant review of the literature indicates that a scarcity of research exists on GenAI, specifically a synthesis of the official views, guidelines and articles of top-ranked universities on the use, limitations, challenges, and opportunities brought by ChatGPT in the early phase when ChatGPT was released in 2022 until early May 2023, which can offer insights into the concerns and recommendations for educators. Using the corpus assisted discourse analysis approach, this study identified the key themes and emotions elicited by evaluations of the ChatGPT situation from a self-built corpus containing 151 articles from 47 top-ranked U.S. universities and 34 U.K. ones. Our findings indicated three prominent themes discussed on official websites, including ChatGPT as a text and content generator, use of ChatGPT in teaching and learning, and potential implications and opportunities of using ChatGPT in HE. Further examination revealed that bias, concern, worry, threat, fear, and trust were the prevailing emotions relating to ChatGPT. Illustrated with examples collected from our corpus, this paper offers an in-depth discussion of universities' first responses to the use of ChatGPT. The insights gained have some pedagogical implications for academics, researchers and educators and may inform policy and practice in HE on the use of GenAI.

For over two decades now, the application of Citizen Science to Education has been evolving, and fundamental topics, such as the drivers of motivation to participate in Citizen Science projects, are still under discussion. Some recent developments, though, like the use of Artificial Intelligence to support data collection and validation, seem to point to a clear-cut divergence from the mainstream research path. The objective of this paper is to summarise the development trajectory of research on Citizen Science in Education so far, and then shed light on its future development, to help researchers direct their efforts towards the most promising open questions in this field. We achieved these objectives by using the lens of the Affordance-Actualisation theory and the Main Path Analysis method.

Purpose: This paper aims to explore how critical digital and data literacies are facilitated by testing different methods in the classroom, with the ambition to find a pedagogical framework for prompting sustained critical literacies. Design/methodology/approach: This contribution draws on a 10-year set of critical pedagogy experiments conducted in Denmark, USA and Italy, and engaging more than 1,500 young adults. Multi-method pedagogical design trains students to conduct self-oriented guided autoethnography, situational analysis, allegorical mapping, and critical infrastructure analysis. Findings: The techniques of guided autoethnography for facilitating sustained data literacy rely on inviting multiple iterations of self-analysis through sequential prompts, whereby students move through stages of observation, critical thinking, critical theory-informed critique around the lived experience of hegemonic data and artificial intelligence (AI) infrastructures. Research limitations/implications: Critical digital/data literacy researchers should continue to test models for building sustained critique that not only facilitate changes in behavior over time but also facilitate citizen social science, whereby participants use these autoethnographic techniques with friends and families to build locally relevant critique of the hegemonic power of data/AI infrastructures. Originality/value: The proposed literacy model adopts a critical theory stance and shows the value of using multiple modes of intervention at micro and macro levels to prompt self-analysis and meta-level reflexivity for learners. This framework places critical theory at the center of the pedagogy to spark more radical stances, which is contended to be an essential step in moving students from attitudinal change to behavioral change.

Chatbots have experienced significant growth over the past decade, with a proliferation of new applications across various domains. Previous studies also demonstrate the trend of new technologies, especially artificial intelligence, being adopted in libraries. The purpose of this study is to determine the current research priorities and findings in the field of chatbots in libraries. A systematic literature review was performed utilising the PRISMA checklist and the databases Scopus and Web of Science, identifying 5734 records. Upon conducting the first screening, abstract screening, full-text assessment, and quality assessments guided by the CASP appraisal checklist, 19 papers were deemed suitable for inclusion in the review. The results of the review indicate that the majority of the existing studies were empirical in nature (primarily adopting qualitative methods) and technology reviews with a focus on reviewing the implementation and maintenance, design, evaluation, characteristics, and application of chatbots. The chatbots of interest were mainly text-based and guided chatbots, with closed-source tools with access portals mostly built on library web pages or integrated with social software. The research findings primarily concerned the development models and necessary tools and technologies, the application of chatbots in libraries. Our systematic review also suggests that studies on chatbots in libraries are still in the early stages. [This paper was presented at the 2023 Libraries in the Digital Age (LIDA) International Conference (Osijek, Croatia, May 24-26, 2023).]

This study explores the trends of chatbots in education studies by conducting a literature review to analyze relevant papers published in the Social Science Citation Index (SSCI) journals by searching the Web of Science (WoS) database. From the analysis results, it was found that the United States, Taiwan and Hong Kong are the top three contributing countries or regions. In addition, most studies adopted quantitative methods in their research design, such as ANOVA (Analysis of variance), descriptive statistics, t test, and correlation analysis. ANCOVA (Analysis of covariance) was the most frequently adopted approach for comparing the performances or perceptions of different groups of students. From the analysis results, the greatest proportion of studies adopted guided learning, followed by no learning activities. It was determined that the studies related to chatbots in education are still in an early stage since there are few empirical studies investigating the use of effective learning designs or learning strategies with chatbots. This implies much room for conducting relevant research to drive innovative teaching in terms of improving the learning process and learning outcomes. Finally, we highlight the research gaps and suggest several directions for future research based on the findings in the present study.

This is a conceptual paper that examines the origin and development of universities and their current role in global society. There has been an unprecedented and exponential growth of technology and artificial intelligence capabilities over the past ten years which is challenging current working practices and affecting all areas of society. The paper examines how this role may change to match the new demands placed on them by a digitally enabled society that has greater leisure time. The design of the paper is first to detail some of the changes in work practices that are taking place and how these will impact on society. It then offers several ways in which universities could modify their role to respond to these emerging challenges. This could include new courses, new organisational structures and new pedagogical practices. The paper provides a platform for discussion and debate around the strategic vision and direction of travel for higher education.

Purpose: This paper aims to present a comprehensive review of the present state and trends of smart education research. It addresses the need to have a systematic review of smart education to depict its research landscape in view of the growing volume of related publications. Design/methodology/approach: A bibliometric analysis of publications on smart education published in 2011 to 2020 was conducted, covering their patterns and trends in terms of collaboration, key publications, major topics and trends. A total of 1,317 publications with 29,317 cited references were collected from the Web of Science and Scopus for the bibliometric analysis. Findings: Research on smart education has been widely published in various sources. The most frequently cited references are all theoretical or discussion articles. Researchers in the USA, China, South Korea, India and Russia have been most active in research collaborations. However, international collaborations have remained infrequent except for those involving the USA. The research on smart education broadly covered smart technologies as well as teaching and learning. The emerging topics have addressed areas such as the Internet of Things, big data, flipped learning and gamification. Originality/value: This study depicts the intellectual landscape of smart education research, and illustrated the evolution and emerging trends in the field. The results highlight its latest developments and research needs, and suggest future work related to research collaborations on a larger scale and more studies on smart pedagogies.

As documented in this special issue, all over the world hard choices must be made in education, government, business, and medicine. Intelligence tests, used intelligently and with appropriate ethical safeguards, are one tool of many that help make hard choices work out well, or at least better than the next-best alternative (Kaufman, Raiford, & Coalson, 2016). The reliability of intelligence tests is far from perfect. Complaining about IQ tests is the privilege of those who have them. It is probably no accident that intelligence tests were invented in France, not long after a series of reforms from 1881 to 1901 made education free and compulsory for all children (Harrigan, 2001). It is likewise probably not an accident that intelligence testing was then adopted most enthusiastically in the world's wealthiest countries in the midst of similar attempts to raise educational standards. Among the countries featured in this special issue, Greece, Japan, the Netherlands, and the United States achieved near-universal education decades ago; Brazil, India, Mexico, Oman, and Peru have done so only within the last 15 years. Once a government takes on the responsibility of educating all of its citizens and then successfully achieves near-universal school attendance, it is confronted with the magnitude of population-wide individual differences in academic aptitude. The most pressing concern has to do with vulnerable children with intellectual disabilities. It is true that every child can learn, but not every child learns best in regular education. In this commentary, the author addresses the relation between universal education and the need for intelligence testing. The article goes on to compare U.S. developments in intelligence testing with those in Oman, Greece, India, Japan, Brazil, Peru, and the Netherlands. It concludes with three themes that emerged among the articles in this issue.