Peter Bannister
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Peter Bannister has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Alexandra Santamaría Urbieta
Universidad Internacional de La Rioja, Logroo, Spain
5 shared publications - Elena Alcalde Peñalver
Universidad de Alcalà
4 shared publications Alexandra Santamaría Urbieta
1 shared publication- Jack Walton
Centre for Research in Assessment and Digital Learning (CRADLE), Deakin University
1 shared publication - Jean-Philippe Deranty
Macquarie University
1 shared publication - José María Ariso
Universidad Internacional de la Rioja, Logroño, Spain
1 shared publication
Latest Publications
- Open Access
- 28 January 2026
- Published by Universidad de Alicante Servicio de Publicaciones in Revista Alicantina de Estudios Ingleses
- No. 44,p. 7-34
Exploring Legitimation Code Theory and TikTok: Knowledge Codes and Semantic Profiles in Finfluencer Discourse
This exploratory study employs Legitimation Code Theory (LCT), a theoretical framework for analysing knowledge practices across disciplinary and social contexts, to elucidate the construction and legitimisation of financial knowledge on TikTok, a platform witnessing a surge in informal financial literacy education. A meticulously curated multimodal corpus of financial influencer (‘finfluencer’) videos (n=100; 50 English, 50 Spanish) underwent rigorous analysis utilising key LCT constructs. Specifically, we examined semantic gravity (the degree to which meaning relates to specific contextual referents) and semantic density (the concentration of meaning within symbols or practices) to generate nuanced semantic profiles that illuminate knowledge-building practices. Empirical semantic profile mapping revealed a striking dichotomy: English-language videos predominantly exhibited prosaic codes, characterised by high interconnectivity between financial concepts and everyday applications, facilitating accessibility whilst maintaining conceptual coherence. Conversely, Spanish-language videos manifested primarily rarefied codes, distinguished by abstract financial terminology with limited contextual grounding and reduced practical interconnections. This linguistic divergence intimates that cultural and linguistic milieus profoundly shape the contours of financial knowledge dissemination across digital platforms. Further comparative analysis revealed a significant correlation between finfluencer popularity metrics and discourse characteristics; those with larger followings tended towards weaker semantic gravity and greater abstraction, corroborating theoretical postulations that elevated platform status bolsters claims to knowledge legitimacy. The findings underscore LCT’s efficacy in illuminating the intricate interplay among cultural contexts, platform-specific power hierarchies, and symbolic capital in the online propagation of financial concepts. This research offers valuable insights into the multifaceted dynamics of knowledge legitimisation in digital spaces, potentially informing future strategies for developing cross-cultural financial literacy programmes that acknowledge diverse knowledge-building practices.
- 28 January 2026
- Published by Universidad de Alicante Servicio de Publicaciones in Revista Alicantina de Estudios Ingleses
- No. 44,p. 7-34
Open AccessExploring Legitimation Code Theory and TikTok: Knowledge Codes and Semantic Profiles in Finfluencer Discourse
This exploratory study employs Legitimation Code Theory (LCT), a theoretical framework for analysing knowledge practices across disciplinary and social contexts, to elucidate the construction and legitimisation of financial knowledge on TikTok, a platform witnessing a surge in informal financial literacy education. A meticulously curated multimodal corpus of financial influencer (‘finfluencer’) videos (n=100; 50 English, 50 Spanish) underwent rigorous analysis utilising key LCT constructs. Specifically, we examined semantic gravity (the degree to which meaning relates to specific contextual referents) and semantic density (the concentration of meaning within symbols or practices) to generate nuanced semantic profiles that illuminate knowledge-building practices. Empirical semantic profile mapping revealed a striking dichotomy: English-language videos predominantly exhibited prosaic codes, characterised by high interconnectivity between financial concepts and everyday applications, facilitating accessibility whilst maintaining conceptual coherence. Conversely, Spanish-language videos manifested primarily rarefied codes, distinguished by abstract financial terminology with limited contextual grounding and reduced practical interconnections. This linguistic divergence intimates that cultural and linguistic milieus profoundly shape the contours of financial knowledge dissemination across digital platforms. Further comparative analysis revealed a significant correlation between finfluencer popularity metrics and discourse characteristics; those with larger followings tended towards weaker semantic gravity and greater abstraction, corroborating theoretical postulations that elevated platform status bolsters claims to knowledge legitimacy. The findings underscore LCT’s efficacy in illuminating the intricate interplay among cultural contexts, platform-specific power hierarchies, and symbolic capital in the online propagation of financial concepts. This research offers valuable insights into the multifaceted dynamics of knowledge legitimisation in digital spaces, potentially informing future strategies for developing cross-cultural financial literacy programmes that acknowledge diverse knowledge-building practices.
- 9 December 2025
- Published by Taylor & Francis in Higher Education Research & Development
- Vol. 45 (3) , 847-856
ParadAIse L0st?
- 9 December 2025
- Published by Taylor & Francis in Higher Education Research & Development
- Vol. 45 (3) , 847-856
ParadAIse L0st?
- Open Access
- 24 November 2025
- Published by Springer Nature in AI & SOCIETY
- Vol. 41 (4) , 3475-3487
‘AI lost the prompt!’ Replacing ‘AI hallucination’ to distinguish between mere errors and irregularities
One of the principal areas of current AI research concerns what are termed “hallucinations”. Whilst hundreds of different definitions and classifications of “AI hallucination” have been published, none have yet considered the distinction between errors and irregularities in Wittgenstein’s sense. This article provides a straightforward explanation of this distinction, illustrated through examples of AI outputs drawn from various publications. We then examine the terms proposed as alternatives to “hallucination” and highlight both their strengths and weaknesses. Drawing upon this analysis, we establish criteria for proposing alternative terms that encompass both errors and irregularities in Wittgenstein’s sense. Our aim is not to definitively resolve the ongoing debate surrounding suitable replacements for “AI hallucination”, but rather to provide a comprehensive overview of the characteristics and nuances that this distinction brings to the discussion. For unlike errors, irregularities prove entirely incomprehensible to users, owing to the grammatical gap created when fundamental certainties that underpin meaningful language use are violated. Given that irregularities prove incomprehensible, as they violate the certainties underlying meaningful language use, the most trustworthy AI systems may ultimately be those that recognise their own epistemic boundaries rather than those that produce seemingly perfect outputs.
- 24 November 2025
- Published by Springer Nature in AI & SOCIETY
- Vol. 41 (4) , 3475-3487
Open Access‘AI lost the prompt!’ Replacing ‘AI hallucination’ to distinguish between mere errors and irregularities
One of the principal areas of current AI research concerns what are termed “hallucinations”. Whilst hundreds of different definitions and classifications of “AI hallucination” have been published, none have yet considered the distinction between errors and irregularities in Wittgenstein’s sense. This article provides a straightforward explanation of this distinction, illustrated through examples of AI outputs drawn from various publications. We then examine the terms proposed as alternatives to “hallucination” and highlight both their strengths and weaknesses. Drawing upon this analysis, we establish criteria for proposing alternative terms that encompass both errors and irregularities in Wittgenstein’s sense. Our aim is not to definitively resolve the ongoing debate surrounding suitable replacements for “AI hallucination”, but rather to provide a comprehensive overview of the characteristics and nuances that this distinction brings to the discussion. For unlike errors, irregularities prove entirely incomprehensible to users, owing to the grammatical gap created when fundamental certainties that underpin meaningful language use are violated. Given that irregularities prove incomprehensible, as they violate the certainties underlying meaningful language use, the most trustworthy AI systems may ultimately be those that recognise their own epistemic boundaries rather than those that produce seemingly perfect outputs.
- Open Access
- 16 October 2025
- Published by Taylor & Francis in Educational Philosophy and Theory
- Vol. 58 (3) , 198-210
On the essay in a time of GenAI
- 16 October 2025
- Published by Taylor & Francis in Educational Philosophy and Theory
- Vol. 58 (3) , 198-210
Open AccessOn the essay in a time of GenAI
- Open Access
- 20 August 2025
- Published by Wiley in Higher Education Quarterly
- Vol. 79 (4)
Validating the PANDORA GenAI Susceptibility Rubric for Higher Education Assessment: A Field Test of All Translation and Interpreting BA Assignments
This study, conducted at a fully online Spanish higher education institution, documents the validation of a bespoke quality assessment tool designed to measure the susceptibility of formative assignments to AI academic misconduct. The research explored the impact of Generative AI (GenAI) technologies in the Humanities. The framework study consisted of four stages: the design of a rubric (Stages 1–3) and its large‐scale validation (Stage 4) through a field test in the Translation and Interpreting Studies Bachelor's Degree. This paper presents stage 4 results where lecturers ( n = 29), using a bottom‐up approach, voluntarily applied the tool to their teaching contexts and analysed assignments ( n = 151) using the rubric, revealing significant vulnerabilities in assessments easily converted by GenAI or lacking originality and collaboration. The findings guided AI‐integrated assessment designs that encourage complexity, creativity, and ethical engagement. This study outlines effective GenAI practices in assessment design and emphasises innovative methods for academic integrity in higher education.
- 20 August 2025
- Published by Wiley in Higher Education Quarterly
- Vol. 79 (4)
Open AccessValidating the PANDORA GenAI Susceptibility Rubric for Higher Education Assessment: A Field Test of All Translation and Interpreting BA Assignments
This study, conducted at a fully online Spanish higher education institution, documents the validation of a bespoke quality assessment tool designed to measure the susceptibility of formative assignments to AI academic misconduct. The research explored the impact of Generative AI (GenAI) technologies in the Humanities. The framework study consisted of four stages: the design of a rubric (Stages 1–3) and its large‐scale validation (Stage 4) through a field test in the Translation and Interpreting Studies Bachelor's Degree. This paper presents stage 4 results where lecturers ( n = 29), using a bottom‐up approach, voluntarily applied the tool to their teaching contexts and analysed assignments ( n = 151) using the rubric, revealing significant vulnerabilities in assessments easily converted by GenAI or lacking originality and collaboration. The findings guided AI‐integrated assessment designs that encourage complexity, creativity, and ethical engagement. This study outlines effective GenAI practices in assessment design and emphasises innovative methods for academic integrity in higher education.
- Open Access
- 2 January 2025
- Published by Kaplan Business School, Australia in Journal of Applied Learning & Teaching
- Vol. 8 (1)
Appraising higher education assessment validity: Development of the PANDORA GenAI Susceptibility Rubric
This paper presents the development and application of the PANDORA GenAI Susceptibility Rubric, a novel tool designed to assess the susceptibility of higher education assessments to the undeclared use of generative artificial intelligence (GenAI) tools. In response to growing concerns about academic integrity and the rising sophistication of GenAI technologies, the rubric provides educators with a structured framework to critically evaluate the validity of their assessments across key criteria, including collaborative authorship, intellectual task complexity, and the opportunity for creativity. Through a mixed-methods design, the rubric was refined to include expert-informed modifications and validated through end-user application across various arts and humanities courses. Results highlight how assessment design can either mitigate or exacerbate GenAI susceptibility, revealing that tasks requiring genuine collaboration, creative thinking, and process-oriented evaluation offer greater resistance to AI manipulation. The rubric also emphasises balancing detailed guidance with student autonomy to avoid facilitating GenAI prompt formulation. This study contributes to the field by offering a practical instrument that promotes more robust, ethically sound, and future-proofed assessment practices. It serves as a critical response to the pressing challenges posed by GenAI in higher education and informs ongoing discourse on academic integrity.
- 2 January 2025
- Published by Kaplan Business School, Australia in Journal of Applied Learning & Teaching
- Vol. 8 (1)
Open AccessAppraising higher education assessment validity: Development of the PANDORA GenAI Susceptibility Rubric
This paper presents the development and application of the PANDORA GenAI Susceptibility Rubric, a novel tool designed to assess the susceptibility of higher education assessments to the undeclared use of generative artificial intelligence (GenAI) tools. In response to growing concerns about academic integrity and the rising sophistication of GenAI technologies, the rubric provides educators with a structured framework to critically evaluate the validity of their assessments across key criteria, including collaborative authorship, intellectual task complexity, and the opportunity for creativity. Through a mixed-methods design, the rubric was refined to include expert-informed modifications and validated through end-user application across various arts and humanities courses. Results highlight how assessment design can either mitigate or exacerbate GenAI susceptibility, revealing that tasks requiring genuine collaboration, creative thinking, and process-oriented evaluation offer greater resistance to AI manipulation. The rubric also emphasises balancing detailed guidance with student autonomy to avoid facilitating GenAI prompt formulation. This study contributes to the field by offering a practical instrument that promotes more robust, ethically sound, and future-proofed assessment practices. It serves as a critical response to the pressing challenges posed by GenAI in higher education and informs ongoing discourse on academic integrity.
- 14 November 2024
- Published by Taylor & Francis in Professional Development in Education
- Vol. 51 (3) , 547-565
‘I don’t need professional development; I want institutional development’: legitimising marginalised epistemic capital that disrupts generative AI discourse
- 14 November 2024
- Published by Taylor & Francis in Professional Development in Education
- Vol. 51 (3) , 547-565
‘I don’t need professional development; I want institutional development’: legitimising marginalised epistemic capital that disrupts generative AI discourse
- Open Access
- 7 May 2024
- Published by Forum XXI in Vivat Academia
- No. 157,p. 1-25
Estrategias metacomunicativas y de multimodalidad en inglés y en español de los medinfluencers en Instagram
Introduction: Social media has become a means of communication that citizens frequently turn to not only for entertainment but also to obtain updated information on a specific topic. Medfluencers, individuals dedicated to disseminating health and wellness-related topics, often appear on worldwide social platforms. Methodology: In order to discern the discourse, both written and spoken, used by these individuals in both Spanish and English language, a total of 200 posts from Instagram have been analyzed, comprising 100 posts in Spanish and 100 posts in English. Results: The results indicate that there are differences between the two languages, as the Spanish language tends to be more detailed, while in English, there is a preference for being more direct and less exhaustive in the information provided to the audience. Discussion: Studies like the ones presented here, as well as their findings, are not intended to be a mere data describing exercise of a specific type of discourse. Instead, the purpose here is to be valuable for the pedagogy of specific fields of knowledge, such as translation, interpreting, linguistics, text editing, or communication. Conclusions: Some of the main conclusions drawn from this study include the trend of the Spanish language to be more didactic in its explanations, while the English language is more direct and concise. Likewise, regarding multimodal elements, once again, the Spanish language uses them more frequently than the English language. All of this demonstrates that the Spanish-speaking medfluencer aims to be didactic and educate, while in English, their primary goal is simply to provide information. Introducción: Las redes sociales se han convertido en medios de comunicación a los que los ciudadanos recurrimos con frecuencia no solo para entretenernos, sino también para obtener información actualizada sobre un tema concreto. Los medinfluencers, personas que se dedican a la divulgación de temas relacionados con la salud y el bienestar, aparecen con frecuencia en redes sociales. Metodología: Con el objetivo de determinar el discurso, tanto escrito como oral, que estos emplean tanto en lengua española como en lengua inglesa, se han analizado un total de 200 publicaciones de Instagram, 100 en español y 100 en inglés. Resultados: Los resultados dan muestra de que existen diferencias entre ambas lenguas, puesto que la lengua española tiende al detalle, mientras que en lengua inglesa se prefiere ser más directo y no ser tan exhaustivo en cuanto a la información que se le da al receptor. Discusión: Estudios como los que aquí se presentan, así como sus resultados, no pretenden ser únicamente datos que describan un tipo de discurso concreto, sino que buscan servir de utilidad para la didáctica de determinados campos de conocimiento como pueden ser la traducción, la interpretación, la lingüística, la corrección de textos o la comunicación. Conclusiones: De entre las principales conclusiones que se extraen en este estudio, destacamos la tendencia de la lengua española por ser más didáctica en sus explicaciones, mientras que la lengua inglesa resulta ser más directa y concisa. Asimismo, en cuanto a los elementos multimodales, nuevamente, la lengua española los emplea con mayor frecuencia que la inglesa. Todo esto da muestra de que el medinfluencer en español pretende ser didáctico y enseñar, mientras que en inglés este tiene como objetivo únicamente informar.
- 7 May 2024
- Published by Forum XXI in Vivat Academia
- No. 157,p. 1-25
Open AccessEstrategias metacomunicativas y de multimodalidad en inglés y en español de los medinfluencers en Instagram
Introduction: Social media has become a means of communication that citizens frequently turn to not only for entertainment but also to obtain updated information on a specific topic. Medfluencers, individuals dedicated to disseminating health and wellness-related topics, often appear on worldwide social platforms. Methodology: In order to discern the discourse, both written and spoken, used by these individuals in both Spanish and English language, a total of 200 posts from Instagram have been analyzed, comprising 100 posts in Spanish and 100 posts in English. Results: The results indicate that there are differences between the two languages, as the Spanish language tends to be more detailed, while in English, there is a preference for being more direct and less exhaustive in the information provided to the audience. Discussion: Studies like the ones presented here, as well as their findings, are not intended to be a mere data describing exercise of a specific type of discourse. Instead, the purpose here is to be valuable for the pedagogy of specific fields of knowledge, such as translation, interpreting, linguistics, text editing, or communication. Conclusions: Some of the main conclusions drawn from this study include the trend of the Spanish language to be more didactic in its explanations, while the English language is more direct and concise. Likewise, regarding multimodal elements, once again, the Spanish language uses them more frequently than the English language. All of this demonstrates that the Spanish-speaking medfluencer aims to be didactic and educate, while in English, their primary goal is simply to provide information. Introducción: Las redes sociales se han convertido en medios de comunicación a los que los ciudadanos recurrimos con frecuencia no solo para entretenernos, sino también para obtener información actualizada sobre un tema concreto. Los medinfluencers, personas que se dedican a la divulgación de temas relacionados con la salud y el bienestar, aparecen con frecuencia en redes sociales. Metodología: Con el objetivo de determinar el discurso, tanto escrito como oral, que estos emplean tanto en lengua española como en lengua inglesa, se han analizado un total de 200 publicaciones de Instagram, 100 en español y 100 en inglés. Resultados: Los resultados dan muestra de que existen diferencias entre ambas lenguas, puesto que la lengua española tiende al detalle, mientras que en lengua inglesa se prefiere ser más directo y no ser tan exhaustivo en cuanto a la información que se le da al receptor. Discusión: Estudios como los que aquí se presentan, así como sus resultados, no pretenden ser únicamente datos que describan un tipo de discurso concreto, sino que buscan servir de utilidad para la didáctica de determinados campos de conocimiento como pueden ser la traducción, la interpretación, la lingüística, la corrección de textos o la comunicación. Conclusiones: De entre las principales conclusiones que se extraen en este estudio, destacamos la tendencia de la lengua española por ser más didáctica en sus explicaciones, mientras que la lengua inglesa resulta ser más directa y concisa. Asimismo, en cuanto a los elementos multimodales, nuevamente, la lengua española los emplea con mayor frecuencia que la inglesa. Todo esto da muestra de que el medinfluencer en español pretende ser didáctico y enseñar, mientras que en inglés este tiene como objetivo únicamente informar.
- Open Access
- 21 April 2024
- Published by STAR Scholars Network in Journal of International Students
- Vol. 14 (3) , 149-170
International Students and Generative Artificial Intelligence
An exploratory analysis of tertiary education GenAI academic integrity policies has been undertaken, to examine provision for international students, having been identified as vulnerable to bias and integrity challenges associated with these technologies. A qualitative document analysis of 131 HE institutions (HEI) policies across 11 countries reveals mostly scant consideration of international students' needs, with the majority lacking any explicit provision. A subsequent SWOT analysis of exceptions to this generally indicates strengths in acknowledging diverse learner assumptions but weaknesses regarding specificity in articulating tailored expectations and support. These novel findings suggest international students have been largely overlooked by HEIs here despite generally significant economic contributions, reflecting the tenets of the commodification of HE. Results also highlight an urgent need to address gaps by incorporating international student voices in policy development,. Not addressing these issues risks exacerbating frictions, undermining academic integrity, and further disadvantaging this group in policy responses to GenAI.
- 21 April 2024
- Published by STAR Scholars Network in Journal of International Students
- Vol. 14 (3) , 149-170
Open AccessInternational Students and Generative Artificial Intelligence
An exploratory analysis of tertiary education GenAI academic integrity policies has been undertaken, to examine provision for international students, having been identified as vulnerable to bias and integrity challenges associated with these technologies. A qualitative document analysis of 131 HE institutions (HEI) policies across 11 countries reveals mostly scant consideration of international students' needs, with the majority lacking any explicit provision. A subsequent SWOT analysis of exceptions to this generally indicates strengths in acknowledging diverse learner assumptions but weaknesses regarding specificity in articulating tailored expectations and support. These novel findings suggest international students have been largely overlooked by HEIs here despite generally significant economic contributions, reflecting the tenets of the commodification of HE. Results also highlight an urgent need to address gaps by incorporating international student voices in policy development,. Not addressing these issues risks exacerbating frictions, undermining academic integrity, and further disadvantaging this group in policy responses to GenAI.
- Top Cited PapersOpen Access
- 20 December 2023
- Published by Universidad de Oviedo in Aula Abierta
- Vol. 52 (4) , 401-409
A Systematic Review of Generative AI and (English Medium Instruction) Higher Education
This systematic review investigates the current state of research on Generative Artificial Intelligence (GenAI) and its implications for (EMI) Higher Education. The study employs a methodology based on an evidence-informed and theoretically credible framework to answer two research questions: (1) What studies of relevance to (EMI) Higher Education have been published thus far, considering the most recent developments of GenAI? and (2) Which key areas are currently lacking in extant literature and in need of further scholarly exploration in this regard in (EMI) Higher Education research? The results of the study reveal a limited number of pertinent publications, indicating a sparse scholarly landscape with a dearth of work on the implications of Generative AI in EMI Higher Education. Based on these findings, preliminary recommendations have been made to guide future research in this area. This study contributes to the literature by highlighting the need for further research on the potential of GenAI to enhance the teaching and learning experience in (EMI) Higher Education and provides a theoretical framework to guide future research. These findings may inform researchers and educators interested in exploring how GenAI may be leveraged from different educational perspectives.
- 20 December 2023
- Published by Universidad de Oviedo in Aula Abierta
- Vol. 52 (4) , 401-409
Top Cited PapersOpen AccessA Systematic Review of Generative AI and (English Medium Instruction) Higher Education
This systematic review investigates the current state of research on Generative Artificial Intelligence (GenAI) and its implications for (EMI) Higher Education. The study employs a methodology based on an evidence-informed and theoretically credible framework to answer two research questions: (1) What studies of relevance to (EMI) Higher Education have been published thus far, considering the most recent developments of GenAI? and (2) Which key areas are currently lacking in extant literature and in need of further scholarly exploration in this regard in (EMI) Higher Education research? The results of the study reveal a limited number of pertinent publications, indicating a sparse scholarly landscape with a dearth of work on the implications of Generative AI in EMI Higher Education. Based on these findings, preliminary recommendations have been made to guide future research in this area. This study contributes to the literature by highlighting the need for further research on the potential of GenAI to enhance the teaching and learning experience in (EMI) Higher Education and provides a theoretical framework to guide future research. These findings may inform researchers and educators interested in exploring how GenAI may be leveraged from different educational perspectives.