View in EDS

APLICACIÓN DE UN MODELO GPT EN EL ESTUDIO DE CONSTELACIONES MULTIMODALES.

Saved in:
Bibliographic Details
Title: APLICACIÓN DE UN MODELO GPT EN EL ESTUDIO DE CONSTELACIONES MULTIMODALES.
Alternate Title: APPLICATION OF A GPT MODEL IN THE STUDY OF MULTIMODAL CONSTELLATIONS.
Authors: Rovira-Collado, José1, Miras, Sebastián1, Cantó-Delgado, Rocío1
Source: Perspectiva Educacional. nov2025, Vol. 64 Issue 3, p122-150. 29p.
Subject Terms: *Educational literature, *Critical literacy, *Artificial intelligence, ChatGPT, Intertextuality, Fictional characters, Multimodal user interfaces, Thematic analysis
Abstract (English): Artificial Intelligence (AI) tools enable content to be evaluated in a more comprehensive and effective way. The design and training of new virtual assistants allows us to structure content and offer more complete and consistent evaluations and analyses. This research analyzes the possibilities of a virtual assistant created with ChatGPT, designed specifically for the analysis and evaluation of multimodal constellations, an educational practice that improves students' critical literacy. These constellations integrate a multitude of elements such as literary works, film, television, music, digital references, and other Internet elements. A map of multimodal elements displays different intertextual relationships that are subsequently reflected in the creation of new literary and audiovisual narratives that show these intertextual links between the different elements. This practice has been carried out at the Faculty of Education of the University of Alicante since 2016, adding more and more elements and educational dynamics to improve literary education and multimodal reading in people who design these constellations. On this occasion, 58 constellations focusing on female protagonists in children's and young adult fiction have been analyzed. These were created by university students taking the Reading and Writing Didactics course in October 2024. All these constellations have been published openly on a teaching blog where more than 300 students participate: (https://constelacionesmultimodales4dle24.blogspot.com/) so that they can be used in the future by anyone. Subsequently, the research team created a specific ChatGPT model for the evaluation of the practice, also available on the blog. This virtual assistant has been fed with all the information published on the blog in October 2024, classified into different lists to identify each element of the constellation. This research conducts a qualitative content analysis of the responses provided by the specific GPT and the validity of the different prompts designed by the research team for the evaluation of multimodal constellations. The importance of focusing on female protagonists in different literary and audiovisual narratives is also analyzed. The results indicate that ChatGPT can locate data, point out trends in the design of constellations, identify the type of relationships between different elements, and offer suggestions for improving the different proposals. It also has limitations in its responses when analyzing aspects such as critical reflection, as well as some errors or hallucinations when presenting blog data, which must always be checked in the different entries. Thus, although the tool can support the evaluation process, its use should be a complement to the teacher's judgment. As future lines of research, the expansion of the sample with constellations from different educational stages is proposed, to determine whether ChatGPT is capable of detecting intertextual relationships established by the participants. It would also be interesting to use this ChatGPT model to support the process of designing new constellations focused on the same topic and to compare the sample from this study with productions from previous courses. [ABSTRACT FROM AUTHOR]
Abstract (Spanish): Las herramientas de Inteligencia Artificial permiten evaluar contenidos de una forma más amplia y eficaz. Esta investigación analiza las posibilidades de un ChatGPT creado específicamente para la evaluación de constelaciones multimodales, una práctica que permite mejorar la literacidad crítica del alumnado. A través del mapa de elementos multimodales se reflejan distintas relaciones intertextuales que posteriormente se plasman en relatos literarios y audiovisuales que aprovechen esas relaciones. Se han analizado 58 constelaciones elaboradas por alumnado universitario en octubre de 2024. Se realiza un análisis de contenido de las respuestas ofrecidas por el GPT específico y la validez de los distintos prompts diseñados para la evaluación de las constelaciones. Los resultados indican que ChatGPT puede localizar datos y ofrecer propuestas de mejora pero presenta limitaciones para analizar aspectos como la reflexión crítica. Así, aunque la herramienta puede apoyar el proceso de evaluación, su uso debe ser un complemento al juicio docente. [ABSTRACT FROM AUTHOR]
Copyright of Perspectiva Educacional is the property of Perspectiva Educacional and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Education Research Complete
Description
Abstract:Artificial Intelligence (AI) tools enable content to be evaluated in a more comprehensive and effective way. The design and training of new virtual assistants allows us to structure content and offer more complete and consistent evaluations and analyses. This research analyzes the possibilities of a virtual assistant created with ChatGPT, designed specifically for the analysis and evaluation of multimodal constellations, an educational practice that improves students' critical literacy. These constellations integrate a multitude of elements such as literary works, film, television, music, digital references, and other Internet elements. A map of multimodal elements displays different intertextual relationships that are subsequently reflected in the creation of new literary and audiovisual narratives that show these intertextual links between the different elements. This practice has been carried out at the Faculty of Education of the University of Alicante since 2016, adding more and more elements and educational dynamics to improve literary education and multimodal reading in people who design these constellations. On this occasion, 58 constellations focusing on female protagonists in children's and young adult fiction have been analyzed. These were created by university students taking the Reading and Writing Didactics course in October 2024. All these constellations have been published openly on a teaching blog where more than 300 students participate: (https://constelacionesmultimodales4dle24.blogspot.com/) so that they can be used in the future by anyone. Subsequently, the research team created a specific ChatGPT model for the evaluation of the practice, also available on the blog. This virtual assistant has been fed with all the information published on the blog in October 2024, classified into different lists to identify each element of the constellation. This research conducts a qualitative content analysis of the responses provided by the specific GPT and the validity of the different prompts designed by the research team for the evaluation of multimodal constellations. The importance of focusing on female protagonists in different literary and audiovisual narratives is also analyzed. The results indicate that ChatGPT can locate data, point out trends in the design of constellations, identify the type of relationships between different elements, and offer suggestions for improving the different proposals. It also has limitations in its responses when analyzing aspects such as critical reflection, as well as some errors or hallucinations when presenting blog data, which must always be checked in the different entries. Thus, although the tool can support the evaluation process, its use should be a complement to the teacher's judgment. As future lines of research, the expansion of the sample with constellations from different educational stages is proposed, to determine whether ChatGPT is capable of detecting intertextual relationships established by the participants. It would also be interesting to use this ChatGPT model to support the process of designing new constellations focused on the same topic and to compare the sample from this study with productions from previous courses. [ABSTRACT FROM AUTHOR]
ISSN:07160488
DOI:10.4151/07189729-Vol.64-Iss.3-Art.1762