Reducing interpretative ambiguity in an educational environment with ChatGPT

Abstract

The study posits that both concrete and abstract words are crucial for effective communication, particularly in educational contexts where the interplay between these forms of language intersects with linguistic, cognitive, and social stratification theories. A key challenge is balancing the efficiency of abstract language in conveying complex concepts with the accessibility of concrete language, which enhances student comprehension. Generative languages, with their capacity to manipulate symbols, offer a way to navigate this challenge by facilitating the structured and systematic representation and exploration of abstract concepts within their contexts. The central research question was: “How can generative languages assist educational stakeholders in articulating their ideas and actions more clearly by identifying and refining abstract terms?” To explore this, a protocol in English was developed for ChatGPT-4, featuring structured guidelines and prompts aimed at helping users achieve specific educational goals. In a pilot study involving 13 participants, ChatGPT-4 provided feedback, suggested improvements, and guided users through text interactions. One of the authors observed the participants, took notes on their behavior, and conducted brief post-exercise discussions to gauge their experiences. After the session, participants were asked to reflect on their experience and share their thoughts via email. The process helped participants refine their responses from abstract to more concrete terms, enhancing clarity and engagement with educational content. The ChatGPT-4 protocol effectively bridges the gap between abstract pedagogical theories and practical classroom application, training teachers to use vivid descriptions, relatable scenarios, and tangible examples. This study illustrates how artificial intelligence can successfully integrate teaching principles and learning theories to enhance educational practices.