Speakers
Description
The impact of artificial intelligence on language learning tools and specifically dictionaries has seen a significant shift with the advent of generative AI and chatbot technologies (De Schryver, 2023; Lew, 2023; Łodzikowski et al., 2024; Rees & Lew, 2024). We report on a study comparing the use of a mobile dictionary (Longman Dictionary of Contemporary English) and ChatGPT—an innovative conversational agent—in assisting advanced English learners in lexical tasks involving both reception and production. To assess the effectiveness of these tools, participants engaged in a series of paper-based lexical exercises designed to evaluate how each resource supports advanced students of English in solving lexical problems. For the study, a cohort of 223 advanced college-level learners of English were divided into two groups: one utilizing ChatGPT-3.5 (freely available version) and the other using a mobile dictionary (Longman Dictionary of Contemporary English) to complete the same set of tasks. Performance was measured in terms of accuracy and time-on-task. In the reception task, which involved understanding, interpreting, and translating English words into Polish, participants using ChatGPT demonstrated a significantly higher (assuming an alpha level of α = 0.05) rate of correct answers compared to those using the mobile dictionary (Odds Ratio 3.46 [CI95% 2.48–4.83], p < 0.001). Similarly, in the production tasks that required active completion of partial translations, the ChatGPT group outperformed the dictionary group (Odds Ratio 5.38 [CI95% 3.54–8.16], p < 0.001). Time-on-task was also recorded, with results showing ChatGPT to be a more efficient tool in terms of the time needed for participants to answer questions in the production task, but not in reception. This suggests that the instantaneity and interactive dialogue offered by ChatGPT enables learners to process and apply lexical items more swiftly than with conventional dictionary searches, thereby streamlining the consultation process. Apparently, LDOCE users were not always able to locate the relevant information even though it was present in the dictionary (we made sure that this was the case). The analysis extends beyond performance metrics to explore practical advantages and potential drawbacks of using ChatGPT in an educational setting. Among the benefits, the research highlights ChatGPT’s ability to provide immediate, context-relevant assistance and to engage learners through interactive and adaptive dialogue. This can potentially lead to enhanced retention and application of new language forms. Conversely, potential limitations include the risk of overreliance on AI for language inputs, reduced memorization efforts, and the challenge of ensuring accurate and pedagogically appropriate responses from the chatbot. The superior performance in both reception and production tasks underscores the importance of considering AI-driven resources as serious competitors to traditional dictionaries in language-learning contexts. However, this study has not examined bilingual dictionaries, which, as earlier studies with rigorous design show (e.g., Lew, 2004), may well be better suited to language learners at all levels of proficiency. Indeed, early results from follow-up study in progress suggest that a bilingual dictionary may result in at least as good success rates as ChatGPT, at least in Reception. We stress the need for a balanced approach, incorporating AI while maintaining essential strategies for independent language learning and critical thinking. The study advocates for further exploration into the optimized integration of AI-based tools into the language learning curriculum, ensuring they complement rather than replace learner autonomy and traditional pedagogical resources such as dictionaries.
