The development of corpus-based genre-specific word list and the self-study materials for eap teaching: The case of metallurgical and materials engineering

A significant and complex aspect of English for Academic Purposes (EAP) research is vocabulary instruction. The instruction and acquisition of EAP vocabulary, learners' needs, the lexical demands of academia, and the teaching of technical terminology have been the focus of extensive scholarly investigation. Metallurgical and Materials Science Engineering English, as a subfield-specific component of EAP vocabulary, is no exception and holds critical importance for students and professionals in the discipline. Therefore, the primary aim of this thesis is to develop discipline-specific technical word lists for students studying Metallurgical and Materials Science Engineering, to design classroom activities generated through Generative Artificial Intelligence (GenAI), and to evaluate the effectiveness of these GenAI-based instructional materials. To achieve this aim, a corpus of 1,788,459 words was compiled from research articles published in SCI-indexed journals, representing four subfields of Metallurgical and Materials Science Engineering. The corpus was analyzed to identify the most frequently occurring words, and AntWordProfiler software was employed to compare the corpus with three widely recognized word lists: the Academic Word List (AWL), the General Service List 1 (GSL1), and the General Service List 2 (GSL2). Following this comparison, the 800 most frequent terms were selected and uploaded to GenAI "Gemini" along with three technical dictionaries. GenAI was instructed to compare the word lists with the discipline-specific dictionaries in order to identify terms referenced in them. The researcher then manually reviewed the output and removed words that did not appear in at least one of the dictionaries. The refined lists were subsequently validated by two experts, resulting in four discipline-specific technical vocabulary lists of 200 words each, labeled CMWL, MSWL, MWL, and PSTWL. Random sampling was used to determine which list would form the basis of the classroom activities, and MWL was selected. A pre-test containing MWL items was administered to nine participants, followed by three weeks of GenAI-generated classroom instruction, after which the same test was administered as a post-test. Results showed improvements ranging from 5.5% to 20.5%, with the post-test scores averaging 10.44 points higher. These findings highlight the necessity for more genre-specific vocabulary instruction in this field.