AN OBSERVATIONAL STUDY OF LARGE LANGUAGE MODELS (LLM) FOR ACADEMIC WRITING METHOD

ABSTRACT

Large language models (LLMs) for academic writing are the subject of this project's observational study, which aims to assess how well they improve writing quality. The study uses a mixed-methods approach to gather information from a wide range of academic writers, including questionnaires, writing samples, and user feedback. The process entails choosing pertinent LLMs, such GPT-3, according to standards including applicability, availability, and relevance for assignments involving academic writing. Writing samples are used to evaluate changes in writing quality before and after using LLMs, and surveys are used to collect feedback and demographic data. Improvements in writing quality, favorable user opinions, and comparisons of LLM efficacy according to participant and model size are among the anticipated outcomes. The study also looks at moral issues like plagiarism detection that are connected to using LLM in academic writing. Overall, this study contributes to the understanding of LLMs' impact on academic writing and provides insights for educators, researchers, and developers in the field of data science and academic writing.