ChatGPT as a pre-submission writing support tool: A qualitative document analysis of published academic texts

This paper examines how English-Language academic manuscripts may benefit from a writing support tool like ChatGPT in detecting minor language-level errors and formatting issues. Using a qualitative document analysis design, this study explored some evidence of minor grammatical, typographical, formatting, citation, and repetition-related issues in six peer-reviewed English-language research articles. This selection of articles comprises a variety of applied linguistics, English language education, AI-assisted writing, GenAI-supported ELT, and research engagement contexts. The data collection began with ChatGPT as the initial screening tool, which identified possible issues in the selected texts. The researchers’ manual cross-checking of each identified issue extended the process. The data analysis noted observations on missing articles, awkward phrasing, word-choice problems, typographical errors, punctuation and spacing issues, capitalization inconsistencies, APA-style concerns, figure-caption problems, metadata issues, and repeated expressions or concepts. The findings revealed that figures, captions, references, and publication details also contain minor issues, simultaneously with the main body of the text. This suggests an additional pre-submission review stage of academic manuscript submission. ChatGPT can assist at this point by contributing ethically as a supportive proofreading and manuscript-refinement tool, not to replace human authorship, scholarly judgment, interpretation, or final decision-making, but to help the researchers’ pre-submission review.