Corpus-Based Error Analysis in Accounting Students’ English Report Writing

Abstract

English for Specific Purposes is essential for accounting students to achieve professional communicative competence in financial and audit contexts. However, recurrent language errors in report writing reduce clarity, accuracy, and workplace readiness. This study investigates error patterns in accounting majors’ English report writing through a corpus-based approach. A corpus of 300 reports (=108,000 words) produced by 100 undergraduate students was compiled across three controlled genres: financial analysis reports, internal memos, and audit-style findings notes. Errors were manually annotated using a taxonomy covering grammar/morphosyntax, lexis and word formation, mechanics, and discourse/genre features. Data analysis employed normalized error rates per 1,000 words, concordance software for contextual examination, and statistical tests for comparative profiling. Techniques of analysis combined quantitative profiling to measure frequency and distribution of errors, comparative analysis across proficiency bands and task types, qualitative concordance analysis to identify meaning risks, and genre-move analysis to evaluate rhetorical appropriateness. Results revealed grammar errors as most frequent, particularly articles, prepositions, and subject–verb agreement. Lexical and collocational errors posed high risks in financial reporting, such as confusion between revenue vs. profit and cost vs. expense. Genre-related problems were evident in maintaining professional tone and register, especially in memo and audit tasks. Findings suggest that instruction should emphasize phrase-level reporting frames, disciplinary vocabulary, and genre-move awareness to enhance accuracy, communicative competence, and workplace readiness in accounting-specific writing.