A CORPUS-BASED ERROR ANALYSIS OF WRITTEN ESSAYS BY UZBEK ENGLISH LEARNERS

This study reports a corpus-based error analysis (EA) of 25 short written essays produced by Uzbek learners of English as a foreign language (EFL) under a CEFR-aligned Task 1 prompt (functional writing). Following the EA procedure described in Kutlimuratova’s (2021) corpus-based methodology, all identifiable errors were classified and quantified using a synthesized 13-tag scheme adapted from Divsar and Heydari’s (2017) IELTS learner-corpus coding model. The scheme is theoretically grounded in (i) Chuang and Nesi’s (2006) hierarchical error coding, (ii) Dagneaux et al.’s (1998) computer-aided error analysis approach, and (iii) Hou’s (2016) ten-category system for learner writing. Descriptive statistics show 211 errors in total (M = 8.44 errors per essay, range = 1–16). Verb-related errors (V) were the most frequent (n = 40, 19.0%), followed by article errors (A) and sentence structure errors (SS) (n = 23 each, 10.9%). Spelling (S) and deletion (D) errors were also prominent (n = 19 each, 9.0%), and confusing/unclear statements (CU) accounted for 18 instances (8.5%). Findings are discussed in relation to Uzbek learners’ developing interlanguage and the genre demands of short functional writing, and pedagogical implications are proposed for form-focused instruction, corpus-informed materials, and targeted feedback.