A Learner Corpus-Based Study on Verb Errors of Turkish EFL Learners.

摘要: As learner corpora have presently become readily accessible, it is practicable to examine interlanguage errors and carry out error analysis (EA) on learner-generated texts. The data available in a corpus enable researchers investigate authentic their respective frequencies terms of types tokens as well contexts which they regularly occur. need consider these the design useful language learning programs remedial teaching materials has been widely emphasized by many (see e.g., Juozulynas, 1994; Mitton, 1996; Cowan, Choi, & Kim, 2003; Ndiaye Vandeventer Faltin, Allerton et al., 2004). This study aims at analyzing inflectional, derivational word form for verbs produced Turkish EFL learners across six distinct proficiency levels, A1-A2; B1-B2; C1-C2, defined Common European Framework Reference Languages (henceforth CEFR) (Council Europe, 2001). used this Cambridge Learner Corpus (CLC), largest annotated test performance corpora enables investigation linguistic rhetorical features performances above stated bands. findings from seem indicate that, different levels registers genres, most common verb categories are incorrect tense (TV), wrong choice (RV), (FV), missing (MV), agreement (AGV) errors. study’s approach uses techniques computer linguistics its roots Error Analysis framework proposed Corder (1971): identification, description, classification explanation