Use of corrective feedback in efl classes

This study aims to explore the correlation between students? errors and teachers? use of corrective feedback at tertiary level EFL speaking classes. The frequency and distribution of error types together with the distribution and frequency of corrective feedback types are identified in the present study. Teachers? tendency of employing the different types of corrective feedback after students? different types of errors made in speaking classes is aimed to be revealed, first, to provide a picture whether these tendencies match teachers? teaching purposes in terms of facilitating successful self repair. The subsequent concern is whether the type of error determines the type and frequency of corrective feedback teachers incorporate in speaking classes. The database consists of 24 class hours of interactions between three EFL instructors and their students in three different intermediate level speaking & listening classes at the school of foreign languages at Anadolu University, Turkey. The verbal interaction was videotaped and transcribed, and coded according to Lyster and Ranta?s (1997) corrective ii discourse model. The results reveal that (1) recast (56.2 %), an implicit corrective feedback move that repeats the learners utterance by correcting only the inaccurate item in an unobtrusive manner, was employed the most frequent by all the participating teachers in the present study. Followed by translation (26.2 %), similar to recast which translates students use of L1 into the target language, and explicit correction (9.5 %) which corrects the ill formed utterance obtrusively. Other corrective feedback types such as, metalinguistic feedback, elicitation, clarification request, repetition, that promote `negotiation of form?, proposed by Lyster & Ranta (1997) accounted for 8.1% within all feedback provided. (2) The correlation between corrective feedback type and its preceding error type indicates that although grammatical errors (342 grammatical errors) were the most occurring error type, they were the least (14%) corrected. In contrast, lexical errors that occurred the least (99 times) were corrected as the most frequent error type (43.4%). This finding might be interpreted that the tendency of all instructors was less to correct grammatical errors but the most frequent lexical errors since fluency (i.e. meaning) is the primary focus of speaking classes. Similarly the total amount of corrective feedback (210 turns) provided to the total amount of errors (856 turns) reveals that speaking teachers are unwilling to correct students? errors in speaking classes where fluency is the primary focus. These findings attest to the assumption that speaking teachers are unwilling to employ corrective feedbacks (24.5%) in general and the least to employ prompting (i.e. negotiation of form) types of corrective feedback in speaking classes, which might lead to self repair. However when they do employ corrective feedback, they have a tendency to employ recast (i.e. negotiation of meaning) due to the orientation of fluency that is the focus of communicative classes.