---

The present study is an attempt to investigate whether online dyadic text-based discussion can help language learners develop their oral proficiency due to the fact that classroom interaction barriers have been eliminated and a less threatening, less stressful environment for learner-learner interaction has been created. A total of 30 college freshmen students at the University of Tehran participated in weekly online discussions using the 'Yahoo Messenger' platform for dyadic discussions during six consecutive weeks. The results of participants' oral proficiency test showed that there was a statistically significant difference between the learners' scores in the pre-test and those in the post-test. In addition, the results of both questionnaires and oral proficiency tests employed in the present study showed that online discussion (S-CMC) is more beneficial for the linguistically insecure and shy learners as well as those who had positive attitudes toward online discussion. Moreover, analysis of learners' chat logs regarding the number of turns and number of words per student per session indicated that learners' participation in dyadic online discussion was not equal. This is, in fact, in contrast to what has been reported in previous studies done in this field.

---

This article examines various features of classroom discourse in a communicative EFL classroom. The class was observed and audio-taped during five class sessions with the total recordings of 4 hours of classroom interactions. An analytic framework was developed to examine these features in four major areas of teaching exchanges, characteristics of input, error treatment, and question types. The analysis revealed that the database comprised 52 teaching exchanges, of which 73% contained the F-move with evaluative function, that the teacher modified his speech in accordance with the learners’ language proficiency level, and that there was a clear preference for recasting (51%) and explicit correction (22%), leaving little opportunity for other effective corrective feedback strategies to encourage learner uptake and self-repair. The database was also examined for question types.  Although referential questions are believed to be valuable in promoting communicative interactions, it was found that the teacher asked proportionately more display questions (57%) than referential questions (21%).