This study investigated whether providing the EFL learners with grammar error feedback would result in the improvement of the accuracy of their writings. As such, two groups of Iranian EFL university students, both males and females, were randomly assigned to a "correct group" (n=31) and a "no correct group" (n=25). They were both required to write four in-class essays and revise them (eight essays, on the whole) throughout an academic semester (a period of four months). The Correct group received the feedback in form of underlining and coding of the errors (based on the category presented by Ferries and Roberts, 2001). The no correct group, however, did not receive any correction. They were just provided with some very implicit and general summary end notes regarding the grammar of their writing, without referring to any explicit error or its location.

The results of the study showed a significant difference between the first and the last essays in each group (both in "correct group" and in "no correct group") in error means, indicating that the students improved the accuracy of their writing from the first to the last essay. In other words, the results showed that both the correct group and the no correct group made fewer errors from the first to the last essay. However, no significant difference was found between the error means of the first essay of the correct and that of the no correct group. The same was true with regard to the last essays of the two groups. On the whole, the results indicated that EFL learners would improve their writing accuracy in the long run, even without receiving precise error feedback through their instructor.