This research examined the effectiveness of direct and indirect written CF on improving the grammatical accuracy of Omani EFL students regarding two newly-learned linguistic structures: the comparative and prepositions of space. The research employed mixed methods whereby a quasi-experiment and think aloud protocol (TAP) were used to answer various questions about written CF. In the quasi-experiment, the participants were assigned to a control group and two treatment groups: 1) one group received direct corrections written above their errors and 2) one group received the underlining of errors only. Since the linguistic structures had not been previously introduced to the participants of this study, the researcher provided the students with an instruction lesson on them, a week prior to the data collection. A week later, the students received a pre-test and subsequent revision, where they were asked to revise their initial task. Three days after the revision, the students performed a new task focused on the same linguistic structure (immediate post-test) to measure the short-term learning effect of the written CF. The delayed post-tests were administered six weeks after the pre-test in order to measure the long-term effect of the treatment. In all of the tests, the students were required to describe pictures. The findings of the quasi-experiment show that the direct and indirect written CF improved the grammatical accuracy of the students during revision for both linguistic structures, but a significant effect was found for direct written CF only. The improved accuracy during revision for both the direct and indirect written CF groups was sustained in the new writing task (immediate post-test) regarding the comparative but not prepositions of space. The improvement of the indirect group was even better than that of the direct CF group on the new task (immediate post-test) with regard to the comparative. This might suggest that the indirect CF group processed the feedback in greater depth. No long-term effect was found for direct and indirect written CF for either linguistic structure. The quasi-experiment findings suggest that written CF had a short-term effect when targeting not only already-learned linguistic structures (as the majority of the previous research found) but also newly-learned linguistic structures as well. While many studies have examined the effectiveness of written CF on improving the grammatical accuracy of learners during revision and in new writing using quasi-experiment designs, little research has been conducted to explore how students engage with and process this feedback. In the current study, a think aloud protocol (TAP) was used to explore how the students processed and repaired their errors in response to both direct and indirect written CF. The findings show that both the direct and indirect written CF groups generated a similar amount of repair with understanding during their subsequent revision. About 35% of the direct group’s repair was without understanding. This result suggests that not all of the feedback that was noticed and incorporated into the students’ subsequent revision might be understood. Furthermore, the retrospective TAP produced some data that helped to identify some possible reasons why some students repeated their errors and failed to incorporate the written CF into their subsequent revision. One of the most important findings of this research is that certain factors, such as the type of error and the proficiency level of the student, were found to impact on the effectiveness of written CF. Moreover, the combination of a quasi-experiment and TAP in this research is unique, as it helped to understand the written CF from different perspectives.

This research examined the effectiveness of direct and indirect written CF on improving the grammatical accuracy of Omani EFL students regarding two newly-learned linguistic structures: the comparative and prepositions of space. The research employed mixed methods whereby a quasi-experiment and thin...