Grammatical Accuracy in L2 Writing: Comparing Different Types of Electronic Corrective Feedback

 
The usefulness of error correction in improving students’ grammatical accuracy has been the focus of attention in the past decades, and hot debates have raged on over this issue. There is yet no clear answer, to date, as to the efficacy of feedback and its various types. The aim of this study is, thus, to shed more light on the relative effectiveness of feedback per se and also the efficiency of some types of grammatical feedback, delivered electronically through MS Word software, over the others in improving students’ written accuracy.
Is there any effect, whatsoever, for different types of feedback (i.e. direct, indication only & indication plus location) when delivered electronically in improving students’ level of grammatical accuracy?
Is there any priority for each of the above feedback types over the others?
 
The participants of this study included 85 Iranian English majors. Of the total participants, 53 were female and 32 were male. The number of females and males were 15 and 11, 13 and 6, 12 and 8 and, 13 and 7 for control group, direct feedback group, indication group and indication and location group respectively.
This study employed a pretest-treatment-posttest format. Of the four groups involved in the study, three were treatment groups and one was the control group. In the first treatment group, direct feedback group, the correct form of the students’ grammatical errors was provided. The two other treatment groups were, however, both provided with indirect feedback. In one of them, the indication-only group, the students were provided with an indication in the margin of the line in which the error was committed to show that an error or errors have occurred. The indication and location group was provided with feedback as to the exact word or phrase in the text that included a grammatical error.
To answer the first research question (i.e. its three sub-questions), three paired sample T-test were used. The second research question (i.e. its three sub-questions) was answered using one-way ANOVA.

Significant difference was found between the pretest and the posttest of the direct feedback group in the mean rate of errors (t=3.475; p‹.05). It means that the provision of direct written corrective feedback has been effective in improving students’ level of accuracy to a statistically significant level. However, the second group of the study, the indication only group, did not show any significant improvement in accuracy from the pretest to the posttest (t=1.627; p›.05). No significant improvement in grammatical accuracy was also observed for the indication and location group.
The primary result of the analysis indicated that there was a significant difference between the four groups with respect to their improvement from the pretest to the posttest (F=6.771; p‹.001). To further investigate the details of this comparison Tukey’s post hoc measure was used. Tukey’s index indicated that there were significant differences between direct feedback group on the one hand and control group (p‹.001), indication only group (p‹.05) and indication and location group (p‹.05). It was, in fact, the direct corrective feedback group whose improvement in accuracy was statistically significantly higher than the other groups. No other significant difference was found between other pairs of the groups with respect to accuracy improvement. An explanation for the findings of the first question can be offered with regard to Schmidt (1990) noticing hypothesis. Among the three types of feedback offered to the groups of the study, direct feedback is apparently the most noticeable. This characteristic may lend this type of feedback to longer retention and quicker internalization. Comparison of the means of improvement for the three treatment groups clearly indicates that the direct group made the most substantial improvement of all. The second substantial improvement is made by indication and location group and the lowest improvement was made by the indication only group. This ranking of improvement is in other words a ranking of noticeability of feedback, and although not consistent with many parts of the literature, is totally consistent with some others especially Schmidt’s noticing hypothesis.
The findings of the second research question can also be interpreted and justified in light of the noticing hypothesis. This point that the direct feedback turned out to make a significant difference and is significantly different from other types of feedback can be justified in light of the fact that the participants of this study were roughly (and not definitely) of the intermediate level of proficiency. For these students, getting involved in problem solving (as the indirect types of feedback requires) while at the same time involved in the quite demanding task of following the ideas might be rather over-demanding, hence distracting their attention. Also, there are many grammatical points which are yet totally unknown to intermediate students and therefore any involvement in problem solving will lead nowhere, no matter how much effort the student makes. This point is also well recognized by Ferris and Roberts (2001) who suggest that direct feedback is perhaps more efficient than indirect corrective feedback with writers of low levels of proficiency.