How often have you seen a writing assignment covered in red pen with numerous little symbols and codes highlighting comma splices or subject-verb disagreement. I am guilty of doing this myself (though on Google Docs, not with red pen). Most of us agree that content and organization are primary, but accuracy does play an important role in communication. So, we persist in written error correction. And students seem to persist in written errors. Are we wasting our time? Are students not taking our feedback into consideration? What is going on with written corrective feedback (for grammar) and how can this inform our teaching practice? This article is Part I in a two-part series on error correction. Part I looks at the lack of evidence that proves its effectiveness and why that may be so. Part II will look at a meta-analysis and describe the circumstances under which written corrective feedback is effective.
Ferris, D. R. (2004). The “grammar correction” debate in L2 writing: Where are we, and where do we go from here?(and what do we do in the meantime…?). Journal of Second Language Writing, 13(1), 49-62. http://www.sciencedirect.com/science/article/pii/S1060374304000086
Ferris writes this article prompted by a back-and-forth (in the literature) response to Truscott’s (1996) assertion that “error correction is harmful and should be abolished”. After examining the research on this subject, as well as conducting his own studies, Ferris comes to the following conclusions:
1. The Research on Error Correction Is Inadequate
The best way to test whether error correction is effective is by setting up simple correction/no correction treatment groups. If the group that received correction has more accurate writing, then we can say that error correction is helpful. However, Ferris found (in 2005) that few studies had actually done this. One reason is possibly the ethical dilemma in not providing error correction, which most researchers believe is beneficial. And, according to Ferris, the results of these six studies show positive effects, no effects, and one is inconclusive.
2. Research is Incomparable Because of Design Differences
Research on effective feedback has taken place in a variety of different settings, making it hard to generalize. However, Ferris argues that if research showed the same results under various conditions, that would be an argument for generalization. Yet, this is not the case. The results differ widely. Furthermore, there has been (as of 2005) no replication of previous studied.
3. Research Predicts Positive Effects
Despite the issues with the research, extant evidence certainly predicts (though does not prove) positive effects. Ferris, in summarizing the possible effects of corrective feedback, writes (p. 56):
- Adult acquirers may fossilize and not continue to make progress in accuracy of linguistic forms without explicit instruction and feedback on their errors.
- Students who receive feedback on their written errors will be more likely to self correct them during revision than those who receive no feedback—and this demonstrated uptake may be a necessary step in developing longer term linguistic competence.
- Students are likely to attend to and appreciate feedback on their errors, and this may motivate them both to make corrections and to work harder on improving their writing. The lack of such feedback may lead to anxiety or resentment, which could decrease motivation and lower confidence in their teachers.
Ferris states that, “at minimum it can be said that if the existing longitudinal studies do not reliably demonstrate the efficacy of error feedback, they certainly do not prove its uselessness, either” (p. 55). Here, she begins to focus more on what the research is lacking. This includes not only experimental designs and replication but also new ways to design research that attempts to answer the big questions directly. To avoid ethically dilemmas of withholding potential positive treatment, she gives an example design: two courses, taught by the same instructor. In one, grammar notes are given at the end of the text. In the other, in-text corrective feedback is given. She also argues that questions need to be more specific, looking at the effects of revision after corrective feedback, grammar-instruction combined with feedback, charting errors (e.g. in grammar logs), which types of errors, and how explicit.
So, what do we do?
Ferris claims that, as a research and teacher, intuition vs (lack) of evidence certainly makes the job of writing instruction more difficult, making her ask herself which should she do: stay rigid until evidence shows positive effects or work based on experiences, intuitions, and student desires. She recommends being careful and systematic in providing feedback as well as sensitive to not overwhelming or discouraging students. In addition, she also points out that error correction is not the only approach: consciousness-raising, grammar instruction, practice, accountability, and seeing editing opportunities as problem-solving opportunities.
I certainly understand the contention that Ferris feels. There has been a lack of evidence supporting corrective feedback. Anecdotal evidence reinforces this: how many times do I need to correct the same errors? Where is the uptake. However, without more research, it’s hard to say what is or is not effective. Should we keep trucking along with providing feedback, or should we wait for a complete picture, if ever a thing could be true. In my own opinion, we should provide feedback, but be more strategic about it: targeted at only a few issues, quantitatively less, and always as secondary to working with content, organization, structure, and critical thinking.
Truscott, J. (1996). The case against grammar correction in L2 writing classes. Language Learning, 46, 327–369.