Action research, also referred to as teacher research, teacher reflection, or practitioner research (Pappas & Raymond-Tucker, 2011), is the recursive process of teachers identifying a problem, introducing solutions, and documenting and reflecting upon the results (Bailey, 2001). It is argued that engaging in action research increases instructors’ awareness of their teaching practice (Atay, 2007) and encourages “reflective, data-driven instruction” (Pappas & Raymond-Tucker, 2011, p. 4). For me, the decision to engage in action research emerged from identifying a knowledge deficit in both my understanding of written feedback pedagogy and my own feedback practices. Therefore, the goal of action research was to analyze the written feedback I gave students in order to understand my feedback practice more fully and compare it more accurately to expert views of best feedback practice. The questions that guided my research were: How do I comment on student writing? How do my findings compare to expert research? Which feedback practices should be continued and which should be changed? Because action research is a recursive endeavor, I repeated the study a year later to understand if I had indeed changed my feedback practice. In this report, I present the research methodology, results, reflections, and resulting actions of both studies.

RESEARCH DESIGN

To examine my feedback on student writing, I selected a small subset of papers on which I had provided feedback during my first year teaching a university-level reading and writing course. This was an intermediate-level course for matriculated students at a large Midwestern university. In total, 18 papers were selected, which comprised nine sets of first and final drafts—three sets of papers from three similar assignments. All nine sets received final grades between 75 and 85 percent. This grade range was chosen because the median and mode score of each assignment fell into this grade range; I therefore considered these papers to be representative of the average paper. As teacher-researchers are advised by Richards and Lockhart (1996) to start with small research projects, I used this limited grade range as one way to reduce the scope of the research project. In addition, I examined only papers for which I had both the first and final drafts.

After compiling the papers to be analyzed, I then conducted a preliminary analysis of the comments I made in order to inductively select relevant categories to research (see Connors & Lunsford, 1993; Reid, 1990). The relevant categories that emerged from this process related to grammatical style (questions, statements, or commands); content (organization, content, vocabulary, meaning clarification, grammar, or praise); and error identification method (whether or not grammar errors were pointed out or corrected). I also noted which comments were made with a mitigation strategy (see definition and discussion below). The final step was to analyze every comment on all 18 papers according to these categories. For each comment I recorded which categories applied to that comment so the content of both the comments and categories could be examined in order to understand features of my feedback.

RESULTS AND ANALYSIS

After determining the categories and features of each comment and compiling the frequency of each comment type, I analyzed these findings according to second language writing research and teacher-training resources; I also reflected on the data in relation to my own teaching goals and teaching context. This process highlighted both positive and negative features of my feedback.

Positive features included the comment variety and quantity. Ferris and Hedgcock (2005) encouraged instructors to provide holistic feedback rather than limiting feedback to just one issue. The fact that each category examined in the study accounted for between 11 and 26 percent of the total comments shows that I did not greatly favor one issue but instead gave feedback on various components of writing including grammar, organization, content, and vocabulary. Another positive feature found in the analysis was the number of comments. I averaged 15 comments on each paper, which is similar to the number of comments given by other teachers (see Ferris, Pezone, Tade, & Tinti, 1997; Treglia, 2008). Apart from being similar in quantity to other instructors’ feedback, the amount seemed appropriate for this teaching context. Fifteen comments on papers of 450 to 800 words seemed like an amount that could provide revision guidance and writing instruction to the intermediate students but would not be too overwhelming. In sum, I felt positive about the variety of issues addressed and the quantity of comments and planned to continue these features of my feedback.

Other results, namely the errors on which I focused and the manner of dealing with grammatical errors, warranted change. Multiple authors argue that correcting mistakes does not improve students’ grammatical competence; instead they advocate using other methods that communicate the existence of an error without providing a correction (Ferris, 2006; Lalande, 1982). Methods of doing this include underlining, highlighting, or using a code (e.g., v.t. for verb tense error). My research revealed that I corrected 66 percent of grammatical errors. I therefore determined to discontinue this practice in favor of pointing out the errors with a code that denoted the type of grammatical error without providing the correct answer.

In addition to changing my manner of dealing with grammatical errors, another concern was the types of grammatical errors on which I focused. The research results revealed that approximately 15 percent of my grammar comments focused on prepositions and articles. Although 15 percent may not be a large percentage, for intermediate students, who are still making errors that impede meaning comprehension, article and preposition mistakes, which rarely affect meaning, should not be a focus. Instead, using the findings and advice of Ferris (2006) to focus error treatment on a small set of grammatical forms and Hinkel (2004) to focus on errors that content professors find egregious, I now focus on five types of grammatical errors: word form, word meaning, verb tense, subject-verb agreement, and word order. Examples of each error can be seen in Table 1.

Each of these errors is marked with a code that is explained to the students. For example, a word form error is marked “w.f.” Most important, I now focus on these errors in class throughout the semester so that the students not only see these errors on their papers but also learn about why these errors are important, how to find them while editing, and how to correct them.

The third change involved adjusting my use of mitigated critiques. Mitigation is frequently used in written feedback in many U.S. educational contexts to curb students’ negative reactions to criticism. Mitigated feedback often includes pairing criticism with praise (e.g., This is good information, but you could explain it more) or pairing a criticism with a suggestion (e.g., The topic sentence is not clear. Since the paragraph is about cultural change, you could start the paragraph by introducing and defining it), and using questions instead of direct criticism (e.g., Can you say this more directly?). Although both praise and critiques (any form of negative feedback) can be mitigated, in my research I focused on the mitigation of critiques. Hyland and Hyland (2001) found that mitigated comments, which are often grammatically and pragmatically difficult, frequently confuse low-level university English language learners. However, Treglia (2008) found that both native and nonnative intermediate-level university students felt mitigation was a valuable way to reduce the sting of criticism. Key differences in the amount of mitigation—teachers in Hyland and Hyland’s study mitigated just over 75 percent of criticisms whereas in Treglia’s study the teachers used mitigation strategies in just 17 percent of criticisms—and the students’ ability level in each study may account for the divergent findings. For instructors, these findings can be reconciled by viewing mitigation as a valuable yet potentially confusing form of feedback, especially for low-level students.

From these results, I developed a heuristic to guide when and how much to utilize mitigation strategies: Mitigation should not be used for highly important comments. What is “highly important” will, of course, vary based on the teaching context but may include issues of plagiarism, thesis statements, or topic sentences. One way I define “highly important” is by thinking about how I will grade the final draft. If a student, for example, would receive a significant grade reduction for failing to understand that he plagiarized part of the paper, mitigation should not be used. A student may not be able to interpret a comment such as “Are some parts of this paragraph copied from an online source?” to mean that the teacher knows that part of the paragraph was copied and that the student should paraphrase or use quotations on the final draft. Instead, a more direct comment such as this should be used: “Part of this paragraph is copied from an online source. Copying directly from a source is not permitted; please paraphrase these ideas as we learned in class last week.” However, for other issues that are less important, mitigation could be used. Some of these less important issues may include adding examples, further explanation, or colorful description. For example, if an idea is pretty clear but would be even better with more description, an appropriate comment may be “This is a really interesting idea. However, a bit more explanation would be useful.” This style of commenting should reduce potential confusion regarding important issues, but will allow for some mitigation to be used because students seem to appreciate this style of comments.

STUDY REPLICATION

After a year, I replicated the study to determine if I had enacted these proposed changes. Table 2 shows that I did accomplish some of my goals.

I no longer corrected grammatical errors or commented on articles and prepositions and I reduced mitigation use. One caveat is that, although I did mitigate fewer critiques (39% instead of 69.5%), it was difficult to determine if I used mitigation only on very important issues. I had some ideas about how this heuristic applied to my teaching situation as explained above, but I still found it difficult to analyze that heuristic in a way that was consistent. When I retrospectively analyzed my comments, it was difficult to take into account all the aspects of the decision to use mitigation and systematically determine which comments addressed very important issues. This second study reveals that continual refinement of my understanding and use of mitigation is needed. As I have continued to review my research findings, monitor my ongoing feedback practice, and read other research studies, I have also begun to realize that potentially not all mitigation strategies are equally confusing. Providing a suggestion may not cause the same confusion as making a critique in the form of a question. Therefore I need to research and understand the effects of mitigation strategies separately as I continue to refine my use and understanding of mitigated comments.

CONCLUSION

The benefits of this research have been my improved understanding of writing feedback literature, the confidence I have in my work, and the hopefully increased clarity of my comments. Despite these benefits, the study was not without limitations and difficulties. Some of the specific difficulties included establishing relevant categories for investigation, defining comment category completely, and consistently applying the definitions. As a novice researcher, I felt overwhelmed at times. Limitations included the fact that the actual data is not widely generalizable because of the small data set. However, given the classroom focus of this research, sound research methodology and generalizability were goals, but not the ultimate goals; the ultimate goals were to gain understanding of my feedback practices, acknowledge and question my assumptions, and engage fully in my teaching practice. For instructors thinking about embarking on an action research voyage, one way to have an even richer experience while also mitigating some of these difficulties is to research collaboratively (Allwright, 2005). This enables deeper and more thorough discussions about methodology, results, and analysis. Although researching alone or collaboratively is not easy, it is worth the effort. The heart of teacher research is that it fosters learning—a process of “knowing and experimenting, reflection and change” (Reid, 1993, p. 257). It is in this process that great personal and professional reward is found.

AUTHOR NOTE

This article is adapted from a longer version published in TESOL Journal. See: Best, K. (2011). Transformation through research-based reflection: A self-study of written feedback practice. TESOL Journal 2(4), 492-509. doi:http://dx.doi.org/10.5054/tj.2010.271901

REFERENCES

Allwright, D. (2005). Developing principles for practitioner research: The case for exploratory practice. Modern Language Journal, 89, 353–366. doi:10.1111/j.1540-4781.2005.00310.x

Atay, D. (2007). Teacher research for professional development.ELT Journal, 62, 139-147. doi:10.1093/elt/ccl053

Bailey, K. (2001). Action research, teacher research, and classroom research in language teaching. In M. Celce-Murcia (Ed.), Teaching English as a second or foreign language (3rd ed., pp. 489–498). Boston, MA: Heinle & Heinle.

Connors, R. J., & Lunsford, A. A. (1993). Teachers’ rhetorical comments on students’ papers. College Composition and Communication, 44, 200-223.

Ferris, D. (2006). Does error feedback help student writers? New evidence on the short and long-term effects of written error correction. In K. Hyland & F. Hyland (Eds.), Feedback in second language writing: Contexts and issues (pp. 81-104). New York, NY: Cambridge University Press.

Ferris, D. R., & Hedgcock, J. S. (2005). Teaching ESL composition: Purpose, process, and practice (2nd ed.). Mahwah, NJ: Lawrence Erlbaum.

Ferris, D. R., Pezone, S., Tade, C. R., & Tinti, S. (1997). Teacher commentary on student writing: Descriptions and implication. Journal of Second Language Writing, 6), 155-182.

Hinkel, E. (2004). Teaching academic ESL writing: Practical techniques in vocabulary and grammar. Mahwah, NJ: Lawrence Erlbaum Associates.

Hyland, F., & Hyland, K. (2001). Sugaring the pill: Praise and criticism in written feedback. Journal of Second Language Writing, 10, 185-212.

Lalande, J. (1982). Reducing composition errors: An experiment.The Modern Language Journal, 66, 140-149.

Pappas, C. C., & Raymond-Tucker, E. (2011). Becoming a teacher researcher in literacy teaching and learning: Strategies and tools for the inquiry process. New York, NY: Taylor & Francis.

Reid, J. (1990). The dirty laundry of ESL survey research. TESOL Quarterly, 24, 323.

Reid, J. (1993). Teaching ESL writing. Englewood Cliffs, NJ: Prentice Hall Regents.

Richards, J. C., & Lockhart, C. (1996). Reflective teaching in second language classrooms. New York, NY: Cambridge University Press.

Treglia, M. (2008). Feedback on feedback: Exploring student responses to teachers’ written commentary. Journal of Basic Writing, 27, 105-137.

Karen Best teaches English as a second language at the University of Wisconsin-Madison. She holds an MA in applied English linguistics from the same institution. Her research interests include writing feedback, student views of feedback, and teacher action research.