Abstract
Reflection has often been considered a powerful tool for students in composition, helping them develop rhetorical awareness and the ability to transfer their knowledge to future writing tasks. However, the methods that promote reflection have often been debated, and students have considered the process both puzzling and difficult. Furthermore, few studies have directly compared the reflections of native (L1) and non-native (L2) speakers to determine if they are likely to reflect in the same ways and if methods to improve reflective writing would be effective for both groups. To assist in this endeavour, the current study builds on work by Krol (1996) and Kathpalia and Heah (2008) to categorize student reflections and the developments they select to discuss in the small-group tutorial program for L1 and L2 speakers at a large, research university. The aim of this work is to better understand how and to what extent these students are reflecting to help guide the program and offer suggestions that could be more widely applied to peer review groups and composition classrooms in general. This study also indicates future areas of inquiry, including analysing how the number of guiding questions provided influences reflection, assessing student reflections before and after modelling, and comparing the reflections of L2 students with different native languages.
Introduction
Ideally, reflection and self-assessment are tasks that enhance student writing by interrupting ‘autopilot’ behaviour, assisting students in recognizing standards, and increasing agency and motivation (Hilgers et al, 2000). Thus, many studies have suggested that reflection facilitates knowledge transfer and decision-making, allowing students to prepare for professional careers. Pedagogical strategies that value reflective practice and link self-assessment and evaluation, such as portfolios, have been brought into many composition courses, and Kathleen Yancey (1998), author of Reflection in the Writing Classroom, expounded its transformative nature for both students and teachers. In the context of writing centres, reflection is often considered an integral part of practice where tutoring can ‘provide a space where students see writing and reading as reflective processes’ (Mullin, 1994: 181). In this framework, students can develop rhetorical awareness instead of simply mimicking conventions. However, the definition of reflection, when and how it occurs, and the best methods for promoting such work have been debated, making reflection difficult to recognize and assess (Bell et al., 2011).
Because of such ambiguity, both instructors and students at all stages may view reflection as both mysterious and challenging, creating hurdles for those aiming to integrate reflective practice into their curriculum. When given a reflective journal assignment, even graduate students handed in their work with an optimistic trepidation: ‘I hope this is what you want’ (Spalding and Wilson, 2002: 1393). According to Boud, Keogh, and Walker (1985), this uncertainty can lead to deficits in production. For example, students may be unaware of the difference between telling and reflecting in their writing or offer distinct interpretations of a prompt that diverge from its intended meaning (Ruth and Murphy, 1984; Spalding and Wilson, 2002). Thus, following researchers such as Wear and her colleagues (2012), who have called for a better understanding of reflection and the techniques that promote it, I examine student responses to a reflective prompt provided at the end of a small-group tutorial program in the Writing Centre. This analysis builds on previous work by explicitly examining how multilingual and native speakers respond to a reflective prompt, testing assumptions about L2 reflective writing and offering ideas to improve future practice. The fact that all students were given the same assignment offers a valuable opportunity to better understand how language influences reflection.
Reflection and Multilingual Students
This study is important because universities aim to equally serve multilingual students as well as native speakers; to do so they must consider how multilingual users reflect and if reflective assignments are serving their needs. Research examining reflective writing in the English as a Second Language (ESL) composition classroom has largely focused on the role reflection plays in preparing pre-service teachers for their careers. While valuable, some researchers cautioned against uncritically accepting reflective practice, as it led to justifying current methods or failed to help teachers, especially novice ones, recognize problems in the classroom (Ramin, 2007). Therefore, a more comprehensive consideration of the costs and benefits of reflection is critical for L1 and L2 students, especially given that, in the context of this study, the majority of them are, as freshman and sophomores, new to the university environment. This analysis will not only help determine if reflective writing also shifts to something less useful for students, but also test several assumptions about our understanding of reflection for these two groups.
One such area of contention is whether or not multilingual students will reflect as much in their second language as native speakers. Because interpreting writing tasks in general, and reflection in particular, is difficult even for writers using their first language (Ruth and Murphy, 1988), we might assume that such tasks will be more demanding for L2 users learning how to write reflectively while also grappling with the language, resulting in less reflection. In a study by Roux, Mora, and Tamez (2012), only 3% of the papers analysed from multilingual speakers were considered reflective. One student whose papers were categorized as non-reflective commented that writing in a second language did not allow her to express what she wanted to in her reflection, stating, ‘My mind was focused on how to say things in English and I lost track of what I wanted to say’ (2012: 7). Furthermore, though early work indicated that self-assessment was useful for increasing L2 students’ motivation and sense of responsibility (Brindley, 1989), it also suggested that cultural differences and lack of training could negatively impact the process of reflection (Blanche and Merino, 1989). For example, some international students were not familiar with the Western concept of writing as it related to their opinions or reflection as an academic genre, making reflecting difficult (Thonus, 1993; Roux et al., 2012). On the other hand, reflective journals and other tools for reflective practice have become popular activities in classrooms with L2 students (Dantas-Whitney, 2002), perhaps suggesting that L2 students in a traditional classroom setting eventually have more practice, and thus more proficiency, working in this model than L1 students.
This comparison can also provide insight into attitudes and themes discussed in the reflections, including similarities and differences between L1 and L2 students, as it is important to understand not only if students are reflecting or not but also what they are reflecting on. More specifically, another assumption made about multilingual students is that they are more likely to spend time on micro-level linguistic improvements in their written work than macro-level improvements when compared to native speakers. For instance, researchers have shown that L2 students’ focus might remain on lower order concerns such as grammar, instead of higher order aspects such as organization and persuasiveness (Schoonen et al., 2003). In a study by Silva (1992), many students discussed the differences between writing in their native and second languages when reviewing their work, including an increased attention to spelling, grammar, and word choice. One student remarked that this focus took time away from considering his essay’s organization. Yet, this focus on ‘error-free composing’ is not the best method for improving the writing, despite student perceptions (Leki, 2008). Examining L1 and L2 reflections will provide a greater understanding of the level of development students are focusing on and if L2 students are more likely to be concerned with improvements that may be less beneficial.
Therefore, this study aims to address three research questions:
Are students reflecting in their final responses?
Do the amount and categories of reflection differ significantly between L1 and L2 students?
Do L1 and L2 students choose to reflect on different skills or experiences in the program?
Methods
Defining Reflection
To better assess reflection and its role in student learning, it is important to provide an operational definition of the term and how it functions in a given context (Farrell, 1999). For the purpose of this study, I consider reflection to be an examination of prior practice (Bolton, 2001), where students select what to discuss about their experience and its effects, including positive, negative, and neutral impressions. To differentiate between telling and reflection, I use Krol’s (1996) distinction between pre-reflection, narrating what occurred or how the student felt, and reflection, a demonstrated awareness of those thoughts and feelings, including how they may connect to other experiences or be used in the future. Drawing on the metacognitive definitions of declarative, procedural, and conditional knowledge, markers of reflection may involve a change of tense (e.g. I thought X, but now I think Y) or a consideration of the strategies used in the writing process (see Table 1 for student examples). I chose this definition because it helps empirically characterize written reflective work. While I agree that reflection should be performed carefully, leading to development and transformative practice, these outcomes have been excluded due to their inability to be measured with the data presented here.
Explanations and Examples of the Pre-reflective and Reflective Modes Coded for in Student Responses.
Background and Participants
Participants were 15 L1 and 15 L2 students from a large, research institution in the Pacific Northwest in the fall semester of 2014. In that year, the university accepted approximately 4,000 freshman and transfer students, putting overall enrollment around 20,000 students. Of the 4,000 incoming students, 3% were international, meaning that a little over 1,000 international undergraduate students were attending the university. The largest populations were from China (393 students), Saudi Arabia (94 students), and Oman (89 students). Minority students with potential multilingual backgrounds represented 33.8% of the student population. All participants successfully completed the small-group tutorial program, also known as English 102 (for L1 students) or English 107 (for L2 students). This one credit program supplements the first-year composition course at the university, providing students with weekly guided peer review sessions of their written work for the course with a trained tutor and up to four peers. Students may be placed into the small-group tutorial program based on their performance on a placement exam or elect to take it; recent in-house reports indicate that approximately 70% of students present in the tutorial sessions are required to be there. The course is completed on a pass/fail basis that accounts for attendance and the completion of two one-page written responses: a mid-semester evaluation of their progress in the program and a final reflective essay. Specifically, the final reflective essay asks students to write about their experiences in the small group tutorial and provides a series of optional guiding questions for response (Appendix 1). Examining these final reflective essays is important because of the growing size of the program—almost 450 students enrolled in Spring 2013 compared to 330 in 2008—as well as the growing pressure on writing centres to produce quantifiable evidence addressing programmatic outcomes (Ernest et al., 2011), and the potential to apply these results to other contexts (e.g. reflective assignments in first-year composition).
Coding and Analysis
Six 107 tutorials and six 102 tutorials were randomly chosen for analysis, and two to four student essays were analysed per tutorial based on group size. Reflection was coded for using the modified scheme of Krol (1996). Her scheme was chosen because her work examined reflective journal entries most similar to the assignment analysed here, and her coding allowed for an examination of different modes within a piece of writing. Other schemes, while useful under different circumstances, were either too fine-grained to be effective here or used the highest level of reflection from a hierarchy to define an entire piece. This whole-paper assessment does not allow for the linking of questions and reflection or analysing patterns, important aspects of this study, and other researchers have also suggested that the spectrum of reflection it ignored was problematic (Bell et al., 2011). While I maintain the categories of pre-reflection and reflection Krol provides (Table 1), I elaborate on definitions to include specifics relevant to final responses, and I do not conserve the hierarchal distinction between quasi-reflection (meta-cognitive and meta-affective modes) and substantive reflection (connective and pro-reflective modes), as I see all four modes as equally valuable for students, and reflecting on thoughts and feelings was an important outcome of the assignment (Annie Ernest, personal communication, 31 March 2014). This decision aligns with other work that has stressed seeing categories of reflection as an integrated whole rather than a hierarchy (Bell et al., 2011). Responses were coded into these categories at the level of idea units, or units addressing a single topic at the sentence-level. Total word count determined the percentage of student response in each category, and the groups were compared using either the Student’s T-Test (when data were normally distributed) or the Mann Whitney-U Test (when data were not normally distributed). Correlations between reflective categories were examined using Pearson’s Correlation Coefficient.
Final reflections were also scored based on what aspects of the small-group tutorials were discussed based on the categories provided by Kathpalia and Heah (2008): linguistic, cognitive, affective, and social (Table 2). The linguistic category was further broken up into micro-level and macro-level changes, and the affective category was divided into positive outlook, negative outlook, and changed outlook to account for the different responses observed. All reflections were scored based on presence/absence data, though a change in the affective category was also marked as either positive (negative view of course to positive view of course over time) or negative (positive view of course to negative view of course over time).
Explanations and Examples of the Categories Discussed in Student Responses.
Amount and Categories of Reflection in Student Responses
Overall, students dedicated 50% of their responses to pre-reflection and 50% to reflection, proportions similar to students’ final responses about the small-group tutorials in the previous semester (unpublished data). All students dedicated at least a portion of their response to the affective and metacognitive modes, and most students also used the descriptive and meta-affective modes (25 and 23 respectively). On the other hand, only five students used the connective or pro-reflective modes. Thus, students were likely to comment on their feelings and the skills they learned in the program, but were unlikely to comment on the relationship between their experiences and other parts of their lives or their futures (Table 1). These results parallel Krol (1996), who found that students were unlikely to write in those two modes without guidance.
When students were reflecting, they were predominantly working in the metacognitive mode (Fig. 1). The fact that students wrote a considerable amount in this mode is reassuring, as this awareness has been shown to contribute to learning how to learn, which may result in increased performance (Surat et al., 2014). However, overall reflection was still limited to approximately half of a page of writing. Examinations of correlations between the categories indicated that writing in the descriptive mode is negatively correlated with the metacognitive mode (p < 0.01) and the meta-affective mode (p < 0.01). Additionally, writing in the affective mode is negatively correlated with the metacognitive mode (p < 0.001; Figure 2).
Percent of Responses Dedicated to Each of Krol’s Categories of Pre-reflection and Reflection by Fifteen L1 and Fifteen L2 Students.
Negative Correlation between the Percentage of Response Dedicated to the Affective Mode and the Percentage of the Response Dedicated to the Metacognitive Mode Across All Student Responses (n = 30).
While Krol (1996) stated the descriptive and affective modes were necessary to initiate reflection, these correlations suggest that, in the context of a page, writing in these pre-reflective modes does not leave much room for reflection. Furthermore, word count was positively correlated with the connective mode (p = 0.067), and students who included that mode wrote an average of 446 words compared to the 344 word average of all the students together. This evidence suggests that many students may not be writing enough to address later questions that attempt to promote the connective mode; e.g. What are some specific tools or strategies for revising or editing your writing that you will carry beyond this class? Overall, do you believe this class was helpful to you, both as a college student and as a writer?
Comparison of L1 and L2 Student Reflections
L1 students wrote slightly more (368 words) than L2 students (319 words) in their responses; however, the difference between the two groups was not significant. Similarly, the amount of pre-reflection and reflection did not differ between L1 and L2 students except in the meta-affective category, where L1 students were likely to respond more in the meta-affective mode than L2 students (Figure 1; p < 0.01). Almost as many L2 students wrote in the meta-affective mode as L1 students, demonstrating that they were still using this mode. However, L1 students were more likely to elaborate. For instance, an example of the meta-affective portion of an L2 student’s response was: ‘I feel more confident with writing now’ compared to this response by an L1 student: ‘As a freshman, engaging in a small-group conversation about writing was a relief because I recognized I was not the only one struggling with their writing’. Despite the longer responses by L1 students, the meta-affective mode did not constitute more than 11% of their responses on average.
The similarities between L1 and L2 students are encouraging, because they suggest that L2 students are as capable as L1 students of writing reflectively and preliminarily indicate improvements stated here may be effective for both groups. However, it is also important to compare the categories of development that students selected to discuss to strengthen this recommendation. As previous literature suggested, L2 students were more likely to remark on micro-level improvements than L1 students. However, they were also almost as apt to discuss macro-level developments as L1 students (Table 3). Thus, studies that cautioned against peer review groups for L2 students because they focused on grammar and error-finding (Nelson and Carson, 1998), are not supported here. Commenting on sentence-level issues did not stop students from also considering the bigger picture in terms of their writing. For example, one L2 student stated: ‘Last but not least English 107 was a very helpful class to me. It didn’t help only with grammar and sentence structure. It also helped me to improve my critical thinking, which was the weakest part in my writing’. Another student claimed: ‘It [the small-group tutorial] definitely help me a lot in writing and English logical thinking. Now, my teacher praise my essay better than before…some little grammar problems, they also can find them out. It reminds me to not make these stupid mistakes again in next time’. Furthermore, all L1 and L2 students had some metacognitive reflection in their response, indicating that they were also considering cognitive skills that could help them evaluate their work on the macro and micro-level (Table 3). According to one L2 student: ‘I am an audio learner, and whenever I re-read my papers, I try to play the role of someone who has no idea what my paper is about, and, if there are sentences that I have to read multiple times, then I try to remake them’. This technique could help with grammar, but also larger concepts of organization and content based on audience awareness. These comments demonstrate that L2 students can balance working with macro and micro-level linguistic issues, and that guided peer review groups can be useful for non-native speakers in terms of big picture considerations.
Percent of L1 and L2 Students who Mentioned Each Category of Development in Their Response (n = 15 per Group).
L2 students were also more likely than L1 students to comment on the social benefits of the small group tutorial program (Table 3). While other studies have remarked that L2 students were less likely to trust or value peer feedback and instead wanted instructor’s comments (Zhang, 1995), the responses given here often mention the positive influence of peers. For example, one student commented: ‘I can get a lot of suggestion from not only the teacher, but also other students with different thinking and culture’. Another remarked: ‘[Our peer] discussion help me to open my mind and come up some ideas out of the limit of original article’. These responses help demonstrate that knowledge can be cooperatively assembled through dialogues between peers (Cross, 2010) which may lead to gains in ‘language ability, awareness, accuracy, and self-confidence’ (2010: 284). The view of these peer interactions as helpful may have contributed to the overwhelmingly favourable affective response by L2 students about their experiences in the program: 80% of them indicated it was a positive experience, while the other 20% stated that their negative incoming perception was transformed into a positive one, a very similar outlook to L1 students (Table 3). This positive attitude concerning the small-group tutorial program is promising; though the relationship between writing, attitude, and motivation is complex, studies have shown that motivation, self-efficacy, and positivity were linked to stronger academic writing and achievement in a second language. According to a study by Hashemian and Heidari (2012), L2 students who had a positive attitude performed better on a writing proficiency test than students with a negative attitude.
Yet, it is possible that students were simply responding in a way that they thought their audience, the Tutor and the Small-Group Coordinator, would want. Scholars such as Orland-Barak (2007) have indicated that the system of reflection in this type of educational context may lead to false positivity and congeniality, and Barkhuizen (1995) discussed how this inability to be completely candid in a reflective journal read by others was problematic. Even the L1 student who had a negative experience, for instance, remarked at the end that: ‘I never want to go to English 102, but after I do actually go I look back and it is a great experience’. However, two L2 students and one L1 student also mentioned improvements that might be made to the sessions, suggesting that at least some students were considering both productive moments and what could be done better.
Recommendations
One recommendation for increasing the amount of reflection provided would be to give students more feedback and practice with this mode. Revising final reflections in one of the final tutorial sessions would be a method for incorporating guidance in this program. Farrell (2001) suggested that self-reflection can be difficult without input from others, and that helping others reflect can be useful for one’s own reflective practice, indicating that discussions could be useful for students for several reasons. Additionally, students could be asked to do short reflections throughout the semester. For example, the last five minutes of every tutorial session could be dedicated to asking students to write what they learned, what went well, what could be improved, and what they still wanted to know. These responses could be collected and skimmed by tutors to help facilitate future sessions and demonstrate what constitutes reflection; they could subsequently be returned to students at the end of the semester to help them write their final reflections. The fact that this change would not drastically increase time spent by students or tutors is also encouraging, as many students discussed the increased effort reflection requires and its negative effect on motivation (Barkhuizen, 1995).
Modelling by the instructor or the use of previous students’ work could also encourage reflection. In general, this study suggests that the prompt should clearly define reflection, a definition that is reinforced through feedback and examples. However, instructors should also decide what categories of reflection they want to emphasize in their practice, and use tools and techniques to encourage that type of response. Both L1 and L2 students reflected the most in the metacognitive mode in this study; the meta-affective, pro-reflective, and connective modes were all small percentages of the response in their essays. If the goal was to have students write more in the connective mode, concept maps, representing knowledge visually, could be used to help students think about how their experiences intersect. A study by Farrell (2008) showed that students who made concept maps before and after their experience in a course were able to describe how their beliefs had changed, and they seemingly allowed L2 students to discuss their experiences more easily, because they could draw on language from the map. Thus, this practice may especially help multilingual students expand on their ideas in the meta-affective category, the only mode that differed significantly from students writing in their first language. Experimenting with different modalities may influence the categories of reflection received as well (Huang, 2010).
Instructors and tutors should also consider if length requirements are long enough to sufficiently promote the reflection and promote transfer through the connective mode, a goal of many universities. Yancey (1998) spoke of the problem of length in her text as she addressed disappointing reflections in response to the prompt: ‘The first problem seemed simple enough: depth, which you couldn’t obtain without sufficient length’ (1998: 160). In her classroom, students were subsequently asked three questions to address in approximately three pages. This differs considerably from the six questions that students could choose to answer here in one page. Final reflections for 102 tutorials could be lengthened, or, alternatively, students could be specifically asked to select one or two questions from a list of options, so prompts could still allow for student choice and thus motivation and engagement. This possibility may be particularly appealing because researchers have often pointed out the importance of selectivity in producing complex, reflective writing. According to David Walker (1985), a proponent of the reflective portfolio: One of the most important things that a participant has to learn in using this exercise is to be selective. It is better to select a few things to record, and work with them in greater depth, rather than record many things. I felt the inclination of the participants initially was to record as much as possible, and the recognition that they had to record less, and work with it more, was a very important development within them (1985: 62).
By asking students to choose, the prompt would stop students from recording ‘as much as possible’ and provide more detail while still increasing face-validity, students’ ability to write about a topic they have something to say about (Polio and Glew, 1996).
On the other hand, instructors and tutors also need to consider how guiding questions are influencing responses. Because the question that most often prompted the meta-affective mode (i.e. For example, Do you feel more confident in your abilities to recognize issues in your own writing, or that of a peer?), could be answered with a simple yes or no response, it should be reworked if the program aims to obtain a fuller picture of L2 (and even L1) student’s views. This question could also be removed entirely, given that it may have disproportionally shaped student responses. Almost half of the L2 students and six of the 15 L1 students specifically used the word ‘confident’ in their responses, and all of them were now ‘more confident’ in their abilities. One L1 student used ‘confident’ three times, reforming the question as a statement by stating, ‘This course not only made me more confident in my own writing but also a more confident editor for both myself and others’. Other questions from the prompt also were directly reiterated by students; four L2 students mentioned being able to recognize issues or mistakes, and words such as ‘helpful’ and ‘useful’ were used 15 and 18 times respectively. Smith (2000) has noted that students were likely to write what they thought the instructor wanted to hear, stating that the goal of the instructor should not be ‘to engender some ‘correct’ answer or even a general type of response, but rather to encourage the process itself’ (2000: 128). Eliminating that question would not only cut down on responses structured by the prompt, but also help limit the number of questions students might consider in one page. Moreover, these results indicate that the language used in the prompt’s questions need to be carefully considered overall to avoid students simply stating the response the prompt promotes.
Implications and Future Directions
Both L1 and L2 students reflect approximately 50% of the time in their final responses with most of that reflection representing the metacognitive mode. If the Writing Centre wants to help students reflect more, I suggest that weekly smaller reflections could be integrated into the program or a small-group tutorial session could be dedicated to discussing or revising the final responses. Furthermore, I recommend that the length of the reflection and the number of questions asked on the prompt could be reconsidered to promote student choice and more detailed answers. Though this study does indicate that L1 and L2 students chose to reflect on the same developmental areas and in much of the same way, it also reveals that L1 students were likely to use more of the meta-affective mode than L2 students. Thus, L2 students may need more explanation regarding how to reflectively talk about their views and emotions. Because the sample size here was limited and inter-rater reliability was not assessed, more studies are necessary to support these results and produce more detailed recommendations, especially because at the university studied here the L2 student label applies to international students from over 100 different countries as well as domestic multilingual students. Additionally, future work may consider other modes of reflection or how students reflect as they compose these pieces. Overall, this work demonstrates the need for more research and greater interaction between scholars in composition and applied linguistics to make reflection understandable and effective for both L1 and L2 students.
Footnotes
Appendix 1. Copy of the Instructions for the Final Reflection Assignment in the 102 and 107 Small-group Tutorials
Acknowledgements
I would like to thank Dr. Kristin L. Arola, Dr. William Condon, and Dr. Nancy Bell for their invaluable advice and support regarding this manuscript. I would also like to thank Stacy Wittstock, Lisa Johnson-Shull, and the students at the Writing Centre for their assistance and the resources necessary to perform this research.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.