Carlo Domingo Casinto<\/strong> Although studies on peer feedback have identified several advantages for second language writing classes, the benefits of peer feedback to low-proficiency students have not been thoroughly investigated. This current study aims to determine how low-proficiency second-language students of English can improve their writing by giving and receiving peer feedback. Utilizing mixed methods, data collection and analyses were primarily conducted through posttest-only control-group design and focus group discussion (FGD). There were twenty university student participants randomly selected from two writing mixed-proficiency level classes. Framed under Vygotsky\u2019s sociocultural theory, an analysis of the gains in writing ability measured from formative and summative writing samples indicated that the low-proficiency students, specifically in the experimental group, made more significant gains in their writing over the course of the semester than the high-proficiency students in both groups. Even though the results generally showed no significant difference, except for higher error avoidance efficiency between experimental and control groups and among the high-proficiency students in both groups, a significant difference was observed between the low-proficiency students in both groups (control group M = 81.6, experimental group M = 89.8, control group SD = 8.3, experimental group SD = 9.5, t(4) = -3.23, p = .032). This leads to the conclusion that low-proficiency students benefit from scaffolded peer feedback.<\/p>\n Keywords<\/em><\/strong>: Scaffolded peer feedback (SPF); Peer review; Peer editing; Second language writing<\/p>\n Writing, as a productive language skill, is one of the most challenging abilities to teach in a second language (L2) class. Peer review, as a teaching approach in writing, enables teachers to give their students more feedback in terms of vocabulary and word expressions, grammar, structure, and content. These are the local and global aspects of writing that most L2 learners struggle with and need more support on. Moreover, peer review affords L2 learners an opportunity for meaningful interaction with other learners, active collaboration with peers such as brainstorming on content and ideas, and critical discussion on the writing itself such as editing word usage, cohesion, and organization.<\/p>\n However, some challenges confront peer review in the writing process. First, there is an attitude of distrust (Alsehibany, 2021). Students perceived peer review as a counterproductive classroom activity, as some peer reviewers are seen as less competent than the writers (Saeli & Cheng, 2021). As reported by Salih (2013), generally peer review tends to be focused on grammar rather than content, ideas, structure, organization, and cohesion. Teachers, on the other hand, perceived it as having no impact on writing quality because most students do not stay on task during the activity (Brammer & Calera, 2007). Because of these challenges, the pedagogical value of peer review in the L2 classroom is questioned by some critics.<\/p>\n Nevertheless, the idea that peer review can be a useful teaching approach in L2 writing is advanced by many studies (Bagheri & Rassaei, 2022; Kim, 2012; Lundstrom & Baker, 2009; Min & Chiu, 2021; Saeli & Cheng, 2021). This is particularly true when students are given sufficient training on how to give scaffolded peer feedback (SPF) (Alnasser, 2018; Alsehibany, 2021; Do, 2020; Leijen, 2017; Sippel, 2021; Zhao, 2018). SPF can be integrated into the writing process promoting a student-centered L2 classroom where learners are empowered to critically evaluate their writing (Brammer & Calera, 2007; Nassaji, 2016; Tajabadi et. al, 2020). SPF sessions can provide L2 learners with a social platform to interact with an audience. Through SPF, L2 learners can see the perspectives of others apart from their own (Zhu, 2001), and negotiate and collaborate on improving their writing (Nguyen, 2021; Tajabadi, et. al, 2020).<\/p>\n Finally, SPF can stimulate L2 students, especially those with low proficiency and those who are not confident with their language abilities, to develop a more positive attitude toward writing as SPF can serve as their classroom social support system in accomplishing the language task (Bolourchi & Soleimani, 2021).<\/p>\n The potential benefits of peer review to low-proficiency L2 students are the dimensions that have rarely been studied empirically in L2 writing research. For instance, the ability to critically evaluate writing and provide effective feedback, especially on a global level such as content and organization as essential writing skills is understudied (Sotouddehnama & Pilehvari, 2016).<\/p>\n One aspect of peer review that should be investigated is how it improves the evaluation skill of L2 learners. Evaluation skill enables L2 students to develop their critical eyes and see logical gaps and problems with cohesion and organization. They are also able to identify other local issues that impact the clarity of their own and their classmates\u2019 writings (Lundstrom & Baker, 2009; Zhao, 2018). In the process, as reported by Hu and Lam (2010), the dyadic and collaborative interactions involving learners regardless of their levels of proficiency during the peer review would lead to developing stronger writing competency and better writing quality because they exchanged mutually beneficial feedback.<\/p>\n This occurrence in second language learning is explicated by Vygotsky\u2019s sociocultural theory which postulates that learners\u2019 ability to obtain and process information is limited to their zone of proximal development (ZPD) or the space between what a learner can do independently and what a learner can do with guidance and assistance (Lantolf & Beckett, 2009). However, with peer review, the learner\u2019s ZPD is expanded, and they can perform a task on their own interdependently with other learners (Hossein & Eva, 2017). Thus, this current study is anchored on Vygotsky\u2019s sociocultural framework which posited that language acquisition is facilitated through learners\u2019 social interaction in which they negotiate and elucidate to confirm, clarify, or recast, and comprehend information mutually given and received (Vygotsky, 1986).<\/p>\n This communicative interaction between learners can provide them with scaffolded support to give feedback especially when it is focused on a particular language form during the negotiation process (Nassaji, 2016). Consequently, this expands learners\u2019 zone of proximal development (ZPD) specifically those with low-proficiency and enables them to improve comprehension and acquire the target language subsequently (Vygotsky, 1986). When ZPD is expanded, learners can also improve their uptake or their ability to respond after the feedback (Nassaji, 2016). This response is a learner\u2019s action to repair or correct the errors in the original output following feedback. Hence, social interaction is the baseline from which learners\u2019 ZPD can be enlarged to clear the way for language acquisition.<\/p>\n Benefits and Limitations of SPF<\/strong><\/p>\n Previous studies elucidate that the benefits of peer review are derived from joint scaffolding or the extra support that learners provide to help each other. For instance, Zheng (2012) demonstrates how two students with different levels of L2 proficiency learn from each other during peer review sessions. In the analysis, the students demonstrate how their dual roles in the peer review process, such as serving as a reviewer and as a writer, enable them to scaffold the learning of their peers. Hence, as asserted by Umaz (2019) and Tajabadi et. al, (2020), it can be contended that even low-proficiency students can scaffold their peers\u2019 learning, and when they can do that, it may benefit them in their writing.<\/p>\n Consequently, Sotouddehnama and Pilehvari (2016) and Kim and Emeliyanova (202) explicated that an L2 student who can critically evaluate others\u2019 writing may develop self-review proficiency, or a learner\u2019s ability to probe into their writing and objectively identify areas that need improvement. This is so because scaffolding requires mainly a collaborative pattern of interaction between the reviewer and the writer (Storch & Suzuki, 2020). The type of feedback generated from a collaborative interaction is the outcome of the negotiation, recast or direct correction, clarification requests, and elicitation prompts (Nassaji, 2016). Zhao (2018) expounded that the negotiation process of a scaffolded peer review facilitates the development of a set of mediating strategies that includes clarification, justification, and confirmation of the feedback given.<\/p>\n Based on the foregoing studies, L2 learners can potentially develop their sociolinguistic competence in the peer review process. When L2 students learn to assess and evaluate their own and their peers\u2019 papers and to identify areas that require improvement, Fatoni and Tauchid (2020) and Nguyen (2021) elucidated that it is a categorical indicator of the positive value that peer review brings to the whole writing process.<\/p>\n Although there has been sufficient empirical evidence pointing towards the positive impact of peer review among low-proficient L2 learners, the study of Kamimura (2006) revealed that peer review activities are more advantageous to highly efficient L2 learners who exhibit greater improvements in the overall quality of their writings. This study, nevertheless, does not contradict the preceding findings as it does not refute the benefits of peer review to low-proficiency students.<\/p>\n In addition, some studies cited the limited sociolinguistic value of peer review. In the studies of Salih (2013) and Saeli and Cheng (2021), critics argued that low-proficiency students tend to focus more on the micro\/local aspects of writing (e.g., grammar, vocabulary, word choice) instead of its macro\/global aspects (e.g., content, organization, cohesion), which would prevent them from developing essential critical evaluation skills. Conversely, the studies of Sanchez-Naranjo (2019), Do (2020), and Bolourchi and Soleimani (2021) explained that these issues confronting the pedagogical value of peer review can be addressed by proper training and sufficient scaffolding.<\/p>\n There are several requisites in developing the ability to scaffold feedback that must be addressed by training. According to Alnasser (2018), Sippel (2021) and En-Chong (2022), these include providing form-focused instructions such as evaluation forms and checklists, and setting up a simulation or a mock review that models the process by which sociolinguistic strategies are employed consistently throughout. Unless these conditions are satisfied, L2 students may not be able to scaffold through their peer feedback since the collaborative pattern of interaction is not established to stimulate negotiation and mediation (Brammer & Calera, 2007; Chang, 2016).<\/p>\n Training, therefore, is indispensable in SPF (Leijen, 2017). In the studies of Zhao (2018) and Tajabadi, et. al,<\/em> (2020) questions emerged, however, as to whether training can foster a collaborative pattern of interaction between peers. In addition. the study of Sato and Lyster (2012) on Japanese EFL learners suggests that training can effectively nurture this kind of interaction during the review process. Remarkable improvements in accuracy and fluency in terms of the ability to recast and provide prompts are observed. In two separate studies of peer review training on L2 learners in Saudi Arabia, an in-class modeling of good peer review practices was implemented in which an instructor provided students a “walk-through” or a class demonstration on how to evaluate a draft using a prepared evaluation form or checklist. Then, a simulation exercise or practice was conducted. It was found that peer feedback had improved its quality and coverage both at the micro\/local level (e.g., grammar, mechanics) and at the macro\/global level (e.g., idea, content, organization) (Alnasser, 2018; Alsehibany, 2021). Moreover, in two separate experimental studies of peer review training on L2 learners in Vietnam and in New Zealand, it was observed that learners achieved a higher total gain score and developed positive attitudes task completion, and ideas development respectively (Do, 2020; Hislop & Stracke, 2017). Hence, this current study intends to incorporate training in its investigation of SPF on low-proficiency students.<\/p>\n This current study investigates whether low-proficiency L2 learners who have been given social support through the scaffolded review process would develop sociolinguistic competence and higher proficiency in L2 writing. Specifically, it will address the following questions:<\/p>\n (RQ1) Which group and proficiency level of L2 students have a higher average gain score and a higher error avoidance efficiency rate in their writing?<\/p>\n (RQ2) If low-proficiency students in the experimental group do improve their writing ability more than those in the control group, which writing aspects (both global and local) do they improve?<\/p>\n (RQ3) Is there a significant difference in the writing mean scores of control and experimental groups and between low-proficiency students and high-proficiency students in both groups?<\/p>\n (RQ4) What are the recurring themes in SPF as perceived and experienced by the students in the experimental group?<\/p>\n (RQ5) What are the implications of these questions to ESL teaching?<\/p>\n Twenty students enrolled in two classes of English Communication for Research, Reading and Writing in the Discipline course offered in semester two of the academic year 2020-2021 of a co-educational university in the Kingdom of Bahrain participated in this study. The number of participants represented a 50% sampling rate of the cohort population. They were chosen based on either of the following criteria: having a CEFR (Common European Framework of Reference for Languages) level at B1 or lower or must have a level at B2 or higher. The other essential criteria for participants in the experimental group were: having attended the SPF training sessions (demonstration and simulation); having reviewed three peer\u2019s formative papers; having received feedback from three formative papers; having attended the post-review conference; and having done three paper revisions.<\/p>\n For this study, B1 or lower level was categorized as a low-proficiency L2 learner while B2 or higher level was categorized as a high-proficiency L2 learner. The language proficiency levels of the participants of this study were generated from their assessment records at the School of Foundation of the university. There were ten participants of mixed levels in each class. To eliminate proficiency bias in each class, five participants had B1 level while the other five had B2 level. After the equal number of mixed-level participants was ascertained, the experimental and the control groups were randomly determined through a lottery of the two EL5002 classes.<\/p>\n Under the Posttest-Only Control-Group Design, the participants who expressed in writing their prior, free and informed consent to participate in the study were randomly assigned and divided into the experimental group (n=<\/em>10) and the control group (n=<\/em>10). All participants in both groups were given alphanumeric codes as their unique identifiers to protect their personal information and to ensure their confidentiality and privacy.<\/p>\n Following strictly COVID-19 safety protocols for face-to-face learning modality, on the second week of the semester, participants in the experimental group were given a series of SPF training sessions. The first session was an orientation and a demonstration on how to use the checklist and the evaluation form that covered global and local aspects of writing. The second session was a simulation or mock exercise in which the participants were immersed in the SPF process.<\/p>\n In the fourth week, participants from both groups started to work on the first of the three formative writing drafts after receiving similar lessons on writing in the preceding weeks. The lessons included how to write the background and the thesis statement in the introduction, comparison and contrast analysis in the body paragraphs, and the CERA (claim, evidence, reason, and action) framework in the conclusion. This is to ensure that all participants in both groups receive the same amount and level of instruction, except for the SPF procedure which only the experimental group had done. Thus, the formative writing activities required that all participants had to listen and take down relevant notes from an audio text at the target course exit C1 level. Based on their notes, they were tasked to write a 300 to 350- word four-paragraph expository essay utilizing mainly persuasive writing methods, namely: comparison and contrast analysis and CERA within a 90-minute time limit.<\/p>\n This was followed by a peer review using the SPF process during the second session for those participants in the experimental group. The participants were purposively paired according to their proficiency level, which requires pairing of B1 level and B2 level participants (Storch & Aldosari, 2012; Zheng, 2012). The first hour of the two-hour SPF session was focused on giving feedback on the global or macro aspects and local or micro aspects of the writing drafts. The subsequent hour was for the peers to meet in a post-review conference for them to negotiate and mediate meanings, to clarify, justify, and confirm content and language, and to elicit and prompt ideas from each other before they act on the feedback for revisions (Nassaji, 2016; Zhao, 2019).<\/p>\n On the other hand, during the second session of the fourth week, the participants in the control group reviewed and revised their papers on their own guided only by the checklist. By depriving the participants in the control group of SPF sessions, this current study was able to investigate the research questions of whether SPF benefits low-proficiency L2 students\u2019 writing.<\/p>\n Writing on two other different topics, the same procedures were also conducted in two separate formative writing sessions conducted on weeks eight and twelve respectively. Subsequently, on the fifteenth week, the summative assessment which served also as the posttest to measure writing proficiency the dependent variable for this study was conducted for both groups.<\/p>\n To guarantee the integrity and reliability of the test results, all submissions by both groups were scanned through a plagiarism checker utilizing software installed as a plugin in the learning management system of the course. Submissions with a similarity rating of 26%-35% are penalized with a 5% deduction from the task fulfillment band score. Moreover, a similarity rating of 36% or higher is outrightly rejected. No submission from the participants was meted with any of the penalties.<\/p>\n In the sixteenth week, a series of focus group discussions (FGD) of the participants in the experimental group was held. To eliminate hesitation among low-proficiency participants and domination of opinion by strong participants, two separate sessions for the group based on their proficiency levels were conducted. Each session lasted for an hour eliciting participants\u2019 experiential opinions based on the perceived strengths and weaknesses of SPF.<\/p>\n The FGD sessions observed the following guidelines: First, the facilitator introduced the FGD and its rationale and objectives; second, the participants were asked to introduce themselves by stating their names and their ID numbers; third, the participants were asked about their SPF experience: during the peer review and the peer review conference. Each participant was given time to speak and the opportunity to agree or disagree with the other participants\u2019 opinions. On the other hand, the facilitator took cues from the explanation of each participant and drew more insights and probing responses. Finally, the facilitator summed up the proceedings and asked if participants had other things to add. After this, the facilitator thanked the participants and closed the FGD. To ensure the validity of the summative assessment or posttest, it had to conform with the learning outcomes, task requirements, time limit, and the exit level of the course. It was similar to the three writing drafts that participants in both groups had written during the formative sessions. As stated, the assessment required that participants had to listen and take down relevant notes from an audio text at the target exit C1 level. Based on their notes, they were tasked to write a 300 to 350- word four-paragraph expository essay utilizing mainly persuasive writing methods, namely: comparison and contrast analysis and CERA within a 90-minute time limit (Please see Appendix A<\/a>). To assess the writing proficiency of the students, the writing rubric was developed by the faculty of English language of the university (Please see Appendix B<\/a>). There were nine criteria in the rubric covering the learning outcomes at the CEFR C1 target exit level. The rubric was used to assess both the local and the global aspects of formative and summative writing assessments. It was based on a five-point scale per criterion, and student essays were assigned a score on the scale on the following writing aspects: task fulfillment (criterion 1); organization, cohesion, and coherence (criterion 2); grammar (criterion 3); CERA structure (criterion 4), vocabulary (criterion 5), tone and register (criterion 6), punctuation and capitalization (criterion 7); spelling (criterion 8); and word count (criterion 9).<\/p>\n At the highest point of the scale, task fulfillment refers to the content coverage that has at least six (6) accurate and relevant points from the listening audio text (three for each body paragraph: one similarity and two differences) and effectively paraphrases this information with similarity rating of 20% or below. Secondly, organization, cohesion, and coherence refer to an effective introduction, the body paragraphs, and recommended course of action; the relationships between ideas and content effectively indicated; a wide range of cohesive devices used effectively; and the variety of referencing. Thirdly, grammar refers to the accurate use of a broad range of appropriate grammatical structures (e.g., passive voice and complex sentences). Fourthly, the CERA structure refers to the relevant claim given and supported with adequate evidence and reasoning in the recommended next course of action paragraph. Fifth, vocabulary refers to a wide range of word usage that is appropriate and accurate. Sixth, tone, and register refer to appropriateness to the audience and purpose consistently, no inappropriate use of slang or idiomatic expressions, and no contractions. Seventh, punctuation and capitalization refer to no mechanical errors in writing. Eight, spelling refers to no spelling error in writing. Finally, word count refers to the number of words in writing by the set limit. The criteria scores on the rubric were then added and converted into percentage score which is the overall writing grade.<\/p>\n Conversely, the error avoidance efficiency rating (EAER), which in theory refers to the learners\u2019 uptake of each participant in both groups, is determined by the local and global errors committed during the series of formative assessments that were coded together and compared with those in the summative assessment. The difference between recurring and corrected errors expressed in percentage scores refers to the error avoidance efficiency rate of each participant. Moreover, the groups (experimental and control) and the clusters (B1 and B2) EAER averages were also ascertained to serve one of the research objectives of this current study (Please see Appendices G and H).<\/p>\n Furthermore, to warrant the validity of the rubric, it underwent pre-moderation and post-moderation. The first was to have senior tenured language faculty members examine the constructive alignment of each criterion in the rubric to the learning outcomes, the assessment task, performance, and the target exit level of the course. The second was to have two of the faculty members, who were teaching the course, pilot test the rubric in the first session of the formative assessment. Ambiguous scaled-criteria descriptors in the rubric were noted, corrected, amended, or revised collectively in a post-moderation meeting. Subsequently, before marking, the assessors conducted a norming session to regulate the interpretation of the writing rubric. In the session, sample essays that were deemed high, average, and low were randomly chosen and rated by each assessor. Then, the assessors compared their scores and discussed the reasons for their ratings until all of them agreed on a given score. This process was repeated for all other criteria in which there were disagreements on their rating scores. The first question of this study inquired on which group and proficiency level of L2 students had a higher average gain score and a higher error avoidance efficiency rate in their writing. To answer the sub-question on the gain score, the average scores in the three formative assessments were compared with the summative scores of both groups which were clustered according to their proficiency levels. Moreover, to answer the sub-question on error avoidance rate, the coded errors identified in the three formative assessments were compared to the coded errors in the summative assessments.<\/p>\n These comparative analyses served two purposes: First to determine how much gain scores were achieved by both groups and their respective clusters, and second to ascertain how efficiently the errors were avoided by both groups and the clusters. The results of these analyses revealed that the mean gain score of the experimental group at 10.91% is higher compared to the mean gain score of the control group at 6.90%. Specifically, the mean gain score of low-proficiency students in the experimental group at 13.81% was higher than the mean gain score of low-proficiency students in the control group which registered at 9.20%. In addition, the mean gain score of high-proficiency students in the experimental group at 6.30% was higher than the mean gain score of high-proficiency students at 4.61% (Please see Appendices E and F).<\/p>\n In the aspect of error avoidance efficiency rate (EAER), the experimental group had 23.78% while the control group had a 13.33% efficiency rating. In terms of proficiency level, the EAER of low-proficiency students in the experimental group at 22.57% was higher than the EAER of low-proficiency students in the control group at 6.66%. Similarly, the EAER of high-proficiency students in the experimental group at 25% was higher than the EAER of high-proficiency students in the control group at 20%. These findings supported the contentions of Do (2020) and Alsehibany (2021) who both asserted that SPF had helped students achieve higher gain scores, more efficient task completion, and better writing quality (Please see Appendices G and H).<\/p>\n The second question of this research study probed whether low-proficiency students in the experimental group did improve their writing ability more than those in the control group and on which writing aspects (both global and local) did they improve. An inventory of the errors avoided revealed that the low-proficiency students were able to improve relatively on content, spelling, vocabulary, lexical semantics, and syntax. Conversely, recurring errors were noted in organization, cohesion, and complex and run-on sentences. By contrast, the low-proficiency students in the control group had a dismal improvement which was limited to spelling and vocabulary or word choice. This explained the wide gap at 15.84% in the EAER of B1 clusters in both groups. These findings confirmed the assertions of Lundstrom & Baker (2009), Fatoni & Tauchid (2020), and Saeli & Cheng (2021) who contended that SPF had supported students, especially those at the lower proficiency level, made significant gains in most aspects of their writing.<\/p>\n Table 1. Significance of the Difference in Writing Proficiency according to Group<\/strong><\/p>\n
\nBahrain Polytechnic, Bahrain
\n<carlo.casinto![\"atmark\"](\"http:\/\/www.tesl-ej.org\/atmark.png\")
polytechnic.bh><\/p>\nAbstract<\/h3>\n
Literature Review<\/h3>\n
Training as a pre-conditioned to an effective SPF<\/h3>\n
Research Problems<\/h3>\n
Methodology<\/h3>\n
Participants<\/h4>\n
Research and Ethical Procedures<\/h4>\n
![](\"https:\/\/tesl-ej.org\/ej104\/a8pix\/image1.png\")
\nFigure 1. Posttest-Only Control-Group Design Flow Chart<\/strong><\/p>\n
\n<\/a><\/p>\nInstruments, Assessors, and Markings<\/h4>\n
\n<\/a>
\nThe posttest was developed and moderated by tenured English language faculty members of the university in semester two of the academic year 2020-2021. Two of the faculty members, who were not involved in moderation, served as assessors. Before marking, the assessors conducted a norming session to standardize the interpretation of the writing rubric. Double-blind average marking was conducted for the twenty papers of participants from both groups. A third assessor from the faculty was required in instances when there was a discrepancy of more than five marks between the first and the second assessors. The average of the three ratings was taken as the writing grade in this case.<\/p>\n
\n<\/a>
\nTo scaffold the peer feedback process, a review checklist and an evaluation form were developed by the proponent of this current study. Both SPF documents were designed according to the assessment criteria in the rubric, and pilot-tested in two phases (Please see Appendix C<\/a>). The first phase of the pilot test was by the tenured faculty members of the English Program of the university. The second phase of the pilot test was by the participants of the experimental group during the simulation session on the second week. Equivocal areas in both documents were identified and revised, Consequently, the final drafts of the two SPF documents were rolled out on the fourth week which was the first of the three formative session-week incorporating SPF as part of the writing process. The checklist prompted participants in the experimental group to examine their peer\u2019s writing and to look for evidence in terms of content, organization, and cohesion requirements to support their feedback. On the other hand, the evaluation form elicited from the participants their qualitative opinions about the quality of writing of their peers and how their peers can improve on their writing. Their feedback was supported by evidence indicated in the checklist. The two forms were the basis of the post-review conference.
\n<\/a>
\nFinally, to facilitate the focus group discussions (FGD) of the participants of the experimental group, an FGD guide was developed by the proponent of this study and validated by the manager of the English program of the university. The guide contained open-ended questions inquiring about the participants\u2019 perception of their experience of the SPF process, namely: during the review and during the conference (Please see Appendix D<\/a>). During the review, the participants were asked about how reviewing their peers\u2019 writings had helped them review and revise their writing. On the other hand, during the conference, the participants were questioned about how discussing their feedback with their peers stimulated them to reflect on their writings but from a different perspective.<\/p>\nResults and Discussion<\/h3>\n