THE EFFECT OF MENTOR TEXT MODEL ON IMPROVING EFL LEARNERS’ WRITING FLUENCY AND ACCURACY

Amir Reza Nemat Tabrizi¹

Assistant Professor, Department of English Language, Payame Noor University, Iran

Keywords : Mentor text, Writing, Fluency, Accuracy, Process, Mentor-based Instruction

Article History: Received: 24 March 2017, Revised: 3 May 2017, Accepted: 30 May 2017, Published: 5 June 2017

Contribution/ Originality: This study is one of very few studies which have investigated the effect of Mentor text model on improving English language learners’ writing fluency as well as writing accuracy through an experimental study in EFL context.

1. INTRODUCTION

Writing is an important productive skill. It may be considered as the most important skill that second language students should enhance (Hyland, 2003). It is a means of communication that helps people share ideas, arouse emotions, and defend opinions (White and Arndt, 1991). According to White and Arndt (1991) writing is a thinking process which demands intellectual effort, and it involves generating ideas, planning, goal setting, monitoring, evaluating what is going to be written as well as what has been written, and using language for expressing exact meanings.

On the other hand, language as the chief means of communication is not spoken in a vacuum. Knowing a language is not merely knowing the grammatical rules but also knowing when to say what and to whom, that is knowledge of how the system is put to use in the performing of social actions of different kinds. In this regard, accuracy and fluency are the two factors which determine the success of English language students in communication. It is a general problem faced by language teachers today, whether to focus on accuracy or fluency. Srivastava (2014) clarified that accuracy refers to the ability of the learners to produce grammatically correct sentences. The learner should not only know correct grammatical rules of the language but also able to write accurately, then, he described fluency as a level of proficiency in communication. It is the ability to produce written sentences with ease and efficiency without pauses or a breakdown of communication. Of course, every student wants to be accurate as well as fluent in writing, but there are many variables and kinds of learners and differences in classrooms that make teaching sometimes very challenging. It is a general problem with language teachers that they prefer focusing on grammar activity than on writing activity. They believe that to learn a second language, grammar is the most important thing to learn first.

In addition, in the case of foreign language, English teaching methods differ in different classes; in foreign language contexts the problem is more enormous. In some classes in EFL context the focus is on grammar activities as well as on writing correct sentences because it is the demand of the syllabus which is always accuracy-based. In other classes, teachers focus on fluency. They pay more attention to meaning and context and are less concerned with grammatical errors. But, instructors have to deal with the students of both kinds. However, teachers have to make a balance between accuracy and fluency.

In reality accuracy and fluency are closely related, which lead to the notion that accuracy as well as fluency is necessary for successful communication. Therefore, language teachers should be able to explore along with his/her students not only grammar of forms but also grammar of functions. According to Terjo (2014) a grammar of forms makes us familiar with the grammatical structures and rules designed to show how the systems and subsystems of a language work. But a grammar of function puts together (a) the grammatical structures of a language and (b) how these can be used by a variety of people in a variety of situations for interpersonal and organizational communication. In light of this issue, Srivastava (2014) stated that language should be learned within a social and cultural setting and it cannot be learned in isolation. So, for teaching grammar of forms the teacher should focus on grammatical practice and task-based activities, and for teaching grammar of functions various activities can be adopted. Using a qualified model of writing, mentor-based text can be an option (Hestenes, 2006) in it learners benefit class discussion, mediation of a good model as a guide in writing, collaboration or team working of the students, teacher-student cooperation, and teacher and peer feedback. This approach of teaching is based on socio-affective strategy which emphasizes the interaction, pair/group work, and cooperation with other students in order to reduce the learners’ stress and to learn them how to write a piece of accurate and fluent writing.

Among a variety of techniques that have been used on improving students’ writing skills, just a few of them have been devoted to investigate the effect of mentor text model in comparison with process and product writing on EFL learners’ writing. So, this study tried to fill this gap by conducting the current study and answering the following research questions:

RQ1. Does mentor text model have any effect on fluency of EFL learners’ writing?

RQ2. Does mentor text model have any effect the accuracy of the EFL learners’ writing?

Concerning the questions proposed above, the following hypotheses were developed:

H01. Mentor text model doesn’t have any effect on fluency of EFL learners’ writing.

H02. Mentor text model doesn’t have any effect the accuracy of the EFL learners’ writing.

Accordingly, in the present study, the researcher decides to conduct the study with the EFL university learners from the English B.A. department. The reason is that students from this level have encountered academic writing tasks that require a lot of critical thinking, so they are more conscious of the vital use of writing strategies. In line with this view, the fundamental goal of the current study is to investigate whether manipulating the mentor text have any substantial effects on writing skill in terms of fluency and accuracy of texts produced by EFL learners to enable them to reach a high level of writing quality.

Putting all things together, concerning the significance of the current study, it should be mentioned that collaborative tasks such as brainstorming, shared planning, multiple drafts, peer and teacher feedback and revision are relevant activities within the cycle of writing process. However, with reference to the related literature and the results of the previous studies in this field, it seems that using a newly approach, mentor-based is more influential for EFL learners in that they not only produce a fluent and accurate piece of writing but involve in activities such as collaboration, mediation, feedback, and teacher- student cooperation.

2. LITERATURE REVIEW

To determine the possible benefits of using the writing workshop model with the implementation of mentor texts, effects of the writing workshop approach were explored through students’ growth in writing strategies and improved attitudes. Researchers also examined using writing by ear approach and teaching students to read like writers. Mentor texts were viewed as writing guides, sources of literature that can be used to teach the craft of writing. These texts can be viewed as the “more knowledgeable other” when teaching students to write.

Mentor texts provide students with opportunities to hear various authors’ sounds and styles. The mentor texts also offer students the chance to emulate these writing strategies within their own writing. Furthermore, Culham (2014) emphasizes the importance of teaching students “how to think about how text is constructed” through the use of mentor texts (p. 32).

Powerful writers can develop from an emphasis on writing by ear and teaching students how to think about texts. Focusing on the particular writing traits, which Coe et al. (2011) found, within a mentor text can further develop students’ writing abilities and skills and promote deep writing.

Doorfman and Cappeli (2007) stated that mentor texts can “help writers notice things about an author’s work that is not like anything they might have done before, and empower them to try something new” (p. 3). They (2007) further explained, “Mentor texts help students find ideas and breathe courage into their writing by helping them take risks and think outside their writing box. The mentor is used as a resource and guide for the students when writing. They can find ideas and strategies within the text that can be used in their own writing. The findings of the study showed that mentor texts and the writing workshop helped to foster an awareness in writing among the students as well as create a sense of confidence in their abilities. Students learn best through encouragement, which occurred during student-teacher conferences, writing sessions, and sharing sessions, and through modeling, which happened with the use of mentor texts. The mentor texts and support helped to develop a sense of confidence among the students and allowed them to view themselves as authors. They felt empowered by their new writing knowledge. Over and above, Adams et al. (1996) found a positive aspect of the mentor texts: “The increased amount of time spent sharing original works, both in small and large groups, seemed to have stimulated imagination and enthusiasm within the class” (p. 46). The students were interested in impressing their audience - their peers. When writing, the students were motivated to work to their fullest potential because they wanted to showcase their abilities. This enthusiasm was seen in my study as well. The students were motivated to not only do their best, but to also help their classmates do their best. Each of the students wanted to see their classmates succeed. The sharing aspect of the writing workshop inspired the students to read their favorite and best written selections during this time. The students enjoyed listening to what others wrote and providing feedback on the writing elements and strategies used. Because the students were so supportive and encouraging, the writing community became a place where students could seek help without being judged. The sense of community allowed students to experiment with their writing. Because students felt comfortable with the setting and the ability to try new writing techniques, they were able to focus on their writing in a way that would not be possible if they did not have the support. The use of mentor texts and the writing workshop along with the newly developed writing community allowed the students to just write. They were able to do this on a daily basis which increased their motivation and desire to write.

3. METHODOLOGY

3.1. Participants

A total of 25 EFL participants, out of 120 students, took part in this study; all of them were B.A. students of university in Tabriz-Iran majoring in English Language Translation. They have passed courses in reading comprehension and grammar. No one denies the in/direct influence of these courses on writing by FL learners. During the first week of the semester, the researcher determined the participants’ homogeneity through Nelson proficiency test according to the level of the learners, intermediate level, and 75 participants whose scores fall one SD above and below the mean were selected and finally, among them 25 students were randomly selected as the main participants of the study.

3.2. Instrumentation

The Nelson proficiency test, a writing task as the pretest, a writing task as the posttest to measure the performances of the participants in writing ability were used in this study.

3.2.1. Nelson Proficiency Test

Nelson (350 A) test was used for homogenizing the subjects regarding their proficiency level. This test consists of 50 multiple choice items. The test was implemented for the purpose of homogenizing the sample of the study and to make sure that the study enjoys homogeneous and identical participants with respect to the participants’ English language proficiency. The internal consistency of the Nelson scores gained from the participants in the piloting phase was estimated through using Cronbach's alpha coefficient as 0.87.

3.2.2. Pre-Test and Post-Test of Writing

Learners’ writing performance concerning accuracy and fluency was measured by the pretest. The participants were also invited to write another essay: as the post-test at the last session. They were required to write their composition in about 300-400 words within 40 minutes and the researcher and another colleague rated the writings based on the scoring rubrics offered by Jacobs et al. (1981).

3.2.3. Raters

The researcher himself was the first rater of the learners’ writings. A second rater then scored the total written data from each test (pre- and post-tests). In order to do this the second rater was provided with the guidelines of scoring. The researcher calculated the correlation between the scores as the inter-rater reliability.

3.3. Procedure

During the first week of the semester 120 participants were taken an English proficiency test and a writing pretest. According to the students’ performances on the tests, 25 students were randomly placed Mentor Text group. There was no sex or age limitation in this study.

In the present study, the instructional unit lasted 10 sessions. In the first teaching session, teacher presented a mentor text as a model and students work in collaborative groups to examine additional sections of the text. Their ideas and thinking then were discussed as a whole class. For the next sessions, students read, analyzed, and deconstructed the mentor text under the supervision of the instructor. Each session began with a teacher leading and modeling the analysis of the text for both content and form. During this time, the teacher read sections of the text aloud and periodically stopped and highlighted specific elements of the writing, such as: transitional phrases, important vocabulary words, and statistics that reinforce the author’s argument. The teacher also asked specific questions and made specific comments about the writing. To assess the students’ proficiency level in writing English text, focusing on accuracy and fluency, at the end of the semester the researcher gave a posttest as their final term examination.

3.3.1. Analysis of the Mentor Texts

A piece of writing as a mentor or modeling text was chosen upon which to build a lesson. In this kind of writing, the instructor tries to analyze the text to teach the learners how a qualified writing text is produced. To teach learners how to write an academic text, the instructor should draw learners attention to: a) the interpretation of words and phrases as they are used in text including determining technical, connotative, figurative meanings, and analyzing the text to show how specific word choices shape meaning or tone, b) the formal connecting words in the text, c) the impersonal constructions, d) the complex sentences, e) the transition markers, reported speech, use of ellipsis, f) the style of writing, how point of views or purpose shapes the content and style of the text, g) the structure of writing by analyzing the structure of the text, including how specific sentences, paragraphs, and large portions of the text relate to each other and the whole, h) the effect of the author’s decisions, i) the rules of language such as spelling, grammar/usage, paragraphing, punctuation, j) the fluency of language: the writing flows with a rhythm and cadence, varied sentence length and style, purposeful sentence beginnings, anaphors, linking words / transitional words, Mackenzie (2013). The language performance features the study were investigated include, accuracy, and fluency of language production in the EFL writings. Specific measures used for each area, as well as the tools to obtain the measures, are outlined below.

3.3.2. Fluency

Fluency was measured following the recommendations by Wigglesworth and Storch (2009). It was measured by: (a) total number of words (fluency I); (b) total number of T-units (fluency II); and (c) total number of clauses in each text (fluency III).

3.3.3. Grammatical Accuracy

There are different measures for grammatical accuracy in task-based research. In this study, “to enhance both the validity of the assessments and the comparability of the results” (Ahmadian and Tavakoli, 2011) some of the measures used by Wigglesworth and Storch (2009) applied for measuring grammatical accuracy.

4. RESULTS

In order to select the participants who were homogenous in terms of their language ability, the researcher used a Nelson proficiency test. Then, using the homogenous participants’ scores in writing section of Nelson, the initial homogeneity of the participants in terms of writing was also ensured.

4.1. Using the Nelson Test for Selecting Homogenous Participants

Initially, the piloted Nelson test was administered among 120 individuals in order to enable the researcher to choose the homogenous participants of the study. The descriptive statistics pertinent to the 120 test takers is presented in Table 1. Also, Figure 1 presents the actual shape of the distribution of the scores.

Table-1. Descriptive Statistics of Nelson Scores for the Initial Group

Figure-1. Histogram of the distribution of Nelson scores for the initial group

Based on the values reported in Table 2, the skewness ratio value .068/.209 = .325) fell within the range of -1.96 and +1.96. This point provides support for the normality of distribution for the scores (Tabachnick and Fidell, 2007). Moreover, the minimum score was 3 and the maximum score was 47 Mean = 25.01, SD = 8.14). Following this, in order to select the participants of the study, the researcher selected those individuals whose Nelson scores fell within the range of -1 SD and +1 SD (16.87 to 33.15). Following this procedure resulted in keeping 75 individuals as the homogenous participants of the study. Table 2 presents the descriptive statistics pertinent to the remaining 75 test takers. Also, Figure 2 presents the actual shape of the distribution of the remaining scores.

Table-2. Descriptive Statistics for the Homogenous Participants

Figure-2. Histogram of the distribution for the homogenous participants

4.2. Pre-test of Writing

Table 3 presents the descriptive statistics of the groups’ accuracy and fluency scores.

Table-3. Descriptive Statistics of the Pretest Scores of Writing

4.3. Answering the Research Questions

4.3.1. First Research Question

RQ1. Does mentor text model have any effect on the fluency of EFL learners’ writing?

In order to answer the first research question, a paired-samples t-test was run. Before running the test, the distribution of the difference of the scores obtained from pretest and posttest had to be subject to some tests. Table 4 shows the scores obtained by the participants in the posttest of writing fluency and accuracy.

Table-4. Descriptive Statistics; Posttest of Writing Accuracy and Fluency

Table 5 shows the scores obtained by the participants of the mentor-text group in the pretest and posttest of writing fluency.

Table-5. Descriptive Statistics of Pretest and Posttest Fluency Scores

In order to see if there is any outlier in the distribution of the difference score, the normality plot was inspected (Figure 3). Furthermore, the normality of the distribution of difference scores was checked by inspecting the skewness and kurtosis ratios (Table 6).

Table-6. Descriptive Statistics: the Gain Scores of Writing Fluency

The inspection of figure 3 showed that there was no outlier in the distribution of the difference scores. Moreover, as reported in Table 6, the distribution for difference scores was normal as both skewness ratio and kurtosis ratio values fell within the range of -1.96 and +1.96, supporting the normality of distribution for the scores.

Figure-3. Normality Plot: Checking Outliers

In order to test the first null hypothesis, the writing fluency scores of the participants in the pretest and posttest were compared using a paired samples t-test (Table 7).

Table-7. Paired-Samples T-Test: Gain Scores of Writing Fluency

As it is evident from Table 7, there was no significant difference (t₍₂₉₎ = 1.884, p>.01) in the gain scores of the participants. Therefore, it can be concluded that mentor-text model had no significant effect on EFL learners’ writing fluency; thus, the first null hypothesis was supported.

4.3.2. Second Research Question

RQ2. Does mentor text model have any effect on the accuracy of the EFL learners’ writing?

In order to answer the second research question, a paired-samples t-test was run. Before running the test, the distribution of the difference of the scores obtained from pretest and posttest had to be subject to some tests. Table 8 shows the scores obtained by the participants of the mentor-text group in the pretest and posttest of writing accuracy.

Table-8. Descriptive Statistics of Pretest and Posttest Scores of Accuracy

In order to see if there is any outlier in the distribution of the difference score, the normality plot was inspected (Figure 4). Furthermore, the normality of the distribution of difference scores was checked by inspecting the skewness and kurtosis ratios (Table 9).

Table-9. Descriptive Statistics: the Gain Scores of Writing Accuracy

Figure-4. Normality Plot: Checking Outliers

The inspection of figure 4 showed that there was no outlier in the distribution of the difference scores. Moreover, as reported in Table 10, the distribution for difference scores was normal as both skewness ratio and kurtosis ratio values fell within the range of -1.96 and +1.96, supporting the normality of distribution for the scores (Tabachnick and Fidell, 2007). In order to test the first null hypothesis, the writing accuracy scores of the participants in the pretest and posttest were compared using a paired samples t-test (Table 10).

Table-10. Paired-Samples T-Test: Gain Scores of Writing Accuracy

As it is evident from Table 10, there was a significant difference (t₍₂₉₎ = 3.853, p<.01) in the gain scores of the participants. Therefore, it can be concluded that mentor-text model had significantly positive effect on EFL learners’ writing accuracy; thus, the second null hypothesis was rejected.

5. DISCUSSION AND CONCLUSION

5.1. Discussion

Considering the questions of this study, the obtained results for each of them will be discussed below.

In order to answer the firs research question, a paired-samples t-test was run to test the first null hypothesis, the writing fluency scores of the participants in the pretest and posttest were compared using a paired samples t-test.

As it is evident from the results, there was no significant difference in the gain scores of the participants’ writing fluency. Therefore, it can be concluded that mentor-text model had no significant effect on EFL learners’ writing fluency; thus, the first null hypothesis was supported. Regarding fluency, this study was in contrast with that of Ahmadian and Tavakoli (2011) in that the opportunity to engage simultaneously in careful online planning and task repetition (the least complex tasks) enhanced fluency outstandingly. The results of the present study are in line with Ong and Zhang (2010) study in which the group with the advantage of availability of the drafts during writing (the least complex group) did not produce more fluent language. Coe et al. (2011) claimed that mentor text can further develop students’ writing abilities and skills and promote deep writing. Adams et al. (1996) found a positive aspect of the mentor texts.

In order to test the second null hypothesis, the writing accuracy scores of the participants in the pretest and posttest were compared using a paired samples t-test. As the gain results revealed, there was a significant difference in the gain scores of the participants. Therefore, it can be concluded that mentor-text model had significantly positive effect on EFL learners’ writing accuracy; thus, the second null hypothesis was rejected. The results of the previous studies showed that some of the approaches and strategies have similarities, as explained by Graham and Perin (2007) the use of mentor texts can encourage students to experiment with their writing even further by using the same strategies and techniques that authors’ use in their writing. According to Tsang and Wong (2000) instruction affects student accuracy in the use of the target language in their writing and also the range of choice of structure and vocabulary available to them for use in writing. The results of the study are also in line with, what Coe et al. (2011) found. According their findings a mentor text can further develop students’ writing abilities and skills and promote deep writing. Doorfman and Cappeli (2007) further explained the importance of Mentor texts in fostering awareness in writing. The results of this study showed, working on the least complex task (with the more writing assistance of mentor modeling), learners were able to put together their memory and cognition resources and paid their full attention to the form of their productions which, on the whole, led to more accurate language. The results of measuring the grammatical accuracy (via the ratio of error-free T-units to total T-units (EFT/T) and the ratio of error-free clauses to total clauses (EFC/C)) are supportive results to both Robinson’s Cognition Hypothesis Robinson (2005) and Skehan and Foster (2001) Limited Attentional Capacity Model. They believe that increasing cognitive task complexity (along with the mentor texts) will lead to a decrease in the level of grammatical accuracy. This was the case in this study, that is, the mentor group produced the texts which were more accurate than other two groups who were involved in more complex tasks. Similarly, the findings of this study converge with those of Ahmadian and Tavakoli (2011). They reported that learners who enjoyed more careful on-line planning and task repetition (the least complex task) produced more accurate language than those who performed the task under pressured planning and without task repetition (the most complex task). Contrarily, the results of the current study contrast with those of the study by Kormos (2011) who manipulated the task complexity through the planning time. He found that the two tasks (± planning time) displayed highly similar grammatical accuracy values. His results did not back up Skehan and Foster (2001) Limited Attentional Capacity Model because the two tasks displayed highly similar accuracy scores. Mainly, the findings of the present study support those which have found a significant impact of task complexity on grammatical accuracy of L2 production when increasing task complexity results in less grammatically accurate language.

5.2. Conclusion

In the light of the researchers’ observation and experience in the field of English language teaching, it seems that the little attention is attached to mentor text type in EFL composition classes in EFL settings. Following this, teachers need to pay more attention to learners' composition classes and the quality of learner’s writings. However, the present researcher as the university instructor thinks that applying mentor texts, qualified written texts by experts, as a model will help learners have a better performance. This research, therefore, tries to prove the point that the mentor-based instruction is a method that can enhance the learners’ writing skills and that it can help to support the English communication in the EFL classroom through the interaction between the students, modeling texts, and the instructor. As this instruction increases the teacher-students cooperation, teacher-students feedback, mediation, and collaboration which follows socio-affective strategies in language learning and teaching realm.

REFERENCES

Adams, D., B. Power, M. Reed, P. Reiss and J. Romaniak, 1996. Improving writing skills and related attitudes among elementary school students. Retrieved from http://search.ebscohost.com [Accessed October 18, 2014].

Ahmadian, M.J. and M. Tavakoli, 2011. The effects of simultaneous use of careful online planning and task repetition on accuracy, complexity, and fluency in EFL learners’ production. Language Teaching Research, 15(1): 35-59. View at Google Scholar | View at Publisher

Coe, F., S. Hanita, A. Nishioka and M. Smiely, 2011. An investigation of the impact of the 6+1 trait writing model on grade 5 student writing achievement (NCEE 2012–4010). Washington, DC: National Center for Education Evaluation and Regional Assistance, Institute of Education Sciences, U.S. Department of Education.

Culham, G., 2014. Putting vygotsky to work: The change laboratory as an application of double stimulation. In H. Daniels, M. Cole, & J. Wertsch (Eds.), The Cambridge Companion to Vygotsky. New York: Cambridge University Press.pp: 363-382.

Doorfman, P. and Cappeli, 2007. Introduction. In C. Bazerman & P. Prior (Eds.), What writing does and how it does it: An introduction to analyzing texts and textual practices. Mahwah, NJ: Lawrence Erlbaum. pp: 1-10.

Graham, D. and A. Perin, 2007. Using mentor texts to improve student writing. Created by PCG education for the maine content literacy project october 2012. MCLP Common Core Writing Workshop.

Hestenes, K., 2006. The study of second language acquisition. Oxford: Oxford University Press.

Hyland, K., 2003. Second language writing. NY. Cambridge University Press. Instruction. School Library Monthly, 26(7): 17-19.

Jacobs, J., E. Zinkgarf, A. Wormuth, S. Hartifel and M. Hughey, 1981. Testing ESL composition: A practical approach. Rowley, MA: Newbury House.

Kormos, J., 2011. Task complexity and linguistic and discourse features of narrative writing performance. Journal of Second Language Writing, 20(2): 148-161. View at Google Scholar | View at Publisher

Mackenzie, D., 2013. The emergence of complexity, fluency, and accuracy in the oral and written production of five chinese learners of english. Applied Linguistics, 27(4): 590-619. View at Google Scholar | View at Publisher

Ong, J. and L. Zhang, J., 2010. Effects of task complexity on fluency and lexical complexity in EFL students’ argumentative writing. Journal of Second Language Writing, 19(4): 218-233. View at Google Scholar | View at Publisher

Robinson, P., 2005. Cognitive complexity and task sequencing: A review of studies in a componential framework for second language task design. International Review of Applied Linguistics, 43(1): 1–32. View at Google Scholar | View at Publisher

Skehan, P. and P. Foster, 2001. Cognition and tasks. In P. Robinson (Ed.), Cognition and second language instruction. New York: Cambridge University Press. pp: 183–205.

Srivastava, S.R., 2014. Accuracy vs fluency in english classroom. New Man International Journal of Multidisciplinary Studies, 1(4): 53-58. View at Google Scholar 

Tabachnick, B.G. and L.S. Fidell, 2007. Using multivariate statistics. Boston, MA: Pearson Education, Inc.

Terjo, W., 2014. Sociolinguistic patterns. Philadelphia: University of Pennsylvania Press.

Tsang, R. and A. Wong, 2000. Narrative competence in second language. Educating for intermediate foreign language capacities. Washington. D. C: Georgetown University Press.

White, R. and V. Arndt, 1991. Process writing. London: Longman.

Wigglesworth, P. and K. Storch, 2009. The war between the vowels and the consonants. New York: Farrar Straus Giroux.