教育研究集刊
第五十九輯第二 期 2013年6月 頁115-151
文類本位寫作教學對大學生學術英語 讀寫能力之影響
鄒文莉、林愛恩 摘要
本文首先探討當前臺灣某國立大學學生的學術英文文體類型認知水平，並衡 量以文類本位寫作教學對於習得論說文體（argumentative）類型的認知與學術英 文寫作品質的影響。86名修讀學術英語（EAP）課程的大二學生隨機編入成控制 與實驗兩組，並在課程前、後分別予以英文文體類型認知及寫作測試。針對上下 文、修辭結構及語彙文法特點間的關聯性，實驗組接受八大關鍵學術文章類型之 訓練課程共七週。透過小組活動，學生們分析並練習寫作不同文體類型。最後，
再針對兩組學生寫作的能力給予評鑑。結果顯示，實驗組的文體類型認知水平於 七週後顯著提升。再者，實驗組在寫作組織度與連貫性方面之得分亦大幅進步。
以文體類型為本的寫作教學針對文體結構、發展與內文連貫性的寫作訓練成效顯 著。因此，本文驗證了以文體類型為本的讀寫訓練課程有助於提升臺灣大學生的 學術英文寫作能力。
關鍵詞：文步、文類本位教學、文體類型、摘要寫作
鄒文莉，國立成功大學外國語文學系教授（通訊作者）
林愛恩，國立成功大學外國語文學系博士班研究生 電子郵件：wtsou@mail.ncku.edu.tw
投稿日期：2012年10月10日；修改日期：2013年2月21日；採用日期：2013年4月23日
Bulletin of Educational Research June, 2013, Vol. 59 No. 2 pp. 115-151
The Effects of Genre-based Writing Instruction on College English Academic
Literacy
Wenli Tsou Vivien Lin A b s t r a c t
This study first investigates the current level of awareness towards genres of academic writing among students at a public university in Taiwan, and then measures the effect of genre-based writing instruction on the students’ genre awareness and the quality of their writing for the argumentative text type. Eighty-six sophomores enrolled in an English for Academic Purposes (EAP) course were randomly divided into control and experimental groups, and completed tests on genre awareness and argumentative writing before and after the genre-based writing instruction courses. Over seven weeks, the experimental group received weekly training sessions on the relationship between text and context, rhetorical structure, and lexico-grammatical features of eight key academic text types. Through small group activities, they analyzed and practiced writing the text types. Finally, both groups were measured on their writing abilities.
Wenli Tsou, Professor, Department of Foreign Languages & Literature, National Cheng Kung University, Correspondence Author
Vivien Lin, PhD Student, Department of Foreign Languages & Literature, National Cheng Kung University
Email: wtsou@mail.ncku.edu.tw
Manuscript received: Oct. 10, 2012; Modified: Feb. 21, 2013; Accepted: Apr. 23, 2013.
Results showed that although both groups displayed a low level of genre awareness, after instruction, the experimental group increased their genre awareness significantly.
Furthermore, the experimental group showed significant improvement in their textual organization and coherence based on analytical scoring. Genre-based training also resulted in significant improvements in text development, structure, and cohesion. The study proved the effectiveness of academic literacy training sessions on improving academic writing at the university level.
Keywords: moves, genre-based instruction, text types, abstract writing
Introduction
Enhancing academic literacy (Lea & Street, 1998) has become an important goal in higher education worldwide (Jones, 2004). With the popularity of English for Academic Purposes (EAP) programs aiming to increase writing competence for international students at British and American universities (Cheng, 2007, 2008a;
Lillis & Scott, 2007), genre-based pedagogy has also gained importance in providing knowledge of the university academic discourse to second/foreign language learners (Hyland & Hamp-Lyons, 2002; Jones, 2004). In the Australian context, genre-based teaching relies on M. Halliday’s Systemic Functional Linguistics (SFL) model focusing on text types to allow students to examine the role of language in the construction of social life (Christie, 1999). In North America, genre-based instruction is used to enhance writing competence for non-native English speakers in university settings (Cheng, 2007, 2008a; Gardner, 2012; Henry & Roseberry, 1998; Jones, 2004). An example in Taiwan is the introduction of the Research Article (RA) genre to Taiwanese graduate students for succeeding in the academic community (Liou, Chang, Kuo, Chen,
& Chang, 2005).
As a branch of English for Specific Purposes (ESP), EAP focused specifically on language research and instruction to meet the communicative needs and practices of learners in academic contexts. It also lays the foundation for learners’ understanding on the cognitive, social and linguistic demands of specific genres in the academic community (Hyland & Hamp-Lyons, 2002). In the last decade, EAP has caught the attention of EFL practitioners in Taiwanese universities at the graduate level (Hsieh
& Liou, 2005; Liou et al., 2005). However, no large-scale EAP program at the undergraduate level has been implemented. In order to help Taiwanese undergraduates produce better academic writing, it is worthwhile to consult the British academic literacy programs. Because university students in the UK specialize in specific disciplines from their first year, Genre-based Writing Instruction (GBWI) has proven
most suitable for them as they learn to write discipline-specific texts (Wingate, 2012).
GBWI thus plays a vital role in improving students’ academic literacy by teaching learners to identify and use linguistic resources to produce academic texts (Gardner, 2012).
Looking at Taiwan, although English-mediated instruction has become widely implemented, most undergraduates are still not equipped with necessary academic literacy skills in English (Min, 2006). It is important to familiarize students with academic genres in order to enhance their understanding of the contexts in which writing takes place (Paltridge et al., 2009). As put forth by Connor (2009), good writing quality is characterized by a fit of a text to its context, including the writer’s purpose, the discourse medium, and audience. It is therefore useful to employ genre- based instruction to make students aware of the privileged forms of the academic discourse and learn about the relationships between a text’s form, rhetorical functions, and community of users (Tardy, 2011). It is believed that teaching academic genres commonly used in students’ undergraduate studies will make them more competent writers in higher education. This research thus bears two purposes: (1) to evaluate the current level of genre awareness for college EAP students in Taiwan; and (2) to examine the effect of genre-based writing instruction on raising genre awareness and enhancing writing quality for the argument text. The results of the academic literacy training provide relevant insights for EFL teachers and researchers who wish to improve students’ academic literacy through genre-based writing instruction.
Literature Review
Traditional approaches in teaching L2 writing
Second language (L2) writing instruction is a complex task which involves teaching non-native English speakers (NNES) from various cultural and linguistic backgrounds whose textual development is greatly affected by their first language
(L1) (Silva, 1993). Moreover, compared to L1 writing, higher error rate and less fluency in L2 writing pose greater challenges for ESL/EFL instructors (Silva, 1993).
While numerous approaches to teaching L2 writing prevail, three common approaches include pattern-centered, content-based and process approaches. To begin, under the pattern-centered approach, students are given a pattern and asked to find topics to produce essays that fit the organizational patterns of academic discourse, including process analysis, comparison/contrast, cause and effect analysis, as well as pro-and- con argument (Shih, 1986). However, the approach was criticized for its focus on form of the final product rather than the pre-writing and rewriting process (Shih, 1986), resulting in a reversal from the normal writing process and lacking meaningfulness in communication in a social context (Zamel & Spack, 1998). Next, the content-based approach employs intensive reading and writing at the beginning of undergraduate study. A content-based English course usually includes selected topical readings as narrow input, followed by academic writing assignments such as summary, personal response, synthesis, and critique/evaluation as a post-reading task (Shih, 1986). In content-based writing instruction, instructors use the discipline content as the core material to improve students’ ability in writing academic texts (Shih, 1986).
Finally, the process approach, which is based on the cognitive process theory of writing (Flower & Hayes, 1986), guides students through the mental recursive process of planning, translating, and reviewing, where thinking was required for the organization of high-level goals and sub-goals. The aim is to allow students to develop ideas with their own expressions rather than controlling them (Silva, 1990). The sequence of
“pre-writing,” “planning,” “drafting,” “revising” and “editing” assumes a learner- centered focus. Some university writing courses in Taiwan have adopted this approach, incorporating multiple drafts through teacher and peer feedback sessions in the writing cycle (Min, 2006). Yet, under the process-writing approach, students still lack the skills needed for facing the complex linguistic and rhetorical expectations of the academy in their university experience (Zamel & Spack, 1998). Furthermore, although process approaches have allowed students to become more independent writers, there is little
evidence that their L2 writing has become significantly better (Hyland, 2003). Since no single approach is perfect, it is important that EFL writing instructors in Taiwan adopt methods that most suit student’ needs from the various approaches in L2 writing.
Academic literacies and genre approaches
Academic literacy, in its singular form, is narrowly defined as the ability to read and write the various assigned texts in college (Spack, 1997). In its plural, academic literacies (Lea & Street, 1998) refers to a critical discourse approach that emerged in the 1970s from new literacies studies to solve problems in the university writing curriculum from cultural and institutional perspectives (Zamel & Spack, 1998). In this framework, teachers explain the academic culture with its norms and practices in order to help students gain access to the academic institution (Paltridge et al., 2009).
Since the approach emphasizes conceptualizing and influencing teaching and learning by considering the cultural and contextual aspects of texts, writing becomes not only a set of acquired skills, but also an awareness of how a text may be used in particular contexts. In action, teachers can conduct a present situation analysis to find out students’ current abilities, familiarity with writing processes and written genres, and their skills and perceptions (Hyland, 2009).
Among the three major approaches to genre, namely the new rhetoric, ESP and SFL (Hyon, 1996), SFL is most widely used for teaching genre with an explicit linguistic approach that allows researchers to assess teaching practices (Hyland, 2009).
Through a top-down analysis of language, the SFL system allows users to choose language features according to the demands of the situation (Perrett, 2000). In relation to EFL writing, SFL teaches learners to treat linguistics as a set of systems consisting of genre, register, and grammar for conveying meanings in different social contexts (Hyland, 2004, 2009). The application of SFL to writing pedagogy as well as the connection between the SFL and genre-based instruction will be discussed in the next section on GBWI.
More recently, EAP practitioners have discussed the possible advantage of
combining the academic literacies approach with SFL for a more holistic model of teaching second language (L2) writing in EAP (Coffin & Donohue, 2012a). While SFL, as a text-in-context approach, employs textual analysis of disciplinary discourses, thereby allowing learners to draw on knowledge of prior texts to appeal to the culture and institution of the academia (Hyland & Hamp-Lyons, 2002), academic literacies adopts a practice-in-context approach to investigate the socio-political processes in which academic texts are situated (Coffin & Donohue, 2012b). The two approaches differ in that SFL focuses on how meaning is constructed and developed through language in a discipline (Gardner, 2012), whereas academic literacies places emphasis on reader-writer identity and power balances in the academy (Coffin & Donohue, 2012b) for understanding university writing practice (Gardner, 2012). It is crucial, when adopting writing pedagogy in university settings, to consider both practice from a wider perspective and text at a discipline-specific level.
Genre-based writing instruction (GBWI)
A genre is a written or spoken text composed of a series of segments called moves that serves a specific purpose in society (Henry & Roseberry, 1998). The development of genre-based approach in language teaching began in the 1980s with the creation of functional grammar by Halliday (1985) who represents the Sydney School (Ahn, 2012), and Swales (1990) who represents the ESP school (Ahn, 2012) with a genre-based approach to college curriculum (Hopkins & Dudley-Evans, 1988) and the analysis on the research article (RA) genre (Henry & Roseberry, 1998). In the ESP school, genre analyses on different sections of the RA were carried out by Hopkins and Dudley- Evans (1988) focusing on the Discussion and by Brett (1994) on the results section.
In the Sydney School, the trend of teaching writing through genre-based approach began in the late 1990s by scholars such as Martin (1999) and Paltridge (1996). Deriving the meta-functional aspect of language from SFL, GBWI combines text analysis with the teaching-learning cycle of modeling, joint construction, and independent construction (Ahn, 2012; Wingate, 2012). The instruction follows explicit
teaching of linguistic features for particular text types with the teacher playing a visible and interventionist role (Ahn, 2012; Martin, 1999). As opposed to traditional EFL teaching where language is taught without strong linkage to context, GBWI allows the teacher to tie lexis, grammar, and discourse structures to genre to make students conscious of text in context (Martin, 2009). In essence, students not only analyze a written text highlighting its specific communicating function, but also learn to use the appropriate lexical, grammatical and discoursal features of each text type to suit their writing needs. These functions thus distinguish GBWI from traditional writing approaches discussed earlier in Section 2.1.
In GBWI, different text types are introduced with their distinctive patterns of vocabulary, grammar, and cohesion to show different kinds of academic texts (Hyland, 2004). Although the terms “genre” and “text type” are often used interchangeably (Lee, 2001), the distinction between the two has been made (Lee, 2001; Paltridge, 1996). While “genre” and “text type” encompass overlapping text categories (Lee, 2001), “genre” commonly refers to the social context and stages that fulfill different communicative purposes in a text, and “text types” are categorized based on the co- occurring linguistic features that form rhetorical patterning in a text (Paltridge, 1996).
The notion of text types was particularly useful when genre analysis was applied to writing pedagogy by scholars such as Derewianka (1991), Martin (1989) and Hammond, Burns, Joyce, Brosnan, and Gerot (1992) who labeled text categories of recounts, narratives, reports, procedures, explanations, arguments, and news articles as text types. In this study, genre-based writing instruction refers to teaching the linguistic features of specific academic text types listed in the next paragraph.
A genre-based writing syllabus may include a list of text types necessary for academic success in students’ own contexts at relevant discourse and language levels (Paltridge et al., 2009). More recently, eight key text types have been identified as recount, information report, explanation, exposition, discussion, procedure, narrative, and news story (Johns, 2011). As each type varies in complexity of form, to successfully engage in text type teaching, teachers must plan a sequence of text types
for writing courses for students to follow a progression of difficulty in their learning (Hyland, 2004). To help students produce a coherent text, a macro-level focus should be placed on the overall organization of the text which is broken into introduction, body, and conclusion. On the micro-level, the linguistic features for each target text type should also be explicitly presented in a genre-based curriculum (Paltridge et al., 2009). For example, writing an argument as the most complex text among the eight key text types may involve using cause-effect conjunctions and clauses to help the writer to show reason and persuade. Writers may use modals such as “should” and emotive expressions such as “I strongly believe...eventually” to reiterate the claim in the thesis statement.
To improve students’ academic literacy in college, teachers can introduce strategies for understanding, discussing, organizing and producing texts in relation to the social context of a text (Johns, 1997). According to Johns (2011), the focus of GBWI could be placed on both genre acquisition and genre awareness. For a clear distinction, “genre acquisition” involves the direct training for the ability to reproduce a specified text type (Johns, 2008; Yasuda, 2011), while “genre awareness” is realized as learners develop the flexibility to adapt to different rhetorical purposes and the broader contexts in which texts may appear (Johns, 2008, 2011). Studies on training genre awareness therefore involve introducing multiple genres to raise students’ awareness of different conventions across various genres (Yasuda, 2011). Instructors can also examine students’ current level of genre knowledge as they conduct needs analysis.
Cheng’s (2007, 2008a, 2008b) studies on genre-based teaching showed that students’
ability to connect genre texts to purpose, writer role, and audience facilitated their rhetorical reading and evaluative reading. Therefore, effective writers must learn how the writer, reader, and purpose interact with one another to make meaning in a text.
Over the last decade, universities in Taiwan (Liou et al., 2005), China (Gao, 2007), and Japan (Yasuda, 2011) have also attempted to use the genre-based approach to teach academic writing and have yielded successful results. Recent studies on the genre-based approach in Taiwan include essay writing, narrative writing, and web-
based instruction for journal article writing. Liou et al. (2005), for example, conducted a 20-month collaborative project online to familiarize graduate students with genres of journal papers via a web-based learning environment. Through corpus analysis of research article sections of introduction, methods, results, discussion, and abstracts, the instructors provided a platform for students to meet the required writing convention of knowledge production in the academic community (Liou et al., 2005). In Japan, Yasuda (2011) used genre-based tasks to engage EFL Japanese undergraduates in email writing and successfully improved students’ genre awareness, linguistic knowledge and writing competence.
Previous genre studies have focused mainly on advanced learners writing in ESP contexts (Yasuda, 2011), with graduate students learning to write their discipline- specific genres. However, these findings cannot be applied to undergraduate learning to write in EFL college settings. Genre training thus needs to reach novice writers in undergraduate programs since it has been argued that teaching with the SFL model may help to create meaningful foreign language education to adult L2 learners from intermediate to advanced levels (Byrnes, 2009). Although Yasuda (2011) has focused on EFL novice writers at the undergraduate level, the teaching was on a non-academic genre and a control group was not available for comparison effect. This study therefore aims to fill previous research gap by focusing on undergraduate academic text types with the inclusion of both experimental and control groups. To explore the current level of genre awareness and facilitate genre acquisition for Taiwanese undergraduates in southern Taiwan, two research questions follow: (1) What is the current level of genre awareness for college EAP students in Taiwan? and (2) Does genre-based teaching lead to heightened genre awareness and better writing quality for the argument text?
Methodology
Context and participants
The study was conducted in an undergraduate EAP program at a national university in southern Taiwan. The program included 250 sophomores from schools of medicine, engineering, bioscience and biotechnology. After an entry TOEIC exam, 86 participants who were in the same score range were chosen among the 250 students.
The 86 participants were then randomly assigned to one control and one experimental group, each consisting of 43 students. Independent Sample T-tests of the students’
TOEIC Reading scores (M = 253 for Experimental Group, M = 254 for Control Group) showed no significant difference between the two groups, t (84) = 0.048, p = .962. One EAP instructor taught both groups using the textbook “English for Bio-technology”
designed to engage students in reading passages related to biotechnology. The course syllabus included both reading and writing. The reading component focused on learning the target vocabulary and grammar for both groups. The writing component differed between the two groups as the control group received content-based writing instruction and the experimental group received genre-based teaching.
Treatment
The treatment involved a GBWI syllabus designed for the experimental group.
In Week 1, both groups received a pre-test on the argument text type. During training sessions from Week 2 to Week 5, the experimental group received 2-hour weekly teaching sessions where GBWI was incorporated into the syllabus with explicit teaching, modeling, and practice on text in context and academic text types. With power point visual aids, the instructor first introduced the concept of using text in context with news story, narrative, procedure and report. Then she modeled the eight academic text types in sequence from recount, narrative, procedure, description, report, explanation,
discussion to exposition/argument. The modeling was followed by activities for participants to explore the purpose and structure of each text type. At the same time, lexico-grammatical features for producing each text type were explicitly taught. From Week 3 to Week 5, the instructor focused on the argument text type since it is one that is commonly used in college academic writing as a culmination of the explanation, discussion, and exposition texts. In small groups, the participants completed text analysis and practiced writing the argumentative text on the topic of “Eugenics.” The syllabus is shown in Appendix A.
In the development of the argumentative essay, the participants used the text type knowledge they acquired to shape the structure of their essays and increase textual cohesion and coherence. As claimed by Connor (2009), textual cohesion and coherence in an argumentative text are closely related. The cohesive ties are lexical and grammatical items which make sentences well connected to serve the purpose of forming a discourse that is specific to an argument (Connor, 2009). Coherence, on the other hand, reflects the sequence of parts controlled by a text-type-specific principle that makes the text meaningful to the reader (Connor, 2009). For an argument text to be coherent, moves such as the “making the claim statement” in the introduction,
“providing sub-topics that justify the claim,” and “inducing the original claim in the concluding statement” are necessary for fulfilling the communicative purpose (Connor, 2009). Following this rationale, the experimental group was instructed to begin the argument with a thesis statement that presented their viewpoint as the introduction.
Then, in the body part, students learned to present their reasons and evidence as arguments to support the thesis. For the conclusion, students practiced inducing the original claim by reinforcing the main argument. In terms of lexico-grammatical features that serve as cohesive devices for this text type, emphasis was placed on presenting cumulative arguments with various types of conjunction, namely temporal (including numerations such as first, second, last), cause-effect (because, therefore), adversative (but, although), and means-purpose (through, by means of) conjunctives.
Throughout the entire text, simple present tense was advised for maintaining coherence
of the argumentative text (Connor, 2009). Different from traditional error-correcting or process-oriented writing instruction, the training sessions provided the necessary cohesive devices aimed to construct a coherent text in the form of an argument, thereby making the writing instruction genre-based.
For the control group, content-based summary writing related to Bio-technology was used as instruction. After read a 600-word passage on topics such as The Human Genome Project, Brain, and Darwin’s Evolution Theory, students were asked to write a 60-word summary about the passage. The instructor used a three-step summarizing guideline: (1) underlining important concepts; (2) writing one sentence about each paragraph; and (3) shortening each sentence to describe the main ideas of the entire passage to teach summary writing. The content and the ability to summarize effectively was the focus in the control group.
Instruments Genre awareness test
To find out the level of the participants’ genre awareness, and to examine the effect of genre-based writing instruction, a genre awareness test was given to both groups along with the pre- and post-writing tests. The genre awareness test focused on the four simple text types commonly encountered in students’ daily lives-narratives, procedures, reports, and news articles by asking about the purpose, context, audience and possible source of each (See Appendix B for the Genre Awareness Test). The test construct aimed to measure students’ familiarity with the four text types and their ability to infer the communicative purpose and context of each example text. The sample texts were excerpts from authentic sources such as English newspapers, story books, encyclopedias and instructional manuals.
Writing test
For both the control and experimental groups, the pre- and post-tests used the same essay topic asking students whether they agree or disagree with the quote on
language learning (See Essay Prompt in Appendix C). Designed to reflect the approach of academic literacies in EAP, the writing prompt related closely to students’ English learning in institutional settings throughout their academic career. Students were encouraged to think and write from personal experience about situated language learning. Students had 30 minutes to compose a short essay of about 150 words.
Treatment evaluation questionnaire
The experimental group completed an additional questionnaire in seventh week after the post-test to evaluate the training sessions. To ensure the validity of the evaluation questionnaire, the designed responses were matched to a five-point scale with descriptors that reflect the extent to which they agree on the clarity, usefulness, effectiveness, and necessity for an extension of the training (See Appendix D for the Questionnaire).
Data analysis
Pre-and post-genre awareness test
Total scores represented students’ performance in text type identification and the ability to indicate possible purpose, author, reader, and source of the text. Since a significant difference was found in the pre-test for genre awareness between the experimental and control groups, MANCOVA was used to account for this covariate in the post-test analysis. The test included 20 questions for students to answer. The first four questions are for identification of the four text types; the remaining 16 questions were Fill-in-the-Blanks questions about purpose, author, audience, and source for the four text types.
The researchers rated the test and assigned five points to each correct answer, amounting to a full score of 100 points. Cronbach’s alpha reliability test showed a high internal consistency for the genre awareness test. The reliability coefficient was close to 1.00 for the pre- and post-tests by both groups (α = .932 (control group pre-test);
α = .961 (control group post-test); α = .885 (experimental group pre-test); α = .916 (experimental group post-test)).
Pre-and post-writing tests
The pre- and post-tests of students’ writing in both control and experimental groups were graded by two raters, with analytical scoring categories of introduction, body, conclusion, and textual coherence. The writing is judged based on the presence of a clear thesis statement in the introduction showing agreement or disagreement with the prompt statement, an argument based on the writer’s English learning experience with supporting details in the body, and a reiteration of the argument about their attitude in the conclusion. Textual coherence is also judged based on overall readability determined by the above-mentioned schematic boundary of the text as well as language use and transitions (Please find the Writing Rubric in Appendix E). The pre-test scores were first analyzed by ANOVA, then, MANCOVA was performed for the analysis on post-test total and categorical scores.
Spearman’s rank order correlation was used to determine the inter-rater reliability.
For the experimental group, positive correlations were found for the pre-test (rs(41) = .890, p = .000) with an large effect size (r2) of .79 and for the post-test (rs(41) = .768, p = .000) with an moderate effect size (r2) of .59. For the Control Group, significant correlation between the two raters was also found for both the pre-and post-tests, with an effect size of .74 for the pre-test (rs(41) = .860, p = .000) and .73 for the post-test (rs(41) = .857, p = .000). Overall, the two raters yielded high inter-rater reliability. To test the intra-rater reliability, both raters scored the pre-and post-tests twice (one week apart) for the control and experimental groups. The intra-rater reliability was strong for both raters for all sets of tests with medium to large effect sizes, though Rater 2 displayed higher stability for most parts as presented in Table 1.
Textual analysis on the writing tests focused on the development, structure and cohesive devices. Text development is reflected through writing fluency, which is indicated by total number of words (Yasuda, 2011) and was analyzed by 2×2 mixed- design ANOVA on SPSS. The structure is judged based on students’ ability to show clear text boundary between introduction, body, and conclusion. The structure of the short argument essay was also analyzed through 2×2 mixed-design ANOVA based on
Table 1
Intra-rater reliabilities
Group N Total Score Introduction Body Conclusion Coherence Experimental 43
Pretest
Rater 1 .881*** .659*** .734*** .861*** .689***
Rater 2 .942*** .892*** .911*** .914*** .936***
Posttest
Rater 1 .679*** .621*** .569*** .543*** .498
Rater 2 .676*** .620*** .561*** .530*** .502
Control 43
Pretest
Rater 1 .682*** .676*** .548*** .597*** .544***
Rater 2 .913*** .860*** .842*** .755*** .835***
Posttest
Rater 1 .823*** .798*** .580*** .674*** .648***
Rater 2 .953*** .944*** .877*** .892*** .890***
***p = .000 (2-tailed)
a 4-point rubric where one point is given respectively to: (1) a clear thesis statement, (2) supporting evidence/arguments, (3) conclusion and (4) clear text boundaries for the three above-mentioned sections marked by transition signals. Finally, the frequencies of cohesive devices, temporal (including numeration), cause-effect, adversative, and means-purpose were measured as text coherence for the argument text and analyzed through MANOVA for multiple dependent variables.
Treatment evaluation questionnaire
Since the post-treatment questionnaire for the experimental group was intended for evaluation purposes, the five questions covered various aspects of the training.
The multifaceted nature of the questionnaire therefore resulted in a fairly low internal consistency (Cronbach’s α = .379).
Results
Genre awareness test results
Pre-test results indicated fairly low levels of genre awareness for both the experimental group (M = 55.2) and the control group (M = 37.7). As the analysis shows, a main effect was found for treatment as represented by “group” in Table 2. The result shows that after the treatment, the experimental group had significant improvement in genre awareness with a large effect size represented by Eta Squared (η2) where .01 constitutes a small effect, .09 a moderate effect and .25 a large effect (Cronk, 2006).
Table 2
MANCOVA on post-test results in genre awareness and writing scores (total and categorical)
Wilk’s Lambda F p η2
Covariate
Genre awareness pre-test .948 .452 .511 .052
Independent Variable
Group .569 11.807 .000 .431
CONTROL EXPERIMENTAL
Dependent Variables M(SD) M(SD) F df p η2
Genre awareness
(post-test) 34.837 (4.599) 77.369 (4.543) 41.362 1 .000 .335†††
Writing score (post-test)
Total 13.976 (.613) 18.070 (.882) 10.175 1 .002 .110††
Introduction 3.990 (.316) 4.266 (.312) .368 1 .546 .004†
Body 4.284 (.220) 5.095 (.217) 6.599 1 .012 .074††
Conclusion 2.545 (.283) 4.119 (.280) 14.956 1 .000 .154††
Textual Coherence 3.157 (.241) 4.591 (.238) 17.157 1 .000 .173††
Note: N = 85; †††large effect size (Eta2); ††moderate effect size (Eta2); †small effect size (Eta2).
Writing test results
For the writing test scores, as Rater Two demonstrated higher intra-rater reliability, results from Rater Two’s scoring for both experimental and control groups are used for analysis. Pre-test scores for both groups were analyzed through ANOVA and no significant difference was found in total (p > .05) and categorical (p > .05 for introduction, body, conclusion, and textual coherence) scores between the control and experimental groups. Therefore, it was unlikely that difference in English writing competence prior to training could be an extraneous variable in the experiment. Since there was no significant difference in the writing pre-tests between the two groups, MANCOVA (as shown in Table 2) was performed with “group” as independent variable, and “post-tests scores in genre awareness and writing” (both total and categorical) as dependent variables. A main effect of treatment was found for post- test total writing scores with a moderate effect. The breakdown of total scores showed significant improvement in body (p < .05), conclusion (p < .05), and textual coherence (p < .05) for the experimental group.
In terms of textual development, mixed-design ANOVA showed an interaction effect between time and group (p < .01) with a large effect size as presented in Table 3. For the structure, mixed-design ANOVA also showed an interaction effect (p < .01) with a moderate effect size. The interaction effects on structure suggest that with time, the treatment had significant impact on students’ writing performance. The average number of words for the experimental group increased 57.7% (from M = 118 to M = 186) while the average word count for the control group increased only half as much by 28.1% (from M = 121 to M = 155). Structurally, 74.1% of the writing improved in text boundary for the experimental group but only 28% of writing displayed clearer text boundaries for the control group in the post tests.
MANOVA results for the frequency of cohesive devices revealed a significant group effect due to training (p < .05). This result matched our expectation that as the experimental group received specific teaching and practice on cohesive devices for the argument text, better command of these cohesive devices would emerge in
their writing. Follow-up univariate tests in Table 3 indicated that the use of temporal/
numeration (p < .05), adversative (p < .05), and means-purpose (p < .05) conjunctives was significantly increased after the training sessions but not for the cause-effect conjunctive (p > .05). These findings suggest that students might already demonstrate sufficient usage of cause-effect conjunctives as they are common linkage devices for essay writing. Therefore, the training did not create a significant effect on the use of this conjunctive. However, students might not have been explicitly taught the usage of the other three types of conjunctives prior to the teaching sessions. After specific training on these cohesive devices, they learned to describe their English learning experience using more temporal conjunctives, sequence their arguments with numeration devices such as “first” and “second,” and explain how to learn English through different means with means-purpose conjunctives.
Treatment evaluation results
The treatment evaluation questionnaire for the experimental group comprised of five questions concerning the effect of the training sessions on: (1) the understanding of text type instruction; (2) the usefulness of text-type writing activity; (3) usefulness of learning distinctive textual organization and lexico-grammatical features; (4) usefulness of text type knowledge in future study; and (5) extending the duration of the teaching sessions. With regard to the understanding of text type instruction, 87.6% showed good understanding of the knowledge transmitted during the teaching sessions (M = 3.53). In terms of the usefulness of text-type writing activity, 89.6% of the respondents think it was useful (M = 3.56). Concerning the usefulness of distinguishing different rhetorical features, 50% of participants rate the activity as useful (M = 4.06). In terms of the use in future study, 77.1% responded that acquiring text type knowledge was somewhat useful (M = 3.36). Finally, 67% of respondents saw a great need for extension of the duration of the teaching (M = 3.94). Overall, the participants gave positive evaluation on the usefulness and effectiveness of the teaching sessions. The results are presented in Table 4.
Table 3
Changes in development, structure, and cohesion for the writing tests
Development Mixed-design ANOVA
Mean SD F p η2
Pre-/Post- Pre-/Post- Experimental 118.518/186.963(+) 32.535/58.148 Control 121.194/155.194(+) 26.485/44.866
Interaction Effect 110.381 .000 .644†††
Structure Mixed-design ANOVA
Mean SD F p η2
Pre-/Post- Pre-/Post- Experimental 2.703/3.818(+) .608/.395
Control 2.920/3.200(+) .640/.707
Interaction Effect 14.600 .000 .226††
Cohesion MANOVA
Wilk’s Lambda F p η2
Group Effect .691 6.692 .000 .309†††
Mean Difference SD F p η2
Temporal/Numeration 4.304 .042 .064†
Experimental 1.851(+) 2.106
Control .921(+) 1.513
Cause-effect 3.154 .081 .048†
Experimental .8148(+) 1.881
Control -.0263(-) 1.881
Adversative 11.194 .001 .151††
Experimental 1.814(+) 1.664
Control .500(+) 1.484
Means-purpose 9.455 .003 .130††
Experimental .296(+) .992
Control -.368(-) .750
Note: (+)mean increased; (-)mean decreased; †††large effect size (Eta2); ††moderate effect size (Eta2);
†small effect size (Eta2).
Table 4
Treatment evaluation questionnaire
Question Mean
1. The understanding of text type instruction 3.53
2. The usefulness of text-type writing activity 3.56
3. Usefulness of learning distinctive textual organization and lexico-grammatical
features 4.06
4. Usefulness of text type knowledge in future study 3.36
5. Extending the duration of the teaching sessions 3.94
Discussion
Overall treatment effectiveness and positive feedback from participants
The findings suggest that both the experimental and control group had fairly low text type awareness to begin with. After the experiment, the control group showed no significant improvement in genre awareness. The possible reason is that the teaching for the control group focused on summary writing only. Without explicit instruction on genre analysis and key academic text types, students in the control group did not gain genre awareness. In contrast, the experimental group that underwent genre- based instruction showed significant improvement in their awareness of possible purpose, author, reader, and source. The pre- and post- writing test results proved the effectiveness of the training sessions in improving textual organization and coherence.
Although the introduction section did not show significant improvement for the experimental group, participants in this group did demonstrate better text structure with clearer text boundary and more use of cohesive devices such as temporal, adversative, and means-purpose conjunctives after the training.
In considering the factors that led to the significant writing improvement in the experimental group, the possible extraneous variable of pre-existing difference
in writing competence between the two groups had been eliminated in the pre-tests.
Moreover, the EAP course was the only English course participants were taking. No other English-mediated instruction was received on campus. Even if the students self-studied outside of class, it was inferred that this factor would affect both groups equally. Therefore, it was reasonable to attribute the writing improvement in the experimental group to the genre-based instruction in academic writing. In terms of the experimental group’s response to the training sessions, an overall positive evaluation was found with regard to the teaching of text type knowledge from the results of the treatment evaluation questionnaire. The usefulness of the workshop is reflected by the participants’ preference to extend the training.
Current level of genre awareness for EFL college students
The genre awareness test helped us to find out about participants’ consideration of the purpose, author, audience, and possible text source for four commonly encountered text types. The present level of genre awareness appears fairly low from the results, thereby informing us that students in Taiwanese colleges are not well-equipped with genre-related skills at present. This might be due to the fact that most students have not been exposed to genre in their previous academic experiences in junior or high schools.
At best, genre-based approach has only been used in a small number of studies in higher education settings in the last decade (Liou et al., 2005; Tsai, 2006), and not yet widely implemented in Taiwan.
Furthermore, with respect to the practical usage of the workshop content, as much as 52% of the students responded negatively or neutrally toward possible future application in their studies. One possible reason is that the students have not had the opportunity to apply academic writing to their specialty areas such as medicine or engineering under the current policy. Since there is no immediate perceived need for genre application, non-English college majors in Taiwan still consider genre knowledge to be unessential for becoming accepted members of the local academic community.
The situation is changing rapidly, however, with globalization and the continued implementation of English-mediated instruction by the ministry of education. EFL
educators in Taiwan are therefore held responsible for enhancing students’ academic writing ability as it is foreseeable that students will soon be required to produce thesis writing in their specialized areas in English. Together, the questionnaire results and the above-mentioned considerations suggest the need to expand the genre-based curriculum to a wider population in Taiwan.
Success of the training sessions on improving writing quality
Within seven weeks of training, students reflected on their English learning experience, expressed their viewpoint on English learning attitudes, and became equipped with the ability to draw on the newly acquired genre knowledge transmitted to them through workshops and genre-analysis activities. On a micro-structural level, the group writing practices allowed students to focus on linguistic features such as cohesive devices by practicing using temporal/numeration, adversative, and means- purpose conjunctives to construct their argument text on English learning. Macro- structurally, students acquired knowledge on the generic features of different text types accumulatively. They also synthesized the functions of explanation, discussion and argumentative text types in the post-test to respond to the writing prompt about their view on English learning. This was an appropriate application of using language in social context since the writing topic was relevant to the students’ academic lives. As put forth by Johns (1997), students come to read and write with confidence after they are able to use the formal characteristics of genre as a member of the international academic community. The participants demonstrated their ability to use the expected conventional features in the academic writing discourse appropriately as a member of the community. The success of the academic literacy training sessions might inspire more EFL practitioners in Taiwan to design and implement similar programs in college writing courses.
Conclusion
From an academic literacy perspective, the current study set out to first discover the current level of genre awareness for college EAP students in Taiwan, and second, test the effects of text type teaching sessions on enhancing both genre awareness and writing for the students. The results indicated an initial low level of genre awareness among participants. After training on academic text types, however, genre awareness and knowledge on text purpose, writer, audience, and source increased significantly for the experimental group. The positive effects of genre-based writing instruction on improving overall writing quality for the argument text in this study provide a template for college EFL instructors who are interested in enhancing students’ textual coherence at the macro-structural level, and fluency, structure, and cohesion at the micro-structural level of conventional academic texts. As globalization has exposed Taiwanese college students to the international academic community, how to gain recognition and membership into the community becomes an important research area for EFL educators in this locale. The findings of the study imply an urgent need for EFL practitioners in Taiwan to raise students’ genre awareness in academic writing instruction. Future research could explore the benefits of genre-based writing instruction to help students achieve long-term academic writing success.
Several limitations from this study could be avoided in future investigations. First, the duration of the training sessions only lasted seven weeks in total. Although the study produced significant results, both we the researchers and the participants would prefer an extension of the treatment to have a more solid understanding and practice on the target text-types. Second, there was only one post-test immediately after the teaching. To ensure treatment validity, a follow-up post-test could take place a few weeks after the workshop to see if students indeed were able to retain and apply the genre knowledge acquired from the seven weeks experimental design. For improving curriculum design, the brief exploration on textual organization and lexico-grammatical
features employed in this study could be developed into more in-depth lessons spread out in a longer training period for more desirable learning outcomes and lasting effects.
Acknowledgments
The authors would like to thank the anonymous reviewers for their valuable comments and suggestions to improve the quality of the paper. The financial supports from National Science Council and English Eagle Project of National Cheng Kung University in Taiwan are gratefully acknowledged.
DOI: 10.3966/102887082013065902004
References
Ahn, H. (2012). Teaching writing skills based on a genre approach to L2 primary school students: An action research. English Language Teaching, 5(2), 1-16.
Brett, P. (1994). A genre analysis of the results section of sociology articles. English for Specific Purposes, 13(1), 47-59.
Byrnes, H. (2009). Systemic-functional reflections on instructed foreign language acquisition as meaning making: An introduction. Linguistics and education, 20(1), 1-9.
Cheng, A. (2007). Transferring generic features and recontextualizing genre awareness:
Understanding writing performance in the ESP genre-based literacy framework. English for Specific Purposes, 26(3), 287-307.
Cheng, A. (2008a). Analyzing genre exemplars in preparation for writing: The case of an L2 graduate student in the ESP genre-based instructional framework of academic literacy.
Applied Linguistics, 29(1), 50-71.
Cheng, A. (2008b). Individualized engagement with genre in academic literacy tasks. English for Specific Purposes, 27(4), 387-411.
Christie, F. (1999). Genre theory and ESL teaching: A systemic functional perspective. TESOL Quarterly, 33(4), 759-763.
Coffin, C., & Donohue, J. P. (2012a). Academic literacies and systemic functional linguistics:
How do they relate? Journal of English for Academic Purposes, 11(1), 64-75.
Coffin, C., & Donohue, J. P. (2012b). English for academic purposes: Contributions from
systemic functional linguistics and academic literacies. Journal of English for Academic Purposes, 11(1), 1-3.
Connor, U. (2009). A study of cohesion and coherence in English as a second language students’
writing. Paper in Linguistics, 17(3), 301-316.
Cronk, B. C. (2006). How to use SPSS: A step-by-step guide to analysis and interpretation.
Glendale, CA: Pyrczak.
Derewianka, B. (1991). Exploring how texts work. Newtown, Australia: Primary English Teaching Association.
Flower, L., & Hayes, J. (1986). Writing research and the writer. American Psychologist, 41(10), 1106-1113.
Gao, J. (2007). Teaching writing in Chinese universities: Finding an eclectic approach. Asian EFL Journal, 20(2), 1-11.
Gardner, S. (2012). Genres and registers of student report writing: An SFL perspective on texts and practices. Journal of English for Academic Purposes, 1(1), 52-63.
Halliday, M. (1985). An introduction to functional grammar. London, UK: Edward Arnold.
Hammond, J., Burns, A., Joyce, H., Brosnan, D., & Gerot, L. (1992). English for social purposes: A handbook for teachers of adult literacy. Sydney, Australia: National Centre for English Language Teaching and Research, Macquarie University.
Henry, A., & Roseberry, R. B. (1998). An evaluation of a genre-based approach to the teaching of EAP/ESP writing. TESOL Quarterly, 32(1), 147-156.
Hopkins, A., & Dudley-Evans, T. (1988). A genre-based investigation of the discussion sections in articles and dissertations. English for Specific Purposes, 7(2), 113-121.
Hsieh, W. M., & Liou, H.-C. (2005, November). Text analyses and online material development for EAP graduate courses. Paper presented at Thirteenth International Symposium on English Teaching of English Teachers’ Association, Taipei, Taiwan.
Hyland, K. (2003). Genre-based pedagogies: A social response to process. Journal of Second Language Writing, 12(1), 17-29.
Hyland, K. (2004). Genre and second language writing. Ann Arbor, MI: The University of Michigan Press.
Hyland, K. (2009). Academic discourse. London, UK: Continuum International.
Hyland, K., & Hamp-Lyons, L. (2002). EAP: Issues and directions. Journal of English for Academic Purposes, 1(1), 1-12.
