• 沒有找到結果。

The Effects of Rhetorical Task Type, English Proficiency, and Writing Anxiety on Senior High School Students' English Writing Performance

N/A
N/A
Protected

Academic year: 2021

Share "The Effects of Rhetorical Task Type, English Proficiency, and Writing Anxiety on Senior High School Students' English Writing Performance"

Copied!
37
0
0

加載中.... (立即查看全文)

全文

(1)

The Effects of Rhetorical Task Type, English

Proficiency, and Writing Anxiety on Senior High

School Students’ English Writing Performance

Pi-chen Tsai

Yuh-show Cheng

Taipei Municipal Da-Zhi High School

National Taiwan Normal University spry.psyche@msa.hinet.net t22035@ntnu.edu.tw

Abstract

English writing tests form a designated section in the General Scholastic Ability English Test (GSAET) and the Department Required English Test (DRET), two nationwide tests for college admission in Taiwan. However, the types of rhetorical writing tasks emphasized in the two tests are different. The GSAET focuses on narrative tasks, whereas DRET, on expository tasks. Because the DRET is designated to assess more advanced knowledge and ability than the GSAET, their difference in focus of rhetorical task suggests that narrative writing is less challenging than expository writing. To validate the assumed relative ease of the two writing tasks, the present study examined the difference the two tasks made in senior high school students’ English writing performance. In light of the potential effects of language proficiency and anxiety on writing performance, this study also explored how the two task types interacted with English proficiency and writing anxiety in influencing the quality of English writing. Students’ opinions about different rhetorical writing tasks were also explored. The results show neither a significant main effect of rhetorical task type nor a significant interaction effect between task type and English proficiency or writing anxiety. In contrast, significant main effects of English proficiency and writing anxiety were observed.

Key Words: rhetorical task type, English proficiency, writing anxiety

(2)

INTRODUCTION

English writing, which was often a neglected part of EFL pedagogy, is receiving greater attention nowadays. As the acquisition of writing skills is deemed important, the testing of writing is in a great demand, and many language tests that enjoy international popularity, such as TOEFL, IELTS, and Cambridge ESOL exams, contain a section on essay writing.

English writing ability is also tested in the General Scholastic Ability English Test (GSAET) and the Department Required English Test (DRET), two nationwide tests for college admission in Taiwan. According to the latest test guidelines for the GSAET (College Entrance Examination Center [CEEC], 2007), the English writing section of the GSAET is aimed to assess students’ “narrative” skills. On the other hand, the test guidelines for the DRET (CEEC, 2007) display a shift in the focus of assessment from “narrative” writing skill to “descriptive” and “expository” writing skills. When taking the GSAET, test-takers are often asked to write a narrative essay. In contrast, an expository writing task is usually designed for the DRET. As the DRET is purported to assess more advanced knowledge and ability than the GSAET (Jeng, Lin, Chen, Hsiao, & Curran, 2001), their different focuses of assessment suggest an assumption that a narrative writing task is less challenging than an expository one. To test the validity of the assumption, this study examined whether the two rhetorical tasks made a significant difference in learners’ writing performance. Because previous research has suggested that general language proficiency and writing anxiety each may affect writing

(3)

type, this study also investigated whether the effect of rhetorical task type on writing performance varied with learners’ general language proficiency and writing anxiety. In addition, learners’ opinions about different rhetorical writing tasks were collected to complement the quantitative data and help interpret the results of statistical analyses. Specifically, the following research questions directed the design of this study:

1. How do the two rhetorical tasks, narration and exposition, affect the quality of English essays written by Taiwanese senior high students in a testing situation?

2. How do the two rhetorical tasks, narration and exposition, interact with language proficiency in affecting the quality of English essays written by Taiwanese senior high students in a testing situation?

3. How do the two rhetorical tasks, narration and exposition, interact with writing anxiety in affecting the quality of English essays written by Taiwanese senior high students in a testing situation?

4. How do Taiwanese senior high school students view different types of rhetorical tasks?

LITERATURE REVIEW

Approaches to Writing Assessment

There has been a growing interest among educators and researchers in approaches to writing assessment. Two methods widely practiced in assessment of writing abilities are direct writing

(4)

assessment and indirect writing assessment. Direct writing assessment requires learners to write an essay or several essays on given topics, which are then evaluated according to certain pre-specified criteria. Indirect writing assessment does not require learners to actually write an essay; instead, it measures writing ability by asking learners to judge the appropriate use of language, usually in the form of multiple-choice questions. In terms of practicality and psychometric features, both methods have their merits and demerits (Ackerman & Smith, 1988; Cooper, 1984; Hyland, 2003; Stiggins, 1981).

For practical reasons, the administration and marking of indirect writing assessment is easy and cost-effective. The rating of direct writing assessment, however, takes more time, and requires some training in advance of the marking (Cooper, 1984; Hyland, 2003). Psychometrically speaking, indirect writing assessment through the use of objective formats can achieve higher reliability and there is better control over the skill to be assessed. On the other hand, direct writing assessment may not ensure high statistical reliability, but it is more likely to attain higher face validity and reflect real-life communicative demands (Hughes, 1989; Hyland, 2003).

The validity issue may account for the current popularity of direct writing assessment over indirect writing assessment. The latter is often criticized as making inferential judgments about learners’ writing ability and confounding learners’ writing ability with reading skills and/or error-recognition skills (Cooper, 1984; Hyland, 2003). Thus, most ESL/EFL professionals today prefer assessing learners’ writing ability directly through an essay writing task (Hamp-Lyons, 2003).

(5)

Effect of Rhetorical Writing Task Type on Writing Performance Writing tasks have been defined differently from different perspectives; so have the components of writing tasks. Weigle (2002) synthesizes these views and proposes a comprehensive theoretical framework, in which a writing task is specified as containing seven components: subject matter, stimulus, scoring criteria, prompt, transcription mode, testing time, and type of rhetorical task. Although each of the components is believed to affect test-takers’ performance on the writing task, Bereiter and Scardamalia (1987) reserve an especially important space for rhetorical problems in the composing process.

Nevertheless, previous research produced mixed results on the effect of rhetorical task type on writing performance. Although most investigations indicated that students performed better on descriptive or narrative tasks that are related to writers’ personal experience (Kegley, 1986; Koda, 1993; Prater & Padia, 1983; Reed, 1992; Way, Joiner, & Seaman, 2000), some studies showed that students tended to do better on expository or argumentative tasks (Faigley, Daly, & Witte, 1981; Quellmalz, Capell, & Chou, 1982). In L1 writing research, Prater and Padia’s (1983) study on elementary school students showed that their subjects’ expressive writing was high in quality, but their explanatory writing and persuasive writing were not equally good. Kegley (1986) investigated secondary school students and reported that the students did best on narrative tasks, followed by descriptive tasks and expository tasks, and they performed worst on persuasive tasks. Similarly, Reed (1992) discovered that college students performed best on the narrative writing task, mediocre on the expository writing task and the descriptive writing task, and worst on

(6)

persuasive writing. However, in Quellmalz et al.’s (1982) study, senior high students tended to produce good expository essays but poor narrative essays, and the group who did well on the narrative writing did not necessarily write good expository essays. Crowhurst (1978) and Faigley et al. (1981) further observed that students’ argumentative essays were syntactically more complex than their narrative essays.

L2 writing research produced a slightly different pattern from L1 research regarding the effect of rhetorical task type on writing performance. Comparing the writing performance of college L2 learners of Japanese on a descriptive task and a narrative task, Koda (1993) discovered that the learners got higher holistic ratings on the descriptive writing task than on the narrative task. Way et al. (2000) investigated three rhetorical tasks: descriptive task, narrative task, and expository task. Holistic scoring of students’ essays also revealed that secondary learners of French performed best on the descriptive task, and worst on the expository task. Koda’s and Way et al.’s findings are slightly contradictory to the task effect displayed in Kegley’s (1986) and Reed’s (1992) L1 studies, where learners excelled in narrative writing rather than descriptive writing. Cheng’s (2006) study on EFL college students’ performance on a narrative writing task and an argumentative writing task revealed an even more complex picture. Although the task effect was not significant in terms of holistic evaluation of essay quality, Cheng found that learners tended to produce longer narrative essays of less syntactic complexity but shorter argumentative essays of higher syntactic complexity. Interestingly, the finding that EFL learners’ argumentative essays

(7)

were syntactically more complicated is consistent with that of Faigley et al.’s (1981) L1 study.

These confusing results could be due to variations in research focus and design across studies, including use of different measurements of writing performance (e.g., evaluation of syntactic complexity of the text vs. holistic rating of writing quality), different categorizations of rhetorical tasks (e.g., the categorization of expressive, persuasive, expository writing vs. that of description, narration, exposition, and argumentation), as well as differences in the rhetorical tasks compared, the language investigated (i.e., L1 vs. L2), and in variables put under control (e.g., with or without the topic controlled).

The issue of research design is worth some elaboration here because methodological problems in some of the previous studies added to confusion regarding the effect of rhetorical task. First of all, the topic effect in some studies was not well controlled. For example, different topics were assigned to different rhetorical tasks in Faigley et al. (1981), Koda (1993), Reed (1992), and Quellmalz et al. (1982). This type of research design confounds topic effect with task effect, making it difficult to conclude what caused the variation in writing performance. The scoring procedures in certain studies could also cause confounding effects. For example, Kegley (1986) evaluated different rhetorical tasks with different scoring guides while Way et al. (2000) adopted different scoring criteria to evaluate the writing of different ability groups. Although these scoring procedures had the advantage of matching scoring rubrics to task features and learner ability levels, they suffered the problem of confounding task effect with rubric or standard effect. That is, quality differences observed in

(8)

learners’ performance on different rhetorical tasks could result from as much rating standard/rubric as rhetorical task type. In view of these pitfalls, the current study took care to control topic effect and opted for using a generic scoring rubric for the two rhetorical tasks under study.

Effect of General Language Proficiency on Writing Performance Like research on the effect of rhetorical writing task type, inconsistent results have been reported on the relationship between general language proficiency and writing performance. Some studies reported that learners’ writing performance was independent of their general language proficiency (e.g., Cumming, 1989; Raimes, 1987; Zamel, 1982), but a considerable number of studies noted that general linguistic competence was positively correlated to writing performance. For example, Reed (1992) recruited college students as participants and observed three proficiency groups’ writing performance. He discovered that the high ability group outperformed the other two groups in holistic writing scores across four types of rhetorical tasks. In a study conducted by Sasaki and Hirose (1996), a strong link was also found between Japanese EFL learners’ expository writing performance and their English proficiency levels. Likewise, Lin’s (2005) research on Taiwanese vocational high school students’ English writing suggests that proficient learners wrote better than less proficient learners. On the whole, highly proficient language learners tended to be good writers that produced longer and better essays, whereas language learners at lower proficiency levels seemed to be generally weak in writing (e.g., Chang, 1995; Lin, 2005; Reed, 1992;

(9)

Sasaki & Hirose, 1996; Way et al., 2000).

However, it is unclear how general language proficiency would influence the effect of rhetorical task type on writing performance. For example, Reed’s (1992) study found no interaction between task type and language proficiency in writing performance. Both low and high proficiency learners were good at narrative writing and received the lowest ratings in persuasive essay writing. Way et al.’s (2000) findings, on the other hand, indicated that students of lower proficiency performed better on the descriptive writing task than the narrative task, whereas students of higher proficiency did nearly equally well on both writing tasks.

Effect of Writing Anxiety on Writing Performance

Despite a general neglect of affective variables in past writing research, writing anxiety has received increased attention in L1 and L2 writing research recently. This body of research presents a complicated relationship between writing anxiety and writing performance. Mixed results have been found on the association between writing anxiety and writing performance, with some studies reporting no significant relationship between the two (e.g., Madigan, Linton, & Johnson, 1996; Richardson, 1981) and many others observing a significant correlation (e.g., Daly & Miller, 1975; Faigley et al., 1981; Faris, Golen, & Lynch, 1999; Lee & Krashen, 2002). What is more, the association between writing anxiety and writing performance seemed to differ with rhetorical task type. For example, in Cheng’s (2006) study, writing anxiety had a significant effect on writing fluency for the argumentative task but not for the narrative task. On the other hand, Faigley et al. (1981) suggested that narrative

(10)

writing tasks that require writers to disclose their personal feelings tended to influence the writing of highly apprehensive writers, whereas expository writing tasks that focus on opinion expression had little effect on the quality of anxious writers’ texts.

In summary, previous research has demonstrated a complex relationship among rhetorical task type, general language proficiency, writing anxiety, and writing performance. Therefore, more research is needed to achieve a better understanding of the roles of rhetorical task type, general language proficiency, and anxiety in EFL learners’ writing performance.

METHOD

Participants

The participants of the main study were 151 (77 male and 74 female) third-year students from four intact classes at a prestigious public senior high school in Taipei. Students admitted to this school generally scored in the 98 percentile in the Basic Competence Test for junior high school students. In consideration of their future academic and career interests, students in three of the classes were on the science track, and students in one of the classes were on the humanities track. At the time of data collection, approximately half of the students were taking additional English courses outside school and nearly 50 percent of them had taken a General English Proficiency Test (GEPT) given by the Language Training and Testing Center (LTTC) in Taiwan.

(11)

Instruments

In the following sections, detailed descriptions of the contents and, whenever appropriate, sources and examples of the instruments are offered.1

Foreign language proficiency test. To assess the participants’

general English proficiency, all participants were given a shortened English version of the intermediate-level Foreign Language Proficiency Test (FLPT) developed by the LTTC. The FLPT was chosen because the researchers assumed that it was less likely for the participants to take this test prior to this study due to the restriction LTTC puts on test-takers: Only adult language learners above 18 are allowed to take this test or to buy the simulated test. The assumption was later confirmed by the participants’ feedback as none of them had taken the test. The English version of the FLPT has four levels: elementary, intermediate, intermediate-high, and high. The practice version of the intermediate-level test was used in this study because LTTC (2005) reports that it has a similar degree of difficulty to the intermediate level of GEPT, whose target test-takers are senior high students. The FLPT test contains four sections: the first three sections involve multiple-choice tests for listening, grammatical usage, as well as vocabulary and reading; the last section is a free-response speaking test. Due to the fact that only a limited amount of class time was available for administration of the test, this test was shortened. First of all, after consulting in-service senior high school English teachers,

1

To observe the Journal’s recommendation concerning length of the manuscript and to avoid potential copyright infringement, the entire set of instruments is not provided. Interested readers could contact the corresponding author (the second author) if they wish to have more information on the instruments.

(12)

the listening and the speaking sections were eliminated for the sake of practicality. The items in the sections of grammatical usage as well as vocabulary and reading were reduced. The final test consisted of a total of 40 items, with 20 items on grammatical usage, 10 items on vocabulary and idioms, and 10 items on reading. To ensure the validity of this shortened FLPT, a pilot test was conducted (see Data Collection Procedures). The participants had to complete the abridged FLPT-intermediate within 30 minutes. A correct response to each item was given one point, so the highest possible score for the adapted FLPT was 40 points.

Second language writing anxiety inventory. This study

employed the Second Language Writing Anxiety Inventory (SLWAI) (Cheng, 2004a) written in Chinese to measure learners’ English writing anxiety. The SLWAI is composed of 22 statements and its Chinese version has been proved to have high reliability and satisfactory criterion-related validity and discriminant validity in Cheng (2004a). The participants were instructed to mark on a 5-point Likert scale to show their agreement to each of the statements, ranging from 1 (strongly disagree), 2 (disagree), 3 (no strong feeling either way), 4 (agree), to 5 (strongly agree). Two examples of the SLWAI items are as follows: “I usually feel my whole body is rigid and tense when I write English compositions.” “My thoughts become jumbled when I write English compositions under time constraints.” A scale score was then created for each participant by averaging his/her responses to all of the 22 items.

Background questionnaire. Written in Chinese, the background questionnaire also contained 22 items. Items 1 to 5 asked

(13)

about the participants’ demographic information such as class, name, and gender. Items 6 to 12 dealt with general English learning experiences, including the time when they started to study English, whether they had been to English-speaking countries, and the length of time they spent on English self-study. Items 13 to 18 focused on learners’ experiences and perceptions of English writing in general, including extracurricular English writing courses they had taken, the frequency they engaged in English mail/letter writing, and how difficult they perceived English writing. Finally, Items 19 to 22 elicited their preference for and perceived ease or difficulty of three rhetorical writing tasks that they were more likely to encounter: expository, narrative, and descriptive.

Writing task. The current study focused on two types of

rhetorical writing tasks that are commonly included in college entrance exams—narrative and expository. To reduce the confounding effect that different topics might induce, both the narrative task and the expository task in this study surrounded the same theme: “recreational activity,” after consulting a bank of topics collected from English composition exercise books in Taiwan. “Recreational activity” was chosen as the topic for writing because it is relevant to senior high students’ life and had not been used in the entrance exams of English writing prior to this study. Moreover, the teachers of the participants in the main study confirmed that the topic had not been practiced in their English classes before.

It is not easy to draw firm conclusions on the distinction between narrative and expository texts because “the labels narrative and

expository encompass a range of specific text types that can have different, and sometimes overlapping, purposes” (Wolfe, 2005, p.

(14)

360). Nevertheless, Berman and Nir-Sagiv (2007) suggest that the two rhetorical texts differ in overall orientation to topic, reference to entities, and text construction. To be specific, narrative texts and stories are agent-oriented. Events are the basic components of text construction in narratives, in which a sequence of events are organized and told according to a temporal and/or causal framework. In contrast, expository texts are topic-oriented and center on concepts and issues. The unfolding of ideas, claims, and arguments is expressed in terms of logical relationships. Based on an analysis of the prompts in US state and national direct writing assessments, Jeffery (2009) has claimed that the two rhetorical tasks differ in the extent they evoke abstract reasoning. A narrative task often evokes more concrete personal experiences, whereas an expository task calls for more abstract thinking on the part of writers. Jeffery’s remark is consistent with Berman and Nir-Sagiv’s (2007) observation that for narrative texts their “participants typically told personalized stories about specific people, objects, and events, whereas they referred to general, abstract ideas in their expository texts” (p. 105).

Against this understanding of narrative and expository texts, the prompts of the two writing tasks in the current study were deliberately written in a way to elicit the intended rhetorical texts (see Appendices A and B). Considering that senior high students in Taiwan may be more familiar with the format of writing tasks in the college entrance exams, the format of the GSAET and DRET was followed in writing the prompts. In the narrative task, the participants were required to provide an unforgettable account of engaging in a recreational activity, in a manner of story telling. Learners were told to be specific about

(15)

the time, the setting, and the events involved in the account. In contrast, the expository task asked the participants to depict their observations about gender preferences for recreational activities and to explain what might cause such gender differences. The minimum length of both essays was set at 120 words. The title and the instructions of the two tasks were written in Chinese for fear that some participants could not understand English instructions well enough. Each of the writing tasks was designed to be completed within 30 minutes. Students were not allowed to use a dictionary.

Post-writing questionnaire. To elicit the participants’ opinions

and feedback on the two writing tasks, each task was followed by a questionnaire that contained three open-ended questions. The first question asked the participants to indicate their preference for the writing task they just took. The second question requested them to self-evaluate the quality of the writing they just produced on the given task. The third question probed if they faced any problems during the writing process. All of the three questions were stated in Chinese, and the participants were allowed to write down their opinions in Chinese.

Data Collection Procedures

Pilot study. Prior to the formal study, a pilot study was

conducted on another group of senior high school students to (1) check the readability of the background questionnaire, the SLWAI, and the shortened FLPT; (2) examine the reliability of the SLWAI and the shortened FLPT; and (3) establish the validity of the shortened FLPT. The three instruments were administered to three classes of senior high students at three different public schools in Taipei (N = 114).

(16)

An analysis of the data collected from the pilot study shows that the SLWAI had good internal reliability (α = .89) and the internal reliability of the shortened FLPT was also adequate (α = .79). The convergent validity of the shortened FLPT was established by computing the Pearson correlation coefficient between students’ FLPT scores and their standardized English course grades (i.e., z scores within each class). The finding of a significant correlation (r = .57, p = .000) between the two scores provides evidence for the convergent validity of the shortened FLPT. Extreme group comparison was then used to examine the discriminating power of the shortened FLPT, following the guidelines in Qiu (2006). Students were ranked according to the total FLPT scores. The average FLPT of the approximately top 27% (n = 28) and bottom 27% (n = 27) of the students was compared via an independent-samples t-test, yielding a significant result (t53 = -5.78, p = .000). The result indicates that the

shortened FLPT could adequately discriminate high achievers from low achievers.

However, some adjustment was made on the background questionnaire based on students’ feedback. For example, an original open-ended question that requested students to report how often they wrote an English essay was changed into one of five closed options. Besides, a brief explanation for “expository essay,” “narrative essay,” and “descriptive essay” was added before the set of questions (Items 19 to 22) designed to elicit students’ opinions about the three rhetorical tasks.

Formal study. The whole data collection procedure in the

(17)

participants received a test packet which contained an adapted version of the intermediate level FLPT and a separate answer sheet. The participants had 30 minutes to complete the test. Immediately after they completed the adapted FLPT, they filled out the SLWAI and the background questionnaire. One week later, the participants were given their first writing test packet. The writing test packet consisted of one writing task and a post-writing questionnaire. All participants had 30 minutes to work on their writing. When they completed their writing task, they were given five additional minutes to fill in the post-writing questionnaire. The order in which the two types of rhetorical tasks were administered was counterbalanced. That is, half of the participants wrote the expository essay first, and the other half wrote the narrative essay first. Since the expository task and the narrative task evolved around the same theme, the implementation of the second writing task took place around three weeks after the first to minimize memory effect. The time limit on the second writing task was also set at 30 minutes.

Procedures for Scoring Writing

To protect the privacy of the participants, the scoring process was anonymous. Holistic rating of writing was used for the sake of efficiency and cost-effectiveness. The rating was based on the five-level scoring schemes utilized in the intermediate level of GEPT. The GEPT scoring rubric is a generic scoring scheme that evaluates test-takers’ fulfillment of a given writing task and language control according to such general criteria as appropriateness and relevance of the content, textual organization, and linguistic range and accuracy. This generic scoring scheme was adopted mainly for four reasons.

(18)

Firstly, the scoring scheme is consonant with the current view of writing as a communicative act (Cumming, Kantor, Powers, Santos, & Taylor, 2000) that entails organizational (i.e., grammatical and textual) and pragmatic (i.e., illocutionary and sociolinguistic) competences, according to Bachman’s (1990) model of communicative language ability. Secondly, as mentioned above, a generic scoring scheme could reduce the potential confounding effect that use of two task-specific rubrics might bring about. Thirdly, since 2000, the GEPT scoring scheme has been used to evaluate test-takers’ performance on various kinds of writing, including expository and narrative writing, and the scores produced based on it has won public recognition. The validity and reliability of the scoring scheme were thus trusted. Fourthly, the intermediate level of GEPT accepts senior high school teachers as raters of the writing tests. This policy made it easier for the researchers to find certified and trained raters to score their participants’ writing samples.

Consequently, two experienced senior high school teachers in Taipei, who had been involved in the official scoring of GEPT writing tests, were invited as the raters. Both of them had participated in the training session held by the LTTC and had rich experiences in rating the GEPT writing tests of intermediate level. In advance of their formal scoring, 20 test papers were selected from the two writing tasks administered in this study, with ten taken from the expository task and ten from the narrative task. Based on the participants’ final English course grades in the previous semester, the ten test papers of each writing task were chosen from students of three different achievement levels: three of the papers were from high-achievers,

(19)

four from middle-achievers, and three from low-achievers. The two raters then scored these 20 test papers together, and discussed reasons for their scores on the basis of the scoring scheme. After the pilot scoring session, the two raters scored the remaining essays independently. Each essay thus received two ratings, and the two ratings were averaged to produce a final score for later analysis. The inter-rater reliability for the narrative task was adequate (r = .76), and that for the expository task was satisfactory (r = .81).

Data Analysis Procedures

Statistical analyses, using SPSS 13, were conducted to answer the first three research questions. To answer the first research question that deals with the effect of rhetorical task type on quality of writing, a paired-samples t-test was run on the whole sample. To answer the second and the third questions, students of high and low English proficiency or writing anxiety were identified before two-way ANOVA with a repeated measure was employed to examine the interaction effect of rhetorical task type with English proficiency and writing anxiety respectively. In addition to statistical analyses, the participants’ responses to both the background questionnaire and the post-writing questionnaires were coded and compared so as to answer the fourth research question.

RESULTS

To answer the first research question regarding the effect of rhetorical task type on learners’ writing performance, the mean scores

(20)

of the two writing tasks were submitted to a paired-samples t-test. The result showed that the quality of narrative writing (M = 3.29; SD = .66) was not significantly different from that of expository writing (M = 3.25; SD = .71), with t 150 = .82, p = .41. This result suggests that

rhetorical task type did not significantly affect learners’ performance on a writing test.

To address the second research question on the interaction effect of rhetorical task type and English proficiency in English writing performance, the participants were first classified into high- and low-proficiency groups based on their performance on the adapted FLPT. Specifically, those scoring in the top third (i.e. above 30 out of the highest possible 40 points) were classified as high-proficiency learners (n = 50), and those scoring in the bottom third (i.e., below 23 out of the highest possible 40 points) were classified as low-proficiency learners (n = 54). Table 1 shows the two proficiency groups’ writing scores on the two rhetorical tasks.

Table 1

Different Proficiency Groups’ Writing Scores on the Two Rhetorical Tasks

Task Proficiency Group M SD

Narrative Task High Intermediate 3.76 0.44 Low Intermediate 2.81 0.68 Expository Task High Intermediate 3.70 0.52 Low Intermediate 2.71 0.69

(21)

The results of two-way repeated measures ANOVA showed a significant main effect of general English proficiency, with F1, 102 = 92.94,

p = .000, in favor of the high-proficiency group. That is, the high-proficiency group earned higher scores than the low-proficiency group on the writing tasks. However, consistent with the above t-test result, no significant task effect on learners’ writing performance was observed, with F1, 102 = 1.59, p = .21; nor was there a significant interaction effect

between the task type and the proficiency level, with F1, 102 = 0.07, p = .79.

The third research question explored the interaction of rhetorical task type and writing anxiety in affecting the quality of English essays. To answer this question, the participants were categorized into high- or low-anxiety groups based on their SLWAI scores. Those who scored in the top third (i.e., receiving a mean scale score above 3.41 on the five-point scale) were categorized into the high-anxiety group (n = 54); those who scored in the bottom third (i.e., receiving a mean scale score below 2.91 on the five-point scale) were classified into the low-anxiety group (n = 52). The two anxiety groups’ scores on the two writing tasks are summarized in Table 2.

Table 2

Different Anxiety Groups’ Writing Scores on the Two Rhetorical Tasks

Task Anxiety Group M SD

Narrative Task Low Anxiety 3.47 0.57

High Anxiety 3.13 0.66

Expository Task Low Anxiety 3.32 0.61

High Anxiety 3.07 0.86

(22)

Two-way repeated measures ANOVA indicated a significant main effect of anxiety, with F1, 104 = 5.70, p = .02, in favor of the low

anxiety group. Specifically, the low anxiety group performed significantly better than the high anxiety group on both tasks. There was no significant interaction effect of rhetorical task type and anxiety level on writing quality, with F1, 104 = .53, p = .47. The effect

of task type on writing performance was again not significant, albeit at a marginal level (F1, 104 = 3.68, p = .058).

The fourth research question was aimed to understand the participants’ perceptions and opinions about the two writing tasks. The information was mainly gleaned from their responses to the two post-writing questionnaires and the background questionnaire. In the post-writing questionnaires, around 63% of the participants expressed that they liked the narrative task; only 25.17% of the participants showed their preference for the expository task. They also stated that the narrative task allowed them to develop their ideas more easily. Their responses to the background questionnaire further revealed that narrative writing was practiced most frequently in English classes. Besides, among descriptive, narrative, and expository writing, narrative writing was considered to be the most popular and the easiest writing task by around 60% of the participants. In contrast, the expository task was thought as the most difficult task by more than 50% of the students.

Although narrative writing was perceived to be relatively easier and was reported to be practiced more often than expository writing, the majority of the participants (approximately 66%) showed similar dissatisfaction with their performance on both writing tests in the

(23)

post-writing questionnaires. It seems that perceived ease and more practice of narrative writing did not ensure more satisfying performance on the narrative writing test. The participants’ negative self-evaluation of their own English writing performance was in line with their general perception of English writing. In the background questionnaire, more than 79% of the participants expressed that English writing was a very challenging task. They further revealed that their difficulties in writing mainly came from limited vocabulary and insufficient knowledge about proper grammatical usage. Taken together, results from the three questionnaires indicate that most of the participants had difficulty writing English even though they felt narrative writing to be relatively easier than expository writing.

DISCUSSION

The Effect of Rhetorical Task Type

This study examined the effect of two types of rhetorical tasks: narrative and expository. The results show that rhetorical task type made no significant difference in L2 learners’ writing performance although students reported practicing more narrative writing at school. The finding of non-significant effect of rhetorical task is in accordance with Cheng’s (2006) study on EFL college students in Taiwan, but it is contradictory to the findings of some L1 studies (e.g., Kegley, 1986; Prater & Padia, 1983; Reed, 1992), which indicate that different types of rhetorical tasks would lead to different writing performances.

There are at least two possible explanations for the finding. Firstly, as suggested in the background questionnaire data, English

(24)

writing itself was deemed very challenging for this group of EFL learners. The inherent difficulty of writing an extended text in a foreign language might override the potential effect of rhetorical task type. Note that the participants were writing in a timed, testing situation. The difficulty and challenge might be compounded in such a situation. As a result, the gains on performance that more practice was assumed to bring to students in writing narratives might be eliminated. Of course, caution should also be exerted in assuming that more practice will lead to better performance because “more” may not be equal to “effective.”

The non-significant task effect could also be explained in terms of EFL learners’ restricted experience and limited ability in English composition writing. Most EFL learners in Taiwan do not start their English composition writing until they enter senior high school, and this was true for most of the participants in the current study. Moreover, as reported in the post-writing questionnaires, many of the participants were still struggling with the problems of limited vocabulary and improper grammatical usage when writing English compositions. It is likely that the effect of rhetorical task type on L2 writing would not manifest itself before learners have sufficient L2 proficiency to express themselves at will.

The Effect of English Proficiency

The importance of L2 proficiency in L2 writing cannot be overemphasized. This study found that, regardless of task type, the high-proficiency group scored significantly higher than the low-proficiency group. Like the majority of the previous research

(25)

concerned with the same issue, this study lends support to those researchers who claim general language proficiency level could make a difference in the quality of learners’ writing (e.g., Chang, 1995; Lin, 2005; Reed, 1992; Sasaki & Hirose, 1996; Way et al., 2000). Since the current study and many other studies have proved that learners’ L2 writing performance is under the influence of their general L2 competence, L2 writing instructors need to recognize the importance of general L2 proficiency in the development of L2 writing skills.

The Effect of Writing Anxiety

In the present study, learners’ writing anxiety was found to have a significant effect on writing performance. Regardless of task type, the low-anxiety group performed significantly better than the high-anxiety group. This finding is in conflict with the results of Madigan et al. (1996) and Richardson (1981), which showed no relationship between writing anxiety and writing quality; but it supports the findings of many others, in which high writing anxiety showed a negative impact on writing (e.g., Daly & Miller, 1975; Faigley et al., 1981; Faris et al., 1999; Lee & Krashen, 2002). As writing anxiety involves negative thoughts about writing that can consume cognitive resources (Cheng, 2004a), it seems reasonable that learners with a high level of writing anxiety would produce essays of poorer quality, for they have diminished cognitive resources available for the writing task at hand. It is noteworthy that most of the previous studies on writing anxiety concern college students (Faigley et al., 1981; Faris et al., 1999; Lee & Krashen, 2002), whereas this study elucidates that senior high students may also be troubled by this anxiety as well. This finding calls for due attention to this affective factor in L2 learners’ writing process.

(26)

CONCLUSION

This study probed into the relationship of rhetorical task type, writing anxiety, and general English proficiency to senior high school EFL students’ English writing performance. Statistical analyses show that rhetorical task type did not account for the variation in the participants’ writing performance. In contrast, English proficiency and writing anxiety each had a significant effect on the quality of learners’ essays. The high-proficiency group outperformed the low-proficiency group; the writing grades of the low-anxiety group surpassed those of the high-anxiety group.

Qualitative data shows that more than 60% of the participants favored the narrative writing task, but the participants held diverse opinions towards the expository writing task. Probably because narrative writing was more frequently practiced in English classes, it was perceived to be easier than expository writing. However, most of the participants thought of English writing as a very challenging and difficult task. The difficulty of composing an extended text in English, particularly in a testing situation, was reflected in the result that around 66% of the participants were not satisfied with their performance on both of the narrative and expository writing tasks. This result is congruent with the finding that their performance on the two tasks did not differ significantly.

Some implications can be drawn from this study. First, the observed significant effects of English proficiency and writing anxiety on English writing performance suggest a need to place a high priority on developing learners’ general English competence in

(27)

tandem with English writing skills and attending to their anxiety about English writing. Particular attention should be paid to strengthening students’ active vocabulary and productive grammatical knowledge, a lack of which was identified by the participants as the major obstacle to English writing. Besides, teachers should help to ameliorate students’ writing anxiety. For instance, teachers should create a friendly learning environment, where anxious writers can move at their pace without worrying too much. Teachers may also need to teach anxious students some relaxation techniques to cope with anxiety under stressful situations, such as in a test, so that anxiety will not seriously impede their performance. Interested teachers could consult Cheng (2004b) and Leki (1995, 1999) for further suggestions on reducing learners’ writing anxiety.

Second, for the testing institutions, the findings suggest a second thought on restricting different writing tests to particular types of rhetorical tasks. To be specific, this study found that it was general English proficiency, rather than rhetorical task type, that made a difference in learners’ writing performance. This result casts doubt on the necessity for the College Entrance Exam Center to set a different priority on rhetorical task types for the GSAET and the DRET, with the former prioritizing narrative writing tasks and the latter, expository tasks. Because the GSAET is a prerequisite for applying for college admission but not every senior high graduate needs to take the DRET, the backwash effect of prioritizing narrative writing in the GSAET might lead to disproportionate instruction and practice of narrative writing in senior high schools. Other types of rhetorical tasks, such as expository, descriptive, and argumentative writing, might receive less attention in English classes. In fact, the

(28)

participants’ responses to the background questionnaire have confirmed this backwash effect on English writing instruction in senior high schools. In the long run, this backwash effect might impede the development of the students’ overall English writing skills. Therefore, the testing agencies need to be careful about the potential negative impact on teaching that their test guidelines could cause.

This study gives some important insights into English writing assessment and instruction in Taiwan. Nonetheless, certain limitations should be recognized. First, the participants in this study came from four intact classes taught by two different teachers. Differences between the two teachers in English writing instruction that might have affected the participants’ performance were not controlled. Second, the generalizability of the findings is relatively limited. For one thing, most of the participants were on the science track. For another, all participants in the present study came from the same elite senior high school. Whether the performance of students on other academic tracks or from other senior high schools is affected by different types of rhetorical tasks remains unknown. Third, due to the packed teaching schedule, the interval between the first writing task and the second writing task was around three weeks, which may not have effectively prevented the participants from memorizing what they wrote for the first task. Lastly, interpretations of the results are limited to holistic scoring of writing, which sees writing performance as “a whole entity” composed of “the parts and their relationships” instead of “the sum of the parts” (Goulden, 1992, cited in Barkaoui, 2007, p. 87). Although holistic scoring has the merits of being efficient and yielding high reliability (Barkaoui, 2007), its

(29)

impressionistic nature may lead the raters to assess learners’ written texts on “impressionistic criteria” rather than those specified in the scoring rubrics. Moreover, holistic scoring fails to distinguish different dimensions or traits of the writing, such as lexical breadth and rhetorical organization. This scoring procedure also fails to take into account the effect that raters’ knowledge about different rhetorical texts might cause. The results on rhetorical task effect might have been different if multiple-trait scoring that focused on rhetorical features of writing had been adopted or the raters’ knowledge about the two rhetorical texts had been considered in the research design. In light of these limitations, future investigators with sufficient resources and support could consider controlling teachers’ and raters’ effects, inviting more schools and students on different academic tracks to join the study, extending the interval between the first and the second writing tasks, and/or resorting to multiple trait scoring or a combination of holistic and multiple trait scoring procedures.

REFERENCES

Ackerman, T. A., & Smith, P. L. (1988). A comparison of the information provided by essay, multiple-choice, and free-response writing tests. Applied Psychological Measurement, 12, 117-128. Bachman, L. F. (1990). Fundamental considerations in language

testing. Oxford: Oxford University Press.

Barkaoui, K. (2007). Rating scale impact on EFL essay marking: A mixed-method study. Assessing Writing, 12, 86-107.

(30)

Bereiter, C., & Scardamalia, M. (1987). The psychology of written

composition. Hillsdale, NJ: L. Erlbaum Associates.

Berman, R. A., & Nir-Sagiv, B. (2007). Comparing narrative and expository text construction across adolescence: A developmental paradox. Discourse processes, 43, 79-120.

Chang, V. W. (1995). Freshman English composition: An error

analysis from the discourse perspective. Taipei, Taiwan: The Crane Publishing Co.

Cheng, Y.-s. (2004a). A measure of second language writing anxiety: Scale development and preliminary validation. Journal of

Second Language Writing, 13, 313-335.

Cheng, Y.-s. (2004b). EFL students’ writing anxiety: Sources and implications. English Teaching & Learning, 29(2), 41-62.

Cheng, Y.-s. (2006). The effects of writing anxiety on writing

performance and self-efficacy. Paper presented at Asia TEFL 2006 International Conference, Fukuoka, Japan.

College Entrance Examination Center. (2007). Introduction to tests. Retrieved May 10, 2007, from http://www.ceec.edu.tw

Cooper, P. L. (1984). The assessment of writing ability. (GRE Board Report No: 82-15R). Princeton, NJ: Educational Testing Service. Crowhurst, M. (1978). Syntactic complexity in two modes of discourse

at grades 6, 10, and 12. (ERIC Document Reproduction Service No. ED168037).

Cumming, A. (1989). Writing expertise and second-language proficiency. Language Learning, 39, 81-141.

Cumming, A., Kantor, R., Powers, D., Santos, T., & Taylor, C. (2000).

(31)

Monograph 18. Princeton, NJ: Educational Testing Service. Daly, J. A., & Miller, M. D. (1975). Further studies on writing

apprehension: SAT scores, success expectations, willingness to take advanced courses, and sex differences. Research in the

Teaching of English, 9, 250-256.

Faigley, L., Daly, J. A., Witte, S. P. (1981). The role of writing apprehension in writing performance and competence. Journal of

Educational Research, 75, 16-21.

Faris, K. A., Golen, S. P., & Lynch, D. H. (1999). Writing apprehension in beginning accounting majors. Business

Communication Quarterly, 62, 9-22.

Goulden, N. R. (1992). Theory and vocabulary for communication assessments. Communication Education, 41, 258–269.

Hamp-Lyons, L. (2003). Writing teachers as assessors of writing. In B. Kroll (ed.), Exploring the dynamics of second language writing (pp.162-189). New York: Cambridge University Press.

Hughes, A. (1989). Testing for language teachers. New York: Cambridge University Press.

Hyland, K. (2003). Second language writing. New York: Cambridge University Press.

Jeffery, J. V. (2009). Constructs of writing proficiency in US state and national writing assessments: Exploring variability.

Assessing Writing, 14, 3-24.

Jeng, H. H., Lin, C. C., Chen, C. R., Hsiao, T. Y., & Curran, J. E. (2001). A study of department required English test and

scholastic aptitude English test. Taipei, Taiwan: College Entrance Examination Center.

(32)

Kegley, P. H. (1986). The effect of mode discourse on student writing performance: Implication for policy. Educational Evaluation and

Policy Analysis, 8, 147-154.

Koda, K. (1993). Task-induced variability in FL composition: Language-specific perspectives. Foreign Language Annals,

26(3), 332-346.

Language Training and Testing Center. (2005). The official guide for

FLPT review. Taipei, Taiwan: LTTC.

Lee, S. Y., & Krashen, S. (2002). Predictors of success in writing in English as a foreign language: Reading, revision behavior, apprehension, and writing. College Student Journal, 36(4), 532-543.

Leki, I. (1995). Academic writing: Exploring processes and strategies. New York: St. Martin’s Press.

Leki, I. (1999). Techniques for reducing second language writing anxiety. In D. J. Young (Ed.), Affect in foreign language and

second language learning: A practical guide to creating a

low-anxiety classroom atmosphere (pp. 64-88). Boston:

McGraw-Hill.

Lin, P. H. (2005). A study of types on EFL writing assessment for

vocational high school students in Taiwan. Unpublished master’s thesis, National Yunlin University of Science and Technology, Yunlin, Taiwan.

Madigan, R., Linton, P., & Johnson, S. (1996). The paradox of writing apprehension. In C. M. Levy & S. Ransdell (Eds.), The science of

writing: Theories, methods, individual differences, and applications

(33)

Prater, D., & Padia, W. (1983). Effects of modes of discourse on writing performance in grades four and six. Research in the

Teaching of English, 17, 127-134.

Qiu, Z.-h. (2006). Quantitative research and statistical analysis in social

& behavioral sciences (3rd ed.). Taipei, Taiwan: Wu-Nan Co., Ltd. Quellmalz, E. S., Capell, F. J., & Chou, C. P. (1982). Effects of

discourse and response mode on the measurement of writing competence. Journal of Educational Measurement, 19, 241-258. Raimes, A. (1987). Language proficiency, writing ability, and

composing strategies: A study of ESL college student writers.

Language Learning, 37, 439-468.

Reed, M. W. (1992). The effects of computer-based writing tasks and mode of discourse on the performance and attitudes of writers of varying abilities. Computer in Human Behavior, 8, 97-119. Richardson, E. M. (1981). Research on writing: Apprehension, quality,

and audience. (ERIC Document Reproduction Service No. ED 204 813).

Sasaki, M., & Hirose, K. (1996). Explanatory variables for EFL students’ expository writing. Language Learning, 46, 137-174. Stiggins, R. J. (1981). A comparison of direct and indirect writing

assessment methods. (ERIC Document Reproduction Service No.

ED204413).

Way, D. P., Joiner, E. G., & Seaman, M. A. (2000). Writing in the secondary foreign language classroom: The effects of prompts and tasks on novice learners of French. Modern Language

Journal, 84, 171-184.

Weigle, S. C. (2002). Assessing writing. Cambridge, U.K.: Cambridge University Press.

(34)

Wolfe, M. B. W. (2005). Memory for narrative and expository text: Independent influences of semantic associations and text organization. Journal of Experimental Psychology: Learning,

Memory, and Cognition, 31, 359-364.

Zamel, V. (1982). Writing: The process of discovering meaning.

TESOL Quarterly, 16, 195-209.

ABOUT THE AUTHORS

Pi-chen Tsai received her master’s degree from the TEFL program in the Department of English at National Taiwan Normal University. Her research interests focus on English writing and language testing. Pi-chen Tsai now serves as an English teacher at Taipei Municipal Da-Zhi High School, Taiwan.

Yuh-show Cheng received her Ph.D. degree in TESOL from the University of Texas at Austin, USA. She is currently a professor of English at National Taiwan Normal University in Taiwan, where she has been teaching various courses related to English language skills, English teacher education, and second language acquisition for more than 12 years. Her research interests include individual differences in second language acquisition, EFL writing instruction/assessment, and foreign language teacher education.

(35)

APPENDIX A

Prompts for the Two Writing Tasks (Original Version)

(A) The prompt for the narrative task (in Chinese)

說明:1. 依提示在「答案卷」上寫一篇英文作文。 2. 文長至少 120 字。 提示:閒暇時間大家都會從事一些自己喜歡的休閒活動,在你所 從事過的休閒活動中,哪一次讓你印象最深刻呢?請用說說說說 故事 故事 故事 故事的方式敘述你最難忘的一次休閒活動經驗,敘述內容 應包括你從事這項休閒活動的時間、地點、過程,以及其 他細節。

(B) The prompt for the expository task (in Chinese) 說明:1. 依提示在「答案卷」上寫一篇英文作文。 2. 文長至少 120 字。 提示:有人說,男女大不同,喜歡的休閒活動也不太一樣。請依 據你對男女休閒活動的瞭解,寫一篇短文。文分兩段,第 一段說明男生和女生喜歡的休閒活動有哪些不同,第二段 說明造成男生和女生對休閒活動各有所好的原因。

(36)

APPENDIX B

Prompts for the Two Writing Tasks (English Version)

(A) The prompt for the narrative task

Directions: Following the prompt, write on the answer sheet an English essay of 120 words or more.

Prompt: Almost everyone does some recreational activities in his/her free time. Among those recreational activities that you have taken part in, which impresses you most? Please provide an account of the most unforgettable recreational activity that you have been engaged in, in a manner of story telling. In the narration, you have to specify the time, the setting, the sequence of events involved, as well as other relevant details.

(B) The prompt for the expository task

Directions: Following the prompt, write on the answer sheet an English essay of 120 words or more.

Prompt: It is said that boys and girls are different in many aspects, including preferences for recreational activities. Based on your understanding of the issue, please write a short essay on gender differences in recreational activities.The essay should include two paragraphs. In the first paragraph, illustrate how the two genders differ in their preferences for recreational activities. In the second paragraph, explain what may cause such gender differences.

(37)

文體類型

文體類型

文體類型

文體類型、

、英語文能力和寫作焦慮

英語文能力和寫作焦慮

英語文能力和寫作焦慮

英語文能力和寫作焦慮

對高中生英文作文表現的影響

對高中生英文作文表現的影響

對高中生英文作文表現的影響

對高中生英文作文表現的影響

摘要 摘要摘要 摘要 在大學入學考試中心所主辦的學科能力測驗(學測) 和指定科目考試(指考)裡,英文寫作是英文考科必 考項目;但學測和指考的英文寫作所著重的文體不盡 相同。學測英文寫作以敘事文為主,而指考的英文寫 作則重說明文和描述文。由於學測所設定評量的知識 和能力比指考來得基礎,學測和指考英文寫作所著重 文體之不同似乎表示試務單位假定敘事文比說明文來 得容易。為了要驗證這種假定,本研究試圖比較敘事 文和說明文兩種文體對高中生英文寫作表現之影響。 有鑒於英語文能力和寫作焦慮對英文寫作也可能造成 影響,本研究同時探究上述兩種文體類型與英語文能 力或寫作焦慮的交互作用。此外,本研究也收集學生 對不同英文寫作文體的看法。研究結果顯示文體類型 的主要效果及其與英語文能力或寫作焦慮的交互作 用,皆未達顯著水準;但英語文能力的優劣和寫作焦 慮的高低則分別對學生的英文寫作表現有顯著的影響。 關鍵詞:文體類型 英語文能力 寫作焦慮

參考文獻

相關文件

- Settings used in films are rarely just backgrounds but are integral to creating atmosphere and building narrative within a film. The film maker may either select an already

After students have had ample practice with developing characters, describing a setting and writing realistic dialogue, they will need to go back to the Short Story Writing Task

Using Information Texts in the Primary English Classroom: Developing KS2 Students’ Reading and Writing Skills (New). Jan-Feb 2016

incorporating creative and academic writing elements and strategies into the English Language Curriculum to deepen the learning and teaching of writing and enhance students’

 A genre is more dynamic than a text type and is always changing and evolving; however, for our practical purposes here, we can take genre to mean text type. Materials developed

Making use of the Learning Progression Framework (LPF) for Reading in the design of post- reading activities to help students develop reading skills and strategies that support their

• School-based curriculum is enriched to allow for value addedness in the reading and writing performance of the students. • Students have a positive attitude and are interested and

To help students appreciate stories related to the theme and consolidate their knowledge and language skills in writing stories, the English Club has organised a workshop on story