An Overview of English Writing Research in Taiwan

This study systematically investigates the English writing research in Taiwan, over the span of time from 1989 to 2008, a 19-year time period. Data collection consisted of five major sources. Guided by Juzwik et al’s (2006) study, the data were analyzed based on the general problems under investigated, the age groups being researched, the methodologies being implemented, and types of research being conducted at different grade levels. Findings revealed that writing instruction, writing and technologies and peer evaluation were the most studied problems in writing research whereas collaborative writing, error analysis, and cultural influences were the least studied problems. The most studied populations were university and senior high school students while the least studied groups were kindergarteners and adults. Most studies were conducted by using qualitative methodology. Writing and technologies was the most studied type of research among elementary school students and university students, whereas writing instruction was frequently studied among senior high school students, graduate students and adult students. The implications and recommendations that emerge out of these results provide possible agendas for writing teachers, researchers and policy makers worldwide.


Introduction
Among the systematic review of English as the first language (L1) writing research, several studies have been published in the United States.However, systematic review of English as a foreign language (EFL) writing research has seldom become the central focus of literature.The reason could be that the review of EFL writing research would involve many non-English speaking countries around the world; therefore, to systematically and thoroughly examine the abundance of published articles related to EFL writing would be a difficult task.Nevertheless, systematical examination of EFL writing research from an EFL speaking country would help fill the gap in the body of FL writing research.This study systematically analyzes and investigates EFL writing research outside the United States.It focuses on the trends and foci of English writing research in Taiwan over the span of time from 1989 to 2008, a 19-year period.The purpose of this study is threefold.First, it helps to fill the gap in the body of foreign language writing research and helps writing instructors, writing researchers, and curriculum developers worldwide to develop the understanding of English writing research in an EFL learning environment.Second, Chinese writing researchers tend to have partial understanding about the bits and pieces of information concerning English writing research conducted in their country.This paper is intended to assist writing researchers to familiarize themselves with research conducted within nearly two decades in Taiwan.Third, it points out a possible future direction for EFL writing research agendas.

Review of literature
Previous research on the overview of English writing in L1 and L2 was conducted by a few scholars in the United States.With regard to the overview of EFL writing research, little information can be acquired in the body of existing literature.It is important to be acquainted with the overview of L1 and L2 writing research because FL writing research has been heavily influenced by L1 and L2 writing research (O'Brien, 2004).Furthermore, English language writing instruction for native speakers has also influenced EFL writing instruction (You, 2006).Accordingly, the writers of L2/EFL are very similar to English native speakers in the respect of learning to write (Williams, 2005).
Among the L1 writing researchers, Hillocks (1984) first conducted a meta-analysis of experimental treatment studies on teaching composition in the United States.The study was an integrative review of experimental studies from 1963 to 1982.In this review, Hillocks investigated experimental studies conducted over this time span by including studies that involved treatments, uses of writing quality scale applied to writing samples, exercises of minimal control for teacher bias, controls for differences among groups of subjects, and insurances on the validity and reliability of the scored compositions.In synthesizing the experimental studies, Hillocks examined the treatments, including four modes of instruction: presentational mode, natural process mode, environmental mode and individualized mode.Hillock's study provides us an insightful view pertaining to what factors of effective writing instruction should be taken into account when teaching writing.
Following Hillocks' studies, Herrington (1989) also published an article entitled the First Twenty Years of Research in the Teaching of English and the Growth of a Research Community in Composition Studies.By using Research in the Teaching of English as the primary source, her historical reflection focused on four periods, inclusive of the initial years of each editor.They were from 1967to 1969, under the editorship of Richard Braddock;from 1973-1975, Alan Purves;from 1977to 1979, Roy O'Donnell;andfinally from 1984 to 1986, Judith Langer and Arthur Applebee.During the first three years, Herrington found the articles were quantitative studies.They focused on investigating writing as texts, describing written texts linguistically or grammatically and evaluating written texts.The next three years was during the "troubled dream" period when researchers were questioning the quantitative research method and called for more qualitative research studies.At the same time, more articles were found to focus on sentence production and measuring writer's development in terms of syntactic complexity.Increased attention was also found on studies about the writing process.The third period of the study found several articles examining "writing as a sentence-generating activity."They included sentence-combining effectiveness, syntactic maturity measurement, and issues relevant to studying syntactic methodology.The authors also questioned "our values as teachers and readers of our students' writing" as well as writing processes.The last period covered in this study demonstrated the findings of using computers to write or to evaluate writing, extending more inquiry in syntactic features and writing quality, and viewing writing from a social perspective approach.Other studies also examined the differences between "basic" and "non-proficient" or "competent" and "proficient" writers.
The next study, conducted by Durst (1990), charted the terrain of writing research by analyzing the empirical studies of composition from 1984 to 1989.Durst analyzed the research on composition by examining what had been learned from the data collected.He specifically emphasized the growth areas issue, less interesting topics, and and topics of more interest.Durst discovered that from a total of 1,577 studies, 969 focused on composition.These studies were categorized into instruction, processes, text analysis, contexts, assessment, status surveys, writing and learning, and rhetoric.Durst further pointed out that almost 90 percent of the research focused on the first five categories.Durst also exclaimed that in his review, he found only about five percent of the research studies centered on minority writers.Additionally, most studies had the tendency of using elementary and college students as their research subjects whereas students from middle and high schools were largely ignored.Furthermore, he pointed out that the number of writing research in contextual studies was relatively small.From his review, Durst informed us of the future direction of research agenda.
Building upon Durst's study, the recent study conducted by Juzwik et al. (2006) took a broader look at the scope of recent research on writing, investigating the studies from 1999 to 2004.In this study, they examined the current trends and foci in writing research in the United States by looking at writing research from three dimensions: problems studied, population age groups studied and methods used.They searched 4,739 articles from ERIC, PsychINFO and Linguistics and Language Behavior Abstract databases but there were only 1,502 relevant articles analyzed.The result showed that, apart from using interpretive methodology in most of the research studies, the most currently studied problems in writing were context and writing practices; multilingualism, bilingualism, and writing; and writing instruction.In contrast, the least researched problems were writing and technologies, writing assessment and evaluation, and relationships among literacy modalities.Findings also indicated that the populated age groups being studied were undergraduate, adult, and postsecondary students, while the least studied age groups were preschool-aged students and middle and high school students.The result of Juzwik et al.'s study provides useful information for contemporary writing researchers and educators.
In L2 and EFL writing research, Silva and Brice (2004) examined published articles regarding research in teaching writing from 2000 to 2004 by including referred journal articles, book chapters, and books.Their study overviewed 1) basic research from composing processes, written texts, and assessment, and 2) applied research from content-based writing instruction, voice and identity, reading and writing, computers and technology, grammar and vocabulary, peer interaction, plagiarism, teacher response, and literature and film.They also discussed general issues and concerns, and pointed out the future direction for researchers.Liou (2008) had recently presented a paper about EFL writing research in Taiwan, focusing only on three years she surveyed papers published on one local journal and proceedings of three English teaching conferences.Liou discovered that 46.9 % of the studies centered on instruction-oriented research, 32.4% focused on text-oriented research, 18.9% were on writer-oriented research and only 1.8% was on reader-oriented research.Liou's small scale study presented a rough picture of English writing research in Taiwan.
The above writing research in L1 and L2 covered studies on historical review (Hillocks, 1984;Herrington,1989), growth and interested research areas (Drust, 1990), teaching pedagogy (Silva and Brice, 2004), and the examination of current trends and foci (Juzwik et al.;2006).To date, there is no thorough historical review of English writing research that examines the writing trends and foci in Taiwan.This paper aims to make significant advances in building a more comprehensive awareness in EFL writing research.

Data collection
In order to make this task manageable and to present an overview of English writing research from 1989 to 2008 in Taiwan, two research assistants and I set limitation on the number of sources ( 5 altogether) from which we investigated the data: (1) Dissertation Abstracts International, (2) Electronic Theses and Dissertations System in Taiwan, (3) EBSCO Databases, (4) ERIC Databases, and (5) Education Full Text, The primary inclusion of these data sources was that authors usually considered these sources (with the exception of Dissertation Abstracts international and Electronic Theses and Dissertations System in Taiwan) as the major outlet for publishing their work internationally.In addition, readers would have easier access to these sources from most of the school libraries around the world.Furthermore, the studies discovered from the five data sources in one way or another have contributed an in-depth treatment in the specific area of English writing in Taiwan.
In general, studies which were conducted in Taiwan were included in the data analysis.In addition, both key words: "writing" and "Taiwan" were used in the databases to make sure the relevant studies were not being overlooked.The data collected from the five major sources sometimes overlapped.Extensive time was spent in order not to include the same theses or articles for data analysis.Table 1 shows the number of articles found from the five major data sources.

Coding
Guided by Juzwik et al's (2006) study, the data were analyzed based on the general problems under investigated, the age groups being researched, the methodologies being implemented, and types of research being conducted at different grade levels.In analyzing the collected data with regard to the problems under investigated, two research assistants (a Ph.D. student from the United States and the other, a Ph.D. candidate from Taiwan) helped worked on and created the categorization.The categories were coded preliminarily based on the title of each article.However, as Durst (1990) has discovered, "Many studies cut across categories; for example, much of the research on writing instruction focuses on the use of writing -process-based pedagogies in teaching composition.A number of writing-instruction studies examine classroom contexts or employ text analysis.In these cases of overlap, studies were categorized based on the dominant focus of the research (p.395)."Thus, in this study, if the title contained peers and computers, the original article would be located and reviewed carefully so as to decide whether it belonged to peer evaluation, or writing and technologies.For example, the study "Students' use of ideas provided by peers during prewriting discussions conducted on networked computers" was coded in the category of peer evaluation rather than writing and technologies.It was found from the original study that the major concern of the research was to examine the writing generated from the talk with their peers during computer-mediated prewriting discussions, and the quality of the peers' comments.Another example "The use of process writing and Internet technology in class: A focus on peer reviews" was not put in the category of writing process or writing and technologies because peer review was the major focus.The coding description for each researched problem can be found in table 2.

(Insert table 2 here)
Similar to the study conducted by Juzwik et al., the age categories were coded based on the level of schooling.In Juzwik et al.'s study, the age category ranges from prior to school to adults.They were prior to school, elementary school, middle school, high school, undergraduate postsecondary and adults.However, in this study, since the educational system in Taiwan is quite different from that of the United States, we coded 12 age groups from the data.They were kindergarten, elementary school, junior high school, senior high school, vocational high school, junior college, military college, university of science and technology, medical college, regular university (science and technology are not the major focus of the university), graduate school, and adults (non-academic writers).
In terms of methodology being implemented, if the title contained a case study, it would be automatically put in the category of qualitative study.In addition, if the title of the study did not have any methodological phrase, the abstract was examined in detail to detect the use of methodology.Furthermore, if the research methodology or subjects were not clearly stated in the abstract, the original study would be located and examined in detail.Finally, if the study was put in the category of both qualitative and quantitative, it means that somewhere in the study, the author had stated that it was a blended study.If not, each abstract or the study itself would be examined carefully to determine which category it belonged in.

The general problems investigated
As a result of analyzing the collected data, the data was divided into fourteen categories.They were writing instruction, writing and technologies, peer evaluation, contrastive rhetoric, writing and metacognition, writing processes, e-mail exchanges, responding to student writing, writing assessment, writing anxiety, journal writing, collaborative writing, error analysis, cultural influences, and others.
From the five major sources of data analysis in this review, the most studied problems on writing research in Taiwan tended to focus on writing instruction, writing and technologies, and peer evaluation, whereas the problems least studied were error analysis and cultural influences.The general problems investigated by this writing research are shown as follows.

The population studied
In Taiwan, university students were found to be the most frequently studied age group in existing English writing research.This age group was dominated as it was examined almost twice as much as the next frequently researched group, the senior high school level.Kindergarteners, military college students, and medical college students were found to be the least studied groups.Other students, such as vocational high school students and non-academic adults seldom became the subjects of writing research studies.The following table summaries the findings of the age groups being studied in Taiwan.

The methodology used
From the data analyzed, with regard to methodology, it was important to note that both quantitative and qualitative research methodologies were almost equally adopted by writing researchers in Taiwan.The data indicated that quantitative method was implemented in 40% of the studies while qualitative method was implemented in 42% of the studies.18% of the studies were found using both qualitative and quantitative methods.

Types of research conducted at different grade levels
From the five major data sources, there was only one study concerning the use of computer technology at kindergarten level.In terms of elementary schools, writing and technologies was still the major type of research, followed by writing instruction.With regard to junior high school levels, of the nine studies, three were about journal writing with the rest concerning writing instruction.Almost half of the research (20) conducted at senior high school level was still dominated by writing instruction.The rest of the studies consisted of writing and technologies, writing and metacognition, writing process, peer evaluation, teacher feedback, e-mail exchanges and journal writing.At vocational high school levels, writing assessment, error analysis, peer evaluation, writing instruction, and contrastive rhetoric were found from the data analysis.Research from the junior, military, medical college levels, and university of science and technology level showed writing and technologies, writing and metacognition were two major types of research.As for the regular university level, writing and technologies, peer evaluation, and contrastive rhetoric were the most studied types of research.At the adult level, two out of the four studies were writing and instruction, the third was e-mail exchanges and the fourth was collaborative writing.At the graduate level, writing and instruction, writing and technologies, and writing process were the most researched topics.The result of the analysis can be found in Appendix A.

Discussion
Focusing on the literature dated from 1989 to 2008, a 19-year time span, this paper examined the trends and foci from the collected data.It investigated the major problems, populations, methodologies and types of research conducted at different age groups in EFL writing research in Taiwan.The findings pointed out several important issues that deserve public attention.
First, in this study, data sources were drawn from Dissertation Abstracts International, Electronic Theses and Dissertations System in Taiwan, EBSCO Databases, ERIC Clearinghouse, and Education Full Text.The result showed a total of 188 studies, all conducted by Taiwanese researchers, on writing research.Among these 188 studies, only a small amount of research studies were published in the internationally referred journals.The rest were theses, dissertations, and reports.The small number of the English writing research articles indicates that more research in English language writing field is further required.This study advocates the growth of scholarship on EFL writing in Taiwan.
Second, from the findings concerning the problems being investigated in this paper, it has demonstrated that more and more writing instructors have become interested in using computer technology to help students compose in English.Given the rapid change of technology in modern times, computer technology has shortened the distance between countries around the world and it has opened up new and exciting possibilities for instructors to teach EFL writing.In recent years e-mail exchanges, online English learning, Plug-in Learning Object Module, Weblog networks implementation have been the central focus of EFL writing literature.It is expected that research in this field will continue to grow in EFL writing research in Taiwan.
Third, the age group being researched most was at the college level and the subjects being studied were mostly from English majors at colleges.This is partly because writing researchers in Taiwan are university professors and they tend to use their students as research subjects.It is also partly because writing is a mandatory course for English majors at universities in Taiwan.This study suggests curriculum developers in Taiwan should consider making non-English major college students take English writing as an obligatory course.This study also points out that increased amount of attention should be devoted to other age group of student writers, especially students from kindergarten, junior high school, vocational high schools, military college, medical college.Furthermore, apart from the above-mentioned age groups, master's students, doctoral students, foreign students, aborigines and adults have received scant attention from EFL writing researchers in Taiwan.More research in these fields also needs to be explored.
Fourth, the study revealed that within the past two decades, qualitative research methodology is being more frequently adopted by writing researchers in Taiwan.Both qualitative and quantitative methods were also used at the same time by 18% of the researchers to conduct their research.The increased number of qualitative research studies indicates that qualitative research has become a significant type of methodology in writing research in Taiwan.Researchers were no longer interested in testing hypotheses, instead, they turned to observe the class or participate in class, interview the students, and analyze the documents and materials so that more in-depth insights and contexts could be obtained from the writing class.
Fifth, from the collected data, research on writing anxiety was only found at college level, it is suggested that more research should be conducted by using younger students, including kindergarteners, elementary, and high school students as research subjects.These groups of students would probably have writing anxiety when asked to compose in English.In addition, from the data collected, of the six studies concerning writing assessment, only four portfolio assessment studies were found in one elementary school, one junior high school and two senior high schools.It is suggested that more research should focus on this field because learning to write is a long term process and it takes time and effort to accomplish a good written product.This study raises the issue of writing assessment for FL educators to think over.
Finally, this study has investigated the trends and foci of EFL writing research from five major sources within nearly two decades in Taiwan.While the study has provided important findings, future research should include more sources, including book chapters, books, local journal articles, and local conference papers devoted to adding information and insights to the existing body of EFL writing research currently available regarding Taiwan.