Prewriting and Drafting Strategies of Graduate Students in Writing Term Papers in English

Researches on writing have found that some writers are successful while others are not. Studies that focus on the texts students produced were unable to reveal the reasons for the phenomenon. Researches on the process of writing attempt to uncover the mental processes that students experience while writing texts. Findings by researchers on mental processes predict that more successful writers employ a battery of effective strategies while writing and they employ a more recursive steps in producing the texts. The present study is an attempt to uncover the mental processes of two graduate students when they were writing papers as a form of academic assignment. The research employed descriptive qualitative design. The data were collected using think-aloud protocol in which the subjects think aloud what came to their mind during the production of the papers and it was tape-recorded. The recorded think-aloud was then transcribed and analysed by categorizing the utterances into cognitive and metacognitive strategies, and by depicting the utterances into a chart that depicted the subjects’ flow of thought during prewriting and drafting stages. The data were collected from 22 and 24 sittings of writing the papers. The result of data analyses shows that: (1) the subjects employ both cognitive and metacognitive strategies when writing, (2) the subjects create mental outline in the prewriting stage and keep on revising it during the drafting stage, and (3) recursiveness is not a mark of successful writers.


Introduction
The ability to write in English is necessary for tertiary level students who pursue their education in academic institutions where English is used as the medium of instruction and evaluation.In these institutions students learn subject matters through English and their mastery of the subject matter are also evaluated through English.To assess the students' level of mastery, lecturers in tertiary education assign students to write short answers to exam questions, to produce summaries, to make reports, and to write term papers.To be able to do the assignments, students need to master the content and the ability to show their mastery in written word.

LSCAC Conference Proceedings
Producing a successful written text is a complex task that requires simultaneous control over a number of language systems.It is not only language systems that a writer has to take care of, but he has to pay attention to content, or the topic he is going to write about, how it is going to be organized, who the audience or the expected reader, for what purpose the text is written, and what processes the writer must undergo to produce it.In other words, when producing a piece of text, The writer has to deal with content, audience, purpose, syntax, grammar, mechanics, organization, and the process of writing (Raimes, 1983).
In producing a scientific written work like term papers students generally follow some stages which include selecting the topic, gathering materials, planning the paper, and writing it down [15].According to [12] there are five steps in producing a scientific written work: selecting the topic, gathering information and materials, evaluating the information and materials, working out ideas and sub-ideas, and writing down.The stages of writing paper mentioned previously can be compressed into three stages of writing, i.e., rehearsing or prewriting, drafting, and revising, following (Murray, 1972) as cited by [1].
Rehearsing, or prewriting, involves finding a topic, thinking about the topic, letting ideas interact, develop and organize themselves; and thinking about the audience and the purpose of the writing task.Drafting involves getting ideas onto paper in a rough form.The writer sketches out an idea, examines it, and follows it through for a while.What has been written serves to generate further ideas, plans, and goals.The writer may also go back to the rehearsing phase and alternate between the rehearsing and drafting phases.Revising involves evaluating what have been written and making deletions and additions as necessary.
Based on the cognitive theory of writing process, (Flower and Hayes, 1980) as cited by [8] say that the drafting phase as identified by Murray consists of three recursive processes: planning, translating, and reviewing.At the planning stage the writer generate materials and organize them.This conceptual planning is termed high-level processes.At the translating stage, the writer concerns with low-level processes, that is generating sentences that express his conceptual plans and this activity results in visible, written text.Reviewing is the stage where the writer rereads the text he has written so far and uses it to generate new materials or new plans.He may also edit the text.
The three processes mentioned move through a recursive manner.The writer may edit the materials even before they are written, or from translating he returns to planning when problems occur.For example, when he is unable to convert ideas that he has generated during planning into appropriate linguistic forms because of his limited linguistic resources, he may go back from the translating stage to planning and generate other materials for which he has language forms to express them.So the writer moves back and forth from planning to translating and to planning again, or Researches on writing have exhibited findings that some writers are successful while others do not.Studies on mental operations writers engaged in while writing show that experienced writers employ strategies that are different from those used by inexperienced writers.Studies on writing processes of inexperienced and experienced writers show that both groups are different in their planning behaviour.Experienced writers have been found to construct more planning episodes [2], (Raimes, 1987).They also spent a longer time before starting to write than the inexperienced [14].They are also more flexible than inexperienced ones in that they are willing to change their plan as they write and come up with new ideas [10] which are termed advanced and emergent planning [14].
Experienced writers also differ from their inexperienced counterparts at the drafting stage.Most inexperienced writers use L1 in planning and have to translate them into L2 so that their drafting process is slow and inefficient.Experienced writers, on the other hand, use L2 both at the planning and drafting stage so that their drafting process is more efficient.In addition, inexperienced writers are concerned to early with sentence and mechanical errors, while experienced ones are more concerned with generating and organizing ideas and save language errors for later revision, so that experienced writers usually produce longer texts than inexperienced ones at first drafting.Another difference is inexperienced writers tend to use a "what-next-strategy" in which they concern with how to produce new sentences related to the previous ones.On the other hand, experienced writers always return to high-level goal of the text when they want to continue to the next sentence.
The study was intended to answer the following questions:

Method
The study is intended to describe writing strategies employed by graduate student writers of English Education major and examine them as reflecting features of writing strategies employed by inexperienced writers when writing term papers as one of academic assignments assigned by their lecturers.The research used a qualitative approach to enable description and examination of prewriting and drafting strategies in depth.
The data in the study took the form of statements generated by the research subjects.The statements came from the verbal protocols of the subjects' think-aloud activities, those generated by the subjects during the semi-structured interviews, and those originated from unstructured interviews carried out whenever the researcher met the subjects in and around the campus building and at their lodging houses.Statements from think-aloud protocols were used to answer research problem no.1 and no. 2.
Statements from semi-structured and unstructured interviews were used to answer research problems no.1.
The subjects of the study were two graduate students of English Education Program of 2006/2007who worked as junior high school teachers of English under the ministry of religious affairs and who had to write term papers not for the sake of learning to write.The two subjects, who were nicknamed D and W, took subject-matter courses during the odd semester of 2006/2007 and were selected for two reasons: (1) As graduate students, they were expected to be able to write papers as one of Academic writing tasks generally assigned to them.(2) While graduate students were usually lecturers teaching at tertiary education, these students were teachers of English at Junior High schools.As it has been described in Chapter 1, secondary teachers in Indonesia generally have limited experience in writing academic texts.Therefore they fitted the description of inexperienced graduate student writers.Interviews with the subjects supported the expectation as described in the next paragraphs.
As the study focuses on students' writing strategies in writing term papers, one subject-matter course selected was one whose lecturer assigned his students to write term-paper as one of academic assignments to decide their final grades.The subject matter course selected was Advanced Linguistics.
Two students, D and W, were selected by considering that a naturalistic, inductive research design usually involves a small number of research subjects for a relatively longer period of time.Then, validity and reliability of the study are the results of manageability of the small number of subjects which permits greater participation of the researcher ([3]: 286).Two research subjects represented one category of writers: the inexperienced writer which was based on the results of their entrance test and an interview prior to the prewriting and drafting periods.Advanced Linguistics course was used as the setting of the study.The course was offered in the odd semester of 2006/2007 academic year.The course was selected as the setting because the lecturer assigned his students to writer term papers.Therefore, the Advanced Linguistic course was considered appropriate as the setting of the study.
As this is a qualitative study, the researcher functioned as the key instrument for data collection.Besides, three other research instruments were employed, namely, 1) a list of effective prewriting and drafting strategies, 2) think-aloud procedures, 3) interview guide, and 4) composition profile.Data collected from the employment of these three research instruments was completed with the researcher's field notes.
Data collection was conducted during the odd semester (from December 2006 to January 2007) at the graduate program of English Education Department of State University of Malang in Malang.The researcher asked the subjects to think-aloud every time they started to draft their papers.As a paper cannot normally be finished in one sitting, the think-aloud activities took place several times until the subjects finished their first draft of their term paper.Only the writing of the first draft of the whole term paper was used as the source of data.
On the whole, the data analysis of the study is done by referring to [11] flow model that consists of data reduction, display, and conclusion drawing and verification.This data analysis was done along and after data collection.Relevant data were selected and separated from irrelevant data.The selected data were in the form of statements from the interviews and generated (in verbal protocol) by the subjects (the smallest units are single-words statements), and the quality of their papers indicating the result of their text production.
The transcribed speeches or the protocols were segmented into separate writing behaviors, i.e., what the subjects do through the utterances they produce.For example, the subject is reading the sentence she has just written, and then this utterance is included in one of the basic cognitive processes as its one of subcategories.Another utterance may indicate that the subject is generating, rehearsing, or evaluating.
One of the most basic decisions in developing a coding system for analyzing protocols involves dividing them into units.To distinguish the stretches of verbalization from one another in an overall protocol, a single criterion that is based on intonation.
A unit of thought was a statement when it started with a rising intonation and a falling one the end of it.Following [4] these stretch of words occurred between the rising and falling intonation were considered as one unit of thought.
The think-aloud activities were recorded by the two subjects in their lodging houses and they were kept in six 60 minutes cassettes.Two cassettes were transcribed by an assistant and were later rechecked by the researcher since some parts were missing.Four other cassettes were transcribed by the researcher herself.

Findings and Discussion
The results of the data analysis reveal the following points.At prewriting stage D did topic selection and external sources collection.She did not read and study the materials and did not do concept mapping, did not narrow the topic, did not create a thesis statement and did not make an outline.W did topic selection, external sources collection, read and studied the materials, talked to collaborators for understanding of what topic was appropriate and what purpose was expected did concept mapping and made an unelaborated outline.He did not create a thesis statement.At drafting stage D did planning, translating, and reviewing.In planning, D employed resourcing strategy most of the time and did copying or "textual borrowing" most of the time.At translating, D did rehearsing to rephrase the copied sentences, but when she reported the small study on pronunciation she conducted, she created her own sentences.When reviewing, D mostly reviewed what she had written previously, and used it to continue drafting.She never reviewed the overall goal or purpose of the paper for which she did not have one since no thesis statement was made.At drafting stage W did planning, translating, and reviewing.W's planning was guided by his outline.He generated materials to be written from him and external sources.He used the retrieved materials to support his statements.When translating he did rehearsing as he formulated his own sentences and did editing after rereading what he had written.Like D, W did reviewing by rereading what he had written previously to continue drafting.He never reviewed the overall goal or purpose of the paper which he could not possibly do since no thesis statement was made.D and W employed a number of local strategies that helped them a) to interact with the materials (Cognitive Strategies), b) to perform control or monitor to the drafting process (Metacognitive Strategies), c) to manage the process of converting ideas to visible sentences (Writing Strategies), and d) to indicate problems, awareness of audience, and to interact socially with people other than themselves (Social and Affective Strategies).The papers produced by D is considered low in quality because a) most of the paragraphs were borrowed from textbooks, so it was plagiarism; b) some important parts were missing, such as a thesis statement, the reason why the study was conducted, and focus of the paper, and c) the abstract was a patchwork of sentences taken from different sources.The paper produced by W was also low in quality but was in some way better from D's because it was not plagued by obvious plagiarism.

Prewriting Strategies
When inexperienced graduate students prepared for writing term papers, they tended to employ an interactive or emergent approach.The topic was not worked out elaborately.One subject, D, did select topic and collected external written sources, but she did not read and studied them before drafting, did not do concept mapping, did not make an outline, and did not create a thesis statement.The other subject, W, did select and change topics for four times, talked to collaborators for information on the audience and the purpose of the paper, collected and studied external written sources, did concept mapping, and wrote an outline which was not elaborated and without a thesis statement.
With regard to the effective prewriting strategies, strategies employed by D at prewriting were not effective for at least three reasons.First, her prewriting strategy was confined to finding a topic and collected five textbooks on pronunciation.Secondly, she did not show any attempt to gain comprehensive knowledge on audience expectation (i.e., the lecturer) and task demand (purpose of the assignment, form, what to evaluate, etc).Third, she did not draw concept mapping and therefore no outline was produced.W, on the other hand, did more effective prewriting strategy than D. He did attempt to gain some knowledge on audience expectation by talking to collaborators, and therefore he showed some awareness of the audience.He also did attempt to gain some knowledge on task demand by reading and studying textbooks and materials from internet.He also did concept mapping and drawing an outline, but the outline did not include a thesis statement, and therefore it is not an effective outline.What W did is in line with [6] statement that less skilled or less experienced writers begin to write much sooner, producing less elaborate "prewriting notes".In conclusion, W employed a different strategy from D but his strategy was not completely effective.It seems everything he did at prewriting was halfway between effective and ineffective.

Drafting Strategies
With regard to the effective writing strategies, D employed ineffective strategies at the drafting stage.The discussion is focused on the three processes of drafting: planning, translating, and reviewing.The drafting process of the two inexperienced graduate students was affected by the interactive approach they used at prewriting stage.The drafting process was classified into three sub processes of planning, translating and reviewing.

Planning
D employed a resourcing strategy for most of the time spent in drafting.To retrieve relevant materials from the written sources she relied on a simple, general outline and assessed the relevance of the materials as she encountered them during resourcing.She retrieved materials that could voice her ideas and then did note taking.The limited guidance she got for planning, i.e., a simple, general outline led to generating materials for the topic in a to-beat-around-the bush manner.
The strategy employed by D at planning was ineffective.When generating materials, she relied heavily on resourcing from external written materials.Her generating search was undirectional; she read, evaluated, and retrieved materials as she encountered them during reading since the search was guided by a very general outline.She did not narrow the topic until she was in the middle of the 22 drafting episodes, something that she should have been done at prewriting.In organizing, she tended to organize and reorganize the content as she gained insights during reading while writing.When setting a goal, she repeatedly set a goal which could not be executed, and then she had to form a new goal which was more achievable.In summary, the strategy employed by D at planning was ineffective because of lacking of guidance.
The ineffective planning caused by the absence of written outline as conducted by D is supported by a study by [9].He conducted an experiment and a field research to examine whether the use of written outline and rough and polished drafts enhance writing performance.The results showed that preparing a written outline,compared with not doing so, increased the time spent translatingideas into text, improved the quality of letters, and failedto enhance overall efficiency.A survey of scienceand engineering faculty revealed that the frequency of usingwritten outlines correlated positively with writing productivity,whereas use of polished drafts was uncorrelated with productivity.Another study by (Galbraith and Torrance, 2001) also confirms the effectiveness of constructing an outline prior to writing.The outline is used to guide retrieval of content during final draft production.
The absence of a thesis statement in W's outline and D's planning was probably caused by the lack of topic knowledge and genre knowledge.D collected five books on pronunciation but she did not read and study them until she was drafting the paper.She did not make an outline either.W did outlining but no thesis statement was created by him.In [7] say that research paper thesis is usually developed only after students have read a number of sources.D and W were limited in their experience in writing papers and had done little reading on their topics; furthermore, they had limited time to produce the paper which was accompanied by the need to finish assignments from other courses.
During planning D employed a number of strategies: cognitive, metacognitive, communicative, and social and affective strategies.The strategies helped her in solving immediate, local problems during planning.For instance, when she found difficulty in generating a specific material, she would delay generating it and created a new goal to be executed.
Unlike D, W started planning guided by an unelaborated outline, but in some way better than D's.Therefore, his planning process was more directional, although it lacked focus because of the absence of a thesis statement, so he was not sure what he was going to prove and why he wanted to prove it.He also did resourcing but relied to it less frequent than what D did.
In terms of time taken in planning, D and W had used a lot of time in planning, but it did not make the drafting process more efficient.Perhaps it is not the length of time that matters, but what is planned and how it is carried out.Planning might take a longer time because D and W encountered problems in generating ideas, in organizing them, and in creating what goals to achieve since they were provided with little guidance caused by the absence of an outline (D) and a thesis statement (W).Length of time taken in planning, therefore, might not the only thing that indicates the planning process is effective, but the content of planning must be considered, too.
In the same vein, recursivity which is widely assumed as a mark of an effective writing process may prove not to be so.In the study, both subjects showed recursivity in their writing process, but it did not result in good quality papers.As the findings indicated, a deeper examination into the processes both subjects underwent showed that their writing processes were recursive because they encountered problems which originated from their deficits in components that contribute to successful writing performance.

Translating
Translating done by D was mostly copying or "textual borrowing".The materials retrieved were modified to avoid "exact copying".Therefore, although she copied from textbooks, she still did rehearsing.For the most part, D was not formulating sentences to express her ideas, but searching for materials in textbooks that voiced her criteria for developing her paper.Only in the part beginning from where she described her small research to the end of the paper did she formulate her own sentences.What D did support ( [16]: 197) who explains that because non-native students do not have the syntactic and semantic skills of native speakers (or the confidence in the skills that they do have), they may "prefer to copy the words of another, which seem so much more accurate and elegant than their own".On the other hand.W did differently from D in that he, for the most part, formulated his own sentences and incorporated materials retrieved from external sources to support his statements.

Reviewing
Based on the list of effective reviewing strategy, both D and W did not employ an effective strategy in reviewing because most of reviewing was done by rereading what has been written previously for the purpose of continue drafting.They never reviewed the overall goal of the paper, something that is not difficult to understand because since the beginning they did not present a thesis; therefore, they did not have one to return to.

Conclusions and Suggestions
The main conclusion that can be derived from the findings is when writing a highly convent-ional zed text like a term paper, the stages of writing, i.e., prewriting and drafting, should be carried out properly by the writers because what a writer does at one stage will have impact on the subsequent stages.When a writer employs an ineffective strategy at one stage, this will likely make the subsequent stage flow haltingly and would lead to low quality of the written product.Other conclusions are based on findings on the strategy graduate students employ at prewriting and drafting stages.
First, the findings suggest that interactive or emergent approach is an ineffective strategy to be used at the prewriting stage when writing a term paper.Second, that inexperienced graduate student writers show the tendency to employ an interactive approach at prewriting may be explained by deficits they have in components that contribute to successful writing performance, i.e., deficits in topic knowledge, genre knowledge, task schema, linguistic knowledge, and audience knowledge.These deficits may be traced back to the scarce writing experiences the students have.Third, the strategy employed at the prewriting stage exerts influence on how the drafting process is carried out.When an interactive strategy is employed at the prewriting stage, it results in the employment of inefficient strategies to carry out the three subprocesses of drafting: planning, translating and reviewing.Fourth, being provided with little guidance to carry out drafting, inexperienced graduate students tend to resort to what-next-strategy during planning, i.e., by focusing on what idea has been written and how to continue writing from it.Fifth, when translating the generated materials into visible text, inexperienced graduate student writers tend to copy sentences from external written sources.Sixth, as a result of the little guidance available to them to carry out drafting, inexperienced graduate student writers do reviewing mostly to continue drafting.They never review the primary goal of the paper.Seventh, when drafting, inexperienced graduate student writers employ strategies that help them to solve immediate problems: cognitive, metacognitive, writing, and other strategies.
Eighth, the ineffective strategy employed at the prewriting stage and subsequently at drafting tends to result in low quality papers.
Students who enter graduate programs are expected to be equipped with skills in writing academic texts.Therefore, no lessons on developing those skills are offered by graduate programs.However, some graduate programs, e.g.English Education major at postgraduate program of State University of Malang, may provide instructions on academic writing skills through the so called pre-postgraduate program.It is at the instructors of pre-postgraduate program and the undergraduate level that the following suggestions are aimed.
To the writing instructors at undergraduate level of English Department and prepostgraduate program.First, it is suggested that students are provided with instructions that can reduce deficits in components that contribute to successful writing performance.Instructions that familiarize them with the structure of a paper and provide them with experiences in writing a paper will improve their genre knowledge.As the students need to master English used in writing academic texts, instructions on academic English would foster improvement on the students' linguistic knowledge.
Another difficult knowledge to grow is audience knowledge.In most school writing assignments, the audience is usually the teacher or the lecturer.Therefore, instructions on how to get information on the audience, i.e., the future readers of the students' papers may be devised.Ways on getting information of the audience, such as talking to seniors, asking the audience him/herself, may be introduced.
Second, in addition to deal with deficits in the students' knowledge, suggestions on how to carry out the writing process are offered.It has been concluded that employing an interactive strategy at the prewriting stage is not effective.Therefore, students should be introduced to advanced strategy when they prepare for their papers.
It has been concluded that drafting is less effective when students employ interactive strategy at prewriting.Therefore, after the students are instructed to employ an advanced strategy at prewriting, which should result in an elaborate outline, drafting can be introduced with the outline guiding the drafting process.Generating, organizing, and goal setting proceed while the students monitor them by referring to the outline and written sources needed.Reviewing to the primary goal of the text can be done by rereading the thesis statement in the outline.
Suggestions to graduate students.The findings of the study show that a number of factors play in producing a paper.One of the factors that contribute in the failure to produce a paper is deficits in the students' knowledge of topic, genre, language, audience, and task schema.Deficits in this knowledge may lead to the students' choosing ineffective strategies at prewriting and drafting stages.It is found out that experience in writing academic papers helps them in developing such knowledge.Therefore, it is suggested that graduate students and those preparing to continue their studies at graduate programs search for ways to develop their knowledge that are necessary to increase their ability in writing a paper.It is advisable that they develop reading habit on topics of their academic field and create opportunities to write papers, e.g.

Figure 1 :
Figure 1: The Cognitive Model of Writing Process (Flower and Hayes, 1980).

D
and W were selected based on the grades of their entrance writing test (Year 2006) and an interview that dug into their writing experience.The entrance tests provided by the graduate program of State University of Malang consisted of written research proposals, writing academic essays, and translating a text from Bahasa Indonesia into English.Only the last two tests were used to determine the subjects' legibility as they were impromptu tests which were considered reliable to assess the subjects' basic writing skills.
1) what writing strategies are employed by inexperienced Indonesian graduate student writers of English Education Major at the prewriting stage?and 2) what writing strategies are employed by inexperienced graduate student writers of English Education Major at the drafting stage?