Finally, the procedures used to analyze the data are stated in the last part of the chapter

(1)

CHAPTER THREE METHOD

This study attempts to investigate how guided writing and dialogue journal

writing influence Taiwanese EFL students’ writing proficiency and writing

apprehension. This chapter describes the research method employed in the study.

First, the research design of the study, including the setting and the participants, is

described. Following the research design, a discussion of the writing instruction and

writing practices is provided. The data collection procedures and instruments

applied in this study are then explained. Finally, the procedures used to analyze the

data are stated in the last part of the chapter.

Research Design

This study is a quasi-experimental research (Larsen-Freeman & Long, 1990)

designed to investigate the effects of two writing practices on students’ writing

proficiency and their writing apprehension. Two groups of students engaged in two

different types of writing practices—guided writing and dialogue journal writing.

This study was conducted during a period of sixteen weeks. At the beginning

of the study, all participants took a writing pre-test and filled in two questionnaires,

the Background Information Questionnaire and the Second Language Writing

Apprehension Test (SLWAT). At the end of the study, a writing post-test was given,

(2)

and the post-SLWAT and the evaluation questionnaire was administered. Between

the pre-test and the post-test, these two groups of participants received different

writing treatments. The guided writing group was required to write seven

compositions, and the dialogue journal group was asked to write fourteen dialogue

journals. The description of the research design is presented later in “Data

Collection Procedures” section. The research design is shown in brief in Figure 1.

Figure 1. The Procedure of the Study

Writing Pre-test

Background Information Questionnaire Second Language Writing Apprehension Test

Treatment Guided Writing Group

Instruction + Writing (7) + Revision (7)

Dialogue Journal Group Weekly Dialogue Journals (14)

Writing Post-test Evaluation Questionnaire

Second Language Writing Apprehension Test

The Setting

This study was conducted in National Kangshan Senior High School (KSSH) in

Kaohsiung County. The researcher chose this site for two reasons. First of all,

since the researcher teaches in KSSH, she can recruit students from her own English

(3)

classes and this will help her understand more about her students’ learning process

and their needs. Secondly, the students in that school can represent the intermediate

learners in Kaohsiung area since the school accepts students who score higher than

190 in the Basic Competence Test for Junior High School Students (BCT). The

results can thus be generalized to a large number of intermediate learners in

Kaohsiung area.

Participants

This study involved two intact classes of first-year senior high school students,

within an average age of 16, at the research site. Participants in one class were

designated as the guided writing group (GW group), writing compositions based on

model essays, and the other class were designated as the dialogue journal group (DJ

group), writing weekly dialogue journals. Originally, each group included 38

students, totaling 76 students, in this study. However, during the experiment, six of

them (3 in the GW group and 3 in the DJ group) did not turn in their writing

assignments as required. Two students (one in the GW group and one in the DJ

group) were also excluded because of their low English proficiency. As a result, the

data were analyzed in terms of the compositions written by and the responses obtained

from 70 students, 35 (16 males and 19 females) in the GW group and 35 (15 males

and 20 females) in the DJ group.

(4)

All of the participants were taught by the same English teacher, that is, the

researcher. They had learned English for at least three years. The average time of

learning English was 5.4 years. The average BCT English scores of the GW group

and the DJ group were 44.1 and 45.5 respectively (the full score is 60). Seven

participants in the GW group and 10 participants in the DJ group had received some

English writing instruction in either junior high school or cram school. Only two of

them (one in the GW group and one in the DJ group) had had the experience of

writing English to foreigners.

Writing Instruction and Writing Practices

Participants of both groups received basically the same instruction in their

English classes except that students in the guided writing group received some extra

instruction on English writing organization using seven reading passages selected

from their English textbook as model texts. These two groups of participants were

assigned to work on different writing exercises.

Classroom Instruction

The guided writing group and the dialogue journal group received basically the

same instruction in their English classes. There are totally twelve lessons in the

selected English textbook, and each lesson contains one reading passage. The

teacher conducted the class using the following procedures: introducing the new

(5)

vocabulary and phases, introducing the focused sentence patterns, explaining the

reading passage, checking students’ comprehension, discussing particular language

use or function in the lesson, and guiding students to do exercises in the workbook.

The only difference in the teaching process between the two groups was in the reading

passage explanation procedure. In this procedure, the teacher taught the reading

passage paragraph by paragraph, identifying and explaining some important sentences,

phrases, and word usage. During this process, for the DJ group, the teacher only

focused her teaching on the explanation of the text itself, but for the GW group, the

teacher further explained the text organization of the reading passage, and helped

students identify the topic sentence, supporting sentences and main idea for each

paragraph.

Writing Practice for the Guided Writing Group

As described above, besides the basic instruction on the reading passage,

participants of the guided writing (GW) group received explicit instruction on text

organization of seven model texts chosen from the twelve reading passages in their

English textbook. The organizational instruction was given while the teacher was

explaining the reading text. Some basic knowledge about writing an English

composition was introduced in the first period of the English class and was reminded

constantly during the writing experiment. The concept of paragraphs, topic sentence

(6)

and supporting sentences, for example, was taught (See Appendix A for the material

used).

After the teacher’s explanation and analysis of one model text, the participants

were asked to write one composition of about 100 words based on the model text.

The assigned topic for the writing exercise was related to the model text (See

Appendix C for the list of topics assigned for the GW group). The participants were

told to write by imitating the organization of the model text with the structure

guidelines (See Appendix B for a sample of the structure guidelines). They were

also encouraged to use the vocabulary and sentence patterns they had learned in their

writing.

The collected compositions were not graded; however, they were returned to the

participants with problems or errors underlined and some suggestions from the teacher.

Tense problems and misspellings were pointed out. Ungrammatical sentences were

underlined, and suggested revised sentences were provided (See Appendix E for an

example). The participants then revise their compositions according to the

underlying cues and teacher’s suggestions, and handed in the revised version the

following week. They were asked to write on a new topic every other week. Every

composition was revised only once. Throughout the study, each participant in the

guided writing group produced totally fourteen drafts, including seven first drafts and

(7)

seven revisions.

Writing Practice for the Dialogue Journal Group

The participants of the dialogue journal (DJ) group, on the other hand, did not

receive any explicit instruction on writing, nor did they need to write any

compositions. Instead, they were engaged in writing dialogue journals, and were

required to hand in one journal entry every week. They wrote on their own topics or

on topics chosen from the list provided by the teacher (See Appendix D for the list of

topics). For each journal entry, they were asked to write about 100 words.

Unlike the guided writing group, there were no guidelines for the DJ

participants to follow, and they were allowed to write anything they liked in their

journals. The journals were not corrected or graded since the focus was the content;

the teacher provided only positive comments and/or personal responses to the entries

(See Appendix G for an example). The participants did not need to revise the

journals. Since this journal writing practice lasted for fourteen weeks, there were

totally fourteen weekly dialogue journals at the end of the study.

Data Collection Procedures

This study lasted for sixteen weeks. In the first meeting of the first week, a

writing pre-test was administered to assess participants’ English writing ability. The

participants were asked to write a composition of about 100 words in 40 minutes on a

(8)

topic “A Good Friend” (See Appendix L). After the writing pre-test, the background

information questionnaire (BIQ) and the pre-SLWAT were distributed for the

participants to complete.

For the following fourteen weeks, the guided writing group received extra

instruction on English writing organization and wrote compositions based on model

texts, while the dialogue journal group received no instruction on English writing and

wrote weekly dialogue journals.

The post-test was administered in the sixteenth week. The participants were

given 40 minutes to write an essay of 100 words on the topic “An Unforgettable Day”

(See Appendix M). On completing the test, they responded to the post-SLWAT and

the evaluation questionnaire (EQ). Table 1 presents the weekly schedules for the

two groups.

(9)

Table 1

Weekly Schedules for the Dialogue Journal Group and the Guided Writing Group

Week DJ group GW group

09/04~09/08 BIQ, pre-SLWAT & pre-test BIQ, pre-SLWAT & pre-test

09/11~09/15 Journal 1 Composition 1

09/18~09/22 Journal 2 Composition 1 revision

12/18~12/22 EQ, post-SLWAT & post-test EQ, post-SLWAT & post-test Instruments

The instruments used in this study included: the English composition grading

criteria developed by the College Entrance Examination Center (CEEC); and three

questionnaires, which contain the background information questionnaire, the Second

Language Writing Apprehension Test, and the evaluation questionnaire.

Composition Rating Scale

The compositions collected from the pre-test and the post-test were graded

using the CEEC English composition grading criteria (See Appendix N). This rating

scale is used to assess senior high school graduates’ compositions in the College

Entrance Examination in Taiwan. The scale includes five components: content (5

points), organization (5 points), grammar/syntax (4 points), diction and spelling (4

(10)

points), and mechanics (2 points). The possible total score is 20 points. The scale

classified compositions into five level of proficiency; namely, Excellent (19-20

points), Good (15-18 points), Fair (10-14 points), Weak (5-9 points), and Poor (0-4

points). The English version of the CEEC composition grading criteria is presented

in Table 2.

(11)

Table 2

The CEEC English Composition Grading Criteria

Excellent Fair Weak Poor

Content Entirely relevant.

Supported by concrete,

complete details.

（5-4 points）

Mostly relevant.

Lacks supporting details.

（3 points）

Topic not well addressed.

（2-1 points）

Topic not addressed.

（0 point）

Organization

Ideas logically organized and well developed.

Wide and appropriate use of linking devices.

Appropriate paragraphing.

（5-4 points）

Ideas not fully developed.

Inappropriate use of linking devices.

Disproportional paragraphing.

（3 points）

Ideas confused or disconnected.

Lacks logical sequencing and developing.

（2-1 points）

Does not communicate.

No organization.

（0 point）

Grammar (Syntax) Wide range of structures is used.

Structures are sophisticated and effective.

Very few grammatical errors.

（4 points）

Structures are accurate but simple.

Errors do not hinder

expression of ideas.

（3 points）

Wrong sentence structures hinder comprehension.

Incomplete sentences.

Inappropriate use of

conjunctions.

（2-1 points）

Does not communicate.

No mastery of sentence construction rules.

（0 point）

Diction & Spelling Accurate and

appropriate use of vocabulary.

Very few spelling errors.

（4 points）

Repetitive or limited use of vocabulary.

Errors are rare.

（3 points）

Inappropriate use of

vocabulary.

Meanings are obscure or incomprehensibl e.

Errors are many.

（2-1 points）

Does not convey complete ideas.

（0 point）

Mechanics Demonstrate mastery of conventions.

No errors in punctuation, capitalization.

（2 points）

Errors are rare and do not hinder

comprehension.

（1 point）

Ignorance of conventions.

Numerous errors.

（0 point）

In addition to the grading criteria, the raters were notified that if the content

students wrote was totally irrelevant, no scores would be given to the other four

(12)

features as well.

Questionnaires

Three questionnaires were used in the current study, namely, the background

information questionnaire, the Second Language Writing Apprehension Test (SLWAT),

and the evaluation questionnaire.

Background Information Questionnaire

In order to collect participants’ English learning experience, their attitudes

toward English writing and their perceptions of English writing, the background

information questionnaire (See Appendix H) was designed and implemented.

The questionnaire was composed of two parts, including fifteen items. The

first part (from item 1 to item 8) probed the participants’ background information,

such as their names, age, gender, English learning experience as well as their English

writing experience. The second part of the questionnaire, containing eight items,

elicited information about participants’ attitudes toward English writing and their

perceptions of English writing. Two items (items 9 and 15) examined participants’

perceptions about importance of English writing; three items (items 10, 11 and 13)

assessed participants’ motivation for learning English writing. The remaining two

items (items 12 and 14) assessed participants’ difficulty of English writing and

evaluate their English writing proficiency.

(13)

Second Language Writing Apprehension Test

For investigating the participants’ writing apprehension before and after the

writing practices, the adapted Second Language Writing Apprehension Test was

employed.

The Second Language Writing Apprehension Test (SLWAT) used in the present

study was adapted from Cheng’s (1998) Chinese version of the Second Language

Writing Apprehension Test, which is a second language version of Daly and Miller’s

(1975) Writing Apprehension Test. The questionnaire used a 5-point Likert scale,

ranging from “strongly agree,” “agree,” “neutral,” “disagree,” and “strongly

disagree,” and it elicited information related to writing apprehension in four

categories: (1) anxiety about writing in general, (2) teacher evaluation of writing, (3)

peer evaluation of writing, and (4) professional (e.g., publishers and magazine editors)

evaluation (Daly & Miller, 1975).

Cheng’s (1998) Second Language Writing Apprehension Test is composed of 26

items, but four items were dropped in the present study because they were related to

English composition courses and the participants did not take any English

composition course in their first year in the senior high school. Therefore, the

questionnaire used for the current study included only 22 items (For questionnaire

items, see Appendix I).

(14)

Evaluation Questionnaires

In order to elicit participants’ opinions and feedback on the two writing

practices—guided writing and dialogue journal writing—the researcher designed two

versions of evaluation questionnaires, and the question items in these two versions

were parallel (See the questionnaire for the guided writing group in Appendix J; see

the questionnaire for the dialogue journal group in Appendix K). Both versions of

the questionnaires consisted of two parts. The first part contained eight 5-point

Likert scale items, ranging from “strongly agree,” “agree,” “neutral,” “disagree,” and

“strongly disagree.” These items assessed participants’ preference on the writing

practices (items 1~4), their views on topics (items 5~7), and their views on writing

frequency (item 8). The second part included eight open-ended questions. They

investigated participants’ preference on the writing practices (items 9~13), and

elicited their comments and suggestions on the writing practices (items 14~16).

Data Sources and Data Analysis

The data collected in this study included pre-test and post-test compositions,

and four questionnaire responses from the background information questionnaire, the

pre- and post-SLWAT, and the evaluation questionnaire. All of the data were

analyzed with different methods. The ways of data analysis are illustrated in the

following sections.

(15)

Analysis of Compositions

In order to detect whether the two writing practices exert any influence on the

participants’ writing proficiency, the pre-test compositions and post-test compositions

were graded and compared. All the compositions were mixed and then evaluated by

two raters using the CEEC English composition grading criteria. The two raters

were English teachers (the researcher and her colleague) at National Kangshan Senior

High School (KSSH) who had at least two years of teaching experience. Both raters

had the experience of teaching and grading English compositions for senior high

school students. To ensure higher reliability, the two raters agreed on the criteria and

discussed how the rating scale was used before they started grading students’

compositions. Also, they were reminded to keep their own scoring stable.

The rating scale developed by the CEEC can be used for holistic scoring as well

as analytical scoring. In the current study, the compositions were graded analytically

in terms of five components (content, organization, grammar/syntax, diction and

spelling, and mechanics), and points were given for each component. The five

sub-scores were then added as a total score for the composition. The sub-scores

were used to compare participants’ performance in each component, and the total

scores were used to compare their overall writing proficiency. Each composition

was read by the two raters, and the average scores were used as final scores. The

(16)

inter-rater reliability between the two raters, calculated by performing the Pearson

product-moment correlation coefficients, was .875.

The scores of the pre-test and post-test compositions were calculated and

computed using the Statistical Package for Social Science (SPSS). Independent-

sample t-test was performed to compare the pre-test scores of the two groups to see

whether the writing proficiency of the two groups differed, and the result showed that

one group performed significantly better than the other group in the pre-test. As a

result, analysis of covariance (ANCOVA) was conducted to determine if there were

significant differences in the writing performance of the two groups in the post-test.

In addition, paired-sample t-test was employed to detect whether there was any

significant improvement between the two writing tests.

In order to measure the participants’ writing fluency, the number of words in

both pre- and post-tests was counted. A paired-sample t-test was then used to

examine whether the students improved in their writing fluency.

Analysis of Questionnaires

Four questionnaires were collected in this study, two before the study (the

background information questionnaire and the pre-SLWAT) and two after the study

(the post-SLWAT and the evaluation questionnaire).

The questionnaire data collected by the background information questionnaire

(17)

included participants’ learning experience and their perceptions about and attitudes

towards English writing. Participants’ personal background information was

summarized earlier in “Participants” section. The Likert-scale items were analyzed

by a chi-square test.

The questionnaire data from the pre-SLWAT, the post-SLWAT and the

evaluation questionnaire was organized and compared. First of all, scores of both

pre-SLWAT and post-SLWAT were calculated. All of the negatively worded

statements were reverse scored: the answers for items 1, 4, 6, 7, 12, 15, 20, 21, and 22

will be given 5 points for “strongly agree,” 4 points for “agree,” 3 points for “neutral,”

2 points for “disagree,” and 1 point for “strongly disagree.” In contrast, the answers

for items 2, 3, 5, 8, 9, 10, 11, 13, 14, 16, 17, 18, and 19 were given 1 point for

“strongly agree,” 2 points for “agree,” 3 points for “neutral,” 4 points for “disagree,”

and 5 points for “strongly disagree.” Next, paired-sample t-test was applied to

determine, within each group, whether there were any significant differences in the

participants’ writing apprehension in the pre-SLWAT and the post-SLWAT. For an

across group comparison, independent-sample t-test was performed.

Finally, the Likert-scale items in the evaluation questionnaire were analyzed by

a chi-square test to examine whether the two groups differed in their responses, and

the responses from the open-ended questions were categorized and compared.