In this chapter, the results of the two data sources, the writing tests and

CHAPTER FOUR

RESULTS AND DISCUSSION

In this chapter, the results of the two data sources, the writing tests and

questionnaire data, are analyzed and reported. The results of the data analyses are

presented first, and then the researcher will interpret the findings and attempts to

answer the research questions listed at the outset of the study.

Results

In the following sections, the results of the participants’ writing performance in

the writing pre-test and post-test are presented. The total scores and the subscores

for five components, namely content, organization, diction and spelling, and

mechanics, are compared within and between the two groups. Also, students’

writing fluency improvement between the two writing tests is reported. Finally, the

analyses of participants’ responses to the three types of questionnaires, including the

background information questionnaire, the Second Language Writing Apprehension

Test, and the evaluation questionnaire, are reported.

Analysis of Compositions

The effects of the guided writing and dialogue journal writing practices on

students’ writing proficiency are presented in this section.

Independent t-test was performed to examine the writing ability of the two

groups at the beginning of the study. Participants’ mean scores obtained in the

writing pre-test were compared. The result is displayed in Table 3. The scores

were measured by average of the sums of the two rater’s scores.

Table 3

Comparison of the Writing Scores in the Pre-test Between the Two Groups Mean (SD)

DJ (n=35) GW (n=35) t-value

Content 2.17 (.64) 1.79 (.46) 2.90

Organization 2.03 (.48) 1.66 (.58) 2.91

Grammar/Syntax 1.56 (.51) 1.37 (.49) 1.55

Diction & Spelling 1.56 (.51) 1.39 (.42) 1.53

Mechanics 1.01 (.09) 1.09 (.19) -2.02

Overall 8.39 (1.74) 7.29 (1.69) 2.68

Note. The full score is 20 and the discrete scores for the five components are as follows.

Content: 5 points; organization: 5 points; grammar/syntax: 4 points; diction & spelling: 4 points; mechanics: 2 points.

* p ＜.05

Based on Table 3, the dialogue journal (DJ) group demonstrated slightly, but

significantly, better writing proficiency in their pre-treatment writing test than the

guided writing (GW) group. The average overall scores of the DJ group and the GW

group were 8.39 and 7.29 respectively. Besides, the average scores gained on the

five components indicated that participants in the DJ group performed significantly

better than those in the GW group in content and organization, while students in the

GW group performed slightly better in mechanics.

In order to examine the effects of the two writing practices on the participants’

writing abilities, the scores gained by each group in the writing pre- and post-tests

were calculated and compared using paired sample t-tests. The comparison between

the pre-test and the post-test for the DJ group is shown in Table 4 and for the GW

group, in Table 5.

Table 4

Comparison of Writing by the Dialogue Journal Group Between Pre-test and Post-test Mean (SD)

Pre-test Post-test Mean Gain (SD) t-value

Content 2.17 (.64) 2.37 (.48) .20 (.71) 1.67

Organization 2.03 (.48) 2.10 (.36) .07 (.58) .72

Grammar/Syntax 1.56 (.51) 1.90 (.43) .34 (.62) 3.30

Diction & Spelling 1.56 (.51) 1.77 (.37) .21 (.62) 2.04

Mechanics 1.01 (.09) 1.10 (.20) .08 (.23) 2.24

Overall 8.39 (1.74) 9.30 (1.14) .91 (1.86) 2.90

Note. The full score is 20 and the discrete scores for the five components are as follows.

Content: 5 points; organization: 5 points; grammar/syntax: 4 points; diction & spelling: 4 points; mechanics: 2 points.

* p ＜.05

As Table 4 illustrates, participants in the DJ group exhibited improvement in

not only their overall writing performance but also in all writing components although

the improvement in organization and content did not reach the significant level. It is

also noted that students made a more remarkable progress in grammar.

Table 5

Comparison of Writing by the Guided Writing Group Between Pre-test and Post-test Mean (SD)

Pre-test Post-test Mean Gain (SD) t-value

Content 1.79 (.46) 2.31 (.58) .53 (.67) 4.64

Organization 1.66 (.58) 2.19 (.62) .53 (.82) 3.80

Grammar/Syntax 1.37 (.49) 1.96 (.39) .59 (.55) 6.31

Diction & Spelling 1.39 (.42) 1.90 (.36) .51 (.51) 6.00

Mechanics 1.09 (.19) 1.24 (.28) .16 (.32) 2.95

Overall 7.29 (1.69) 9.60 (1.70) 2.31 (1.91) 7.17

Note. The full score is 20 and the discrete scores for the five components are as follows.

Content: 5 points; organization: 5 points; grammar/syntax: 4 points; diction & spelling: 4 points; mechanics: 2 points.

* p ＜.05

As indicated in Table 5, students in the GW group made a significant progress

in their overall performance as well as in all the discrete components of writing. A

comparison of their improvement in these components reveals that the improvement

in content, organization, grammar, and diction and spelling is quite remarkable.

Finally, in order to explore the effects of the two different writing practices on

students’ writing, the scores of the two groups gained in the writing post-test were

compared. Since the two groups’ performance in the pre-writing test was

significantly different, instead of the independent-sample t-test, the analysis of

covariance (ANCOVA) was employed. Table 6 shows the results of the ANCOVA

and the post-test mean scores as adjusted by the covariates are also displayed here.

Table 6

Summary of ANCOVA for the Writing Scores in the Post-test

Source of Variance SS df MS F Sig.

Content

Covariance (Pre-test) .712 1 .712 2.580 .113

Between Groups .003 1 .003 .011 .918

Error 18.502 67 .276

Organization

Covariance (Pre-test) .066 1 .066 .254 .616

Between Groups .179 1 .179 .692 .409

Error 17.377 67 .259

Grammar/Syntax

Covariance (Pre-test) .446 1 .46 2.680 .106

Between Groups .128 1 .128 .772 .383

Error 11.140 67

Diction & Spelling

Covariance (Pre-test) .075 1 .075 .558 .458

Between Groups .335 1 .335 2.494 .119

Error 8.997 67

Mechanics

Covariance (Pre-test) .033 1 .033 .546 .463

Between Groups .289 1 .289 4.770 .032

Error 4.053 67 .060

Overall

Covariance (Pre-test) 12.463 1 12.463 6.446 .013

Between Groups 5.223 1 5.223 2.701 .105

Error 129.537 67 1.933

* p ＜.05

Adjusted Mean

DJ GW

Content 2.34 2.35

Organization 2.09 2.20

Grammar/Syntax 1.89 1.97

Diction & Spelling 1.77 1.91

Mechanics 1.11 1.24

Overall 9.16 9.74

It is evident that the adjusted mean score obtained by the DJ group on the

overall performance is slightly lower than that of the GW group, but there is no

significant difference. Similarly, the adjusted mean scores obtained by the DJ group

on all the five writing components are lower than those of the GW group. However,

none of them reach the significant level except mechanics.

In order to investigate the effects of the two writing practices on the

participants’ writing fluency, the total number of words written by the DJ group and

the GW group in the pre- and post-tests were counted, and the mean word per

composition was calculated.

Paired-sample t-test was used to examine whether the students improve their

writing fluency after the writing experiment. The comparison of mean word count

between the pre- and post-tests is presented in Table 7.

Table 7

Comparison of Mean Word Count Between Pre-test and Post-test Mean (SD)

Pre-test Post-test

Mean Gain

(SD) t-value

DJ (n=35) 81.40 (19.30) 101.74 (25.06) -20.34 (24.18) -4.98 * GW (n=35) 71.89 (23.03) 83.17 (23.10) -11.29 (23.60) -2.83 *

* p ＜.05

As shown in Table 7, participants of both groups did write more words in the

post-test than in the pre-test, and the improvement was statistically significant. In

other words, their writing fluency improved significantly at the end of the study.

Analysis of Questionnaire Data

In the following sections, the results of the participants’ responses to the three

questionnaires are analyzed and presented. The questionnaire data are then

compared between and within the two groups.

Results of Participants’ Responses to the Background Information Questionnaire

The background information questionnaire contains two parts. The first part

concerns participants’ personal information and their English learning experiences,

and the results have been summarized in “Participants” section in chapter three.

The second part of the background information questionnaire includes seven

Likert-scale items, eliciting information about participants’ perceptions about and

attitudes toward English writing. A Chi-square test was employed to examine if

participants from the two groups differed in their perceptions about and attitudes

toward English writing at the beginning of the experiment. The results are illustrated

in Table 8. Table 8 also shows a comparison of the percentages of the participants’

positive responses (including “agree” and “strongly agree”) and negative responses

(including “disagree” and “strong disagree”) to each item.

Table 8

Comparison of Participants’ Responses to the Background Information Questionnaire

DJ GW

Item χ

P

(%) N (%)

P (%)

N (%) 10. I think English writing is very important. 3.27 88.5 5.7 88.6 0.0

11. I’m eager to learn English writing. 2.31 82.9 5.8 88.5 0.0

12. I’m interested in English writing. 8.53 28.5 17.2 57.1 17.1

13. I think English writing is difficult. 2.54 62.8 11.5 51.5 8.6 14. I’m willing to spend time practicing English

writing. .39 68.6 5.7 68.6 2.9

15. I think my English writing ability is good. 3.45 11.5 71.4 2.9 77.1

16. I think learning English wiring is necessary. 5.25 57.1 5.7 80.0 0.0 Note. P = positive (“agree” and “strongly agree”); N = negative (“disagree” and “strongly disagree”)

* p ＜.05

As shown in Table 8, there was no significant difference in participants’

responses to any of the seven Likert-scale items in the background information

questionnaire between the two groups.

The comparison of the positive and negative responses indicates that the

majority of participants in both groups responded that English writing is important

(88.5% and 88.6%) and necessary (57.1% and 80.0%). Nearly 70% of them

demonstrated willingness to learn English writing, and about 30% of participants in

the DJ group and over 50% of the participants in the GW group displayed interest in

learning English writing. Besides, more than two-thirds of the participants in both

groups expressed their willingness to spend time practicing English writing.

However, more than half of the participants in both groups (62.8% and 51.5%)

responded that English writing was difficult to them, and more than 70% of

participants in both groups thought that their writing ability was poor.

In sum, although participants in both groups expressed similar opinions about

the importance of English writing and demonstrated willingness to learn English

writing, the comparison of their responses to items 12 and 16 showed that students in

the GW group seemed to have stronger motivation to participate in the writing

activity.

Results of Participants’ Responses to the Second Language Writing Apprehension Test

In order to investigate whether participants from the two groups differ in their

degrees of writing apprehension at the beginning of and the end of the study, the data

obtained through the pre- and post-Second Language Writing Apprehension Tests

(SLWAT) was compared and analyzed.

First, t-tests were employed to compare the scores gained in the pre-SLWAT

and post- SLWAT between the two groups, and the results are presented in Table 9.

Table 9

Comparison of the Scores Between Two Groups in the Pre-SLWAT and Post-SLWAT Mean (SD)

DJ (n=35) GW (n=35) t-value

Pre-SLWAT 74.06 (14.43) 72.29 (13.54) .53

Post-SLWAT 69.17 (12.11) 71.63 (15.46) - .74

* p ＜.05

The mean scores of the pre-SLWAT for the DJ group and the GW group were

74.06 and 72.29 respectively. The result indicates that though participants in the DJ

group showed relatively higher anxiety than those in the GW group, there was no

significant difference between the two groups at the beginning of the experiment.

Similarly, no significant difference was found between the writing apprehension of the

two groups in the post-SLWAT after the writing practices. However, participants in

the DJ group felt relatively less anxiety about English writing than those in the GW

group.

A paired sample t-test was used to examine the effect of the two writing

practices on the participants’ writing apprehension. The scores gained by each group

in the post-SLWAT were calculated and compared with those gained in the

pre-SLWAT. The results are displayed below in Table 10.

Table 10

Comparison of the Scores in the Pre- and Post-SLWAT Between the Two Groups Mean (SD)

Pre-SLWAT Post-SLWAT Mean Change (SD) t-value DJ (n=35) 74.06 (14.43) 69.17 (12.11) -4.89 (12.73) 2.27

GW (n=35) 72.29 (13.54) 71.63 (15.46) - .66 (8.55) .46

* p ＜.05

According to Table 10, the mean changes from the pre-SLWAT to post-SLWAT

for the DJ group and the GW group were -4.89 and -.66 respectively, and the results

show a reduction of writing apprehension in both groups. The table also indicates

that scores of writing apprehension obtained by the participants in the DJ group

dropped significantly from 74.06 to 69.17.

Results of Participants’ Responses to the Evaluation Questionnaire

In order to elicit participants’ feedback on the two writing practices, the

evaluation questionnaire was employed at the end of the experiment. The

questionnaire includes 16 items; eight of them are Likert-scale questions and the rest

were open-ended questions. The results of the Likert-scale items are presented first.

The eight Likert-scale items are classified into three categories. Items 1, 2, 3,

and 4 assess participants’ preference on the writing practices; item 5, 6, and 7 explore

their views on topics; and item 8 concerns their views on writing quantity. A

Chi-square was performed to examine whether the two groups differed in their

responses. The results are displayed in Table 11. Table 11 also shows the

comparison of the percentages of the participants’ positive responses and negative

responses.

Table 11

Comparison of Participants’ Responses to the Evaluation Questionnaire: The

Likert-scale Items

DJ GW

Item χ

P

(%) N (%)

P (%)

N (%) 1. I like this semester’s writing practice. 1.57 37.1 17.1 45.7 11.4 2. This writing practice cultivates my English

writing ability. 2.58 68.6 14.3 71.4 5.7

3. I look forward to writing the English

journal/composition every week. 2.90 20.0 31.4 31.4 22.8 4. I look forward to reading the teacher’s

responses/marked composition every week. 6.86 65.7 2.9 57.2 8.6 5. I prefer to write on an assigned topic. 8.81 25.8 31.5 48.6 11.5

6. I prefer to write on a suggested topic. 5.39 68.6 8.6 80.0 5.7

7. I prefer to write on a topic of my own choice. 9.51 48.6 8.6 22.8 31.5

8. I think it is best to write once a week. 3.48 65.8 20.0 54.3 17.2 Note. P = positive (“agree” and “strongly agree”); N = negative (“disagree” and “strongly disagree”)

* p ＜.05

As shown in Table 11, no significant difference was observed between the two

groups’ responses to all the items.

In terms of participants’ attitudes toward the two writing practices, though less

than half of the participants in both groups enjoyed the writing practices, more than

two-thirds of the participants in both groups (68.6% and 71.4%) affirmed that both

dialogue journaling and guided writing practice could cultivate their writing ability.

In addition, more than half of the participants in both groups agreed that they looked

forward to reading their teacher’s responses or marked compositions from their

teacher (65.7% and 57.2%); however, less than one-third of the participants looked

forward to writing English journals or compositions to their teacher, and it was

observed that a slightly higher percentage of participants in the GW group (31.4%)

showed their willingness to do the weekly English writing.

Similarly, there was no significant difference in participants’ views on the topics.

Nearly half of the participants in the GW group (48.6%) preferred to write on

assigned topics, while only about one-fourth of the participants in the DJ group

(25.8%) liked to write on assigned topics. On the contrary, nearly half of the

participants in the DJ group (48.6%) preferred to write on topics of their own choice,

and only about one-third of the participants in the GW group (31.5%) preferred to

write on topics of their own choice. However, nearly 70% of the participants in the

DJ group and 80% of the participants in the GW group preferred to write on suggested

topics.

As for participants’ views on writing quantity, more than half of the participants

in both groups (65.8% and 54.3%) agreed that writing one English journal entry or

composition a week was reasonable.

The following paragraphs summarize the participants’ responses to the eight

open-ended questions. The open-ended questions are categorized into two categories:

items 9 to 13 probed participants’ preference on the two writing practices, and items

14 to 16 asked for their comments and suggestions on the two writing practices.

Students’ Preference on the Writing Practices. For item 9 (“Do you like this

semester’s writing practice? Why or why not?”), the majority of the participants in

both DJ and GW groups (83% and 80%) liked the writing practices they had been

doing for the past 14 weeks.

The reasons why they liked the writing practice provided by the participants in

the DJ group can be categorized into four aspects: that the writing practice improved

their writing skills as well as their English, that the writing practice encouraged them

to think and to try to write down their ideas, that the writing practice was helpful for

their future writing skills development, and that they looked forward to reading

responses from the teacher. One student further explained that he became more

confident when writing in English, and that he had learned how to clearly express

himself in English. However, ten of the students also specified that they gradually

got tired of the dialogue journal writing activity due to their heavy schoolwork and

lack of ideas to write.

Likewise, the GW group also reported that the guided writing practice

improved their writing skills and their English, and that the guided writing practice

encouraged them to think and to try to write down their ideas. Two of the students in

the GW group specified that they felt a sense of achievement when they wrote down

their ideas in English. However, among these positive responses, five of them noted

that they disliked the writing task when the schoolwork was too heavy.

As for the reasons why the students disliked the two writing practices,

participants in both groups provided similar responses. The reasons included that

English writing was time-consuming and that they were unable to put their ideas into

English due to their low English proficiency. Moreover, three students in the DJ

group stated that the practice was meaningless if the teacher did not correct their

writing.

As for the benefits of the two writing practices (item 10: “What have you

learned from this writing practice?” and item 11: “Which part(s) in this writing

practice have you found most helpful?”), the participants from the DJ group in general

expressed unanimously that dialogue journal writing benefited their language

improvement as well as writing ability development. Almost one-third of the DJ

participants indicated that they had learned how to express themselves and share their

own ideas or life experiences with others in English. Students’ appreciation of

dialogue journal writing was also displayed by their comment that they became more

active in thinking and learning. They tried to read different English reading

materials to help them enrich their content. Some responses also indicated the

affective advantage of dialogue journal writing. For example, six students responded

that they gained more confidence in writing English because the teacher did not

correct their journal entries.

Similarly, participants from the GW group articulated their positive evaluation

for the benefits of writing activity used. Majority of the students in the GW group

stated that their language ability improved in that they not only learned new words

and phrases, but also were able to use what they had learned in their writing. They

found comments and suggestions provided by the teacher helpful. Also, they paid

more attention to the sentences they had written in terms of grammaticality and tense.

About half of the students wrote that they had learned some writing skills as well.

Particularly, nine of them specified that they had learned how to write and organize an

English composition, and three of them expressed that they were able to write more

fluently. Students also agreed that guided writing practice helped cultivate their

thinking ability. They became more aware of the mistakes they made in writing and

would try to avoid them. Finally, some students learned that English writing was not

as difficult as they had thought, and became less anxious about writing. Five of

them even specified that when they saw fewer red-ink marks in their writing, they

grew more confidence and became more motivated.

As for item 12 (“Which part(s) in this writing practice have you found not

helpful?”), only a few participants in both groups specified some parts that they had

not found helpful. Three participants in the DJ group responded that they hoped

their journal entries could be corrected so that they knew what their problems were.

On the other hand, three of the participants in the GW group thought that the outlines

were useless and wished to write more freely, and one wrote that revising the

composition was useless and boring.

When asked to state the difficulty they experienced (item 13: “What was the

biggest difficulty you found during the writing practice?”), more than half of the

participants in the DJ group and almost all the participants in the GW group stated

that the greatest difficulty they encountered during the writing practice was about

vocabulary and grammar. Many students in both groups expressed that they were

unable to convey their meaning correctly in English and they did not know how to use

the vocabulary or sentence patterns properly as well. Some of them even responded

that they did not know what and how to write. In addition, one in the DJ group and

two in the GW group wrote that their biggest difficulty was for them to turn in the

writing assignments on time.

Next, participants’ comments and suggestions on the two writing practices

elicited by items 14 to 16 are summarized in the following paragraphs. Feedback

from the participants of the DJ group is presented first.

Comments and Suggestions from the Dialogue Journal Group. The majority

of the participants in the DJ group thought that dialogue journal writing was

advantageous in both cultivating their writing skills and improving their language

ability. Eighteen of them stated that they sensed their improvement in English

writing. Among them, ten of them responded that they were able to write more

fluently, and five said that they were no longer afraid of English writing and were able

to express their ideas in English. More than half of the participants also found that

reading responses from the teacher was beneficial to their writing. When reading the

responses from the teacher, they not only reviewed what they had learned before but

also learned new vocabulary. Moreover, they became more aware of their own

writing. They had tried to make themselves clear in order to communicate with the

teacher. Many of them further pointed out that they looked forward to reading the

teacher’s responses because the teacher’s words encouraged and motivated them to

write more. They, therefore, grew more confident in writing English. Additionally,

three students pointed out that dialogue journal writing developed their creativity and

encouraged them to think.

Although most participants in the DJ group held a positive attitude toward

dialogue journal writing, a small number of participants (5 out of 35) mentioned that

the writing practice did not have any effect on their English writing ability. They

explained that since the teacher did not correct their writing, they did not know

whether their sentences were correct and what their problems were. They enjoyed

reading the teacher’s responses, but they thought the responses did not help them

much in writing development. As a result, 12 out of 35 students wrote in item 16

that they hoped the teacher could correct their grammatical errors and point out their

problems.

In the following two paragraphs, feedback from the participants in the GW

group is reported.

Comments and Suggestions from the Guided Writing Group. Thirty out of 35

participants in the GW group responded positively about the use of model texts in the

writing practice. These students found it helpful for three reasons: that it helped

them understand the content of the texts better, that they could imitate the

organization of the texts, and that they could practice using the vocabulary and

sentence patterns appeared in the texts. They also expressed that writing a

composition by imitating the text model had a positive influence on their writing

ability in several ways. They found imitating the model texts not only helped them

familiarize the vocabulary and sentence patterns they had learned but also cultivate

their writing ability. They had learned how to write an English composition and

organize their ideas, and they had become more fluent in writing English. Besides,

all of the participants in the GW group affirmed that text organization instruction was

very helpful. They had learned how to organize an English composition, and the

instruction helped them understand more about text structures in different writing

styles. Also, the instruction had made English writing much easier for them.

For those who responded negatively to the use of model texts, three of them

wrote that they felt restricted and wanted to have more freedom in the content they

wrote, and two of them thought the texts were too difficult to imitate. In the end,

some participants made suggestions for the guided writing practice. Four

participants thought that it would be better to allow them to choose their own topics

and write without the outlines. Four of them suggested that the teacher could assign

some particular sentence patterns for them to use in their writing, and three of them

proposed that the writing practice could stop once before the midterm exams.

Discussion

The results of the data analysis are discussed in this section. The section starts

with a discussion on the performance of the two groups in their two writing tests

intending to answer the first three research questions. The analyses of the

participants’ responses to the two Second Language Writing Apprehension Tests are

then discussed in order to answer the other three research questions.

Dialogue Journal Group’s Writing Improvement Between the Pre- and Post-tests

The comparison of the writing scores gained by the DJ group indicated that the

participants’ writing improved significantly in their overall writing performance as

well as grammar, diction and spelling, and mechanics. This result showed that the

dialogue journal writing certainly helped improve students’ writing, and it supports

what Heath (1988), Reid (1997), and Song (1997) have found in their studies.

Specifically, it was observed that the participants’ improvement in grammar was

more remarkable than that in the other writing components. The result is in

accordance with Heath’s (1988) observation on her students’ improvement in

grammar skills after they participated in the dialogue journal writing with the teacher.

It was also noticed that dialogue journal writing benefits students’ writing more in

content than in organization. The DJ group’s score gain in content, though not

significant, was slightly greater than that in organization. In other words, for the

purpose of improving students’ organization, explicit instruction on organization

would be helpful as Chiang (1999) suggested.

Guided Writing Group’s Writing Improvement Between the Pre- and Post-tests

The comparison of the writing scores gained by the GW group showed

significant progress in their overall writing performance and all the five writing

components. The result indicated that the guided writing practice and organizational

instruction indeed exerted influence on students’ writing in all aspects, and it gives

support to the findings reported in Chiang’s (1999) study that the instruction in

organization helped students produced compositions of better writing quality.

Further examination of the score gains by the GW group in the five components

demonstrated that students improved most in grammar. This result conformed to the

participants’ responses in evaluation questionnaire that they felt the guided writing

practice helped improve their English in terms of vocabulary and grammar. Besides,

the GW group’s content and organization both improved significantly, and the result is

slightly different from Wu’s (2002) finding that model texts helped students most in

the organizational part, because the GW group obtained the same gains in content and

organization.

Comparison of Participants’ Writing Improvement Between Two Groups

Table 12, based on the results presented in Tables 4 and 5, is a summary of the

comparison of the mean scores gained by the two groups between the pre-test and the

post-test.

Table 12

Comparison of the Score Gaines Between Two Groups

DJ GW

Mean Gain Sig Mean Gain Sig.

Content .20 1.67 .53 4.64

Organization .07 .72 .53 3.80

Grammar/Syntax .34 3.30

.59 6.31

Diction & Spelling .21 2.04

.51 6.00

Mechanics .08 2.24

.16 2.95

Overall .91 2.90

2.31 7.17

* p ＜.05

The comparison of the score gains revealed that both groups made significant

improvements in their overall writing performance. This suggested that both writing

practices were effective in improving students’ writing quality. This result,

nevertheless, showed that the GW group made remarkably more progress than the DJ

group in all the five components and the overall performance. The overall writing

proficiency of the GW group improved more than that of the DJ group.

It was expected by the researcher that both groups would improve in their

grammar, diction and spelling, and mechanics since the participants of both groups

received basically the same instruction on English reading texts in their English

classes. It was also predictable that the GW group would outperform the DJ group

in organizational improvement because the GW group received explicit instruction on

organization. However, the results tended to contradict the researcher’s expectation

that the DJ group would perform better than the GW group in content.

The possible reasons for the less improvement observed in content in the DJ

group include the type of writing practices they were involved in, the topic of the

writing post-test, and the teacher’s feedback on students’ writing assignments.

In terms of the type of writing practices, the GW group, after receiving explicit

instruction on organization, was asked to write compositions by imitating the text

structure of model texts with the help of structural outlines. Model-based

instruction on the organization, as Chiang (1998) found, helped writers produce

descriptive and narrative compositions with better organization and richer content.

The practice of composing the essays based on an outline very likely enabled the GW

group participants to develop their topic with relevant ideas, and further helped them

produce more substantial content. The DJ group, on the contrary, did not receive any

explicit writing instruction nor did they understand the rhetorical conventions of an

English composition. Thus, when they wrote, they just wrote their ideas down

without considering the relevance of the ideas, and this may have resulted in the DJ

group’s failure to make more gains in content in the post-test.

Another factor that influences students’ improvement in content may have been

attributed to the topic of the writing post-test. During the writing experiment, the

students in the DJ group were allowed to write whatever they liked while the

participants in the GW group were required to write on assigned topics imitating the

model texts. As the researcher observed, there was a lack of variety in content

produced by the DJ group in their journal entries since they mostly shared their

thoughts from their lives in the journals. The content that the DJ group produced

was mostly exposition, and narration was relatively practiced less frequently. But,

among the six topics assigned for the students in the GW group, four of them were

narrative (“The Most Unforgettable Experience,” “Meeting A Keypal,” “I Learned A

Lesson,” and “A Fable”). This suggests that the GW group was able to practice

writing narrative topics more frequently than the DJ group during the writing

experiment. This may have led to the result that the GW group improved more in

content in the post-test.

Teacher’ feedback on participants’ writing assignments may also have

contributed to the differences in improvement in content. The DJ group, during the

writing practice, received only the teacher’s positive responses and reflections on their

journal entries; thus, they might not know what their problems were in the writing.

The GW group, in contrast, received the teacher’s comments and suggestions on their

organization, grammar, diction and spelling, and mechanics. Even though the

teacher did not comment on their content, when the participants in the GW group read

the underlined sentences and the comments, it is very likely that they understood that

the sentences were problematic and failed to convey the meaning correctly. Their

attempt to clarify and correct the sentences may have resulted in better and clear

presentation of ideas. In other words, students in the GW group learned to write

more correctly and clearly.

In addition to the improvement in the overall writing performance and the five

writing components, students also showed improvement in writing fluency at the end

of the study. The comparison of mean words produced by the DJ group and the GW

group on the pre- and post-tests revealed that both dialogue journal writing and guided

writing practices exerted considerable impact on students’ writing fluency.

Participants in both groups produced significant more words in their post-tests. That

is, both dialogue journal writing and guided writing practice are helpful in terms of

writing fluency improvement.

Comparison of Participants’ Writing Apprehension

The comparisons of the scores in the pre-SLWAT and the post-SLWAT for the

DJ group and the GW group indicated that there was no significant difference in their

writing apprehension between the two groups at the beginning and the end of the

study. However, the mean changes in writing apprehension from the pre-SLWAT to

the post-SLWAT showed reduction for both groups. It can thus be concluded that

both writing practices helped reduce students’ writing apprehension.

The significant mean change observed in the DJ group, not in the GW group,

suggests that guided writing practice was not as effective in reducing students’ writing

apprehension as dialogue journal writing. This result conforms to previous findings

that dialogue journal writing was helpful in reducing students’ writing apprehension

(e.g. Ho, 2004; Holmes & Moulton, 1997; Reid, 1997; Song, 1997; Wang, 2003).

Moreover, as the previous studies reported, the DJ participants in the present study

became more willing to write in English, and their motivation to write was also

enhanced.

Finally, the researcher noticed that frequent practice in writing helps build

students’ confidence in English writing. Although the guided writing practice did

not significantly reduce students’ writing apprehension as the dialogue journal writing

did, it was observed that the students in the GW group also became less afraid of

English writing at the end of the study.