




































Putri Anggraeni, et al / Journal of English Language Teaching 6 (1) (2017) 

37 

 

 
ELT FORUM 6 (1) (2017) 

 

Journal of English Language Teaching 

 
http://journal.unnes.ac.id/sju/index.php/elt 

 
 

CLUSTERING TECHNIQUE AND PEER 

ASSESSMENT IN TEACHING WRITING RECOUNT 

TEXT TO JUNIOR HIGH SCHOOL STUDENTS 

 

Eka Yunita Widyawati, Novia Trisanti 
 

English Department, Faculty of Languages and Arts, Universitas Negeri Semarang, Indonesia  

 

Article Info 
________________  
Article History: 

Received in June 

2017  
Approved in July 2017 

Published in August 2017 
________________  
Keywords: clustering; peer 
assessment; recount text; 
writing  
  
____________________ 

 
Abstract  
___________________________________________________________________  
This study was conducted under the consideration of the observation in SMPN 39 Semarang that many 

students faced problems in learning to write a text. This research is aimed at investigating the effectiveness 

of using clustering technique and peer assessment in teaching writing recount text compared to three-phase 

technique and the students’ perception of the lesson. A quasi-experimental was used as the research design 

in this study. The subjects were the eighth graders of SMPN 39 Semarang in the academic year of 

2016/2017 with 60 students as the samples. The control group (VIIIC) was treated using three-phase 

technique and teacher assessment, while the experimental group (VIII D) was taught using clustering 

technique and peer assessment. The findings revealed that the use of clustering technique and peer 

assessment significantly improved the students’ achievement in writing recount text. It was proven by the 

result of t-test in which the t-value (4.31) was higher than t-table (2.00). In addition, based on the result of 

the questionnaire, 90% of the students agreed that the use of clustering technique and peer assessment 

improved the students’ writing achievement. All in all, the use of clustering technique and peer assessment 

is more effective for teaching writing recount text to junior high school students than three-phase technique 

and teacher assessment. Moreover, most of the students have a positive perception to the lesson. 

  
 © 2017 Universitas Negeri Semarang 

  


 Correspondent Address: ISSN 2252-6706 
B3 Building FBS Unnes  

Sekaran, Gunungpati, Semarang, 50229  

E-mail: unnes_english@yahoo.com  



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

38 

 

INTRODUCTION 

As one of the four language skills which should be mastered by junior school students, writing 

is different since the writers should arrange the words and phrases into some sentences which can be 

clearly understood by the readers. Meyers (2005) stated that writing is a way to produce language 

which you do naturally when you speak. A good writing needs a long process that may include 

planning, drafting, editing (reflecting and revising), and writing final version (Harmer 2004). Some 

elements including a topic sentence, supporting sentences, concluding sentence, unity, and 

coherence also should be considered in writing (Oshima and Hogue 2006). 

According to School-Based Curriculum, any text types in English should be mastered by 

junior school students; one of them is recount text. Recount text is one of the text types which 

should be mastered by the eighth-grade students. According to Anderson (2003), recount text is a 

text that retells past events in the order in which they happen. In constructing a recount text, a writer 

has to consider some steps; (1) first paragraph that gives background information including the 

participants, place, and time, (2) a series of paragraphs that retell the events in the order in which 

they happened, and (3) a concluding paragraph. In addition, the specific language features including 

the use of proper nouns, descriptive words, past tense, and words that show the order of events are 

used in writing a recount text. 

 However, based on the observation in SMPN 39 Semarang, most of the students faced any 

problems in writing, especially in writing a recount text. The main problem was they did not know 

what they had to write. Also, they felt difficult to organize their idea into a good text. Moreover, 

most of the students frequently made errors in grammatical use, punctuation, conjunction, 

mechanics, and spelling. As the solution of these problems, any different techniques should be 

applied in teaching writing. Clustering is one of many techniques that may be useful in writing. 

Clustering is one of prewriting techniques that can be used by a writer to produce ideas (Oshima and 

Hogue 1997). It helps the students to generate any ideas which come to students’ thinking before 

they start writing. According to Noel (2005), there are a number of steps of using clustering 

technique. First, the writer can put a word, phrase, or sentence in a circle in the center of a blank 

page. Then, the writer can put every new idea in smaller circles around the first circle. Lastly, it is 

ended by drawing a line to another circle to its relationship to a previous idea. 

Another technique that may be useful in writing is peer assessment. Roberts (2006) defined 

peer assessment as a process of having students critically reflect upon, and perhaps suggest grades or 

comments for, the learning of their peers. According to Brown (2004), it offers some benefits, such 

as direct involvement of students in the learning process, the encouragement of autonomy, and the 

increase in motivation. However, peer assessment also gives some important drawbacks, such as 

subjectivity and the ability to discern their own errors. Trisanti (2013) explained some general issues 

that should be considered in using peer assessment. First, the teacher should explain to the students 

by giving a copy of form used for assessment and evaluation. Second, the teacher should give an 

opportunity to the students to practice assessing other group members in which their assessment do 

not affect project scores. Third, the teacher should provide feedback in order to improve the 

students’ performance. 

Based on the background above, this research is aimed at investigating the effectiveness of 

using clustering technique and peer assessment for teaching writing recount text compared to three-

phase technique and teacher assessment and describing the students’ perception to the lesson. In 

order to achieve the objectives of this research, a number of theories and procedures are used. 

 

 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

39 

 

METHODOLOGY OF THE RESEARCH 

The research design used in the study was Matching-Only Pre-test-Post-test Control Group Design 

by Fraenkel and Wallen (2008: 271). The design is presented as the following: 

 

Group Pre-test Treatment Post-test 

Experimental O1 X O2 

Control O3 C O4 

 
Table 2.1 Pre-test Post-test Control Group Design 

 

From table 1, it can be described that the subjects are divided into the experimental and 

control groups. The basic quality of the subjects is first checked by giving pre-test to both groups. 

Then, the two groups are taught the same topic, but they are treated by using different techniques of 

teaching. Post-test is given after the treatment and the result can be computed and analyzed. 

The subjects of the study were the eighth-grade students of SMPN 39 Semarang in the 

academic year of 2016/2017 with 274 students. The samples were VIII C, as the control group, 

consisted of 30 students and VIII D, as the experimental group, consisted of 30 students. The control 

group was treated by using three-phase technique and teacher assessment, while the experimental 

group was treated by using clustering technique and peer assessment. 

In collecting the data, two kinds of instrument were used. The first instrument was writing test 

which was divided into try-out, pre-test, and post-test. The writing test was given to investigate the 

effectiveness of using clustering technique and peer assessment for teaching writing recount text to 

junior high school students compared to three-phase technique and teacher assessment. Another 

instrument, questionnaire, was distributed to describe the students’ perception of using clustering 

technique and peer assessment for teaching writing recount text. Questionnaire design used in the 

study was closed-ended questions consisting of ten statements.  

A number of steps were done in order to achieve the objectives of the study. Before 

conducting the research, the try-out test was conducted to see the reliability and validity of the test. 

It was needed because a good instrument should be reliable and valid. After calculating and 

analyzing the reliability and validity of the instrument, the pre-test was given to the experimental 

and control groups. After scoring the pre-test of the students and analyzing them, the same material 

about recount text was given to both groups. However, they were taught by using the different 

teaching techniques. The experimental group students were taught by using clustering technique and 

peer assessment, whereas the control group was treated by using three-phase technique and teacher 

assessment. The treatments were given in four meetings.  

After completing the treatments, the post-test was given to both groups. In addition, the 

questionnaire was distributed to the students of the experimental group to see the students’ 

perception of using clustering technique and peer assessment for teaching writing recount text. After 

calculating and analyzing the post-test of the students, the t-test was calculated to see the significant 

improvement of the students in writing recount text. In addition, the students’ responses to each 

statement were computed and interpreted to describe the students’ perception of using clustering and 

peer assessment techniques for teaching writing recount text. 

In order to find out the validity of the data, this study used methods triangulation. According 

to Hales (2010:14), methods triangulation is the use of multiple methods to study a situation or 

phenomenon. Some methods in collecting the data, such as observation, test, and questionnaire 

were used. These multiple methods were used to make the methods completed each other. 

 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

40 

 

RESULT AND ANALYSIS 

 

In this part, the result and analyses of try-out, pre-test, post-test, and questionnaire are 

provided. Moreover, the discussion of the research findings is presented clearly at the end of this 

part. 

Try-out 

The try-out was conducted to find out the validity and reliability of the instruments. It was 

conducted in VIII B with 29 students who took the try-out test. To find out the validity of the test, 

the writer used content validity. The test was compared with the materials dealing with the 

curriculum requirement. The School-Based Curriculum states that the-eighth grade students are 

supposed to write a simple recount text. In reveals that the test is valid since the test is compatible 

with the curriculum requirement. 

Another characteristic of a good test is reliable. Since the instrument was writing test, the 

writer used inter-rater reliability. According to Nunnaly (1960) in Ghozali (2006), an instrument is 

considered as reliable if the value of Cronbach’s Alpha is higher than 0.6 (>0.6). The computation of 

inter-rater reliability of the test using SPSS 21.0 program is presented as follows: 

 

Reliability Statistics 

Cronbach's Alpha 
Cronbach's Alpha Based on 

Standardized Items 
N of Items 

,770 ,776 2 

 

Table 3.1 The Reliability Analysis of the Test 

From table 2, it can be seen that the value of Cronbach’s Alpha is 0.77. It means that the test 

is reliable since the value of Cronbach’s Alpha (0.77) is higher than 0.6. In addition, the reliability of 

peer assessment rubric used as an instrument in this research was also checked. The calculation of 

inter-rater reliability of peer assessment rubric used as instrument is presented in the following: 

 

Reliability Statistics 

Cronbach's Alpha 
Cronbach's Alpha Based on 

Standardized Items 
N of Items 

,830 ,835 2 

 

Table 2.2 The Reliability Analysis of the Instrument (Peer Assessment Rubric) 

 

Table 3 shows that the value of Cronbach’s Alpha is 0.83. It can be concluded that peer 

assessment rubric used an instrument in this research is reliable since the value of Cronbach’s Alpha 

is higher than 0.6 (0.83>0.6). From the overall analysis above, the instruments can be used in the 

real research since they fulfill the requirements of a good instrument 

Pre-test 

Pre-test was given to the experimental and control groups in order to know the basic ability of 

the students in writing recount text. From the pre-test result, the mean scores of the students in both 

groups were analyzed. The mean score of the control group students was 66.73, while the 

experimental group students was 66.97. It could be concluded that the experimental and control 

group students had almost equal achievement in writing recount text before getting the treatment. 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

41 

 

However, the homogeneity and normality of the test should be computed to support the 

statement. The homogeneity of the pre-test is computed in order to know whether the data of each 

group are homogeneous. The data are homogeneous if the significance value is higher than the 

significance level (0.05) (Ghozali 2006). The pre-test homogeneity computation is presented as the 

following: 

 

Test of Homogeneity of Variances 

Pretest   

Levene Statistic df1 df2 Sig. 

,000 1 58 ,989 

  

Table 3.3 The Homogeneity Test of the Experimental and Control Groups Pre-test 

 

From table 4, it can be seen that the significance value (0.99) is higher than 0.05. It reveals 

that the pre-test data are homogeneous. In other words, the population of the control and 

experimental groups has the same achievement before receiving the treatment. 

The normality of the pre-test also should be analyzed to find out whether the data are normal 

or not. According to Ghozali (2006), the data are considered normally distributed if the value of 

Asym.Sig.(2-tailed) of Kolmogorov-sminorv test is higher than the significance level (0.05). The 

computation of normality test using Kolmogorov-sminorv test in SPSS 21.0 program is presented 

below: 

 

One-Sample Kolmogorov-Smirnov Test 

 Experimental Group 

Pretest 

Control Group 

Pretest 

N 30 30 

Normal Parametersa,b 
Mean 66,97 66,73 

Std. Deviation 5,881 6,209 

Most Extreme Differences 

Absolute ,099 ,099 

Positive ,099 ,099 

Negative -,081 -,090 

Kolmogorov-Smirnov Z ,540 ,544 

Asymp. Sig. (2-tailed) ,932 ,928 

Table 3.4 The Normality Test of the Experimental and Control Groups Pre-test 

From table 5, it can be seen that the result of One-Sample Kolmogorov-Smirnov Test 

(Asymp.Sig.2-tailed) of experimental group pre-test is 0.93. It indicates that the data of the 

experimental group pre-test are normally distributed since the value of Asymp.Sig.(2-tailed) (0.93) is 

higher than 0.05. In addition, the value of Asymp.Sig.(2-tailed) of control group pre-test is 0.93. It 

reveals that the data of the control group pre-test are also normal because the Asymp.Sig.(2-tailed) 

value is higher than 0.05 (0.93>0.05). From the overall analysis above, the treatments can be given 

to the experimental and control groups since the data of pre-test are normally distributed and 

homogeneous. 

Treatments 

The different treatments were given to the control and experimental groups. The treatment 

using clustering technique and peer assessment was given to the experimental group, while the 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

42 

 

control group was treated by using three-phase technique and teacher assessment. The treatments 

were given in four meetings. At the first meeting, the material about recount text was explained to 

both groups, and they did the same activity and task. At the second meeting, both groups wrote a 

recount text in groups. At the third meeting, they did the activity in pairs and at the fourth meeting, 

they did it individually.   

Post-test 

Post-test was conducted to both groups after completing all the activities above to measure the 

students’ achievement after getting the treatment. Based on the post-test result, the mean score of the 

experimental group was 80.33, while the control group was 74.17. The result indicates that the 

students’ improvement of the experimental group is higher than the control group. 

However, the t-test should be computed as a statistical evidence to prove the statement. Before 

calculating the t-test, the homogeneity and normality of the post-test should be calculated and 

analyzed first to know whether the data are homogeneous and normal or not. The result of 

homogeneity post-test computation is presented as follows: 

 

Test of Homogeneity of Variances 

Levene Statistic df1 df2 Sig. 

,000 1 58 ,989 

Table 3.5 The Homogeneity Test of Post-test 

From table 6, it can be seen that the significance value (0.99) is higher than 0.05. It reveals 

that the post-test scores have the same variance. It can be concluded that the population of the 

control and experimental groups is homogeneous. 

The normality of post-test also should be computed first in order to see whether the data are 

normal or not. The normality test of post-test is provided below: 

 

One-Sample Kolmogorov-Smirnov Test 

 Experimental Group 

Posttest 

Control Group 

Posttest 

N 30 30 

Normal Parametersa,b 
Mean 80,33 74,17 

Std. Deviation 5,313 5,760 

Most Extreme Differences 

Absolute ,143 ,122 

Positive ,126 ,089 

Negative -,143 -,122 

Kolmogorov-Smirnov Z ,786 ,667 

Asymp. Sig. (2-tailed) ,568 ,765 

Table 3.6 The Normality Test of the Experimental and Control Groups Post-test 

Table 5 shows that the value of (Asymp. Sig. 2-tailed) of the experimental group post-test is 

0.57. It indicates that the data of the experimental group post-test are normally distributed because 

the significance value (0.57) is higher than 0.05. Moreover, the Asymp.Sig.(2-tailed) value of the 

control group post-test (0.77) is also higher than 0.05. It reveals that the data of the control group 

post-test are also normal. 

The Mean Scores Difference of Pre-test and Post-test 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

43 

 

The improvement of the experimental and control groups in writing recount text can be seen 

through the mean scores difference of pre-test and post-test which is presented in the following chart: 

 

Diagram 3.1 The Mean Scores Difference of Pre-test and Post-test 

Chart 1 shows that the mean score of the experimental group pre-test is 66.97. Meanwhile, the 

mean score of the post-test is 80.33. It reveals that there is significant improvement of the pre-test 

and the post-test scores achieved by the experimental group students. On the other hand, the writing 

achievement of the control group students also improves. It is 66.73 for the pre-test and 74.17 for the 

post-test. In indicates that there is less improvement than the experimental group. The difference of 

the experimental group mean scores (13.36) is higher than the control group (7.44). It can be 

concluded that there is a better improvement of the experimental group’s writing achievement after 

receiving the treatments by using clustering technique and peer assessment. 

T-test Analysis 

The purpose of calculating the t-test is to examine the hypotheses of this study. The 

improvement of the experimental and control groups is crucial to be tested as a quantitative evidence 

to find out whether the two groups have a significant difference in writing recount text. The result of 

t-test is presented as the following: 

 

Group Statistics 

Class N Mean Std. Deviation 
Std. Error 

Mean 

Posttest 
Experimental Group 30 80,33 5,313 ,970 

Control Group 30 74,17 5,760 1,052 

 Levene's Test 

for Equality of 

Variances 

t-test for Equality of Means 

F Sig. t df Sig. 

(2-

tailed) 

Mean 

Difference 

Std. Error 

Difference 

95% 

Confidence 

Interval of the 

Difference 

Lower Upper 

66.73 66.97 
74.17 

80.33 

0

10

20

30

40

50

60

70

80

90

Control Group Experimental Group

The Mean Scores of Pre-test and Post-test 

Pre-test Post-test



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

44 

 

Posttest 

Equal 

variances 

assumed 

,00 ,99 4,31 58 ,00 6,17 1,43 3,30 9,03 

Equal 

variances 

not assumed 

  4,31 57,63 ,00 6,17 1,43 3,30 9,03 

 

Table 3.7 Independent Sample T-test of the Experimental and Control Groups Post-test 

Independent Samples Test 

 

Based on the independent sample t-test result above, the test is significant, t(58)=4.31, p<0.05, 

d=95. It means that the t-value from 58 degree of freedom is 4.31. The probability of obtaining t-

value is 0.5 and the effect size is 0.95. The 95% confidence interval for the average percentage of 

post-test score ranges from 3.30 to 9.03. An examination of the group mean scores indicates that the 

post-test score of the experimental group (M=80.33, SD=5.31) is significantly higher than the 

control group (M=74.17, SD=5.76).  

In sum, after analyzing the result of t-test calculation, the use of clustering technique and peer 

assessment is more effective for teaching writing recount text to the eighth-grade students of a junior 

high school than three-phase technique and teacher assessment. The students’ achievement in 

writing recount text significantly improves after getting the treatment by using clustering technique 

and peer assessment. 

Questionnaire Result 

After conducting post-test, the questionnaire was given to the experimental group. The 

purpose of giving the questionnaire is to see the students’ perception of using clustering technique 

and peer assessment for teaching writing recount text. The result of the calculation of the students’ 

responses to each statement in the questionnaire can be seen in the table below: 

 

No Pernyataan 
Jawaban 

1 2 3 4 5 

1 Saya menyukai pelajaran bahasa Inggris. 0% 0% 30% 40% 30% 

2 
Saya menyukai pelajaran menulis dalam 

bahasa Inggris (writing). 
0% 3% 30% 57% 10% 

3 
Teknik clustering sangat mudah digunakan oleh 

siswa. 
0% 0% 17% 63% 20% 

4 
Teknik clustering membantu saya untuk 

mengembangkan ide-ide. 
0% 3% 10% 70% 17% 

5 

Dengan menggunakan teknik clustering, saya 

dapat mengembangkan ide-ide menjadi sebuah 

paragraf dengan lebih mudah. 

0% 3% 10% 63% 23% 

6 
Saya menyukai teknik peer assessment dalam 

pembelajaran menulis teks recount. 
0% 10% 33% 53% 3% 

7 

Suasana kelas menjadi menyenangkan setelah 

diberikan pembelajaran dengan menggunakan 

teknik peer assessment. 

0% 7% 33% 60% 0% 

8 
Penggunaan teknik peer assessment dapat 

meningkatkan pemahaman bagaimana cara 
0% 3% 3% 80% 13% 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

45 

 

menulis yang baik. 

9 

Penggunaan teknik peer assessment membantu 

saya dalam memperbaiki penulisan teks recount 

menjadi lebih baik. 

0% 7% 7% 67% 20% 

10 

Penggunaan teknik clustering dan peer assessment 

dapat meningkatkan kemampuan menulis 

saya. 

0% 7% 3% 57% 33% 

Table 3.8 The Result of the Questionnaire 

 

From the table above, the calculation result of the students’ responses to each statement in the 

questionnaire can be drawn as the following chart: 

 

Diagram 3.2 The Result of the Questionnaire 

Based on the table 9 and chart 2, the result of the questionnaire can be interpreted. The first 

statement is intended to know whether the students are interested in English or not. From the result, 

it indicates that most of the students are interested in English. However, there are some students that 

do not quite agree about it. The second statement in the questionnaire is to see the students’ interest 

in writing. From the result, it indicates that most of the students like writing. Nevertheless, there are 

some students that do not quite agree and few students disagree about the statement. 

The third statement in the questionnaire is intended to find out whether clustering technique 

is easy to be applied by the students or not. From the table and chart, it can be concluded that 

clustering technique is easy to be applied by almost all students. Nonetheless, there are a few 

students that do not totally understand how to apply clustering technique in writing. The fourth 

statement in the questionnaire is to see whether clustering technique helps the students to produce 

and explore their ideas or not. The result shows that most of the students agree that clustering 

technique helps them so much to produce and explore their ideas. By using clustering technique, 

they can write any ideas come into their mind before writing a draft. However, there are some 

students that do not quite agree and disagree that the technique helps them to produce and explore 

their ideas. 

The fifth statement in the questionnaire is intended to find out whether clustering technique 

facilitates the students to develop their ideas into a recount text. From the result, it reveals that most 

of the students agree that clustering technique helps them to develop their ideas into a recount text 

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

Strongly Disagree Disagree Not quite agree Agree Strongly agree

Questionnaire Findings 

Question 1 Question 2 Question 3 Question 4 Question 5

Question 6 Question 7 Question 8 Question 9 Question 10



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

46 

 

easily. Clustering technique helps them in developing words or phrases into sentences. Then, it helps 

for developing some sentences into some paragraphs and a text. However, some students argue that 

clustering technique does not facilitate them to develop their ideas into a recount text easily. The 

sixth statement in the questionnaire is to describe the students’ opinion of using peer assessment for 

teaching writing recount text. The result shows that more than a half of the number of the students 

like using peer assessment technique for teaching writing recount text. Nonetheless, almost a half of 

the number of the students argue that they do not quite agree and disagree with the statement. They 

feel that they are less interested in peer assessment technique because it is not easy to do. Also, they 

sometimes feel reluctant when giving assessment and suggestion to their peers. 

The seventh statement in the questionnaire is to know the class condition when using peer 

assessment technique. By seeing the result of the questionnaire, more than a half of the number of 

the students argue that teaching writing recount text using peer assessment is enjoyable because they 

have time to discuss with their classmates to evaluate their peers’ work. They can read it and then 

give assessment and suggestions. However, almost a half of the number of the students do not quite 

agree and disagree with the statement. They think that using peer assessment in writing recount text 

is challenging because they do not have good knowledge in writing. The eighth statement in the 

questionnaire is intended to find out whether peer assessment improves the students’ understanding 

and knowledge of a good writing or not. The result reveals that more than three-quarter of the 

students agree that their understanding of a good writing improves after they receive the treatment 

by using peer assessment. They can learn from the assessment and suggestions given by their friends. 

Nevertheless, few students are not sure that their understanding of a good writing improves after 

peer assessment is implemented for teaching writing recount text. 

The ninth statement in the questionnaire is to find out whether peer assessment technique 

facilitates the students to revise and edit their own writing. From the calculation result, it indicates 

that most of the students agree that peer assessment technique facilitates them to revise and edit their 

writing. They may be aware of the mistakes and errors they make so that they can revise their 

writing into a better result. Nonetheless, some students do not quite agree and disagree with the 

statement. The last statement in the questionnaire is to see whether the use of clustering technique 

and peer assessment improves the students’ achievement in writing recount text or not. From the 

students’ responses to the item, it indicates that the use of clustering technique and peer assessment 

significantly improves the students’ ability in writing recount text. Each component of writing 

improves after receiving the treatment by using clustering technique and peer assessment. However, 

there are few students argue that the use of clustering technique and peer assessment does not 

improve their achievement in writing recount text. 

Discussion of Research Findings 

The first objective of this study was to investigate the effectiveness of using clustering 

technique and peer assessment for teaching writing recount to junior high school students compared 

to three-phase technique and teacher assessment. The research findings revealed that the use of 

clustering technique and peer assessment was more effective for teaching writing recount text to 

junior high school students than three-phase technique and teacher assessment. It was proven by the 

result of pre-test, post-test and t-test. The mean score of the experimental students in pre-test was 

66.97, while the control group was 66.73. It indicated that both groups had almost the same 

achievement in writing recount text before the treatment given. 

After getting the treatment, the mean scores of the two groups improved. The mean score of 

the experimental group was 80.33, while the control group was 74.17. The data proved that the 

experimental group achieved a better result than the control group after receiving the treatment by 

using clustering technique and peer assessment for writing a recount text. 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

47 

 

Moreover, the t-test was used to prove the statement. Based on independent sample t-test, the 

t-value was 4.31. It revealed that the t-value (4.31) was higher than t-table (2.00). In addition, the 

significance value was lower than 0.05. It indicated that there was a significant difference in writing 

recount text between the experimental and control groups. Hence, the working hypothesis (Ha) 

which stated that “the use of clustering technique and peer assessment is effective for teaching 

writing recount text to the eighth-grade students of SMPN 39 Semarang in the academic year of 

2016/2017” was accepted and the null hypothesis (Ho) which stated that “the use of clustering 

technique and peer assessment is not effective for teaching writing recount text to the eighth-grade 

students of SMPN 39 Semarang in the academic year of 2016/2017” was rejected. 

Second, the study was aimed at describing the students’ perception of using clustering 

technique and peer assessment for teaching writing recount text. Based on the result of the 

questionnaire, most of the students had positive perception to the use of clustering technique and 

peer assessment for teaching writing recount text to junior high school students. From the 

calculation, 70% of the students were interested in English and 67% of the students were interested 

in writing. However, there were few students that stated that they did not quite agree with these 

statements. 

In addition, most of the students had a positive response to the use of clustering technique. 

From the questionnaire finding, 83% of the students agreed that clustering technique was easy to be 

applied by the students. 87 % of the students agreed that it helped them much more to produce and 

explore their ideas. Moreover, it facilitated 86 % of the students to develop the ideas into a recount 

text easily. By using this technique, the content and organization of the text improved. However, 

there were some students argued that the use of clustering technique did not help the students to 

explore ideas and develop the ideas into a recount text easily. 

Moreover, even though some students argued that they were not interested in using peer 

assessment because it was difficult, 93% of the students agreed that the use of peer assessment 

improved their understanding and knowledge of a good writing. They could learn from the 

correction, assessment and suggestions given by their peers so that their understanding of correct 

grammar, punctuation, capitalization and spelling improved. It also facilitated 87% of the students 

to revise and edit their own writing. They could rewrite their writing based on the peer assessment 

rubric so that the result was better than the first writing. In overall, 90% of the students agreed that 

the use of clustering technique and peer assessment improved the students’ achievement in writing 

recount text. 

However, there were still some drawbacks when using clustering technique and peer 

assessment for teaching writing text to junior high school students. The students needed much time 

when using clustering technique since they started by writing some words or phrases, and then they 

organized them into some sentences. Then, they developed the sentences into some paragraphs. 

Besides, peer assessment could be applied successfully when the students had a good background 

knowledge so that it also took much time to prepare the students. Subjectivity was also a problem 

when using peer assessment. Some students felt reluctant when giving low scores to their peers’ 

work.   

Triangulation Result 

The result of the test indicated that the experimental group students, which were treated by 

using clustering technique and peer assessment, achieved a better improvement than the control 

group, which was taught by using three-phase technique and teacher assessment. It was supported 

by the result of questionnaire in which most students agreed that their writing achievement 

significantly improved after they were taught by using clustering technique and peer assessment. 

Moreover, when the researcher observed the class activities and situation, most of the students were 



E. Yunita Widyawati & N. Trisanti / Journal of English Language Teaching 6 (1) (2017) 

48 

 

enthusiastic during the lesson since the techniques were easy, helpful, and pleasing. From the overall 

explanation, it could be concluded that the data were valid, and they supported one another. 

  

CONCLUSIONS 

 After scrutinizing the data, the writer concludes that she found 146 errors in all derivational 

categories. The errors were classified based on the four types of Surface Structure Taxonomy as 

proposed by Dulay, Burt and Krashen (cited in James, 1998:106). They are omission, addition, 

misordering and misformation. The research findings indicated that the most frequent errors 

occurred in misformation with 98 (67,1%) errors. It was followed by omission  ( 34 errors or 23,3%), 

and addition ( 14 errors or 9,6%).  

Out of the total errors (146), it showed that the majority of errors occurred in deriving noun to 

noun and verb to adjective with 30 errors. It was followed by 25 errors in deriving verb to verb, 23 

errors in deriving adjective to adjective, 11 errors in deriving adjective to adverb, 10 errors in 

deriving verb to noun, 9 errors in deriving noun to verb, 5 errors in driving noun to adjective and 3 

errors in deriving adjective to noun. This is relevant to the questionnaire result which showed the 

highest percentage (60 % or 15 students) found most of students agreed they found difficulty in 

deriving noun to noun. It was followed by 12 students or 48% stated that they found difficulty in 

deriving verb to adjective. It proved that both the test and questionnaire result indicated the most 

frequent errors occurred in deriving noun to noun.  

 

REFERENCES 

Akmajian, A., Demers, R.A., and Harnish, R.M. (1984). Linguistics : An Introduction to Language and 

Communication. Cambridge : The MIT Presss. 

Ellis, Rod. (1997). Second Language Acquisition. Oxford : Oxford University Press. 

Haspelmath, M. and Sims, A.D. (2010). Understanding Morphology. London : Hodder Education. [Online.] 

Available at http://morphology.xyz/pdf/HaspelmathSims2010.pdf [accessed 1/10/2016]. 

Fromkin, V., Rodman, R., and Hyams, N. (2011). An Introduction to Language. Wadsworth : Cengage Learning. 

James, Carl. (1998). Errors in Language Learning and Use.  Exploring Error Analysis. USA : Longman. 

Katamba, Francis. (1993). Morphology. New York: St. Martin’s Press. 

Kothari, C. R. (2004).  Research Methodology, Methods, and Techniques (Second Revised Edition). New Delhi : New 

Age International. 

Priyatmojo, A. S. (2011). Political Discourse : Obama’s Appraisal Attitude. LANGUAGE CIRCLE Journal of 

Language and Literature Vol. VI/. Available at : 

http://journal.unnes.ac.id/nju/index.php/LC/article/view/2043 [accessed 08/15/16] 

 


