168 | JISAE. Volume 6 Number 2 September 2020. https://doi.org/10.21009/JISAE JISAE (Journal of Indonesian Student Assessment and Evaluation) ISSN : P-ISSN: 2442-4919│E-ISSN: 2597-8934 Vol 6 No 2 (2020) Website : http://journal.unj.ac.id/unj/index.php/jisae A COMPARISON STUDY: TEACHER, PEER AND SELF-FEEDBACK ON STUDENTS ESSAY WRITING Imam Sudarmaji 1* , Marlien Bunga Lifanie 2 Universitas Islam Syekh-Yusuf 12 ABSTRACT The aim of this research was to know whether there are significant affection of reduction language error in students essay writing by using peer, teacher and self-feedback and to investigate which the best one of the strategies on reduction language error in students essay writing. The 126 tenth grade students of one senior high school in Tangerang were chosen as the sample which taken by non-random using purposive sampling. This research used experimental method with three experimental groups design. Students’ test for pre-test and post-test were used by the researcher to collect the data. The hypothesis data was tested using formula of t-test. The result showed that there was a significant effect used peer, teacher and self-feedback on reduction language error in students writing essay at the tenth grade of senior high school, but teacher feedback is better strategy than peer and self-feedback. It means that the three feedback strategies in reducing language error are effective on improving students writing skill and the teacher feedback strategy is the most effective one among the other strategies. Keywords: Comparison Study, Peer Feedback, Self-Feedback, Teacher Feedback, Writing Essay Address for Correspondence: 1 isudarmaji@unis.ac.id* ; 2marlienbunga@gmail.com INTRODUCTION Writing is one of English skills that should be mastered by the students. It’s related to daily lives. When they learn how to write it means that the students learn many things. As Raimes (1983) stated that using writing the student can learn how to amplify grammatical structures, idioms, and vocabularies, then they also learn about how to be a responsible and brave to take risk, and the last the most important when they write something they will embroil with that language and they will make hard effort to explain their ideas which connected by the senses (eyes, hand, brain) into a learning. As explanation by Nik et al. (2010) writing is not just putting the text into a paper but it is how effective ideas are presented. And this is in line with the research by Björk et al. (2003) who said that writing is a way of focusing on methods, practices, socio-psychological processes of intellectual inquiry, innovation, and learning. According to Brown (2003) writing was exclusive ability to have and all people use oral form to communicate with each other while for the written transactional like the business transactions, legal documents, agreement letters in all sections wrote by people who expert in their field. But look at this era, writing grow fast from the special ability become general skill that everyone must have that can not change as literacy in the global community. Nowadays, writing is a communicative activity that necessary to push and nurture during the process of language learning. Considering it Graham & Perin (2007) said writing is a flexible tool which can be used to fulfill various purposes. Writing process can help the writers to develop their intelligence http://journal.unj.ac.id/unj/index.php/jisae mailto:1%20isudarmaji@unis.ac.id mailto:2marlienbunga@gmail.com 169 | JISAE. Volume 6 Number 2 September 2020. and way to thinking, it can make the writers to be critical thinker. If this happen to the students, so they can be a creative person of thinking and be a critical thinker in all aspects. It means the purposes of government on created 2013 curriculum in 21st century is success. In addition by Dehkordi & Allami, (2012), writing is a process making use of the ideas of the people. This is important as writers’ need to show that they have understood the materials which they have studied and they can use the other writers/speakers ideas and findings in their own way. Writing process is not as simple as we see. The students get confused to write anything related the topic given by teacher. They should make a topic clear enough for them (Gould & Gould, 2004). Because it is difficult process even students in college will get confused when their lecturer asks them to make an essay or some stories or dialogues moreover the senior high school students. Salima (2013) argued that some students said they usually felt blank in mind so it made them less of concentration and they also said that they less of mastery grammatical rules of writing. It happened because they think writing is not too important. It can be called if writing is the most difficult skill than other skills. In the process, writing is also not an easy process. But through writing, students’ logical will explore all of their ideas and make it to be meaningful (Setyowati et al. 2017). Before that, the students also should make clear their idea first before they start writing. After that they can explain their ideas into a written form. They may get the point in Bahasa but when they want to translate into English the students do not know the proper words in English. Considering the problems above, both teacher and students have an important role for developing the writing skill. The teacher’s role is helping the students to improve their writing skill by introducing several methods which appropriate on improving students’ writing skill. The students’ role is pushing themselves to be aware if writing skill is important to be mastery and increasing the motivation to learn more the writing skill. Where it can be called if the feedback from students and teacher have influence to increasing the learning successful. Where it all can be formulated become the comparison between teacher, peer and self-feedback on reduction. Peer feedback is any comment that is given by the students to the students. Yu (2019) claimed, the findings show, peer feedback it also can be seen as a learning activity that not only focus on the cognitive development of participants, but also increase their metacognitive development. It is also regarded as a social activity and really useful in writing classes because of the cognitive, and social benefits of peer feedback (Bijami et al. 2013). Indirectly, the students encourage to make a written work in early and make it seriously and the result is the students getting better. Because as explained by Boud et al. (1999) that peer feedback is an activities which include of working collaboratively with others, taking responsibility for their own learning and deepening their understanding of specific course content. As explained by Smemoe (2018), in peers feedback process helps the students to aware of the point where in writing a multiple draft is really necessary to increase their paper quality therefore should not be seen as a threat to the face, and that there are ways to talk about making changes to write that are more positive than others. The students can develop their analytical ability from peer, as Baker (2016) claimed that with peers, the students review the comment of feedback identified by their peers problem and try to looking for the solving problem for that problem. The students can give the suggestion for their peers how to solve their writing problem Teacher feedback is very important feedback where students expect to receive. The students need to get the feedback or comment from the teacher about their writing. It can help the students to know about their mistakes. As stated by Hyland & Hyland (2014) that teacher feedback can be used as an active strategy by academic writers as they develop their own 170 | JISAE. Volume 6 Number 2 September 2020. voices and their familiarity with different genres. In same line, teacher feedback is important aspect in learning process. Because from teacher feedback the students can make improvement on their skill in writing. They can learn where the incorrect grammatical, or organization in their own written. According to Ruegg (2015) if teacher feedback has an important role in improving students writing skill and also increasing grammatical accuracy, the students who get the teacher feedback as directly lead the increasing grammatical accuracy and get score higher than peers feedback. Teacher feedback also has a benefit for the students. According to Hyland (2016) teacher feedback is very important for the development of writing skills, both for their potential for learning and for student motivation. Besides the advantages of the teacher feedback, there are also the disadvantages that appear in the process. Like we know now why the government print out the k 13 curriculum that argued that students can’t reliant on themselves, the students always depend on the teacher and the teacher is only one of the sources information. The students can’t develop their ability by themselves, lack of creativities and critical thinker. And according to Razali & Jupri (2014) if teacher feedback get negative impact on students revision, because on teacher written feedback the comments is unclear and too general. It can make the students felt confused and difficult to respond and include comments in their revision process. The most important is if in students writing there are many correction, it can make demotivate student to do revision. Self-feedback means that the students make a judgement on their own work. The students can make a marks as assessment and feedback from their own work. This method make a student learn about assessing process. Before the students try to make a marks as a feedback in their written, the teacher will taught them about the required of assessment process and what kind of criteria in the assessing process ( Diab, 2016). According to Xiang (2004) it is an effective way to increasing student organization of compositions and very useful for students with higher skills. Esfandiari & Myford (2013) argued that self-feedback make the student learn to depend on themselves, be responsible of their own learning, looking for the strengths and weakness of their own individual pattern, help them to be reflective and involved students. It seem that give the students opportunity of self-feedback/ assessment can help them to sharper their metacognitive, where it can give effect of lead they thinking and learning (Birjandi & Tamjid, 2011). As same as Nicol & MacFarlane-Dick (2006) statement that when the students assessing their own writing, they can decided criteria or standard of achievement which want they achieve. But the students become more independence it make them feel an arrogant person (can’t be an objective) or even they will get confused related to the content of criteria like the result research of Lindblom-Ylanne, Pihlajamaki, & Kotkas, (2014). As we as known in EFL learner, they get difficulties to find out the right words when they write the essay. For write down all their ideas into the written they still get hard moreover they should assess their own. Give peer, teacher and self-feedback, it means the students can easier on improve their writing skills. Because of that assumption above, the researcher will find out the most effective methods to reduction language errors in students’ essay with quasi experimental with pre-test and post-test design as a model of this research methods. Where the researcher has an expected fin ding of the research to be useful for the teacher and for other teacher about the best methods for reduction students’ language error in essay. METHOD 171 | JISAE. Volume 6 Number 2 September 2020. This research used the quasi experimental. The researcher cannot choose randomly samples or the other hand is the students can be isolated in certain times. On the stage, the researcher gave treatment in three experimental classes. There would be different treatment that was given in each class. First experimental class used peer feedback, the second experimental class used self-feedback and the third experimental class used teacher’s feedback. The researcher was use target of population and accessible population. The target population is a large population where writers usually limit data sources to accessible populations. While, the population that can be accessed is a population that can be generalize data to be more specific. In this research a target of population was all of tenth student on SMAN 3 Kab. Tangerang but before the condition not allow for do this strategies in target population so the researcher determine was used accessible population The researcher determined purposive sampling because this study had a purpose and aspects that should be investigated. The researcher chose 3 classes on tenth grade of 10 math and science 5, 10 math and science 6 and 10 math and science 7 as sample. This research consists of quantitative data. It was obtained of students’ score from the task. Pre-test and post-test in essay form was as instrument of test that was given by the researcher. Pre-test was given before the researcher did the treatment. The purposes of give a pre-test in a first process is to find out and collecting data of the students’ ability in written text and to know how far the first students’ ability of writing in a simple text. After it, the researcher did the treatment by using peer, self and teacher feedback in experimental class. And the last process gave the post-test to compare the students’ ability after the treatment was given by the researcher. And to look out whether there are any significant differences between peer, self and teacher feedback on students’ writing skill. RESULTS AND DISCUSSION The researcher collaborated with the students to do the research. The researcher and the students conducted the research by giving treatment in both of classes. The treatment given was teaching descriptive text to the students using teacher feedback, peer feedback and self- feedback. It was different strategies in each class. Where the researcher used the strategy with circling, categorizing the errors, and giving corrections, and then the students used the strategy with circling and giving corrections. The researcher used the written test in the form of essay with the instrument that has been valid and reliable. To find out the validity and the reliability of the instrument, the researcher validated before it was used by conducted the content of validity. The researcher gave the content of validity to the two English teachers and two Lecturers of English Education. The test was given to find out the effect of teacher feedback, peer feedback and self-feedback in student’s writing skill. After giving the pre-test and post-test in three experimental classes, the researcher analyzed the result to get the student’s score. The result of the test was called data and it will be process by using SPSS 24 in analyzing the data, the researcher used the difference score of the students. It was obtained from looking for the differences between pre-test and post test score of each class. First Experimental Class Based on the writing test that was given to the tenth grade students of SMAN 3 Kab. Tangerang, the researcher analyzed the data of the scores of the students‘ writing skill by statistical result SPSS 24 of pre-test and post-test there was difference 408. The data was taken from students‘ answer of the first experimental class in the pre-test which was found the minimum score was 19 and the maximum score was 60 with mean 41,44, and standard deviation 9,560. While, the post-test which was found the minimum score was 33 and the 172 | JISAE. Volume 6 Number 2 September 2020. maximum score was 82 with mean 51,39, and standard deviation 10,327. It can be seen in the table of descriptive statistics as follows: Table 4.1 Descriptive Statistics First Experimental Class Source: Statistical result SPSS 24 The Minimum Mastery Criterion (KKM) in English Language especially in writing at the school is 72. Based on the result of post-test, there were almost students less than 72 (40 students) as KKM. It showed that 40 students not reach the KKM on writing. While the total of the students were who got than 72 as KKM was only 2 students. If we can see the posttest in students’ percentages as follows: Figure 1. Figure of Percentages Score of Posttest in First Experimental Class Based on the Figure above, 4.76% students can reach the KKM values and 95.24% students cannot reach the KKM values. Second Experimental Class Based on the writing test that was given to the tenth grade students of SMAN 3 Kab. Tangerang, the researcher analyzed the data of the scores of the students‘ writing skill by statistical result SPSS 24 of pre-test and post-test there was difference 655. The data was taken from students‘ answer of the first experimental class in the pre-test which was found the minimum score was 30 and the maximum score was 80 with mean 41,24, and standard deviation 10,552. While, the post-test which was found the minimum score was 40 and the Elements PreTest1 PostTest 1 Mean 41,44 51,39 Std. Deviation 9,560 10,327 Minimum 19 33 Maximum 60 82 Sum 1699 2107 173 | JISAE. Volume 6 Number 2 September 2020. maximum score was 87 with mean 57,22, and standard deviation 10.135. It can be seen in the table of descriptive statistics as follows: Table 4.2. Descriptive Statistics Second Experimental Class Elements PreTest PostTest Mean 41,24 57,22 Std. Deviation 10,552 10,135 Minimum 30 40 Maximum 80 87 Sum 1691 2346 Source: Statistical result SPSS 24 Based on the result of posttest, there were 37 students less than 72 as KKM. It showed that 37 not reach the KKM on writing. While the total of the students were who got than 72 as KKM was only 5 students. If we can see the posttest in students’ percentages as follows: Figure 2. Figure of Percentages Posttest in Second Experimental Class Based on the Figure above, 11.91% students can reach the KKM values and 88.09% students cannot reach KKM values. Third Experimental Class Based on the writing test that was given to the tenth grade students of SMAN 3 Kab. Tangerang, the researcher analyzed the data of the scores of the students‘ writing skill by statistical result SPSS 24 of pre-test and post-test there was difference 870. The data was taken from students‘ answer of the first experimental class in the pre-test which was found the minimum score was 17 and the maximum score was 63 with mean 36,71, and standard deviation 9,220. While, the post-test which was found the minimum score was 18 and the maximum score was 72 with mean 57,93, and standard deviation 10,417. It can be seen in the table of descriptive statistics as follows: Table 3. Descriptive Statistics Third Experimental Class Elements PreTest PostTest Mean 36,71 57,93 Std. Deviation 9,220 10,417 Minimum 17 18 Maximum 63 72 Sum 1505 2375 Source: Statistical result SPSS 24 Based on the result of posttest, there were 39 students less than 72 as KKM. It showed that 39 not reach the KKM on writing. While the total of the students were who got than 72 as KKM was only 3 students. If we can see the posttest in students’ percentages as follows: 174 | JISAE. Volume 6 Number 2 September 2020. Figure 3. Percentages Posttest in Third Experimental Class Based on the Figure above, 7% students can reach the KKM values and 93% students cannot reach KKM values. Difference Value of First, Second and Third Experimental Class There were many strategies which used teacher teaches their students. The strategies can be depend or not depend on the teacher and the strategies. Based on this reason, the teacher should have a good strategy to teach their students. Here, the researcher uses comparison between feedback by teacher with students’ feedback (peer feedback and self-feedback) with different strategies. It also aimed to know which one of the strategies is better. Source: Statistical Result SPSS 24 Based on the table above, the researcher got the data between the difference score in first experimental class, second experimental class and third experimental class using different feedback strategy. It shows from the mean score of posttest in first experimental class is 9,95 then the mean score of posttest in second experimental class is 15,98 and the last in the mean core of posttest in third experimental class is 21,22. The standard deviation of first experimental class is 11,063 then the standard deviation score of second experimental class is 8,448 and the last is the standard deviation score of third experimental class is 10,106. The difference minimum score of first experimental class is -14, second experimental class is -1 and third experimental class is 1. The difference maximum score of first experimental class is 39, second experimental class is 34 and third experimental is 43. Total score differences of first experimental class is 408, second experimental class is 655 and third experimental class is 870. Testing of the Research Hypothesis After tested for normality and homogeneity, it can be said that the data in first eperimental class was normal while the data in second and third experimental class was not normal. For homogeneity test, all of data are homogeneous. And then, the researcher testing the hypothesis test (t-test) formula to find out the effect of feedback strategy on the reduction Table 4.4 Descriptive Statistics of Difference Scores Elements Difference 1 Diffrence 2 Difference 3 Mean 9,95 15,98 21,22 Std. Deviation 11,063 8,448 10,106 Minimum -14 -1 1 Maximum 39 34 43 Sum 408 655 870 175 | JISAE. Volume 6 Number 2 September 2020. of language error in students writing essays by teaching descriptive text and also aim of this test was to compare the pre-test and post-test score in each class (first experimental class using peer feedback, second experimental class using teacher feedback and third experimental class using self-feedback). There are three testing the hypotheses test (t-test): First Experimental Class The result t-test about writing skill for the pre-test and post-test showed that the significant result in the first experimental class, it was found that the result of the test of homogeneity could be seen as follows: Table 7. Paired Samples TestFirst Experimental Class Elements Paired Differences T Df Sig. (2- tailed) Mean Std. Deviatio n Std. Error Mean 95% Confidence Interval of the Difference Lower Upper Pair 1 PreTest1 - PostTest1 -9,951 11,063 1,728 -13,443 -6,459 -5,759 40 ,000 Source: Statistical result SPSS 24 Based on table above, it is known –tcount < -ttable (-5,759< -2,026) with sig = -0,000 < 0,05, so it can be concluded that Ho is rejected and Ha is accepted, it means there is a significant difference between pre-test and post-test in first experimental class using peer feedback. So, applied peer feedback can reduction of language error on students writing essay at the tenth of SMAN 3 Kab. Tangerang. Second Experimental Class The result t-test about writing skill for the pre-test and post-test showed that the significant result in the second experimental class, it was found that the result of the test of homogeneity could be seen as follows: Table 8.Nonparametric Test Second Experimental Class Null Hypothesis Test Sig. Decision 1 The median of differences between Pre2 and Post2 equals 0. Related- Samples Wilcoxon Signed Rank Test .000 Reject the null hypothesis. Asymptotic significances are displayed. The significance level is .05. Source: Statistical result SPSS 24 176 | JISAE. Volume 6 Number 2 September 2020. Based on table above, in this experimental class the researcher got not normal result, so she did Nonparametric Test named Wilcoxon Signal Rank Test and get the result is reject the null hypothesis with the significance value 0,000 < 0,05. Ha is accepted and Ho is rejected, where it means there is a significant difference between pre-test and post-test in second experimental class using self-feedback at the tenth grade of SMAN 3 Kab. Tangerang. So, applied self-feedback can reduction of language error on students writing essay at the tenth of SMAN 3 Kab. Tangerang. Third Experimental Class The result t-test about writing skill for the pre-test and post-test showed that the significant result in the third experimental class, it was found that the result of the test of homogeneity could be seen as follows: Table 9. Nonparametric Test Third Experimental Class Null Hypothesis Test Sig. Decision 1 The median of differences between Pre3 and Post3 equals 0. Related- Samples Wilcoxon Signed Rank Test .000 Reject the null hypothesis. Source: Statistical result SPSS 24 Based on table 4.10 above, in this experimental class the researcher also got not normal result, then she did Nonparametric Test named Wilcoxon Signal Rank Test same like before and get the result is reject the null hypothesis with the significance value 0,000 < 0,05. Ha is accepted and Ho is rejected, where it means there is a significant difference between pre-test and post-test in third experimental class using self-feedback at the tenth grade of SMAN 3 Kab. Tangerang. So, applied self-feedback can reduction of language error on students writing essay at the tenth of SMAN 3 Kab. Tangerang. Different Values of First, Second and Third Experimental Class Result t-test about writing skill for the final test shows the significant result in first experimental class, second experimental class and third experimental class as follows: 177 | JISAE. Volume 6 Number 2 September 2020. Table 4.10 Multiple Comparisons Dependent Variable: Different23 Bonferroni (I) Class (J) Class Mean Difference (I-J) Std. Error Sig. 95% Confidence Interval Lower Bound Upper Bound Peer Feedback Self-Feedback -5.849* 2.203 .027 -11.20 -.50 Teacher Feedback -11.311* 2.176 .000 -16.59 -6.03 Self-Feedback Peer Feedback 5.849* 2.203 .027 .50 11.20 Teacher Feedback -5.462* 2.190 .042 -10.78 -.15 Teacher Feedback Peer Feedback 11.311* 2.176 .000 6.03 16.59 Self-Feedback 5.462* 2.190 .042 .15 10.78 *. The mean difference is significant at the 0.05 level. Source: Statistical result SPSS 24 The researcher analyzed the result 126 of the differences score first experimental class, differences score second experimental class and differences score third experimental class. In this data analyzed, the researcher used multiple comparisons of Bonferroni to compare all of feedback strategy. Based on table above, sig. all of experimental class < 0,05, so it can be concluded that there is significant differences on improving student writing essay between first experimental class, second experimental class and third experimental class using peer feedback, self – feedback and teacher feedback strategy at the tenth grade of SMAN 3 Kab. Tangerang. So, all of feedback strategy that applied by the researcher has differences on improving the students writing skill at the tenth grade of SMAN 3 Kab. Tangerang. From the students’ analysis of writing skill taught by using peer, teacher and self- feedback in the different experimental class, the researcher has found the result of students’ writing skill. The researcher has given pre-test, different treatment, and post-test in order to know the students’ score in writing skill. In first experimental class used peer feedback strategy, where the student got the treatment from their friends which focused on error feedback the composition ―circle the errors, and provide the corrections. The researcher was started by giving pre-test, applying treatment, and giving post-test. The test is done in order to know the students‘ writing skill. The result of test was computer by using SPSS 24 version that can be seen the lowest score is 33 and the highest score is 82. Then the mean is 51,39, the median is 49,00 and the standard deviation is 10,327. The result of teaching writing was given feedback by their friends using peer feedback strategy is good because the students get significant differences between pre-test and post-test score. It can be proved with score of post-test was high than pre-test, although the score still below the KKM. 5 % students can pass KKM values and 95 % students cannot pass KKM values. Based on theory stated before, Bijami, Kashef, & Nejad (2013) concluded that peer feedback has effect in writing classes because of cognitive and social benefit which is obtained from the peer feedback. Almost same with first experimental class that stated above, in third experimental class used self-feedback strategy with same composition with peer feedback, where the student got the treatment from their self which focused on error feedback the composition ―circle the errors, and provide the corrections. The researcher was started by giving pre-test, applying treatment, and giving post-test. The test is done in order to know the students‘ writing skill. The result of test was computed by using SPSS 24 version that can be seen the lowest score is 178 | JISAE. Volume 6 Number 2 September 2020. 18 and the highest score is 72. Then the mean is 57,93, the median is 60,00, and the standard deviation is 10,417. The result of teaching writing was given feedback by their self using self-feedback strategy is good because the students get significant differences between pre- test and post-test score. It can be proved with score of post-test was high than pre-test, although the score still below the KKM. 7 % students can pass KKM values and 93 % students cannot pass KKM values. Based on the result and theory above. Nicol & MacFarlane-Dick (2006) concluded that by using self-feedback the students can lead to re- interpretation of tasks or adjusting internal goals, tactics and strategies. Different with first experimental class and second experimental class that stated above, in third experimental class used teacher feedback strategy with different composition. Third experimental class taught by the researcher using teacher feedback strategy focused on error feedback with the composition ―circle the errors, categorize the errors and provide the corrections. The researcher started by giving pre-test, applying treatment, and giving post- test. The test was done in order to know the students‘ writing skill. The result of test was computer by using SPSS 24 version that could be seen the lowest score is 40 and the highest score is 87. Then the mean is 57,22, the median is 55,00, and the standard deviation is 10,135. The result of teaching writing taught by the researcher using teacher feedback strategy is good because the students get significant differences between pre-test and post-test score. It can be proved with score of post-test was high than pre-test, although the score still below the KKM. 12 % students can pass KKM values and 88 % students cannot pass KKM values. Based on explanation above Ruegg (2015) stated that teacher‘s feedback can be more effective when students are given the freedom to decide what kind of feedback they would like. Furthermore, this research supported the previous research entitled; “A Comparison of Peer, Teacher and Self-Feedback on the Reduction of Language Error in Student Essay at the Tenth Grade of SMA Negeri 3 Kabupaten Tangerang Academic Year 2019-2020” is an effective way to teach writing essay. In summary, using peer, teacher and self-feedback on reduction language error in students’ essay give positive effect, but teacher feedback given more affection on reduction language error of students essay, especially for the tenth grade students of SMA Negeri 3 Kabupaten Tangerang. CONCLUSION Based on the formulation of the problem, the research objective, the hypothesis testing and analysis result, it can be concluded that: 1. The students’ score in first, second and third experimental class are increased, but many students have a score below the minimum criteria (KKM). In first experimental class, 5% students could pass KKM values and 95% students could not pass KKM values. Then, in second experimental class only 7% students could pass KKM values and 93% students could not pass KKM values, and in second experimental class only 12% students could pass KKM values and 88% students could not pass KKM values. It means, almost all of students have score under KKM. 2. The hypothesis data was tested using formula of t-test, it was known that in first experimental class Ha is accepted and Ho is rejected. It means, there is a significant difference between pre-test and post-test in first experimental class using peer feedback at the tenth grade of SMAN 3 Kab. Tangerang in Academic Year 2019/2020. 3. The hypothesis data was tested using formula of t-test, it was known that in first experimental class Ha is accepted and Ho is rejected. It means, there is a significant difference between pre-test and post-test in first experimental class using teacher feedback at the tenth grade of SMAN 3 Kab. Tangerang in Academic Year 2019/2020. 179 | JISAE. Volume 6 Number 2 September 2020. 4. The hypothesis data was tested using formula of t-test, it was known that in first experimental class Ha is accepted and Ho is rejected. It means, there is a significant difference between pre-test and post-test in first experimental class using self-feedback at the tenth grade of SMAN 3 Kab. Tangerang in Academic Year 2019/2020. Teacher feedback has significant effect than peer and self-feedback on reduction of language errors in students writing essays. So, its better teacher feedback on reduction of language errors in students writing essays. REFERENCES Baker, K. M. (2016). Peer review as a strategy for improving students’ writing process. Active Learning in Higher Education, 17(3), 179–192. https://doi.org/10.1177/1469787416654794 Bijami, M., Kashef, S. H., & Nejad, M. S. (2013). Peer Feedback in Learning English Writing: Advantages and Disadvantages. Journal of Studies in Education, 3(4), 91. https://doi.org/10.5296/jse.v3i4.4314 Birjandi, P., & Tamjid, N. H. (2011). Assessment & Evaluation in Higher Education The role of self- , peer and teacher assessment in promoting Iranian EFL learners ’ writing performance. Assessment and Evaluation in Higher Education, 37(5), 37–41. https://doi.org/10.1080/02602938.2010.549204 Björk, L., Bräuer, G., Rienecker, L., & Jörgensen, P. S. (2003). Teaching Academic Writing in Europe (G. Rijlaarsdam (ed.); First). Kluwer Academic Publishers. Boud, D., Cohen, R., & Sampson, J. (1999). Peer learning and assessment. Assessment and Evaluation in Higher Education, 24(4), 413–426. https://doi.org/10.1080/0260293990240405 Brown, H. D. (2003). Language Assessment Principles and Classroom Practices (Third). Longman. https://doi.org/10.1017/CBO9781107415324.004 Dehkordi, M. E., & Allami, H. (2012). Evidentiality in Academic Writing. Theory and Practice in Language Studies, 2(9), 2–11. https://doi.org/10.4304/tpls.2.9.1895-1904 Diab, N. M. (2016). A comparison of peer, teacher and self-feedback on the reduction of language errors in student essays. System, 57, 55–65. https://doi.org/10.1016/j.system.2015.12.014 Esfandiari, R., & Myford, C. M. (2013). Severity differences among self-assessors, peer- assessors, and teacher assessors rating EFL essays. Assessing Writing, 18(2), 111–131. https://doi.org/10.1016/j.asw.2012.12.002 Gould, J. S., & Gould, E. J. (2004). Four Square Writing in the Content Areas (first). Teaching & Learning Company. Graham, S., & Perin, D. (2007). A Meta-Analysis of Writing Instruction for Adolescent Students. Journal of Educational Psychology, 99(3), 445–476. https://doi.org/10.1037/0022-0663.99.3.445 Hyland, K. (2016). Methods and Methodologies in Second Language Writing Research. System, 59, 116–125. https://doi.org/10.1016/j.system.2016.05.002 Hyland, K., & Hyland, F. (2014). State of the art article Feedback on second language students ’ writing. Language Teaching, 39, 83–101. https://doi.org/10.1017/S0261444806003399 Lindblom-Ylanne, S., Pihlajamaki, H., & Kotkas, T. (2014). Self-, peer- and teacher- assessment of student essays. Active Learning in Higher Education, 7(1), 51–62. https://doi.org/10.1177/1469787406061148 Nicol, D., & MacFarlane-Dick, D. (2006). Formative assessment and selfregulated learning: A model and seven principles of good feedback practice. Studies in Higher Education, 31(2), 199–218. https://doi.org/10.1080/03075070600572090 180 | JISAE. Volume 6 Number 2 September 2020. Nik, Y. A., Hamzah, A., & Hasollah, H. F. bin. (2010). Why Writing in ESL is Difficult for Undergraduates in A Public University in Malaysia. International Conference on Educational and Information Technology (ICEIT), 2–7. Raimes, A. (1983). Techniques in Teaching Writing (R. N. Campnell & W. E. Rutherford (eds.); First). Oxford University Press. Razali, R., & Jupri, R. (2014). Exploring Teacher Written Feedback and Student Revisions on ESL Students’ Writing. IOSR Journal of Humanities and Social Science, 19(5), 63– 70. https://doi.org/10.9790/0837-19556370 Ruegg, R. (2015). The relative effects of peer and teacher feedback on improvement in EFL students’ writing ability. Linguistics and Education, 29, 73–82. https://doi.org/10.1016/j.linged.2014.12.001 Salima, R. (2013). Measures of Eliminating EFL Students’ Errors in Writing. Procedia - Social and Behavioral Sciences, 69, 318–327. https://doi.org/10.1016/j.sbspro.2012.11.416 Setyowati, L., Sukmawa, S., & Latief, M. A. (2017). Solving the Students’ Problems in Writing Argumentative Essay Through the Provision of Planning. In Celt: A Journal of Culture, English Language Teaching & Literature (Vol. 17, Issue 1). https://doi.org/10.24167/celt.v17i1.1140 Smemoe, W. B. (2018). Peer Feedback. The TESOL Encyclopedia of English Language Teaching, 0(0), 1–5. https://doi.org/10.1002/9781118784235.eelt0540 Xiang, W. (2004). Encouraging self-monitoring in writing by Chinese students. ELT Journal, 58(3), 238–246. Yu, S. (2019). Learning from giving peer feedback on postgraduate theses: Voices from Master’s students in the Macau EFL context. Assessing Writing, 1–11. https://doi.org/10.1016/j.asw.2019.03.004