1 THE EFFECT OF TEACHERS’ DIRECT AND INDIRECT FEEDBACK ON STUDENT’S WRITING ABILITY Yuli Latifah1, Bambang Suwarno2, Irma Diani 3 123English Education Postgraduate Program Bengkulu University, Indonesia E-mails: Yulilatifah22@gmail.com, wdsaraswati@gmail.com, irmabengkulu @yahoo.com This study is intended to find out any difference in effectiveness between direct and indirect feedback on students’ writing ability. The design was Quasi Experiment. The population consisted of the students of the X class of Senior High School 4, Rejang Lebong, Curup, Indonesia. The samples comprised 32 students in group 1 and 32 Students in group 2. The instrument was a writing test. In the data analysis, the researcher used the normality, homogeneity, and t-test.These were calculated and analyzed by using SPSS 20. The tests consisted of pre-test and post-test. In post-test, there was no significant difference between groups on all aspect. There were two results of this study. Firstly, in the post test there was a difference of mean score between experiment group 1 and experiment group 2. In general ability, in group 1 increased by 15.59. Meanwhile, the group 2 increased by 2.60. This indicated that there was an effect of indirect feedback on students’ writing ability. From t-test calculation, t-count was 3.274 which t- table was 1.670. It showed that t obtained was bigger than t-table (3.274 > 1.670). H0 was rejected and H1 was accepted. In other words, indirect feedback technique was effective on students’ writing ability. Based on the data analysis, the indirect feedback technique was positively effective in increasing on students’ writing ability, on general writing ability. There were also significant differences between both groups in the writing ability aspects such as organization, language use and vocabulary. Keyword: Writing, Indirect Feedback, Direct Feedback, Experimental Study. Penelitian ini bertujuan untuk menemukan perbedaan efektivitas antara respon langsung dengan respon tidak langsung pada kemampuan menulis siswa. Penelitian ini menggunakan Quasi Eksperimen. Populasi dari penelitian ini adalah seluruh siswa kelas X di SMA N 4, Rejang Lebong, Curup, Indonesia. Sampel penelitian ini terdiri atas 32 siswa dari group 1 dan 32 siswa dari group 2. Instrumen yang digunakan adalah tes menulis. Dalam menganalisis data, peneliti menggunakan normality, homogeniti, dan t-test. Mereka dihitung dan dianalisis dengan menggunakan SPSS 20. Tes terdiri dari pre-test dan post-test. Ada dua hasil dari penelitian ini. Pertama, ada perbedaan skor rata-rata antara kelompok eksperimen group 1 dan kelompok eksperimen group 2. Dalam kemampuan secara umum, peningkatan skor dari kelompok 1 adalah 15,59. Sementara itu, grup 2 meningkat sebnyak 2,60. Ini menunjukkan bahwa ada pengaruh umpan balik tidak langsung pada kemampuan menulis siswa. Dari perhitungan t-test, nilai t-hitung adalah 3,274 dan t-tabel adalah 1,670. Ini menunjukkan bahwa t yang diperoleh lebih besar dari t-tabel (3,274> 1,670). Dapat disimpulkan bahwa H0 ditolak dan H1 diterima. Dalam kata lain, teknik umpan balik tidak langsung efektif pada kemampuan menulis siswa. Kedua, ditemukan bahwa motivasi menulis siswa juga meningkat setelah pelaksanaan perlakuan. Berdasarkan analisis data, teknik umpan balik tidak langsung secara positif efektif dalam meningkatkan kemampuan menulis siswa. Pada kemampuan menulis umum, ada juga perbedaan signifikan antara kedua kelompok dalam aspek kemampuan menulis seperti organisasi, penggunaan bahasa dan kosa kata. Keyword: Menulis, Umpan balik tidak langsung, Umpan balik langsung, Penelitian eksperimen. mailto:Yulilatifah22@gmail.com, mailto:wdsaraswati@gmail.com, mailto:wdsaraswati@gmail.com, mailto:dank_aron@yahoo.com 2 INTRODUCTION Writing is one of the activities that form art. It has ability to describe and to explain about opinion or idea in our brain. It gives new information to readers about knowledge him or herself but also can tell about characters of a writer without meeting the writer. Moreover, writing as communicative process; the students can express their feelings, their opinions, or their ideas to the reader. Their writing will be read by the reader. The reader can know about writing quality of someone. So, to make good writing, the writer or the students must know the process of writing. Astuti (2013) states that many grammatical mistakes are made in their writing such as spelling, fragment, run ones, concord, punctuation, and organization. It makes the students confused. They do not really know how to correct their mistakes in their writing work. As a result, the students do not make any progress in writing skills. According to Adas and Bakir (2013), there are many reasons that causing difficulties in writing of second language learners. Those are: (1) Lack of students‘motivation and teachers‘interests. (2) Learners have limited vocabulary of second language. Therefore, students end up repeating the same words; this hinders creativity. And then (3). English language learners don‘t use invented spelling and their written texts are restricted to words which they know. (4) Then, The present tense is the only tense used in their writing. (5) The students‘writing is difficult to understand because of the illstructured sentences in composition. And also, (6) students are unwilling to share their work with other students and they don‘t get the suitable feedback. (7)The last, when the learners read their writing aloud, they couldn‘t distinguish whether what they read or write is right or wrong. Riani (2016) found that in order to solve the students’ difficulties in writing. Teachers can modify their assessment in correcting students’ work. Direct and indirect feedback is believed to contribute in improving student’s writing ability. Jamalinesari (2015) found his study that direct feedback was more effective than indirect feedback. The students can find error on their writing quickly. So direct feedback method can used by teacher. Eslami (2014) said that implementation of indirect method of error correction will necessarily call for sufficient linguistic knowledge possessed by students to self- correct errors and also getting used to self edit their own text. Using indirect feedback strategies may strongly demand somewhat focused error correction especially with low-level-of proficiency learners.Utami (2012) found that teachers’ direct feedback improved the students writing skill in a short period and also teachers’ direct feedback need a long time in the class with a low proficiency level. As stated by Sokolik (2003), there are four principles of teaching writing. The principles are understanding students’ reason for writing, providing many opportunities for the students to write, making feedback helpful and meaningful, and clarifying how the students’ writing will be evaluated. The first principle is understanding students’ reason for writing. It is about the importance of both teacher and students’ goals. Teacher needs to understand and convey goals to the students so that the students can apply the writing skills that 3 they learn. The second principle is providing many opportunities for the students to write. Since writing is in part a physical activity, the teacher should give more practice for the students in the learning activities and the practice itself should provide different types of writing. By practicing more, the students can construct correct English words and also sentences. The third principle is making feedback helpful and meaningful for the students. In writing, students need feedback from their teacher. Teacher should give clear feedback to the students in which the students understand the vocabulary or symbol that is used by the teacher. If it is necessary, teacher can discuss the feedback with the students in the class so that the students can see the errors on their writing. In this way, the students can learn from their mistakes and be more aware of making errors. The last principle is clarifying how the students’ writing will be evaluated. The teacher should give an evaluation which is clear for the students such as how important creativity or originality of ideas is; how important following a particular written format is; how important grammatical accuracy is; how important that the assignment include recently taught material is; and how important accuracy is in spelling and punctuation. Besides, the teacher needs to make a scoring rubric so that the students know what kind of aspects and requirements that will be assessed in writing. Feedback as an exact word is the teachers offer new information or specific information to their students that related about process in learning teaching. A crucial question is what this feedback should look like. A feedback type commonly used in classroom is corrective feedback: the marking of a student‘s error by the teacher. Recently, there has been quite some disagreement in the academic field on the benefits of this kind of feedback on learners‘written output. Feedback is defined as teacher's input to a writer's composition in the form of information to be used for revision (Keh, 1990). It is also defined as information provided by teachers to help students trouble-shoot their performance (Nicole and Macfarlane, 2006). I would define it as teacher's response to students' writing in the form of oral or written comments that aim to help them improve their writing performance. Feedback has an important role in teaching technique that is given by teacher. However, feedback is important to increase students‘learning process, this statement similar with Aghajanloom et al (2016) said that a clear implication of the study is that written corrective feedback types (specially unfocused direct corrective feedback) should be emphasized as an essential tool for developing writing ability of intermediate EFL learners. Its mean that, the teacher should give students the appropriate feedback to achieve the goal of feedback itself. The aim of feedback is to deliver students with perception that helps them to increase their writing, the students can learn more not only know new vocabulary but also they can memorize verb 2 that form irregular or regular verb. Then, they can arrange word by word with good grammar. According to Lewis (2002: 3-4), feedback has several purposes such as; a. Feedback provides information for teachers and students 4 Teacher able to get information about the students progress and directly is a form of evaluation for the teacher teaching. For the learners, feedback is ongoing form assessment which is more focused on the progress rather than value or marks. By highlighthing strengths and weakness, the comments provide information about individual progress, unlike marks or rank. And then compare one student with another. The comments can give direction about language, by stating a rule or giving an example. b. Feedback provides students with language input Illustrate how language is used in a communication. That is why it is important, to extend student’ language by writing comments in language use. In this way, the students can learn new vocabulary and structures in structures in context. c. Feedback is a form of motivation Feedback can encourage the students to learn and to use English language for the best the students’ ability. As the teacher find out more about the students like as encouragement can take personal situations report. d. Feedback provides students with advice about learning The teachers can provide students with more than simply description of their language use. Comments can be made for the students in learning process. e. Feedback able to lead students toward autonomy. One long term purpose of feedback is to lead to the point when they can find their own mistakes. Direct feedback is offered when the teacher writes the correct form on the student’s paper, while indirect error feedback is offered when the teacher indicates the location of the error on the paper by underlining, highlighting or circling it with providing the correct form (Lee, 2004). Ferris and Roberts (2001) said Indirect feedback is a type of written feedback in which teacher indicates an error has been made by students on their writing, but the teacher does not offer or provide the correct form of the error. The teacher only gives correction on students’ writing. Indirect corrective feedback is commonly presented by giving indicators. The indicators may be in one of four ways: recording in the margin the number of errors in a given line, using a code to show where the error has occurred and what type or error it is or underlining or circling the errors. Ferris and Roberts (2001) said Indirect feedback is a type of written feedback in which teacher indicates an error has been made by students on their writing, but the teacher does not offer or provide the correct form of the error. The teacher only gives correction on students’ writing. Indirect corrective feedback is commonly presented by giving indicators. The indicators may be in one of four ways: recording in the margin the number of errors in a given line, using a code to show where the error has occurred and what type or error it is or underlining or circling the errors. Therefore, it needs to find the effective method between direct or indirect feedback used by teacher in encouraging student to do good writing by posing four questions: (1) Is there any difference in effectiveness between direct and indirect feedback on students’ writing ability? (2) Is there any difference in effectiveness between direct and indirect feedback on students’ writing ability in the aspect of content? (3) Is 5 there any difference in effectiveness between direct and indirect feedback on students’ writing ability in the aspect of language use? (4) Is there any difference in effectiveness between direct and indirect feedback on students’ writing ability in the aspect of vocabulary? METHOD This research use, the reseacher conducted a quasi experimental research design because according to gay (2000: 367) experimental research was the only type of research that can test hypotheses to establish cause and effect relationship. The design of this research was as follow Groups Pre- test Treatments Post- test (E1) Y1 X1 Y2 (E2) Y1 X2 Y2 Note: E1 = Experimental group 1 E2 = Experimental group 2 Y1 = Pre-test Y2 = Post-test X1 = Using indirect feedback X2 = Using direct feedback The participant of this research were all the students X of Senior High School namely was there were 137 students. The researcher used cluster sampling. The sample of this research consist of the students of two classes, which are selected after pre-test. The sample of this research was the students of X1 (social science class) were 32 and X2 (Social science class) were 32, the total sample 64 students Table 1 : Sample of The Research No Class Total 1. X1 (Social science class) 32 students 2. X2 (Social science class) 32 students The instrument of this research was a writing test. Students’ scores are rated by two (2) raters. The scores from two (2) raters are validated by using product moment formula: Note: r count is compared with r table if r count was greater than r table then there was a correlation between r1 (the score from rater 1) and r2 (the score from the rater 2). Then, the scores are averaged. If r count was less than r table then scoring is repeated. There were four stages for these research procedures; (a) Planning is including the drafting and preparation of study design to obtain data, such as the researcher gave pretest for the students to write a recount text. (b) Pre-Test, before started to give a treatment, the researcher gave a pre-test to the students. The purpose of pre-test was to know the students writing ability. First, the researcher gave the theme past experience and the students write it on their sheet. (c) Treatment gave through direct and indirect feedback for at least 8 meetings. In treatment, the researcher gave steps in direct and indirect feedback. The researcher explained about recount text materials. Then, the researcher asked the students to write down recount text. (d) Post-test was given after the treatment to the students. Post- test was given at the end of treatment to know the improving writing ability of recount text and two classes. The researcher gave the post-test and the post- test test was same with the pre-test test. After the researcher gave post-test, the writing test are checked and scored. Researcher calculated the test and analyzes it by using the formula state in the data analysis. The researcher generalized the data based on the analysis. (e) Report writing, the researcher wrote the result of comparison between indirect and direct feedback after did pre-test, treatment, and post-test. FINDINGS AND DISCUSSION The result and findings from this research, the first phase was finding 6 students’ scores are rated by two (2) raters. The scores from two (2) raters are validated by using product moment formula: Table 2. Validation of Scores Group rcount rtabel Result Indirect Feedback Pre-Test 0.414 0.2461 Valid Indirect Feedback Post-Test 0.451 0.2075 Valid Direct Feedback Pre-Test 0.859 0.2075 Valid Direct Feedback Post-Test 0.583 0.2075 Valid Based on the table above, r count was compared with r table, so the result rcount was greater than rtable then there was a correlation between r1 (the scores from rater 1) and r2 (the scores from the rater 2). Then, the scores are averaged. Its mean that, scores of the two raters on the pre-test and post- test had statistically significant correlation. The second phase, the reseacher checked the correlation between two inter- raters of writing test before the reseacher gave the result of pre test and post test. On pre- test, the Pearsons’ correlation revealed for experiment group 1 (Indirect Feedback) r= -0.105, p= 0.567 and also in Experimental group 2 was r = -0.105, p = 0.567. It means that the scores of two raters on the pre- test had significant correlation. On the post- test result, the Pearsons’ correlation revealed for experiment group 1 (Indirect Feedback) r= 0.558, p= 0.001 and also in Experimental group 2 was r = 0.558, p = 0.001. It means that the scores of two raters on the pre- test had statisticaly significant correlation on the post test. a. Result of pre-test and post-test in overall aspect The result of pre-test was calculated by using the formula of F value. From the F value calculation, F obtained was smaller than F table (0.579 < 3.996), in other words, both sample have equal variances. Thus, analysis was continued by using the t- test for equal variances. The result of pre-test for experiment group1 was calculated by using the formula of normality test. From the Chi Square Value, χ2 obtained was smaller than χ2 table (0.127 < 46.194). And experiment group 2, χ2 obtained was smaller than χ2 table (0.111 < 46.194). The result of post-test for experiment group 1 was calculated by using the formula of normality test. From the Chi Square Value, χ2 obtained was smaller than χ2 table (0.127 < 46.194). And experiment group 2, χ2 obtained was smaller than χ2 table (0.111 < 46.194). From the calculation, it was found that score of experiment group 1 was 60.29 and experiment group 2 was 62.11. The following hypothesis (this hypothesis for the research question number one is reviewed: There is no a significant difference in student’s total score average between the class that uses direct feedback and the class that uses indirect feedback. Then, there is a significant difference in student’s total score average between the class that uses direct feedback and the class that uses indirect feedback. From t-test calculation for total pre- test, the value of t-count was -0.513 and t-table was 1.670. It showed that t obtained was smaller than t-table (t-count < t-table; -0.513 < 1.670). Based on the explanation above, the differences between the two groups were small. Briefly, these two groups have similar ability and therefore those groups can be accepted as the sample of the research. Based on the result, In order to to analyze the pre-test and post-test result the score of experimental group 1 and experimetal group 2 were compared by using t-test was used to see whether the treatment could improve students’ writing ability or not. From t-test calculation for total post- test, the value of t-count was 3.274 and t- table was 1.670. It showed that t obtained was bigger than t-table (t-count > t-table; 3.274 > 1.670). It can be concluded that H0 was rejected and H1 was accepted. In other words, there was a difference in post-test score average between the 7 experimental group 1 and group 2. Shortly, this research showed that indirect feedback can be effective for students’ writing ability. b. Difference on Performance Aspects 1. Content Aspect Score Analysis The result of pre-test for group 1 was calculated by using the formula of normality test. For group 1, χ2 obtained was smaller than χ2 table (0.263 < 46.194). And group 2, χ2 obtained was smaller than χ2 table (0.208 < 46.194). Thus, the data from both groups were normal. Table 3. Homogeneity test The result of post-test for experiment group 1 was calculated by using the formula of normality test. From the Chi Square Value, χ 2 obtained = 0.210 dan χ 2 tabel = 46.194 (χ 2 hitung < χ 2 tabel ; 0,210< 46.194) it can be concluded that the post test score distribution is normal, And experiment group 2, χ2 obtained was smaller than χ2 table (0.271 < 46.194). Table 4. Homogeneity test The result of post-test was calculated by using the formula of F- value. From the F-value calculation, the F-count was 3.138, whereas the F- table was 3.996. It means that F-count was smaller than F-table (3.138 < 3.996. In other words, both samples have equal variances. The following hypothesis (this hypothesis for the research question number two point b) is reviewed. In the pre test result before the reseacher gave treatments were there was no a difference in student’s score average for content aspect between the class that uses direct feedback and the class that uses indirect feedback. However, in the post test after treatments were there was a difference in student’s score average special for content aspect between the class that uses direct feedback and the class that uses indirect feedback. It can be concluded that H0 was rejected and H1 was accepted. Based on the summary on table 11, indirect feedback was effective for the whole aspect of writing ability. Table 5. Content Pre-Test Result Class Mean score t-count t-table Difference Remark E1 68.75 -0. 089 1.670 No Significant difference Two tailed α = 0.05 E2 69.14 Note : E1 : Experiment group 1 E2 : Experiment group 2 From t-test calculation, the value of t-count was -0.089 and t-table was 1.670. It showed that t obtained was smaller than t-table (-0.089 < 1.670). Based on the explanation above, the differences between the two groups were small. Briefly, these two groups have similar ability and therefore those groups can be accepted as the sample of the research. Based on the summary on table 10, indirect feedback was not effective yet for pre test result. From the pre-test result, the effectiveness between experiment group 1 and experiment group 2 have not looked yet because the students forgot of generic structure or language features from the text. Table 6. Content Post-test Result Clas s Mean score t-count t-table Difference Remark E1 84.77 3.138 1.670 Significant difference Two tailed α = 0.05 E2 71.48 Note : E1 : Experiment group 1 E2 : Experiment group 2 From the calculation, it was found that mean score of experimental group 1 was 84.77 and the mean score of experiment group 2 was 71.48. Thus, the analysis was continued by using t-test formula for equal variances. From t-test calculation, the value of t-count was 3.610 and t-table was 1.670. It showed that t obtained was smaller than t-table (t-count > t-table; 3.610 > 1.670). It can be concluded that H0 was rejected and H1 was accepted. In other Levene's Test for Equality of Variances F Sig. .257 .614 Levene's Test for Equality of Variances F Sig. 3.138 .081 8 words, there was a difference in post-test score average between the experimental group and control group. Based on the summary on table 11, indirect feedback was effective for the whole aspect of writing ability. 2. Language Use Aspect Score Analysis In order to answer the third problem or research question 2(b), the researcher analyzed the result of Language use aspect score analysis of pre-test. The result of pre-test for experiment group was calculated by using the formula of normality test. From the Chi Square Value, χ2 obtained was smaller than χ2 table (0.190 < 46.194). And experiment group 2, χ2 obtained was smaller than χ2 table (0.169 < 46.194). Then the result of pre-test was calculated by using the formula of F value. From the F value calculation, the F obtained was smaller than F table (1.445 < 3,996), in other words, both sample have equal variances. Thus, analysis was continued by using the t-test for equal variances. Table 7. Homogeneity test of Language Use Pre-Test The following hypothesis (this hypothesis for the research question number two point b) is reviewed. In the pre test result before the reseacher gave treatments were there was no a difference in student’s score average for language use aspect between the class that uses direct feedback and the class that uses indirect feedback. However, in the post test after treatments were there was a difference in student’s score average special for language use aspect between the class that uses direct feedback and the class that uses indirect feedback. It can be concluded that H0 was rejected and H1 was accepted. Its mean that, indirect feedback was effective for the whole aspect of writing ability. Table 8 Language Use Pre-Test Result Group Mean score t-count t-table Difference Remark E1 54.69 -1.072 1.670 No Significant difference Two tailed α = 0.05 E2 59.38 Note : E1 : Experiment group 1 E2 : Experiment group 2 From t-test calculation, the value of t-count was -1.072 and t-table was 1.670. It showed that t obtained was smaller than t-table (t-count < t-table; - 1.072 < 1.670). Based on the explanation above, the differences between the two groups were small or not significant briefly, these two groups have similar ability. Table 9. Language Post-Test Result Note : E1 : Experiment group 1 E2 : Experiment group 2 From the calculation, it was found that mean score of experimental group 1 was 70.70 and the mean score of experimental group 2 was 60.94. From t- test calculation, the value of t-count was 2.773 and t-table was 1.670. It showed that t obtained was smaller than t-table (t-count > t-table; 2.773 > 1.670). The result language use post-test score was used to find out maximal score, minimal score, and mean score from both groups that can be seen in the following table. It can be concluded that H0 was rejected and H1 was accepted. In other words, there was a difference in post- test score average between the experimental group 1 and experiment group 2 In other word. Its mean that, indirect feedback was more effective than direct feedback. Levene's Test for Equality of Variances F Sig. 1.445 .234 Group Mean score t-count t-table Difference Remark E1 70.70 2.773 1,670 Significant difference Two tailed α = 0.05 E2 60.94 9 3. Vocabulary Aspect Score Analysis The result of pre-test for experiment group was calculated by using the formula of normality test. From the Chi Square Value, χ2 obtained was smaller than χ2 table (0.191 < 46.194). And experiment group 2, χ2 obtained was smaller than χ2 table (0.228 < 46.194). Then the result of pre-test was calculated by using the formula of F value. From the F value calculation, the F obtained was bigger than F table (0.062<3.996 ). In other words, both samples have no equal variances. Table 10. Homogeneity test of Vocabulary pre-test The result of post-test for experimental group 2 was calculated by using the formula of normality test. From the Chi Square Value, χ2 obtained was smaller than χ2 table (0.168 < 46.194). And experiment group 2, χ2 obtained was smaller than χ2 table 0.173 < 46.194). The result of post-test was calculated by using the formula of F- value. From the F-value calculation, the F-count was 0.539 whereas the F- table was 3.996. It means that F-count was smaller than F-table (0.539 < 3.996. In other words, both samples have equal variances. Table 11. Homogeneity test of Vocabulary post- test The following hypothesis (this hypothesis for the research question number two point (c) is reviewed. In the pre test result before the reseacher gave treatments were there was no a difference in student’s score average for content aspect between the class that uses direct feedback and the class that uses indirect feedback. However, in the post test after treatments were there was a significant difference in student’s score average special for content aspect between the class that uses direct feedback and the class that uses indirect feedback. It can be concluded that H0 was rejected and H1 was accepted. Based on the summary on table 17, indirect feedback was effective for the whole aspect of writing ability. In order to answer the last problem or research question 2c, the researcher analyzed the result of Vocabulary score analysis of pre-test. The result was compared in the following table. Table 12. Vocabulary Pre-Test Result Group Mean score t-count t-table Difference Remark E1 57.42 -0,102 1,670 No Significant difference Two tailed α = 0.05 E2 57.81 Note : E1 : Experiment group 1 E2 : Experiment group 2 Thus, analysis was not continued by using the t-test for equal variances.From t-test calculation, the value of t-count was -0.102 and t-table was 1.670. It showed that t obtained was smaller than t-table (t-count < t-table; - 0,102 < 1.670). Based on the explanation above, the differences between the two groups were small or not significant briefly, these two groups have similar ability. The result vocabulary post-test score was used to find out maximal score, minimal score, and mean score from both groups that can be seen in the following table. Table 13. Vocabulary Post-Test Result Group Mean score t-count t-table Difference Remark E1 71.88 2.133 1,670 Significant difference Two tailed α = 0.05 E2 61.72 Note : E1 : Experiment group 1 E2 : Experiment group 2 Thus, the analysis was continued by using t-test formula for equal variances. and t-table = 1,670. t –count = 2.133 and t-table = 1,670 from t-test calculation, the value of t-count was 2.133 and t-table was 1.670. It showed that t Levene's Test for Equality of Variances F Sig. .062 .804 Levene's Test for Equality of Variances F Sig. .539 .465 10 obtained was bigger than t-table (t-count > t-table; 2.133 > 1.670). It can be concluded that H1 was accepted and H0 was rejected. In other words, there was a significant difference in post-test score average between the experimental group 1 and experiment group 2. Shortly, this research was successful and showed that applying indirect feedback could improve students’ writing ability at the second grade of SMA Negeri 4 Rejang Lebong. DISCUSSION As previously mentioned, the main purpose of this research was to find out differences in effectivenes between direct and indirect feedback on students’ writing ability. Based on the data analysis, the findings are;The reseacher found that the pre-test should that there was no significant differerences between experiments group 1 and experiment group 2 .However, in the post-test, there was significant between experiment group 1, experiment group 2. It’s mean that the reseacher could answer for number one that was indirect feedback was more effective than direct feedback. In this research the researcher discusses about three aspects in writing ability such as, content aspect, language use aspect and vocabulary aspect. From the result, there was significant difference in mean score between the experimental group 1 and experimental group 2 for all aspect. For the second questions, the reseacher found that the result of post test from experiment group 1 and group 2 was different. The hyphothesis from the result was H1 was accepted and then H0 was rejected because the score from experiment group 1 that uses indirect feedback was higher than group 2. Its show that if indirect feedback was more effective than direct feedback in writing ability of content aspect. The reseacher found the same result for language use aspect. In here, the result showed the students’ score in experiment group 1 was bigger than students’ score in group 2. Its mean that H1 was accepted and H0 was rejected. The reseacher concluded that indirect feedback was more effective than direct feedback to increase the students’ writng ability. The reseacher could answer the last question that indirect feedback was more effective than direct feedback in writing ability of vocabulary aspect. The reseacher found in this research, the scores from students that used direct feedback in group 2 was smaller than the scores’ students in group 1. However, in the treatment stages, when the students wrote the text there were some students were noisy, and also they wrote slowly. It made their writing was not maximal. Beside that some students were passive in class and it caused of some factors, such as they were not confident enough to their text. Most of them were afraid of making mistakes. Sometimes, the teaching learning process was also dominated by several students only. Furthermore, after some treatments, the researcher found that the students can understand more about their writing after the researcher gave indirect and direct feedback. The students got treatment that used indirect feedback more active to ask the reseacher about the circling symbol on their paper and then, this technique was more encourage the students to revise their writng so that the students could increase their writng. On condition that the students’ got indirect feedback from the reseacher, they found that indirect feedback given was clear enough. Moreover, the symbols that used were also comprehensible and they were satisfied with it. Then, the thing that they liked the most from indirect feedback was they would be able to know their mistakes that they made in their writing. It was significantly helpful for the students to improve their writing. Besides, after receiving indirect feedback the students were motivated to revise their writing. The students wanted to decrease the correction or the sign on their writing that was given by the reseacher. Additionally, the result of this research also clarified some theories and ideas from the expert. The result of this research was also similar to the result of 11 some previous study. The finding of this study seems to support the theories that there was a significant between indirect feedback and direct feedback. In general, indirect feedback was more effective than direct feedback to improve students’ writing ability or its aspects of writing. CONCLUSION This study aimed to find out whether there was a significant difference on writing ability between students at the tenth grade students of SMA Negeri 4 Rejang Lebong who were taught by using indirect feedback and that were taught using direct feedback. Some conclusions can be drawn. Based on the findings and discussion it could be concluded that, first, indirect feedback was more effective than direct feedback and increasing students’ writing ability. This increasing may result from the types of feedback that encourages learners to know and understand their mistakes so that the students tried to learn independent and they became actively asked questions for the researcher. Moreover, the students became more aware of their errors. In addition, the students felt motivated and challenged after receiving the indirect feedback from the reseacher. Secondly, indirect feedback was more effective to improve the students’ writing ability in the content aspect. It was proven by the scores differences between two groups in the post test scores. The result of post-test analysis showed that It means indirect feedback was more effective than direct feedback. Thirdly, from result of statistical calculation of indirect feedback was more effective than direct feedback in the language use aspect. Its means that indirect feedback was more effective than direct feedback. Furthermore, from result of statistical calculation of indirect feedback was more effective than direct feedback in the vocabulary aspect. The overall conclusion is that, in each aspect of Indirect Feedback was more effective than direct feedback. REFERENCES Adas, D & Ayda, B.(2013). Writing difficulties and new solution: blended learning as an approach to improve writing abilities .International Journal of Humanities and Social Science, 3(9), 255. Astuti, Puji A. (2013).The effectiveness of peer feedback to improve the writing ability of the tenth grade students of SMA Kanisius Harapan Tirtomoyo in the academic year of 2012/2013 (Skripsi). Pend. Bahasa Inggris, Universitas Yogyakarta. Indonesia. Ellis, R.(2008). A typology of written corrective feedback types. ELT Journal, 28(2), 97-107. Eslami, E.(2014). The effects of direct and indirect corrective feedback techniques on EFL students’ writing. Procedia-Social and Behavioral Sciences, 98.445-452. Ferris, D., & Roberts, B. (2001). Error feedback in L2 writing classes. How explicit does it need to be?. Journal of Second Language Writing. 10(3), 161- 184. Jamalinesari, A & friends.(2015). The effect of teacher-written direct vs indirect feedback on students’ writing. Procedia-Social and Behavioral Sciences, 192.116-123. Gay, L.R & Peter A. (2000). Educational Research: Competencies for Analysis and Application. (6thed). Lee, I. (2008). Understanding teachers ’written feedback practices in Hong Kong secondary classrooms. Journal of Second Language Writing, 17, 69-85. Sokolik, M. (2003). Writing. In D. Nunan (ed.), Practical English language teaching. New York, NY: McGraw Hill. Utami, U (2012).Improving students’ writing skill through teacher’s direct feedback in SMA n 1 Jogonalan.skripsi. Pend. Bahasa Inggris, Universitas Yogyakarta, Indonesia. 12