Teaching Grammar-in-context and Its Impact in Minimizing Students' Grammatical Errors

This study is conducted to determine theeffectiveness of teaching grammar-in-context tominimize students‘ grammatical errors in writing. Thedesign of the study was a quasi-experimental with anon-randomized pretest-posttest control group. Thesamples of the study were taken from the population ofthe tenth-grade students. The control group was taughtby conventional grammar which was separately givenwith writing skills. Likewise, the experimental one wastreated by teaching grammar-in-context. The results ofthe study showed that the mean score in the post-testwas higher than that in the pretest; and the mean scoreof experimental group increased 16.20 point after thetreatment. This result indicated that teaching grammarin-context is considered to be effective in minimizingstudents‘ grammatical errors in writing.


INTRODUCTION
There is no doubt that writing is the most difficult skill for L2 learners to master. The difficulty lies not only in generating and organizing ideas, but also translating these ideas into readable text with good grammar. Theoretically, the skill of composing English sentences is much determined by the ability of identifying or recognizing syntactic or writing units: the letter itself, word, phrase, and sentence (Shaw, 1986). The students cannot create correct writing if they do not master grammar. Gebhard (1996) asserts that teachers generally agree that beginning level EFL/ESL writers need to learn the basic conventions of writing that consists of letters, words, and sentences. And one of teacher activities involves teaching students grammatical patterns and functional rules. Because of the shortage of grammar knowledge, the students, even writers, still make many grammatical errors in their writing. So, it is reasonable that grammatical errors in the process of learning English as foreign language needs more attention to correct or, at least, reduce to minimum.
The problem can be revealed through the errors posited as indication that the problem or difficulty in learning English still occurs. Thus, apparently, a correct grammar also plays an important role in EFL writing. As a proof relating to grammar, Mukminatien (1999) obserYHG WKH VWXGHQWV· ZULWLQJ SHUIRUPDQFH LQ WKLV FDVH LQ PDETP, S1-Equivalent Program and Regular S1 students, showing that they were grammatically incompetent due to the errors they had made. Furthermore, having conducted an observation by giving writing test to the tenth-grade students of MAN Lasem on the 7 th of February 2015, it was found that the students still felt confused of what they were going to write; they wrote with many grammatical errors on account of very little mastery in grammar. This is one of H[DPSOHV RI WKHLU ZULWLQJ ´Last semester, I had Ø semester holiday. I *wented to my grandmother* house. I took* a hour to go there. There * is many mango*es tree*. ØMango*es trees * is *more bigger than my mango*es tree. My grandmother *gived a mango*es to me. I did not *ate it because my stomach Ø full. I also *visit rice field. I * sawed *harvest rice there. I *am happy *spend P\ KROLGD\ WKHUHµ.
The example indicated that the student made both morphological errors and syntactical errors. The incorrect use of past tense such as wented, gived, did not ate, visit, sawed; the incorrect use of plural and singular such as tree for plural, trees for singular; the incorrect use of possessive such as grandmother house; the incorrect use of article such as a hour; and the incorrect use of comparative such as more bigger belong to morphological errors. Meanwhile, the disagreement, omission and incorrect use of to be for past tense such Amin, ´«P\ VWRPDFK IXOO µ,,DP KDSS\ «µ,etc. belong to syntactical errors. In accordance with the background and problem faced by the VWXGHQWV WKH UHVHDUFK SUREOHP LV VWDWHG DV ´'R WKH VWXGHQWV WDXJKW by grammar-in-context make less grammatical errors in their writing than those who are taught by grammar FRQYHQWLRQDOO\"µ This research is intended to find out the effectiveness of teaching grammar-in-context to minimize students· grammatical errors in writing. The findings of this study are expected to provide useful information about the effectiveness of teaching grammar-incontext so that the students can learn grammar associated with writing skill (in association with genre). Other significance is to give a practical contribution to teaching learning English in secondary schools, mainly the teaching of grammar in context associated with the teaching of writing skill. Furthermore, it is hoped that other researchers in the same areas can use the findings as additional references in their research.

LITERATURE REVIEW
Teaching Grammar-in-Context Lock (1996) stated that practice exercises in many textbooks of the 1950s, 1960s, and 1970s typically involved manipulation of sentence-level structures, with little or no context. At that time, structural grammar influenced language teaching. In some cases, the exercises could be successfully completed without the learners even understanding the meanings of the forms they were manipulating. As a result, in writing or speaking the learners did not understand what they wrote or spoke, so that many grammatical errors were made.
In rejecting structural language teaching, Communicative Language Teaching (CLT) has been proposed to develop communicative competence in which activities have been designed to maximize opportunity for communication without ignoring grammar. Widdowson (Lock, 1996) has written that a proper understanding of the concept of communicative competence would have revealed that it gives no endorsement for the neglect of grammar.
Unfortunately, for some teachers, the teaching grammar has come to be seen as a minor part of the development of communicative competence, even this time when Competence ² Based Curriculum (CBC) has been applied. In CBC (actually has same SXUSRVH DV &/7 ´JHQUHµ ² some call text type ² is put forward. Grammar can, in fact, play its role in text level practice because it is easier to build up strong associations between structures and their meaning in context, which makes it likely that the students are able to select appropriate structures in similar context. In this case, the teaching grammar in context is meant to relate grammar to text type. For example, when a teacher explains recount text, s/he also explains past tense, and so forth. Teaching grammar-in-context is actually much more inspired by Contextual Teaching Learning (henceforth, it is called CTL). CTL is an approach of teaching and learning that relates the materials and classroom activities to real situation and actual experience focusing on the learning process leading to creativity, critically thinking, and problem solving and being able to apply their knowledge in their daily lives (Nurhadi, 2004).
In relation to the principles of CTL, teaching grammar-incontext attempts to adapt its principles. First, in inquiry, the teachers can observe and progress sWXGHQWV· XQGHUVWDQGLQJ RI JUDPPDU DQG the students analyze grammar from text given by the teachers. 6HFRQG LQ TXHVWLRQLQJ WKH\ JXLGH DQG DVVHVV WKH VWXGHQW·V understanding of grammar in text. Third, in constructivism, they guide the students to construct their prior knowledge of grammar with new experience in analyzing grammar in text. Fourth, in modeling, they give the students examples or models how to construct grammar to be a sentence, a paragraph, even a text. Fifth, in learning community, they allow the students to share ideas with each Amin, Teaching Grammar-in-Context and Its Impact in Minimizing 6WXGHQWV· *UDPPDWLFDO (UURUV 72 other, and collaborate with others to produce some texts. Sixth, in authentic assessment, they measure their knowledge and skill (grammar knowledge in writing skill) from relevant and contextualized tasks; process and products can both be measured. The last, in reflection, both of them review and respond to activities and experiences, and record what they have learned. In this case, the WHDFKHUV FRUUHFW WKH VWXGHQWV· JUDPPDWLFDO HUURUV DQG WKH VWXGHQWV improve their grammatical accuracy in their writing.
In concordance with the text, teaching grammar-in-context could make the students to be critical toward writing text, such as: (1) in text completion, the students have to pay attention to features of context in order to select appropriateness items to complete a text. Some possibilities are: selecting between two or more structures at points in the text (e.g. selecting tense), creating larger blocks of text (e.g. expanding a narrative with descriptive background, focusing on relational process clauses and past continuous tense with action process); (2) in text sequencing, the students have to pay attention to the relationships between grammar and context to sequence out-oforder units (e.g. clauses, sentences, or paragraph) of a text; (3) in text formation, the students recast texts for different contexts and communicative purpose, for example: rewriting a set of rules or formal command focusing on mood and modality; recasting a spoken explanation by someone about how s/he makes something as a written text about how something is made, focusing on voice choice (active or passive voice); and (4) in text reconstruction, the students can reconstruct a sequence of pictures matched with appropriate verb groups to be a text, mainly for the reconstruction of a recount or a narrative. In the reconstruction, grammatical features are the focus of the reconstruction of the text. In text creation, the students produce complete texts, either individually or collaboratively. This begins with work on an input text, focusing on the language features typical of the text type, and leads to the students creating their own texts of the same types (Raimes, 1983).
In addition, as a part of language, grammar-in-context can relate grammar teaching to situational context. It is because language is used in context of situation as well a context of culture (Hammond, 1992). It means that language can be easily understood in relation to the context in which it is used. So does grammar; it can be easily understood or used by students in their writing if it is taught in relation to the context in which grammar is used.

Writing, Grammar, and Errors
Writing is considered as such a complex subject that it is sometimes difficult to teach and to learn (Heaton, 1989). It requires mastery not only of grammatical and rhetorical devices but also of conceptual and judgmental elements. It can also be meant that writing has very close relationship with grammar. Raimes (1983) proposes that writing is considered as a means of reinforcing and manipulating grammatical and rhetorical structures, not as a tool for communication. Through writing, students can reinforce the grammatical structures, idioms, and vocabulary that they have learned. Thus, grammar plays the most important aspect in writing. Bad grammar makes writing imprecise or ambiguous. Collinson (1986) states that without grammar, that is without rules to govern the arrangement of words and making of their plurals, tense, and so on, meaning could not be made clear by writers or understood by readers.
Grammar is worth studying because it can help us to express our ideas clearly and effectively in both speech and writing. Weakness in writing ² particularly incorrectness of sentence structure ² is often due to insufficient understanding of grammar. However, the complexity of grammar leads the students to make errors both in spoken and written English. In an observation, a Senior High School student who has been learning English is good at vocabulary mastery and at developing his ideas, on the contrary, he is bad at grammar. So, when he expressed his ideas ´%HEHUDSD 0XVOLP WLGDN VHULQJ SHUJL NH PDVMLG VHWLDS -XPDW µ in written form, he wrote ´6RPH 0RVOHPV QRW JR WR Amin, Teaching Grammar-in-Context and Its Impact in Minimizing 6WXGHQWV· *UDPPDWLFDO (UURUV 74 the mosque every Friday µ After being taught some rules of present tense, and regularly memorizing and practicing the rules, he could produce negative sentences of present tense such as ´6RPH 0RVOHPV GR QRW JR WR WKH PRVTXH HYHU\ )ULGD\ µ However, he just remembered putting do not in the negative sentence and ignored the changing of Vs/-es into LQILQLWLYH YHUE $V D UHVXOW KH ZURWH ´My friend *do not *goes to the PRVTXH HYHU\ )ULGD\ µ instead of ´0\ IULHQG GRHV QRW JR WR WKH PRVTXH HYHU\ )ULGD\ µ This illustration shows that we need to know grammar when we want to write a sentence or text correctly. In other words, grammar is also very important in writing. Nevertheless, the complexity of grammar often leads the students to make some errors. In fact, making errors in grammar is common problem not only for the students of ESL/EFL but also for the native speakers, and even the professional writers never free of grammatical errors in their writing (Brereton, 1986). Politzer and Ramirez, Burt and Kiparsky (in Dulay et. al., 1982) had proved that several thousand English errors had been made by students learning English in foreign as well as host environments.
Likewise, it does not mean that grammatical errors made by the students can be taken heed. Grammatical errors should be reduced or minimized, and reducing grammatical errors is one of the WHDFKHU·V UHVSRQVLELOLWLHV 0XNPLQDWLHQ VXJJHVWV WKDW WR KHOS the students improve grammatical accuracy, a teacher can make use RI WKH VWXGHQWV· HUURUV DV D base to determine the types of activities that the students should do in class. This is because errors analysis based on an adequate data will show the common weaknesses with which the students need help. Thus, it is worthwhile for the teacher to diagnose WKH VWXGHQWV· JUDPPDWLFDO HUURUV LQ RUGHU WR GHWHUPLQH the kind of activities to give for the students for the sake of PLQLPL]LQJ WKH VWXGHQWV· JUDPPDWLFDO HUURUV LQ ZULWLQJ %HFDXVH RI grammatical problem, of course, treatment to reduce the problem should be grammar. Concisely, bad grammar in writing can result a lot of errors. It means that teaching grammar in better technique can help the students minimize their grammatical errors, and one of the techniques is teaching grammar-in-context.

METHOD
This study was a quasi-experimental research which applied a non-randomized pretest and posttest control group design as introduced by Ary, et al. (2006). They state that research in classroom setting is an example of settings in educational research in which it is difficult to conduct a true-experimental research. The experimental group was taught by using a method of teaching grammar-in-context integrated to writing skill, and the control group was taught by grammar in conventional technique separated from writing skill. Before the experiment was conducted, a pretest was administered to the control and experimental group, and then treatment was given to the experimental group. Meanwhile, the control group was about to be taught by conventional grammar which was separately given with writing skills. In the end, a posttest was given to both groups.
The target population of this study was the tenth-grade students of one State Islamic Senior High School in Rembang -Indonesia in the academic year 2015/2016. Because of too large number of population, the sampling technique was implemented to make the study feasible and representative. The results of the randomization were class X-Science-1 as experimental group and class X-Science-2 as control group.
Based on the research design chosen, the procedures of the research consisted of four phases: pre-test, diagnosing, treatment, and posttest. Pretest was given for both groups, but designing lesson plans and applying strategy were only addressed to the experimental group. After a certain period of treatment, posttest was administered for the two groups. Briefly, the procedures of the research were as follows: (1) the administration of pretest was to measure grammatical errors of the two groups, and enabled the researcher to get LQIRUPDWLRQ DERXW WKH VWXGHQWV· JUDPPDWLFDO HUURUV LQ WKHLU ZULWLQJ WKH DVVHVVLQJ SURFHVV RI WKH ERWK JURXSV· ZULWLQJ UHVXOWV ZDV WR Amin, Teaching Grammar-in-Context and Its Impact in Minimizing 6WXGHQWV· *UDPPDWLFDO (UURUV 76 FRPSDUH EHWZHHQ SUHWHVW DQG SRVWWHVW· VFRUHV ZKHWKHU WKHUH ZDV D significant difference between the result of control group and that of experiment one, (3) the designing of lesson plan of teaching grammar in context was focused on the fields in which the students were taught using grammar-in-context teaching, (4) exposing the experimental group of the research by teaching the students using grammar in context integrated with writing skill for 6 meetings, (5) the administration of posttest for the two groups was to measure grammatical accuracy in their writing, and (6) the degree of the effectiveness of teaching grammar in context was determined by using ANOVA to compare the mean scores of the two groups. 7R REWDLQ WKH VWXGHQWV· GDWD RQ JUDPPDWLFDO DFFXUDF\ RQ writing skill through teaching grammar in context, writing test was decided to use as the instrument of this study. The instruction of the test was asking the students to write a recount text. Based on the research design chosen, the procedures of the research consisted of four phases: pre-test, diagnosing, treatment, and posttest. Data were collected from writing products made by both groups. The data of grammatical errors are, technically, collected from each writing product by doing the following activities. Each piece of writing was first read thoroughly, accurately and critically to identify the errors as questioned in the research problem. Then, previous identified errors were rewritten in the separate lists of data corpus for further analysis. The list of data corpus was respectively the data corpus of errors in the tenth-JUDGH VWXGHQWV· ZULWLQJ IURP ERWK JURups in pretest and posttest.
The data obtained from the pre-test and post-test were analyzed for the sake of (1) finding out the degree of the score reliability; (2) data normality testing; and (3) testing the hypothesis of the research. To avoid the errors that might result from manual data computation, the data gathered were computerized by using SPSS 11.5 for windows.

FINDINGS AND DISCUSSION Findings
The discussion of this research finding is to know the difference between the mean score of the control group and that of the experimental group. If the mean score of the experimental group is higher than that of the control one, the strategy proposed is considered as an effective strategy. To make an ease in analyzing statistically the result of the study, the hypothesis is conversely put LQWR QXOO K\SRWKHVLV WKDW LV ´7KH VWXGHQWV WDXJKW JUDPPDU-incontext do not make less grammatical errors in their writing than WKRVH ZKR DUH WDXJKW JUDPPDU FRQYHQWLRQDOO\ µ

Pretest Results from Control and Experimental Groups
The pretest was given at the very first meeting in preliminary VWXG\ ,W ZDV FRQGXFWHG WR VHH WKH VWXGHQWV· JUDPPDWLFDO HUURUV DQG grammatical accuracy at once in their writing in both control and experimental group before the treatment was given to the experimental group. In this stage, the researcher gave the students three different topics and they chose one. The result of the pretest showed that the students from the control group made 1008 errors out of 1600 total use of words (63% errors) in their writing; specifically 719 errors were syntactical errors and 289 errors were morphological ones. Meanwhile, 976 errors out of 1600 total use of ZRUGV ZHUH HUURUV GHWHFWHG LQ WKH H[SHULPHQWDO JURXS·V compositions; specifically 660 errors of syntactical errors and 316 words errors of morphological ones. These results indicated that the students in both group still made many grammatical errors in their writing.

Posttest Results from Control and Experimental Groups
The posttest was given to both the control group and the experimental group at the end of meeting in the experiment. It was FRQGXFWHG WR VHH WKH VWXGHQWV· JUDPPDWLFDO HUURUV LQ WKHLU ZULWLQJ both in the control and experimental groups after the treatment was Amin, Teaching  given to the experimental group. In this stage, the researcher also gave the students three different topics and they chose one. The result of the posttest detected that grammatical errors from the experimental group were quite significantly reduced to the minimum level compared to the control one. In this posttest, the students from the control group made 1067 errors out of 1700 total use of words (62.76% errors) in their writing; specifically 795 errors were syntactical errors and 272 errors were morphological ones. Meanwhile, 882 errors out of 1800 total use of words (49% errors) ZHUH GHWHFWHG LQ WKH H[SHULPHQWDO JURXS·V FRPSRVLWLRQV VSHFLILFDOO\ 606 errors were syntactical errors and 276 errors were morphological errors. The results show that the experimental group made less grammatical errors than the control one.
In conjunction with the results of the posttest, the data were analyzed to find out the result of reliability, data normality testing, hypothesis testing, and general tendency of the research. Those results are then used to prove the effectiveness of teaching grammarin-FRQWH[W LQ UHODWLRQ WR PLQLPL]LQJ VWXGHQWV· JUDPPDWLFDO HUURUV LQ their writing.

Results of Scoring Reliability Analysis
The main objective of scoring reliability analysis is to find out whether the scores produced by two raters were consistent or not. By observing the scores produced by the two raters in the experimental group, it was found that the difference of the set of the score was 0.4 in the pretest and 0.2 in the posttest. It could be said that the two products were about similar in the mean score of grammatical DFFXUDF\ LQ VWXGHQWV· ZULWLQJ )RU FODULILFDWLRQ 7DEOHV SUHVHQWV WKH comparison between the mean score produced by the two raters.  Table 1 reveals that the difference between the mean scores based on the judgment of the rater 1 and rater 2 for the experimental group in the pretest was 0.4 and in the posttest was 0.2. Meanwhile, for the control group, the difference between the mean scores in the pretest was 0.2 and in the posttest was also 0.2. It seems that the scores produced by the two raters were not too far different. &RQVHTXHQWO\ WKHUH ZDV QR QHHG WR UHDVVHVV WKH VWXGHQWV· SURGXFWV then one of them was reliable to be used for further data processing. In this case, for the sake of hypothesis testing, the product of the first rater was chosen. It is also proven by the coefficient of interrelation reliability (r) as shown in Table 2.  Table 2 shows that the coefficients obtained from the experimental group were respectively 0.89 from pretest and 0.73 from posttest. Meanwhile, the coefficient from the control group was 0.99 ERWK LQ SUHWHVW DQG SRVWWHVW %DVHG RQ :HLJOH·V VFDOH RI LQWHU-rater coefficient correlation, it could be interpreted that the obtained coefficient from the experimental group indicated the high positive interrelation reliability, and from the control one indicated very high positive interrelation reliability. In other words, the scores produced by the two raters were quite consistent.

The Result of Data Normality Testing
The criteria of data normality testing in the study used the level significance .05 in Kolmogorov ² Smirnov Scale. The coefficients of the normality of the data in this scale are between .05 and .200. If the obtained coefficient is more than or equal to .05, it can be interpreted that the distribution of the data is normal.
In this study the data comprised four groups of data: two groups of the scores yielded by the experimental group (pretest and posttest) and two groups produced by the control one. The result of the testing is summarized in Table 3.  Table 3 shows that the obtained value was .079 for experimental group and .092 for control group in the pretest. Furthermore, in the posttest, the experimental group got .063 and .128 for the control one. Since the obtained value exceeded .05 (the level of significance; in this case. .079, .092, .063, and .128 > .05), this revealed that the data did not deviate from a normal distribution. As the distribution of data was normal, then, the data fulfilled the criteria to be used for testing the hypothesis.

Result of Hypothesis Testing
The hypothesis was formulated as tentative answer of the research problem. It was tested using statistical procedure. The presentation of the hypothesis can be stated in the form of statistical hypothesis as follows:

Ho is accepted if F-ratio < F-table Ho is rejected if F-ratio > F-table
To make an ease in analyzing the results of the study, the hypothesis ZDV FRQYHUVHO\ IRUPXODWHG LQWR WKH QXOO K\SRWKHVLV ´7KH students taught grammar-in-context do not make less grammatical errors in their writing than those who are taught grammar conventionally.
The result of the analysis indicated that the obtained F-ratio was 17.969, and F-table was 3.963. By using the same way, comparing the F-ratio to F-table as used in testing the hypothesis, the obtained Fratio was higher than F-table (17.969 > 3.963, sig. .000 < .050). Thus, there was enough evidence to reject the null hypothesis; as a result, the hypothesis was accepted. Therefore, it could be concluded that the students treated by teaching grammar-in-context make less grammatical errors in their writing than those who are taught grammar conventionally. In other words, teaching grammar-incontext was accepted as an effective strategy to minimize JUDPPDWLFDO HUURUV LQ VWXGHQWV· ZULWLQJ

General Tendency of the Research
The general tendency of the research in this study is an illustration of the degree of the stuGHQWV· ZULWLQJ DELOLW\ LQ minimizing grammatical errors before and after treatment. The procedures used in analyzing the tendency were (1) computing the mean score and standard deviation of grammar component in writing obtained by each group in the pretest and posttest; (2) dividing the mean score obtained by each group by the number of subjects in each group. The results of the computation for the two groups are as follows: (1) the mean scores of grammatical component in the pretest and posttest were respectively 37.80 and 54.00, meaning that the mean score in the posttest was higher than that in the pretest and the difference was 16.20; (2) the mean score of grammatical component in writing achieved by the control group in the posttest was not significantly higher than that in the pretest (37.00 in the pretest and 37.20 in the posttest) and the difference was just .20. It means that there was no significant improvement since the result of the posttest was still nearly the same with that in the pretest.

Discussion
The effectiveness of teaching grammar in context can be seen from the results of the test revealing that the experimental group, which was taught grammar-in-context integrated with writing skill, makes less grammatical errors in writing. This is because the students in this group directly practice grammatical items after they analyze and learn the grammatical items in the text. In this study, the students were given some recount texts. Then, they paid attention a lot to the grammatical items used in the recount texts. The following activity was that the students were guided by the teacher to deepen their knowledge about those grammatical items in the recount texts by doing some exercises. After that, they were asked to practice those grammatical items in writing a recount text. As a result, most of students make less grammatical errors. It was shown by the increase of mean score and their grammatical errors in writing were significantly reduced. In addition, the control group made 62.76% of grammatical errors while the experimental one made 49% of grammatical errors in the posttest. Furthermore, the experimental group made 61% in the pretest and 49% of grammatical errors in the posttest. It means that the experimental group reduced the grammatical errors to 12%. Thus, it indicated that teaching grammarin-context is proved to be an effective method to minimize JUDPPDWLFDO HUURUV LQ VWXGHQWV· ZULWLQJ In relation to the principles of CTL, teaching grammar-incontext adapts its principles, those are inquiry, questioning, constructivism, modeling, learning community, authentic assessment, and reflection. Grammar can, in fact, play its role in text level practice because it is easier to build up strong associations between structures and text types, which make it likely that the students are able to select appropriate structures in recount text. In this study, teaching grammar-in-context relates grammar to text type, in this case, grammar is taught in relation to the context in which grammar is used in recount text. As a result, the students get easier in applying grammar directly in writing a recount text.
In connection with the theory of grammar, writing, and errors in which writing, as a complex skill, needs grammar to construct sentences and as students still make some errors because of the complexity of grammar, teaching grammar-in-context can afford to PLQLPL]H VWXGHQWV· JUDPPDWLFDO HUURUV LQ WKHLU ZULWLQJ

CONCLUSION
Based on the findings and discussion presented in the previous explanation, it can be concluded that teaching grammar-in-context is HIIHFWLYH LQ PLQLPL]LQJ JUDPPDWLFDO HUURUV LQ VWXGHQWV· ZULWLQJ ,W LV SURYHG E\ WKH VLJQLILFDQW LPSURYHPHQW RI WKH H[SHULPHQWDO JURXS·V mean score. Additionally, the F-ratio obtained from the analysis exceeded the critical value. Clearly, since there is enough evidence to reject the null hypothesis and accept the working hypothesis, it could be concluded that teaching grammar in context can reduce more JUDPPDWLFDO HUURUV LQ VWXGHQWV· ZULWLQJ Based on the findings, some suggestions are recommended for English teacher, for Islamic Senior High School students, and for the future researchers. It is suggested that English teachers implement this technique to their writing classes by teaching grammatical items integrated with writing skill, and the students need special efforts to find an appropriate technique in order to able to practice correct grammar in their writing. To have the paragraph correct, they can find the grammatical items in the recount texts made by outstanding writers in many textbooks.
The study has provided the basis for future researchers since teaching grammar-in-context has been proved effective in reducing JUDPPDWLFDO HUURUV LQ VWXGHQWV· ZULWLQJ WKH ILQGLQJV RI WKH UHVHDUFK will be very heOSIXO IRU WKH LPSURYHPHQW RI WKH VWXGHQWV· ZULWLQJ ability in the future. At last, the future researchers may confirm, modify, or add the findings of the study.