Many researchers have been paying their attention to pair-work strategy in EFL writing and found several research findings. Baleghizadeh (2009) highlighted that learners who completed the task in pairs outperformed those who attempted it individually. Storch (1999) also found that learners who worked in pairs had more opportunities to communicate in the target language than those in teacher-fronted classrooms. These findings implicitly imply EFL teachers on how to create a strategy in which students have more opportunities to communicate in the target language. Baleghizadeh and Farhesh (2014) investigated the power of pair work to students's motivation and found that pair work had positive contributions to learners' motivations. This means that pairwork can be employed to stimulate students's motivation. Shin, et al (2015) found that students who worked together in pairs assisted each other, thereby recalling more correct idea units from the texts and eliminating extraneous information in their writing. These research findings have shown that pair work really, in general, gives a lot of advantages if it is employed in EFL classrooms. The findings discussed above on the merits of pair work are also related to aspects required in writing.
Although the use of pair work in the classroom is relatively limited in use (Storch, 2011), this strategy is believed to have beneficial points. Storch (2007:143) underlines states that pair-work has strong pedagogical and theoretical supports. From the pedagogical aspect, Biria and Jafari (2013:166) affirm that pair-work offers language learners more chances to use the language. From a theoretical point, this strategy is in line with language learning that emphasizes the importance of interaction for learning in order to develop learner's language competence. Dobao (2012:41) asserts that the role of interaction and peer collaboration are considered important in L2 development. In addition, Xiao (2008:106) also points out that paired-peer review gives students benefits in terms of EFL writing proficiency, transferable skills, and self-efficacy. Besides, Mulligan and Garofalo (2011:9) emphasize that collaborative writing (pairwork) is a non- threatening approach for students that results in purposeful usage of the target language across skills and demonstrable improvements in writing. In short, pair-work does not only function as an activity to be employed in teaching-learning process, but it also gives advantages to students in writing performance.
Many research findings have showed that pair work had a significant effect or contribution on students' writing performance. Sorch (1999:363) asserted that collaboration had a positive effect on overall grammatical accuracy. In other study, Storch (2005:168) made a comparison of the products (completed texts) of pairs and individuals. The pairs produced shorter and better texts that had greater grammatical accuracy and linguistic complexity, and were more succinct. They seemed to fulﬁll the task more competently. In a collaborative study, Wigglesworth and Storch (2009:445) found that collaboration (pair work) impacted positively on accuracy although it does not affect fluency and complexity. Indirectly, these three research findings reveal that pair work is able to facilitate students to improve their skill in using language component (grammar) required in producing good written texts. In addition, pair work also helps students to write effectively in presenting ideas in their essays.
Some researchers also found the effect of pair work toward students' writing performance. Shehadeh (2011:286) found that collaborative writing had an overall significant effect on students' L2 writing. However, this effect varied from one writing skill area to another. Specifically, the effect was significant for content, organization, and vocabulary, but not for grammar or mechanics. Globally, her findings are the same as Storch's. However, she also found a different result from Storch's study (2005). Her finding showed that collaborative writing did not have a significant effect on grammar. Meanwhile, Storch found that pairs produced shorter and better texts that had greater grammatical accuracy. The difference was assumed by the fact that they used the subject with different proficiency (based on IELTS score). The subjects' score (average) in Shehade's study was 4. Meanwhile, the subjects' score (average) in Storch's study was 6.5. Although research findings by Shehade on grammar accuracy is different from Storch's, Shehadeh found the significant effect of pair work on some other aspects of writing. These aspects were found to be a problem for students working individually (Sabarun:2006, Attamim: 2007, Ulfiati: 2010, and Isnawati: 2010).
Jafari and Ansari (2012: 128) found that students working in pairs had better writing accuracy than those working individually. This finding can enrich the theory on the merrits of pair work found by Shehadeh (2011) and Storch (2005). In addition, this finding (students working in pairs had better writing accuracy) also has answered research finding by Kasman (2004) and Irawati (2008) in which college students and university graduates were found to have low academic writing. It means that having students to work in pairs can help them to minimize their problem in writing.
Biria and Jafari (2013:164) found that practicing in pairs really improved the overall quality of the learners' writing even though the fluency of written texts did not change significantly. Studets need to be encouraged in order to have good writing quality and this research finding has showed that pair works can meet the target. In other words, pair work can guide students to have a good quality writing.
At last, Meihami, Meihami, and Varmaghani (2013:47) found that collaborative work (pair work) could improve students' grammatical accuracy in their upcoming writings. Their findings match what Storch (1999:363) has found on the effect of pair work on students grammar in writing. Having grammatical accuracy is a must in writing activities because sentences with wrong grammatical patterns can mislead and will not be understandable. Working in pairs was found to be useful for students to improve their grammatical accuracy because writing activites where students correct each other and share ideas with one another guide students to better grammatical accuracy.
The target to be reached in this study is the empirical evidence of the effect of pair work types on students' writing quality. Therefore, the researcher formulates the research problem as follows: Is there a different writing quality between the students who work in randomized group and those in heterogeneous one?
The researcher carried out this present study at Islamic University of Malang. This campus was chosen because it was the only campus which offered writing III class (argumentative essay is one of genres discussed) when the researcher conducted this study. The target population of this study was the English Department students of the Faculty of Teacher's Training and Education at Islamic University of Malang. The accessible population was sophomores who were taking Writing III course in semester 3, 2016/2017.
This study was carried out for six meetings and each group experienced the same treatment. The following table shows the treatment procedure.
The research treatment was carried out in the second to fifth meetings. On the second meeting, the researcher explained the concept of argumentative, including the components of argumentative essay and how to develop each component. The third meeting was used to discuss the concept of an introductory paragraph and complete the task on an introductory paragraph. The fourth meeting was used for a body paragraph. The fifth meeting was used for the concluding paragraph.
In the third meeting, the subjects were given an argumentative topic. Then, they were asked to work in pair to develop the introductory paragraph of the topic. In the process of developing the introductory paragraph, they were asked to employ writing stages (planning, drafting/writing, and editing). In the stage of planning, everyone in each pair was asked to produce at least two or three sentences individually that would be used to be general statements and a thesis statement. This means that each group had at least 4-6 sentences. Then, each pair discussed the sentences produced to be an introductory paragraph. In this step, they could add necessary sentence/s and eliminate improper sentence/s in order that their introductory paragraph was acceptable and proper. While they were discussing the sentences in order to be introductory paragraph, they were reminded that their introductory paragraph had to consist of some general statements and a thesis statement that had to be put as the final sentence of introductory paragraph.
In the 4 meeting, the researcher first discussed some points related to the introductory paragraphs developed by the subjects for 10-15 minutes in previous meeting. After that, the researcher presented and discussed the concept of a body paragraph and how to develop the body paragraph. Then, the subjects were asked to develop a body paragraph. In this phase, they were reminded that their body paragraph had to develop the thesis statement formulated at the end of introductory paragraph. In the process of developing the body paragraph, the subjects were also asked to employ writing stages (planning, drafting/writing, and editing). In the stage of planning, everyone in each pair was asked to produce at least three sentences individually. In drafting/writing and editing stages, each pair discussed sentences produced (at least 6) for the body paragraph. Here, they were also reminded that their body paragraph had to consist of a topic sentence and several supporting sentences.
In the 5 meeting, the researcher first discussed some points related to the body paragraphs developed by the subjects for 10-15 minutes in the previous meeting. As soon as the discussion was completed, the researcher presented and discussed the concept of a concluding paragraph and how to develop the concluding paragraph. Here, they were informed that they could use a conclusion transition signal such as in conclusion, in summary, or tosummarize. Next, they could either summarize the main points of the essay or rewrite the thesis statement in different words. After that, the researcher had the subjects develop concluding paragraph. In the process of developing the concluding paragraph, the subjects were asked to make the concluding paragraph individually based on the thesis statement and body paragraph. Then, everybody brought the concluding paragraph developed to the group and compared it to one developed by their partner. The comparison was carried out in order that they could produce a proper concluding paragraph.
When the subjects were developing introductory, body, and concluding paragraphs, the researcher had to walk around the classroom, moving from one group to another. This activity was carried out to make sure that the subjects really worked in pairs as intended. In addition, this was conducted to minimize the chance of any pair dominating his/her partner.
To collect the required data (students' score) for this study, the researcher used a writing test as his research instrument. The writing test was used because the writing test result could measure students' performance. This is in line with Zare-ee's statement (2013) saying that EFL students' writing quality can be drawn as their EFL writing skill through their performance.
The test to be used in this study was a direct writing. This kind of format was a better way to test students' writing ability than indirect one because the direct writing test could provide better accuracy. This is in line with Latief (2014:234) stating that students writing skill can only be assessed from students' actual piece of writing.
In developing the writing prompt for this study, the researher adapted one available in IELTS test for two reasons. First, IELTS is a standard test recognised internationally. As a standard test recognised internationally, IELTS must have a clear writing prompt. In addition, it also shows that the test takers are required to do the same task. Second, IELTS also contains an argumentative writing test that requires test takers to present a solution to a problem, present and justify an opinion, and evaluate and challenge an ideas, evidences or arguments (Cambrridge Practice Test for IELTS 9, 2013:3).
Finding and discussion
In line with the research problem formulated previously (to investigate different writing quality among students working with heterogeneous pairs and those working with homogeneous pairs), the researcher analyzed the data to detect if there were different mean among the two groups. Based on the result of data analysis, it was found that the subjects in the randomized group performed better significantly than those in heterogeneous group. The result of data analysis also revealed that there were different means among two groups (Randomized pair/RD and Heterogeneous pair/HT). In addition, the table also showed that the subjects in the randomized pair performed the higher mean (M = 68.87, SD = 13.815) than those in heterogenous group (M = 57.13, SD = 11.190). Next, the table also revealed the standard error for each group. The standard error is the standard deviation of sample means. A small standard error indicates that most sample means are similar to the population mean and so the sample is likely to be an accurate reflection of the population (Field, 2009:43).
Then, the researcher also investigated which groups had the significant difference. The significant difference of means can be detected if the sign star (*) is available in Mean Difference column. The following Table 1 showed clearly which groups had the sign (*) in Mean Difference column.
Based on the Table 3.4, it was found that ramdomized group was significantly different from heterogeneous one. Tukey's test and Bonferroni's test showed the same result related to the significant difference.
What was found in this study reveals that there is a strategy, randomized grouping that can help EFL students perform better in EFL writing than heterogeneous grouping. Previously, Eckley (2014) conducted a study to discover what type of cooperative learning has the best result /educational outcomes. His research finding was that students from heterogeneous groups performed better working than homogenous groups. Then, Fauziah and Latief (2015) also investigated the effect of working in Heterogeneous and Homogeneous pairs on the students' writing skill. Their research finding was that the quality of descriptive essays produced individually by the students from heterogeneous group was on the average significantly better than the quality of descriptive essays produced by the students from the homogeneous group. What was found by Eckley and Fauziah & Latief above means that heterogeneous group has better result than homogeneous group to students' achievement.
3. Conclusion and Recommendation
This study was aimed at investigating the comparison of writing quality of students working in randomized group and those in heterogeneous group. Based on the result of data analysis, it was found that the subjects in the randomized group performed better than those in heterogeneous group. Based on the result of data analysis, the researcher draws three recommendations. First, the randomized pair strategy is recommended to be employed in EFL writing class because it is proven that the students using this strategy have higher score than heterogeneous group. Second, the following study on pair work is recommended to employ for one semester. Third, it is also recommended to employ these three strategies in genres other than argumentative text in order to compare the effect of pair work in different genres.