Peer Interaction by English as a Foreign Language (EFL) Students in a Decision Making Task

Tasks are frequently used in English as a Foreign Language (EFL) classes. The sociocultural theory considers tasks to mediate learners through language. However, how a specific task can embrace such a claim remains an issue to explore. This study was set out to address the questions of how a decision making task stimulated interaction among peers and how it mediated knowledge construction. Thirty EFL junior college students forming fifteen pairs participated in the study. The recorded videos of their decision-making task performance were used for data analysis. The findings reveal that cumulative talk was present in all pairs’ recordings. While exploratory talk was not generated as frequently as cumulative talk, it tended to conduce to more language learning and knowledge construction. The findings also indicate that with the same task, peers interacted in different patterns and mediated each other by providing immediate assistance with English vocabulary and ideas.


INTRODUCTION
Interaction among peers in language classes has been emphasized for its benefits of promoting idea exchanges, language use and collaborative work. When facilitated effectively, peer interaction is supposed to result in language learning. The appearance of the communicative language teaching approach in the 1980s was a catalyst for the creation and use of language tasks that put peers into communication in the classroom. Seen from sociocultural theory (SCT) perspectives (more discussion of the theory is presented in the literature review of this paper), peer interaction plays critical role in students' language learning because their language use serves to scaffold one another in performing language tasks. Peer scaffolding may even mediate and assist students to do the language activities beyond their current abilities.
This paper reports an empirical study on how a decision making task stimulated interaction among peers and how it mediated knowledge construction. Given the fact that tasks are vital in EFL classes for promoting real-world processes of language use, opportunities to speak, and co-constructing meaning (Sato & Ballinger, 2016), exploring the processes of students learning with English tasks will provide insights into how and what students might be able to achieve from tasks. As tasks are considered to perform a mediating role in SCT through peer interaction with the use of explicit language structures and forms emerging from difficulties experienced during the completion of tasks (Ogilvie & Dunn, 2010), analyzing the recorded peer interaction when working on tasks is of significance.

LITERATURE REVIEW
This paper adopted SCT as the theoretical framework; thus, the theory is introduced in the following section. Besides, how peer interaction and tasks seen from SCT are presented, followed by the discussion of the previous studies to identify the gaps that the current study to fill.

VIETNAM JOURNAL OF EDUCATION
 33  continually on the meaning which demands spontaneous and structurally dissimilar samples of language that have not been rehearsed" (Coultas & Booth, 2019). In SCT, although tasks may provide opportunities for learners to extend their L2 knowledge, these opportunities are not only created by tasks themselves but also by the way in which learners perform them.
SCT emphasizes the role of learners, more knowledgeable peers, the use of language and other mediational tools, social interaction and mutual scaffolding in language learning. With an emphasis on the participants and their social interaction through the mediational tool as language, peer interaction tasks should be chosen with the concerns of the language learners themselves so that "Students act as language users with the explicit analysis of language structures and forms emerging from difficulties experienced during the completion of tasks" (Ogilvie & Dunn, 2010) SCT suggests using the tasks which can create the learning environment as in real-world in which language learning is a social process between language learners and others within learners' reach and language is one of the semiotic tools (Lantolf & Thorne, 2006). That is to say, individual language learners in their learning environment should have the chance to take active participation in peer interaction and use language as a mediational tool to solve language problems.
Speaking tasks are supposed to create the most interactive opportunities in order to involve students verbalizing their thoughts while working together. According to Mackey and Gass (2005, p. 65), decision making tasks (also called consensus tasks) are to manipulate the kinds of interactions in which learners are involved, the kind of feedback they receive during the interaction, and the kind of output they produce in order to determine the relationship between the various components of interaction and second language learning.
The tasks that have been investigated varied from vocabulary (Ahmadian & Tajabadi, 2017), grammar (Ohta, 2000), reading, writing (Roberson, 2014;Storch, 2001a;Storch, 2002;Watanabe & Swain, 2007;Zheng, 2012) and speaking (Le, 2004;Le, 2007) which are classified into two types of oral and written performance of language. In other words, collaborative interaction and scaffolding provide opportunities for learning through the use of language in oral and written performance. Thus, the speaking and writing tasks are all in the paradigm of SCT and confined within its tenets.

Decision Making Tasks
The aim of this study is to scrutinize how peers interact and mediate with each other in language learning. The adoption of a task that involves dyads is thus necessary. A decision making task usually triggers more talk and the exchange of thinking in the process of "two heads are better than one". Besides, Nunan (2006) pointed out that tasks create an enhancement of the learner's own personal experiences as important contributing elements to classroom learning. There is also the linking of classroom language learning with language use outside the classroom when learners work on a language task.
In this study, the decision-making task was adapted from Pica et al. (1993) in which learner dyads are given a problem for which there are a number of possible outcomes and they must choose one through discussion. "The Desert Island" (Appendix A) by Sadow (1982) provides a situation of two people on a sinking boat and they were allowed to bring five out of twelve items, namely torchlight, pillows, canned food, clothes, freshwater, knives, map, family documents, handphone, first-aid kit, matches and gun, for their survival on a desert island until they were rescued. The task required both participants to use their critical thinking to choose five items and persuade their partners to agree with the items.

Related Studies
Studies on peer interaction have been conducted to explore the interaction patterns established and to show the relationship between the two participants in pair work. For example, Storch (2001a) investigated the patterns of dyadic interaction in the context of an advanced level adult English as a second language class, the formation of different patterns of interaction and the links between these patterns and second language development. The analysis of data from audio-recorded pair talks by 10 pairs completing three different language-focused tasks of writing, observation notes, a survey on attitudes to pair and group work showed four distinct patterns of dyadic interaction: collaborative, dominant/dominant, dominant/passive and expert/novice. The results revealed evidence of a transfer of language knowledge from pair to individual performance in the data of the collaborative and expert/novice dyads than in the other two patterns of dyadic interaction.
Studies on peer interaction in the small group have also been carried out. For example, the study by Shima (2008) focused on peer interaction in a small group work task of a pre-intermediate level Japanese language classroom at an Australian university. The findings from audio-recordings showed the positive results of mutual

VIETNAM JOURNAL OF EDUCATION
 34  assistance in peer interaction in terms of providing various learning opportunities and two-way direction of assistance. The study, however, revealed the changeable nature of expert and novice roles in peer interaction, which came not only from different levels of expertise but also from the learners' relative positioning of themselves in interaction with their peers.
In an investigation on how pair work influences language learning of six Korean college pairs in EFL lessons, Kim (2016) explored the type of peer interaction, learners' perception of the interaction, and the relationship between the interaction and language learning opportunities. Data from transcripts of audio-recorded pair talk in completing seven different activities, from observation notes, and from interviews indicated that the learner's engagement is important in pair interaction with a collaborative orientation because it can lead to the language development of socially co-constructed and mutually accepted language knowledge.
Dyadic interaction, or interaction in pairs, has been explored in the sociocultural framework. For example, Ahmadian and Tajabadi (2017) documented the dyadic interaction in an EFL course for young learners at the preelementary school level in Iran. Data were analyzed from recorded dialogues of 9 pairs completing a recognition task and a production task to identify the four patterns: collaborative, expert/novice, dominant/dominant, and dominant/passive patterns of the interaction coding scheme. The comparison between the patterns and the participant's performance on a vocabulary test revealed that collaborative and expert/novice were associated with better learning outcomes. Another example was shown in the study by Chen (2018) which examined the interaction of dyads on a dictogloss task, focusing on the transition process of learners' role relationship. The results revealed that a collaborative pattern of interaction is more conducive to L2 learning than other patterns such as dominant/dominant, dominant/passive. During an interaction in pairs, students adjusted their relationship to become more collaborative, which resulted in increased frequency of language-related discussion.
Despite the rather large volume of research using SCT to investigate mediation in peer interaction and discussing how different factors including language, tasks, materials, textbooks lead to new learning in learners, a study that delves into a specific task for dyads to interact has not been carried out. The current study focuses on peer interaction with very minimal or no participation from the teacher. The teacher just asked students to do the tasks at the beginning. It studies the interaction patterns by EFL students and how peer interaction mediates language learning in a decision making task. The study adopted SCT as a theoretical framework to analyze and interpret peer interaction.

MATERIALS AND METHODS
Research Setting and Participants: This study took place at a college in Vietnam where students studied for three years to obtain a college degree. Responding to the invitation from the researchers, thirty ELF college students volunteered to participate in the study. Their English proficiency was determined as "pre-intermediate" level or A2 level based on CEFR. Their ages ranged from 20 to 22. They were in their first year at a junior college of pedagogy and were expected to obtain the B2 level of English up on graduation from the junior college. All the participants were female due to the fact that few male students enrolled in the program at the research site and none of them volunteered to join the study. This fact, however, does not affect the research outcomes as this study aims to investigate peer interaction in a given task and does not purpose to compare how different genders interact in pair work.
Observation with Video-Recording: This study used classroom observation with video-recording on the ground that this method is able to capture what is actually going on when students are working in pairs in the class. For the purpose of the observed recording, smartphones with video functions were used. Particularly, classmates were asked to do the recordings while the researcher remained outside the classrooms for any instant help. In this way, the discussions were recorded in a comfortable atmosphere without any interference from the researcher. Due to their sitting postures, the visual recordings focused from the level of the table surface upward, which showed the upper parts of the bodies of the two participants. Each pair was recorded separately by their friends acting as cameramen who had some technical training.
The recordings with can-be-heard sound, clear images were chosen for transcribing. However, gestures of the participants were not examined due to the fact that the focus of the study was on verbal interaction. The images helped to identify the speaker in the talks. Task achievement is the primary criterion to extract the transcript. That means each pair had to complete the given task once or the participants had to come to a conclusion of the five things to bring to the desert island. Pair discussion of the problem solving task was then transcribed by the researcher, and

VIETNAM JOURNAL OF EDUCATION
 35  certain sections re-transcribed and analyzed using conventions from sociocultural discourse analysis (Mercer, 2004, p. 141). According to Mercer (2004), sociocultural discourse analysis focuses on the use of language as a social mode of thinking and a tool for teaching and learning, constructing knowledge, creating joint understanding and tackling problems collaboratively. Transcription of classroom interactions was extracted as excerpts for commentaries.
The researcher and a separate coder having understanding and experience in coding the talks worked independently on the analysis of the interaction transcript. The coding process started with the interaction transcripts of three pairs piloted and was then compared to see any mismatches. According to Storch (2001a, p. 134), the reliable norm of coding interaction is acceptable with more than 80% of results in consensus. The two coders in this study reached a higher level of consensus for all the recordings. In general, there was not any difference of coding results between the two coders thanks to the specific examples.
The transcription in this study follows the verbatim in which fillers were noted with "…" before and after the non-word utterances. The participants sometimes used Vietnamese for the vocabulary and structures they did not know or remembered. The phrases or sentences produced in Vietnamese were translated into English, italicized and put in parentheses next to the utterances. The unclear words were followed by a question mark in parentheses (?) while utterances that could not be understood are marked [unintelligible].
Sociocultural Discourse Analysis: Sociocultural discourse analysis (SDA) was coined by Mercer (2004) and his colleagues in the educational research for analyzing talks to understand how spoken language is used as a tool for thinking collectively. SDA with the focus on the use of language as the social model of thinking provided qualitative and quantitative data for interaction-based research. The current study adopted this approach to analyze peer talks for the understanding of peer interaction in an EFL context. Mercer's (2004)  In accordance with theoretical framework of SCT, the concept of mediation was also examined to clarify the features of the peer talks, the techniques the participants used to interact with their peers and to mediate language learning when completing the speaking tasks.

Results
This study's main aim is to identify the interaction features in peer interaction on completing a decision making task, more specifically, talk types and how peers mediate each other when performing the task.

VIETNAM JOURNAL OF EDUCATION
 37 

The Types of Talk
The types of talk were identified in accordance with Mercer's convention of sociocultural discourse analysis (Mercer, 2004). The following table shows the types of talks found in the task transcriptions of all pairs.

Pair
Talk Pattern S1-S2 Cumulative Exploratory → Cumulative → Exploratory → Cumulative S11-S12 Exploratory → Disputational → Exploratory → Disputational → Exploratory → Cumulative S13-S14 Cumulative → Exploratory → Cumulative S15-S16 Disputational → Exploratory → Disputational → Cumulative → Disputational → Cumulative S17-S18 Cumulative → Exploratory → Cumulative S19-S20 Cumulative The table above reveals the predominance of the cumulative talk type. The general interaction among the pairs started with cumulative talks (except pairs S9-10, S11-12, and S25-26 with an exploratory pattern and pair S15-16 a disputational one) and ended with a cumulative one. There is evidence of three types of talk in each pair's interaction. The task involves language learners into various steps of thinking, reasoning and challenging with different skills of negotiation, communication and collaboration, which entails the occurrence of different talk types and a complicated combination of talk types within each pair in this speaking task. Besides, with the same task, talk patterns varied among the pairs, which indicates that peers are a decisive factor creating the patterns of talks. It should also be noted that all the pairs ended their talks with a cumulative talk. This could be explained by the fact that students had to reach certain confirmation related to the decision that the speaking task requires.
Below are examples of the three kinds of talk. Cumulative Talk Excerpts 1 and 2 below illustrate the cumulative talk type. Excerpt 1: S2: How about clothes? What's your opinion? S1: You're right. ..mm.. We need clothes to wear… ..mm.. during the days we live on the island. S2: I feel the same way. S1: What about a hand phone? S2: ..mm.. I think we don't need… it because on the… island ..mm.. unable to connect ..mm.. to with anyone. S1: ..mm.. Yeah, I… completely agree with you. There was no evidence of dispute in the talk between S1 and S2 about clothes and a hand phone. In fact, they seemed to agree with each other's ideas quickly and share the common solutions through confirmation of the other's ideas. The cumulative talk type in the task was produced in the harmonious agreement with confirmation and elaboration.
Excerpt 2 Both of the participants in excerpt 02 above accumulated their "common knowledge" of the torch light with repetition and confirmation. Each echoed the other, for example, on what the other had said. S7 paraphrased the idea of S8 without any explicit critical opinions.

Exploratory Talk
The excerpts below represent the exploratory talk type. Excerpt 3: S11: I think a hand phone is unnecessary to bring it. S12: Yes, I have the same with you. S12: Because.. maybe on the desert island we don't… we can't… we can't connect to the other people. S11: Yes S12: So we needn't to bring it. S11: and I know that, in there, there isn't the electricity. So we cannot charge the hand phone. The exploratory type indicates both participants' efforts to solve the linguistic problems with explanations for the final shared decision of whether to bring a hand phone to the island or not. S11 and S12 provided their reasonable opinions on hand phone with explicit reasoning for their ideas and even build upon each other's opinions. Thus, they can make joint understanding through the use of a cultural tool, namely language and a mediational tool of hand phone as shown in the cue card of the speaking task for their knowledge con-construction. There is evidence of mediation in the exploratory talk above with visible explanations, explicit arguments, rejection and reasons by both peers. In the beginning, when being asked for opinions about canned food, S25 explained the reasons for bringing canned food due to the lack of food on the island and got assistance from S26 to complete the idea. S26 immediately rejected with some linguistic difficulties but could come to the solution with the support from S25. Consequently, S26 continued to persuade S25 by explicitly reasoning the opportunity to catch fish in the sea so that they can have food instead of canned food. S26 was mediated and they came to the shared solution of how to have food on the island. In general, the knowledge of food on the desert island was co-constructed with

VIETNAM JOURNAL OF EDUCATION
 39  the collaborative contribution and critical analysis of looking for food. The mutual mediation via the talk and the task resulted in the ability to use the correct phrase "catch fish".

Disputational Talk
The excerpts below characterize the disputational talk type. Excerpt 5: S11: You bring water from the sea and you warm it to drink. S12: you drink the sea? (laugh) S11: yes, you can (laugh) S12: no, I don't think so. S11: yes, I think you can. Both S11 and S12 cannot give any explanations or reasons for their opinions. The short exchanges and quick rejection of the other's ideas resulted in no further interaction. The excerpt displays little discussion of language and ideas between the two participants during the process of working through the task. Thus, they cannot gain any knowledge co-construction with such short utterances and typical disagreement in this type of talk. S4: I don't agree with you. We will bring a torchlight so we don't need to bring matches. S3: I don't think so. The disputational exchanges between S3 and S4 include rejection and disagreement. They challenged their peer's decisions but did not provide any logical arguments. In spite of S1's efforts to mention his/her strength to swim for fish, S2 still disregarded his/her peer's opinions.
Excerpt 7: S15: And the other thing I will choose is matches because you use matches to light a fire to cook. And I heard that wild animals are scared fire. So what do you think? It's a good idea? S16: Đâu có thức ăn đâu mà nấu ăn. (There is no food to cook) S15: Thì ở trên đó mầy giết động vật rồi mầy phải nấu chứ đúng không? Không lẽ mầy giết rồi mầy để mầy ăn sống à? (So on the island you have to kill the animals and cook them, right? How can you kill them and then eat them raw?) S16: Chưa chắc là đã giết được. (You can't be sure that you can kill the animals) S15: Mầy phải nấu. Mầy nói đi. (You have to cook. Your turn to say) S16: I don't agree. S15 and S16 in the excerpt above make their arguments in their mother tongue. They tried to defend their opinions and abruptly objected to their partner's ideas. Although both participants took their active parts during the talk, there was little evidence of joint or cooperative engagement. Much of the interactional talk consisted of commands and assertion. They hardly considered the other's ideas and the excerpt ends with the individual decision of not to bring matches on the island.
The excerpts above document the change of talk types in one speaking task. The most common pattern is cumulative → exploratory → cumulative, which was identified in six out of fifteen pairs, including pairs S1-S2, S13-S14, S17-S18, S19-S20, S21-S22 and S29-S30. The most complicated pattern belongs to pair S3-S4 with all types of talk shifting within the pair talk. This could be explained that the same task when carried out by different peers and prior knowledge of peers, the talk types would be different.

Features of Peer Interaction in the Decision Making Task
SCT considers interaction as the fundamental factor in language learning and highlights the roles of peer interaction in the process of learning language via mediation. The students in this study completed their speaking tasks by assisting each other to solve the linguistic problems with unknown words and elaborate the ideas generated by the task. This indicates that language learners mediate each other to co-construct knowledge. In other words, language mediation and task mediation may result in collaborative thinking and language understanding. This section shows the results of interaction features to better understand how EFL college students interacted with their peers in the decision making task.

VIETNAM JOURNAL OF EDUCATION
 40  The recorded data of the decision making task performance disclose that the students used various types of questions to encourage and involve the partners in the talk. They also used feeding back and explaining techniques to facilitate each other to discuss the appropriate solutions for survival on the desert island. While they were reasoning the arguments, they were also building up their thoughts on each other's opinions. Initially, one student could choose one item and the reason for their choice. Then the other student was encouraged to clarify the meaning through scaffolding to reach their final decisions. Besides, it is also found that feeding back played a surpassing role to become the most popular technique with high percentages in most of the pairs in this study.
The following examples illustrate the mediational features in the speaking task. In excerpt 09, S25 used feeding back to give information to S26's question to pronounce a word in English. S26 then asked the question of what to use the first-aid-kit for. S25 then gave hint in the first language with the equivalent word in English "solve". S25 also used questioning to understand S26's problem and then to provide linguistic help. S25 was then able to continue with the talk.
The excerpt below shows how S21 scaffolded S22 to use a knife in response to the requirement of the task. Excerpt 10: S21: And next I will choose a knife S22: Knife? S21: Yes S22: Why? S21: A knife, it's the basic thing you need when you in a jungle, when you go to travel in a jungle or in something, some dangerous area. First, it can protect, you can use it to protect yourself, it is the first. And second, you can use it to hunt, you can hunt some small animals to make food and the third it is the.. the item to making food S22: Ok. That's good. S21: Basic item S22: I agree with you. I think maybe there is a lot of danger ..mm.. in the island. ..mm..I should bring knives to defend ourselves. S21: Yes Excerpt 10 reveals the process of thinking when S21 provided critical reasons with a daily experience of using knives to persuade his/her peer to bring it to the island. Thanks to the explanations of S21 that S22 could develop his/her ideas drawing on S21's opinions about the importance of knives and consequently S22 agreed with "bringing knives to defend ourselves".

VIETNAM JOURNAL OF EDUCATION
 41 

Discussion
This study adopted SCT as the theoretical framework to explore peer interaction in a decision making task. The mediation taking place in peer interaction was scrutinized from the transcription of the recorded tasks. It is argued that peer interaction may have no identifiable or constant "experts" (Ohta, 1995;Villamil & De Guerrero, 1996). Even when this is the case, the transcript from this study shows that mediation can occur in a collaborative form where students can exchange their roles frequently and get mutual benefits from providing "peer scaffolding" (Donato, 1994) because each learner has his/her prior knowledge and sociocultural background relevant to the task at hand. The EFL college students in the present study used various scaffolding techniques to support their partners to use their language and thinking. The findings show that peer interaction results in confirmation and repetition on receiving the mediation and assistance from their peers. Overall, the co-construction of knowledge present in the excerpts in the current study derived from two language learners with similar proficiency working collaboratively and scaffolding mutually. In other words, the unclear and unidentifiable role of experts or more knowledgeable peers in peer interaction on completing given speaking tasks could also lead to better talk outcomes, which is in line with Edwards (2005) but unlike other studies (Kasper, 2001 for example) prioritizing the superior roles of experts in boosting peer learning. This raises the focus to the fact that language learning can occur in interactive contexts where peers with similar proficiency interact. The results revealed that three types of the talk were evident in peer interaction in the decision making task but with different numbers in each pair. Cumulative talk plays a dominant role, which indicates that the peers and their proficiency do not influence much on the clustering and combining of talk types. Besides, the qualitative investigation of each pair's discussion revealed that the exchanges that exhibited exploratory talk showed more in-depth reasoning than other exchanges did, which is consistent with Mercer (2004). The participants in exploratory talks used their prior knowledge for reasoning and critical explanations. As a result, this type is more conducive to language learning and knowledge construction.
This study also examined tasks that are considered to be mediating artifacts in SCT. More specifically, the decision making task in this study provides an authentic learning context for the meaningful communicative interaction. With the use of questions, students draw their partners into a shared understanding of the tasks in which they are engaged (Mercer, 2004). Wegerif and Mercer (1997) found the transition from one type (exploratory) to another type (disputational) in the interaction of three nine-year-old pupils working on a series of graphical puzzles taken from a reasoning test. Similarly, Polo, Lund, Plantin and Niccolai (2015) reported the shifting natures from disputational talks to exploratory talks continuously under the circumstance of arguing and claimed that engaging in a given type of talk is not only a matter of cognitive ability. In the same vein, Saadi's (2016) investigation into the talk types pointed out that participants make meaning through seven types of talk consisting of uni-level (Disputational, Exploratory, and Cumulative), bi-level (Disputational/ Exploratory, Disputational/ Cumulative, and Exploratory/ Cumulative), and multi-level (Disputational/ Exploratory/ Cumulative) categories. The results in the present study are in line with previous studies, which claim that natures of talks are changeable (Polo et al., 2015;Saadi, 2016;Wegerif & Mercer, 1997) with rather unidentifiable routes. Moreover, interaction in the current study facilitates the students in the task performance through finding unknown words, elaborating the ideas and also providing critical analysis of the arguments in order to reach the consensus. Therefore, peer interaction results in the co-construction of knowledge and language learning, at least in terms of the vocabulary and structure to express the ideas to complete the given task.

CONCLUSION
This study aims to investigate the types of talks in peer interaction and the features of peer talks. Data were recorded from 15 pairs doing a decision making task and transcribed for analysis. The results revealed that three types of talk, namely cumulative, exploratory and disputational were present in all pairs' talks; however, the cumulative talk was generated with the highest percentage. The analysis of the recordings revealed that the exchanges that exhibited exploratory talk showed more in-depth reasoning than the other exchanges did because peers resorted to their prior knowledge for reasoning and critical explanations to complete the task. As a result, this type tended to be more conducive to language learning and knowledge construction in peer interaction.
Another finding of the current study is that the most salient feature of peer interaction is feeding back followed by the technique of questioning. These two features facilitated peers into a shared understanding of the tasks in which they were engaged. With these two techniques, peers were able to clarify unclear information, seek help with VIETNAM JOURNAL OF EDUCATION  42  language and ideas, encourage participation, and even correct the wrong use of words. The task itself played a critical role in generating thoughts in peer interaction and the immediate need arising from the task promotes peer scaffolding, benefiting both partners in the interacting process.
It should be noticed that the peers in this study were not trained to follow or use certain types of talk, or scaffolding techniques. However, they naturally resorted to the interaction to mediate with the other peer to complete the task. These findings suggest language teachers pair students with similar language abilities. It is thus worth investigating how to direct peers to talk in a given task that brings about linguistic and cognitive development as well as effective use of peer interaction.
The current study did not aim to explore whether peers were able to retain the language structures, vocabulary and ways to elicit other peers to talk during tasks. It did not either focus on the comparison of peer interaction in different language tasks. The further study thus can dwell into these topics to gain more insights into EFL peer interaction.