1.2 How to Pair Learners in Pair Work

One of the big concerns about pair work for teachers in a second language class is how to pair students, especially when students’ second language proficiency is varying. And the findings of the existing researches are mixed. The early researches by Long and Porter (1985) suggested that mixed proficiency pairs can be more effective than similar proficiency pairs because the former pairs usually have more deliberations and discussions than the later pairs. Nevertheless, the later research made by Kowal and Swain (1994) showed that making mixed proficiency students into pairs can be risky because while doing the pair work, the high level proficiency students usually dominate the task while the participation of low level proficiency students farmers can be very low. And this finding suggested that mixed proficiency pairs, especially those with difference in proficiency level, may not be more effective than similar proficiency pairs. Baines, Blatchford, and Kutnick (2008) put forward another idea. In their study, they argued that the best way to pair students is putting high- to middle proficiency students into pairs and putting low-to middle proficiency students in to pairs. However, as Neomy Storch and Ali Aldosari (2012) have recently argued, how to best pair students in heterogeneous classes depend on the aim of the activity, and that the dyadic relationship may be of greater significance than proficiency pairing. 

1.3 The Advantages of Pair Work

Many researches have suggested that the use of pair work can bring many advantages compared with inpidual learning in second language classes. The advantages can be concluded as follows: firstly, pair work can give each student more opportunity to practice second language. (e.g.; Harmer 2001)Secondly, it enables students to discuss and learn independently, so that it enhances students’ autonomy. Thirdly, pair work allows teachers to give more attention to each student. While teachers are helping or assessing one pair, the rest of pairs can still go on their work. (e.g.; Lu Chen, 2014) Fourthly, pair work is easy to be applied in second language classes as it can be organized quickly and easily.

1.4 The Problems Associated with Pair Work 

In spite of these advantages, there are also some disadvantages about pair work that we cannot ignore. Among these problems, the most common one is that students are inclined to use their native language when they do pair tasks. One of the main reasons for teachers to give pair tasks is that they believe that pair work can increase the amount of speaking time in English each student gets in second language classes. However, if students use these opportunities to speak native language, then pair work turns into an disadvantage in this respect. (e.g. Kellerman, 1995) The second problem is about class management. In many second language classes, there are more than 40 students. Therefore, it is extremely difficult for one or two teachers to handle this kind of class, especially when students do their pair work. For example, teachers are not able to comment on every pairs and give them applicable advices in that circumstance. Besides, it is really hard for teachers to keep an eye on every pairs. And some pairs may digress from the given task while doing the pair work. As a result, the merits of pair work start to fade in that situation. Another limitation of it is that some students can develop a strong dependency upon their peers. In other words, some students place too much reliance on their partner, and their partner turn to dominate the pair work.

1.5 Previous researches about pair work 源^自·吹冰|文\论]文'网[www.chuibin.com

However, most studies so far have showed that pair work is more effective than inpidual learning in SEL. Previous studies also showed that pair work and other kind of collaborative tasks enable students to accomplish the performance that is beyond their inpidual ability (eg. Swain, M. (2001)). N. Storch’s (1999) grammatical tasks showed that the overall accuracy of the grammatical exercises (close, text reconstruction and composition) completed by pairs is significantly higher than those completed by inpiduals. Later, in the text editing study, N. Storch (2007) tried to find out the benefits of pair work by comparing the accuracy of texts edited by pairs and that by inpiduals. The result of this study showed that there is no significant difference in the accuracy. Yet, the pair talks suggested that learners have more opportunities to reflect about their language use and are more aware of the feedbacks from their peers. And another word-building study by Sasan Baleghizadeh (2008) suggested that the pair work group achieved significantly higher marks on word-building tasks than the inpidual group.   .