EFL learners talking to each other: Task types and the quality of output

Focus of the study: Unlike previous studies, this study questions a more claim of the Output hypothesis: the potential occurrence of pushed output when learners interact with each other in an English as Foreign Language context rather than a second language or immersion context. As a result of the nature of EFL context, the research aims to explore if learners get engaged in output production when their peers speak the same L1 as theirs. Due to this common L1, it is hypothesized that learners will not  produce ouput in L2 and even if they do, the grammaticality of their output will not be target-like as they will be able to understand well due to their common linguistic background.
Instructional settings and participants: 12 senior high school students enrolled in an English conversation class at a private highscool in Japan. The participants had been taking English classes for six years. Besides this conversation class, they were also enrolled in other English courses including reading and grammar.
Instrumentation and procedure: The tasks used in this study were information-gap, jig-saw, and an open-ended discussion. In the information-gap activity, students worked in pairs and one student described a picture in detail for the other participant to draw the item on a paper. All pairs in a classroom worked on the same picture. In the jigsaw activity, students worked on reconstructing story using pictures. In the open-ended discussion, participants talked about their favorite memory. All peer interactions during these tasks were recorded and transcribed. Data was analyzed by c-units. To measure grammaticality, the ratio of the grammatical c-units to total number of c-units was calculated. 
Results and discussion: Regarding the first research question, the study found that the grammaticality was highest during the open-ended discussion and lowest in the information-gap task. The groups did not differ significantly in terms of reformulated responses during peer-interactions through peer-feedback. The main goal of the study to find out if non-native speakers of L2 would produce output even though their L1s were the same and how peers would help each other to reformulate their utterances and eventually produce pushed-modified output. The study showed that nonnative speakers produced output as much as they would to with a native speaker of L2. This study is important as it shows the effects of output in EFL context and output even in such context promotes production; however to what extent output is grammatically produced needs further investigation.